Handbook of Exoplanets [1 ed.] 9783319553320, 9783319553337

Citation preview

Hans J. Deeg Juan Antonio Belmonte Editors

Handbook of Exoplanets

Handbook of Exoplanets

Hans J. Deeg • Juan Antonio Belmonte Editors

Handbook of Exoplanets With 910 Figures and 107 Tables

123

Editors Hans J. Deeg Instituto de Astrofísica de Canarias La Laguna, Tenerife, Spain

Juan Antonio Belmonte Instituto de Astrofísica de Canarias La Laguna, Tenerife, Spain

Departamento de Astrofísica Universidad de La Laguna La Laguna, Tenerife, Spain

Departamento de Astrofísica Universidad de La Laguna Tenerife, Spain

ISBN 978-3-319-55332-0 ISBN 978-3-319-55333-7 (eBook) ISBN 978-3-319-55334-4 (print and electronic bundle) https://doi.org/10.1007/978-3-319-55333-7 Library of Congress Control Number: 2018950058 © Springer International Publishing AG, part of Springer Nature 2018 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Foreword

In the search for other worlds, the last decades have probably been among the most exciting over the past centuries, possibly since the years of the Copernican heliocentrism and the discovery by Galileo of the Moons around Jupiter. The large series of breakthroughs in the search for exoworlds make this recent period a rather remarkable time in the history of astronomy which appears to be as fascinating as the one about 400 years ago when humankind started to abandon geocentrism. During the past 25 years, we have witnessed the detection of planets orbiting thousands of nearby and distant stars. Since the discovery of the first planets around pulsars in the early 1990s and the first Jupiter-mass planet around the solar-type star 51 Peg in 1995, a large diversity of planetary systems, has been identified in the nearby universe. Efficient hunting programs have provided increasing statistical evidence that planets are very common around stars. More than 50% of the stars in our galaxy may host planetary systems and therefore, tens of billions may await discovery. The detection rate of exoplanets has only increased with time, reaching values above one exoplanet discovery per day. The number of known exoplanets, several thousand, will considerably increase in the coming decade thanks to the many search programs already started or planned for ground and space telescopes. The study of this extremely rich population of planetary systems will lead to a better understanding of their architecture and the physics involved in the formation processes. Ultimately, the ongoing search and characterization work may unveil planets with adequate conditions to sustain the development of life and will pave the road to the discovery of exolife. Planets with masses similar to those existing in the Solar System are frequently found in other planetary systems, displaying very different physical conditions. Exoplanets appear in a large range of orbital separations around a variety of stars and therefore are subject to very different stellar irradiations. The properties of the planets depend heavily on their mass, chemical composition, stellar irradiation, and on their interaction with the host stars’ gravity, radiation, and magnetic field. Observations have revealed and will continue bringing to light an enormous diversity of planets and planetary systems conforming an exceptional set of laboratories which will challenge our knowledge on physics, chemistry, geology, and biology. The masses of known exoplanets span the range between the mass of the Earth and several times the mass of Jupiter. While our Solar System provides useful v

vi

Foreword

guidance to establish the minimum mass of a planet, observations do not offer a strong indication about the value of a maximum mass. A widely adopted criterion for such value is the deuterium burning limit (13 MJup /. However, objects with masses slightly above, generally designated as brown dwarfs, and below this limit have been found orbiting stars and could in principle form via the same mechanism of gravitational instability in protoplanetary discs, blurring a distinction. Brown dwarfs are defined as self-gravitating objects unable to sustain stable hydrogen burning which according to evolutionary models have masses below 75 MJup (for solar metallicity). Discovered in 1995, free-floating brown dwarfs are known to populate the galaxy in a comparable number to stars, but are rarely found around stars (with occurrence rate of a few percent). In year 2000, free-floating objects with only a few times the mass of Jupiter were discovered in star clusters via imaging and spectroscopy. Subsequent searches have revealed that free-floating super-Jupiters compare in number to solar-type stars and are far more common as free-floaters than orbiting stars. Establishing an upper limit to the mass of planets will have to await until an adequate understanding of the formation mechanisms of these superJupiters is achieved. Doppler radial velocity measurements provided the first exoplanet discoveries around solar-type stars, the so-called Hot Jupiters, close-in orbit gas giants dominated by a hydrogen-helium envelope with a rocky core. This type of exoplanets was also the first detected to produce eclipses of their stars. While cold Jupiterlike planets of much longer orbital periods appear to orbit around 3% of solar-type stars, their hot counterparts are present only around less than 1%. Hot Jupiters are likely formed via core accretion at much higher separation from their host stars suffering subsequent migration to their observed orbits. Planets with such very close orbits (P < 7 days) offer a high probability (10%) of producing eclipses, and many have been the subject of extensive atmospheric characterization via differential photometry and spectroscopy during transits. These observing techniques have provided some initial insight on their atmospheric chemical composition, vertical pressure-temperature profiles, albedos, and circulation patterns. Among the identified new types of planets, super-Earths, which have several times, the mass of the Earth and sizes up to twice its radius, are remarkably different to the planets in the Solar System. Besides, they are the most abundant planets with orbital periods of less than 100 days and are frequently found in compact multiple-planet systems. More than 50% of the stars seem to host a super-Earth or a smaller planet. The generation of these planets is expected to occur through the formation of a rocky core and subsequent accretion of a gas envelope. The envelopes can be massive enough to notably contribute to the total radius of the planet. However, many processes (photoevaporation, collisions, etc.) contribute to eroding the atmosphere during evolution causing a large diversity of these envelopes, which observations are starting to unveil. Some super-Earths could in principle form a crust and host liquid water, if they are located at suitable orbital separations. Several have been detected in the habitable zone of stars producing eclipses. They are very

Foreword

vii

attractive targets for atmospheric characterization via transit spectroscopy with the new suite of large diameter ground and space telescopes. Evidence for the existence of terrestrial planets is compelling, and planets with similar mass, size, and physical conditions potentially similar to the Earth have already been discovered. Planet Proxima b in the nearest star to the Sun, detected using Doppler radial velocity measurements, is the closest example of a continuously increasing family. Such rocky planets may host liquid water, and the characterization of their thin atmospheres will be an extraordinary challenge, even for the new generation of extremely large telescopes. Proxima b is not known to transit its parent star, and direct imaging and spectroscopy with coronographs assisted by Adaptive Optics on very large and extremely large telescopes is a promising way to obtain information on its atmospheric properties. Identifying tracers of biological activity will possibly require new technological advances. The Kepler space observatory and other ground-based observatories have identified a large number of transiting planets, including those of Earth-size. Series of radial velocity measurements of the host stars could in principle achieve a determination of the masses for these small planets, which typically induce radial velocity semi-amplitudes of tens of cm/s in solar type stars. The advent of a new generation of ultra-stable high dispersion spectrographs at very large telescopes (ESPRESSO is the first to achieve 10 cm/s) will make possible such measurements in a fraction of the detected systems, leading to the obtainment of planet densities and further insight on the formation processes of terrestrial planets. In multiple transiting planet systems, transit time variability observations can also provide a determination of masses. Bright stars with transiting Earth-size planets offer an excellent opportunity to study planet atmospheric properties with JWST and the ELTs. A large effort is currently undertaken to search for transiting planets in the habitable zone of nearby stars using a series of dedicated ground-based telescopes (MEarth, SPECULOOS, etc.) and space observatories (TESS). In the future, other space telescopes like JWST, CHEOPS, and PLATO and the extremely large telescopes (EELT, TMT, GMT) will bring exceptional capacities for the characterization of the atmospheres of a large variety of exoplanets, including the new terrestrials. This Handbook of Exoplanets provides an outstanding vision on the state of the art of exoplanet research, as well as describes the historical evolution of the field from first discoveries to the most recent detections. It includes a revision of the theories of formation and evolution for the various types of planets and the on going effort to characterize both planet interiors and their atmospheric properties. Current knowledge on exoplanet properties is confronted with the detailed information provided by planets in the Solar System and by brown dwarfs. The atmospheres of nearby free-floating brown dwarfs can be studied in great detail and offer important insight and guidance for the exploration of exoplanet atmospheres in a large range of temperatures extending below the temperature of the atmosphere of the Earth.

viii

Foreword

This book also offers an overview of recent advances in the various techniques employed in the field and shows how progress on direct imaging, radial velocities, transit photometry and spectroscopy, microlensing, astrometry, etc., will enable the path to understanding the origin, evolution, and the physical/chemical properties of the large diversity of planets so far discovered, including those similar to Earth. Instituto de Astrofísica de Canarias, La Laguna, Spain February 2018

Rafael Rebolo

Preface

About 25 years after the discovery of the first exoplanets by a few scattered pioneers, the field of exoplanetology has developed into a principal branch of astronomy, producing over a thousand scientific articles every year. The underlying central question that motivates most of its activity, “Are we alone in the Universe?” and “What are the origins of our and of other Worlds?” can now be illuminated from several angles, but a conclusive answer remains in the distance. The present work is a first attempt to summarize the current status of the science driven by these questions. The idea for it started almost like a joke during a dinner in a sympathetic Korean Restaurant during the 29th IAU General Assembly in Honolulu. Three years later, that embryo has developed into four heavy volumes, with contributions by over 200 scientists. We, as Editors-in-Chief of this project, are very proud of how our colleagues, partners, friends, and even some scientific rivals have taken a substantial part of their more than busy lives to make this possible. A big “Thank you” to all of them! The Handbook of Exoplanets, like other major reference works by Springer, has been organized into Sections. Each of them was developed under the supervision of one or more dedicated Section Editors. The work of these scholars has been absolutely fundamental for the success of the project. Dear Tsevi, Agustín, María Rosa, Alex, Norio, Malcolm, Roi, Hans, Nuccio, Natalie, Sara, Ralph, Pedro, Vikki, Rory, and Jean, you cannot imagine how thankful we are! The Handbook is organized along both a chronological and thematic perspective. The first section “Exoplanet Research: A History of Discovery” serves as an introduction for the Handbook. Then, two sections follow that contextualize exoplanets within the wider field of astronomy: “Solar System–Exoplanet Synergies” and “Between Planets and Stars,” devoted to the celestial bodies of our vicinity, including the Earth and objects like free-floating planets or brown dwarfs. The major part of the Handbook describes the observational efforts of the last 25 years, namely “Planet Discovery Methods,” “Ground-Based Instrumental Projects for Exoplanet Research,” “Space Missions for Exoplanet Research,” and “Exoplanet Characterization.” The central stars are fundamental for our understanding of planet systems, hence the sections devoted to: “Characterizing Planet Host Stars” and “Planets and Their Stars: Interactions.” The next section of the Handbook introduces the status of interpretative work in exoplanet science. First, the major global results ix

x

Preface

are given in the section “Catalogues, Planet Abundances and Statistics.” The most detailed knowledge we have about exoplanets is about their atmospheres, hence the section “Exoplanet Atmospheres.” The question about our and other worlds’ origins is directly confronted in “Formation and Evolution of Planets and Planet Systems.” One of the major observational results is our awareness of the variety of other worlds that exist in the Universe, which motivated the section named “The Diversity of Worlds: An Exoplanet Fauna.” The largest section of this book “Where Life May Arise: Habitability” is directly dedicated to the fundamental question “Are we alone?” We do not have a crystal ball suggesting what will be next in our field. However, we felt the necessity to envisage how it may develop; hence, the book concludes with “The Future: What Will Be Next?” Certainly, there will be colleagues who point out that important topics have been omitted and they will likely be correct. However, this work has been envisaged as a living document in which future developments, as well as updates of current ones, will be addressed in its electronic edition. So, it is open to suggestions and improvements, and we invite readers to provide feedback. Our ultimate hope is that sometimes in the future there will be chapters or whole sections, not devoted to remote observations and exoplanet habitability as of today, but rather to results from in situ missions and to exoplanet habitats. Future provides indeed a wide open window to our understanding of the Universe! Tenerife, Spain April 2018

Hans J. Deeg Juan Antonio Belmonte

Contents

Volume 1 Section I Exoplanet Research: A History of Discovery . . . . . . . . . . . . . Tsevi Mazeh 1

The Discovery of the First Exoplanets . . . . . . . . . . . . . . . . . . . . . . . . Davide Cenadelli and Andrea Bernagozzi

2

PSR B1257+12 and the First Confirmed Planets Beyond the Solar System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Alexander Wolszczan

1 3

21

3

Prehistory of Transit Searches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Danielle Briot and Jean Schneider

35

4

Discovery of the First Transiting Planets . . . . . . . . . . . . . . . . . . . . . . Edward W. Dunham

51

5

The Way to Circumbinary Planets . . . . . . . . . . . . . . . . . . . . . . . . . . . Laurance R. Doyle and Hans J. Deeg

65

6

The Naming of Extrasolar Planets . . . . . . . . . . . . . . . . . . . . . . . . . . . Frederic V. Hessman

85

7

Impact of Exoplanet Science in the Early Twenty-First Century . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hans J. Deeg and Juan Antonio Belmonte

Section II Solar System–Exoplanet Synergies . . . . . . . . . . . . . . . . . . . . Agustín Sánchez Lavega 8

The Solar System: A Panorama . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Katherine de Kleer and Imke de Pater

9

Interiors and Surfaces of Terrestrial Planets and Major Satellites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Alberto G. Fairén

95

115 117

141

xi

xii

10

11

Contents

Internal Structure of Giant and Icy Planets: Importance of Heavy Elements and Mixing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ravit Helled and Tristan Guillot

167

Composition and Chemistry of the Atmospheres of Terrestrial Planets: Venus, the Earth, Mars, and Titan . . . . . . . . . . . . . . . . . . . Thérèse Encrenaz and Athena Coustenis

187

12

Tenuous Atmospheres in the Solar System . . . . . . . . . . . . . . . . . . . . Emmanuel Lellouch

13

Temperature, Clouds, and Aerosols in the Terrestrial Bodies of the Solar System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . F. Montmessin and A. Määttänen

215

235

14

Temperature, Clouds, and Aerosols in Giant and Icy Planets . . . . Robert A. West

265

15

Atmospheric Dynamics of Terrestrial Planets . . . . . . . . . . . . . . . . . . Peter L. Read, Stephen R. Lewis, and Geoffrey K. Vallis

285

16

Atmospheric Dynamics of Giants and Icy Planets . . . . . . . . . . . . . . A. Sánchez-Lavega and M. Heimpel

317

17

Upper Atmospheres and Ionospheres of Planets and Satellites . . . Antonio García Muñoz, Tommi T. Koskinen, and Panayotis Lavvas

349

18

Rings in the Solar System: A Short Review . . . . . . . . . . . . . . . . . . . . Sébastien Charnoz, Aurélien Crida, and Ryuki Hyodo

375

19

The Diverse Population of Small Bodies of the Solar System . . . . . Julia de León, Javier Licandro, and Noemí Pinilla-Alonso

395

20

The Solar System as a Benchmark for Exoplanet Systems Interpretation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pilar Montañés-Rodríguez and Enric Pallé

Section III Between Planets and Stars . . . . . . . . . . . . . . . . . . . . . . . . . . . María Rosa Zapatero-Osorio

421

445

21

Brown Dwarf Formation: Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . Anthony P. Whitworth

22

Brown Dwarfs and Free-Floating Planets in Young Stellar Clusters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V. J. S. Béjar and Eduardo L. Martín

469

Large-Scale Searches for Brown Dwarfs and Free-Floating Planets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ben Burningham

503

23

447

Contents

24

25

xiii

Spectral Properties of Brown Dwarfs and Unbound Planetary Mass Objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jacqueline K. Faherty

531

Y Dwarfs: The Challenge of Discovering the Coldest Substellar Population in the Solar Neighborhood . . . . . . . . . . . . . . Sandy K. Leggett

543

26

Variability of Brown Dwarfs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Étienne Artigau

555

27

Metal-Depleted Brown Dwarfs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nicolas Lodieu

575

28

Radio Emission from Ultracool Dwarfs . . . . . . . . . . . . . . . . . . . . . . . Peter K. G. Williams

589

29

Definition of Exoplanets and Brown Dwarfs . . . . . . . . . . . . . . . . . . . Jean Schneider

611

Section IV Planet Discovery Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . Alexander Wolszczan

617

30

Radial Velocities as an Exoplanet Discovery Method . . . . . . . . . . . . Jason T. Wright

619

31

Transit Photometry as an Exoplanet Discovery Method . . . . . . . . . Hans J. Deeg and Roi Alonso

633

32

Finding Planets via Gravitational Microlensing . . . . . . . . . . . . . . . . Virginie Batista

659

33

Astrometry as an Exoplanet Discovery Method . . . . . . . . . . . . . . . . Fabien Malbet and Alessandro Sozzetti

689

34

Direct Imaging as a Detection Technique for Exoplanets . . . . . . . . Laurent Pueyo

705

35

Pulsar Timing as an Exoplanet Discovery Method . . . . . . . . . . . . . . Michael Kramer

767

36

Timing by Stellar Pulsations as an Exoplanet Discovery Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . J. J. Hermes

787

Transit-Timing and Duration Variations for the Discovery and Characterization of Exoplanets . . . . . . . . . . . . . . . . . . . . . . . . . . Eric Agol and Daniel C. Fabrycky

797

37

38

Radio Observations as an Exoplanet Discovery Method . . . . . . . . . T. Joseph W. Lazio

817

xiv

39

Contents

Detecting and Characterizing Exomoons and Exorings . . . . . . . . . René Heller

835

Volume 2 Section V

Ground-Based Instrumental Projects for Exoplanet Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Norio Narita

40

High-Precision Spectrographs for Exoplanet Research: CORAVEL, ELODIE, CORALIE, SOPHIE, and HARPS . . . . . . . Francesco Pepe, François Bouchy, Michel Mayor, and Stéphane Udry

41

ESPRESSO on VLT: An Instrument for Exoplanet Research . . . . Jonay I. González Hernández, Francesco Pepe, Paolo Molaro, and Nuno C. Santos

42

SPIRou: A NIR Spectropolarimeter/High-Precision Velocimeter for the CFHT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jean-François Donati, D. Kouach, M. Lacombe, S. Baratchart, R. Doyon, X. Delfosse, Étienne Artigau, Claire Moutou, G. Hébrard, François Bouchy, J. Bouvier, S. Alencar, L. Saddlemyer, L. Parès, P. Rabou, Y. Micheau, F. Dolon, G. Barrick, O. Hernandez, S. Y. Wang, V. Reshetov, N. Striebig, Z. Challita, A. Carmona, S. Tibault, E. Martioli, P. Figueira, I. Boisse, Francesco Pepe, and the SPIRou Teams

43

44

855

883

903

HiCIAO and IRD: Two Exoplanet Instruments for the Subaru 8.2 m Telescope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Motohide Tamura

931

Imaging with Adaptive Optics and Coronographs for Exoplanet Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Olivier Guyon

937

45

The HATNet and HATSouth Exoplanet Surveys . . . . . . . . . . . . . . . Gáspár Á. Bakos

46

KELT: The Kilodegree Extremely Little Telescope, a Survey for Exoplanets Transiting Bright, Hot Stars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Joshua Pepper, Keivan G. Stassun, and B. S. Gaudi

47

853

Small Telescope Exoplanet Transit Surveys: XO . . . . . . . . . . . . . . . Nicolas Crouzet

957

969 981

Contents

xv

48

SPECULOOS Exoplanet Search and Its Prototype on TRAPPIST . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1007 Artem Burdanov, Laetitia Delrez, Michaël Gillon, Emmanuël Jehin, and the SPECULOOS and TRAPPIST Teams

49

Microlensing Surveys for Exoplanet Research (OGLE Survey Perspective) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1025 Andrzej Udalski

50

Microlensing Surveys for Exoplanet Research (MOA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1045 Philip Yock and Yasushi Muraki

51

Korea Microlensing Telescope Network . . . . . . . . . . . . . . . . . . . . . . . 1065 Byeong-Gon Park, Andrew P. Gould, Chung-Uk Lee, and Seung-Lee Kim

52

Exoplanet Research with the Stratospheric Observatory for Infrared Astronomy (SOFIA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1085 Daniel Angerhausen

53

Exoplanet Research in the Era of the Extremely Large Telescope (ELT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1105 Florian Rodler

Section VI Space Missions for Exoplanet Research . . . . . . . . . . . . . . . . 1121 Malcolm Fridlund 54

Space Missions for Exoplanet Research: Overview and Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1123 Malcolm Fridlund

55

CoRoT: The First Space-Based Transit Survey to Explore the Close-in Planet Population . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1135 Magali Deleuil and Malcolm Fridlund

56

Space Missions for Exoplanet Science: Kepler/K2 . . . . . . . . . . . . . . 1159 William J. Borucki

57

Observing Exoplanets with the Spitzer Space Telescope . . . . . . . . . 1179 Charles A. Beichman and Drake Deming

58

Space Astrometry Missions for Exoplanet Science: Gaia and the Legacy of Hipparcos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1205 Alessandro Sozzetti and Jos de Bruijne

59

Interferometric Space Missions for Exoplanet Science: Legacy of Darwin/TPF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1229 Denis Defrère, Olivier Absil, and Charles A. Beichman

xvi

Contents

60

CHEOPS: CHaracterizing ExOPlanets Satellite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1257 Willy Benz, David Ehrenreich, and Kate Isaak

61

Observing Exoplanets with the James Webb Space Telescope . . . . 1283 Charles A. Beichman and Thomas P. Greene

62

Space Missions for Exoplanet Science: PLATO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1309 Heike Rauer and Ana M. Heras

63

Future Astrometric Space Missions for Exoplanet Science . . . . . . . 1331 Markus Janson, Alexis Brandeker, Celine Boehm, and Alberto Krone Martins

64

Future Exoplanet Space Missions: Spectroscopy and Coronographic Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1343 Artie P. Hatzes and René Liseau

Section VII Exoplanet Characterization . . . . . . . . . . . . . . . . . . . . . . . . . 1355 Roi Alonso 65

Mass-Radius Relations of Giant Planets: The Radius Anomaly and Interior Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1357 Gregory Laughlin

66

The Rossiter–McLaughlin Effect in Exoplanet Research . . . . . . . . 1375 Amaury H. M. J. Triaud

67

Stellar Limb Darkening’s Effects on Exoplanet Characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1403 Szilárd Csizmadia

68

Exoplanet Phase Curves: Observations and Theory . . . . . . . . . . . . 1419 Vivien Parmentier and Ian J. M. Crossfield

69

Characterization of Exoplanets: Secondary Eclipses . . . . . . . . . . . . 1441 Roi Alonso

70

Mapping Exoplanets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1469 Nicolas B. Cowan and Yuka Fujii

71

Spectroscopic Direct Detection of Exoplanets . . . . . . . . . . . . . . . . . . 1485 Jayne L. Birkby

72

Characterizing Evaporating Atmospheres of Exoplanets . . . . . . . . 1509 Vincent Bourrier and Alain Lecavelier des Etangs

73

Disintegrating Rocky Exoplanets . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1527 Rik van Lieshout and Saul A. Rappaport

Contents

xvii

74

Characterizing the Chemistry of Planetary Materials Around White Dwarf Stars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1545 B. Zuckerman and E. D. Young

75

Bayesian Methods for Exoplanet Science . . . . . . . . . . . . . . . . . . . . . . 1567 Hannu Parviainen

76

Tools for Transit and Radial Velocity Modeling and Analysis . . . . 1591 Hans J. Deeg

Section VIII Characterizing Planet Host Stars . . . . . . . . . . . . . . . . . . . . 1613 Hans Kjeldsen 77

Characterizing Planet Host Stars: Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1615 Hans Kjeldsen

78

Accurate Stellar Parameters for Radial Velocity Surveys . . . . . . . . 1623 Nuno C. Santos and Lars A. Buchhave

79

The Combined System of Microlensing Exoplanets and Their Host Stars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1641 Uffe Gråe Jørgensen and Markus Hundertmark

80

Characterizing Host Stars Using Asteroseismology . . . . . . . . . . . . . 1655 Mia Sloth Lundkvist, Daniel Huber, Víctor Silva Aguirre, and William J. Chaplin

81

Ages for Exoplanet Host Stars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1679 Jørgen Christensen-Dalsgaard and Víctor Silva Aguirre

Volume 3 Section IX Planets and Their Stars: Interactions . . . . . . . . . . . . . . . . . . 1697 Antonino F. Lanza 82

Planet and Star Interactions: Introduction . . . . . . . . . . . . . . . . . . . . 1699 Antonino F. Lanza

83

Rotation of Planet-Hosting Stars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1705 Pierre F. L. Maxted

84

Stellar Coronal Activity and Its Impact on Planets . . . . . . . . . . . . . 1723 Giuseppina Micela

85

Signatures of Star-Planet Interactions . . . . . . . . . . . . . . . . . . . . . . . . 1737 Evgenya L. Shkolnik and Joe Llama

86

Magnetic Fields in Planet-Hosting Stars . . . . . . . . . . . . . . . . . . . . . . 1755 Claire Moutou, Rim Fares, and Jean-François Donati

xviii

Contents

87

Star-Planet Interactions in the Radio Domain: Prospect for Their Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1775 Philippe Zarka

88

The Impact of Stellar Activity on the Detection and Characterization of Exoplanets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1791 Andrew Collier Cameron

89

Tidal Star-Planet Interactions: A Stellar and Planetary Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1801 Stéphane Mathis

90

Models of Star-Planet Magnetic Interaction . . . . . . . . . . . . . . . . . . . 1833 Antoine Strugarek

91

Stellar Coronal and Wind Models: Impact on Exoplanets . . . . . . . 1857 Aline A. Vidotto

92

Electromagnetic Coupling in Star-Planet Systems . . . . . . . . . . . . . . 1877 Joachim Saur

93

Accretion of Planetary Material onto Host Stars . . . . . . . . . . . . . . . 1895 Brian Jackson and Joleen Carlberg

94

Planetary Evaporation Through Evolution . . . . . . . . . . . . . . . . . . . . 1913 Travis S. Barman

Section X Exoplanet Catalogs, Abundances, and Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1931 Natalie Batalha 95

Exoplanet Catalogs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1933 Jessie Christiansen

96

Planet Occurrence: Doppler and Transit Surveys . . . . . . . . . . . . . . 1949 Joshua N. Winn

97

Occurrence Rates from Direct Imaging Surveys . . . . . . . . . . . . . . . 1967 Brendan P. Bowler and Eric L. Nielsen

98

Populations of Extrasolar Giant Planets from Transit and Radial Velocity Surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1985 Alexandre Santerne

99

Planet Populations as a Function of Stellar Properties . . . . . . . . . . 2009 Gijs D. Mulders

100

Populations of Planets in Multiple Star Systems . . . . . . . . . . . . . . . 2035 David V. Martin

Contents

xix

Section XI Exoplanet Atmospheres . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2061 Sara Seager 101

The “Spectral Zoo” of Exoplanet Atmospheres . . . . . . . . . . . . . . . . 2063 Aki Roberge and Sara Seager

102

Exoplanet Atmosphere Measurements from Transmission Spectroscopy and Other Planet Star Combined Light Observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2083 Laura Kreidberg

103

Exoplanet Atmosphere Measurements from Direct Imaging . . . . . 2107 Beth A. Biller and Mickaël Bonnefoy

104

Radiative Transfer for Exoplanet Atmospheres . . . . . . . . . . . . . . . . 2137 Kevin Heng and Mark S. Marley

105

Atmospheric Retrieval of Exoplanets . . . . . . . . . . . . . . . . . . . . . . . . . 2153 Nikku Madhusudhan

Section XII Formation and Evolution of Planets and Planetary Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2183 Ralph E. Pudritz 106

A Brief Overview of Planet Formation . . . . . . . . . . . . . . . . . . . . . . . . 2185 Philip J. Armitage

107

Dust Evolution in Protoplanetary Disks . . . . . . . . . . . . . . . . . . . . . . . 2205 Sean M. Andrews and Tilman Birnstiel

108

Chemistry During the Gas-Rich Stage of Planet Formation . . . . . . 2221 Edwin A. Bergin and L. Ilsedore Cleeves

109

Instabilities and Flow Structures in Protoplanetary Disks: Setting the Stage for Planetesimal Formation . . . . . . . . . . . . . . . . . . 2251 Hubert Klahr, Thomas Pfeil, and Andreas Schreiber

110

Planetary Migration in Protoplanetary Disks . . . . . . . . . . . . . . . . . . 2287 Richard P. Nelson

111

Formation of Giant Planets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2319 Gennaro D’Angelo and Jack J. Lissauer

112

Formation of Super-Earths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2345 Hilke E. Schlichting

113

Formation of Terrestrial Planets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2365 André Izidoro and Sean N. Raymond

114

Planetary Population Synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2425 Christoph Mordasini

xx

Contents

115

Connecting Planetary Composition with Formation . . . . . . . . . . . . 2475 Ralph E. Pudritz, Alex J. Cridland, and Matthew Alessi

116

Dynamical Evolution of Planetary Systems . . . . . . . . . . . . . . . . . . . . 2523 Alessandro Morbidelli

117

Debris Disks: Probing Planet Formation . . . . . . . . . . . . . . . . . . . . . . 2543 Mark C. Wyatt

Volume 4 Section XIII The Diversity of Worlds: An Exoplanet Fauna . . . . . . . . . . 2569 Pedro Figueira 118

HD189733b: The Transiting Hot Jupiter That Revealed a Hazy and Cloudy Atmosphere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2571 François Bouchy

119

WASP-12b: A Mass-Losing Extremely Hot Jupiter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2585 Carole A. Haswell

120

Transiting Disintegrating Planetary Debris Around WD 1145+017 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2603 Andrew Vanderburg and Saul A. Rappaport

121

Proxima b: The Detection of the Earth-Type Planet Candidate Orbiting Our Closest Neighbor . . . . . . . . . . . . . . . . . . . . 2627 Guillem Anglada-Escudé, Mikko Tuomi, Ignasi Ribas, Ansgar Reiners, Pedro J. Amado, and Guillem Anglada

122

HR8799: Imaging a System of Exoplanets . . . . . . . . . . . . . . . . . . . . . 2645 Quinn M. Konopacky and Travis S. Barman

123

Fomalhaut’s Dusty Debris Belt and Eccentric Planet . . . . . . . . . . . 2669 Paul G. Kalas

124

55 Cancri (Copernicus): A Multi-planet System with a Hot Super-Earth and a Jupiter Analogue . . . . . . . . . . . . . . . . . . . . . . . . . 2677 Debra A. Fischer

125

Planets in Mean-Motion Resonances and the System Around HD45364 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2693 Alexandre C. M. Correia, Jean-Baptiste Delisle, and Jacques Laskar

126

Tightly Packed Planetary Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . 2713 Rebekah I. Dawson

127

Circumbinary Planets Around Evolved Stars . . . . . . . . . . . . . . . . . . 2731 T. R. Marsh

Contents

128

xxi

Two Suns in the Sky: The Kepler Circumbinary Planets . . . . . . . . 2749 William F. Welsh and Jerome A. Orosz

Section XIV Where Life May Arise: Habitability . . . . . . . . . . . . . . . . . . . 2769 Victoria Meadows, Rory Barnes 129

Factors Affecting Exoplanet Habitability . . . . . . . . . . . . . . . . . . . . . 2771 Victoria S. Meadows and Rory K. Barnes

130

Life’s Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2795 Tori M. Hoehler, Sanjoy M. Som, and Nancy Y. Kiang

131

Earth: Atmospheric Evolution of a Habitable Planet . . . . . . . . . . . 2817 Stephanie L. Olson, Edward W. Schwieterman, Christopher T. Reinhard, and Timothy W. Lyons

132

The Habitability of Icy Ocean Worlds in the Solar System . . . . . . . 2855 Steven D. Vance

133

Planet Formation, Migration, and Habitability . . . . . . . . . . . . . . . . 2879 Yann Alibert, Sareh Ataiee, and Julia Venturini

134

Volcanic-Tectonic Modes and Planetary Life Potential . . . . . . . . . . 2897 A. Lenardic

135

Planetary Interiors, Magnetic Fields, and Habitability . . . . . . . . . . 2917 Peter E. Driscoll

136

Planetary Interior-Atmosphere Interaction and Habitability . . . . 2937 Norman H. Sleep

137

Stellar Composition, Structure, and Evolution: Impact on Habitability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2959 Patrick A. Young

138

The Habitable Zone: The Climatic Limits of Habitability . . . . . . . 2981 Ravi Kumar Kopparapu

139

Star-Planet Interactions and Habitability: Radiative Effects . . . . . 2995 Antígona Segura

140

Gravitational Interactions and Habitability . . . . . . . . . . . . . . . . . . . 3019 Rory K. Barnes and Russell Deitrick

141

Habitability of Planets in Binary Star Systems . . . . . . . . . . . . . . . . . 3041 Siegfried Eggl

142

Habitability in Brown Dwarf Systems . . . . . . . . . . . . . . . . . . . . . . . . 3069 Emeline Bolmont

143

Galactic Effects on Habitability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3091 Nathan A. Kaib

xxii

Contents

144

Assessing the Interior Structure of Terrestrial Exoplanets with Implications for Habitability . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3111 Caroline Dorn, Dan J. Bower, and Antoine Rozel

145

Characterizing Exoplanet Habitability . . . . . . . . . . . . . . . . . . . . . . . 3137 Tyler D. Robinson

146

Atmospheric Biosignatures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3159 John Lee Grenfell

147

Surface and Temporal Biosignatures . . . . . . . . . . . . . . . . . . . . . . . . . 3173 Edward W. Schwieterman

148

Biosignature False Positives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3203 Chester E. Harman and Shawn Domagal-Goldman

149

The Detectability of Earth’s Biosignatures Across Time . . . . . . . . . 3225 Enric Pallé

Section XV The Future: What Will Be Next? . . . . . . . . . . . . . . . . . . . . . . 3243 Jean Schneider 150

Future Exoplanet Research: Science Questions and How to Address Them . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3245 Jean Schneider

151

Future Exoplanet Research: Radio Detection and Characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3269 J.-M. Griessmeier

152

Future Exoplanet Research: High-Contrast Imaging Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3285 Pierre Baudoz

153

Future Exoplanet Research: XUV (EUV and X-Ray) Detection and Characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3301 Graziella Branduardi-Raymont, William R. Dunn, and Salvatore Sciortino

154

Circumstellar Discs: What Will Be Next? . . . . . . . . . . . . . . . . . . . . . 3321 Quentin Kral, Cathie Clarke, and Mark C. Wyatt

155

Solid Exoplanet Surfaces and Relief . . . . . . . . . . . . . . . . . . . . . . . . . . 3353 Jean-Loup Bertaux

156

Exotic Forms of Life on Other Worlds . . . . . . . . . . . . . . . . . . . . . . . . 3375 Louis N. Irwin

157

Multi-Pixel Imaging of Exoplanets with a Hypertelescope in Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3389 Antoine Labeyrie

Contents

xxiii

158

Exoplanets and SETI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3405 Jason T. Wright

159

Direct Exoplanet Investigation Using Interstellar Space Probes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3413 Ian A. Crawford

160

Special Cases: Moons, Rings, Comets, and Trojans . . . . . . . . . . . . . 3433 Juan Cabrera, María Fernández Jiménez, Antonio García Muñoz, and Jean Schneider

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3451

About the Editors

Hans J. Deeg is staff astronomer at the Instituto de Astrofísica de Canarias (IAC) in Tenerife, Spain. Born in Würzburg, Germany, he obtained a Master’s in Physics from SUNY Buffalo, USA, in 1986 and a Ph.D. from the University of New Mexico, USA, in 1993. Previously, he held posts at the Rochester Institute of Technology; the SETI Institute (both USA); the Centro de Astrobiología, Madrid; and the Instituto de Astrofísica de Andalucía (both Spain). Deeg’s principal interests are the detection and characterization of exoplanets, for which he has been working since 1994 on a wide range of ground- and space-based projects. He was the principal Spanish investigator for the exoplanet detection with the CoRoT space mission (2006–2014) and is currently coordinating several tasks for ESA’s next-generation PLATO space mission. He is also a habitual user of the large telescopes installed at Roque de los Muchachos Observatory on the island of La Palma. During his career, he has authored over 300 scientific articles mostly related to exoplanets, organized several conferences on this topic, and been member on several review panels for funding agencies or telescope time allocation. Deeg has also supervised several Ph.D. theses and is currently teaching a Master-level course on exoplanets at the University of La Laguna, Tenerife.

xxv

xxvi

About the Editors

Juan Antonio Belmonte is Research Professor at the Instituto de Astrofísica de Canarias (Tenerife, Spain) where he investigates exoplanets, stellar physics, and cultural astronomy. He has published or edited a dozen books and authored more than 200 publications on those subjects. He has been the Director of the Science and Cosmos Museum of Tenerife from 1995 to 2000 and President of the European Society for Astronomy in Culture (SEAC) from 2005 to 2011 and of the Spanish Time Allocation Committee (CAT) of the Canarian observatories from 2003 to 2012. In 2012, he received the “Carlos Jaschek” Award of the European Society for Astronomy in Culture for his contributions to that discipline. He has been editor of two sections and the author of 12 contributions in the previous Springer’s Handbook of Archaeoastronomy and Ethnoastronomy. In the early 2000s, he got involved in exoplanet research being one of the founders of the TrES Network and co-Director of one of the first international schools in the field in 2004. He has supervised three Ph.D. students in Exoplanets, including that of Roi Alonso, one of the Section Editors of this Handbook of Exoplanets. A member of the Exoplanet Project at the IAC, he is currently teaching a master-level course on extrasolar planets at the University of La Laguna. He is now President of the International Society for Archaeoastronomy and Astronomy in Culture and Advisory Editor of the Journal for the History of Astronomy. Born in Murcia (Spain) in 1962, he studied physics and got his master thesis in 1986 at Barcelona University and obtained his Ph.D. in Astrophysics from La Laguna University in 1989.

Section Editors

Exoplanet Research: A History of Discovery Tsevi Mazeh School of Physics and Astronomy, Tel Aviv University, Tel Aviv, Israel Solar System–Exoplanet Synergies Agustín Sánchez Lavega Escuela de Ingeniería de Bilbao, Universidad del País Vasco UPV/EHU, Bilbao, Spain Between Planets and Stars María Rosa Zapatero-Osorio Centro de Astrobiología, Madrid, Spain Planet Discovery Methods Alexander Wolszczan Department of Astronomy and Astrophysics and Center for Exoplanets and Habitable Worlds, The Pennsylvania State University, University Park, PA, USA Ground-Based Instrumental Projects for Exoplanet Research Norio Narita Department of Astronomy, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan. Space Missions for Exoplanet Research Malcolm Fridlund Leiden Observatory, Leiden, The Netherlands Department of Space, Earth and Environment, Chalmers University of Technology, Onsala, Sweden Exoplanet Characterization Roi Alonso Instituto de Astrofísica de Canarias, La Laguna, Tenerife, Spain Departamento de Astrofísica, La Laguna, Universidad de La Laguna, La Laguna, Tenerife, Spain xxvii

xxviii

Section Editors

Characterizing Planet Host Stars Hans Kjeldsen Stellar Astrophysics Centre, Aarhus University, Aarhus, Denmark Planets and Their Stars: Interactions Antonino F. Lanza INAF-Osservatorio Astrofisico di Catania, Catania, Italy Exoplanet Catalogs, Abundances, and Statistics Natalie Batalha NASA Ames Research Center, Mountain View, CA, USA Exoplanet Atmospheres Sara Seager Department of Earth, Atmospheric and Planetary Sciences, Massachusetts Institute of Technology, Cambridge, MA, USA Formation and Evolution of Planets and Planetary Systems Ralph E. Pudritz Department of Physics and Astronomy, McMaster University, Hamilton, Ontario L8S 4K1, Canada The Diversity of Worlds: An Exoplanet Fauna Pedro Figueira European Southern Observatory, Santiago, Chile Where Life May Arise: Habitability Victoria Meadows Astronomy Department, University of Washington, Seattle, WA, USA Rory Barnes Astronomy Department, University of Washington, Seattle, WA, USA The Future: What Will Be Next? Jean Schneider Paris Observatory, LUTh Departement, Meudon, France

Contributors

Olivier Absil F.R.S.-FNRS Research Associate, Space Sciences, Technologies and Astrophysics Research (STAR) Institute, University of Liege, Liege, Belgium Eric Agol Department of Astronomy, University of Washington, Seattle, WA, USA S. Alencar UFMG, Belo Horizonte, Brazil Matthew Alessi Department of Physics and Astronomy, McMaster University, Hamilton, ON, Canada Yann Alibert Space Research and Planetary Sciences, Physikalisches Institut, Universitaet Bern, Bern, Switzerland Roi Alonso Instituto de Astrofísica de Canarias, La Laguna, Tenerife, Spain Departamento de Astrofísica, Universidad de La Laguna, La Laguna, Tenerife, Spain Pedro J. Amado Glorieta de la Astronomía S/N, Instituto de Astrofísica de Andalucía – CSIC, Granada, Spain Sean M. Andrews Harvard-Smithsonian Center for Astrophysics, Cambridge, MA, USA Daniel Angerhausen Center for Space and Habitability, University of Bern, Bern, Switzerland Guillem Anglada Glorieta de la Astronomía S/N, Instituto de Astrofísica de Andalucía – CSIC, Granada, Spain Guillem Anglada-Escudé School of Physics and Astronomy, Queen Mary University of London, London, UK Philip J. Armitage JILA, University of Colorado and NIST, Boulder, CO, USA Étienne Artigau Institut de Recherche sur les Exoplanètes, Département de Physique, Université de Montréal, Montréal, QC, Canada Sareh Ataiee Physikalisches Institut, Universitaet Bern, Bern, Switzerland xxix

xxx

Contributors

Gáspár Á. Bakos Department of Astrophysical Sciences, Princeton University, Princeton, NJ, USA S. Baratchart IRAP/OMP, Toulouse, France Travis S. Barman Lunar and Planetary Lab, University of Arizona, Tucson, AZ, USA Rory K. Barnes Astronomy Department, University of Washington, Seattle, WA, USA G. Barrick CFHT, Waimea, HI, USA Virginie Batista Institut d’astrophysique de Paris, Paris, France Centre National d’Etudes Spatiales, Paris Cedex 1, France Pierre Baudoz LESIA, Observatoire de Paris, PSL Research University, CNRS, Sorbonne Universités, UPMC Universitié Paris 06, Universitié Paris Diderot, Sorbonne Paris Cité, Paris, France Charles A. Beichman NASA Exoplanet Science Institute, California Institute of Technology and Jet Propulsion Laboratory, Pasadena, CA, USA V. J. S. Béjar Instituto de Astrofísica de Canarias, C. Vía Láctea S/N, Tenerife, Spain Juan Antonio Belmonte Instituto de Astrofísica de Canarias, La Laguna, Tenerife, Spain Departamento de Astrofísica, Universidad de La Laguna, Tenerife, Spain Willy Benz Physikalisches Institut, Universität Bern, Bern, Switzerland Edwin A. Bergin Department of Astronomy, University of Michigan, Ann Arbor, MI, USA Andrea Bernagozzi Scuola di Scienze e Tecnologia, Sezione di Geologia, UNICAMearth Working Group, Università degli Studi di Camerino, Camerino, Italy Osservatorio Astronomico della Regione Autonoma Valle d’Aosta, Nus (Aosta), Italy Jean-Loup Bertaux CNRS/LATMOS/UVSQ, Paris, France Laboratory for Atmospheres of Planets and Exo-Planets, IKI-RAS, Moscow, Russia Beth A. Biller Institute of Astronomy, University of Edinburgh, Edinburgh, UK Jayne L. Birkby Anton Pannekoek Institute of Astronomy, University of Amsterdam, Amsterdam, The Netherlands

Contributors

xxxi

Tilman Birnstiel Faculty of Physics, Ludwig-Maximilians-Universität München, University Observatory, Munich, Germany Max Planck Institute for Astronomy, Heidelberg, Germany Celine Boehm Department of Physics, Institute for Particle Physics Phenomenology, Durham University, Durham, UK I. Boisse LAM/OHP, Marseille, France Emeline Bolmont IRFU, CEA, Université Paris-Saclay, Gif-sur-Yvette, France Université Paris Diderot, AIM, Sorbonne Paris Cité, CEA, CNRS, Gif-sur-Yvette, France Mickaël Bonnefoy IPAG, University Grenoble Alpes, Grenoble, France William J. Borucki NASA Ames Research Center, Moffett Field, CA, USA François Bouchy Département d’Astronomie, Université de Genève, Versoix, GE, Switzerland Observatoire astronomique de l’Université de Genève, Versoix, Switzerland LAM/OHP, Marseille, France Vincent Bourrier Observatoire de l’Université de Genève, Sauverny, Switzerland J. Bouvier IPAG, Paris, France Dan J. Bower University of Bern, Bern, Switzerland Brendan P. Bowler Department of Astronomy, The University of Texas at Austin, Austin, TX, USA Alexis Brandeker Department of Astronomy, Institutionen för astronomi, Stockholm University, AlbaNova University Center, Stockholm, Sweden Graziella Branduardi-Raymont Mullard Space Science Laboratory, University College London, Holmbury St Mary, Dorking, Surrey, UK Danielle Briot GEPI, UMR 8111, Observatoire de Paris, 61 avenue de l’Observatoire, Paris, France Lars A. Buchhave Centre for Star and Planet Formation, Natural History Museum of Denmark and Niels Bohr Institute, University of Copenhagen, Copenhagen K, Denmark Artem Burdanov Space Sciences, Technologies and Astrophysics Research (STAR) Institute, Université de Liège, Liège, Belgium Ben Burningham Centre for Astrophysics Research, School of Physics, Astronomy and Mathematics, University of Hertfordshire, Hatfield, UK

xxxii

Contributors

Juan Cabrera Institut für Planetenforschung, Deutsches Zentrum für Luft – und Raumfahrt, Berlin, Germany Andrew Collier Cameron Centre for Exoplanet Science, SUPA School of Physics and Astronomy, University of St Andrews, St Andrews, UK Joleen Carlberg Instruments Division, Space Telescope Science Institute, Baltimore, MD, USA A. Carmona IRAP/OMP, Toulouse, France Davide Cenadelli Osservatorio Astronomico della Regione Autonoma Valle d’Aosta, Nus (Aosta), Italy Z. Challita IRAP/OMP, Toulouse, France William J. Chaplin School of Physics and Astronomy, University of Birmingham, Birmingham, UK Sébastien Charnoz Université Paris Diderot/Institut de Physique du Globe, Paris, France Jørgen Christensen-Dalsgaard Stellar Astrophysics Centre, Department of Physics and Astronomy, Aarhus University, Aarhus C, Denmark Jessie Christiansen IPAC, Pasadena, CA, USA Cathie Clarke Institute of Astronomy, University of Cambridge, Cambridge, UK L. Ilsedore Cleeves Harvard-Smithsonian Center for Astrophysics, Cambridge, MA, USA Alexandre C. M. Correia CIDMA, Departamento de Física, Universidade de Aveiro, Aveiro, Portugal Athena Coustenis LESIA, Observatoire de Paris, CNRS, PSL Universities, UPMC, UPD, Meudon, France Nicolas B. Cowan McGill University, Montréal, QC, Canada Ian A. Crawford Department of Earth and Planetary Sciences, Birkbeck College, University of London, London, UK Aurélien Crida Université Côte d’Azur/Observatoire de la Côte d’Azur, Lagrange, Nice, France Institut Universitaire de France, Paris, France Alex J. Cridland Leiden Observatory, Leiden University, Leiden, The Netherlands Ian J. M. Crossfield Department of Physics, Massachusetts Institute of Technology, Cambridge, MA, USA Nicolas Crouzet Instituto de Astrofísica de Canarias, San Cristóbal de La Laguna, Santa Cruz de Tenerife, Spain

Contributors

xxxiii

Szilárd Csizmadia Institute of Planetary Research, German Aerospace Center, Berlin, Germany DLR, Berlin, Germany Gennaro D’Angelo Theoretical Division, Los Alamos National Laboratory, Los Alamos, NM, USA Rebekah I. Dawson Department of Astronomy and Astrophysics, The Pennsylvania State University, University Park, PA, USA Center for Exoplanets and Habitable Worlds, The Pennsylvania State University, University Park, PA, USA Hans J. Deeg Instituto de Astrofísica de Canarias, La Laguna, Tenerife, Spain Departamento de Astrofísica, Universidad de La Laguna, La Laguna, Tenerife, Spain Denis Defrère Space Sciences, Technologies and Astrophysics Research (STAR) Institute, University of Liege, Liege, Belgium Russell Deitrick Center for Space and Habitability, University of Bern, Bern, Switzerland Magali Deleuil LAM (Laboratoire d’Astrophysique de Marseille), CNRS, CNES, UMR 7326, Aix Marseille Université, Marseille, France X. Delfosse IAP/IdF, Paris, France Jean-Baptiste Delisle Observatoire de l’Université de Genève, Sauverny, Switzerland Laetitia Delrez Astrophysics Group, Cavendish Laboratory, Cambridge, UK Jos de Bruijne European Space Agency (ESA-ESTEC), Noordwijk, The Netherlands Katherine de Kleer Division of Geological and Planetary Sciences, California Institute of Technology, Pasadena, CA, USA Julia de León Instituto de Astrofísica de Canarias, La Laguna, Tenerife, Spain Imke de Pater Department of Astronomy, The University of California at Berkeley, Berkeley, CA, USA Faculty of Aerospace Engineering, Delft University of Technology, Delft, The Netherlands SRON Netherlands Institute for Space Research, Utrecht, The Netherlands Drake Deming Department of Astronomy, University of Maryland, College Park, MD, USA F. Dolon LAM/OHP, Marseille, France

xxxiv

Contributors

Shawn Domagal-Goldman Planetary Systems Laboratory, NASA Goddard Space Flight Center, Greenbelt, MD, USA NASA Astrobiology Institute–Virtual Planetary Laboratory, Seattle, USA Sellers Exoplanets Environments Collaboration, NASA Goddard Space Flight Center, Greenbelt, MD, USA Jean-François Donati CNRS, Institut de Recherche en Astrophysique et Planétologie, Toulouse, France Caroline Dorn University of Zürich, Zurich, Switzerland Laurance R. Doyle Institute for the Metaphysics of Physics, One Maybeck Place, Principia College, Elsah, IL, USA Carl Sagan Center, SETI Institute, Mountain View, CA, USA R. Doyon UdeM/UL, Montréal, QC, Canada Peter E. Driscoll Carnegie Institution for Science, Washington, DC, USA Edward W. Dunham Lowell Observatory, Flagstaff, AZ, USA William R. Dunn Mullard Space Science Laboratory, University College London, Holmbury St Mary, Dorking, Surrey, UK Siegfried Eggl Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA, USA David Ehrenreich Astronomical Observatory of the University of Geneva, Sauverny, Switzerland Thérèse Encrenaz LESIA, Observatoire de Paris, CNRS, PSL Universities, UPMC, UPD, Meudon, France Daniel C. Fabrycky Department of Astronomy and Astrophysics, University of Chicago, Chicago, IL, USA Jacqueline K. Faherty Department of Astrophysics, American Museum of Natural History, New York, NY, USA Alberto G. Fairén Department of Planetology and Habitability, Centro de Astrobiologia (CSIC-INTA), Madrid, Spain Department of Astronomy, Cornell University, Ithaca, NY, USA Rim Fares INAF – Osservatorio Astrofisico di Catania, Catania, Italy María Fernández Jiménez Institut für Planetenforschung, Deutsches Zentrum für Luft – und Raumfahrt, Berlin, Germany P. Figueira CAUP, Porto, Portugal Debra A. Fischer Yale University, New Haven, CT, USA

Contributors

xxxv

Malcolm Fridlund Leiden Observatory, Leiden, RA, The Netherlands Department of Space, Earth and Environment, Chalmers University of Technology, Onsala, Sweden Yuka Fujii Earth-Life Science Institute, Tokyo Institute of Technology, Tokyo, Japan Antonio García Muñoz Zentrum für Astronomie und Astrophysik, Berlin, TU, Germany Technische Universität Berlin, Berlin, Germany B. S. Gaudi Department of Astronomy, The Ohio State University, Columbus, OH, USA Michaël Gillon Space Sciences, Technologies and Astrophysics Research (STAR) Institute, Université de Liège, Liège, Belgium Jonay I. González Hernández Departamento de Astrofísica, Universidad de La Laguna (ULL), La Laguna, Tenerife, Spain Instituto de Astrofísica de Canarias, La Laguna, Tenerife, Spain Andrew P. Gould Ohio State University, Columbus, OH, USA Thomas P. Greene NASA Ames Research Center, Space Science and Astrobiology Division, Moffett Field, CA, USA John Lee Grenfell Department of Extrasolar Planets and Atmospheres (EPA), German Aerospace Centre (DLR), Berlin Adlershof, Germany J.-M. Griessmeier LPC2E-Université d’Orléans/CNRS, Orléans, France Station de Radioastronomie de Nançay, Observatoire de Paris, PSL Research University, CNRS, University of Orléans, OSUC, Nançay, France Tristan Guillot Observatoire de la Cote dAzur, Nice Cedex 4, France Olivier Guyon Astrobiology Center, National Institutes of Natural Sciences, Mitaka, Japan Steward Observatory, University of Arizona, Tucson, AZ, USA National Astronomical Observatory of Japan, Subaru Telescope, National Institutes of Natural Sciences, Hilo, HI, USA Chester E. Harman Department of Applied Physics and Applied Mathematics, Columbia University, New York, NY, USA NASA Astrobiology Institute–Virtual Planetary Laboratory, Seattle, USA NASA Goddard Institute for Space Studies, New York, NY, USA Carole A. Haswell School of Physical Sciences, The Open University, Milton Keynes, UK

xxxvi

Contributors

Artie P. Hatzes Thüringer Landessternwarte, Tautenburg, Germany G. Hébrard IAP/IdF, Paris, France M. Heimpel Department of Physics, University of Alberta, Edmonton, AB, Canada Ravit Helled Institute for Computational Sciences, University of Zurich, Zurich, Switzerland René Heller Max Planck Institute for Solar System Research, Göttingen, Germany Kevin Heng Center for Space and Habitability, University of Bern, Bern, Switzerland Ana M. Heras Science Support Office, Directorate of Science, European Space Agency, ESTEC/SCI-S, Noordwijk, The Netherlands J. J. Hermes Department of Physics and Astronomy, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA O. Hernandez UdeM/UL, Montréal, QC, Canada Frederic V. Hessman Institut für Astrophysik, University of Göttingen, Göttingen, Germany Tori M. Hoehler Space Sciences and Astrobiology Division, NASA Ames Research Center, Mountain View, CA, USA Daniel Huber Institute for Astronomy, University of Hawai‘i, Honolulu, HI, USA Sydney Institute for Astronomy, School of Physics, University of Sydney, Sydney, NSW, Australia Markus Hundertmark Astronomisches Rechen-Institut, Zentrum für Astronomie der Universität Heidelberg (ZAH), Heidelberg, Germany Ryuki Hyodo Earth-Life Science Institute/Tokyo Institute of Technology, Tokyo, Japan Louis N. Irwin University of Texas at El Paso, El Paso, TX, USA Kate Isaak Science Support Office, European Space Agency - ESTEC, AZ, Noordwijk, The Netherlands André Izidoro UNESP, Universidade Estadual Paulista – Grupo de Dinâmica Orbital Planetologia, São Paulo, Brazil Brian Jackson Department of Physics, Boise State University, Boise, ID, USA Markus Janson Department of Astronomy, Institutionen för astronomi, Stockholm University, AlbaNova University Center, Stockholm, Sweden Emmanuël Jehin Space Sciences, Technologies and Astrophysics Research (STAR) Institute, Université de Liège, Liège, Belgium

Contributors

xxxvii

Uffe Gråe Jørgensen Centre for Star and Planet Formation, Niels Bohr Institute, University of Copenhagen, Copenhagen, Denmark Nathan A. Kaib HL Dodge Department of Physics and Astronomy, University of Oklahoma, Norman, OK, USA Paul G. Kalas Astronomy Department, University of California, Berkeley, CA, USA SETI Institute, Mountain View, CA, USA Nancy Y. Kiang NASA Goddard Institute for Space Studies, New York, NY, USA Seung-Lee Kim Korea Astronomy and Space Science Institute (KASI), Daejeon, Republic of Korea Hans Kjeldsen Stellar Astrophysics Centre, Department of Physics and Astronomy, Aarhus University, Aarhus C, Denmark Hubert Klahr Max Planck Institut for Astronomy, Heidelberg, Germany Quinn M. Konopacky Center for Astrophysics and Space Sciences, University of California, San Diego, La Jolla, CA, USA Ravi Kumar Kopparapu NASA Goddard Space Flight Center, Greenbelt, MD, USA University of Maryland, College Park, MD, USA Tommi T. Koskinen Lunar and Planetary Laboratory, University of Arizona, Tucson, AZ, USA D. Kouach IRAP/OMP, Toulouse, France Quentin Kral Institute of Astronomy, University of Cambridge, Cambridge, UK Michael Kramer MPI für Radioastrononomie, Bonn, Germany Laura Kreidberg Harvard Society of Fellows, Cambridge, MA, USA Antoine Labeyrie Collège de France and Observatoire de la Côte d’Azur, Caussols, France M. Lacombe IRAP/OMP, Toulouse, France Antonino F. Lanza INAF-Osservatorio Astrofisico di Catania, Catania, Italy Jacques Laskar ASD, IMCCE-CNRS UMR8028, Paris, France Gregory Laughlin Department of Astronomy, Yale University, New Haven, CT, USA Panayotis Lavvas GSMA, UMR 7331, CNRS, Université de Reims, ChampagneArdenne, Reims, France

xxxviii

Contributors

T. Joseph W. Lazio Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA, USA Alain Lecavelier des Etangs CNRS, Sorbonne Universités, UPMC Univ Paris 06, Institut d’Astrophysique de Paris, Paris, France Chung-Uk Lee Korea Astronomy and Space Science Institute (KASI), Daejeon, Republic of Korea Sandy K. Leggett Gemini Observatory, Northern Operations Center, Hilo, HI, USA Emmanuel Lellouch Laboratoire d’Etudes Spatiales et d’Instrumentation en Astrophysique (LESIA), Observatoire de Paris, Meudon, France A. Lenardic Department of Earth Science, Rice University, Houston, TX, USA Stephen R. Lewis School of Physical Sciences, Faculty of Science, Technology, Engineering and Mathematics, The Open University, Milton Keynes, UK Javier Licandro Instituto de Astrofísica de Canarias, La Laguna, Tenerife, Spain René Liseau Department of Earth and Space Sciences, Onsala Space Observatory, Chalmers University of Technology, Onsala, Sweden Jack J. Lissauer Space Science and Astrobiology Division, NASA Ames Research Center, Moffett Field, CA, USA Joe Llama Lowell Observatory, Flagstaff, AZ, USA Nicolas Lodieu Instituto de Astrofísica de Canarias, La Laguna, Spain Mia Sloth Lundkvist Zentrum für Astronomie der Universität Heidelberg, Heidelberg, Germany Stellar Astrophysics Centre, Aarhus University, Aarhus C, Denmark Timothy W. Lyons NASA Astrobiology Institute and Department of Earth Sciences, University of California Riverside, Riverside, CA, USA A. Määttänen LATMOS/IPSL, UVSQ Université Paris-Saclay, UPMC Univ. Paris 06, CNRS, Guyancourt, France Nikku Madhusudhan Institute of Astronomy, University of Cambridge, Cambridge, UK Fabien Malbet University of Grenoble Alpes, CNRS, IPAG, Grenoble, France Mark S. Marley NASA Ames Research Center, Moffett Field, CA, USA T. R. Marsh Department of Physics, University of Warwick, Coventry, UK Alberto Krone Martins CENTRA/SIM, Faculdade de Ciencias, Universidade de Lisboa, Lisboa, Portugal

Contributors

xxxix

Eduardo L. Martín CSIC-INTA Centro de Astrobiología, Madrid, Spain David V. Martin Swiss National Science Foundation, University of Chicago, Chicago, USA E. Martioli LNA, Itajubá, Brazil Stéphane Mathis Laboratoire AIM Paris-Saclay, IRFU/DAp Centre de Saclay, CEA/DRF – CNRS – Université Paris Diderot, Gif-sur-Yvette Cedex, France LESIA, Observatoire de Paris, PSL Research University, CNRS, Sorbonne Universités, UPMC Univ. Paris 06, Univ. Paris Diderot, Meudon, France Pierre F. L. Maxted Astrophysics Group, Keele University, Keele, UK Michel Mayor Département d’Astronomie, Université de Genève, Versoix, GE, Switzerland Victoria S. Meadows Astronomy Department, University of Washington, Seattle, WA, USA Giuseppina Micela Istituto Nazionale di Astrofisica – Osservatorio Astronomico di Palermo Giuseppe S. Vaiana, Palermo, Italy Y. Micheau IRAP/OMP, Toulouse, France Paolo Molaro INAF Osservatorio Astronomico di Trieste, Trieste, Italy Pilar Montañés-Rodríguez Instituto de Astrofísica de Canarias, La Laguna, Tenerife, Spain F. Montmessin LATMOS/IPSL, UVSQ Université Paris-Saclay, UPMC Univ. Paris 06, CNRS, Guyancourt, France Alessandro Morbidelli Observatoire de la Côte d’ Azur, Boulevard de l’ Observatoire, Université Côte d’ Azur, CNRS, Nice, France Christoph Mordasini Physikalisches Institut, University of Bern, Bern, Switzerland Claire Moutou CNRS/CFHT, Kamuela, HI, USA CNRS, LAM, Laboratoire d’Astrophysique de Marseille, Aix Marseille University, Marseille, France UdeM/UL, Montréal, QC, Canada Gijs D. Mulders Lunar and Planetary Laboratory, The University of Arizona, Tucson, AZ, USA Yasushi Muraki Institute for Space-Earth Environment Research, Nagoya University, Nagoya, Japan Richard P. Nelson Astronomy Unit, Queen Mary University of London, London, UK

xl

Contributors

Eric L. Nielsen Kavli Institute for Particle Astrophysics and Cosmology, Stanford University, Stanford, CA, USA Stephanie L. Olson NASA Astrobiology Institute and Department of Earth Sciences, University of California Riverside, Riverside, CA, USA Jerome A. Orosz Astronomy Department, San Diego State University, San Diego, CA, USA Enric Pallé Instituto de Astrofísica de Canarias, La Laguna, Tenerife, Spain L. Parès IRAP/OMP, Toulouse, France Byeong-Gon Park Korea Astronomy and Space Science Institute (KASI), Daejeon, Republic of Korea Vivien Parmentier Lunar and Planetary Laboratory, University of Arizona, Tucson, AZ, USA Hannu Parviainen Instituto de Astrofísica de Canarias, La Laguna, Spain Departamento de Astrofísica, Universidad de La Laguna, La Laguna, Spain Francesco Pepe Département d’Astronomie, Observatoire de l’Université de Genéve, Versoix, GE, Switzerland Joshua Pepper Department of Physics, Lehigh University, Bethlehem, PA, USA Thomas Pfeil Max Planck Institut for Astronomy, Heidelberg, Germany Noemí Pinilla-Alonso Florida Space Institute, UCF, Orlando, FL, USA Ralph E. Pudritz Department of Physics and Astronomy, McMaster University, Hamilton, ON, Canada Origins Institute, McMaster University, Hamilton, ON, Canada Laurent Pueyo STScI, Baltimore, MD, USA P. Rabou IPAG, Paris, France Saul A. Rappaport Department of Physics, Massachusetts Institute of Technology, Cambridge, MA, USA Heike Rauer Institute of Planetary Research, German Aerospace Center, Berlin, Germany Department of Astronomy and Astrophysics, Berlin University of Technology, Berlin, Germany Sean N. Raymond Laboratoire d’Astrophysique de Bordeaux, University of Bordeaux, CNRS, Bordeaux, France Peter L. Read Atmospheric, Oceanic and Planetary Physics, Clarendon Laboratory, University of Oxford, Oxford, UK

Contributors

xli

Ansgar Reiners Institut für Astrophysik, Georg-August-Universität Göttingen, Göttingen, Germany Christopher T. Reinhard School of Earth and Atmospheric Science, Georgia Institute of Technology, Atlanta, GA, USA V. Reshetov NRC-H, Victoria, Canada Ignasi Ribas C/Can Magrans, Institut de Ciències de l’Espai – IEEC/CSIC, Bellaterra, Spain Aki Roberge Exoplanets and Stellar Astrophysics Lab, NASA Goddard Space Flight Center, Greenbelt, MD, USA Tyler D. Robinson Department of Physics and Astronomy, Northern Arizona University, Flagstaff, AZ, USA Florian Rodler European Southern Observatory, Vitacura, Santiago, Chile Antoine Rozel ETH Zürich, Zürich, Switzerland L. Saddlemyer NRC-H, Victoria, Canada A. Sánchez-Lavega Departamento Física Aplicada I, Escuela de Ingeniería de Bilbao, Universidad del País Vasco UPV/EHU, Bilbao, Spain Alexandre Santerne Aix Marseille Univ, CNRS, CNES, LAM, Marseille, France Nuno C. Santos Instituto de Astrofísica e Ciências do Espaço, Universidade do Porto, CAUP, Porto, Portugal Departamento de Física e Astronomia, Faculdade de Ciências, Universidade do Porto, Porto, Portugal Joachim Saur Institute of Geophysics and Meteorology, University of Cologne, Cologne, Germany Hilke E. Schlichting University of California, Los Angeles, CA, USA Massachusetts Institute of Technology, Cambridge, MA, USA Jean Schneider LUTh, UMR 8102, Observatoire de Paris, 5 place Jules Janssen, F-92195 Meudon Cedex, France LUTH, Observatoire de Paris, PSL Research University, CNRS, Université Paris Diderot, Meudon, France Andreas Schreiber Max Planck Institut for Astronomy, Heidelberg, Germany Edward W. Schwieterman Department of Earth Sciences, University of California, Riverside, CA, USA Blue Marble Space Institute of Science, Seattle, WA, USA Salvatore Sciortino INAF – Osservatorio Astronomico di Palermo “Giuseppe S. Vaiana”, Piazza del Parlamento, 1, Palermo, Italy

xlii

Contributors

Sara Seager Department of Earth, Atmospheric and Planetary Sciences, Massachusetts Institute of Technology, Cambridge, MA, USA Antígona Segura Instituto de Ciencias Nucleares, Universidad Nacional Autónoma de México, Circuito Exterior s/n, Ciudad Universitaria, Ciudad de México, México Evgenya L. Shkolnik ASU School of Earth and Space Exploration, Tempe, AZ, USA Víctor Silva Aguirre Stellar Astrophysics Centre, Department of Physics and Astronomy, Aarhus University, Aarhus C, Denmark Norman H. Sleep Department of Geophysics, Stanford University, Stanford, CA, USA Sanjoy M. Som Blue Marble Space Institute of Science, Seattle, WA, USA Alessandro Sozzetti INAF – Osservatorio Astrofisico di Torino, Pino Torinese, Italy Keivan G. Stassun Department of Physics and Astronomy, Vanderbilt University, Nashville, TN, USA N. Striebig IRAP/OMP, Toulouse, France Antoine Strugarek Laboratoire AIM, DRF/IRFU/DAp, CEA Saclay, Gif-surYvette Cedex, France Département de Physique, Université de Montréal, Montréal, QC, Canada Motohide Tamura University of Tokyo and Astrobiology Center, Tokyo, Japan the SPIRou Team the SPECULOOS and TRAPPIST Teams S. Tibault UdeM/UL, Montréal, QC, Canada Amaury H. M. J. Triaud School of Physics and Astronomy, University of Birmingham, Birmingham, UK Mikko Tuomi Centre for Astrophysics Research, Science and Technology Research Institute, University of Hertfordshire, Hatfield, UK Andrzej Udalski Warsaw University Observatory, Warszawa, Poland Stéphane Udry Département d’Astronomie, Université de Genève, Versoix, GE, Switzerland Geoffrey K. Vallis College of Engineering, Mathematics and Physical Sciences, University of Exeter, Exeter, UK

Contributors

xliii

Rik van Lieshout Institute of Astronomy, University of Cambridge, Cambridge, UK Steven D. Vance Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA, USA Andrew Vanderburg Harvard-Smithsonian Center for Astrophysics, Cambridge, MA, USA Julia Venturini Physikalisches Institut, Universitaet Bern, Bern, Switzerland University of Zurich, Zurich, Switzerland Aline A. Vidotto School of Physics, Trinity College Dublin, The University of Dublin, Dublin-2, Ireland S. Y. Wang ASIAA, Taipei, Taiwan William F. Welsh Astronomy Department, San Diego State University, San Diego, CA, USA Robert A. West Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA, USA Anthony P. Whitworth School of Physics and Astronomy, Cardiff University, Wales, UK Peter K. G. Williams Harvard-Smithsonian Center for Astrophysics, Cambridge, MA, USA Joshua N. Winn Department of Astrophysical Sciences, Princeton University, Princeton, NJ, USA Alexander Wolszczan Department of Astronomy and Astrophysics, and Center for Exoplanets and Habitable Worlds, The Pennsylvania State University, University Park, PA, USA Jason T. Wright Department of Astronomy and Astrophysics, Center for Exoplanets and Habitable Worlds, The Pennsylvania State University, University Park, PA, USA Breakthrough Listen Laboratory, Department of Astronomy, University of California, Berkeley, CA, USA Mark C. Wyatt Institute of Astronomy, University of Cambridge, Cambridge, UK Philip Yock Department of Physics, University of Auckland, Auckland, New Zealand E. D. Young Department of Earth, Planetary, and Space Sciences, University of California, Los Angeles, Los Angeles, CA, USA

xliv

Contributors

Patrick A. Young School of Earth and Space Exploration, Arizona State University, Tempe, AZ, USA Philippe Zarka LESIA, Observatoire de Paris, CNRS, PSL, UPMC/SU, UPD, Meudon, France Station de Radioastronomie de Nançay, Observatoire de Paris, CNRS, PSL, Univ. Orléans, Nançay, France B. Zuckerman Department of Physics and Astronomy, University of California, Los Angeles, Los Angeles, CA, USA

Section I Exoplanet Research: A History of Discovery Tsevi Mazeh

Tsevi Mazeh is Professor Emeritus at Tel Aviv University, where he has served as researcher and lecturer since 1979. In 1984, Mazeh initiated the first radial-velocity search for extrasolar planets in pursuit of his hypothesis that massive planets could exist close to their parent stars, at that time widely believed to be impossible. The observations, carried out with David Latham (CfA) and in cooperation with Michel Mayor (Geneva), provided the discovery in 1989 of the first known massive candidate for an extrasolar planet – HD114762b. Since then he is devoting his career to the study of binary stars and extrasolar planets.

1

The Discovery of the First Exoplanets Davide Cenadelli and Andrea Bernagozzi

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Improvements in Doppler Measurements from the 1950s to the 1970s: The Pioneers . . . . . . . The Case of HD 114762 b . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Case of ” Cep b . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Discovery of 51 Peg b, 70 Vir b, and 47 UMa b . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 Peg b Questioned and Finally Confirmed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross-References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

4 6 8 11 12 15 17 18 18

Abstract

In this chapter, we will deal with the discovery of the first extrasolar planets around normal stars. This discovery took place in the mid-1990s thanks to the analysis of periodic Doppler shifts in stellar spectra and turned out to be a landmark achievement, in that it established a new field of research that is growing at full speed since then, and at the same time, it answered a question – do other worlds exist in the cosmos? – that dates back to ancient times. This major result was made possible by the impressive improvement in Doppler analysis techniques during the second half of the twentieth century, until the precision

D. Cenadelli () Osservatorio Astronomico della Regione Autonoma Valle d’Aosta, Nus (Aosta), Italy e-mail: [email protected] A. Bernagozzi Scuola di Scienze e Tecnologia, Sezione di Geologia, UNICAMearth Working Group, Università degli Studi di Camerino, Camerino, Italy Osservatorio Astronomico della Regione Autonoma Valle d’Aosta, Nus (Aosta), Italy e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_134

3

4

D. Cenadelli and A. Bernagozzi

boosted to values suitable to detect massive planets in close-in orbits. In the late 1980s, two objects came to the limelight: HD 114762 b and ” Cep b. They were suspected to be exoplanets, but it was not possible to rule out alternative explanations, so no undisputable discovery could be claimed. A few years later, in 1995, the discovery of 51 Peg b was announced and confirmed as the first secure one. However, the features of this planet were quite weird as compared to those of the planets of the Solar System, as it was massive and very close to its parent star. Its discovery overturned the previously widespread idea that other systems had to be made like ours and opened the floodgates to the spotting of many other planets. In this chapter, we will tell this story, mainly dwelling upon the historical and epistemological aspects. Keywords

First exoplanets · HD 114762 b · Cep b · 51 Peg b · 70 Vir b · 47 UMa b · radial velocities · dynamics of scientific discovery

Introduction The idea of innumerable worlds existing in the cosmos is venerably old. It dates back, at least, to the ancient Greek civilization. It is remarkable that this idea belonged for millennia to the realm of pure speculation and eventually entered the territory of real science in the second half of the twentieth century, to culminate in the discovery of the first exoplanets in the 1990s. We are the first generation in the history of mankind to know that the ancient hypothesis of countless planets existing in space is actually true. Finding out these bodies has been a landmark achievement, for several reasons. It brought to light a new field of research that is impressively growing over time, it set up a new community of planet hunting astronomers, it boosted the interest in our own Solar System and the mechanism of its formation, it renewed the interest in the search for life in the Universe and in the feasibility of interstellar flight (could we ever visit any of these planets?), and of course, it entered popular culture at large. The establishment of the new field of research that led to the discovery of the first exoplanets was, no surprise, a gradual process that was undertaken by some pioneers after the mid of the twentieth century and gradually attracted more and more astronomers in the following decades. As many other scholars entering new territories before them, these astronomers seldom had to confront skepticism about the real existence or, at least, the actual possibility of detecting exoplanets. They persevered and finally gained their goal, although the first planets they found had really strange characteristics as compared to those of the Solar System, and to some extent, they came unexpected. The technique through which the first exoplanets were discovered is the spectroscopic method, based on the observation of small Doppler shifts in the spectrum of a star. These shifts are due to the motion of the star around the center of mass with a planet that is not directly visible at the telescope but whose presence can be inferred thanks to its gravitational pull on its parent star. In other words, if a star happens

1 The Discovery of the First Exoplanets

5

to show a periodic Doppler shift, we can deduce that the star is wobbling back and forth around a center of mass with another object. Knowing the mass of the star and its velocity, it is possible to calculate the mass of this companion object and, if it is small enough, to infer that it is a planet. If it is bigger, it can be a brown dwarf or a faint star, and in this latter case, we can try to see its dim light blended with the one of its brighter companions. The spectroscopic technique is also called the method of radial velocities (shortened in RV), because what we actually measure is the radial component of the velocity of the star, i.e., the component along the line of sight. This entails that we don’t gain information about its tangential component. This is an important issue: the star’s real speed triggered by the invisible companion will be larger than the one we measure through this method and equal only if its orbit is seen perfectly edge on. Thus, we can assess a minimum value for the companion’s mass, not the real value. If we get a mass estimate of – say – 1 MJ (here and henceforth MJ D mass of Jupiter), the unseen companion will have such a mass if the orbit is seen edge on, some MJ if it seen with intermediate orientation, and many MJ if the orbit is almost face-on. In this latter case, the actual value of the mass can increase to the point the companion is a brown dwarf or even a small star. If we observe a transit of the planet in front of the star, we can be sure that the orbit is seen edge on (or almost edge on), and this is the best possible case, as the mass we get is the real one, but usually that does not happen. This is something we must bear in mind when speaking of planets found via the spectroscopic method. Which precision is necessary to reach to detect the oscillations of a star around the center of mass with a planet? If we consider the Solar System, the movement the Sun undergoes because of the gravitational pull of Jupiter – by far the largest among all planets – is around 12 m s1 , and a great precision is needed to spot such a tiny speed. If we imagine that other systems are made like ours, with giant, Jupiter-like planets far from their parent stars and small-mass ones closer, it is necessary to attain stunningly high precisions to discover exoplanets. Before the first exoplanets were eventually discovered, the idea of a similarity between other systems and ours was the prevailing one. Although this idea was pretty natural, as we shall see, it turned out to be wrong and rather led astronomers astray than helping them out. The story of the discovery of the first exoplanets got started several decades ago, when astronomers began to improve the precision reachable by Doppler measurements, in a first time with the aim of studying the motion of stars in space to investigate the galactic dynamics or to detect new binaries and then, as the attainable precision grew higher, gradually having in mind also the possible existence of substellar mass bodies, companion to real stars. The first clues of their possible existence were collected in the late 1980s, and the first discoveries came a few years later, in the 1990s. This happened thanks to a sizeable improvement in Doppler techniques that was rooted in some pioneering works performed in the previous decades. In this chapter, we will first introduce the work of these pioneers (section “Improvements in Doppler Measurements from the 1950s to the 1970s: The Pioneers”) and then focus on the first objects suspected to be exoplanets around

6

D. Cenadelli and A. Bernagozzi

normal stars (sections “The Case of HD 114762 b” and “The Case of ” Cep b”) to come finally to the first recognized discoveries (sections “The Discovery of 51 Peg b, 70 Vir b and 47 UMa b” and “51 Peg b Questioned and Finally Confirmed”). Some final remarks (section “Concluding Remarks”) will follow. We shall not speak of the discovery of planets orbiting pulsars (the first were discovered around pulsar PSR B1257 C 12; see Wolszczan and Frail 1992; Wolszczan 1994) but just normal stars. The main focus of this chapter will be on the dynamics of the discovery, from the scientific, historical, and epistemological point of view. We shall not dwell too much on the technological and instrumental facets that will be just hinted at when necessary for a deeper understanding.

Improvements in Doppler Measurements from the 1950s to the 1970s: The Pioneers The age of pioneers began in 1953, when Peter B. Fellget came about with the idea of a new equipment (Fellget 1953) aimed at improving the Doppler precision attainable at the time. He suggested to convey the light of a stellar spectrum onto a mask, manufactured in opaque glass, with tiny slits traced on it in correspondence to the dark absorption lines in the stellar spectrum (template spectrum). The mask blocked out starlight, except in correspondence to the slits where light could pass through the mask and reached a photomultiplier placed behind it. When the spectrum of a star is collected, according to its spectral type, it is well known which are the exact wavelengths of the absorption lines; using a laboratory spectrum as a wavelength calibrator (reference spectrum), it is possible to place accurately the mask in such a way that the slits perfectly match the absorption lines. When that happens, a minimum amount of light reaches the photomultiplier, and this brings about a null or very weak electric signal. But this does not happen if the lines are Doppler-shifted: in that case, the signal turns out to be stronger. Hence, moving back and forth the mask until the signal comes to a minimum permits to obtain a differential measurement of the wavelength shift of the spectral lines of the star, i.e., its radial velocity. For the sake of precision, it is necessary to state that the Doppler-shifted wavelengths are not simply translated, as the shift is proportional to the wavelength, so that the spectrum is actually rescaled, but it is possible to correct for this. A drawback of this method is the fact that the slits on the mask are fixed, so they simulate just the spectrum of a star of a given spectral class and are suitable only for stars of that class or not too different. The method proposed by Fellget has the further advantage that all the light across the spectrum – not a single line at a time – is used to work out a measurement of radial velocity, so it works also for faint stars. Fellget put forth the idea but did not put in practice. It was Roger F. Griffin, over a decade later, who did it (Griffin 1967). He built a spectrophotometer completely devoted to stellar radial velocity measurements. Griffin’s instrument was called “cross-correlation spectrograph” because it cross correlated the stellar spectrum

1 The Discovery of the First Exoplanets

7

and the template spectrum on the mask. In 1973, he, together with Rita E. Griffin, perfected the apparatus (Griffin and Griffin 1973) considering the different path of the stellar light and the reference spectrum it was compared to as a possible source of error. To avoid this, they proposed to use as comparison spectrum terrestrial atmospheric bands. The Griffins estimated this could yield a remarkable 10 m s1 precision for selected stars. To attain such a precision, not only was it necessary to correct for the orbital and diurnal motion of the Earth but also for the motion of our planet around the Earth-Moon center of mass and the motion of the Sun around the center of mass with Jupiter, as they are of the same order of magnitude of the equipment’s precision. It is possible to reverse this consideration: if we must account for the motion of the Sun, perhaps we can spot a similar motion undergone by another star. In other words, the Griffins verged on the precision limit necessary to discover a Jupiter-mass planet orbiting a Sun-like star with a similar mean distance. The Griffins perfectly realized this and explicitly claimed that one of the merits of such a precision was the possibility to enter the unexplored territory of the search for planetary companions to other stars. They also realized that a velocity of around 10 m s1 can be comparable to that of gas motions in stellar photospheres, so caution was needed in assessing the real origin of a Doppler shift of this magnitude. However, they realized that these shifts would be perfectly periodic if due to a planet, and this could help to tell apart the case of a planetary companion from that of motions of stellar gases. The Griffins did build their own instrument and performed measurements with it, achieving a precision better than 100 m s1 for selected stars (Griffin and Griffin 1973). That was considered very high precision at the time. The issues raised by the Griffins are very relevant. The first is the supposed similarity of another planetary system to ours in terms of distance between a star and its giant planets. The second is the existence of stellar photospheric phenomena able to trigger Doppler shift similar to the ones due to a planet. These issues are of different kinds and unrelated to each other, but we will come back extensively to both of them as they are continually intertwined with the search for exoplanets until their discovery. Griffin contributed to another big step forward: he inspired Michel Mayor to pursue high-precision radial velocity measurements. The two met in 1970 at Cambridge where both were attending a conference on stellar dynamics. At the time, Mayor was deeply involved in the tricky issue of working out a model of the spiral arms of galaxies. Their existence and stability imply that they are not material arms, because in that case, the differential rotation of galaxies would mix them up over time. In 1964, a paper by Lin and Shu (1964) had argued that they can be interpreted as density waves, like modes of the gravitational potential of a galaxy, with stars and nebulae entering and exiting them. Mayor undertook the task of confirming this theory investigating the motion of stars in our galaxy to see if it was possible to work out a density wavelike structure. Thus, he needed good quality spectra to investigate the radial velocities via the Doppler effect. While musing over a way to improve the existing precision of radial velocities, he bumped into Griffin tinkering with his newly conceived spectrograph. Such an instrument was what

8

D. Cenadelli and A. Bernagozzi

Mayor needed, and with the fundamental aid by Prof. André Baranne, an eminent optician of Marseille University, he built his own, named CORAVEL: it was an innovative-design spectrograph that reached a precision of 0.2 km s1 . CORAVEL employed a high-dispersion échelle diffraction lattice and a mask reproducing 1.500 lines of the giant star Arcturus. CORAVEL was mounted at the Cassegrain focus of the 1 m telescope of the Observatoire de Haute-Provence (OHP). In January 1981, a twin device was mounted at the 1.5 m Danish telescope at La Silla. An observational campaign was planned from 1977 to 1990. During the 1970s, another step ahead was performed by K. Serkowski and coworkers at the Lunar and Planetary Laboratory of the University of Arizona (Serkowski 1978; Serkowski et al. 1979a, b). They devised a farseeing design that employed for wavelength calibration “a NO2 absorption cell illuminated by a lamp giving a continuous spectrum” (Serkowski et al. 1979b, p. 131) and a fiber for good image scrambling. This boosted precision beyond the limit of 0.1 km s1 and up to 20 m s1 . Although it was an innovative result by the time, it was still not enough to spot a Jupiter-like planet orbiting a Sun-like star at a few AU distance. But what if a massive planet were much closer? It would trigger a larger, and easier to detect, movement on the star. Unfortunately, at the time, the prevailing view was that typically giant planets can only be found far away from their stars. This was a natural belief, because the only known system to be used as a guidance was ours. Moreover, it was theoretically claimed that in protoplanetary disks giant planets are typically created beyond the so-called ice line, where the majority of the material of the disk is accumulated during the turbulent early stages of formation of a new planetary system. To be in close-in orbits, they must migrate from their birthplace, and almost nobody was thinking of this possibility at the time. In truth, the newfangled idea of “planetary migration” had already been put forth theoretically since the 1980s (Goldreich and Tremaine 1980; Lin and Papaloizou 1986; Artymowicz 1993), before any planet with this far-fetched feature was discovered. However, these ideas were not widespread and basically excluded from common opinion about planetary formation (see, e.g., Mayor et al. 2014, p. 329).

The Case of HD 114762 b As time went by, the increasing precision attained by spectrographs was pushing the astronomers into the substellar mass domain. Here, before seeing planets, they expected to bump into another kind of object, whose existence was theoretically conjectured in the 1960s but which had never been found yet: brown dwarfs. Brown dwarfs are intermediate objects between planets and stars: more massive than a planet and less massive than a star. They are not massive enough to ignite hydrogen fusion in their core and become real stars. They shine very faintly and typically in the infrared, so they are hardly observable without powerful infrared telescopes or satellites. For example, today the closest known brown dwarfs are the two components of the double system Luhman 16, two companion objects that lie a

1 The Discovery of the First Exoplanets

9

mere 2 pc away and constitute the third nearest system to ours, after Alpha Centauri and Barnard’s Star. Although they are so near, they were discovered on images taken by the WISE infrared space telescope not earlier than 2010–2011 and announced in 2013 (Luhman 2013). With state-of-the-art technology in the 1980s, the easiest way to spot a brown dwarf was to resort to the Doppler technique. The search for brown dwarfs contributed to foster the development of sensitive equipment. As there were no constraints as to where a brown dwarf orbiting a star should lie, i.e., closer or further, it was expected that in some cases such objects could be in close-in orbits and trigger large motions on their companion stars (to elaborate upon the brown dwarf issue, see Mayor and Frei 2003, pp. 113–133; Joergens 2014). Mayor himself, when using CORAVEL and then its successor ELODIE (about which, see section “The Discovery of 51 peg b, 70 vir b, and 47 uma b”), thought of brown dwarfs as a primary goal of his research, although he left the door open to exoplanets as well. In fact, as we have seen, in the early 1980s, the precision in radial velocities attainable by available spectrographs was of the order of 0.1 km s1 , with some attempts to go beyond that barrier and reach a few tens of m s1 , and that was thought to fall short of the precision requested to discover another planetary system, although it was deemed to be enough to seek brown dwarfs. At this point, a young scholar, Tsevi Mazeh, came up proposing to see things from a different point of view. Mazeh knew that the existing theories claimed that massive planets should not form close to stars. However, having been trained as a theoretician himself, he knew that theories can be challenged. In Mazeh’s words: “I did not trust the theories. Maybe this is because of my background. Usually, observers tend to trust theory, and theoreticians tend to trust the observers. I did my Ph.D. as a theoretician, and only then moved to do observations, so I did not trust the theories. I claimed that the only example we have is the Solar System, and it is not justified to assume that all solar systems in the Universe are the same. Therefore, I postulated that there are Jupiters that are very close to their parent stars, and let us look for them” (Cenadelli and Bernagozzi 2015, p. 532). Furthermore, Mazeh suggested to seek planets around low-mass stars like M dwarfs, because we can expect a given planet to induce a larger reflex motion on a smaller than on a higher mass star. In 1984, Mazeh went to Harvard University visiting David W. Latham, an associate director for the Optical and Infrared Astronomy at the Oak Ridge Observatory, operated by the Harvard-Smithsonian Center for Astrophysics (CfA). The two met at the observatory in September 1984, and Mazeh proposed to Latham to use the facilities of the CfA to seek planets around red dwarfs (Latham 2012). At the CfA the so-called digital speedometers attained a precision of a few hundred m s1 . A few tens of stars of this kind were looked at in the following years, but nothing was found. However, as a byproduct of these observations, something was noticed about HD 114762, an F9 dwarf star, similar in mass to the Sun. This star was looked at because it is an IAU Radial-Velocity Standard Stars, and Latham proposed to take spectra of two such stars close to any red dwarf to correct for possible systematic errors. The idea was that if all three stars showed the same velocity drift, that would be the sign of a systematic error, whereas a velocity shift of the target red dwarf not shared by the two comparison stars would be an evidence of

10

D. Cenadelli and A. Bernagozzi

real motion. This work was also aimed at providing better radial velocity standards to the community of astronomers. Thanks to a newly built fiber feed for the digital speedometer, precision increased, and Latham could see that these higher-precision measurements for HD 114762 were a few hundred m s1 different from the already available lowerprecision ones. An overall analysis of 66 spectra of that star collected over 8 years, performed thanks to codes written by Mazeh, allowed Latham to extract a periodic signal of 84 days with amplitude of velocity variation of around 0.5 km s1 . It should be noticed that 66 measurements over 8 years was an unusually dense time sampling. In fact, the common views about the occurrence of giant planets at that time favored long, multiyear orbits, and astronomers deemed that a proper strategy to find them was to take one or a few spectra per year, and that was enough. According to Latham: “In retrospect, I adopted the correct observing strategy for finding short-period planets for a completely unrelated reason. Sometimes altruism pays off in unexpected ways” (Cenadelli and Bernagozzi 2015, p. 533). The 84-day signal suggested the presence of an unseen companion, called HD 114762 b, with that orbital period. The orbital period was similar to Mercury’s, and the same happened for the semimajor axis, as the stellar mass was similar to the Sun’s. The velocity of 0.5 km s1 implied a minimum mass of 11 MJ , and the eccentricity had a sizeable value around 0.3, much larger than any planet in the Solar System. Latham alerted Mayor to have other spectra about this star. Being it an IAU standard, anybody working with radial velocities was supposed to have looked at it and that was the case: Mayor had already collected about 100 radial velocities over several years and could confirm the signal. What was orbiting around HD 114762? As the suggested minimum mass was 11 MJ , it turned out to be a big planet in the case the orbit were almost face-on, a brown dwarf for higher angles of the orbital plane to the line of sight, or even a small star if the Earth were looking straight down on the orbital plane. Latham went public at the August 1988 IAU meeting in Baltimore, underlining the uncertain nature of this object, and then the discovery was published in a paper in Nature with the cautious title “The unseen companion of HD 114762: a probable brown dwarf” (Latham et al. 1989). The discussion about the real nature of HD 114762 b began, and in lack of precise information about its actual mass, plausibility considerations were put forth. Against the planetary hypothesis stood some considerations, namely, (1) the short semimajor axis, as giant planets were not supposed to exist inside the ice line; (2) the great mass, by far bigger than Jupiter’s even in the most favorable case; and (3) the large eccentricity. Large eccentricities are not typical of planets in the Solar System, but are common in binary stellar system; thus this feature seemed to be more likely for a brown dwarf that, in principle, could share this characteristic with stars (no brown dwarfs had been yet discovered and anyone could think whatever they wanted in this respect). Today, many massive planets with high eccentricities and close-in orbits are known, and all the aforementioned objections appear to be just prejudices, albeit natural at the time. The real point is the assessment of 11 MJ as the minimum mass that would push the object into the brown dwarf domain even for moderate inclinations of the orbit to the line of sight. It should be noticed, however, that

1 The Discovery of the First Exoplanets

11

the boundary between the dominion of big planets and that of brown dwarfs is still debated. In terms of mass, a “shady area” could exist between the two, with some degree of overlapping, and the mass criterion is not the only possible one: another way to distinguish between the two categories of objects could be based on the mechanism of formation. Another issue was not mentioned at the time, but came about in the following years, when evidence accumulated about the positive correlation between the presence of planets and the metallicity of stars. HD 114762 is a metal-deficient star, and today scientists reckon that massive planets are less likely to form in an environment poor of metals; less likely but not impossible. In other words, the objections that were put forth just after the discovery are dismissed today, while this latter one, that was not known by the time emerged later. This suggests that not even the latter is a conclusive one: scientific knowledge increases and conceptual paradigms can be overturned, and this is particularly true in the field of exoplanets, a young field that is undergoing an impressive growth these days. Latham states this very well: “It is a lot of fun to challenge theoretical ideas with new observational results that do not seem to fit the generally accepted paradigm. A lot of the progress in the field of exoplanets has come this way. It is wise to keep an open mind and be on the lookout for anomalies” (Cenadelli and Bernagozzi 2015, p. 534).

The Case of ” Cep b Together with HD 114762 b, another intriguing object came to the limelight in the late 1980s: ” Cep b. At the time, besides the scholars we already met, other groups were engaged in the search for exoplanets. Some of them stand out in the pursuit of innovative solutions in high-precision research: in Canada Bruce Campbell and Gordon A.H. Walker, in Texas Artie P. Hatzes and William D. Cochran, and in California Geoffrey W. Marcy and R. Paul Butler, together with several coworkers. Walker and colleagues started the first high-precision radial velocity search for exoplanets using the 3.6 m Canada-France-Hawaii Telescope (CFHT). Exploiting a laboratory reference made of hydrogen fluoride (HF) in a vessel that was passed through by the stellar light, they assessed that they could attain a 15 m/s precision (Campbell and Walker 1979; Walker 2012). The observational schedule was made up of just a few observations per year. This is an unfit choice based on what we know today of exoplanets but was deemed to be acceptable at the time, because “it was assumed that extra-solar planets would most likely resemble Jupiter in mass and orbit” (Walker 2012, p. 9). Besides this scientific assumption, there were other reasons, more cultural than strictly science-related, tied to the skepticism about the actual possibility of discovering exoplanets that was widespread in the community of astronomers at the time and the consequent difficulty in allocating telescope time. In 1988, when putting forth the results of their observations of 16 dwarfs and subgiants, each observed a few times per year, Campbell, Walker, and Yang noticed that seven of them could have a substellar mass companion (Campbell et al. 1988). The most interesting one was the K1 star ” Cep: it turned out to be orbited by a small-mass, red dwarf companion but also by another object with an estimated

12

D. Cenadelli and A. Bernagozzi

minimum mass around 1.7 MJ and an orbital period of about 2.7 years. The Canadians were cautious about this possible discovery because they were perfectly aware – as the Griffins before them – that stellar photospheric phenomena could trigger Doppler shifts similar to the one brought about by a planet. In fact, after taking new spectra of ” Cep, a few years later, they questioned themselves the real existence of the planet (Walker et al. 1992) putting forth a new hypothesis: the observed pattern in radial velocity could be due to gaseous motions in the stellar atmosphere triggered by its rotation. The stellar rotational period, as worked out from an analysis of the spectral lines profile, turned out to be pretty close to the period of the supposed-to-be planet. Moreover, Walker and coworkers made another interesting point: “Formally, an explanation [.. .] in terms of the reflex motion about a Jupiter-mass planet [.. .] is still viable but, to survive so close to a stellar K giant, the planet would need to be solid. A Jupiter mass of solid material would greatly exceed the combined masses of the terrestrial planets and the cores of the solar system gas giants” (Walker et al. 1992, pp. L93–94). To sum up, the 1992 paper by Walker and coworkers is remarkable because it discusses both the issues we saw to be strictly related with the search for exoplanets: (1) the idea that gas movements on the stellar surface could mime a nonexisting planet in Doppler measurements and (2) the fact that finding something considerably different from the Solar System was thought to be an unlikely possibility. Hence, the case of ” Cep b is very intriguing, and much the more so if we consider that over a decade later these scientists, together with others (Hatzes et al. 2003), came to the conclusion that the planet is really there, and today the existence of ” Cep b is generally acknowledged.

The Discovery of 51 Peg b, 70 Vir b, and 47 UMa b In the late 1980s, while the Canadians were carrying out these high-precision measurements, it became clear to Mayor that it was necessary to devise a new instrument with better performance than CORAVEL. CORAVEL’s successor, named ELODIE, was designed with the fundamental aid by Baranne, and it was a newly conceived instrument, in that it replaced the fixed template of CORAVEL with a software-generated one. This entailed a much greater flexibility, because the software template could account for the dependence of the line shift upon the temperature of the air and the atmospheric pressure, whereas the physical template could not be changed in dependence of such conditions. Moreover, it employed optical fibers. The use of fibers allowed another major improvement: the spectrograph could be detached from the telescope and light simply conveyed to it via the fibers. This eliminated any problem due to mechanical instabilities; in other words, the perfect control of the gravity center of the equipment was no more a critical point. Instead of resorting to the technique of using iodine cell along the optical path for wavelength calibration, a second optical fiber conveyed the light of a thorium lamp continuously in parallel to the star.

1 The Discovery of the First Exoplanets

13

ELODIE was mounted on the 1.93 m telescope at the OHP, and the commissioning was done by Mayor and his Ph.D. student Didier Queloz. They realized that the jump ahead in precision as compared to CORAVEL was really great: from 300 to somewhere between 10 and 15 m/s. Another outcome of this instrument design was that it supplied real-time reduced data instead of a typical CCD raw image that needs to go through the data reduction and analysis pipeline. In other words, the measured radial velocities were available at once. This is a key point. Was this expected to be a major advantage? If one expects to find a planet orbiting a star in a Jupiter-like orbit, it is not, because the orbital period lasts several years. But if one does not know what he is going to find and leaves all possibilities open, in the case of a short-period orbit, it can be. Mayor was not seeking only planets but also brown dwarfs, for which there were no prejudices about the possible orbital periods, as there were no theoretical constraints from the one side, and no one had been found yet from the other. Moreover, in the previous years with CORAVEL, Mayor and coworkers had carried out an important survey of solar-type stars in the galactic neighborhood, to search for binaries and get constraints on the mechanism of formation of stars (Duquennoy et al. 1991; Duquennoy and Mayor 1991). In other words, they were not starting a search for exoplanets from scratch; rather, they had been involved for a long time in an investigation of stars that could be extended to brown dwarfs and possibly planets. It was a matter of extrapolating a research already in focus to the smaller-mass domain. The search for exoplanets was a direct continuation of this previous work, once the attainable precision grew enough to bring planets within reach or, at least, massive planets in close-in orbits. This had a major bearing on the observational time schedule: as the observational strategy perfected for stars was made up of a dense schedule, with frequent observations, it was a natural choice to go on like this also for smaller-mass bodies. Finally, there was a more cultural reason that played in favor of Mayor and Queloz. At the time the search for exoplanets was hampered by the idea that no one had been ever found and thus there was no guarantee of success. This played against the allocation of telescope time and brought about a sizeable pressure upon the protagonists of their search. But the search for low-mass stars was considered legitimate astronomy, and nobody could criticize the program of Mayor and Queloz if no planet was found in the end. ELODIE started the operations in April 1994 and it did not take long for this observational strategy to pay off. The following year, Mayor and Queloz bumped into a periodic Doppler oscillation in the spectrum of 51 Peg, a solar-type star. The amplitude of the oscillation was around 60 m/s, and a careful analysis suggested the existence of a planet with orbital period of 4.23 days, minimum mass 0.4–0.5 MJ , semimajor axis 0.05 AU, and eccentricity around 0 (Mayor and Queloz 1995; see Fig. 1). The discovery of 51 Peg b was published on the Nature issue of November 23, 1995, and constituted by all means a groundbreaking achievement, opening up a new era in modern astronomy. Moreover, the existence of a planet with such weird features overturned the current paradigm about the way a planetary system could be made.

14

D. Cenadelli and A. Bernagozzi

100

Vr (ms−1)

50

0

−50

−100 0

0.5

1

f

Fig. 1 The phased plot showing the oscillations in radial velocity of 51 Peg detected by Mayor and Queloz. The solid line is the orbital fit with a period of 4.23 days. (From Mayor and Queloz 1995, p. 357). Reprinted by permission from Springer Nature, © 1995

While Mayor and Queloz were undertaking their research, the other scholars were not keeping on the shelf. In those same years, Marcy and Butler, using an iodine cell for calibration, boosted the attainable precision to the astonishing value of 3 m/s. That was possible thanks to the identification of new possible Doppler error sources previously unknown, such as changes in the instrumental profile of the spectrometer every second. Fixing these errors was quite a time-consuming task, and data reduction was performed later, even years after data collection. According to Marcy: “We felt little urgency for two reasons. First, the orbital periods of Jupiters were surely at least 5 years (which was not true, of course). And all astronomers in the world thought we would never succeed in finding planets anyway. There’s no rush, when the prospect is simply failure” (Cenadelli and Bernagozzi 2015, p. 538). When Mayor and Queloz announced the discovery of 51 Peg b, and it became clear that exoplanets can have very short orbital periods, Marcy and Butler sifted their already taken data and found planets in them: in 1996, they announced the discovery of 70 Vir b (Marcy and Butler 1996; see Fig. 2) and 47 UMa b (Butler and Marcy 1996; see Fig. 3), orbiting two solar-type stars, like 51 Peg. 70 Vir b has strange characteristics as compared to the planets of the Solar System: it is massive and close to the star, with a minimum mass of 6.6 MJ , semimajor axis of 0.43 AU, orbital period of 116.7 days, and sizeable orbital eccentricity e D 0.40. On the other hand, 47 UMa, with a with a minimum mass of 2.39 MJ , a semimajor axis of 2.1 AU,

1 The Discovery of the First Exoplanets

15

Fig. 2 The phased plot showing the oscillations in radial velocity of 70 Vir detected by Marcy and Butler. Error bars are smaller than the data points. The solid line is the orbital fit with a period of 116.67 days. (From Marcy and Butler 1996, p. L149). © AAS. Reproduced with permission

Fig. 3 The plot showing the oscillations in radial velocity of 47 UMa (34 measurements over 8.7 years), detected by Butler and Marcy. The solid line is the orbital fit with a period of 1090 days. The residuals to the orbital fit (crosses) have an rms D 10.9 m/s. (From Butler and Marcy 1996, p. L155). © AAS. Reproduced with permission

an orbital period of 1090 days, and a low eccentricity e D 0.03, is more ordinary. All in all, the picture that was emerging was that a new class of planets existed, massive and close to their stars. They were named hot Jupiters. Moreover, the two American astronomers were able to confirm the existence of 51 Peg b (Marcy and Butler 1995). A further analysis on 51 Peg confirmed both the oscillation in the spectrum and the orbital data of the planet (Marcy et al. 1997).

51 Peg b Questioned and Finally Confirmed After these discoveries, there was some degree of resistance in admitting that such unexpected features as those of 51 Peg b (and the hot Jupiters at large) were real. This spurred astronomers to discuss about their reliability. As we saw, the alternative explanation of photospheric gas movements had already been put forth. In particular, the observed Doppler shifts could be triggered by stellar pulsations, although the periods involved were not typical of any known pulsation mode – supposed to

16

D. Cenadelli and A. Bernagozzi

last minutes or hours, not days. Mayor and Queloz had carefully analyzed that hypothesis and felt confident they could rule it out, but in 1997, David F. Gray of the University of Ontario took issue with the idea that 51 Peg b actually existed and rather advocated the stellar oscillation position (Gray 1997). Looking closely the spectra he owned of 51 Peg, he noticed that the profile of such lines as FeI6252.57 Å and VaI-6251.83 Å were tilted one way and then the other, with a period consistent with the 4.23 period of the supposed planet. To wit, these lines didn’t simply translate, but they were rather warped to an asymmetric profile. This phenomenon was confirmed by Gray by a check of the vanadium line at 6251.83 Å, and he concluded that 51 Peg was showing a pulsation, because no planet could warp the spectral line profile. Clearly, the stellar pulsation hypothesis brought about problems in its turn, because such long pulsations were of unprecedented nature. Moreover, no overtones or harmonics were detected, an unexpected feature according to current plasma physics theories. There was something puzzling in both cases: either it was necessary to modify the current planetary formation theories to account for planets with unexpected characteristics or it was necessary to change the theories of stellar structure to explain very long pulsations – if not general physics itself to explain the absence of harmonics. Nonetheless, the two explanations were not on the same ground from the epistemological point of view: from the one hand, stellar structure theories were a heavier corpus of knowledge to challenge, and from the other, planetary migration and the possible existence of hot Jupiters was not a thoroughly new idea, as we have already seen at the end of section “Improvements in Doppler Measurements from the 1950s to the 1970s: The Pioneers.” Moreover, their discovery spurred theoreticians to investigate further the mechanisms of planetary systems formation. In 1996, Lin, Bodenheimer, and Richardson (1996) pointed out that the actual orbital configuration of 51 Peg b was possible after a proper migration. This put the planetary hypothesis in a more favorable position. Not surprisingly, in the following years, this hypothesis finally overcame. Besides the aforementioned reasons, other observations bolstered it. First, Hatzes, Cochran, and Johns-Krull observed the spectrum of 51 Peg with a resolution higher than ever, and they could not see the warping of spectral lines observed by Gray (Hatzes et al. 1997). A further study with even higher spectral resolution confirmed that the supposed warping was not present. Gray took other spectra himself and came to the same conclusion. The supposed warping of the spectral lines had disappeared, and in the end, it seemed to be nothing but a casual distribution of errors. In the end, both Gray and Hatzes and coworkers acknowledged the existence of the planetary companion to 51 Peg (Gray 1998; Hatzes et al. 1998). Furthermore, other (supposed-to-be) planets were discovered spectroscopically in the meantime, and it was much more straightforward to explain the data with the idea of planets of different mass and orbital periods than with stellar oscillations of different periods. Doppler shift with periods lasting tens of days were discovered, and this is easily explainable with planets further from their stars, while it is difficult to imagine such long-lasting pulsations. Furthermore, the existence of multi-planetary systems was inferred from multiple periodicities discovered in one single stellar spectrum (the first multi-planetary system discovered around a main

1 The Discovery of the First Exoplanets

17

sequence star was the one of ¤ And, and a first planet was announced in 1997 (Butler et al. 1997), two others in 1999 (Butler et al. 1999), and a fourth one in 2011 (Curiel et al. 2011)). This was a pretty natural discovery: if we accept that exoplanets do actually exist, it is of no surprise that some stars are accompanied by several planets, as it happens in the Solar System. The announcement, in year 2000, of the discovery of transiting planets (Charbonneau et al. 2000; Gregory et al. 2000) was a sort of final confirmation: two different and independent methods gave results perfectly matching with each other, and the existence of exoplanets could not be doubted anymore. In any case, the doubts raised by Gray epitomized the difficulty the community of astronomers had in accepting that something unexpected like the hot Jupiters had been discovered. Moreover, in the past, there had been false alarms as regards the existence of exoplanets, so some degree of skepticism was natural. In 1991, for example, the discovery of a planet orbiting the pulsar PSR 1829-10 was announced (Bailes et al. 1991), but it was just an error in the data analysis software (Lyne and Bailes 1992; for this episode, see also Croswell 1997, pp. 132–60). To sum up, it was natural that such a groundbreaking discovery as the existence of other worlds underwent a careful inspection. In Marcy’s words, the “plethora of interpretation represents science at his healthiest. [ : : : ] It was right that the planet interpretation should not go unchallenged” (Marcy 1998, p. 127).

Concluding Remarks As we have seen, the characteristics of the first exoplanets were unexpected, because they challenged the common views about what a planetary system should be made like. The theoretical views about the formation of planets seemed to rule out the possibility of Jupiter-like planets lying near their parent stars. A few suggestions about planetary migration had already been put forth, but at the time, these voices did not have widespread recognition. Mayor and Queloz were searching for brown dwarfs, as well as extrasolar planets; hence they were more likely to discover planets with features – very closein orbits – that, at the time, were reckoned to be proper of stars. In other words, the fact that Mayor and Queloz placed themselves between two fields of research – planets and brown dwarfs – fostered their success. They were not looking too closely and did not immerse themselves completely in one particular field of research and so were less subject to the prejudices inherent to that particular field. When they bumped into the signal in the spectrum of 51 Peg, they were puzzled but very confident in the quality of their instrument and took a step ahead accepting the existence of something so unexpected. Marcy and Butler, in their turn, had chosen a proper observational strategy and attained very high precision. They were likely to discover something, and they actually did, finding out several planets in the following years. What played against them was the delay in data analysis. After the discovery of 51 Peg b, they found planets in the data they had already collected, but not analyzed yet.

18

D. Cenadelli and A. Bernagozzi

On balance, we can underline how all the main characters we spoke of came to major results. The discovery of the first exoplanets was a task undertaken by an entire community of astronomers. It gave birth to a new field of research that is impressively growing over time and attracting an increasing number of scholars. No more than two or three decades have elapsed since the discovery of the first exoplanets, and thousands of them have been found, a lot of investigations have been carried out, and new horizons are continually being opened. Most importantly, the pioneering work we spoke of answered the ancient question about the existence of other worlds in the cosmos, and today we may answer: yes, they do exist, the Galaxy teems with planets and most of them must still reveal their secrets.

Cross-References  Discovery of the First Transiting Planets  PSR B1257+12 and the First Confirmed Planets Beyond the Solar System

References Artymowicz P (1993) Disk-satellite interaction via density waves and the eccentricity evolution of bodies embedded in disks. ApJ 419:166–180. https://doi.org/10.1086/173470 Bailes M, Lyne AG, Shemar SL (1991) A planet orbiting the neutron star PSR 1829-10. Nature 352:311–313. https://doi.org/10.1038/352311a0 Butler RP, Marcy GW (1996) A planet orbiting 47 UMa. ApJ Lett 464:L153–L156. https://doi.org/10.1086/310102 Butler RP, Marcy GW, Williams E, Hauser H, Shirts P (1997) Three new “51 Pegasi-type” planets. ApJ 474:L115–L118. https://doi.org/10.1086/310444 Butler RP, Marcy GW, Fischer DA et al (1999) Evidence for multiple companions to ¤ Andromedae. ApJ 526:916–927. https://doi.org/10.1086/308035 Campbell B, Walker GAH (1979) Precision radial velocities with an absorption cell. PASP 91:540– 545. https://doi.org/10.1086/130535 Campbell B, Walker GAH, Yang S (1988) A search for substellar companions to solar-type stars. ApJ 331:902–921. https://doi.org/10.1086/166608 Cenadelli D, Bernagozzi A (2015) Youth plus experience: the discovery of 51 Pegasi b. Eur Phys J H 40:527–552. https://doi.org/10.1140/epjh/e2015-60041-5 Charbonneau D, Brown TM, Latham DW, Mayor M (2000) Detection of planetary transits across a sun-like star. ApJ 529:L45–L48. https://doi.org/10.1086/312457 Croswell K (1997) Planet quest: the epic discovery of alien solar systems. Oxford University Press, Oxford Curiel S, Cantó J, Georgiev L, Chávez CE, Poveda A (2011) A fourth planet orbiting ¤ Andromedae. A&A 525(A):78. https://doi.org/10.1051/0004-6361/201015693 Duquennoy A, Mayor M (1991) Multiplicity among solar type stars in the solar neighbourhood. II distribution of the orbital elements in an unbiased sample. A&A 248:485–524 Duquennoy A, Mayor M, Halbwachs JL (1991) Multiplicity among solar type stars in the solar neighbourhood. I. CORAVEL radial velocity observations of 291 stars. A&A Suppl Ser 88: 281–324 Fellget PB (1953) A proposal for a radial velocity photometer. Opt Acta 2:9–15 Goldreich P, Tremaine S (1980) Disk-satellite interactions. ApJ 241:425–441. https://doi.org/10.1086/158356

1 The Discovery of the First Exoplanets

19

Gray D (1997) Absence of a planetary signature in the spectra of the star 51 Pegasi. Nature 385:795–796. https://doi.org/10.1038/385795a0 Gray D (1998) A planetary companion for 51 Pegasi implied by absence of pulsations in the stellar spectra. Nature 391:153–154. https://doi.org/10.1038/34365 Gregory HW, Marcy GW, Butler RP, Vogt SS (2000) A transiting “51 Pegasi-like” planet. ApJ 529:L41–L44. https://doi.org/10.1086/312458 Griffin RF (1967) A photoelectric radial-velocity spectrometer. ApJ 148:465–476. https://doi.org/10.1086/149168 Griffin RF, Griffin RE (1973) On the possibility of determining stellar radial velocities to 0.01 km s-1. MNRAS 162:243–253 Hatzes AP, Cochran WD, Johns-Krull CM (1997) Testing the planet hypothesis: a search for variability in the spectral-line shapes of 51 Pegasi. ApJ 478:374–380. https://doi.org/10.1086/303770 Hatzes AP, Cochran WD, Bakker EG (1998) Further evidence for the planet around 51 Pegasi. Nature 391:154–156. https://doi.org/10.1038/34369 Hatzes AP, Cochran WD, Endl M et al (2003) A planetary companion to ” Cephei a. ApJ 599:1383–1394. https://doi.org/10.1086/379281 Joergens V (ed) (2014) 50 years of brown dwarfs – from prediction to discovery to forefront of research, Astrophysics and space science library, vol 401. Springer, Cham Latham DW (2012) The unseen companion of HD 114762. New Astron. Rev. 56:16–18. https://doi.org/10.1016/j.newar.2011.06.003 Latham DW, Mazeh T, Stefanik RP, Mayor M, Burki G (1989) The unseen companion of HD114762: a probable brown dwarf. Nature 339:38–40. https://doi.org/10.1038/339038a0 Lin DNC, Papaloizou J (1986) On the tidal interaction between protoplanets and the protoplanetary disk. III – orbital migration of protoplanets. ApJ 309:846–857. https://doi.org/10.1086/164653 Lin CC, Shu FH (1964) On the spiral structure of disk galaxies. ApJ 140:646–655. https://doi.org/10.1086/147955 Lin DNC, Bodenheimer P, Richardson DC (1996) Orbital migration of the planetary companion of 51 Pegasi to its present location. Nature 380:606–607. https://doi.org/10.1038/380606a0 Luhman KL (2013) Discovery of a binary Brown dwarf at 2 pc from the sun. ApJ lett 767: article id. L1, 6 pp. https://doi.org/10.1088/2041-8205/767/1/L1 Lyne AG, Bailes M (1992) No planet orbiting PSR 1829-10. Nature 355:213. https://doi.org/10.1038/355213b0 Marcy GW (1998) Extrasolar planets: back in focus. Nature 401:127. https://doi.org/10.1038/34298 Marcy GW, Butler RP (1995) The planet around 51 Pegasi. Bull Astron Am Soc 27: 1379 Marcy GW, Butler RP (1996) A planetary companion to 70 Vir. ApJ Lett 464:L147–L151. https://doi.org/10.1086/310096 Marcy GW, Butler RP, Williams E et al (1997) The planet around 51 Pegasi. ApJ 481:926–935. https://doi.org/10.1086/304088 Mayor M, Frei PY (2003) New worlds in the cosmos. The discovery of exoplanets. Cambridge University Press, Cambridge Mayor M, Queloz D (1995) A Jupiter-mass companion to a solar-type star. Nature 378:355–359. https://doi.org/10.1038/378355a0 Mayor M, Lovis C, Santos NC (2014) Doppler spectroscopy as a path to the detection of earth-like planets. Nature 513:328–335. https://doi.org/10.1038/nature13780 Serkowski K (1978) Possibilities of improving the accuracy of stellar radial velocities. In: Hack M (ed) Proceedings of the 4th international colloquium on astrophysics, high resolution spectrometry, Trieste, 3–7 July 1978, pp 245–267 Serkowski K, Frecker JE, Heacox WD, Roland EH (1979a) Fabry-Perot radial velocity spectrometer. In: Proceedings of the seminar on instrumentation in astronomy III, Tucson, 29 Jan–1 Feb 1979. Society of Photo-optical Instrumentation Engineers, Bellingham, pp 130–133. https://doi.org/10.1117/12.957075

20

D. Cenadelli and A. Bernagozzi

Serkowski K, Frecker JE, Heacox WD, KenKnight CE, Roland EH (1979b) Retrograde rotation of the stratosphere of Venus measured with a Fabry-Perot radial velocity spectrometer. ApJ 228:630–634. https://doi.org/10.1086/156888 Walker GAH (2012) The first high-precision radial velocity search for extra-solar planets. New Astron. Rev. 56:9–15. https://doi.org/10.1016/j.newar.2011.06.001 Walker GAH, Bohlender DA, Walker AR et al (1992) ” Cephei: rotation or planetary companion? ApJ Letters 396:L91–L94. https://doi.org/10.1086/186524 Wolszczan A (1994) Confirmation of earth-mass planets orbiting the millisecond pulsar PSR 1257C12. Science 264:538–342. https://doi.org/10.1126/science.264.5158.538 Wolszczan A, Frail DA (1992) A planetary system around the millisecond pulsar PSR 1257C12. Nature 355:145–147. https://doi.org/10.1038/355145a0

PSR B1257+12 and the First Confirmed Planets Beyond the Solar System Alexander Wolszczan

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Timing the Pulsar with Companions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Timing the Pulses from PSR B1257+12 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Confirmation of the PSR B1257+12 Planets and a Third Planet in the System . . . . . . . . . . . . . Discussion and Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

22 23 25 26 28 32

Abstract

We describe the survey and timing observations conducted in early 1990 with the Arecibo radio telescope, which have led to the discovery of the first confirmed extrasolar planetary system consisting of three low-mass planets, orbiting the 6.2-ms millisecond pulsar, PSR B1257+12. The existence of planets around a neutron star has carried with it a now fulfilled promise that planets should be common around the various types of stars. Furthermore, the architecture of the PSR B1257+12 system offered an early preview of the future discoveries of the very common compact systems of superEarth-mass planets by the Kepler telescope and ground-based radial velocity surveys.

A. Wolszczan () Department of Astronomy and Astrophysics, and Center for Exoplanets and Habitable Worlds, The Pennsylvania State University, University Park, PA, USA e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_131

21

2

22

A. Wolszczan

Introduction The discovery of the first confirmed extrasolar planets was intimately related to the events that took place at the Arecibo Observatory in early 1990. When a routine inspection of the 305-m Arecibo radio telescope revealed cracks in its support structure, caused by material fatigue, the obvious decision was made to replace the faulty elements. This would take at least several weeks, and inevitably, normal observations would have to be suspended during that time. Consequently, the telescope, immobilized for repairs, was not able to track radio sources, but it could be used as a transit instrument by parking it at a desired azimuth and elevation and scanning all the objects passing across the beam. At 430 MHz, which was a typical Arecibo pulsar survey frequency at the time, the beam size is about 10 arcmin, which translates to up to about 40 s integration time depending on source declination. Also, because of a large collecting area of the telescope, one can still count on a very good 1 mJy sensitivity while using it for pulsar searching. In early 1990, only four millisecond pulsars (MSPs) had been known, all of them located close to the galactic plane. However, given their most plausible origin from the spin-up of slowly rotating neutron stars in the process of accretion of matter and angular momentum from stellar binary companions and the kinematics of supernova explosions, it was quite plausible that MSPs should be isotropically distributed over the sky (Kulkarni and Narayan 1988). Taking advantage of the availability of large amounts of the observing time on the Arecibo telescope, projected to remain under repair for several weeks, we had proposed to test this idea by using the instrument in the transit mode to search for MSPs away of the galactic plane. The survey took place in January and February of 1990, and upon completion, it had covered 150 square degrees of the sky. Observations were made at 430 MHz, and the received signal was fed into the Arecibo correlation spectrometer used as a 128-channel, 10 MHz bandwidth filterbank, sampled at 506.625 s intervals. This observing setup had generated 65,536 data points per frequency channel, collected over the 34-s integration time for each sky position. The data were recorded on the magnetic tapes and subsequently processed with a 6-node, IBM 3090 computer of the Cornell Theory Center. The data analysis was completed by June 1990, and it had produced two pulsar candidates with the respective periods of 37.9 and 6.2 ms, both located at high galactic latitudes (Wolszczan 1990). Confirmation observations further revealed that the first of them, PSR B1534+12 ,was a 10h, neutron star binary, which has later become one of the best pulsar probes of relativistic gravity (Wolszczan 1991; Fonseca et al. 2014). Both PSR B1534+12 and the second one, PSR B1257+12, the fifth MSP discovered thus far, offered an immediate confirmation of the idea that the old, recycled neutron stars should be detectable all over the sky. This has initiated very successful all-sky MSP surveys, many of which have been continuing to this day, using the modern hardware and data analysis algorithms (Manchester et al. 2001). PSR B1257+12 was discovered on February 9, 1990 and confirmed in June of the same year. The confirmation data are shown in Fig. 1. In the rest of this chapter,

2 PSR B1257+12 and the First Confirmed Planets Beyond the Solar System

23

Fig. 1 A notebook scan of the confirming observation of the PSR B1257+12 position with the Arecibo radio telescope operating at 430 MHz. (Left) The continuously sampled data folded modulo the discovery period of 6.2 ms into 32 phase bins in each of the 128 frequency channels. Both the interstellar dispersion induced delay of the signal and its smearing effect on the pulse profile summed over all channels (upper panel) are clearly visible. (Right) The folded data with the dispersion delay removed. Without dispersion smearing, the sum of all channels displayed in the upper panel reveals the pulse profile of the pulsar. Pulse intensity variations across the band are due to the interstellar scintillation of the pulsar signal (Rickett 1990). (Reprinted from New Astronomy Reviews, Vol 56, A. Wolszczan, Discovery of pulsar planets, Pages 2-8, Copyright (2012), with permission from Elsevier)

we will describe further observations and timing analysis of the pulsar that had eventually led to the discovery of what has become the first confirmed planetary system beyond the Sun (Wolszczan and Frail 1992). For background information on the MSPs, pulsar timing, and the current status of the neutron star planet research, the reader is referred to Lorimer and Kramer (2005), Kramer (2011), Wolszczan (2012), and Manchester (2017) .

Timing the Pulsar with Companions A reflex motion of the pulsar in orbit with a binary stellar companion or planets results in periodic variations of the Römer delay of pulse arrival times. For a single Keplerian orbit, the Römer delay, tR , is a function of the five standard orbital parameters, the orbital period, Pb ; the eccentricity, e; the longitude of periastron, !; the semi-major axis, a1 sin(i ), where i is the orbital inclination; and the time of periastron passage, T0 , as given by Blandford and Teukolsky (1976): tR D x .cos E  e/ sin ! C x sin E

p

1  e 2 cos !;

(1)

24

A. Wolszczan

where x D a1 sin(i )/c, c is the speed of light and E, the eccentric anomaly, is related to the mean anomaly, M D .2=Pb /.t  T0 /, through the equation: E  e sin E D M;

(2)

where t is the epoch of observation. The process of modeling time-of-arrival (TOA) variations of the pulsar pulses involves fitting to data one or more orbits defined by Eq. 1, along with the astrometric and the interstellar propagation parameters, as described in detail by Kramer (2011). To illustrate the precision of pulsar timing, one can use the mass function to express the maximum time delay, tmax , for a circular, edge-on orbit as: tmax

 a  m  M 1 p   1:5 ms ; 1 AU m˚ Mˇ

(3)

where a is the orbital radius in astronomical units; mp is the planet mass in units of Earth mass, m˚ ; and M is the mass of the star in solar units. Assuming M D 1:4Mˇ , which is the neutron star mass typically used in such estimations, one can see that an Earth-mass planet, 1 AU away from the pulsar, would generate tmax 1 ms. The respective delays for a Moon-mass and Ceres-mass bodies in such orbits amount to 10 and 0.2 s. Given a superb precision of the pulsar timing measurements, which is routinely better than 1 s for many MSPs, reaching as far down as 100 ns for some objects (e.g., Hobbs et al. 2012), it is clear that this method of exoplanet discovery should easily detect terrestrial and Moon-mass planets, and it has a capability to reach all the way to asteroid-mass bodies. It is instructive to compare the pulsar timing precision to that of the Doppler velocity method, which, together with transit photometry, remains one of the two most prolific planet detection techniques. To do this, we derive an equation similar to Eq. 3 for the radial velocity variation, Vr : Vr  0:1 m s

1

 a  12  M  12  m  p  ; 1 AU Mˇ m˚

(4)

with the observed Doppler variation of the pulsar period obviously related to Vr through: P D Pı

Vr ; c

(5)

where Pı is the intrinsic period of the pulsar. For an Earth analog, orbiting a solarmass star, one clearly needs a not yet attainable Doppler velocity precision better than 10 cm s1 in the visible part of the spectrum. For comparison, measurements of Doppler shifts of periods of typical MSPs easily achieve an equivalent precision of a few mm s1 , which becomes even better in the case of pulse timing, because of its phase-coherent nature.

2 PSR B1257+12 and the First Confirmed Planets Beyond the Solar System

25

Timing the Pulses from PSR B1257+12 Timing measurements of PSR B1257+12 with the Arecibo radio telescope began shortly after its confirmation, in June 1990. The modeling of these observations originally assumed that this new only fifth MSP, known at that time, would be most likely in orbit with a white dwarf, with the orbital period anywhere between a few days and a few months. This assumption was based on a contemporary theoretical understanding of the formation and evolution of the MSPs and on the available evidence (e.g., Alpar et al. 1982). As illustrated in Fig. 2, the practically implemented observing scheme over the first 8 months was to make a few TOA measurements per month, allowing for the occasional, longer breaks due to the telescope scheduling constraints. Our early attempts to model the TOA variability of PSR B1257+12 in terms of binary motion with a stellar companion, or as a solitary pulsar, clearly showed that such simple models were not able to account for the pulsar’s timing behavior. This fact became even more obvious after a campaign of high-cadence observations conducted later in May and June of 1991 to better constrain our modeling efforts. Most interestingly, the periodogram analysis of all the measurements had made us realize that we were quite possibly looking at a superposition of two or more periodic variations at the amplitudes not exceeding just a few milliseconds. If interpreted in terms of orbital motion, this would necessarily involve orbiting bodies of masses not much larger than that of the Earth (Eq. 3). One important factor that could affect this astonishing conclusion was a possible error in the sky position of the pulsar. Because the timing modeling involves transformation of the topocentric TOAs to the solar system barycenter, a position error manifests itself in the form of a sinusoidal variation of the TOAs with the amplitude of .1 AU =c/cosˇ  500 s cosˇ, where ˇ is the ecliptic latitude. In the case of PSR B 1257+12, ˇ 17ı , meaning that even a 2 arcsec error would make a position error-related residual reach one-half of the pulsar’s 6.2 ms rotation period. This was certainly a possibility, given the Arecibo telescope beamwidth of 10 arcmin at 430 MHz. In this case, one would lose the exact pulse count throughout the entire observing period, with the consequence that any timing model fitted to the data would have no physical meaning. We have solved this problem by making an independent, 0.1 mas precise interferometric measurement of the position of PSR B1257+12 with the Very Large Array (Heinke et al. 1996) and fixing it in the pulsar timing model. By the end of September 1991, we have accumulated enough data to make it possible to fit to the TOAs a phase-connected timing model including two Earth-mass planets on the 66.5- and 98.2-day, slightly eccentric orbits around the pulsar with the post-fit residual of less than 10 s. The actual TOA residuals before and after the fit are shown in Fig. 2, along with the fit of the same model to the measured Doppler variations of the pulsar period. To place this discovery in a context, it is important to mention that the idea that neutron stars could have planetary companions was not entirely new at the time of this detection. In the early 1970s, orbital motion was used to explain the timing variations of the Crab Nebula pulsar in terms of orbiting planets (Richards et al. 1970). Subsequently, Demia´nski and Prószy´nski (1979) proposed that timing varia-

26

A. Wolszczan

Fig. 2 The first 18 months of timing observations of PSR B1257+12 with the Arecibo radio telescope at 430 MHz. (a) Doppler variations of the pulsar period converted to the equivalent changes of the pulsar’s radial velocity with the superimposed best-fit, two-planet model (solid curve). (b) The best-fit TOA residuals for a standard pulsar timing model without planets. (c) As above, with the two terrestrial-mass planets included in the fit

tions observed in a relatively young pulsar, PSR B0329+54, could be caused by an orbiting planet. Unfortunately, these findings have not been confirmed and found an explanation in terms of the timing noise, which reflects rotation irregularities that are typical in young neutron stars (Lyne et al. 1995). Pulsar planets were also considered by theorists, who had put forward a possibility of planet formation out of the matter surrounding neutron stars at various stages of their evolution (Hills 1970; Michel 1970). Perhaps most dramatically, in July 1991, Bailes et al. (1991) announced the discovery of a 10M˚ planet in a 6-month orbit around another young pulsar, PSR B1829-10, which had to be subsequently retracted (Lyne and Bailes 1992).

Confirmation of the PSR B1257+12 Planets and a Third Planet in the System An indirect nature of the detection of the PSR B1257+12 planets left it quite naturally open to questioning and to offers of alternative explanations of the observed TOA variability involving the neutron star physics (Dolginov and Stepinski 1993;

2 PSR B1257+12 and the First Confirmed Planets Beyond the Solar System

27

Gil and Jessner 1993). Fortunately, as discussed by Rasio et al. (1992) and Malhotra et al. (1992), the fact that the orbital periods of the two planets are close to the 3:2 mean motion resonance (MMR) had offered a way to confirm the planets by detecting its signature in the timing data. The predicted, long-term behavior of the timing residuals induced by the MMR was described in detail by Malhotra (1993) and Peale (1993). Two additional years of timing monitoring of PSR B1257+12 proved to be sufficient to accumulate enough data to identify the 3:2 MMR pattern in the timing residuals. This detection has served as a convincing proof that the two periodicities in the TOAs of the pulsar are actually caused by orbital motion rather than by any other unrelated effect (Wolszczan 1994). As the time baseline of the data kept growing, it became possible to implement a timing model that included parametrization of the MMR-related perturbations in terms of planet masses and their mutual orbital inclination and to successfully fit it to the TOA variations (Konacki and Wolszczan 2003). This work has demonstrated that the two planets have nearly coplanar orbits, which, together with their periods being close to the 3:2 MMR, has provided a convincing evidence that the planets must have originated from a disk, very much like in the case of planets around normal stars. Perhaps the most appealing visualization of the MMR effect on the PSR B1257+12 TOAs is provided by a periodogram of the TOA residuals after fitting out all the model parameters other than those related to the MMR perturbations (Fig. 3). In 1993, the continuing, routine periodogram analyses of the growing set of TOA measurements of PSR B1257+12 had revealed another 25.2-day periodicity in the data, which, if interpreted in terms of an orbiting planet, indicated a Moon-mass body in a 0.19 AU orbit around the pulsar (Wolszczan 1994). Because the timing precision achievable at that time was comparable to the ˙3 s amplitude of the signal, it had taken more than 2 years for the relatively low-cadence observations to cover enough orbital phase space for the signature of the third planet in the system to emerge from noise. Another sensitivity limitation was caused by the fact that, from Eqs. 3 and 4, in contrast to the Doppler velocity method, the pulsar timing becomes less sensitive to planet detection for smaller orbital radii. The timing residuals folded modulo the orbital period are shown in Fig. 4, along with the latest measurements, made in 2015 with the upgraded data acquisition hardware. Clearly, a transition from the 3.5 s precision achievable in the 1990s with the Penn State Pulsar Machine (PSPM) to that of 0.7 s possible with the Puerto Rican Ultimate Pulsar Processing Instrument (PUPPI) makes a dramatic difference and reveals that the orbit of the inner planet is not circular and has a significant 0.3 eccentricity (Rivera et al. 2016). The fact that the orbital period of the inner planet is close to the solar rotation period had inspired (Scherer et al. 1997) to propose that, rather than being of a planetary origin, the observed signal could be a dispersive propagation effect generated by the interplanetary plasma as it corotates with the Sun. We have shown that the planetary interpretation of this periodicity is much more plausible, by demonstrating that the amplitude of the periodic signal is the same at 430 and 1130 MHz (Wolszczan et al. 2000). Because a dispersion delay of the radio signal in the cold, tenuous plasma scales as  2 , where  is the observing frequency, it is

28

A. Wolszczan

Fig. 3 Periodograms of the real and the simulated TOA residuals of PSR B1257+12 after fitting out a complete timing model except for the parameters characterizing the MMR between the two superEarth-mass planets. The simulated TOAs are noiseless, sampled once a day over the same 10year period as the real data. The periodograms show MMR-related double peaks around the planets orbital frequencies, fc and fd with the resonance frequency, fR D fc  fd  5  103 day1 . Non-zero spectral power around the second harmonic of the orbital frequency of the outermost planet is also visible in the simulated data, because of the orbit being slightly eccentric (see Konacki et al. 1999 for more details)

evident that the amplitude of the 25-day periodicity at 1130 MHz should be almost an order of magnitude lower than at 430 MHz, which was clearly not observed.

Discussion and Conclusions Systematic timing measurements of PSB B1257+12 with the previous generation backend hardware (PSPM) had ended in 2008, when it became clear that obtaining more information about the pulsar’s planetary system at the available 3.5 s precision was unlikely. The timing model, based on observations made from 1998 to early 2009, which includes the rotational, astrometric, and propagation parameters, as well as the three planets, and the MMR-related perturbations between the two superEarth-mass outer ones, is shown in Table 1 and illustrated in Fig. 5. The architecture of this system and the masses of the planets are fascinatingly

2 PSR B1257+12 and the First Confirmed Planets Beyond the Solar System

29

Fig. 4 A comparison of the recent PUPPI timing observations of PSR B1257+12 (blue dots) with the similar measurements conducted in 1999 using the PSPM pulsar backend (red dots). (Top) Timing residuals from a fit of the model including the two outer planets with the inner one left out. The solid curve traces the simulated, best-fit TOA residuals due to the inner planet in an eccentric, e0.3, orbit. (Bottom) Post-fit residuals for a fit of the model including the complete, three-planet system

reminiscent of the compact, superEarth systems so commonly discovered by the Kepler telescope (Batalha 2014), and as stated earlier, it can be treated as an early preview of these discoveries, which became available already in 1992! At present, five other planet-mass bodies are known to orbit MSPs. All of them have very different evolutionary histories from that of the PSR B1257+12 system, which must have originated from a protoplanetary disk, very much like planets around normal stars (Konacki and Wolszczan 2003). One of them is the result of a multi-body interaction in a globular cluster (Sigurdsson et al. 2003), while the other four are most likely the original stellar companions ablated down to planetary masses by the pulsar energy flux (Bailes et al. 2011; Pletsch et al. 2012; Stovall et al. 2014; Spiewak et al.(2017). So far, no planets have been found around young

30

A. Wolszczan

Table 1 Observed and derived parameters of the PSR B1257+12 planets Parameter Projected semi-major axis, x (ms) Eccentricity, e . . . . . . . . Epoch of pericenter, Tp (MJD) . Orbital period, Pb (d) . . . . . Longitude of pericenter, ! (deg) Mass (M˚ ) . . . . . . . . . . Inclination, solution 1, i (deg) . Inclination, solution 2, i (deg) . Planet semi-major axis, ap (AU)

Planet a 0.0030(1) 0.0 (assumed) 49765.1(2) 25.262(3) 0.0 0.020(2) ... ... 0.19

Planet b 1.3106(1) 0.0186(2) 49768.1(1) 66.5419(1) 250.4(6) 4.3(2) 53(4) 127(4) 0.36

Planet c 1.4134(2)a 0.0252(2) 49766.5(1) 98.2114(2) 108.3(5) 3.9(2) 47(3) 133(3) 0.46

a Figures in parentheses are the 1  errors in the last digits quoted. (Reprinted from New Astronomy Reviews, Vol 56, A. Wolszczan, Discovery of pulsar planets, Pages 2-8, Copyright (2012), with permission from Elsevier)

Fig. 5 The best-fit residuals for the timing model of PSR B1257+12 including the standard pulsar parameters, the three planets, and the perturbations between planets b and c. From top to bottom: residuals at 430 MHz (red dots), residuals at 1400 MHz (blue dots), and the best-fit to dual frequency data (red and blue dots) with the long-term dispersion measure variations (bottom panel) corrected for. (Reprinted from New Astronomy Reviews, Vol 56, A. Wolszczan, Discovery of pulsar planets, Pages 2-8, Copyright (2012), with permission from Elsevier)

2 PSR B1257+12 and the First Confirmed Planets Beyond the Solar System

31

Fig. 6 The discovery space for planets circling a 1.4 Mˇ neutron star. Dashed lines mark the minimum detectable mass as a function of orbital period, for different values of timing precision. Red lines are for a typical range of timing precision for different monitoring programs. The upward turn of these lines around the 1-year period is due to loss of sensitivity, because of a covariance of long orbital period signal with the rotational and astrometric model parameters. The green rectangle approximately delimits the previously unexplored low-mass, short orbital period part of the parameter space

pulsars (Kerr et al. 2016). An important question that remains open is why the neutron star planet systems that form from disks appear to be so rare. Soon after the discovery of the PSR B1257+12 planets, it was generally agreed that they must have been created out of the material provided either by a supernova fallback (Menou et al. 2001) or by some form of destruction of the pulsar’s binary stellar companion. The fact that PSR1257+12 is a solitary MSP suggests that the material from a now nonexistent binary companion could contribute the mass needed to create planets around it (e.g., Podsiadlowski (1993) and references therein; Martin et al. (2016); Margalit and Metzger (2017)). However, it could also be an extreme member of the young pulsar population (Miller and Hamilton 2001) and have planets formed from a supernova fallback event (Hansen et al. 2009). Clearly, more neutron star planet detections are needed to remove the existing ambiguities in our understanding of their formation. One experimentally unexplored possibility is that neutron star planetary systems may typically consist of low-mass bodies in tight orbits around their parent stars, similar to the PSR B1257+12 planets. In fact, it is interesting to speculate that these planets represent a “tip of an iceberg” of a population of low-mass, compact systems, possibly including potentially observable asteroid belts (Shannon et al. 2013). If this is the case, one can argue that the typical long-term, low-cadence timing observations of the MSPs are not optimal for the detection of such systems, because, as discussed above in the context of the inner PSR B1257+12 planet, they cover the orbital phase space very slowly, and their sensitivity degrades with a

32

A. Wolszczan

decreasing orbital size. This is illustrated in Fig. 6, which shows the planet discovery space for hypothetical observations covering a timing precision range from 1 s to 1 ms. Arguably, a logical choice of strategy to explore the low-mass, tight-orbit corner of the parameter space is to conduct high-cadence timing observations of a set of isolated MSPs. Of course, Arecibo currently remains the best choice for such a project, but its limited declination range makes the Green Bank and the Effelsberg 100-m telescopes very attractive choices as well. The need for a large aperture is obviously dictated by the fact that the sensitivity to planet detection of the timing method decreases with the decreasing orbital radius. As it is highly unlikely that PSR B1257+12 is a lone galactic MSP that has disk-formed planets around it, the suggested observations and, possibly, a thorough reanalysis of the existing database of high precision timing measurements of other MSPs (e.g., the NANOGrav data, Arzoumanian et al. 2016) should help resolving the intriguing puzzle of the origin of such systems.

Cross-References  Pulsar Timing as an Exoplanet Discovery Method  Radial Velocities as an Exoplanet Discovery Method Acknowledgements The Center for Exoplanets and Habitable Worlds is supported by the Pennsylvania State University and the Eberly College of Science. The Arecibo Observatory is operated by the SRI International under a cooperative agreement with the National Science Foundation (AST-1100968) and in alliance with Ana G. Méndez-Universidad Metropolitana and the Universities Space Research Association.

References Alpar MA, Cheng AF, Ruderman MA et al (1982) A new class of radio pulsars. Nature 300:728 Arzoumanian Z, Brazier A, Burke-Spolaor S et al (2016) The NANOGrav nine-year data set: limits on the isotropic stochastic gravitational wave background. ApJ 821:1 Bailes M, Lyne AG, Shemar SL (1991) A planet orbiting the neutron star PSR1829-10. Nature 352:311 Bailes M, Bates SD, Bhalerao V et al (2011) Transformation of a star into a planet in a millisecond pulsar binary. Science 333:1717 Batalha NM (2014) Exploring exoplanet populations with NASAs Kepler Mission. PNAS 111:12647 Blandford R, Teukolsky SA (1976) Arrival-time analysis for a pulsar in a binary system. ApJ 205:580 Demia´nski M, Prószy´nski M (1979) Does PSR0329+54 have companions? Nature 282:383 Dolginov AZ, Stepinski TF (1993) On quasiperiodic variations of pulsars’ periods – an alternative to the planetary interpretation of PSR1257+12. In: Phillips JA, Thorsett SE, Kulkarni SR (eds) Planets around pulsars. ASP conference series, vol 36. ASP, San Francisco, p 61 Fonseca E, Stairs IH, Thorsett SE (2014) A comprehensive study of relativistic gravity using PSR B1534+12. ApJ 787:82

2 PSR B1257+12 and the First Confirmed Planets Beyond the Solar System

33

Gil JA, Jessner A (1993) Are there really planets around PSR 1257+12? In: Phillips JA, Thorsett SE, Kulkarni SR (eds) Planets around pulsars. ASP conference series, vol 36. ASP, San Francisco, p 71 Hansen BMS, Shih HY, Currie T (2009) A test case of terrestrial planet assembly. ApJ 691:382 Heinke CO, Frail DA, Wolszczan A (1996) The position and proper motion of PSR B1257+12. BAAS 28:1368 Hills JG (1970) Planetary companions of pulsars. Nature 226:730 Hobbs G, Coles W, Manchester RN et al (2012) Development of a pulsar-based time-scale. MNRAS 427:2780 Kerr M, Hobbs G, Johnston S, Shannon RM (2016) Periodic modulation in pulse arrival times from young pulsars: a renewed case for neutron star precession. MNRAS 455:1845 Konacki M, Wolszczan A (2003) Masses and orbital inclinations of planets in the PSR B1257+12 system. ApJ 591:L147 Konacki M, Maciejewski A, Wolszczan A (1999) Resonance in PSR B1257+12 planetary system. ApJ 513:471 Kramer M (2011) Planets around pulsars. AIP conference series, Vol 1331. AIP Publishing, Melville, p 5 Kulkarni SR, Narayan R (1988) Birthrates of low-mass binary pulsars and low-mass X-ray binaries. ApJ 335:755 Lorimer DR, Kramer M (2005) Handbook of pulsar astronomy. Cambridge University Press, Cambridge Lyne AG, Bailes M (1992) No planet orbiting PSR1829-10. Nature 355:213 Lyne AG, Pritchard RS, Shemar SL (1995) Timing noise and glitches. JApA 16:179 Malhotra R (1993) Three-body effects in the PSR 1257+12 planetary system. ApJ 407:266 Malhotra R, Black D, Eck A et al (1992) Resonant orbital evolution in the putative planetary system of PSR1257 + 12. Nature 356:583 Manchester RN (2017) Millisecond pulsars, their evolution and applications. JApA 38:42 Manchester RN, Lyne AG, Camilo F et al (2001) The Parkes multi-beam pulsar survey – I. Observing and data analysis systems, discovery and timing of 100 pulsars. MNRAS 328:17 Margalit B, Metzger BD (2017) Merger of a white dwarf-neutron star binary to 1029 carat diamonds: origin of the pulsar planets. MNRAS 465:2790 Martin RG, Livio M, Palaniswamy D (2016) Why are pulsar planets rare? ApJ 832:122 Menou, K, Perna R, Hernquist L (2001) Stability and evolution of supernova fallback disks. ApJ 559:1032 Michel FC (1970) Pulsar planetary systems. ApJ 159:L25 Miller MC, Hamilton DP (2001) Implications of the PSR 1257+12 planetary system for isolated millisecond pulsars. ApJ 550:863 Peale SJ (1993) On the verification of the planetary system around PSR 1257 + 12. AJ 105:1562 Pletsch HJ, Guillemot L, Fehrmann H (2012) Binary millisecond pulsar discovery via gamma-ray pulsations. Science 338:1314 Podsiadlowski P (1993) Planet formation scenarios. In: Phillips JA, Thorsett SE, Kulkarni SR (eds) Planets around pulsars. ASP conference series, vol 36. ASP, San Francisco, p 149 Rasio FA, Nicholson PD, Shapiro SL et al (1992) An observational test for the existence of a planetary system orbiting PSR1257 + 12. Nature 355:325 Richards DW, Pettengill GH, Counselman CC III et al (1970) Quasi-sinusoidal oscillation in arrival times of pulses from NP 0532. ApJ 160:L1 Rickett BJ (1990) Radio propagation through the turbulent interstellar plasma. ARA&A 28:561 Rivera R, Wolszczan A, Seymour A (2016) High-cadence timing observations of an exoplanetpulsar system, PSR B1257+12. BAAS 227:241.08 Scherer K, Fichtner H, Anderson JD et al (1997) A pulsar, the heliosphere, and pioneer 10: probable mimicking of a planet of PSR B1257+12 by solar rotation. Science 278:1919 Shannon RM, Cordes JM, Metcalfe TS et al (2013) An asteroid belt interpretation for the timing variations of the millisecond pulsar B1937+21. ApJ 766:5

34

A. Wolszczan

Sigurdsson S, Richer HB, Hansen BM et al (2003) A young white dwarf companion to pulsar b1620-26: evidence for early planet formation. Science 301:193 Spiewak R, Bailes M, Barr ED (2017) PSR J23222650 a low-luminosity millisecond pulsar with a planetary-mass companion. arXiv 1712.04445 Stovall K, Lynch RS, Ransom SM et al (2014) The green bank Northern celestial cap pulsar survey. I. survey description, data analysis, and initial results. ApJ 791:67 Wolszczan A (1990) PSR 1257+12 and PSR 1534+12. IAU Circ 5073:1 Wolszczan A (1991) A nearby 37.9-ms radio pulsar in a relativistic binary system. Nature 350:668 Wolszczan A (1994) Confirmation of Earth-Mass planets orbiting the millisecond pulsar PSR B1257+12. Science 264:538 Wolszczan A (2012) Discovery of pulsar planets. New Astron Rev 56:2 Wolszczan A, Frail DA (1992) A planetary system around the millisecond pulsar PSR1257+12. Nature 355:145 Wolszczan A, Hoffman IM, Konacki M et al (2000) A 25.3 day periodicity in the timing of the pulsar PSR B1257+12: a planet or a heliospheric propagation effect? ApJ 540:L41

3

Prehistory of Transit Searches Danielle Briot and Jean Schneider

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Predictory Studies of Various Methods for Exoplanet Detection . . . . . . . . . . . . . . . . . . . . . . . . Transits in the Solar System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fictional Transit: Solar Spots, Patches, or Planets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Transit of Mercury: The First Transit Really Observed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Some Information About Venus Transits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Research of Transits of Objects Which Actually Do Not Exist . . . . . . . . . . . . . . . . . . . . . . . Other Occultations or Transits in the Solar System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Early Predictions of Detection of Exoplanets by the Transit Method . . . . . . . . . . . . . . . . . . . . Some Hypothesis for Explaining Algol Variations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Predictive Announcements of Discoveries of Exoplanets by Transits . . . . . . . . . . . . . . . . . . Some Twentieth-Century Investigations Before HD 209458 b . . . . . . . . . . . . . . . . . . . . . . . . Back to the Future: “ Pic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Future Is Already Present: Transits and Extraterrestrial Civilizations . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

36 37 38 38 39 41 42 42 42 43 43 45 45 46 47 48

Abstract

Nowadays the more powerful method to detect extrasolar planets is the transit method, that is to say observations of the stellar luminosity regularly decreasing

D. Briot () GEPI, UMR 8111, Observatoire de Paris, 61 avenue de l’Observatoire, Paris, France e-mail: [email protected] J. Schneider LUTh, UMR 8102, Observatoire de Paris, 5 place Jules Janssen, F-92195 Meudon Cedex, France LUTH, Observatoire de Paris, PSL Research University, CNRS, Université Paris Diderot, Meudon, France e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_169

35

36

D. Briot and J. Schneider

when the planet is transiting the star. We review the planet transits which were anticipated and searched and the first ones which were observed all through history. Indeed transits of planets in front of their star were first investigated and studied in the Solar System, concerning the star Sun. The first observations of sunspots were sometimes mistaken for transits of unknown planets. The first scientific observation and study of a transit in the Solar System was the observation of Mercury transit by Pierre Gassendi in 1631. Because observations of Venus transits could give a way to determine the distance Sun-Earth, transits of Venus were overwhelmingly observed. Some objects which actually do not exist were searched by their hypothetical transits on the Sun, as some examples a Venus satellite and an infra-mercurial planet. We evoke the possible first use of the hypothesis of an exoplanet transit to explain some periodic variations of the luminosity of a star, namely, the star Algol, during the eighteenth century. Then we review the predictions of detection of exoplanets by their transits, those predictions being sometimes ancient and made by astronomers as well as popular science writers. However, these very interesting predictions were never published in peer-reviewed journals specialized in astronomical discoveries and results. A possible transit of the planet “ Pic b was observed in 1981. Shall we see another transit expected for the same planet during 2018? Nowadays, some studies of transits which are connected to hypothetical extraterrestrial civilizations are published in astronomical peer-reviewed journals. So we can note that the discovery of exoplanets is modiying in the research methods of astronomers. Some studies which would be classified not long ago as science fiction are now considered as scientific ones.

Introduction The discovery of the planet orbiting the star 51 Pegasi, now planet named 51 Peg b, has been a striking discovery for the whole astronomical community and even more for the mankind. Actually this discovery was expected and hoped for a very long time. Various methods exist for the detection of exoplanets and some of them were anticipated for a long time. A good synthetic presentation of all these methods is given by a figure of Michael Perryman (Perryman 2000). This figure is known as The Perryman tree because the display of the different methods is organized into a hierarchy. The first efficient method was the detection by observations of small periodic variations of radial velocities. During several years, it was the only successful method. Nowadays the most efficient method is the detection of transits of the planets in front of their star, implying a periodic decrease of the star’s luminosity. This method is essential for determining some physical parameters of the planet as period, diameter, mass, and chemical atmospheric composition. The present study is mostly dedicated to preliminary studies predicting, looking for, or

3 Prehistory of Transit Searches

37

observing transits. Before that, we briefly review some methods for planet detection and their precursory studies. The search for other worlds has been existing since antique ages. Nowadays this formulation is interpreted as the search for planets around other stars than the Sun. However, the meaning of other worlds has then been totally different than this representation. Two fundamental astronomical discoveries have been necessary so that the words search for other worlds correspond to the search for extrasolar planets. The first discovery is the heliocentric system established by Nicolaus Copernicus (1473–1543) in the book De revolutionibus orbium coelestium, published in 1543, that is to say the planets are orbiting the Sun, and so Earth is no more the center of the world (Copernicus 1543). The second discovery is the understanding that stars are other Suns. During the seventeenth century, many unsuccessful searches for determination of stellar distances have implied that stars are much more remote than it was supposed. Then stars are intrinsically very luminous, as is our Sun. Because stars are objects similar to the Sun, it is highly likely that they are surrounded by a planetary system. As soon as 1686, Bernard Le Bovier de Fontenelle (1657–1757) wrote in his book the Entretiens sur la pluralité des mondes, i.e., A conversation on the Plurality of Worlds: Every fixed star is a sun, which diffuses lights to its surrounding worlds (Fontenelle 1686). This book was a best seller; it was reedited many times and translated in many languages. Its influence throughout the occidental world was very important. So, as soon as the second part of the seventeenth century, the existence of extrasolar planets was considered. The discovery of these planets was made in 1995, more than three centuries later (Mayor and Queloz 1995). This paper is devoted to some studies in the past which searched for and predicted methods ahead of their time for the detection of possible planets around other stars than the Sun and then from the seventeenth century. After a rapid review of some precursory studies of various methods, we will focus especially on the studies predicting some planetary transits and establishing detection methods for them, first in the Solar System and then outside the Solar System.

Predictory Studies of Various Methods for Exoplanet Detection 1. Imaging – After centuries of philosophical speculations, the first scientific approach to the detection of exoplanets was due to Christiaan Huygens (1629– 1695) as soon as 1698, by imaging (Huygens 1698). In the book Kosmotheoros, Huygens at once admitted that no planet could be seen: “For let us fancy our selves placed at an equal distance from the Sun and fixed Stars; we would then perceive no difference between them. For, as for all the Planets that we know see attend the Sun, we should not have the least glimpse of them, either that their Light would be too weak to affect us, or that the Orbs in which they move would make up one lucid point with the Sun” (Huygens 1698). 2. Astrometry – For example, Kaj Aage Strand (1907–2000) wrote in 1943 about an unseen companion in the double star system 61 Cygni: “With a mass

38

D. Briot and J. Schneider

considerably smaller than the smallest known stellar mass, the dark companion must have an intrinsic luminosity so extremely low that we may consider it a planet rather than a star. Thus planetary motion has been found outside the Solar System : : : ” (Strand 1943). 3. Radial velocity – This method was predicted by Otto Struve (1897–1963) in 1952: “A planet ten times the mass of Jupiter would be easy to detect, since it would cause the observed radial velocity of the star to oscillate with ˙ 2 km s1 ” (Struve 1952). As we know, this method was very successful to detect the first exoplanet and many other ones. It was the only method efficiently used during several years. 4. Multiplanet perturbations – This method was successfully used in the Solar System to discover the Neptune planet by Urbain Le Verrier (1811–1877) in 1846 (Le Verrier 1846).

Transits in the Solar System The history of astronomy mentions several observations of supposed transits on the Sun anterior to the first observations with an optical instrument. The question is was it real transits or more probably sunspots?

Fictional Transit: Solar Spots, Patches, or Planets As soon as sunspots are observed, two hypotheses have been put forward to explain their origins: patches on the Sun and transit in front of the Sun of unknown inframercurial planets. In 1613, Galileo Galilei (1564–1642) announced that he discovered and observed some spots in front of the Sun. At once, a controversy appeared about some anterior observations of sunspots by Thomas Harriot (1560–1621), Christoph Scheiner (1575–1650), David Fabricius (1564–1617), and his son Johannes Fabricius (1586– 1615). Furthermore, historians have made inventories of sunspot observations long ago, in various civilizations, with naked eyes or with a camera obscura. An example of these inventories can be found in Vaquero (Vaquero 2007). Jean Tarde (1561–1636) was a canon in Sarlat, in the Perigord (southwest of France). He went to visit Galileo in 1614. He observed and studied sunspots during 4 years. His observations are probably the most or among the most extended period of observations of sunspots at this epoch. He carried out his observations with a scientific method. He interpreted the sunspots as small planets passing between the Mercury and the Sun. He named the planets that he supposed he observed Borbonia sidera, i.e., Bourbonian planets, from the dynasty name of the king of France, to honor Louis XIII, the king of France, as Galileo named Medicean planets the four Jupiter satellites that he discovered to honor the Medicis princes. He published a book in Latin Borbona sidera in 1620, translated in French in 1623, Les astres de Borbon (Tarde 1620). We have to emphasize that he used the Copernic system, i.e., the Earth orbiting the Sun, whereas he was a priest of the Catholic Church. He

3 Prehistory of Transit Searches

39

carried on his observations with a great perseverance. He noted that those planets move with different velocities and are moving slowly that of Mercury. The third law found by Johannes Kepler (1571–1630) establishing the relation between the period of a planet and its distance to the Sun was published only in 1619 (Kepler 1619), and probably Jean Tarde did not know it when he wrote his book published in 1620. As many scientists of this epoch, he used a religious argument to refute the theory of sunspots belonging to the Sun. He wrote that spots on the Sun are impossible because God chooses the Sun as place of residence: In sole posuit tabernaculum Suum.. The place chosen by God to stay could not be corrupted. Tarde quoted this sentence in Latin even in the French version of his book. He did not indicate the origin of it; that means that this sentence was being very known by every people. Actually this sentence is a part of the psalm 19, in the Bible version called the Vulgate, translated from the Hebrew to Latin by St Jerome at the end of the fourthcentury AD. This very popular version of the Bible was used by the Catholic Church during many centuries, and it was the first book ever printed by Gutenberg in 1455. However, this sentence, which is used by Tarde as a basis for his argumentation denying that the origin of dark patches seen on the Sun are sunspots, results from a mistake in the translation from Hebrew or in a copy of the original translation. The real meaning is something like In the heavens God has pitched a tent for the sun, as it can be seen in any other translation, in many various languages. However, the idea that the Sun is pure and cannot be corrupted nor soiled corresponds to the description of the world according to the Aristotle’s philosophy. Some more information about life and work of Jean Tarde can be found in Baumgartner (Baumgartner 1987).

Transit of Mercury: The First Transit Really Observed In 1627, Johannes Kepler (1571–1630) published Tabulae Rudolphinae, the Rudolphine Tables (Kepler 1627) so-called in honor of the former Emperor Rudolph II of Habsburg (1552–1612). These astronomical tables are based on the three laws concerning the planet motions published by Kepler in 1609 and 1618 and using the observations of Tycho Brahe (1552–1601). The discovery of logarithms by the Scot John Napier (1550–1617) in 1614 (Napier 1614) was greatly appreciated by Kepler and gave facilities for the making of the Tables. In 1630, he published ephemerides for the years 1629 to 1639, based on his Rudolphine Tables in which he included a Reminder for Astronomers and people studying celestial objects (Admonitio ad astronomos, rerumqve coelestium studiosos that we will name simply Admonitio) (Kepler 1629) indicating that it will be a transit of Venus in front of the Sun on the 6th of December 1631, visible from America, according to his calculations. However, as a little mistake could possibly exist in his predictions, he advice European astronomers to observe the Sun during this day. Moreover, on the 7th of November of the same year, it will happen a transit of Mercury in front of the Sun. However, because of difficult Mercury observations, this is no certainty in this date, and Kepler advised to European astronomers to observe from the 6th and continue observations up to the 8th of November. Admonitio is distributed through Europa to educated people.

40

D. Briot and J. Schneider

Using Kepler’s Admonitio, the French scientist Pierre Gassendi (1592–1655) very carefully prepared the observation of the Mercury transit. He was in Paris during the planned days. Kepler has advised to project image of the Sun on a paper with a refracting telescope or with a simple camera obscura in the absence of any telescope. A camera obscura can be equipped with an optical instrument, for example, a single lens or a refracting telescope, as well as a mirror to straighten the images obtained. Gassendi owned already an instrumentation that he used to observe sunspots and solar eclipses. In the camera obscura that he used, luminous rays coming from the Sun passed through a Galilean telescope and formed a Sun image on a sheet of paper. The adjustment was made as this image diameter would be around 25 cm. He draw a similar circle which he divided the diameter in 60 equal parts. In another room immediately below, an assistant using a two-foot quarter-circle instrument was to observe and note the height of the Sun when Gassendi indicated stamping his foot. So it is obvious that this observation has been prepared in a really scientific way. The weather was in part cloudy, and it was impossible to observe the Sun before the 7th of November. On this morning, Gassendi observed a very little black patch that he supposed in first to be a sunspot, because he was surprised by the smallness of this patch, and very soon he realized that for the first time, he observed a planet, that is, Mercury, in front of the Sun. Gassendi described in detail his observations in a book entitled Mercurius in sole visus et Venus invisa (Mercure visible on the Sun and Venus invisible), published in 1632 (Gassendi 1632). However, for several reasons, very few people could observe this first Mercury transit. One of the reasons was the rainy or cloudy weather in these days, which is very frequent in November in a great part of Europe. Another reason was the unexpected smallness of Mercury, so observers who used a camera obscura without any optical instrument could not see Mercury. The transit of Mercury was also observed by Johann Baptist Cysatus, the former pupil of Christoph Scheiner, in Innsbruck (Austria); by Johannes Remus Quietanus, physician and mathematician of the Emperor Mathias in Rouffach (Alsace); and by an anonymous Jesuit in Ingolstadt (Bavaria). We do not know circumstances of any of these observations, so Gassendi’s observation is the only one from which it is possible to deduce some astronomical conclusions and so the only one which can be considered as really scientific. A transit of Mercury was supposedly observed in other circumstances. In 1607, from Tycho Brahe’s observations, Kepler calculated that a transit of Mercury in front of the Sun will happen at the end of May. He observed the Sun on the 28th of May with a makeshift camera obscura, without any lenses. Actually, he detected a black spot on the Sun that he supposed to be Mercury. However, when sunspots have been discovered by observations carried on with optical instruments, Kepler understood then that he observed some sunspots. The observation in 1631 by Gassendi of the Mercury transit is of a real importance. It allowed astronomers to correct Kepler’s data about Mercury. The Mercury inclination on the ecliptic plane and the trajectory of Mercury became much more precise. A very important result was a new estimation of the Mercury diameter, much smaller than though up to this time. This last point

3 Prehistory of Transit Searches

41

implied that the planet diameter could not be deduced from the luminosity or from the telescope observations. Some information about the importance of the first observation of a planet transit can be found, for example, in Van Helden (Van Helden 1976).

Some Information About Venus Transits In his Admonitio, Kepler announced a transit of Venus in front of the Sun predicted for the 7th of December 1631 that is 1 month after the Mercury transit. However, this transit could not be observed from Europe, what explains the second part of the title of the book written by Gassendi: et Venus invisa. Kepler expected that the next Venus transit will happen in 1761, that is to say 120 years later. So the astronomers who were interested directed their studies to other subjects. However, in England, Jeremiah Horrocks (1618–1641) studied the Rudolphine Tables of Kepler, and he determined in October 1639 that the next transit will actually happen on the 4th of December 1639, according to the Gregorian calendar, that is, only 8 years after the transit of 1631. He notified his friend and correspondent William Crabtree (1610– 1644) in order that he would observe this phenomenon. They were the only first observers of a Venus transit. Nowadays, we know that the Venus transits happen four times during a cycle of 243 years, the intervals between the transits being 8 years, 105.5 years, 8 years, and 121.5 years. The Mercury transits are much more frequent because they happen 13 or 14 times during a century. After having observed a Mercury transit in 1677, Edmund Halley (1656–1742) showed that observations of the transits of the inferior planets, Mercury or Venus, from places of different latitudes on Earth, allow to determine the distance SunEarth and then all the distances in the Solar System. The Venus transits are more easy to observe than the Mercury transits. Using the Halley’s method, it would be necessary to precisely determine the times of the contacts between the planet Venus and the limit of the Sun surface, observed from different places on Earth. The following Venus transits happened in 1761 and 1769 and then in 1874 and 1882. To observe these Venus transits, many perilous expeditions were launched worldwide from more and more countries. The relations of these expeditions represent a very interesting part of the history of astronomy, often picturesque and sometimes tragic. Unfortunately, the results were not as good as hoped, because the observers faced the black drop phenomenon. They had to determine the precise times of the “second contact” and the “third” contact. The second contact corresponds to the moment when the surface of Venus appears completely on the Sun surface, the edge of Venus being tangential to the edge of the Sun. The third contact corresponds to the moment when the edge of Venus is tangential to the edge of the Sun, the surface of Venus totally appearing on the surface of the Sun, just before the last phase of the transit. However, the moment of the second contact was quite impossible to precisely determine because the black disc of Venus seemed to remain linked to the edge of the Sun by a dark “neck.” The surface of Venus did no more appear as

42

D. Briot and J. Schneider

circular but almost pear-shaped. The problem existed as well for the determination of the precise moment of the third contact. The accuracy of the time determinations of the second and the third contacts was hoped within about a second by Halley. Due to the black drop effect, the accuracy of timing became like a minute. When due to the instrumental progress, astronomers understood how to eliminate the black drop phenomenon; other more precise ways to determine the distances in the Solar System were discovered. Nowadays, the more accurate methods use some laser.

Research of Transits of Objects Which Actually Do Not Exist The history of searches for transits in the Solar System is not always successful. Numerous observations have been carried out to detect a satellite for the Venus planet as well as some infra-Mercurial planets. A list of Observations or supposed observations of the Transits of Infra-Mercurial Planets or other Bodies across the Sun’s Disk from 1761 to 1865 is displayed by Ledger (1879). The precise trajectory of the planet Mercury, and particularly the advance of its perihelion – the point on its orbit when Mercury is closest to the Sun – cannot been explained using only the Newtonian theory. As the French astronomer Le Verrier discovered in 1846 the planet Neptune by calculations from the trajectories of other planets, he tentatively explained the trajectory of Mercury by a hypothetical planet orbiting between Mercury and the Sun. This infra-mercurial hypothetical planet named Vulcain has been researched for a long time, by astronomers and amateur observers, and sometimes has been believed to be observed. The solution for this problem has been obtained only in 1915 when Albert Einstein discovered the theory of general relativity. Actually, the corrections to Newton’s theory due to the theory of general relativity explain the advance of the perihelion of Mercury. Obviously, Le Verrier could not know this theory.

Other Occultations or Transits in the Solar System Let us briefly recall the importance of scientific discoveries by other transits or occultations in the Solar System. The word occultation is used when the transiting object hides the whole transited object or a large part of it. The regular movements of the Galilean satellites of Jupiter, hidden or shadowed by the planet, allowed the discovery of the velocity of light during the seventeenth century.

Early Predictions of Detection of Exoplanets by the Transit Method Thanks to the Kepler space telescope, the detection of extrasolar planets by observing the periodical decreasing of the luminosity of the star due to the transit of the planets between the star and the observer is now the most efficient way to

3 Prehistory of Transit Searches

43

detect extrasolar planets. Before this method has been used successfully, this way of detection was announced in some very premonitory studies and sometimes a long time ago before the first transit observed.

Some Hypothesis for Explaining Algol Variations Algol is a regularly variable star. The first observations of these variations are generally attributed to Geminiano Montanari (1633–1687) from 1668 to 1677 and Giovanni Filippo Maraldi (1665–1729) around 1693 and 1694. However, these variations are probably known for a very long time because the origin of the name of the star is generally considered as Arabic meaning “The Demon.” This star, as well as many other variable stars, was intensively observed during the eighteenth century by two close friends and collaborators, Edward Pigott (1753–1825) and John Goodricke (1764–1786). Goodricke determined the period of variations as 2 days, 20 h, and 45 min. It is remarkable that this differs only a few minutes from the modern value (Goodricke 1783). These very careful observers made some assumptions about the cause of these regular variations. Goodricke wrote: “If it were not perharps too early to hazard even a conjecture on the cause of this variation, I should imagine it could hardly be accounted for otherwise than either by the interposition of a large body revolving round Algol, or some kind of motion of its own, whereby part of its body, covered with spots or such like matter, is periodically turned towards the earth. But the intention of this paper is to communicate facts, not conjectures : : : ” (Goodricke 1783). However, Michael Hoskin having had the opportunity to study the journal of Piggot attributes to him the hypothesis of a transit of a large body, planet or satellite (Hoskin 1979). Some years later, Piggot wrote “Hitherto the opinion of astronomers concerning the changes of Algol’s light seem to be very unsettled; at least none are universally adopted, though various are the hypotheses to account for it; such, as supposing the star of some other than a spherical form, or a large body revolving round it, or with several dark spots or small bright ones on its surface, also giving an inclination to its axis, &c. : : : ” (Piggot 1785). Let us notice that this argumentation was regarded interesting enough to be published again in a French journal Journal encyclopédique. This is possibly the first use, and at least one of the first ones, of the hypothesis of a planet transit to explain regular variations of a star light, although the existence of planets orbiting around stars was considered as plausible for more than a century (see, e.g., Fontenelle 1686). Nowadays, we know that Algol is a semidetached binary star with a mass transfer from one component to the other one, at present the less massive star being the more evolved. The name Algol is now used for all the stars having the same evolutionary path.

Predictive Announcements of Discoveries of Exoplanets by Transits The first known indication of a possible detection of exoplanets by transit was predicted by Dionysius Lardner (1793–1859) in a book of popular science. He

44

D. Briot and J. Schneider

studied the periodic variable stars and listed all the hypotheses proposed to explain the phenomena. The fifth hypothesis is “It has been suggested that the periodical obscuration or total disappearance of the star may arise from the transits of the star by its attendant planets” (Lardner 1853). Lardner was an Irish popularizer of science. David Belorizky (1901–1982) was an astronomer at the Marseille observatory. In 1938, he wrote a paper about variations of the Sun in L’Astronomie, a magazine for amateurs-astronomers (Belorizky 1938). In a paper entitled “Le Soleil, Etoile variable” (“The Sun, as variable star”), he studied the different ways for discovering other planetary systems. He explained that the variation of the radial velocity of the Sun due to the presence of Jupiter could not be detected by the spectrographs existing at this time. He completely rejected the possibility to observe a planet orbiting a star, neither by eye nor photographically considering the magnitude of a planet located at a stellar distance and considering also the very large ratio between the luminosity of the star and the luminosity of the planet. He studied the variation of the luminosity of the Sun due to some transits of Jupiter observed from another planetary system and wrote: “The only way that we see at the moment to possibly detect existence of planets in other worlds is the photometry with a precision of 1/100 magnitude, which is the precision of current photo-cells.” The life of David Belorizky shows some interesting coincidences. He was born in Russia and emigrated to France during the 1920s. To escape the Jew extermination during World War II and the Nazi occupation in France, he was protected and hidden as a Jew at the Haute Provence Observatory. It is remarkable that the first extrasolar planet 51 Peg b was discovered in this observatory. So there are two commemorative plaques in the Haute Provence Observatory, the first one in memory of the Jews hidden during the war in this observatory and the second one to celebrate the discovery of 51 Peg b, the first exoplanet. Gabriel Rémy (1945) wrote in a science popularization book: “Who knows if we will succeed in some days to detect changes of light emitted by some close stars when an invisible and dark object, like a planet, will periodically cross the field?” (Rémy 1945). Rémy was a priest and amateur-astronomer and interested also by microscopic science. Otto Struve (1897–1963) indicated, in 1952, in the same reference that quoted above: “ : : : the projected eclipse area is about 1/50th of that of the star, and the loss of light in stellar magnitude is about 0.02” (Struve 1952). Struve was born in Russia in a family containing several famous astronomers. He made all his careers in the United States where he was a very important astronomer. Among many other subjects of interest and studies, he was very interested by the research of the life in the Universe, during a time when only very few astronomers were interested by this research. These quoted papers or books clearly indicate that the research of planets orbiting other stars than the Sun was a subject. We emphasize that all these papers, even written by professional astronomers, are more often published in science popularization books or magazines and never published in the most famous peerreviewed astronomical journals.

3 Prehistory of Transit Searches

45

Let us finally note that the spectroscopy of transit as a tool to explore their atmosphere was proposed in 1992, well before the detection of 51 Peg b (Schneider 1994a). Also, the dynamical behavior of circumbinary transiting planets was predicted in 1994 (Schneider 1994b), well before its observational confirmation for Kepler-413(AB) b (Kostov et al. 2014).

Some Twentieth-Century Investigations Before HD 209458 b The first exoplanet transit was discovered for a planet already known from radial velocity detection (HD 209458 b). Before that discovery, some systematic searches for transits were made, with two approaches: (1) search for transits for a few suitably selected stars and (2) systematic search for transits on very large stellar samples. The first approach was proposed by Doyle et al. (1984) and Doyle (1985); the idea was to select stars for which the rotation axis, inferred from the relation i D arcsin (P* Vsini/2R* ) where P* and R* are the stellar rotation period and radius and Vsini is the mean observed velocity at the stellar surface, inferred from the spectral lines width. Assuming that the planet orbit is in the star equatorial plane, potential planets should have a large probability of transits. A special case of an a priori planet favorable orbital orientation was the selection of eclipsing binaries, with the assumption that the planet and binary orbits are coplanar. The latter approach was implemented for CM Dra with the transit of extrasolar planet (TEP) network (Deeg et al. 1997). It was the first systematic detection program of a planetary transit (and by the way, the first systematic search for circumbinary planets and planets around DM stars). The second approach was based on large stellar samples to compensate the low geometric probability R_*/a of transits (where a is the planet orbital radius). The first actual implementation was the FRESIP proposal (Borucki et al. 1996), which was finally launched as the Kepler mission. It was greatly facilitated with the development of CCDs in the 1980s, although it could in principle have been possible much before with a Lallemand electronic camera equipping a wide field telescope (Lallemand et al. 1970). And even, as suggested by Belorizky (1938), it could have detected (by great chance) exoplanet transits on some individual stars decades before the discovery of the transit of HD 209458 b.

Back to the Future: “ Pic The exoplanet “ Pic b was discovered by imaging in 2008 (Lagrange et al. 2009). Because of the presence of a circumstellar disk of gas and dust which is oriented edge-on to Earth, the star “ Pic was considered during several years before the detection of 51 Peg b, as a good candidate for the direct detection of the first exoplanet. So this star has been extensively studied. As early as 1993, some spectroscopic events are interpreted as the signature of the vaporization of cometlike bodies (Lecavelier des Etangs et al. 1993). Moreover, a detailed study of

46

D. Briot and J. Schneider

previous observations revealed that the star showed some light variations in 1981 which can be possibly interpreted as a planetary transit (Lecavelier des Etangs et al. 1995). Another observation of a similar variation is necessary to confirm the exoplanet transit. Studies of the planet “ Pic b indicate that a transit of this planet in front of its star is possible in 2017–2018. This transit prediction is actually depending on the orbital eccentricity of the planet. In case of a low-eccentricity orbit, the expected period is 18 years, and in case of a high-eccentricity orbit, the period is 36 years. We have to wait up to around 2018 to obtain an answer and to obtain a confirmation of a transit. The duration of the transit is estimated to approximatively 10 h. Several observational campaigns are dedicated to accurate ground observations of “ Pic. A nano-satellite PicSat was designed by astronomers of the Paris-Meudon observatory especially to detect and observe accurately a possible transit on “ Pic (see, e.g., Nowak et al. 2017). This satellite was launched on the 12th of January 2018. But the communication was unfortunately lost on March 20 for an unknown reason, so that it will not accomplish its goal. If the transit is confirmed in the near future, we could obtain better observations of the following transit in 2053, when the period will be known more accurately and when we could use very large and extraordinary outstanding future instruments. So, will Beta Pictoris b win the title of the first detected exoplanet?

The Future Is Already Present: Transits and Extraterrestrial Civilizations Nowadays, we can notice that the detection of extrasolar planets has led to some changes in astronomical publications. As we said above, the studies predicting the discovery of extrasolar planets from their transits in front of their star were more often published in books or magazines of popular science. The discovery of exoplanets and the expectation that these studies may allow to detect an extraterrestrial life in a more or less distant future has extended our area of scientific research. Some studies about planet transits, which would considered as pure science fiction until just recently, are now published in astronomical journals with peer review. We give now some examples. In 2005, Luc Arnold published a study about transit lightcurve signatures of artificial objects (Arnold 2005). These artificial objects would be built and put into orbit by some civilizations living on extrasolar planets and willing to make themselves known. These artificial objects would be designed in such a manner their transits are completely different of all natural planet transits as we know them. The cases of natural planets transits include single planets, or with moons or with rings. These artificial objects could be, for example, simple objects of unusual shapes, as triangles, or furthermore a fleet of objects. Obviously, that would imply a very high level of civilization for the extraterrestrial creatures imagining and building some such artificial planets. The Arnold’s paper has inspired other studies. As an example, in 2015, Korpela et al. studied how extraterrestrial civilizations would change the transits that we could observe, by illuminating the dark side of their planet (Korpela et al. 2015). In 2016, Kipping and Teachey considered the

3 Prehistory of Transit Searches

47

point of view of inhabitants of the planet Earth who would broadcast or cloak their existence to extraterrestrial civilizations (Kipping and Teachey 2016). This would be possible with laser emission. These last three papers were published in journals with peer review, i.e., Astrophysical Journal and Monthly Notices of the Royal Astronomical Society. Very recently the hypothesis of artificial transits or structures made by extraterrestrial civilizations has been considered to explain some observations of a star (KIC 8462852) still unexplained with purely physical processes. Let us notice that when the first pulsars have been discovered but not yet explained by astrophysical theories, they were called LGM1 and LGM2, as Little Green Man 1 and Little Green Man 2, but it was only something as a private joke, far from being published in a scientific journal. To be complete, we wish to recall that of course a scientific explanation was found to explain pulsar observations. We know now that pulsars are neutron stars, the end of the life of some massive stars.

Conclusion Observations and studies of planetary transits of planets, or other objects, in front of their star form a very fruitful part of astronomical research. Studies of transits in the Solar System began a long time ago, and their history is very interesting because it contains a lot of unexpected episodes. The first observation of a planet transit in front of the Sun, that is, the observation of the Mercury transit carried out scientifically by Pierre Gassendi in 1631, is an important step in the history of astronomy. Nowadays, more than half of the several thousands of extrasolar planets were discovered by the transit method, particularly thanks to the Kepler space telescope, we found several premonitory and visionary studies, published several decades before the discovery of the first exoplanet in 1995. These studies foresaw that extrasolar planets could be detected by precise and continue observations of luminosities of stars. The regular and periodic decrease of luminosities of these stars due to the transits of a planet could be detected by some instruments existing at this time. One can wonder why these premonitory studies were ignored. Actually, they were not published in the most famous referred journals. Maybe the reason is that they were too different and too new in comparison to the research results published at these times. Maybe as well because the subject was not considered as a serious one. However, if the study of Belorizky (Belorizky 1938) published in a French journal for amateur astronomers was not ignored and if some survey observation programs were carried out, perhaps extrasolar planets would be discovered much earlier than the discovery of 51 Peg b in 1995. The discovery of exoplanets since the discovery of 51 Peg b not only opened a new research area very fruitful and successful but also changed some study methods at it can be seen in astronomical publications. Imagination is being always a part of the scientific research but occupies now a more significant part, larger than ever. A so rapid evolution of the scientific methods was very rarely observed.

48

D. Briot and J. Schneider

How can we know if these studies which take into account not only the existence but also the intelligence and the intentions of the extraterrestrial civilizations will be considered in the future as very clever and premonitory or on the contrary as somewhat naive and a little ridiculous?

References Arnold LFA (2005) Transit light-curve signatures of artificial objects. Astrophys J 627:534–539 Baumgartner FJ (1987) Sun spots or Sun’s planets: Jean Tarde and the sunspot controversy of the early seventeenth century. J Hist Astron 18:44–54 Belorizky D (1938) Le Soleil, Etoile Variable. L’Astronomie 52:359–361 Borucki W, Dunham E, Koch D et al (1996) FRESIP: a mission to determine the character and frequency of extra-solar planets around solar-like stars. Astrophys Space Sci 241:111 Copernicus N (1543) De revolutionibus orbium coelestium. Johan Petreius, Nuremberg Deeg H, Martin E, Schneider J et al (1997) The TEP network – a search for transits of extrasolar planets: observations of CM Draconis in 1994. Astron Astrophys Trans 13:233 Doyle L 1985 Assisting extrasolar planetary detection through the determination of stellar space orientations. In: IUA Symposium 112, p 97 Doyle L, Wilcox T, Lorre J (1984) The space orientation of stars. Astrophys J 287:307 Fontenelle BLB d (1686) Entretien sur la Pluralité des Mondes. Vve. Blageart C, Paris [English translation used here: (1803) Conversations on the plurality of worlds. translated by Gunning E, Cundee, London] Galilei G (1613) Istoria e dimostrazioni intorno alle macchie solari. Giacomo Mascardi, Roma Gassendi P (1632) Mercurius in sole visus, et Venus invisa Parisiis, anno 1631. Sébastien Cramoisy, Paris Goodricke (1783) A series of observations on, and a discovery of, the period of the variation of the light of a bright star in the head of medusa, called Algol. Phil Trans R Soc London 73:474–482 Hoskin M (1979) Goodricke, Pigott and the quest for variable stars. J Hist Astron 10:23–41 Huygens C (1698) Kosmotheoros: sive de terris coelestibus, earumque ornatu conjecturae. Den Haag [English translation: The celestial worlds discovered: or, conjectures concerning the inhabitants, plants and productions of the worlds in the planets. London] Kepler J (1619) Harmonices mundi. Libri V, Linz Kepler J (1627) Tabulae Rudolphinae. J. Saurii, Ulm Kepler J (1629) Admonitio ad astronomos, rerumque coelestium studiosos, de raris mirisq[ue] anni 1631. Jakob Bartsch, Leipzig Kipping DM, Teachey A (2016) A cloaking device for transiting planets. Mon Not R Astron Soc 459:1233–1241 Korpela EJ, Sallmen SM, Greene DL (2015) Modeling indications of technology in planetary transit light curves-dark-side illumination. Astrophys J 809:139. (13pp) Kostov VB, McCullough PR, Carter JA et al (2014) Kepler-413b: a slightly misaligned, Neptunesize transiting circumbinary planet. Astrophys J 784:14. (18pp) Lagrange AM, Gratadour D, Chauvin G et al (2009) A probable giant planet imaged in the “ pictoris disk. Astron Astrophys 493(2):L21–L25 Lallemand R, Renard L, Servan B (1970) Sur une caméra électronique à grand champ destinée à la photométrie astronomique. CR Acad Sci Paris Dsérie B 270:385 Lardner D (1853) Hand-book of natural philosophy and astronomy. Walton and Maberly, London, p 771 Lecavelier des Etangs A, Perrin G, Ferlet R et al (1993) Observation of the central part of the Beta-Pictoris disk with an anti-blooming CCD. Astron Astrophys 274:877–882 Lecavelier des Etangs A, Deleuil M, Vidal-Madjar A et al (1995) “ Pictoris: evidence of light variations. Astron Astrophys 299:557–562

3 Prehistory of Transit Searches

49

Le Verrier UJ (1846) Recherches sur les mouvements d’Uranus. Astron Nachr 25:85–92 Ledger E (1879) Transits of intra-mercurial planets. Observatory 3:251–252 Mayor M, Queloz D (1995) A Jupiter-mass companion to a solar-type star. Nature 378:355–359 Napier J (1614) Logarithmorum Canonis descriptio. Andrew Hart, Edinburgh Nowak M, Lacour V, Lapeyrère V, David L, Crouzier A, Schworer G, Perrot P, Rayane S (2017) A compact and lightweight fibered photometer for the PicSat mission. https://arxiv.org/abs/1708.04015 Perryman MAC (2000) Extra-solar planets. Rep Prog Phys 63:1209–1272 Piggot E (1785) Observations of a new variable star. Phil Trans R Soc London 75:127–136 Rémy G (1945) Clarté sur la route. Casterman, Paris, p 41 Schneider J (1994a) On the search for O2 in extrasolar planets. Astrophys Space Sci 212:321–325 Schneider J (1994b) On the occultations of a binary star by a circum-orbiting dark companion. Planet Space Sci 42:539–544 Strand KAa (1943) 61 Cygni as a triple system. Publ Astron Soc Pac 55:29–32 Struve O (1952) Proposal for a project of high-precision stellar radial velocity work. Observatory 72:199–200 Tarde J (1620) Borbonia Sidera. Jean Gesselin, Paris. [French translation (1623) Les astres de Borbon et apologie pour le soleil. Jean Gesselin, Paris] Van Helden A (1976) The importance of the transit of mercury of 1631. J Hist Astron 7:1–10 Vaquero JM (2007) Historical sunspot observations: a review. Adv Space Res 40:929–941

4

Discovery of the First Transiting Planets Edward W. Dunham

Contents Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Early Ground-Based Transit Searches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . OGLE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Proto-TrES, Vulcan, and Kepler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . HD209458 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . TrES-1, TrES-2, and False Positives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Parting Thoughts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross-References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

52 53 53 55 60 61 61 62 62

Abstract

Early thinking about detecting extrasolar planets was largely circumscribed by the expectation that other solar systems would be similar to our own, the only known example at the time. Given this mind-set, transit detections were expected to be exceedingly difficult for small planets and rarely seen for larger ones. The discovery of 51 Peg and subsequent hot Jupiters by the radial velocity method completely upended our thinking – transits were suddenly practical, perhaps even easy! This immediately led to follow-up searches for transits in systems discovered by the radial velocity technique and, conversely, to wide-field groundbased transit search programs with radial velocity follow-up observations. As is usually the case, transit work turned out to be harder than initially expected but was still possible and productive. This chapter reviews the circumstances leading to the first transit observations of HD 209458, the early OGLE exoplanets,

E. W. Dunham () Lowell Observatory, Flagstaff, AZ, USA e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_170

51

52

E. W. Dunham

and TrES-1 and TrES-2, as well as some of the frustrations and difficulties encountered along the way. Keywords

Transit photometry · History · Discovery · TrES · OGLE · HD209458

Background The idea of detecting extrasolar planets by means of transit observations dates back at least to Otto Struve (Struve 1952) in a paper discussing the potential for radial velocity and “eclipse” detections. Struve, in fact, alludes to the possibility of what later became known as hot Jupiters, by analogy with spectroscopic binary stars. In a sense he was prescient, though we don’t believe that hot Jupiters form the same way close binaries do. Subsequently such thoughts disappeared, and approaches to detect extrasolar planets were conceived and judged on the basis of their ability to find planets like those in our own solar system (Rosenblatt 1971; Borucki and Summers 1984; COMPLEX 1990). This was reasonable enough – giant planets would form beyond the ice line where more abundant materials in the solar nebula could condense and smaller planets made of more refractory, and less abundant, materials would form closer to the star. In this sensible and comfortable paradigm, large transiting planets would be hard to find because of (1) the low probability of suitable orbital plane alignment for their large orbital radii and (2) infrequent transits due to long orbital periods. Small planets (with presumed smaller orbital radii) were less bad in these areas, but the transit depth would be 0.01%, which seemed to most people to be impossibly small in those days. It was in this environment that what eventually became the Kepler mission was first proposed as FRESIP (Borucki et al. 1996) to a rather unreceptive audience, 1 year before everything changed. The discovery of 51 Peg (Mayor and Queloz 1995) and the subsequent hot Jupiters that came close on its heels (see http://exoplanet.eu and sort the catalogue by discovery date) had a galvanizing effect on the field. For these objects the orbital alignment probability was 10%, transit depths were 1%, and orbital periods were only a few days! Within a year or so, enough radial velocity planets had been found that the idea that one of them would soon show transits had occurred to many people. This can be thought of as transit follow-up confirmation of planets discovered by radial velocity variations. The converse idea of undertaking transit searches with radial velocity confirmation was also current and led to the application of the OGLE system to transit searching (Udalski et al. 2002a) and development of several widefield ground-based transit search programs including, but certainly not limited to, the Vulcan (Borucki et al. 2001), Planet Search Survey Telescope (PSST; Dunham et al. 2004), and Stellar Astrophysics and Research on Exoplanets (STARE; https:// www.hao.ucar.edu/research/stare/overview.html) projects. The actual situation was less clean than this with interconnections among projects, including the developing Kepler concepts, and with radial velocity and photometric observations sometimes going on simultaneously. The HD 209458

4 Discovery of the First Transiting Planets

53

transit discovery was tangled up with the development of STARE in particular, but also Vulcan and PSST, and in turn provided motivation for the OGLE work that was soon to come. Nothing happens in a vacuum.

Early Ground-Based Transit Searches One of the first questions to resolve when designing a transit search project is how to maximize the number of targets the system can observe. Roughly speaking, maximizing the product of the telescope’s light-collecting area and the solid angle of the field of view, the A product, does this. Larger A allows more common fainter stars to be observed, and larger  allows more stars to be observed simultaneously. For a given detector area, the fastest f/ratio optical system available is best. The location of the target field is obviously also important with crowded fields being best, at least at first glance. What was less obvious was where on the A continuum would be best. Smaller telescopes with wider fields would be inexpensive and relatively easy to set up with brighter limiting magnitudes and easier follow-up, but the efficacy of differential photometry was expected to deteriorate with wider field systems. There was really no experience to guide this decision, so the first systems were developed around the equipment at hand. For OGLE the decision was preordained since the system was in operation already. The proto-TrES projects, STARE and PSST, achieved comparable A product by using much smaller “telescopes” with much wider fields of view and large pixel size mapped to the sky. Initially these projects used 300 mm focal length f/2.5 Aero-Ektar lenses that had been in storage in a garage at Lowell in combination with then-common 2Kx2K CCDs. Vulcan initially used the same lens type but a different CCD format. Although this lens type was later abandoned, its focal length and f/ratio remained.

OGLE The Optical Gravitational Lensing Experiment (OGLE) commenced in 1992 with a single 2Kx2K CCD detector (Udalski et al. 1992) and the goal of discovering gravitational microlensing events by photometric monitoring of millions of stars in crowded southern Milky Way fields. The success of the project, and its limitations, led to implementation of an 8Kx8K CCD array (Udalski et al. 2002a) on the 1.3-m Warsaw telescope at Las Campanas Observatory with a field of view of about 35  35 arcmin. This was put into operation in 2001. Figure 1 shows the telescope with this camera installed. In addition the previous DoPHOT-based profile-fitting data analysis pipeline was changed to one based on image subtraction (Alard and Lupton 1998; Alard 2000; Udalski 2003a) resulting in much better photometric precision. The third phase of the OGLE survey, OGLE-III, began in June 2001 with the intent accomplishing a much more complete gravitational microlensing survey, a search for exoplanet transits, and other research areas that could be addressed

54

E. W. Dunham

Fig. 1 The 1.3-m Warsaw telescope with the 8 K square camera installed, taken in 2006. (Image by OGLE team member Krzysztof Ulaczyk)

with the anticipated rich dataset. The OGLE-III survey was successful in every respect (Udalski,  Microlensing Surveys for Exoplanet Research (OGLE Survey Perspective), this volume). When the HD 209458 transit was announced in 1999, the large mosaic camera for OGLE-III was well underway with plans to begin operating it in mid-2001. The A product and photometric precision of the final system would be sufficient for the task, and the OGLE team had a great deal of experience already with crowded field photometry and what we would now call “big data.” Nobody had yet tackled the problem of a transiting exoplanet survey, and a better match between hammer and nail could hardly be imagined. The next question was which field to choose. The Galactic disk fields would be preferred because they were less crowded than the Galactic bulge fields, but the timing argued otherwise. The new system was ready to go in June 2001, at the end of the observing season for the disk fields, and the decision was made to try the bulge fields first. The intended transit targets would be disk stars in the closest Galactic arm, 1–2 kpc distant, while the other stars would be good microlensing targets. Observations commenced, running for 7 weeks with generally good weather. A large dataset was obtained with 15 min cadence on several fields. By the end

4 Discovery of the First Transiting Planets

55

of the year, the photometry was complete, and searching the light curves for transits was the next step. The OGLE-III survey was successful in identifying a large number of candidate transiting objects. These were published (Udalski et al. 2002a, b, c, Udalski et al. 2003b) without follow-up observations so that others in the community could contribute to the follow-up spectroscopy effort. The decision to proceed this way rather than attempt the follow-up work within the OGLE team was not a small one. On the one hand, it did result in speedy follow-up with substantial resources in the community being brought to bear, but on the other hand proper attribution of the original data source was sometimes lacking. The first successful confirmation of an OGLE exoplanet, OGLE-TR-56b, occurred shortly thereafter (Konacki et al. 2003a). This was the first example of an exoplanet discovered by transit and confirmed with radial velocity observations. The same paper also identified almost all of the other initial candidates as either unsuitable for follow-up (early spectral types, too faint, obvious binaries from light curve morphology, single transits) or false positives (blends, spectroscopic binaries), giving an early sense of the high false-positive rate among these objects. This problem was certainly not unique to OGLE. Over the next year or so, four more OGLE candidates were confirmed to be exoplanets: OGLE-TR-10b, which didn’t give up its identity without putting up a good fight, (Konacki et al. 2003b, 2005; Bouchy et al. 2005), OGLE-TR-111b (Pont et al. 2004), OGLE-TR-113b (Konacki et al. 2004; Bouchy et al. 2004), and OGLETR-132b (Bouchy et al. 2004). The faint targets characteristic of OGLE’s position in the A continuum required extensive effort and significant observational resources for follow-up observations.

Proto-TrES, Vulcan, and Kepler The Vulcan project was closely tied to the developing Kepler mission concept, and there were ties between the principals of Vulcan (W. Borucki), STARE (T. Brown), and PSST (E. Dunham) from previous and ongoing collaborations and a common interest in precise photometry. A workshop held by Tim Brown at the High Altitude Observatory in the fall of 1996 on a related topic had the side effect of kicking off development of these ground-based projects, ending with a to-do list for all of us to work on. Progress was fast at first, and a system was mostly put in place in a small roll-off building at Lowell in front of the Trustee’s house by late winter that used an existing 2Kx2K CCD system and one of the Aero-Ektar lenses from a garage at Lowell. The CCD part of the system was needed for occultation prediction work in Australia in February so our first test data were obtained in March, 1997. During the summer the same CCD system was needed again for occultation prediction work. Brown obtained internal HAO funds for a CCD camera and assembled a second system with the other Aero-Ektar lens from the garage at Lowell. This system was named STARE and was operated on a friend’s farm nearby with the computers being in an adjacent repurposed turkey coop. These two systems are shown in Fig. 2. The

56

E. W. Dunham

Fig. 2 The original Aero-Ektar test systems: the proto-PSST system left, in the Trustee’s front yard at Lowell and STARE right, at the Foothills Lab site after the farm had been sold. (STARE picture from Tim Brown)

Vulcan team set up shop in the previously empty Crocker dome at Lick Observatory on Mount Hamilton. Our first serious test data were taken in the winter of 1997– 1998 with three systems based on the Aero-Ektar lenses as well as a test by the Vulcan team using a shorter focal length f/1.5 lens that was less successful. The test results were encouraging enough to obtain NASA funding to build and operate systems suitable for a serious transit search project. A photometry workshop was held at the SETI institute in the fall of 1998 during which we shared results (Borucki and Lasher 2000). Among other things this workshop revealed two problems, one recognized and one only partially appreciated. The first problem was that the image quality of the Aero-Ektar lenses was not good, which Brown subsequently found was due to spherical aberration. The only way to cure this was to stop down the lenses, but this had a fatal impact on sensitivity. The only solution was to abandon the old lenses. Brown dealt with this by making a Schmidt of the same focal length and aperture for the STARE system (because, as an old amateur telescope maker, he had always wanted to make a Schmidt), while the Vulcan and PSST systems used commercial camera lenses with the same focal length and f/ratio. The second problem was that our observed noise level was higher than expected from photon noise from the target star and sky background when combined with scintillation. At the time we thought this was due

4 Discovery of the First Transiting Planets

57

Fig. 3 Locations of the three sites in the TrES network, circa 2003–2004

to inadequacies in our lenses and data analysis software, which it largely was, but in retrospect it was also related to low-frequency noise that crept into the differential photometry (Pont et al. 2006) and that afflicts wide-field ground-based transit search programs in varying degrees. Through 1999 our new systems were taking shape with STARE being in a workable state by the fall (when it was used for the discovery of transits in HD209458). The Lowell system, now named PSST and located at its dark Anderson Mesa site, was delayed till the spring of 2000 by an obscure problem that slowed delivery of its CCD control electronics. Automated operation was implemented and we began more or less routine operation. The STARE system was moved to Tenerife in 2001, providing dark sky and much-needed longitude coverage. During 2003 Dave Charbonneau assembled another system at Palomar known as Sleuth. This provided a little additional longitude coverage but also a very different weather pattern than Flagstaff and was an extremely important addition to the collaboration, all the more so as our collaboration with the Vulcan team waned. The final configuration of the TrES network after moving STARE and adding Sleuth is shown in Fig. 3 while the systems themselves are shown in Fig. 4. Data analysis was a continuing problem because of a shortage of computing power, and our noise performance was still not as expected in spite of the improved optics. At that time our data analysis pipeline was based on weighted aperture photometry. We had experimented with DAOPHOT, but our computing capacity was inadequate for such a massive application. Georgi Mandushev joined the PSST team in 2002 and, armed with faster computers, was able to implement an image subtraction or differential image analysis (DIA) pipeline in 2003 (Dunham et al.

58

E. W. Dunham

Fig. 4 The three TrES transit photometry systems. STARE (upper left), Sleuth (upper right), and PSST (lower left) in their final configurations and locations at Tenerife, Palomar, and Anderson Mesa, respectively. (STARE image from STARE team; Sleuth image from Dave Charbonneau)

2004) based in part on the algorithms of Alard (2000) together with decorrelation based on the approach of Jenkins et al. (2000). This change in our data analysis pipeline was crucial to the success of the enterprise as it improved the noise performance by a factor of 2–3 for the brighter stars. Our final noise performance, illustrated with PSST data, is shown in Fig. 5. In this figure the vertical axis is the v1/2 (n) function defined by Pont et al. (2006). It is the fractional uncertainty in the depth of a given transit. The horizontal axis is the number of in-transit data points for the same transit. In the case of PSST, a 2.5-h transit duration occurs at the right edge of this plot, and the overall uncertainty in the transit depth, integrated over the entire transit, can be read from the vertical axis. Figure 5 also shows how decorrelation and SYSREM (Tamuz et al. 2005) perform for bright (left) and faint (right) stars. For bright stars decorrelation was superior, but for faint stars the two approaches were similar, given that decorrelation also reduces the transit amplitude by 20%. Although we didn’t look at our data this way until late 2006, the same data analysis pipeline was used back to 2003 so the same performance applies.

4 Discovery of the First Transiting Planets

59

Fig. 5 The final TrES noise performance is shown for the example of PSST for stars at the bright (left) and faint (right) ends of the PSST dynamic range of about 3.4 magnitudes

Early on we also recognized that the difficult photometric problems we had been dealing with were the easy part; the complexity and effort needed for follow-up imaging and spectroscopic observations to weed out false positives and then confirm transit candidates as real planets were another major undertaking. Rather than following OGLE’s approach of publishing candidates and allowing the community to take on the follow-up work, we expanded our team to deal with the problem. Dave Latham’s group took on the bulk of the “reconnaissance” spectroscopy using the CfA Digital Speedometer (Latham 1992) to weed out hot stars, fast rotators, and spectroscopic binaries. Willie Torres began applying his TODCOR and line bisector analyses and astrophysical modeling of multiple systems as he had done previously with OGLE candidates to help eliminate triple systems masquerading as exoplanets. In fact he was so successful in this enterprise that Latham nicknamed him “Killer.” The problem of follow-up and false positives also impacted our field choices causing us to prefer fields somewhat off the Galactic plane to reduce the number of giants in our fields. This was completely mixed up with discussions related to the change in the planned location of the Kepler field at the same time for the same reason.

60

E. W. Dunham

HD209458 In the summer of 1999, Dave Charbonneau, then a Harvard graduate student, drove to Colorado to take on his thesis research with Tim Brown working with the STARE system at the High Altitude Observatory (HAO). Having never done time series photometry before, he contacted Dave Latham before driving west to see if there might be a single star that was interesting so he could get some experience before tackling thousands of stars at a time. As it happened Latham was involved in a radial velocity collaboration that, among other targets, was investigating HD 209458, a good fall object. This star was a very promising radial velocity candidate, but some nagging details regarding correlations with stellar activity indicators had so far prevented its announcement. This met the need – a single, bright, interesting star. The plan was to use the newly improved STARE system, now located at the Foothills Lab site because the farm had been sold, for Dave’s time series photometry trial run. Observations were attempted in August but were foiled by bad weather. The first predicted transit that occurred during clear weather was on September 9, 1999, UT and a subsequent transit 1 week (two orbital periods) later on the 16th also had good weather. In addition many non-transit nights of photometry had been obtained in late August and the first half of September. However, the analysis needs of Ron Gilliland’s 47 Tuc HST data (Gilliland et al. 2000) intervened along with other problems, and the HD 209458 data weren’t analyzed till November. The transit was clearly there in both datasets. Charbonneau (2017, private communication) points out that Oban is the official whisky of transiting planets because that is what Tim served when they finally had a chance to look at the plots together at his home. Meanwhile HD 209458 was also on Geoff Marcy’s target list, and Greg Henry was observing radial velocity candidates with the automatic photoelectric telescopes (APT) at Fairborn Observatory (Henry 1999). An APT observed a partial transit on November 7 with timing consistent with the radial velocity orbit. This prompted a story in the New York Times. Later in the month, several additional transit observations appeared in the IAU Circulars (7314, 7317, and 7323) together with an ephemeris provided by Latham and collaborators (IAUC 7315). Papers were written, reviewed, and published at warp speed with the two groups’ papers appearing back-to-back in the January 20, 2000, Ap. J. Letters (Charbonneau et al. 2000 with two full transits; Henry et al. 2000 with one partial transit). This discovery had a truly galvanizing effect on the field. Although the subsequent HST observation of several HD 209458 transits (Brown et al. 2001) wasn’t involved in its discovery, it showed the promise of high-quality observations with large facilities for these bright transiting targets and led the way into a whole new realm of detailed study of exoplanets.

4 Discovery of the First Transiting Planets

61

TrES-1, TrES-2, and False Positives With the advent of the new DIA/decorrelation data processing pipeline in 2003, faster computers, and the new STARE and PSST systems running routinely at good sites, the stage was set for the proto-TrES collaboration to discover an exoplanet. The discovery observations occurred during the summer of 2003, while the Sleuth system was still under development. Our first transit detection turned out to have a period of almost exactly 3 days so that transits repeated again and again at a given longitude. This had the effect of increasing the number of observed transits and our overall confidence in the detection. In this case, the favored longitude was that of Tenerife, and STARE was the system that observed all of the transits. Although PSST observed no transits, the non-detections were important in ruling out other possible periods that would have resulted in transits in the PSST data, and it did observe transits in the 2004 season when the preferred longitude happened to be in the Western USA. The problems of blends, multiple star systems, grazing eclipses, etc. were well known in the community by this time (e.g., Konacki et al. 2003b; Brown 2003) so substantial effort was devoted to ruling out possible non-planetary explanations for the observed light curves. The photometric and spectroscopic follow-up observations are well described in the discovery paper (Alonso et al. 2004) and won’t be repeated here. It is worth noting that this exoplanet was discovered via transits with a 0.1 m telescope, while the high-precision radial velocity confirmation came from the 10-m Keck telescope and HIRES spectrograph. As a side note, it wasn’t until we knew that we had a real planet that a name for our collaboration was needed. We were writing a paper and had to call the planet something, after all. Thus the Transatlantic Exoplanet Survey (TrES) came into existence. At this point optimism that additional planets would soon be found was on the rise, but 2 years would pass before TrES-2 was found (O’Donovan et al. 2006a), the first transiting planet in the Kepler field. Many very promising candidates were found in the meantime, but all fell victim to the vetting process (Mandushev et al. 2005; O’Donovan et al. 2006b, 2007). The Vulcan project had a similar experience (Jenkins et al. 2002) as did OGLE, in spite of the very different place that OGLE occupied along the A continuum. The main difference is that while the workload for vetting wide-field survey transit candidates is high, it borders on prohibitive for the fainter OGLE candidates (Pont et al. 2008).

Parting Thoughts The early days of exoplanet transit searches were exciting and heady but also frustrating and a lot more work than we imagined at the outset. On the other hand,

62

E. W. Dunham

the additional complications meant that many more smart, creative, and committed people became involved and the whole field flourished. Alas, in a short survey paper, it is impossible to mention all of the people who played important roles. The reader may wonder why the OGLE and TrES planets are as few as they are and why these transit search projects stopped. For OGLE the principal reason was that wide-field search programs matured and became very productive, with targets that were much easier to confirm. The OGLE team moved on to other important undertakings that their considerable capability could contribute to. In the case of TrES, the principals were involved in other projects, notably Kepler and MEarth (https://www.cfa.harvard.edu/MEarth/Welcome.html) but also SOFIA and the developing LCOGT, and needed to focus on those priorities. At the same time, it was clear that the way forward with ground-based wide-field transit searches was to grow them to an industrial scale, with many cameras at several locations, as the HAT and WASP projects have done. We also look forward to the results from TESS, which is a logical successor to both the ground-based wide-field transit search programs and Kepler.

Cross-References  HD189733b: The Transiting Hot Jupiter That Revealed a Hazy and Cloudy

Atmosphere  Microlensing Surveys for Exoplanet Research (OGLE Survey Perspective)  Prehistory of Transit Searches  Space Missions for Exoplanet Science: Kepler/K2  Space Missions for Exoplanet Research: Overview and Introduction  The Discovery of the First Exoplanets  The HATNet and HATSouth Exoplanet Surveys  Transit Photometry as an Exoplanet Discovery Method

References Alard C (2000) Image subtraction using a space-varying kernel. A&AS 144:363–370 Alard C, Lupton RH (1998) A method for optimal image subtraction. ApJ 503:325–331 Alonso R, Brown TM, Torres G et al (2004) TrES-1: the transiting planet of a bright K0 V star. ApJ 613:L153–L156 Borucki WJ, Lasher LE (2000) Third workshop on photometry. NASA/CP-2000-209614 Borucki WJ, Summers AL (1984) The photometric method of detecting other planetary systems. Icarus 58:121–134 Borucki WJ, Dunham EW, Koch DG et al (1996) FRESIP: a mission to determine the character and frequency of extra-solar planets around solar-like stars. Ap&SS 241:111–134 Borucki WJ, Caldwell D, Koch DG, Webster LD et al (2001) The Vulcan photometer: a dedicated photometer for extrasolar planet searches. PASP 113:439–451 Bouchy F, Pont F, Santos NC et al (2004) Two new “very hot Jupiters” among the OGLE transiting candidates. A&A 421:L13–L16

4 Discovery of the First Transiting Planets

63

Bouchy F, Pont F, Melo C et al (2005) Doppler follow-up of OGLE transiting companions in the galactic bulge. A&A 431:1105–1121 Brown TM (2003) Expected detection and false alarm rates for transiting Jovian planets. ApJ 593:L125–L128 Brown TM, Charbonneau D, Gilliland RL, Noyes RW, Burrows A (2001) Hubble space telescope time-series photometry of the transiting planet of HD 209458. ApJ 552:699–709 Charbonneau D, Brown TM, Latham DW, Mayor M (2000) Detection of planetary transits across a sun-like star. ApJ 529:L45–L48 COMPLEX (1990) Strategy for the detection and study of other planetary systems and extrasolar planetary materials: 1990–2000. National Academy Press, Washington, DC Dunham EW, Mandushev GI, Taylor BW, Oetiker B (2004) PSST: the planet search survey telescope. PASP 116:1072–1080 Gilliland RL, Brown TM, Guhathakurta P et al (2000) A lack of planets in 47 Tucanae from a Hubble Space Telescope search. ApJ 545:L47–L51 Henry GW (1999) Techniques for automated high-precision photometry of sun-like stars. PASP 111:845–860 Henry GW, Marcy GW, Butler RP, Vogt SS (2000) A transiting “51 Peg-like” planet. ApJ 529: L41–L44 Jenkins JM, Witteborn F, Koch DG et al (2000) Processing CCD images to detect transits of earthsized planets: maximizing sensitivity while achieving reasonable downlink requirements. Proc SPIE 4013:520–531 Jenkins JM, Caldwell DA, Borucki WJ (2002) Some tests to establish confidence in planets discovered by transit photometry. ApJ 564:495–507 Konacki M, Torres G, Jha S, Sasselov DD (2003a) An extrasolar planet that transits the disk of its parent star. Nature 421:507–509 Konacki M, Torres G, Sasselov DD, Jha S (2003b) High-resolution spectroscopic follow-up of OGLE planetary transit candidates in the galactic bulge: two possible Jupiter-mass planets and two blends. ApJ 597:1076–1091 Konacki M, Torres G, Sasselov DD et al (2004) The transiting extrasolar giant planet around the star OGLE-TR-113. ApJ 609:L37–L40 Konacki M, Torres G, Sasselov DD, Jha S (2005) A transiting extrasolar giant planet around the star OGLE-TR-10. ApJ 624:372–377 Latham DW (1992) Surveys of spectroscopic binaries at the center for astrophysics. In: HA MA, Hartkopf WI (eds) Complementary approaches to double and multiple star research. IAU Colloquium 135, ASP Conference Series 32, San Francisco, pp 110–118 Mandushev G, Torres G, Latham DW et al (2005) The challenge of wide-field transit surveys: the case of GSC 01944-02289. ApJ 621:1061–1071 Mayor M, Queloz D (1995) A Jupiter-mass companion to a solar-type star. Nature 378:355–359 O’Donovan FT, Charbonneau D, Mandushev G et al (2006a) TrES-2: the first transiting planet in the Kepler field. ApJ 651:L61–L64 O’Donovan FT, Charbonneau D, Torres G et al (2006b) Rejecting astrophysical false positives from the TrES transiting planet survey: the example of GSC 03885-00829. ApJ 644:1237–1245 O’Donovan FT, Charbonneau D, Alonso R et al (2007) Outcome of six candidate transiting planets from a TrES field in Andromeda. ApJ 662:658–668 Pont F, Bouchy F, Queloz D et al (2004) The “missing link”: a 4-day period transiting exoplanet around OGLE-TR-111. A&A 426:L15–L18 Pont F, Zucker S, Queloz D (2006) The effect of red noise on planetary transit detection. MNRAS 373:231–242 Pont F, Tamuz O, Udalski A et al (2008) A transiting planet among 23 new near-threshold candidates from the OGLE survey – OGLE-TR-182. A&A 487:749–754 Rosenblatt F (1971) A two-color photometric method for detection of extra-solar planetary systems. Icarus 14:71–93 Struve O (1952) Proposal for a project of high-precision stellar radial velocity work. Observatory 72:199–200

64

E. W. Dunham

Tamuz O, Mazeh T, Zucker S (2005) Correcting systematic effects in a large set of photometric light curves. MNRAS 356:1466–1470 Udalski A, Szymanski M, Kaluzny J et al (1992) The optical gravitational lensing experiment. AcA 42:253–284 Udalski A, Paczynski B, Zebrun K et al (2002a) The optical gravitational lensing experiment. Search for planetary and low-luminosity object transits in the galactic disk. Results of 2001a campaign. AcA 52:1–37 Udalski A, Szewczyk O, Zebrun K et al (2002b) The optical gravitational lensing experiment. Planetary and low-luminosity object transits in the Carina fields of the galactic disk. AcA 52:317–359 Udalski A, Zebrun K, Szymanski M et al (2002c) The optical gravitational lensing experiment. Search for planetary and low-luminosity object transits in the galactic disk. Results of 2001b campaign - supplement. AcA 52:115–128 Udalski A (2003a) The optical gravitational lensing experiment. Real time data analysis systems in the OGLE-III survey. AcA 53:291–305 Udalski A, Pietrzynski G, Szymanski M et al (2003b) The optical gravitational lensing experiment. Additional planetary and low-luminosity object transits from the OGLE 2001 and 2002 observational campaigns. AcA 53:133–149

5

The Way to Circumbinary Planets Laurance R. Doyle and Hans J. Deeg

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Circumbinary Planet Detection Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eclipse Timing Variation Based on the Light Time Effect . . . . . . . . . . . . . . . . . . . . . . . . . . . Transits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gravitational Lensing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eclipse Timing Variation from Dynamical Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Radial Velocity Variations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reflected Light or Eclipse Echos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Habitability of M-Dwarfs and Circumbinary Planets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross-References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

66 68 69 69 72 74 74 75 76 77 79 81 81 82

Abstract

The historic quest to detect circumbinary planets (CBPs) dates back to a time before the first extrasolar planets were detected. Eclipsing binary star systems (EBs) were considered prime targets for the detection of CBP transits, as it was

L. R. Doyle () Institute for the Metaphysics of Physics, One Maybeck Place, Principia College, Elsah, IL, USA Carl Sagan Center, SETI Institute, Mountain View, CA, USA e-mail: [email protected] H. J. Deeg Instituto de Astrofísica de Canarias, La Laguna, Tenerife, Spain Departamento de Astrofísica, Universidad de La Laguna, La Laguna, Tenerife, Spain e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_115

65

66

L. R. Doyle and H. J. Deeg

considered likely that the planetary orbits would also be close to edge on to our line of sight and so cross (transit) the stellar discs of the eclipsing stars. The presence of CBPs remained however doubtful until the unequivocal detection, by transit, of Kepler-16b and of further CBPs with the NASA Kepler space telescope. Stellar eclipses were also timed for about a dozen small-mass main-sequence EBs as well. In this chapter we discuss the history of theory and observations in the search for CBPs and the various techniques that have been applied, as well as several methods that might provide results in the future.

Introduction Most of what we know about the sizes of main sequence stars comes from the study and modeling of eclipsing binary (EB) stellar systems – stars that orbit in front of (i.e., eclipse) each other across our line of sight. And most of what we know about the sizes of extrasolar planets comes from their transits across the stellar disc(s) around which they orbit. Thus circumbinary planets (CBPs) combine the best of both worlds (so to speak) and have resulted in the most precisely measured planetary systems outside of our own solar system. This is due not just to direct sampling of stellar and planetary size ratios during eclipses and transits, respectively, but also due to the predictable – but quasi-periodic – nature of these orbital events. While CBP transits can be quite complex, their interpretation is less equivocal once detected, as we shall see. In this chapter we will review the short history of the quest to detect CBPs, culminating in the detection of Kepler-16b – the first direct detection of a CBP. The term “direct” we feel applies to either the direct detection of reflected light from an extrasolar planet by imaging, or to the direct detection of a planet’s shadow, seen when the planet transits its star across our line of sight. (Upon consideration, the term “direct detection” might also apply to microlensing, as well; see below.) There are also indirect methods for detecting CBPs by, for example, their gravitational displacement of the central parent stars, and we shall touch on these methods as well. The nature of eclipsing binary stars was first correctly explained by the Dutch astronomer John Goodricke working in England in May of 1783. He was the first to propose that variable stars like Algol were actually binary systems whose orbital plane caused them to regularly eclipse each other across our line of sight (Goodricke 1784). The idea that planets could also be orbiting edge on across our line of sight, and so be detected photometrically, appears to have been first suggested by the German astronomer Otto Struve (1952). The idea that these two features could be combined in a search for CBPs originated with Borucki and Summers (1984) and, in more detail, with Schneider and Chevreton (1990). They argued that the orbital plane of EBs, already being edge on to our line of sight, would optimize the likelihood of any planetary orbital planes also being edge on across our line of sight – citing models of protoplanetary discs, and the close alignment of the solar equator with the planetary plane of our solar system, as suggestive examples.

5 The Way to Circumbinary Planets

67

In addition, it was pointed out (Schneider 1994; Schneider and Doyle 1995) that a CBP’s orbit would precess so that the rotating line of nodes would cause even noncoplanar CBP transits to appear and disappear, albeit with periods significantly longer than the CBP’s orbital period itself (see Martin and Triaud 2015; Martin 2017 for a more detailed development of this idea). It was also recognized that the smallest-sized (i.e., very late-type dwarf) stellar discs – like the CM Draconis eclipsing binary system – would optimize any planetary transit signals, as the photometric detectability of such planets depends upon the ratio of the planetary disc to the stellar disc area. (Both components of the CM Dra system combined are still only about 12% the area of the solar disc.) Also such small-mass systems would optimize any timing offsets caused by a third body mass in a circumbinary orbit (see, e.g., Doyle et al. 1998; Deeg et al. 2000, 2008; Morales et al. 2009). Thus the fairly bright twelfth magnitude (in the visual) M4-M4 dwarf EB system, CM Draconis, became the primary target for the first global search for CBPs. The search consisted of an international network of observatories at different longitudes, called the TEP (transit of extrasolar planets) observing network, which operated from 1993 to 1999. TEP observatories were located in California, Korea, Russia, France, Spain, New York, and New Mexico (Deeg et al. 1998). A CBP transit detection algorithm (Jenkins et al. 1996; Doyle et al. 2000) was developed which cross-correlated the observational light curves with all reasonable planetary transit sizes (i.e., planetary radii) and orbital characteristics (i.e., period, inclination, eccentricity, etc.). Falsealarm probabilities were quantified using statistics generated by the insertion of artificial planets in between eclipse features in the observational light curves. (This approach was eventually extended to the quantification of false alarms for the NASA Kepler mission.) Finally, an equation for timing precision of stellar eclipses based upon observables, for the detection of non-transiting CBPs, was found to be fairly robust when tested over many different types of EB systems (Doyle and Deeg 2004; Sybilski et al. 2010; Deeg and Tingley 2017). The TEP project represented the first thorough search for CBPs within the habitable zone of a main-sequence EB system, reaching a detection limit of super-Earth-sized planets. A detection probability of greater than 90% was achieved for any transiting CBPs of 3.0 Earth radii or greater after correlating over 400 million models with more than 1,000-hours of observations (see Fig. 1). For a CBP size of 2.5 Earth radii, about 80% of any transiting CBP systems would have been detected (Deeg et al. 1998; Doyle et al. 2000).

Fig. 1 A light curve from CM Draconis showing a possible planetary transit feature (solid line) along with the model light curve (dashed line) generated by the transit detection algorithm (TDA). In the left panel is a miss-fit of the orbital phase of the model, while in the right panel is a good model fit to the possible transit signal. (Based on Doyle et al. 2000)

68

L. R. Doyle and H. J. Deeg

Background Circumbinary planets were still hypothetical until the first years of the twentyfirst century, although an early candidate existed as early as 1993. However, long before the discovery of the first extrasolar planets, their perceived strangeness led to speculation about their actual ability to form and if so, about their possible nature – in both science and science fiction. Fiction, as usual, preceded science, with the most famous fictional CBP system certainly being the planet “Tatooine” from the movie series Star Wars. The desert planet Tatooine, which circled two solarlike suns, provided the opening setting for George Lucas’ original 1977 Star Wars movie, with several more appearances in later sequels of that space opera. Further examples of CBPs in fiction can be found in Wikipedia at https://en.wikipedia.org/ wiki/Circumbinary_planet. The first remarks about planets in binary systems that are closer to the scientific realm might have come from Camille Flammarion (1874, 1884), who wrote “Les étoiles doubles sont donc en réalité des groupes de deux soleils. Ces soleils gravitent l’un autour de l’autre, et il est bien probable, pour ne pas dire certain, qu’autour de chacun de ces foyers une famille de planètes est suspendue” (“Double stars are therefore actually groups of two suns. These suns revolve around each other, and it is very likely, not to say certain, that around each of them there is a system of planets”). While this does apparently refer to planets around each of a binary’s component, Flammarion might have had circumbinary planets in mind when he wrote in 1884: “Quelles merveilleuses années, quelles singulières saisons, quels jours et quelles nuits fantastiques sont le partage des planètes inconnues qui gravitent autour de ces [deux] étoiles colorées” (“What wonderful years, what singular seasons, what days and what fantastic nights are sharing unknown planets which revolve around these [two] colored stars”). Arriving at a modern scientific viewpoint, the presence of CBPs was not entirely unexpected, due to the known existence of circumbinary (possibly protoplanetary) dust discs. For short-periodic evolved binary systems, evidence for such discs had been accumulating since the 1970s (e.g., the V471 Tau system, Paczynski 1976), with some systems also giving direct observational evidence from spectroscopy (e.g., epsilon Aurigae, see Castelli 1977; Beta Lyrae, see Kondo et al. 1983). Direct evidence for such circumbinary envelopes was later detected in images taken by the Hubble Space Telescope (HST) – the circumbinary disc around the GG Tau system being the best example (Krist et al. 2002; McCabe et al. 2002, see also Fig. 2). Although it was argued much later (Nelson and Marzari 2016) that GG Tau’s disc is not a suitable place for planet formation, these images certainly raised the expectation that CBPs might actually be able to form. Obviously, it only made sense to speculate seriously about the existence of CBPs if it could be demonstrated that such planetary systems could accrete from circumbinary discs. The first investigations to directly focus on CBP formation were investigations into their orbital stability, a field which had evolved from more general considerations of the orbital stability of the three-body problem (e.g., Black 1982; Pendleton and Black 1983). The pioneering work by Rudolf

5 The Way to Circumbinary Planets

69

Fig. 2 The central wavelengths referred to are in millimeters: 1.03 mm, 1.55 mm, and 1.9 mm, respectively

Dvorak (1986) introduced the still-used nomenclature of “P-type” (planet-type) orbits for CBPs, and “S-type” (satellite-type) orbits for objects orbiting only one of the two wide binary components (the “circumstellar” planets). In the same paper, Dvorak established that a CBP will have a stable orbit if its distance from the common barycenter between the EB and the CBP is more than 2.4 times the distance between the two (equal-mass) binary components (with a slight dependence on the eccentricity). Further works along this line can be found in Dvorak et al. (1989, with more details on the unstable region between S- and P-type orbits). Holman and Wiegert (1999) further established the stability criteria for CBPs given different mass ratios of the stellar components of the binary system, while Pilat-Lohinger et al. (2003) investigated the stability of CBPs for different orbital inclinations. Doolin and Blundell (2011) considered stability for both various binary mass ratios and varying eccentricities, and Morais and Giuppone (2012) considered the orbital stability of both prograde and retrograde planets. Chavez et al. (2014) considered the long-term stability of some of the CBPs found by the Kepler mission. (See also  Chap. 141, “Habitability of Planets in Binary Star Systems” for a more thorough discussion of CBP stability.)

Circumbinary Planet Detection Methods Eclipse Timing Variation Based on the Light Time Effect Similar to an earlier pulsar CBP candidate (Thorsett et al. 1993), in addition to transits, earlier researchers working on evolved short-periodic binaries had also realized that orbiting third bodies could be detectable from induced slight periodic variations in the stellar binary eclipse minima times through the light time or Rømer effect (LTE). This effect had been known since the late seventeenth century from Rømer’s timing of observations of the Jovian satellites (mainly Io) as they were occulted by Jupiter. In this way Rømer first determined the speed of light in 1676 (as reported by Huygens in Huygens 1677).

70

L. R. Doyle and H. J. Deeg

In the case for the application of this method to the detection of CBPs, one deduces a variation in the distance of the EB toward and away from the observer by a periodic variation in the times of eclipses caused by an offset of the EB system by the CBP, although the overall period of the EB is not changed (as opposed to the case with the dynamical detection method, which we discuss below). In the case of a CBP, the periodic timing variations of the eclipses of the EB would be induced from the motion of the binary around a barycenter between it and a third body (i.e., third star or CBP) and the consequent reduced or increased light travel over that displacement distance (distance to the barycenter). However, such binary eclipse time variations usually require measurement precisions of seconds or smaller in order to detect planetary mass third bodies. Only with the development of fast-reading CCDs and the ready availability of precise time standards did such precision become viable (again, with the exception of pulsars who provided their own intrinsically precise “clocks”). A candidate CBP around the pulsar white dwarf binary PSR B-1620-26 in the globular cluster M4 was proposed as one of two possibilities in 1993 (Thorsett et al. 1993; Backer et al. 1993). Secondary variations in the pulsar timing indicated that the system was orbited by either a third star with a semimajor axis of 50 AU or a CBP with a semimajor axis of 10 AUs. In 2003 a third body in circumbinary orbit was characterized – a 2.5 Jovian mass CBP orbiting with a semimajor axis of 23 AU from the EB. Thus this system was confirmed as a CBP a decade after its first announcement (Sigurdsson et al. 2003). Thus, overall, it was really not until the twenty-first century that the first claims for the detection of CBPs from timing measurements could be firmly established. Thus, similar to pulsars but not as accurate, the eclipses of an EB system can also be thought of as its own “clock” in which the mutual eclipses are timed to determine if there are any variations in the actual observed (O) compared to the calculated (C) times of the eclipses. Measured variations in the O-C times can also alternatively be caused by such things as stellar mass transfer (in the case of contact or semi-contact EB systems), star spot activity (and the related Applegate effect), dynamic gravitational interaction of a third body with the individual binary system components, or effects of a third body displacing the EB around the EB-CBP barycenter – the latter being the LTE. As mentioned, in the LTE, the observed times of the stellar eclipses are periodically offset toward and away from the observer by the light travel time across the EB-CBP barycenter. A key to this effect is that the O–C timing of the eclipses of the EB are periodic – eclipses occur on time, too late, on time, too soon, on time, etc., in a periodic manner – while the period of the EB remains (over an entire cycle) essentially constant, that is, the times between secondary and primary eclipses do not change (disregarding precession, etc.). It is clear that this method works best for long-period CBPs because they will produce the largest barycenter displacement between the EB and the CBP for any given CBP mass. One can also see that this method will be most effective for small stellar mass EB systems, as mentioned, since they will be displaced by the largest amount about the EB-CBP barycenter. The expected light time effect can be approximated by:

5 The Way to Circumbinary Planets

O C D

Mp Mp C MA C MB

71



Pp PAB

2=3 

GMA 2 PAB c 3

1=3 sin i;

(1)

where O–C is the “Observed- Calculated” difference in the time of an eclipse between a strictly periodic (calculated) one and the observed one, Mp is the CBPs mass, MA is the mass the primary (usually higher-mass) star, MB is the mass of the secondary star, Pp is the period of the CBP, PAB is the EB period, i is the inclination of the planet’s orbital plane and G and c are the universal gravitational constant and the speed of light, respectively (e.g., Deeg et al. 2008, Eq. 10). As can be seen from Eq. 1, the detectability of the O–C variations increases with the increase in the planetary orbital period, and so the CBP system must usually be observed for a fairly long period of time in most cases to be able to detect any periodicities in the variations in the O–C timings at all. The periodic offset of the timing can also be seen as directly proportional to the mass of the CBP, as well, with stellar third masses being fairly easily detectable. Planetary transits for such systems, of course, also become less likely to align with the EB orbital plane as the planetary orbital period increases, so that these two methods, – LTE and transits – for the detection of CBPs, are complimentary. Unlike transits, however, the LTE method only detects a “projected” mass of the displacing third body: Mp sin i (again, as can be seen in Eq. 1). In addition to the PSR B-1620-26 system discussed above, the other current CBP systems detected using the LTE effect are: – DP Leonis which is a 6.1 Jovian-mass CBP in a 28-year period around a white dwarf/Roche lobe-filling binary pair. (Beuermann et al. 2011). – NN Serpentis which is a 6.9 M sin i Jovian-mass CBP with a 15.5-year with 15.5year period and a 2.2 M sin i Jovian-mass CBP with a 7.7-year period around an M4 dwarf and white dwarf EB. (Beuermann et al. 2010). – NY Virginis which is an sdB C M-dwarf system with a CBP of 2.8 Jovian mass with an 8.2-year orbit and a CBP of 4.5 Jovian masses with 27-year orbital period. (Lee et al. 2014). – RR Caeli which is a 4.2 Jovian-mass CBP, at 5.3 AU distance from the EB, which is an M-dwarf/white dwarf stellar pair. (Qian et al. 2012). – MXB 1658-298 is a compact X-ray binary with a binary orbital period of about 7.1 h. The residual O-Cs indicate a 20.5–26.9 Jovian-mass object to exist around it with a period of 760 days (Jain et al. 2017). Whether this can be called a CBP may be a question of the indicated mass, the canonical upper limit for a planet being below about 13 Jovian masses. If it is a CBP, then it would be an important counter-example to CBPs not found around short-period EB systems. We refer also to a list of 236 candidate three-body systems that have been identified from applying the LTE to precise eclipse minimum timings of binaries in the Kepler sample (Conroy et al. 2014).

72

L. R. Doyle and H. J. Deeg

The currently known CBPs detected by the LTE are all cases of binaries with relatively long periods orbiting short-periodic evolved binary systems. To some extent this is due to the predisposition of the LTE method to detect CBPs with such parameters. However, the CBPs currently known from the LTE may also pertain to a fundamentally different planet population, which might have arisen as a consequence of the evolution of the central binary. Further details of the CBPs detected by the LTE method are reviewed in  Chap. 127, “Circumbinary Planets Around Evolved Stars” in this handbook.

Transits As of this writing, the most successful CBP detection method has been the transit method (see  Chap. 31, “Transit Photometry as an Exoplanet Discovery Method” for an introduction), with ten CBPs having been detected by their transits across EB systems by the Kepler spacecraft (see  Chap. 128, “Two Suns in the Sky: The Kepler Circumbinary Planets” for a more thorough discussion of the CBPs discovered by Kepler). Contrary to transits across single stars (e.g., Hale and Doyle 1994), transits from CBPs had been predicted (Jenkins et al. 1996) to produce unique brightness variations with patterns that depends principally on the phase of the central binary during the planet transit. These transits are semi-periodic, with characteristic transit timing variations; for details see Armstrong et al. (2011). Among the future space missions, from PLATO we may expect that the sample of transiting CBPs gets several times larger (Rauer et al. 2014), given its much larger field of view and temporal coverages that are similar to those of Kepler. The TESS mission might also lead to the discovery of some transiting CBPs, mainly in the polar zones where its survey coverage will be about 1 year, while for most of the sky, the coverage by TESS of only 28 days is too short for efficient CBP discovery. Insufficient length of coverage, of 156 days or less (depending on the survey field), is also the principal cause for the non-detection of CBPs from the CoRoT mission, which surveyed about the same number of EBs as Kepler (Klagyivik et al. 2017). In the case of the first CBP detected by transits, Kepler-16b transited the stellar disc of Kepler-16A, a K-type main-sequence star, followed by a transit across the stellar disc of Kepler-16B, an M-type main-sequence star. Kepler-16A and B orbit around each other with a period of about 41 days, while Kepler-16b orbits around both with a period of about 229 days (Doyle et al. 2011). The next time Kepler-16b came within the transit “window,” the EB had changed configuration as projected toward the Earth, and the CBP transited first Kepler-16B and then Kepler-16A. During the third transit pass of Kepler-16b across our line of sight, the EB system again had switched its projected configuration on the sky so that it was again in the original configuration and the primary transit across Kepler16A preceded the secondary transit across Kepler-16B again (see in Fig. 3). This confirmed the circumbinary nature of the events since such a switching of the order of the CBP transits – primary transit, secondary transit, secondary transit, primary transit, primary transit, and secondary transit – would not lend itself to any other

0.85

0.90

0.95

1.00

0.85

0.90

0.95

170.9 171.0 171.1 171.2

Star B eclipses Star A

0

0.985

0.990

0.995

1.000

1.005

100

0.980

0.985

0.990

0.995

1.000

1.005

400

425.0 425.1 425.2 425.3 425.4

Planet b transits Star A

300

Time [BJD – 2,455,000]

273.1 273.2 273.3 273.4 273.5

Star A occults Star B

200

0.9990

0.9992

0.9994

0.9996

0.9998

1.0000

1.0002

–18.6 –18.5 –18.4 –18.3 –18.2

Planet b transits Star B

500

Fig. 3 Light curve of the Kepler-16AB-b system – transits of the CBP across the brighter binary component A are in green, while CBP transits across the secondary component B are in red. The order of these transits can be seen (as indicated by the colored arrows at the base of the light curve) to reverse in the second pair of transits (red then green), compared to the first and last pair (green then red). Such a reversal is impossible to explain in terms of any other known astrophysical event (i.e., a background EB could not produce such a series of events in such a switching order), and so this allowed the confirmation that these events were caused by a circumbinary body

Relative flux

1.00

5 The Way to Circumbinary Planets 73

74

L. R. Doyle and H. J. Deeg

astrophysical explanation. It is of note that the large majority of the CBPs discovered by Kepler orbit close to their inner stability limit. The diversity of the Kepler CBPs was surprising and is expected to increase with future discoveries.

Imaging Direct imaging has long been a goal of extrasolar planet detection, in part because it would allow spectra to be taken of planetary bodies with, hopefully, little interference from the parent stars (see  Chap. 34, “Direct Imaging as a Detection Technique for Exoplanets”). (Difference spectra can also be taken during transits.) Similar to planets around single stars that were detected by imaging, at present this method permits only the detection of CBPs that are relatively distant from their central star, on the order of 10–1000 AU. Also, the imaged CBPs are all in young systems, where the planets still have an elevated temperature from their formation process and are notable emitters of thermal radiation, which permits their detection. The direct detection of temperate CBPs in reflected light is expected to become possible only with the operation of space based coronographic or interferometric imagers (see  Chap. 64, “Future Exoplanet Space Missions: Spectroscopy and Coronographic Imaging”). With respect to CBPs, we include here a couple of selected examples of systems that have been imaged. – ROXs 42b is a 9 jovian-mass sub-deuterium-burning CBP around the M-dwarf binary ROXs 42 detected by direct imaging (Currie et al. 2014). The CBP is at a projected distance of about 157 AU from the central binary. Infrared photometry and K-band spectroscopy indicate that the CBP has a dusty atmosphere. – Ross 458 AB (Burgasser et al. 2010, also known as DT Virginis) is an eclipsing binary system consisting of two red dwarf stars, with an 8.5 Jovian-mass CBP at a distance of 1100 AU and an age of 150–800 M years. In the near future, many more CBP systems may be expect to be imaged with the consequent ability to take spectra of the CBPs directly, thus allowing the important determination of CBP atmospheres.

Gravitational Lensing Due to the warping of space-time, as explained by general relativity, a planet can momentarily (hours) focus light onto Earth in a highly columnated way. Distances are usually significant (kiloparsecs) and so CBPs can be sampled in this way only at great distances. In 2016 OGLE-2007-BLG-349L(AB)c became the first CBP found by microlensing (Bennett et al. 2016). It is a planet of mass mc D 80 ˙ 13 M˚ , orbiting a pair of M-dwarfs with masses of MA D 0.41 ˙ 0.07 Msolar masses

5 The Way to Circumbinary Planets

75

and MB D 0.30 ˙ 0.07 Msolar masses . The distance from the planet to the system barycenter is about 40 times the distance between the binary components, so unlike most of the CBPs found by Kepler, this planet does not orbit near the orbital stability limit. The preceding methods are those through which CBPs have been discovered to date. But there are additional methods by which CBPs might also be discovered or that might be complimentary to the investigation of known CBPs. Besides those described below, we also note that the detection of CBPs by astrometry in data from the GAIA mission has been evaluated by Sahlmann et al. (2015)

Eclipse Timing Variation from Dynamical Effects While no CBPs have yet been initially discovered using the dynamical detection method, this effect was essential to confirming the presence of CBPs after detection by their transits (e.g., Kepler-16b). If the orbital period of the CBP is not long, then the offset of the EB around the EB-CBP mutual barycenter – and consequent light travel time over the offset barycenter interval – may not be sufficient to detect a periodic variation in eclipse times (the LTE). However, the closer the CBP is to the EB, the more gravitational pull it does exert on the stellar components directly. Unlike the LTE, this “dynamical” effect will change the period of the EB by direct interaction with the components of the EB. Also, this interaction is not a function of the sine of the CBP’s orbital inclination. The dynamical effect can be characterized by the equation: Mp 3 O C D 8 Mp C MA C MB

PAB 2 Pp

! ;

(2)

with the meaning of the variables the same as in Eq. 1 (Borkovits et al. 2003). Here, however, we can see that has the period of the CBP gets shorter, the O–C values increase, in the opposite direction of the LTE. Let us illustrate with an EB system in which the stellar mass-ratio of the components are about one-third. Let us further assume that the orbital period of the CBP is less than a couple of hundred days. The offset of the whole EB system, due to the sub-year period of the CBP, will not offset the EB system around the mutual EB-CBP barycenter enough to be as measurable as the direct dynamical effect on the different stellar components. The barycenter between the two stars will be three times farther from the less massive (secondary) star than from the more massive (primary) star. Thus the gravitational tug on the secondary star by the CBP, each time it orbits the EB, will be greater than on the primary star since it is closer to the secondary star, causing a tidal change in the period of the EB itself because of the greater effect on the (usually less massive) secondary star, as well. Thus the dynamical effect begins to dominate in measurability with decreasing orbital period of the CBP, while the LTE begins to diminish in measurability (see Fig. 4). Also, since the gravitational effect of the

76

L. R. Doyle and H. J. Deeg

Fig. 4 Comparison of changes in the binary period between the dynamical and the LTE effect. The dynamical effect is most sensitive to the detection of shorter periodic CBPs, while the LTE is only measurable for longer periodic CBPs. The dynamical effect was used to confirm the first directly detected CBP, Kepler-16b

third body (e.g., circumbinary planet) is not dependent upon the line of sight (i.e., it is not a radial velocity or LTE “projected” sine-function effect), the direct mass of the CBP can be derived from these periodic changes in the O–C values of the EB. An example of the utility of this method was demonstrated with the confirmation of the first direct detection of a CBP, Kepler-16b. While examining the LTE, which had constrained the mass of a third body to less than that of about a T-type brown dwarf, the dynamical effect of the third body – especially on the drift in the timing of the secondary eclipses – was able to tightly confirm the mass of Kepler-16b as about that of Saturn. A similar detection of the dynamical effect has also been reported for Kepler 34 and 35 (Welsh et al. 2012), and it is to be expected that this method will become used for primary detection of CBPs. While the above discussion refers to timing variations from longer-term orbital evolution, shorter-term timing variations on time scales of the orbital periods may also occur; for these we refer to Borkovits et al. (2011) and references therein.

Radial Velocity Variations In contrast to the detection of planets around single stars, where radial velocity (RV) measurements are one of the dominant detection methods (see  Chap. 30, “Radial Velocities as an Exoplanet Discovery Method”), RVs have played only a minor role in the discovery of CBPs to date. This is due to the difficulties in detecting RV variations caused by a CBP in the presence of the much larger RV variations from the stellar motion in the binary system. Assuming a binary with equally massive components and a total mass of MAB , and with both binary and CBP being on orbits with identical inclination and eccentricity, the ratio of RV amplitudes due to the orbiting CBP and due to the stellar orbits is given by:

5 The Way to Circumbinary Planets

1=3  Kp =KAB D Pp =PAB 2 Mp =MAB ;

77

(3)

where Kp is the RV amplitude of either of the stellar components due to the presence of the planet and KAB is the RV amplitude of either stellar component due to the other component. The dominant factor in Eq. 3 is clearly the mass ratio. For a binary consisting of two solar-mass components, orbited by a Jupiter-mass CBP with small period ratios, the ratio of RVs is then about 1/2000. Even for M-star binaries orbited by massive CBPs of several Jovian masses, the above ratio would not be more than about 1/100. The extraction of a CBP’s RV signal, of typically 1–100 m/s, is therefore very difficult in the presence of the much larger signal from the central binary, which would have typical RV amplitudes of 10–100 km/s. Notwithstanding this obstacle, serious efforts have been made to use RV variations for the detection of CBPs. Their principal advocate has been Konacki et al. (2009, 2012) with their TATOOINE search. Since 2003, they employed a specific radial velocity technique using several spectrographs with iodine absorption cells (such as the Keck/HIRES instrument or the TNG/Sarg instrument), surveying a set of double-lined (SB2) binaries with a precision of up to 2 m/s. However, to date no detection of a CBP has been claimed from this survey or this technique. The only RV-discovered system for which the existence of a CBP had been proposed is HD 202206, presented by Correia et al. (2005) in a paper entitled “A pair of planets around HD 202206 or a circumbinary planet?” They describe the system with these components: A solar-like G6V central star, a b component with a mass of m sin i 17.4Mjup , and a c component with a mass of m sin i 2.44Mjup , with both b and c on markedly eccentric orbits. Hence, it depends on whether component b is a planet or a brown dwarf (see  Chap. 29, “Definition of Exoplanets and Brown Dwarfs”) in concluding if planet c should be considered as a CBP or not. This system also presents serious issues on its long-term stability, which favors a 5:1 mean motion resonance between the two orbiting bodies (Correia et al. 2005; Couetdic et al. 2010). However, Benedict and Harrison (2017) showed from astrometric parallaxes that HD 202206 is a nearly face-on system, with small values of sin i and hence, much larger orbital masses of about 94 Mjup and 18Mjup . Therefore they conclude that HD 202206 has neither a pair of planets nor a CBP and rather, describe it as a G8V C M6V binary orbited by a brown dwarf third body.

Reflected Light or Eclipse Echos The eclipse echo method proposes to detect the “echos” of binary eclipses that are reflected (Jenkins and Doyle 2003) by the CBPs in orbit around them (Fig. 5), with the resultant delay times allowing the derivation of several of the CBP orbital parameters (Deeg and Doyle 2011). The delays of the eclipse echos are given by the difference in the period of the echos (Pre ) versus the period of the CBP, given by: 1=Pre D 1=PAB ˙ 1=Pp

(4)

78

L. R. Doyle and H. J. Deeg

Reflection (ppm)

Total flux

1.2 1.0 0.8 0.6

Pb

0.4 0.2 0.0 2.0

Pre

15 10 5 0

0

2

4

6

8

10

t (day)

Fig. 5 Simulated light curve from a system with a binary of 1-day orbital period (Pb ), orbited by a prograde planet with a period of Pp D 10d, both with an inclination of 90ı . The upper panel shows the total observed flux, which is dominated by the eclipses observed directly from the binary. The lower panel shows the reflected light from the planet. Note that its amplitude is on the order of ppm. The general shape of this sinusoidal light curve is due largely to the planetary phase function, upon which the eclipse echos (positions indicated by vertical red lines) are imprinted, with a periodicity of Pre D 1.11d. (From Deeg and Doyle 2011; see also for more details)

where the ˙ sign indicates prograde () or retrograde (C) planetary orbits. Considering a minimum in the planet/binary period ratio of about 3.7, we may constrain Pre to values fairly close to the binary period, namely, 0.78PAB  Pre  1.37PAB . If we consider only prograde orbits, then only echo periods of PAB < Pre  1.37PAB need to be considered. This restricted range of possible periods should facilitate any echo detection attempts. We note that the echos suffer also an LTE effect, which is however very small (on the order of seconds for potentially detectable configurations) compared to the delay from the period difference. This method is most sensitive to short-periodic CBPs around binaries that themselves have very short periods. The echo method could, however, detect binaries and CBPs over a wide range of inclinations, and if successful, the orbital characteristics of the CBP system could be obtained solely from the photometric light curve. However, shortly after the development of this method, it became evident from searches for transiting CBPs in Kepler data that CBPs around binaries with periods of less than 7 days are rare or even absent. This method has therefore not yet been applied to the detection of CBPs. Martin et al. (2015) and Hamers et al. (2016) argue that short-periodic binaries form in triple systems and are followed by a dynamical evolution that either ejects their planets or moves them to wide and potentially inclined orbits (see also  Chap. 141, “Habitability of Planets in Binary Star Systems”). Whether short-period CBPs really exist around short-period EBs is however still an open issue, and the eclipse echo method described here might aid in the resolution of this issue.

5 The Way to Circumbinary Planets

79

Habitability of M-Dwarfs and Circumbinary Planets The habitable zones (HZs) around CBPs differ from those around single stars mainly by the fact that the CBP, in a given orbit, nevertheless receives quite variable stellar flux because of the rapidly changing distances between the stellar components and the planet, as well as the regular dimming due to stellar eclipses for CBP in or near the EB orbital plane (see, e.g., Welsh et al. 2012 for a calculation of the changing insolation reaching Kepler 34b and 35b). For a more detailed evaluation of the habitability of CBPs, we refer to Haghighipour and Kaltenegger (2013) and  Chap. 141, “Habitability of Planets in Binary Star Systems”. With the beginning of the first transit searches for CBPs by the TEP network, interest in the habitability of both M-star planets and the CBP HZ arose. Although about 75% of all stars in the Milky Way galaxy are M-dwarf stars, it was, at the time, thought that planets within the region receiving Earth-equivalent flux could not be habitable because planets this close to their stars would be tidally locked,

Fig. 6 The observations of the TEP network led to the modeling of the potential habitability of M-dwarf planets using 3-D climate models. An artist’s conception (Lynette Cook, reproduced with her permission) is shown of a tidally locked “Earthlike” planet orbiting within the HZ around the CM Draconis eclipsing binary system. CM Draconis was the first binary star system searched for circumbinary planets within its habitable zone

80

L. R. Doyle and H. J. Deeg

i.e., in synchronous rotation. This is because, while the insolation increases with the inverse square of the planet’s distance from the star(s), the tidal effects scale as the inverse cube of this distance. However, such synchronously rotating Earthlike planets were shown, using 3-D atmospheric circulation models, (Haberle et al. 1996) to allow a low thermal atmospheric gradient with a reasonable increase of additional greenhouse gases (e.g., 0.1 bar of carbon dioxide). These synchronously rotating planets were also shown to have some unusual weather patterns that propagate from the substellar point (where a continuous hurricane rages) to the anti-substellar equatorial ice cap (Heath et al. 1999, Joshi et al. 1997, Joshi 2003). A planetary hydrological cycle would also be possible with temperate zones at mid-substellar “latitudes,” (i.e., latitudes defined as 90 degrees from the rotational poles; Heath et al. 1999). In Fig. 6 artist Lynette Cook illustrates a hypothetical tidally locked Earthlike planet in circumbinary orbit around the CM Draconis system. We see that the weather patterns propagate not across lines of latitude but in streams from the substellar point, and there is an equatorial ice cap at the anti-substellar point. In the case of CM Draconis, the backside of the Earthlike planet is illumined by a white dwarf proper motion companion that travels in a highly eccentric galactic orbit with CM Draconis and allows a more familiar blue light scattering off the

Fig. 7 The full set of CBPs, known in early 2018, is plotted by planetary orbital period versus mutual binary orbital period. The detection methods, as well as the region of unstable CBP orbits, are indicated. The CBPs detected by imaging all have very long periods with rather large uncertainties and are outside of the plotted period ranges. Data is from the NASA Exoplanet Archive, selecting all planets with a circumbinary flag of 1

5 The Way to Circumbinary Planets

81

planet’s hypothetical seas. The galactic orbit of CM Draconis indicates an age older than our Solar System, and this is indicated by a highly impact-cratered foreground asteroid in the artist’s conception.

Conclusions For various reasons there had remained doubts as to the existence of CBPs until the discovery of Kepler-16b. The system was nicknamed “Tatooine” after one of the planets in the Star Wars movie because – like the sunset scene in the movie – it was the first CBP system where the binary pair could be seen as approximately solar-sized discs in the sky from the perspective of the planet. (George Lucas, the producer of Star Wars, gave unofficial permission to use the nickname “Tatooine” for this system, and the name stuck.) The CBPs detected to date by various methods are shown in Fig. 7. With the current discoveries of CBPs, it has been estimated that there must be at least hundreds of millions of such planetary systems in the Milky Way galaxy (for a more detailed discussion, see  Chap. 141, “Habitability of Planets in Binary Star Systems”). CBPs are among the most precisely measured planetary systems available, giving insights into star-planet spectral type and size distributions and putting constraints on star and planet formation with, for example, a determination in some cases of the rotation axis of the stars with the orbital plane of the CBP. Outstanding questions include why many CBPs are near their orbital stability limit, why CBPs seem to have intermediate masses (i.e., Neptune to Saturn masses; see Armstrong et al. 2014), and why close-orbit CBPs have not been found around shortperiodic (contact and semi-contact) binary systems. The stability of “misaligned” CBPs with orbits that are strongly inclined to the binary orbit has been demonstrated (Martin and Triaud 2014, 2016), but their existence, for which several detection methods such as eclipse timing variations (Borkovits et al. 2011) or occasional transits (Martin 2017) are possible, has yet to be demonstrated. We also need to determine if there is a continuous CBP population between the CBPs already discovered by the various methods and, if not, which groups relate to different populations. In a historic context, the study of CBPs has only just begun.

Cross-References  Circumbinary Planets Around Evolved Stars  Direct Imaging as a Detection Technique for Exoplanets  Habitability of Planets in Binary Star Systems  Populations of Planets in Multiple Star Systems  Radial Velocities as an Exoplanet Discovery Method  Space Missions for Exoplanet Research: Overview and Introduction  Space Missions for Exoplanet Science: Kepler/K2  Space Missions for Exoplanet Science: PLATO

82

L. R. Doyle and H. J. Deeg

 Transit Photometry as an Exoplanet Discovery Method  Transit-Timing and Duration Variations for the Discovery and Characterization of

Exoplanets  Two Suns in the Sky: The Kepler Circumbinary Planets Acknowledgements HD acknowledges support by grant ESP2015-65712-C5-4-R of the Spanish Secretary of State for R&D&i (MINECO). This contribution has benefited from the use of the NASA Exoplanet Archive and the Extrasolar Planets Encyclopaedia, and the authors acknowledge the people behind these tools.

References Armstrong D, Martin DV, Brown G et al (2011) Placing limits on the transit timing variations of circumbinary exoplanets. MNRAS 434:3047 Armstrong DJ, Osborn H, Brown DJA et al (2014) On the abundance of circumbinary planets. MNRAS 444:1873 Backer DC, Foster RS, Sallmen S (1993) A second companion of the millisecond pulsar 1620 – 26. Nature 365:817 Benedict GF, Harrison TE (2017) HD 202206: a circumbinary brown dwarf system. AJ 153:258 Bennett DP, Rhie SH, Udalski A et al (2016) The first circumbinary planet found by microlensing: OGLE-2007-BLG-349L(AB)c. AJ 152:125 Beuermann K, Hessman FV, Dreizler S et al (2010) Two planets orbiting the recently formed postcommon envelope binary NN Serpentis. A&A 521:L60 Beuermann K, Buhlmann J, Diese J et al (2011) The giant planet orbiting the cataclysmic binary DP Leonis. A&A 526:A53 Black DC (1982) A simple criterion for determining the dynamical stability of three-body systems. AJ 87:1333 Borkovits TB, Erdi E, Forgács-Dajka E, Kovács T (2003) On the detectability of long period perturbations in close hierarchical triple stellar systems. A&A 398:1091–1102 Borkovits TB, Csizmadia SZ, Forgács-Dajka E, Hegedus T (2011) Transit timing variations in eccentric hierarchical triple exoplanetary systems I. Perturbations on the time scale of the orbital period of the perturber. A&A 528:A53 Borucki WJ, Summers AL (1984) The photometric method of detecting other planetary systems. Icarus 58:121 Burgasser AJ, Simcoe RA, Bochanski JJ et al (2010) Clouds in the coldest brown dwarfs: fire spectroscopy of Ross 458C. ApJ 725:1405 Castelli F (1977) The spectrum of e aurigae outside eclipse. Ap&SS 49:179 Chavez CE, Georgakarakos N, Prodan S et al (2014) A dynamical stability study of Kepler circumbinary planetary systems with one planet. MNRAS 446:1283 Conroy KE, Prša A, Stassun KG (2014) Kepler eclipsing binary stars. IV. Precise eclipse times for close binaries and identification of candidate three-body systems. AJ 147:45 Correia ACM, Udry S, Mayor M et al (2005) The CORALIE survey for southern extra-solar planets. XIII. A pair of planets around HD 202206 or a circumbinary planet? A&A 440:751 Couetdic J, Laskar J, Correia ACM, Mayor M, Udry S (2010) Dynamical stability analysis of the HD 202206 system and constraints to the planetary orbits. A&A 519:10 Currie T, Burrows A, Daemgen S (2014) A first- look atmospheric modeling study of the young directly imaged planet-mass companion, ROXs 42Bb. ApJ 787:104 Deeg HJ, Doyle LR (2011) Reflected eclipses on circumbinary planets. In: Bouchy F et al (eds) Detection and dynamics of transiting exoplanets, St. Michel l’Observatoire, France. EPJ Web of conferences, 11, id.05005 Deeg HJ, Tingley B (2017) TEE, an estimator for the precision of eclipse and transit minimum times. A&A 599:93

5 The Way to Circumbinary Planets

83

Deeg HJ, Doyle LR, Kozhevnikov VP et al (1998) A photometric search for transits of extrasolar planets: observations and photometric analysis of CM Draconis. A&A 338:479 Deeg HJ, Doyle LR, Kozhevnikov VP (2000) A search for jovian-mass planets around CM Draconis using eclipse minima timing. A&A 358:L5 Deeg HJ, Ocaña B, Kozhevnikov VP et al (2008) Extrasolar planet detection by binary stellar eclipse timing: evidence for a third body around CM Draconis. A&A 480:563 Doolin S, Blundell KM (2011) The dynamics and stability of circumbinary orbits. MNRAS 418:2656 Doyle LR, Deeg HJ (2004) Timing detection of eclipsing binary planets and transiting extrasolar moons. In: Norris RP, Stootman FH (eds) Bioastronomy 2002: life among the stars, IAU symposium 213:80 Doyle LR, Deeg HJ, Jenkins JM et al (1998) Detectability of Jupiter-to-brown-dwarf-mass companions around small eclipsing binary systems. In Rebolo R et al (eds) Brown dwarfs and extrasolar planets, ASP conference series 134:224 Doyle LR, Deeg HJ, Kozhevnikov VP et al (2000) Observational limits on terrestrial-sized inner planets around the CM Draconis system using the photometric transit method with a matchedfilter algorithm. ApJ 535:338 Doyle LR, Carter JA, Fabrycky DC et al (2011) Kepler-16: a transiting Circumbinary planet. Science 333:1602 Dvorak R (1986) Critical orbits in the elliptic restricted three-body problem. A&A 167:379 Dvorak R, Froeschle C, Froeschle C (1989) Stability of outer planetary orbits (P-types) in binaries. A&A 226:335 Flammarion C (1874) Les Étoiles doubles. La Nat Revue Sci 37:161 Flammarion C (1884) Les Terres du Ciel; voyage astronomique sur les autres mondes et description des conditions actuelles de la vie sur les diverses planètes du système solaire; (11th edition), p 757 Goodricke J (1784) On the periods of the changes of light in the star Algol. In a letter from John Goodricke, Esq. to the Rev. Anthony Shepherd, D.D.F.R.S. Professor of Astronomy at Cambridge. Philos Trans R Soc Lond 74(0):287 Haghighipour N, Kaltenegger L (2013) Calculating the habitable zone of binary star systems. II P-type binaries. ApJ 777:166 Hale A, Doyle LR (1994) The photometric detection of extrasolar planets revisited. Astroph & Space Sci 212:335 Hamers AS, Perets HB, Portegies Zwart SF (2016) A triple origin for the lack of tight coplanar circumbinary planets around short-period binaries. MNRAS 455:3180 Heath M, Doyle LR, Joshi MM, Haberle R (1999) Habitablility of planets around M-dwarf stars. Orig Life 29:405 Holman M, Wiegert PA (1999) Long-term stability of planets in binary systems. AJ 117:621 Huygens C (1677) Lettre Nı 2104. In: Bosscha J (ed) Œuvres complètes de Christiaan Huygens. Tome VIII: Correspondance 1676–1684. Martinus Nijhoff, The Hague (published 1899), p 32 (in Latin) Jain C, Paul B, Sharma R, Jaleel A, Dutta A (2017) Indication of a massive circumbinary planet orbiting the low-mass X-ray binary MXB 1658–298. MNRAS 468:L118 Jenkins JM, Doyle LR (2003) Detecting reflected light from close-in Giant planets using spacebased photometers. ApJ 595:429 Jenkins JM, Doyle LR, Cullers K (1996) A matched-filter method for ground-based sub- noise detection of extrasolar planets in eclipsing binaries: application to CM Draconis. Icarus 119:244 Joshi MM, Haberle RM, Reynolds RT (1997) Simulations of the atmospheres of synchronously rotating terrestrial planets orbiting M dwarfs: conditions for atmospheric collapse and the implications for habitability. Icarus 129:450 Joshi MM (2003) Climate model studies of synchronously rotating planets. Astrobiology 3:415 Klagyivik P, Deeg HJ, Cabrera J, Csizmadia SZ, Almenara JM (2017) Limits to the presence of transiting circumbinary planets in CoRoT data. A&A 602:A117

84

L. R. Doyle and H. J. Deeg

Konacki M, Muterspaugh MW, Kulkarni SR, Hełminiak KG (2009) The radial velocity Tatooine search for Circumbinary planets: planet detection limits for a sample of double-lined binary stars. ApJ 704:513 Konacki M, Sybilski P, Kozłowski SK, Ratajczak M, Hełminiak KG (2012) Circumbinary planets and the SOLARIS project. IAU Symp 282:111 Kondo Y, Parsons SB, Henize KG et al (1983) Skylab ultraviolet stellar spectra: emission lines from the Beta Lyrae system. ApJ 208:468 Krist JE, Stapelfeldt KR, Watson AM (2002) Hubble space telescope/WFPC2 images of the GG Tauri Circumbinary disk. ApJ 570:785 Lee JW, Hinse TC, Youn JH, Han W (2014) The pulsating sdBCM eclipsing system NY Virginis and its circumbinary planets. MNRAS 445:2331 Martin DV (2017) Circumbinary planets II – when transits come and go. MNRAS 465:3235 Martin DV, Triaud AHMJ (2014) Planets transiting non-eclipsing binaries. A&A 570:91 Martin DV, Triaud AHMJ (2015) Circumbinary planets – why they are so likely to transit. MNRAS 449:781 Martin DV, Triaud AHMJ (2016) Kozai-Lidov cycles towards the limit of circumbinary planets. MNRAS 455:46 Martin DV, Mazeh T, Fabrycky DC (2015) No circumbinary planets transiting the tightest Kepler binaries – a possible fingerprint of a third star. MNRAS 453:3554 McCabe C, Duchêne G, Ghez AM (2002) NICMOS images of the GG Tau circumbinary disk. ApJ 575:974 Morais MHM, Giuppone CA (2012) Stability of prograde and retrograde planets in circular binary systems. MNRAS 424:52 Morales JC, Ribas I, Jordi C et al (2009) Absolute properties of the low-mass eclipsing binary CM Draconis. ApJ 691:1400 Nelson A, Marzari F (2016) Dynamics of circumstellar disks. III. The case of GG Tau A. ApJ 93:40 Paczynski B (1976) Common envelope binaries. In: IAU symposium no. 73, p 75 Pendleton YJ, Black DC (1983) Further studies on criteria for the onset of dynamical instability in general three-body systems. AJ 88:1415 Pilat-Lohinger E, Funk B, Dvorak R (2003) Stability limits in double stars: a study of inclined planetary orbits. A&A 400:1085 Qian S-B, Liu L, Zhu L-Y et al (2012) A circumbinary planet in orbit around the short-period white dwarf eclipsing binary RR Cae. MNRAS 422:L24 Rauer H, Catala C, Aerts C et al (2014) The PLATO 2.0 mission. Exp Astron 38:249 Sahlmann J, Triaud AHMJ, Martin DV (2015) Gaia’s potential for the discovery of circumbinary planets. MNRAS 447:287 Schneider J (1994) On the occultations of a binary star by a circum-orbiting dark companion. Planet Space Sci 42:539 Schneider J, Chevreton M (1990) The photometric search for earth-sized extrasolar planets by occultation in binary systems. A&A 232:251 Schneider J, Doyle LR (1995) Ground-based detection of terrestrial extrasolar planets by photometry: the case for CM Draconis. Earth Moon Planet 71:153 Sigurdsson S, Richer H, Hansen BM, Stairs IH, Thorsett SE (2003) A young white dwarf companion to pulsar B1620-26: evidence for early planet formation. Science 301:193 Struve O (1952) Proposal for a project of high precision radial velocity work. The Observatory 72:199 Sybilski PM, Konacki M, Kozłowski SK (2010) Detecting circumbinary planets using eclipse timing of binary stars – numerical simulations. MNRAS 405:657 Thorsett SE, Arzoumanian Z, Taylor JH (1993) PSR B1620-26 – a binary radio pulsar with a planetary companion? ApJ 412:L33 Welsh WF, Orosz JA, Carter JA (2012) Transiting circumbinary planets Kepler-34b and Kepler35b. Nature 481:475

6

The Naming of Extrasolar Planets Frederic V. Hessman

Contents The Names of Stars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Naming Exoplanets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Back to Proper Names . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IAU Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

85 87 91 92 93

Abstract

A short historical introduction to the classical naming of stars and extrasolar planets is given. Keywords

Dynamics · Host stars · Hierarchies

The Names of Stars The naming of astronomical objects has a long and colorful history, and so the names of extrasolar planets (usually called “exoplanets”) – being derived from their host star’s names – share that history. The International Astronomical Union (IAU) is the official curator of star and planet names and naming conventions, but the IAU inherited a very heterogeneous mix of nomenclatures that developed over thousands of years.

F. V. Hessman () Institut für Astrophysik, University of Göttingen, Göttingen, Germany e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_193

85

86

F. V. Hessman

All cultures have given names to the bright planets and stars, usually keeping special names for the planets as moving objects. In modern western culture, one uses the Roman names for the visible planets based on the older Greek names of major gods and goddesses – Mercury, Venus, Mars, Jupiter, and Saturn – but the official names of bright visible stars are a combination of Roman (e.g., Arcturus), Arabic (e.g., Aldebaran, the latinization of the Arabic al dabaran), Greek (e.g., Kornephoros), and even a few artificial names (e.g., Sualocin, the second brightest star in the constellation Delphinus, is an Italian astronomer’s name in reverse order!). As more and more naked-eye stars were identified by their position, it became cumbersome to give all of them an individual name. Johannes Bayer (1572–1625) adopted a naming convention for his Uranometria star charts, most often based on the relative brightness within a constellation: the brightest star was called ˛, the second brightest ˇ, etc., and the genitive form of the constellation was added. Thus, the third brightest object of the constellation Leo with the Arabic name Algieba (latinized from the original al jabhah, the “forehead”) is also called Leonis (usually shortened to Leo). After the invention of the telescope, astronomers quickly ran out of Greek letters, and Latin lower- and uppercase letters were used as well. John Flamsteed (1646–1719) and Jéôme Lalande (1732–1807) adopted a numerical system for the Flamsteed catalogue based on the relative right ascension within a constellation: the first exoplanet around a normal star was discovered in 1995 around the 51st star in Flamsteed’s list for the constellation Pegasus, 51 Pegasi (51 Peg). This system obviously became unwieldy as surveys began to measure the positions of tens of thousands of fainter stars, so one reverted to a naming convention including an abbreviation of the name of the survey and the relatively random catalogue number: the 20794th star in the catalogue of positions created by the Draper Catalogue of Stellar Spectra – named after one of the pioneers of astronomical spectroscopy, Henry Draper (HD), – has the name HD 20794 as well as 82 Eridani. As each new survey was made – including those made at other wavelengths such as radio, infrared, X-rays, and -rays – the number of names for the same object increased. To complicated things further, early observers of variable stars wanted to give these totally unexpected objects their own special names. The pioneer of variable star observers, Friedrich Argelander (1799–1875), didn’t expect to find very many, and so proposed a system related to that of Bayer and Flamsteed, this time using capital Roman instead of Greek letters (Hoffleit 1987). In order to avoid confusion, he started with the relatively unused letter “R”: e.g., R CrB (usually pronounced “R Cor Bor”) is the first variable star discovered in the constellation of Corona Borealis. As the letter “Z” was quickly reached after only nine objects, an arcane system of two letters – “RR,” “RS,” . . . “RZ,” “SS,” “ST,”. . . “ZZ,” “AA,” “AB,”. . . “QZ” – was adopted. For example, a likely exoplanet has been detected around the young variable star TW Hydrae. Of course, eventually even more names were needed, so the variable stars were named by their variable star catalogue number with a “V” prefix and their constellation: e.g., the young variable star V830 Tauri is #830 in

6 The Naming of Extrasolar Planets

87

the list of variable stars in the constellation Taurus and has an orbiting hot Jupiter companion. In modern times, the large number of objects included in catalogues makes even a catalogue number unwieldy – the final Gaia catalogue will contain literally billions of objects – and so many newly catalogued objects are named by the survey and the position on the sky: HD 20794 can also be labeled 2MASS J031955634304112 (“2MASS” is the survey, the letter “J” indicates that an epoch 2000 position follows, consisting of the hour, minute, and decimal parts of the right ascension, 03h 19m 55.63s , and the degrees, arc minutes, and arc seconds of the declination, 43ı 040 ; 11:200 ). While the stars were simply points of light on the sky, these naming systems were sufficient, but when higher angular resolution measurements and/or detailed proper motions became available, some visible stars were recognized actually to consist of two or more stars sharing the same name. Some stars were obviously connected dynamically – binary stars – and even hierarchical systems of objects were discovered (e.g., two or more binary systems orbiting each other). In order to identify the different parts of a common dynamical system, uppercase Roman letters were appended to the system names: e.g., the “star” Sirius actually consists of Sirius A (an A1V dwarf star) and Sirius B (a DA white dwarf star) that can be seen as two separate stars at high angular resolutions. Eclipsing binaries were similarly given uppercase designations, since the eclipse light curves are the immediate sign of the presence of multiple bodies, even if they cannot be seen as separate objects in the sky. For purely historical reasons, spectroscopic components were given lowercase designations – e.g., HD 20794a is the main spectroscopic component in the dynamical system HD 20794. Combinations of suffixes were possible: the secondary spectroscopic component around the primary visual binary component would have the composite suffix “Ab.” Since many visual or eclipsing binaries are also spectroscopic binaries, it became a question of historical chance as to whether the uppercase or lowercase designation was used first. Unfortunately, the naming system for multiple star systems was not meant to express a dynamic hierarchy: if the secondary star in a visual binary itself has a spectroscopic secondary, then the star’s name is formally “Ba”: “B” stands for the visual secondary which is the spectroscopic primary “a” and host of the spectroscopy secondary star “b.” Thus, the additional higher-order members of multiple systems are given names starting with “b” rather than “a.” Note that this naming convention also does not express any explicit dynamical information other than the statistical tendency for the dynamically most obvious planets to be discovered first.

Naming Exoplanets Exoplanets inherited the arcane naming system used for multiple stellar systems. Since the first exoplanets were discovered dynamically using pulsar timing and radial velocity methods and couldn’t be seen directly because of their extreme

88

F. V. Hessman

faintness, it was natural to use lowercase Roman suffixes to designate them. The first exoplanet to be discovered, PSR B1257+12 b (Wolszczan and Frail 1992), was that around the pulsar PSR B1257+12 a (“PSR” is the acronym for a pulsar catalogue, and the rough right ascension and declination are 12h 57m and +12ı ). Starting in 1995, a large number of exoplanet candidates were detected using the radial velocity method, and so a number of objects with a “b” suffix designation were named. As more and more measurements were made, it was even possible to detect planetary systems consisting of the central star (“a”) and whole series of planets named in the order of their discovery. Later, the first direct images of massive sub-stellar and planetary objects were made, leading to the use of the uppercase naming convention: e.g., the object GJ 229 (the 229th object in the Gliese-Jahreiß catalogue of high proper-motion stars) was discovered to contain the M1V red dwarf GJ 229 A and the brown dwarf GJ 220 B (Nakajima et al. 1995). However, the radial velocity method was still the primary means of discovering exoplanets, and so many directly detected extrasolar planets started to carry lowercase rather than capitalized designations in their names, even when – formally – they should have been labeled differently. For example, the first confirmed exoplanet imaged by the Hubble Space Telescope around the dwarf Astar Fomalhaut (˛ Piscis Austrini) was given the name Fomalhaut b even though it formally should have been called Fomalhaut B (Kalas et al. 2008). The same applies to the planetary system directly imaged from the ground around the dwarf F-star HR 8799 (Marois et al. 2008, 2010): the planets were given the names “b,” “c,” “d,” and “e” by their discoverers. When the first planet to show a transit was detected, HD 209458 b (Henry et al. 1999; Charbonneau et al. 2000), it was already known to be a planet via spectroscopy – hence the lowercase designation. When the presence of a planet was first suggested via the transit method only, e.g., the object OGLETR-56, it formally should have been designated OGLE-TR-56 B, i.e., using the uppercase designation used for eclipsing binaries (exoplanet transits are nothing but extremely modest eclipses). However, the slight photometric dips thought to be due to exoplanet transits can also be produced by unresolved background eclipsing binaries: if a foreground star provides most of the light, a normally obvious eclipse light curve can be diluted to where it is barely visible. Certainty comes from detecting the radial velocity signature, and so the object above was given the name OGLE-TR-56 b instead (Konacki et al. 2003). On the other hand, the first exoplanet to be discovered by a dedicated transit survey was given the name TrES-1 (named after the first successful detection by the Trans-atlantic Exoplanet Survey) rather than 2MASS 19040985+3637574 b or GSC 02652-01324 B (Alonso et al. 2004). With NASA’s Kepler satellite, the detection of transits became the most common means of discovering exoplanet candidates. The preliminary candidate host stars – objects that showed potential transit signatures – were simply numbered as “Kepler objects of interest” (KOI) , and the suggested exoplanetary candidates received a two-decimal number starting with KOI*.01. Those that were eventually identified as being excellent candidates were then given a number for the official Kepler

6 The Naming of Extrasolar Planets

89

exoplanet list. The Kepler photometric data are often of such high quality that final spectroscopic confirmation wasn’t as important as for ground-based detections: rather than adopting the eclipsing binary standard using capital letters or an arbitrary numbering system, the Kepler team chose to use the more common lowercase letter suffixes. Thus, the exoplanet candidate KOI 70.01 was eventually found to have reliable transits and became Kepler-20 b (Gautier et al. 2012; Fressin et al. 2012). Given that the use of upper- and lowercase suffix notation was already somewhat arbitrary when applied to stars, the use of the latter for 99% of all exoplanets wasn’t a problem until the planetary systems became more complicated. Within a stellar multiple system whose components are visually resolved, a circumstellar planet around one of the components could formally be given a uppercase designation, since this nomenclature normally doesn’t express any direct dynamic hierarchy, but both common usage and the natural desire to express the secondary dynamic nature of a planet naturally lead to the simple addition of a lowercase suffix to that of the star. For example, the planet orbiting the dwarf G-star 55 Cancri A (obviously in a binary system together with 55 Cancri B) has the proper (i.e., full) name 55 Cancri Ab. This naming convention, while seemingly natural, has its own problems, however: this usage implies that 55 Cancri A also has the name 55 Cancri a (all “b” designations for planets imply that there is a host “a” component) or perhaps 55 Cancri Aa, a name that expresses a complicated dynamical hierarchy but could be mistaken for a system of two objects rather than just one. With the discovery of circumbinary planets like that in the binary system Kepler16 (Doyle et al. 2011) – a planet showing transits superimposed on the eclipse light curves of a close stellar binary system – the problem of a useful naming convention again raised its head: the stellar components could clearly be given the designations Kepler-16 A and B, but the name Kepler-16 b (or even Kepler16 C) doesn’t express any dynamical or hierarchical information. In order to find a more useful nomenclature, Hessman et al. (2010) proposed a variation on the classical stellar naming conventions. This system has effectively been adopted in the literature but awaits formal confirmation/correction by the IAU working group. It is based upon four simple rules.

Rule 1. The formal name of an exoplanet is obtained by appending the appropriate suffixes to the formal name of the host star or stellar system. The upper hierarchy is defined by uppercase letters, followed by lowercase letters, followed by numbers, etc. The naming order within a hierarchical level is for the order of discovery only. This rule corresponds to the present provisional IAU-sanctioned WMC naming convention, including the present use, e.g., of the “Bb” notation for the exoplanets around the secondaries in binaries. Rule 2. Whenever the leading capital letter designation is missing, this is interpreted as being an informal form with an implicit “A” unless otherwise explicitly stated.

90

F. V. Hessman

This rule corresponds to the present exoplanet community usage for planets around single stars (e.g., 51 Peg b  51 Peg Ab). Thus, all of the present names for 99% of the planets around single stars are preserved as informal forms of the IAU-sanctioned provisional standard. Rule 3. As an alternative to the nomenclature standard in rule#1, a hierarchical relationship can be expressed by concatenating the names of the higher-order system and placing them in parentheses, after which the suffix for a lower-order system is added. This rule permits one to keep the lowercase b notation even when the previous hierarchical naming would suggest the use of a different suffix. For example, given an exoplanet in a circumbinary orbit around the fictitious close binary system CT Men, one could, in principle, name the exoplanet with any of the following conventions: CT Men B, the “second” part of the system otherwise consisting of the two stars CT Men Aa+Ab but potentially containing another stellar system CT Men C with a totally different dynamical status; CT Men C, the third body in the system otherwise consisting of the two stars CT Men A+B, placing the circumbinary exoplanet on the same hierarchy as the two stars it orbits; or CT Men (AB)b, the “second” dynamical part of the system otherwise consisting of the two stars CT Men A+B. The addition of the form using parentheses to the provisional IAU standard makes it possible to support the last rule. Rule 4. When in doubt (i.e., if a different name has not been clearly set in the literature), the hierarchy expressed by the nomenclature should correspond to dynamically distinct (sub-)systems in order of their dynamical relevance. The choice of hierarchical levels should be made to emphasize dynamical relationships, if known. This rule exploits the implicit freedom within the IAU provisional standard to help decide which hierarchical scheme to adopt. The examples above clearly show that the new form is the best form for known circumbinary planets and has the nice side effect of giving these kinds of planets an identical sublevel hierarchical label and stellar component names which conform to the usage within the very close binary community.

Figure 1 shows several examples illustrating just how complicated the naming of exoplanets can be, including the use of the third-level numerical nomenclature for an extrasolar system moon (“exomoon”). Indeed, a real candidate exomoon has been proposed by Teachey et al. (2017) in the exoplanet Kepler-1625b, but – in traditional astronomical manner – they have adopted yet another naming nomenclature by calling the exomoon Kepler- 1625 b i (i.e., with lowercase Roman instead of Arabic numbers). Barring a formal acceptance of the system proposed by Hessman et al. (2010) by the IAU, the use of lowercase Roman numbers would be a reasonable alternative, but in keeping with the other notations, one should, of course, start with roman numeral ii!

6 The Naming of Extrasolar Planets

91

Fig. 1 Examples of different exoplanet name suffixes in single and binary systems (taken from Hessman et al. 2015). Upper left: a single exoplanet around a single star (e.g., 51 Peg) plus a moon. Upper right: double star, each with a planet (e.g. HD 41004), plus a circumbinary planet. Lower left: two circumbinary planets (e.g., NN Ser). Lower right: a planet around the secondary star in a binary (e.g., HD 178911)

Back to Proper Names Currently, the vast majority of exoplanet names (95%) are based on survey and catalogue numbers, largely due to the incredible success of NASA’s Kepler mission at finding exoplanets around relatively faint stars. Only 1% of the host stars have Bayer designations, 0.9% Flamsteed, 0.7% variable star, 2% by catalogue plus positions, and only 0.2% have formal names. This situation will change somewhat after the next generation of exoplanetary satellite missions targeting brighter stars: NASA’s Transiting Exoplanet Survey Satellite (TESS, launch 2018) hopes to find about 3000 new exoplanets showing faint transits against relatively bright host stars (< 12t h magnitude), and ESA’s CHaracterizing ExOPlanets Satellite (CHEOPS, launch 2018) will undoubtedly find new planets around the known exoplanetary host stars it targets. Farther in the future, ESA’s PLAnetary Transits and Oscillations of stars satellite (PLATO, launch 2026) will observe nearly a million stars brighter than 11t h magnitude with multiple telescopes that will enable the system to reach the photometric accuracies necessary to detect many rocky planets around brighter stars. Still, the vast majority of stars even in this magnitude range have catalogue rather than “proper” names, so the scientific names of future exoplanets are likely to be as arcane and occasionally confusing as those presently given. We should remember that these names serve only to identify what researchers are talking about – whether an object is called Fomalhaut b or ˛ PsA B is ultimately unimportant as long as we can agree on what object is associated with which name or names. This coldly rational approach to exoplanet naming ignores the fact that each be-itso-arcane name conjures up visions of distant planets, each with their own colorful surfaces and atmospheres. These visions have been fed by untold science fiction

92

F. V. Hessman

stories and films and are easily simulated in planetarium programs and computer games. Thus, it is natural that the IAU has started a campaign to find proper names for many of the thousands of planets known and that such a project has been opened to the public – see the information about the IAU below and  Chap. 7, “Impact of Exoplanet Science in the Early Twenty-First Century” for more details.

IAU Resources The following IAU divisions, commissions, and working groups – all recently reorganized at the 28th general assembly – deal with and/or curate nomenclatures relevant to exoplanets. • The Division C (“Education, Outreach and Heritage”) Working Group Star Names (WGSN) maintains a list of official star names (available online). • The Division F (“Planetary Systems and Bioastronomy”) Commission “Exoplanets and the Solar system” has the formal responsibility for exoplanet nomenclature. • The Division G (“Stars and Stellar Physics”) Commission G1, Binary and Multiple Star Systems, has the responsibility for the nomenclature of multiple star systems. • The Executive Committee (EC) “Public Naming of Planets and Planetary Satellites” has the task of helping to organize the assignment of proper names to exoplanets and their satellites, using public input, but the formal responsibilities lie with the commissions and working groups. Note that the Division F working groups Planetary System Nomenclature (WGPSN) and Small Bodies Nomenclature (SBN) concern themselves with solar system objects only, but that there are obvious joint naming issues shared between the solar system and exoplanet communities.

Cross-References  Discovery of the First Transiting Planets  Exoplanet Catalogs  Impact of Exoplanet Science in the Early Twenty-First Century  Space Missions for Exoplanet Research: Overview and Introduction  Special Cases: Moons, Rings, Comets, and Trojans  The Discovery of the First Exoplanets  The Way to Circumbinary Planets Acknowledgements I would like to thank M. Geffert (Bonn) for trying to find the very first use of the variable star notation by Argelander.

6 The Naming of Extrasolar Planets

93

References Alonso R, Brown TM, Torres G et al. (2004) TrES-1: the transiting planet of a bright K0 V star. ApJ 613:L153–L156 Charbonneau D, Brown TM, Latham DW Mayor M (2000) Detection of planetary transits across a sun-like star. ApJ 529:L45–L48 Doyle LR, Carter JA, Fabrycky DC et al. (2011) Kepler-16: a transiting circumbinary planet. Science 333:1602 Fressin F, Torres G, Rowe JF et al. (2012) Two earth-sized planets orbiting Kepler-20. Nature 482:195–198 Gautier TN III, Charbonneau D, Rowe JF et al. (2012) Kepler-20: a sun-like star with three subNeptune exoplanets and two earth-size candidates. ApJ 749:15 Henry GW, Marcy G, Butler RP Vogt SS (1999) HD 209458. IAU Circ 7307 Hessman FV, Dhillon VS, Winget DE et al (2010) On the naming convention used for multiple star systems and extrasolar planets. ArXiv e-prints Hoffleit D (1987) History of variable star nomenclature. J Am Assoc Var Star Obs (JAAVSO) 16:65–70 Kalas P, Graham JR, Chiang E et al. (2008) Optical images of an exosolar planet 25 light-years from earth. Science 322:1345 Konacki M, Torres G, Jha S Sasselov DD (2003) An extrasolar planet that transits the disk of its parent star. Nature 421:507–509 Marois C, Macintosh B, Barman T et al (2008) Direct imaging of multiple planets orbiting the star HR 8799. Science 322:1348 Marois C, Zuckerman B, Konopacky QM, Macintosh B Barman T (2010) Images of a fourth planet orbiting HR 8799. Nature 468:1080–1083 Nakajima T, Oppenheimer BR, Kulkarni SR et al (1995) Discovery of a cool brown dwarf. Nature 378:463–465 Teachey A, Kipping DM Schmitt AR (2017) HEK VI: on the dearth of Galilean analogs in Kepler and the exomoon candidate Kepler-1625b I. ArXiv e-prints Wolszczan A Frail DA (1992) A planetary system around the millisecond pulsar PSR1257 + 12. Nature 355:145–147

7

Impact of Exoplanet Science in the Early Twenty-First Century Hans J. Deeg and Juan Antonio Belmonte

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Impact of Exoplanet Science Within Professional Astronomy . . . . . . . . . . . . . . . . . . . . . . . . . Exoplanets and the General Public . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Citizen Science . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Public’s View About the Presence of Life in the Universe . . . . . . . . . . . . . . . . . . . . . . The “NameExoWorlds” Contest: Was the Interaction Between the Exoplanet Scientific Community and the Public Successful? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

96 96 99 99 101 103 112

Abstract

The impact of exoplanet science at the beginning of the twenty-first century is studied from two different approaches: the impact in astronomical science both by professionals and citizen scientists and through public outreach. The impact of exoplanetology on the scientific community as well as on the informed public is presented by several indicators and examples. Currently about 3% of all refereed articles in astronomy are focused on exoplanets, whereas about a quarter of the science cases for very large multipurpose astronomical instruments is based on exoplanet science. Interactions with the public occur on several levels; we present

H. J. Deeg () Instituto de Astrofísica de Canarias, La Laguna, Tenerife, Spain Departamento de Astrofísica, Universidad de La Laguna, La Laguna, Tenerife, Spain e-mail: [email protected] J. A. Belmonte Instituto de Astrofísica de Canarias, La Laguna, Tenerife, Spain Departamento de Astrofísica, Universidad de La Laguna, Tenerife, Spain e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_166

95

96

H.-J. Deeg and J. A. Belmonte

an overview of exoplanet-related citizen science and discuss the influence of the public’s exposure to media coverage about exoplanets and life in the universe. This influence is changing the public’s perception about the uniqueness of our Earth and about the presence of extraterrestrial life. In turn, this perception affects the public support for the research infrastructure that is necessary to maintain the growth of the field of exoplanetology. Also, an interactive process between exoplanet science and the public through IAU’s NameExoWorlds Contest is presented in detail. Keywords

Exoplanetology · Planetary systems · Sociology of astronomy publications bibliography · History and philosophy of astronomy

Introduction Without doubt, in the last 25 years, exoplanetology has become an integral part of the sciences of astronomy and astrophysics and one of its most rapidly growing fields. But now, well into the twenty-first century, how far has come the impact and interest for exoplanets, both in the professional domain and in the society at large? Which interactions with society has it generated? These questions should also be of interest for active professionals, for their perception about their own field, and also with towards continued support for professional research activities in exoplanetology. In the first part of this chapter, we provide several estimators for exoplanetology’s current role within professional astronomy. The impact of exoplanets onto the general public is more difficult to evaluate. Scientists interact through a variety of media directly or indirectly with the public. We choose to present this interaction on several levels, discussing first the involvement of the interested public in citizen science projects and then the reaction of the general public to media coverage about exoplanets and the possibility of life in the universe. Using the example of efforts by the IAU to designate more popular names for some exoplanets, the interaction between citizen and the science community is shown in more detail.

Impact of Exoplanet Science Within Professional Astronomy Since the first widely recognized exoplanet in 1995, exoplanet science certainly has become a significant field in astronomy. But how big has it become? Here we give a few indicators that may quantify this issue. Figure 1 shows the growth in published papers that contain the terms “exoplanet” or “extrasolar AND planet” be it in title, abstract, or text body. One of the authors (HJD) revised his own publications against this search, and it provided a rather precise selection of those he considers as related to exoplanets (using the same search terms only on title and abstract led to the omission of about a third of those

7 Impact of Exoplanet Science in the Early Twenty-First Century

97

Fig. 1 Number of publications, in 2-year bins, that contain the terms “exoplanet” or “extrasolar AND planet” for the years 1989–2016. (Source: Astrophysics Data System)

publications, including many fully focused on exoplanets). This search is however not perfect, and some papers (such as Mayor and Queloz’s 1995 discovery of 51 Peg b) are missed, mainly because ADS’s full-text search is not implemented for all journals and years. In any case, the fraction of missed papers should be small and concern mainly older works. Starting from a rather constant level of about 20 refereed papers per year during the 1980s and a moderate rise during most of the 1990s, the field started to take off in 1998 or 1999. This was followed by a strong rise in productivity during the entire first decade of the 2000s, surpassing 1000 yearly papers in 2010. Exoplanets is therefore truly a science branch of the twenty-first century! A slight leveling of the growth can be observed since 2015, which is likely related to the ending of the main part of the Kepler space mission in 2013. To evaluate the productivity of exoplanet science within the whole field of astronomy, we note that the 1530 refereed papers in 2016 found by the search correspond to 5.8% of the 26,092 refereed papers in the entire astronomy collection of ADS for 2016. An attempt to quantify the fraction of the papers with a principal focus on exoplanets was made by revising the last 100 papers of the year 2016 that had been found by the search. Of these, 55 were considered to have exoplanets (as well as analytical or instrumental techniques mainly used in exoplanet science) as their main topic. We may therefore conclude that about 3% of the current technical works in astronomy are focused or driven by exoplanet science. An alternative indicator on the impact of exoplanetology within the entire field of astronomy might be the amount of exoplanet-related sessions in large meetings that cover the entire discipline of astronomy. In Table 1, this fraction is shown for a sample of recent General Assemblies of the IAU, of meetings of the American Astronomical Society, and of the “EWASS” meetings of the European Astronomical

98

H.-J. Deeg and J. A. Belmonte

Table 1 Number and fraction of exoplanetology sessions in major astronomy meetings Year 2018 2015 2012 2009 2006 2017 2017 2016 2016 2017 2016 2015 2014

Name IAU GA, Vienna IAU GA, Hawaii IAU GA, Peking IAU GA, Rio de Janeiro IAU GA, Prague 230th AAS, Austin, TX 229th AAS, Grapevine, TX 228th AAS, San Diego, CA 227th AAS, Kissimmee, FL EWASS, Prague EWASS, Athens EWASS, Tenerife EWASS, Geneva

Ns 29 34 41 37 36 27 107 33 100 53 46 45 24

Nexo_foc 2 2 3 1 0 5 15 4 15 3 1 3 2

Nexo_rel 3 3 0 3 0 5 4 0 5 2 2 0 0

fexo 12% 10% 7% 7% 0% 28% 16% 12% 18% 8% 4% 7% 8%

Ns is a meeting’s total number of sessions. In the IAU General Assemblies and the EWASS meetings, symposia were counted as two sessions. “Joint discussions” and “special sessions” were counted single. In the AAS meetings, only the concurrent talk sessions were counted. Nexo_foc and Nexo_rel give the number of sessions with a topic focused and related, respectively, on exoplanets. fexo is the fraction of sessions on exoplanets, where the focused ones were weighted fully and the related ones were weighted by a half

Society. As can be seen, exoplanetology accounts for about 10% at these meetings with increasing tendency, with the largest devotion to exoplanets occurring in the meetings of the American Astronomical Society. One of the most important consequences of exoplanetology is its share among the science objectives that are the basis for the design of the major forthcoming multipurpose astronomical instruments. This might even be more important than the field’s current productivity, as it is an indicator of the funds that are being expended and of the expected importance of the field’s findings during the coming decades. Here we indicate high-level science cases that are stated for some of the major actual instruments: • Exoplanets – Towards Other Earths is one of six principal cases for the European Extremely Large Telescope (Kissler-Patig and Lyubenova 2011). • The Birth and Early Lives of Stars and Planets and Exoplanets are two out of nine cases for the Thirty Meter Telescope (Skidmore 2015). • Birth of Stars and Protoplanetary Systems and Planets and Origins of Life constitute two of the four main cases for the James Webb Space Telescope (https://jwst.nasa.gov/science.html). • Planet-Forming Disks is given as one of the six science themes of the Atacama Large Millimeter Array (https://almascience.nrao.edu/alma-science). As a very rough estimate, exoplanetology constitutes about a quarter of the motivation for these major facilities in astronomy. This rather large share is certainly

7 Impact of Exoplanet Science in the Early Twenty-First Century

99

also based on the potential that is perceived for this field (with an emphasis on the issue of life in the universe) outside of professional astronomy, among political and financial decision makers and among the public at large. Without doubt, since the first exoplanet in 1995, exoplanetology has become a principal area in astronomy!

Exoplanets and the General Public More difficult is an evaluation of the impact of exoplanet science on the public at large. Here we are limited to give a few types of examples where exoplanetology reaches out and affects the informed general public, that is, the section of the public that displays interest in natural sciences and is capable of understanding their basic concepts.

Citizen Science Under citizen science we may encompass several levels of involvement by nonprofessional actors. This may go from a fleeting interest as might be displayed by the installation of a screensaver for some distributed computing project to a very serious level that approaches the work of professionals, such as displayed by some advanced amateur astronomers. Distributed computing and analysis projects are likely the twenty-first century’s principal involvement of citizens in actual science projects. On the most basic level, also known as volunteer computing, a citizen lends private computing power to a science project employing distributed computing. The project seeks to solve a problem which is difficult or infeasible to tackle using other methods, but it does not require further citizen interaction besides the installation of a software running on a networked computer. The best known example of this is likely SETI@home, a search for signals from extraterrestrial intelligences in data taken by the SETI project. The SETI@home screensaver, released to the public in 1999, got quickly very popular and was a common sight on many computers in the first years of the twenty-first century (see Fig. 2); it remains to date (October 2017) among the top 2 distributed computing projects based on the number of connected CPUs (Wikipedia, list of distributed computing projects). On the next level are projects that are computer-based but with active participant involvement. Typically, repetitive tasks that are difficult for computer codes but easy for the human brain are being distributed to the participants. The Zooniverse (www. zooniverse.org), a web portal operated by the Citizen Science Alliance, has been a pioneer in this field. In October 2017, it hosts 77 such projects from a variety of disciplines. Besides three projects involving solar system planets (“Planet 4” for Mars and “Planet 9” as well as “Backyard Worlds: Planet 9” for the suspected outermost planet), searches for exoplanets are represented by the Planet Hunters

100

H.-J. Deeg and J. A. Belmonte

Fig. 2 The SETI@home classic screensaver and distributed computing project in its original (“classic”) version. (Source Wikimedia, SETI@HOME is licensed under the GNU General Public License)

(www.planethunters.org) and the Exoplanet Explorers (https://www.zooniverse.org/ projects/ianc2/exoplanet-explorers/) projects. For Planet Hunters, the project’s participants are employed to spot transit-like features in chunks of data from the Kepler space mission and, in a second phase, to inspect and confirm the found candidates, with an assessment if further follow-up is warranted (Schwamb et al. 2012, for a description of its initial setup). The Planet Hunters project has led to a series of currently 11 “Planet Hunters” papers in refereed journals, mostly presenting discoveries of exoplanets, including the discovery of a circumbinary planet in a quadruple star system (Schwamb et al. 2013). The Exoplanet Explorers project is rather similar, working on the data by Kepler’s successor, the K2 mission. With over 15,000 volunteers, it has led to the discovery of several planet systems, including one with four transiting planets (Barentsen et al. 2017). The most impacting result of these projects is likely the discovery (Boyajian 2016) of a mysterious object with very unusual brightness variations (see Fig. 3). A variety of explanations have been brought forward (see Wright and Sigurðsson 2016 for an overview), including very speculative ones involving extraterrestrial intelligences, which led to significant media attention. The discovery of Boyajian’s star has been a wonderful result of the power of citizen science: its light curve had been discarded by all the algorithms that had sifted through Kepler data, and only through the viewing by a real person was its strange nature recognized – leading to one of the strangest puzzles in current-day astronomy! Amateur astronomers in their classical form, with observations taken by themselves, are also contributing significantly to exoplanet science. In contrary to

7 Impact of Exoplanet Science in the Early Twenty-First Century

101

Fig. 3 The unusual light curve of Boyajian’s star which was discovered by citizen scientists in data of the Kepler space mission. (Figure by JohnPassos – Own work, CC BY-SA 4.0, https:// commons.wikimedia.org/w/index.php?curid=46685223)

the expectation prior to the discovery of the first transiting exoplanets, useful observation of exoplanets do not always need large professional equipment, but small instruments accessible to amateurs may do so as well. Indeed, several projects to find transiting planets have been based on amateur equipment, although executed by professionals, and the first planet transit, of HD209458 b, was found with such equipment. Besides planet transits, now the main stay of amateur involvement in this field, there have also been some forays into optical (Project Bambi, www.bambi.net/) and radio SETI searches (Boquete Optical SETI Observatory, optical-seti.org). While the independent finding of new exoplanets cannot be expected from amateur observations (but surprises are always possible), at least one exoplanet (XO-1b, McCullough et al. 2007) was discovered with their help. The observation of transits of known exoplanets on brighter stars has become routine, and detailed observing instructions are available (e.g., Gary 2010; Dennis 2016). Numerous light curves of planet transits contributed by amateurs have been collected in the Exoplanet Transit Database (http://var2.astro.cz/ETD/). Their principal use is in the surveying of the constancy of transit shapes and in the tracking of transit and ephemeris, with the best light curves being a good match to data taken by professional astronomers (see Fig. 4). An exciting target that is currently being surveyed by several amateurs is Boyajian’s star, while the 2018 launch of the TESS satellite, dedicated to an all-sky survey of planets transiting bright stars, is expected to provide a multitude of new targets that might be at the reach of these citizen scientists.

The Public’s View About the Presence of Life in the Universe The general public is being informed about exoplanets from a variety of sources. How may this have changed the public’s perception of our Earth’s position in the universe, as well as its perception on the universality of life?

102

H.-J. Deeg and J. A. Belmonte

R-MAG BRIGHTNESS CHANGE [MMAG]

0.015 1

0.010

2

MID

3

4

0.005 0.000 –0.005 –0.010 –0.015 –0.020 –0.025 –0.030 –0.035 –3

–2

1 –1 0 TIME AFTER MID-TRANSIT [HR]

2

3

Fig. 4 Amateur light curve of a planet transit across the 11.3 mag star XO-1, made with a 14-in. telescope. (From Gary 2010, reproduced with permission by the author)

From a historic perspective, our view about life in the universe, or about other planets away from the sun, has changed from one of complete speculation to a semiempirical one, where the famous Drake equation (Drake 1965, 2011) remains as the best-known tool for quantification attempts. This equation, whose original form estimates the abundance of detectable intelligent life, is based on the multiplication of several factors, of which only one, the rate of star formation in our Galaxy, could be estimated reasonably well when the equation was originally presented. Some other factors, such as the fraction of planets with life that develops technological civilizations or the length of time over which such civilizations release detectable signals, remain completely unknown to date. Our increasing knowledge on exoplanets has however raised two of the equation’s factors, namely, the fraction of stars with planets, and the number of habitable planets around such stars, to estimates that may be reliable within a factor of “a few.” These estimates constitute one of the major advances of the past 20 years of research on exoplanets, leading, for example, to an estimate (Petigura et al. 2013) that about 6% of sunlike stars have Earth-like planets (see also  Chap. 96, “Planet Occurrence: Doppler and Transit Surveys” and Section 14, “Where Life May Arise: Habitability” of this handbook). However, currently we may label planets as “Earth-like” solely based on their principal physical parameters but without crucial information for habitability such as the presence of water or the type of atmosphere. The quoted 6% is therefore an upper limit for habitable Earth-like planets. The habitability of planets that are not Earth-like is hypothetical but in many cases reasonable (e.g., on planets around lowmass stars or on planet-sized moons, see also  Chaps. 141, “Habitability of Planets in Binary Star Systems”, and  142, “Habitability in Brown Dwarf Systems”), and such planets might be the vast majority among the habitable ones. It is however safe

7 Impact of Exoplanet Science in the Early Twenty-First Century

103

to assume that habitable planets (in the sense of fulfilling all requirements to develop life) are very frequent and that life – evolved or not – will be frequent as well, unless the remaining factors of Drake’s equation are minuscule small (e.g., Frank and Sullivan (2016) estimate that there is at least one other technological species in the observable universe, unless the probability that a habitable zone planet develops a technological species is below 1024 ). These results from exoplanet science, as well as numerous individual findings of potentially habitable or Earthlike planets, have led to the perception among the informed public that life might be frequent in the universe and is there to be discovered. At the same time, some concrete search activities have come within technological reach. Such searches are supported by a wide interest among the public, which is essential for the support of the many ambitious projects of the coming decades (for an interesting debate about the public perception of exoplanets, see https://www.astrobio.net/meteoritescomets-and-asteroids/the-great-exoplanetdebate-part-8-the-exoplanets-that-cried-wolf/). These in turn will advance the entire field of exoplanet science and humanity’s understanding of the origin and presence of life in the universe. And lastly, the public interest in exoplanets may better prepare society for the day when we find out that we are not alone!

The “NameExoWorlds” Contest: Was the Interaction Between the Exoplanet Scientific Community and the Public Successful? Current exoplanet designations do not exactly sound inviting to the general public! The field of exoplanets has grown so fast that, literally, there was no time to put order in what is probably the largest mess in astronomical nomenclature in history so far: exoplanet names. Although there have been attempts of producing a naming convention (see, e.g., Lyra 2010), the scientific community has resigned itself to designate exoplanets with a system only marginally distinct to the one for the naming of binary and multiple star components (see Hessman et al. (2010) or  Chap. 6, “The Naming of Extrasolar Planets” in this handbook). Names are so variegated that a nonexpert in the field may have the impression that we were not talking of the same class of objects: planets orbiting stars other than the sun. In this sense, 55 Cancri e, or ¡ Cancri e, or Janssen, its most recent name, is one of the thousands of exoplanets stuck with multiple names because of inconsistent naming systems. 55 Cancri e is experiencing a bit of an identity crisis (Koziol 2016). The planet, which orbits a star about 40 light-years away, also answers to 55 Cancri A e, ¡Cnc e, HD 75732 e, and “Janssen.” Janssen and the other planets in its star system – Galileo, Brahe, Lippershey, and Harriot – were part of the 2015 effort by the International Astronomical Union (IAU) to provide common names to 14 stars (out of 19 systems) and the 31 planets that orbit them: the “NameExoWorlds Contest” (see Fig. 5). The process involved public nomination and online voting for the new names. These names were intended as approachable additions to the astronomical designations originally given by scientists.

104

H.-J. Deeg and J. A. Belmonte

Fig. 5 General results of the IAU NameExoWorlds Contest. Panel (a) stands for voting statistic behavior (this includes the large number of votes that were disqualified due to proposing a Hindu guru; see text). Notice the outstanding public participation of Spain (c. 8% of the votes despite this country represents only 0.6% of world population). Panel (b) shows the contest winners by continents. Western Europe and North America, where exoplanets research was born and fructified, promoted more names than the rest of the world all together. (Adapted from IAU’s https://www. iau.org/news/pressreleases/detail/iau1514/)

It has been a small step in the impossible task of bringing order to the night sky that has been one of the long-standing nightmares of astronomy since its inception. As the number of confirmed exoplanets beyond the borders of our solar system has exploded in recent decades from dozens to thousands, the patchwork naming systems developed for them have not kept pace. Even IAU’s yearly efforts approving

7 Impact of Exoplanet Science in the Early Twenty-First Century

105

names for more celestial bodies are a drop in the bucket to the number of known stars and planets out there. All of these newly found exoplanets are being published with designations coming from various star catalogs. But they are unappealing, confusing, and boring. Most of these “scientific names” can be a very long list of numbers and letters or the name of a sophisticated planet finder instrument plus a number. It is as in the R2D2 or C3P0 characters of the Star Wars saga but in a cosmic scale: something that plays like a joke ends in a crunchy nightmare. And to make matters worse, each exoplanet often has more than one scientific name. It is not the fault of the exoplanets, though, because the problem predates their discoveries by centuries. As described by Koziol (2016), exoplanets are named after their host star’s scientific designation. Janssen, for example, orbits the star 55 Cancri or ¡ Cancri or HD 75732, depending on who you ask. Each references a different star catalog. ¡ Cancri comes from Johann Bayer’s Uranometria, first published in 1603, an opus magnum including the designations for over 1500 of the brightest stars visible from Germany in his times. 55 Cancri comes from England’s Royal Astronomer John Flamsteed, who in the early eighteenth century catalogued over 2500 stars visible from Southern England. Finally, HD 75732 comes from the Henry Draper (HD) Catalogue, an early twentieth-century effort to catalogue more than 225,000 stars. Each cataloging effort gave the star another designation in line with its own naming system. This is why the star 55 Cancri, which for the sake of simplicity could now be referred to as Copernicus (thanks to the IAU’s 2015 effort), also has other less commonly used designations like HR 3522, Gliese 324, and HIP 43587, among others (Koziol 2016). Since their earliest discoveries, exoplanets have taken the name of their parent star, plus a lowercase letter to distinguish them as a body on their own. For multiplanet systems, the planets are lettered as they were discovered. The first planet is b, the second c, and so on. This poses a problem since this naming has no physical meaning (e.g., order of planets as they go outward) and in a few systems d or e is closer to the star than, for example, b or c. 55Cnc is a paradigm. The natural guess for Janssen (55Cnc-e) should be that it is the fourth planet from the star at the center of the 55 Cancri system. It is not. It is the closest, orbiting the star once every 18 h – so close that specialists could not decide if it was even real for over 6 years. When it finally received its scientific designations, three other planets in the system had already been identified and named. So even though 55 Cancri e is the first planet out, it is the fourth one in the alphabetical naming lineup. Only when a series of planets are discovered at once, the proximity rule applies. However, in situations when astronomers discover an exoplanet closer to its star than previously discovered planets in the system, they have no choice but to tweak the rules of the convention and upset the alphabetical order. Unfortunately, this means that there is no way to tell, from scientific designations, whether a planet’s letter refers to its orbit, and has a physical meaning, or simply to the order of discovery. Hence, to name an exoplanet, take any one of the host star’s two or three or five catalog names and append a letter to the end to designate where the planet is in

106

H.-J. Deeg and J. A. Belmonte

the order outward from the star – unless it is discovered later and closer to the star than its sibling planets. In that case, just stick the next available letter onto the end. It is indeed a complete and unfortunate mess as has been recently argued (Koziol 2016). In the worst-case scenarios, you may end up with exoplanets like Gliese 436 b, which has 34 different scientific designations, and experts must deal at once with several of them. Just imagine a similar situation in familiar or social relationships. Chaos would be guaranteed. Exoplanet hunting projects have complicated the issue with names such as TrES-1b, Corot-7b, or WASP-20b, just to mention a few of them, which apart from telling us which experiment has discovered the planet (and the order of discovery) does not give any further relevant information. Most recent space projects have continued to complicate matters. NASA’s Kepler space observatory has singlehandedly discovered (either confirmed or validated) over 2300 exoplanets, all of them orbiting stars too distant to be seen with the naked eye. To catalog them, the Kepler team once more created their own designation, adding planets like Kepler22b and Kepler-444c. In case the reader is not exasperated enough, the Kepler team maintains a separate designation for objects that might be planets but have not been validated with sufficient certainty called Kepler objects of interest, meaning there are more celestial objects out there with names like KOI-1118.01, or even KOI961.02, KOI961.01, and KOI961.03, where the numbers after the dot represent the order of discovery and, once more, have no physical meaning at all. Why the team decided to list the accepted planets with lowercase letters and candidate ones with numerals remains a mystery. But despite their inconsistency, the names are obviously needed since they are extremely useful for scientific publications and research. What else can we do? As Prof. Lecavelier des Etangs, former president of IAU Exoplanets Commission, has argued “they are like social security numbers,” this avoids any confusion. However, your mother does not call you by your social security number, but by your name or nickname, with which you feel much more comfortable. This is why there is so much need to giving these stars and planets common names. Even if it is highly probable that some of them will never replace the scientific designations (see Fig. 6), it is worth the effort. The goal behind the NameExoWorlds Contest was to test methods of giving these stars and planets common names. Five hundred seventy-three thousand and two hundred forty-two votes were cast for the 19 planetary systems being considered (see Fig. 5). These 19 systems open to naming had been selected in a first phase of the naming contest from a list of 305 systems with at least one published planet before to the end of 2008. Names were suggested from astronomical outreach institutions and societies (research centers were deliberately avoided) around the world and had to adhere to a strict set of guidelines. The IAU did not consider names of a commercial nature, the names of living individuals, the names of pets, or the names of individuals primarily known for political or religious activities. This is the reason why the announcement of the final results had to be delayed because a huge number of votes had gone to a Hindu guru, a proposal strictly forbidden by the contest rules.

7 Impact of Exoplanet Science in the Early Twenty-First Century

107

Fig. 6 While there is much debate over which exoplanet discovery is considered the “first,” one stands out from the rest. In 1995, Didier Queloz and Michel Mayor discovered 51 Pegasi b, forever changing the way we see the universe and our place in it. The exoplanet is about half the mass of Jupiter, with a seemingly impossible, star-hugging orbit of only 4.2 Earth days. Not only was it the first planet confirmed to orbit a sunlike star (51Peg); it also ushered in a whole new class of planets called Hot Jupiters. Despite efforts on the contrary, it will be difficult that a new name for this system, even if appropriate and appealing, will ever be successful in exoplanet science. (Image taken from https://exoplanets.nasa.gov/alien-worlds/exoplanet-travel-bureau/ by courtesy of NASA)

In the end, the IAU approved names ranging from the most famous astronomers and telescope makers (e.g., the star Copernicus and its cohort: Galilei, Brahe, Janssen, Lippershey, and Harriot) to mythological figures from around the world (a common custom in the designation of the solar system bodies, including planets and their moons, the new class of minor planets and asteroids; people prefer a name like Ogma – an Irish mythological figure – to HD 149026 b) and even old literary characters as we will lengthy discussed later on. The list of approved names can be found at http://nameexoworlds.iau.org/names. The first exoplanets ever discovered, which orbit pulsar PSR 1257 C 12, are now known as Poltergeist, Phobetor (a Greek god of nightmares), and Draugr (a fictional undead creature). These frightening names were proposed because these planets are thought to live in permanent darkness and possibly suffer quite inhospitable

108

H.-J. Deeg and J. A. Belmonte

conditions. Maybe they will be successful. On the contrary, Fomalhaut b, possibly the first exoplanet imaged directly, has received the name Dagon, a Semitic deity that was half man and half fish, certainly an appropriate name for a planet in the system of a star which translation from Arabic reads as the Mouth of the Fish, the brightest in the constellation of Piscis Austrinus (in this case, the star having a well-stablished name did not need to be renamed). Will Dagon prosper? Only time will say. In other cases, names will probably not be successful. The contest did not expressly involve scientific institutions, and many members of the exoplanet community and academic journals may not take note of the results of a public outreach event even if organized by IAU. This will probably be the case of the first sunlike star spotted with an orbiting planet known as 51 Pegasi (51Peg) and its planet 51 Pegasi b. Although several astronomers’ preferred name was Epicurus, their new IAU-given names were Helvetios and Dimidium, respectively. Helvetios refers to a Celtic tribe from the old ages populating what today is Switzerland (akin Helvetia), while Dimidium is the Latin word for “half,” a play on the planet’s mass measuring about half that of Jupiter. Both names were proposed by the Astronomical Society of Luzern in Switzerland. Actually, the first name is rather nationalistic (considering the origin of the discoverers) and will probably pass away without further noise. The second curiously includes the initial letters of the name of these same discoverers: Di[dier Queloz] and Mi[chel Mayor]. We do not know if this was chance or was deliberately chosen by the proposers. Indeed, this could be considered a fanciful and deserved homage to our colleagues (names of living people were strictly forbidden), and maybe it will be worth making an effort from the side of the scientific community to preserve it (perhaps under the catchier nickname “Dimi”). It is now the turn of discussing one of the contest’s proposals with which the authors of this essay were closely familiarized. As a result of the NameExoWorlds IAU Contest, the star and four planets of the Arae system received names in honor of Miguel de Cervantes Saavedra (1547–1616), a famous Spanish writer and author of El Ingenioso Hidalgo Don Quixote de la Mancha, i.e., “The Ingenious Gentleman Don Quixote of La Mancha.” The planets would get names after characters of that novel. Arae (abbreviated Ara), often designated HD 160691, is a main sequence G-type star approximately 50 light-years away from the sun in the constellation of Ara, the Altar, one of the historical constellations very close to Mediterranean southern horizons in antiquity. The star has a planetary system with four known extrasolar planets (designated Arae b, c, d, and e), three of them with masses comparable to that of Jupiter. The system’s innermost planet was the first “hot Neptune” to be discovered. As already discussed, the established convention for extrasolar planets is that the planets receive designations consisting of the star’s name followed by lowercase Roman letters starting from “b,” in order of discovery. This system was used by a team led by Go´zdziewski to designate the system planets (Go´zdziewski et al. 2007). On the contrary, another team proposed a modification of the designation system, where the planets are designated in order of characterization (Pepe et al. 2007). Since the parameters of the outermost planet were poorly constrained

7 Impact of Exoplanet Science in the Early Twenty-First Century

109

Fig. 7 An evocative, not to scale, portrait of the Arae stellar system with the five names received by the system star and planets, after the IAU’s NameExoWorlds Contest. The star became Cervantes, the famous Golden Century Spanish writer, while the orbiting planets received names of his most famous novel characters. (Image by courtesy of the Pamplona Planetarium and the Spanish Astronomical Society)

Fig. 8 A much applauded cartoon by the late ( Feb. 2018) Spanish graphic humorist Antonio Fraguas “Forges,” published in local media in support of the Cervatine naming proposal for the Ara system. Sancho’s humble donkey, Rucio, protests because his name has been ignored. This might be solved in the future if new planets are discovered in the stellar system. (Image by courtesy of the author, kindly provided by the Spanish Astronomical Society archive)

110

H.-J. Deeg and J. A. Belmonte

before the introduction of the four-planet model of the system, this results in a different order of designations for the planets in the Ara system. Both systems agree on the designation of the 640-day planet as “b,” but the old system designates the 9-day planet as “d,” the 310-day planet as “e,” and the outer planet as “c.” Since the International Astronomical Union had not defined an official system for designations of extrasolar planets, the issue of which convention was “correct” remained open, although subsequent scientific publications about this system appear to have adopted the Pepe et al. naming, reflected by the system’s entry in the wellrecognized Extrasolar Planets Encyclopedia. However, the mess was unavoidable; hence, a new appealing proposal was mandatory and the system was included in the IAU contest. The winning Cervantine proposal was presented by the Planetario de Pamplona (Spain) and supported by the Spanish Astronomical Society (SEA), the Spanish National Outreach Contact of the IAU, and the Instituto Cervantes to name the star

Arae and its four exoplanets with the name of Cervantes and those of the main characters of the novel (Gorjas 2016). The idea was to elevate Cervantes to the status of a galactic Homer, lending his name to the system’s central star, while Don Quijote (Quixote), Rocinante, Sancho, and Dulcinea were transfigured into his planetary escort. Quijote ( Arae b), the leading character, in a somewhat eccentric orbit, is befitting to his character, and beside is his faithful companion the horse Rocinante ( Arae d) in the middle of the scene. Good Sancho ( Arae e), the ingenious squire, is moving slowly through the outer insulae of the system. Finally, the enchanted Dulcinea ( Arae c), so difficult for Don Quijote to contemplate in her real shape, would orbit close to the heart of the writer (see Fig. 7). Further characters of the novel such as Clavileño (the wooden horse used by Sancho to travel to the stars) or the squire’s humble donkey Rucio (see Fig. 8), among others, would be reserved for future planet discoveries in the system. The importance of Miguel de Cervantes in the universal culture can hardly be overestimated. His major work, Don Quijote, considered the first modern novel of world literature and one of the most influential books in the entire literary canon, has many times been regarded as the best work of fiction ever written. However, while Shakespeare has his Uranian satellites, such as Miranda or Oberon, Cervantes has been so far excluded from the cosmic spheres. With this proposal, which arrived just in time for the 400 anniversary of the publication of the novel’s second part, the Famous Knight of La Mancha, his comrades and their creator obtained the place that they deserved among the stars. The proposal was supported by a huge social effort (see Fig. 9). It engaged a wide range of different actors from outreach institutions such as the proposer, Pamplona Planetarium, or dozens of Amateur Associations to scientific societies such as SEA and the whole Spanish-speaking astronomical community. The media also strongly supported the candidature (see Fig. 8) leading to a result that exceeded all expectances. The Cervantine proposal was not only the one most voted for the

Ara system but received more public votes than any other planetary system all together (with the exception of the breaking rule guru’s proposal). The following winning names such as Palmyra or Hypatia received a much lower percentage of votes (see Fig. 9). This was an extremely profitable and paradigmatic example of

7 Impact of Exoplanet Science in the Early Twenty-First Century

a

60000 50000 40000

Cervantes

30000 20000 10000 0

b

mu Arae

40000

30000

20000

10000

ira

f

Kh

Ar

ar

ap

iz

oo

a

an

R

u

w

as

la

er

ha

at

D

Am

C

a ni

ag

ta si

Lu

yp H

40000 69% 30000

20000

10000

er al s M

ku ko ai D

in

te n

... R

an um H

ob

ta

er tH

hu

a iz R

ia ta n si Lu

nt

es

0

er va

c

on

ia

a

at

yr lm

Pa

C

er

va

nt

es

0

C

Fig. 9 NameExoWorlds Contest statistics in relationship to the Cervantine proposal as a paradigmatic example of the interaction between social actors and the public. Panel (a): total number of votes for each of the 19 proposed candidates. Panel (b): the ten most voted candidate names. Cervantes is highlighted. Panel (c): votes obtained by the different proposal for the Ara system. The success of Cervantes’s proposal among the public is clearly emphasized. (Images adapted from https://estrellacervantes.es/ cervantes-ya-es-una-estrelladesde-el-15-de-diciembre-de2015/ by courtesy of the Pamplona Planetarium and the Spanish Astronomical Society)

111

112

H.-J. Deeg and J. A. Belmonte

the interaction between social actors, including the Spanish-speaking astronomical community, and the public. Will this effort survive? It will be fascinating to follow the process and to check whether the international astronomical community will adopt or ignore these and other exoplanet system names. At the end of 2017, we could find only one professional publication that mentions 51 Peg b’s new name Dimidium (Jenkins et al. 2017, who label it in the same line also as “Helvetios b”!). It will also be advisable to scrutinise the process so that IAU will decide in the future whether the promotion of a new naming contest is worthwhile or not. Perhaps someone will come out with a proposal that will satisfy both the public and the exoplanet scientific community. For an alternative naming system to become adapted in the scientific community, an IAU-sanctioned procedure to quickly designate such names to newly found planets seems essential, before the usual “b” names become entrenched in the literature. This procedure might be similar to the one by which the science team of the Kepler/K2 space mission rapidly assigns Kepler and K2 planet names, pending on the presentation of an accepted manuscript about a planet discovery from data of these missions. Though it sounds painful to give poor Gliese 436 b – a fascinating exoplanet anyway – a 35th name to answer to, at least this one will need to be catchy. Only future has the responses! Acknowledgments We want to thank the Spanish institutions who supported the Cervantine candidature, notably for their kind offer to use all the relevant material produced for the IAU naming contest. HD acknowledges support by grant ESP2015-65712-C5-4-R of the Spanish Secretary of State for R&D&i (MINECO).

References Barentsen G et al (2017) K2 citizen science discovery of a four-planet system in a chain of 3:2 resonances. American Astronomical Society, AAS Meeting #230, id.118.10 Boyajian TS (2016) Planet hunters IX. KIC 8462852 – where’s the flux? MNRAS 457:3988–4004 Dennis A (2016) Exoplanet observing by amateur astronomers, http://astrodennis.com/ Drake F (1965) The radio search for intelligent extraterrestrial life. In: Mamikunian G, Briggs MH (eds) Current aspects of exobiology. Pergamon, New York, pp 323–345 Drake F (2011) The search for extra-terrestrial intelligence. Phil Trans R Soc Lond A Math Phys Eng Sci 369:633–643 Frank A, Sullivan WT (2016) A new empirical constraint on the prevalence of technological species in the universe. Astrobiology 16:359–362 Gary B (2010) Exoplanet observing for amateurs, 2nd edn. Reductionist Publications, Hereford. available at http://brucegary.net/book_EOA/ExoplanetObservingAmateurs2ndEdition.zip Gorjas J (2016) Estrella Cervantes. http://estrellacervantes.es/estrella-cervantes-unaconferencia-de-francisco-javier-gorgas/ Gozdziewski K, Maciejewski AJ, Migaszewski C (2007) On the extrasolar multi-planet system around HD160691. ApJ 657:546–558 Hessman FV, Dhillon VS, Winget DE, Schreiber MR, Horne K, Marsh TR, Guenther E, Schwope A, Heber U (2010) On the naming convention used for multiple star systems and extrasolar planets. 2010arXiv1012.0707H Jenkins JS et al (2017) New planetary systems from the Calan–Hertfordshire extrasolar planet search. MNRAS 466:443–473

7 Impact of Exoplanet Science in the Early Twenty-First Century

113

Kissler-Patig M, Lyubenova M (eds) (2011) An expanded view of the universe: science with the European Extremely Large Telescope. https://www.eso.org/sci/facilities/eelt/science/doc/ eelt_sciencecase.pdf" Koziol M (2016) No one really knows how to name exoplanets. Naming new worlds is a confusing mess. http://www.popsci.com/how-an-exoplanet-gets-its-names Lyra W (2010) Naming the Extrasolar Planets. Boletim da Sociedade Astronòmica Brasileira 29(19):26 McCullough PR et al (2007) A transiting planet of a sun-like star. ApJ 648:1228–1238 Pepe F et al (2007) The HARPS search for southern extra-solar planets. VIII. Arae, a system with four planets. A&A 462:769–776 Petigura EA, Howard AW, Marcy GW (2013) Proc Natl Acad Sci 110(48):19273 Schwamb ME et al (2012) Planet hunters: assessing the Kepler inventory of short-period planets. ApJ 754:129 Schwamb ME et al (2013) Planet hunters: a transiting circumbinary planet in a quadruple star system. ApJ 768:127 Skidmore W (ed) (2015) Thirty Meter Telescope: detailed science case: 2015. http://www.tmt.org, document TMT.PSC.TEC.07.007.REL02 Wright JT, Sigurðsson S (2016) Families of plausible solutions to the puzzle of Boyajian’s star. ApJ 829:1L

Section II Solar System–Exoplanet Synergies Agustín Sánchez Lavega

Agustín Sanchez-Lavega is Full Professor (Catedrático) of Physics at the Faculty of Engineering of UPV/EHU in Bilbao. He founded and heads the “Aula EspaZio Gela” dedicated to promote studies and research in space science and technology. His research interests are the study of the dynamics, of meteorological phenomena, and of cloud and hazes in planetary atmospheres. He participates in space missions to Venus, Mars, and Jupiter from ESA and NASA, and in different committees for ESA and ESO. Currently, he is the manager of Spain’s national astronomy and astrophysics program. In 2014, the Royal Spanish Society of Physics awarded him the Price in Physics Education and Outreach at university level, while in 2016 he received the Euskadi Research Award from the Basque Country Government.

8

The Solar System: A Panorama Katherine de Kleer and Imke de Pater

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A Brief History of the Solar System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Sun . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Terrestrial Planets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Surfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Atmospheres . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interiors and Magnetic Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Satellites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Giant Planets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Atmospheres . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interiors and Magnetic Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Satellites and Rings of the Giant Planets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Major Satellites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Small Body Populations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Asteroid Belt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Near-Earth Objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Trans-Neptunian Objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Oort Cloud . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Comets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

118 120 122 122 125 125 126 126 127 128 128 129 130 132 134 135 136 136 138 138 139 139

K. de Kleer () Division of Geological and Planetary Sciences, California Institute of Technology, Pasadena, CA, USA e-mail: [email protected] I. de Pater Department of Astronomy, The University of California at Berkeley, Berkeley, CA, USA Faculty of Aerospace Engineering, Delft University of Technology, Delft, The Netherlands SRON Netherlands Institute for Space Research, Utrecht, The Netherlands e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_42

117

118

K. de Kleer and I. de Pater

Abstract

The closest and most extensively studied planetary system, our solar system provides the foundation for understanding the characteristics of planetary and sub-planetary bodies and the processes that shape them. This chapter surveys the diversity of objects orbiting our Sun and what they tell us about the origins and evolution of the solar system. The numerous small bodies populating specific orbits, from the asteroid belt to the far reaches of the Oort cloud, encode information on the solar system’s age and the initial conditions in the solar nebula. The surfaces and atmospheres of the planets and their satellites reveal how the same fundamental physical processes produced bodies with vastly different characteristics, from the dry, metal-dominated composition of Mercury through the storm-wracked hydrogen atmosphere of Jupiter. Finally, the search for liquid water and temperate climates elsewhere in the solar system, past or present, provides context for understanding the origin of life on Earth and the potential for life’s existence elsewhere in the Universe.

Introduction Our solar system consists of a diverse and dynamic collection of objects spanning many orders of magnitude in scale. From the Sun through the smallest atoms and electrons, all components play a role in the complex gravitational, chemical, and electromagnetic interactions that govern the evolution of the system as a whole. Although it is not yet clear whether our own solar system is representative of planetary systems elsewhere in the Universe, or even within our stellar neighborhood, it is the only system that we are currently capable of studying in detail. We study the solar system through a combination of spacecraft missions, ground-based and Earthorbiting telescopes, and laboratory analysis of samples brought back by missions or delivered to Earth’s surface in the form of meteorites. Such studies have shown us the complexity and diversity of solar system bodies, from the giant planets and their satellites down to asteroid fragments and minute ring particles. Through studying the solar system today, we try to piece together the story of the planets’ formation and subsequent evolution. The goal is to understand how the initial conditions, acted on by fundamental physical processes, led to the system we see today. We seek to explain why we have the number of planets we have in the orbits in which we see them and to understand the similarities and differences between them. What differences in the initial conditions of the terrestrial planets, or in their subsequent processing, produced the stark variations in atmospheric temperature and pressure between Earth, Venus, and Mars? Answering questions like these also sheds light on the occurrence and longevity of habitable environments on planets and satellites. Figure 1 shows an inventory of our solar system, where the size of each body is plotted as a function of distance from the Sun (the orbital and rotational properties of the planets are given in Table 1). The eight planets stand out with respect to

8 The Solar System: A Panorama

119

Fig. 1 Inventory of solar system objects. Note that the apparent increase in the minimum size of bodies with heliocentric distance is an observational bias, because it is much harder to find small bodies at large geocentric distances. Data provided by the International Astronomical Union’s Minor Planet Center (Figure after Spencer)

size; Jupiter, followed closely by Saturn, is by far the largest body. Two orders of magnitude down in size from the terrestrial planets, we find a multitude of objects. Some of these are on stable orbits, including the main-belt asteroids between the orbits of Mars and Jupiter, the Trojan asteroids at Jupiter’s L4 and L5 Lagrange points, and the Kuiper belt out beyond the orbit of Neptune. Other bodies are on chaotic or unstable orbits. Near-Earth asteroids may ultimately be on a course to hit a terrestrial planet, the Moon, or the Sun, and the Centaurs, in transit from the Kuiper belt into the inner solar system, will be transformed into comets when ices on their surface start to sublimate under increased solar heating. At roughly 10,000– 50,000 AU, our solar system is enveloped by a shell populated by several billion icy bodies that is referred to as the Oort cloud.

120

K. de Kleer and I. de Pater

Table 1 Orbital and rotation parameters of the solar system planets Planet Mercury Venus Earth Mars Jupiter Saturn Uranus Neptune

a [AU] 0.387 0.723 1.000 1.523 5.203 9.543 19.192 30.069

e 0.205 0.007 0.017 0.094 0.049 0.057 0.046 0.011

i [deg] 7.0 3.4 0.0 1.9 1.3 2.5 0.8 1.8

Obliquity [deg] 0.01 177.4 23.4 25.2 3.1 26.7 97.8 29.3

Porbit [Days] 88.0 224.7 365.2 687.0 4,331 10,747 30,589 59,800

Protation [Hours] 1,407.6 5,832.5 23.9 24.6 9.9 10.7 17.2 16.1

Data from the NASA National Space Science Data Coordinated Archive: http://nssdc.gsfc. nasa.gov/

In order to connect the current state of the solar system to its formation and evolution, we need to understand the processes at work on planetary bodies and calibrate our understanding of measurable signatures in terms of what they reveal about the past. The study of small solar system bodies – asteroids and comets – has taught us the age and initial conditions of the solar system. Computer simulations demonstrate potential evolution scenarios, and studies of planet-forming nebulae and disks around other stars provide the context into which to place our own stellar system.

A Brief History of the Solar System Our Sun is a fairly typical star, and our understanding of the origin and formation of the solar system is based on studies of other systems in the galaxy that we believe represent the preliminary stages of stellar and planetary formation, calibrated to match the observed composition and architecture of our solar system. We believe that the Sun was formed under similar conditions as other similar stars, in a dense core within a giant molecular cloud composed of molecular hydrogen, helium, more complex molecules, and dust grains. This core contained the inventory of volatiles and heavy elements present in our solar system today. At a specific point in time, an external trigger initiated the gravitational collapse of the dense molecular core. The core began to collapse under self-gravity, releasing gravitational potential energy that heated the central regions. Collisions between grains in the increasingly dense central region damped out vertical motion, resulting in a disk of material that spun in conservation of the initial random angular momentum of the core. Through a process that is still not well understood, the small grains agglomerated into large grains, which grew into kilometer-sized bodies known as planetesimals. Gravitational interactions then accreted the planetesimals into larger-sized embryos, or protoplanets.

8 The Solar System: A Panorama

121

The cores of the giant planets likely formed early via runaway solid-body accretion. Once these cores reached masses of order ten times the mass of the Earth, these protoplanets grew rapidly through runaway accretion of gas from the protoplanetary disk. The giant planets likely formed in the first 10 million years, while the terrestrial planets took about ten times longer. Formation stopped when the gas in the protoplanetary disk was cleared away by the strong solar wind during the Sun’s T Tauri phase. Since the abundance of solid material in the disk decreased with heliocentric distance beyond the “ice” line, it took much longer for protoUranus and proto-Neptune to form than Jupiter and Saturn, and hence these planets have much less hydrogen and helium than Jupiter and Saturn. Models for the evolution of the solar system during and after formation must be able to explain the sizes, locations, and compositions of the planets and the small body populations. Most models postulate that the giant planets have migrated significantly since their formation, although the timing of this migration, as well as the planets’ formation locations and migration paths, vary significantly. Migration of the giant planets scattered some planetesimals inward toward the Sun and ejected others into the distant Oort cloud. Those that were scattered into the inner solar system were responsible for the bombardment of the terrestrial planets evident in these planets’ cratering records. The heavily cratered surfaces of most rocky bodies in our solar system, together with the high porosities of many asteroids, suggest that the planets formed in a violent environment in which numerous bodies continued to impact the forming planets and collide with each other, often shattering protoasteroids or planets to pieces. Unless completely dispersed, such fragments accreted to form a new body, sometimes leaving fragments behind in the form of small moonlets. Proto-Earth is thought to have been hit by a Mars-sized object, leading to the formation of our Moon. The gradient in temperature from the hot Sun-forming region to the cold outer reaches is responsible for many of the broad-scale differences between the terrestrial and gas giant planets. Refractory metals and silicates condensed throughout the protoplanetary disk, while the cool temperatures in the outer solar system permitted the condensation of volatiles in the form of water (H2 O), ammonia (NH3 ), methane (CH4 ), and other species. Since the volatile elements are much more abundant than rock-forming elements (e.g., Si, Mg, and Fe), there was much more material available for planet formation in the outer regions of the planetary disk where the giant planets formed, than in the inner solar system where the terrestrial planets formed. These differences are reflected in the sizes and bulk densities of the eight planets, given in Table 2. Bodies over 100 km in size are typically differentiated. That is, their cores consist mostly of the heaviest elements like Fe, while their mantles consist of lighter rocky material that may be overlain, depending on the body’s composition, with an icy outer layer. The differentiation of bodies hints at a molten state early in their formation history, caused by the heat of formation and the decay of radioactive isotopes. Both of these heat sources remain active today, although their magnitude has been decreasing over the solar system’s history. Figure 2 shows the interior structures of the eight planets, which is the product of early interior differentiation.

122 Table 2 Physical parameters of the solar system planets

K. de Kleer and I. de Pater Planet Mercury Venus Earth Mars Jupiter Saturn Uranus Neptune

Mass [1027 g] 0.33 4.87 5.97 0.64 1,898 568 86.8 102

Radius [km] 2,440 6,052 6,378 3,396 71,492 60,268 25,559 24,764

Density [g/cm3 ] 5.427 5.243 5.514 3.933 1.326 0.687 1.271 1.638

The Sun In many respects, the characteristics of our solar system are dominated by the Sun itself. The Sun is the only feature of our solar system that would be visible from a neighboring star. It dominates the mass of the solar system (99.86%), and the solar radiation and solar wind are responsible for radiative and chemical processing of the surfaces and atmospheres of objects out to the farthest reaches of the solar system. Our Sun is a fairly typical G2V-type star; it has a mass of 2  1030 kg, and its luminosity of 3.8  1026 W places it squarely within the main sequence on the Hertzsprung-Russell diagram. The bulk composition of the Sun is representative of the original composition of the molecular cloud from which the solar system formed and is dominated by hydrogen (70% by mass) and helium (28%), with the remaining amount dominated by oxygen, carbon, and nitrogen. Studies of stellar evolution indicate that our Sun has an expected main-sequence lifetime of 9–10 Gyr and that it will then pass through a red giant phase and eventually become a white dwarf. Today, the amount of solar energy incident on the planets matches or exceeds their intrinsic energy production. By comparing the total power emitted by planetary bodies with the solar power at their orbital distances, we can calculate the amount of internally generated energy in each of these bodies and gain insight into their thermal states and histories. The intrinsic heat flow of the terrestrial planets is nonzero but is negligible in comparison with the solar energy they absorb. Of the giant planets, Jupiter, Saturn, and Neptune emit twice the energy they receive from the Sun, while Uranus appears to lack an internal heat source for reasons that remain a mystery.

The Terrestrial Planets Four planets orbit the Sun within 2 AU and constitute the terrestrial planets of the inner solar system: Mercury, Venus, Earth, and Mars (shown in Fig. 3). They are characterized by solid surfaces and dense compositions, in contrast to the

8 The Solar System: A Panorama

a

b Crust here Lithosp

123

Upper mantle

Silicate mantle

c

Crust + Lithosphere Crust

Lithosphere FeO-enriched silicate mantle

Lower mantle silicate

Liquid Fe–FeS outer core

Liquid Fe–FeS core

Core liquid? solid? Fe–Ni

Solid Fe–Ni inner core

Venus

Mercury

Mars

crust

upper mantle transition zone

m) th (k Dep 410 660

silic ates (sol id)

d

lower mantle

Moho 50 –150 km

Fe ñ

outer core

Ni (liqu (+S?) id)

2886

6370

Fe– N (sol i id)

5154 inner core

Earth

atmosphere crust lithosphere asthenosphere upper mantle

e

f

g

h

Fig. 2 Interior structures of the planets. The exact composition of the layers and the locations of their boundaries are still not precisely known; the figure represents our current understanding (Terrestrial planet schematics reprinted with permission from de Pater and Lissauer 2010)

124

K. de Kleer and I. de Pater

Fig. 3 The terrestrial planets. (a) Mercury (Source: NASA/JHU APL/Carnegie Institute of Washington). (b) Venus global view mosaicked from Magellan and Pioneer observations (Source: NASA/JPL). (c) Earth, as viewed by Apollo 17 (Source: NASA). (d) Mars as viewed by the Viking Orbiters in the 1970s (Source: NASA)

volatile- and gas-dominated giant planets. The densities of the inner three planets indicate predominantly rocky compositions, roughly in the range of 5.2–5.5 g/cm3 , while the density of Mars is a markedly lower 3.94 g/cm3 (see Table 2). Due to the proximity of Mars and Venus to Earth, as well as their rough similarity in size, studies of these planets have included investigations into their past and current habitability. In addition, while Mars’ cold, dry climate and tenuous atmosphere contrast starkly with Venus’ hot, dense atmosphere, both planets represent possible phases through which Earth may have passed and end-members that it may ultimately reach.

8 The Solar System: A Panorama

125

Surfaces The surfaces of the terrestrial planets and their satellites are silicate-dominated and have been processed by numerous mechanisms that provide a record of the evolutionary history of the bodies themselves and of the objects that impacted them. New crust was repeatedly created early in the solar system’s history by the emplacement of massive lava flows and was subsequently modified by numerous processes. External processes include sputtering and chemical and radiation weathering, while intrinsic processes include tectonics and atmospheric weathering. Craters on planetary surfaces record the bombardment history as a function of solar distance but can be erased by surface processes and hence provide a tool for dating surface ages. The surface of Mercury is dominated by tectonic features and heavy cratering; the lack of atmosphere and proximity to the Sun led to a high impact rate throughout its history. Mars’ surface, while also heavily cratered (at least in the southern hemisphere), shows evidence of erosion by wind and liquid water. The appearance of Venus’ surface is dominated by volcanism. Impact craters are randomly distributed over its surface, which may only be 300–600 million years in age; debates continue as to the mechanisms of its apparently global resurfacing. Earth’s tectonic features are primarily related to its tectonic plate activity and include trenches at divergent plate boundaries and mountain ranges at interacting plate edges. Although Mars has substantial polar ice deposits of both water and CO2 ice, Earth is the only planet in the solar system with stable liquid water on its surface.

Atmospheres Although all of the planets likely accreted hydrogen and helium envelopes from the solar nebula, the rocky planets lacked the mass to maintain these primary atmospheres. Secondary atmospheres were formed by a combination of large-scale outgassing from early molten interiors and delivery from farther out in the solar system via comet and asteroid impacts. There is evidence that the climates of Earth, Venus, and Mars have evolved significantly over their histories, passing through phases of vastly different atmospheric temperature, density, and composition from what we see today. Mercury’s atmosphere is negligible and is composed of atoms sputtered from the surface and captured solar wind ions. Although both Mars and Venus have CO2 dominated atmospheres with N2 present at the few percent level, Mars’ atmosphere is cold (218 K) and extremely tenuous (6 mbar), while Venus’ dense (93bar) atmosphere produces a greenhouse effect that heats the surface to a mean temperature of 735 K. On Venus, the time variability of atmospheric SO2 points to ongoing volcanic activity. Some evidence indicates that Mars’ past atmosphere may have been warm and dense enough to permit liquid water on the surface. Such an environment may have had all the conditions to support life, but the question

126

K. de Kleer and I. de Pater

of Mars’ climate history is still under active investigation. In contrast, Earth has a moderately dense (1 bar) atmosphere composed predominantly of molecular nitrogen (78% by volume) and oxygen (21%), with contributions from argon, carbon dioxide, water vapor, and other trace species. Earth’s atmosphere is unique in its abundance of free oxygen, which is converted from CO2 by organisms.

Interiors and Magnetic Fields The interior structure of all four terrestrial planets is characterized by a dense core, a mantle, and a crust (see Fig. 2). Both Earth and Mercury have a solid Fe-Ni inner core, a liquid outer core, a silicate mantle, and an outer crust. The core of Mars is likely liquid Fe-FeS throughout, while we still do not know the state (liquid or solid) of Venus’ core. Mercury’s core comprises a strikingly large fraction of its total volume. One of the many hypotheses that attempt to explain the unusual core size postulates that a planetesimal impacted Mercury after its differentiation, stripping off much of its mantle and leaving behind an object dominated by core material. Currents in the liquid outer cores of Earth and Mercury generate these planets’ intrinsic magnetic fields. Earth’s magnetic field protects its surface from the solar wind, which it holds off at a height of 10RE , the location of the bow shock. The ions populating this field originate both in Earth’s ionosphere and in the solar wind. Mercury’s magnetic field is strong enough to protect the surface from solar wind particles under typical conditions but is insufficient to hold off the solar wind at times of increased pressure. During these times, the solar wind impinges directly on Mercury’s surface. Although Mars lacks a magnetic field today, areas of localized remnant crustal magnetism indicate that it possessed a strong magnetic field during the first few hundred million years after its formation. The dynamo was likely shut off when Mars’ interior cooled to the point that turbulent convection ceased in the core. Like Mars, Venus does not currently possess an intrinsic magnetic field, although an induced field is produced by interactions between the solar wind and Venus’ ionosphere.

Satellites The Earth is the only terrestrial planet with a large, regular satellite. It is generally accepted that the Moon was formed by the collision of a Mars-sized object with the proto-Earth, at a late enough stage that differentiation into a metallic core and silicate mantle had already occurred. The cores of the two objects merged to form Earth’s core, while the mantle material was thrown into Earth’s orbit and eventually accreted into the Moon. This process resulted in a body without a significant core and with a bulk composition similar to that of Earth’s mantle. The only other terrestrial planet to host satellites is Mars, whose moons Phobos and Deimos are small in size and irregular in shape. Their compositions appear similar to

8 The Solar System: A Panorama

127

carbonaceous chondrites, suggesting that they may be captured asteroids. However, various other properties (e.g., their close-in orbits) are difficult to reconcile with such a scenario.

The Giant Planets Exploration of the giant planets by spacecraft began with the Pioneer missions in the 1970s. In 1989, Voyager 2 became the first spacecraft to complete a visit to all four of the giant planets and still remains the only spacecraft to have visited Uranus and Neptune. Subsequently, both Jupiter and Saturn were visited by orbiters (the Galileo and Cassini spacecraft, respectively). The data resulting from these missions, as well as from ground- and space-based telescope observations (Fig. 4), have shown us the dynamic atmospheres, the intricate ring and satellite systems, and the complex neutral and plasma environments of the giant planets. The giant planets can be separated into two categories – the gas giants (Jupiter and Saturn) and the ice giants (Uranus and Neptune). Within each category, the two bodies are of comparable size and mass (within a factor of 2), yet the gas giants have 10 times the mass of the ice giants. All four have some form of metallic/rocky core, a convecting conductive layer that generates a magnetic field, and an intricate system of rings and satellites.

Fig. 4 The giant planets. (a) Jupiter and (b) Saturn viewed at optical wavelengths by the Hubble Space Telescope (Source: NASA/ESA/A. Simon). False-color images of (c) Uranus (Source: I. de Pater, H. Hammel, L. Sromovsy, P. Fry) and (d) Neptune (Adapted from de Pater et al. 2014) viewed at near-infrared wavelengths by the Keck telescope

128

K. de Kleer and I. de Pater

Atmospheres The atmospheres of all four giant planets exhibit decreasing temperature with depth down to 0.1 bar (the tropopause), below which temperature increases with depth, following an adiabat below roughly 1 bar. All four atmospheres are dominated by hydrogen (80–90% by volume) and helium (10–15%) captured from the solar nebula during formation, with trace amounts of C, O, N, S, and P in the form of methane, water, ammonia, hydrogen sulfide, and phosphine. Most of the information we have on the giant planets’ atmospheres has been determined by remote sensing. However, the Galileo probe, which was sent directly into Jupiter’s atmosphere, determined the composition of this planet’s “deep” (10–20 bar) atmosphere. It found that all elements, C, N, and S, as well as Ar, Kr, and Xe, were enhanced by a factor of 4 over the solar elemental values. Water, however, was measured at a factor of 3 below the solar O abundance, which was attributed to the fact that the probe descended into an anomalously dry region. Clearly, our understanding of atmospheric composition is limited to the depths probed by the observations, which are limited to the upper few bars for optical and infrared wavelengths. Radio wavelengths extend the observational reach significantly but still leave a large portion of the atmosphere untapped. Remote sensing data of all four planets indicate that the abundances of CH4 and H2 S gases, relative to hydrogen, increase from Jupiter (factor of 4 over solar values) out to Uranus and Neptune (factor of 30–60 over solar values). Within a planetary atmosphere, a gas rises from the deep atmosphere adiabatically until it reaches its condensation temperature and condenses to form a cloud. Such clouds vary from tenuous, localized features to global, optically thick layers. A water-solution cloud is expected to be present at the deepest levels, topped by a water-ice cloud. Above the level of the water cloud, we expect an NH4 SH cloud on all four planets. Jupiter and Saturn’s upper atmospheres contain a top cloud layer of ammonia ice particles, while Uranus and Neptune’s contain a methane-ice cloud at the highest levels with a somewhat deeper (yet above the NH4 SH layer) H2 S cloud. The appearance of Jupiter and Saturn is characterized by massive storm systems and distinctive zonal wind patterns with multiple jets in each hemisphere. Neptune lacks the diverse, distinctive storm patterns of Jupiter but is spotted with omnipresent (but constantly evolving) bright cloud features and an occasional large storm. In contrast, Uranus typically appears free of bright cloud features; this difference may be due to a lesser degree of convective activity, perhaps connected to Uranus’ apparent lack of internal heat production. An excess of CO in the atmospheres of Jupiter, Saturn, and Neptune (as high as 1,000 the equilibrium abundance in the case of Neptune) may indicate external delivery from comets or from materials originating on their icy moons.

Interiors and Magnetic Fields Information on planetary interiors is difficult to obtain and relies on indirect information such as a planet’s oblateness, rotation rate, and heat flow, as well as its

8 The Solar System: A Panorama

129

magnetic and gravitational fields. Properties of the body’s shape and structure can be inferred from its gravity field, as measured in situ by spacecraft or indirectly by its effect on the orbiting moons and rings. Models to match the observed properties of Jupiter and Saturn indicate relatively small, dense cores dominated by iron and silicate, roughly five to ten Earth masses for Jupiter and somewhat larger for Saturn. The bulk of each planet’s volume is dominated by hydrogen and helium, in metallic form for the high-pressure interior and in molecular form in the shallower atmosphere (see Fig. 2 for interior diagrams of these planets). The layer of metallic hydrogen and helium generates the powerful magnetic fields of Jupiter and Saturn. Saturn’s magnetic field is somewhat weaker than Jupiter’s and is distinctive in that it is the only known magnetic field that aligns with its planet’s rotation axis. Jupiter’s magnetosphere is home to vast, dynamic systems of ions and electrons, including radiation belts and a torus of plasma that is sourced from volcanic by-products in Io’s atmosphere. As with Jupiter, Saturn is surrounded by a torus of plasma that is sourced by its satellites and rings, including material jetting out of Enceladus’ geysers. Uranus and Neptune have a larger relative abundance of heavier elements than Jupiter and Saturn and hence contain larger cores (out to perhaps 1/3 of their radii). Their interior pressures are too low for metallic hydrogen to exist, yet the presence of magnetic fields indicates a convecting conductive material, which is thought to be an ionic “ocean” between about 0.3 and 0.7 planetary radii. Although little is known about the magnetic fields of Uranus and Neptune, these planets’ magnetic axes appear to be not only tipped significantly (60ı and 47ı , respectively) relative to their rotation axes, but the magnetic dipole centers are also displaced from their centers by one third of the radius in the case of Uranus and over half the radius in the case of Neptune.

Satellites and Rings of the Giant Planets A system of satellites and ring particles orbiting a central planet resembles a miniature planetary system, with bodies spanning a range of sizes, compositions, and orbital parameters interacting through diverse processes. The giant planets host a small number of large, regular satellites and a network of smaller satellites (more than 60 in number in the case of Jupiter). Nearly all regular and close-in satellites orbit in a prograde sense, in an orbital plane that aligns with the planet’s equatorial plane to within a few degrees. They are also typically in synchronous rotation, so that the orbital and rotation periods are equal. In contrast, many of the small, outer satellites are in retrograde orbits and/or are not synchronously rotating, suggesting that these populations are dominated by captured comets, asteroids, or planetesimals. Several groups of small satellites follow similar orbits around Jupiter, suggesting that these satellites are fragments of larger bodies that were disrupted after capture. All four of the giant planets are also encircled by ring systems, although the appearance of the rings is strikingly different between planets. Most rings can be

130

K. de Kleer and I. de Pater

thought of as failed satellites: the tidal forces close to the planet are sufficiently strong to prevent debris from coalescing into a satellite or to tidally disrupt a satellite in such an orbit if the satellite’s mechanical strength is weaker than the tidal forces. Indeed, most rings are found within or near the Roche limit. Collisions between particles dissipate energy but conserve angular momentum, resulting in the flat, annular appearance of the rings. The dynamical, chemical, and even electromagnetic processes at work in planetary ring systems can be studied as small-scale analogs to the processes that shape the early disks in which planets form. In addition to these “classical” rings, some planets also have tenuous, dusty rings. While cm-m-size ring particle orbits are governed by the planet’s gravitational field, micron-sized dust is influenced by solar radiation, plasma, and electromagnetic forces, which limit their lifetimes to 103 to 105 years (for a 1 m size grained in Jupiter’s rings). Such dusty rings must therefore be young and continuously replenished by new material.

Rings Saturn’s ring system is the most dramatic and well-studied in the solar system. Its appearance is dominated by the two main (A and B) rings, which are separated by a distinctive gap. Interior to the main rings are the C and the tenuous D rings, while 3,000 km beyond the A ring are the narrow, seemingly intertwined, F ring and the dusty G and E rings. The G ring is red due to light scattered off the dust, while the E ring appears blue, indicative of a ring composed of only tiny grains. Enceladus’ water ice geysers are likely the source of the E ring material. The main rings exhibit a wide range of dynamic features caused by complex gravitational interactions between the particles themselves, as well as between the particles and Saturn and its moons. The particles are predominantly pure water ice; slight contaminations of the ice have been used to infer a particle age of 5 Earth masses orbiting on an inclined, eccentric orbit out at hundreds of AU. Such a planet could have been ejected from an orbit closer to the Sun early in the solar system’s formation or could have been captured from and/or perturbed by a nearby star. The chase is still on to actually find Planet Nine and confirm its existence; only then can we say that our solar system has nine planets (again).

The Oort Cloud Out past the scattered disk, beyond even the reaches of the heliosphere (see Fig. 7), resides a population of roughly a trillion objects that constitute the Oort cloud. Extending from 10,000 AU out to 50,000 AU or more, the Oort cloud resides far enough from the Sun that objects are perturbed by passing stars as well as the galactic tide. Such interactions often result in highly eccentric and inclined orbits. The dynamical lifetime of objects in the Oort cloud is estimated at about half the age of the solar system. The classical Oort cloud may occasionally be replenished with objects from an unseen inner Oort cloud, between 1,000 and 10,000 AU.

Comets Objects originating in the Oort cloud or Kuiper belt whose orbits are perturbed such that they pass through the inner solar system become known as comets. Short-period comets (200 year periods) come from the Oort cloud. The composition of comets, including condensed silicate grains and volatile ices, resembles the composition of dense cores within interstellar clouds, and the species present do not indicate significant subsequent processing by the solar nebula. Comets are therefore believed to be the most primitive bodies in the solar system, preserving a record of the initial conditions in the solar nebula and providing insight into the first few hundred million years of the solar system’s history. Comet nuclei have very low material strength and are likely made up of loosely bound material that impacted at low temperatures and velocities and stuck together. Their volatile-rich compositions reflect their formation in the cold outer reaches of the solar system. The volatile components are dominated by water ice but also contain ices involving CO, CO2 , CH4 , and NH3 . Complex molecules, such as methanol, formaldehyde, and even ethylene glycol, have been detected in some comets’ comae. The study of comet composition also has astrobiological implications. Bombardment of the inner solar system by icy planetesimals early in the solar system’s history may have delivered significant quantities of water and/or organic molecules to the surfaces of Earth and the other terrestrial planets. Although it is still unclear

8 The Solar System: A Panorama

139

what fraction of Earth’s water was delivered from external sources, such impacts may have played an important role in creating the life-sustaining environment that set the stage for life on Earth.

Conclusions Our exploration of the solar system we live in is still ongoing. Despite decades (and in some cases, centuries) of observation, many aspects of the planets and other bodies have eluded our understanding. The structure and composition of planetary interiors encode information on planetary histories yet can only be studied through indirect means. Investigations into outer solar system small body populations are in an exciting era of discovery due to emerging telescope technologies; the results of these investigations have the potential to discriminate between solar system formation models and reveal the processes of planet formation and subsequent orbital evolution. The hypothesized Planet Nine, orbiting many times farther out than Neptune, forces us to reevaluate our conception of the solar system’s scope. Finally, recent discoveries have revealed the prevalence of geological activity and subsurface oceans on the outer solar system’s icy worlds, paving the way for us to answer the question of whether life could have evolved, or could exist today, elsewhere in our solar system.

References Apai D, Lauretta DS (eds) (2010) Protoplanetary dust: astrophysical and cosmochemical perspectives. Cambridge University Press, Cambridge Asphaug E (2014) Impact origin of the moon? AREPS 42:551–578 Atreya SK, Pollack JB, Matthews MS (eds) (1989) Origin and evolution of planetary and satellite Atmospheres. University of Arizona Press, Tucson Bagenal F, Dowling T, McKinnon W (eds) (2007) Jupiter: the planet, satellites and magnetosphere. Cambridge University Press, Cambridge Barlow N (ed) (2008) Mars: an introduction to its interior, surface and atmosphere. Cambridge University Press, Cambridge Barucci MA, Boehnhardt H, Cruikshank DP, Morbidelli A (eds) (2008) The solar system beyond Neptune. University of Arizona Press, Tucson Batygin K, Brown ME (2016) Evidence for a distant giant planet in the solar system. Astron J 151:2 Beatty JK, Peterson CC, Chaikin A (eds) (1999) The new solar system, 4th edn. Sky Publishing Corp, Cambridge, MA Brown ME (2012) The compositions of Kuiper belt objects. AREPS 40:467–494 Bougher SW, Hunten DM, Phillips RJ (eds) (1997) Venus II: geology, geophysics, atmosphere, and solar wind environment. University of Arizona Press, Tucson Burbine TH (ed) (2016) Asteroids: astronomical and geological bodies. Cambridge University Press, Cambridge Canup RM, Righter K (eds) (2000) Origin of the earth and moon. University of Arizona Press, Tucson Chiang E, Youdin AN (2010) Forming planetesimals in solar and extrasolar nebulae. AREPS 38:493–522

140

K. de Kleer and I. de Pater

Clark PE (2015) Mercury’s interior, surface, and surrounding environment: latest discoveries. Springer, New York Cruikshank DP (ed) (1995) Neptune and Triton. University of Arizona Press, Tucson de Pater I, Lissauer JJ (2010) Planetary science, 2nd edn. Cambridge University Press, Cambridge de Pater I, Fletcher LN, Luszcz-Cook S et al (2014) Neptune’s global circulation deduced from multi-wavelength observations. Icarus 237:211–238 Elkins-Tanton LT, Weiss BP (eds) (2017) Planetesimals: early differentiation and consequences for planets. Cambridge University Press, Cambridge Esposito LW (2010) Composition, structure, dynamics, and evolution of Saturn’s rings. AREPS 38:383–410 Esposito LW (ed) (2014) Planetary rings: a post-equinox view. Cambridge University Press, Cambridge Festou MC, Keller HU, Weaver HA (eds) (2004) Comets II. University of Arizona Press, Tucson Greenberg R, Brahic A (eds) (1984) Planetary rings. University of Arizona Press, Tucson Hayes AG (2016) The lakes and seas of titan. AREPS 44:57–83 Huebner WF (ed) (1990) Physics and chemistry of comets. Springer, Berlin Irwin P (2009) Giant planets of our solar system, 2nd edn. Springer-Praxis, Chichester Lewis JS (2004) Physics and chemistry of the solar system, 2nd edn. Elsevier/Academic, San Diego Lissauer JJ, de Pater I (2013) Fundamental planetary science. Cambridge University Press, Cambridge Matthews MS, Bergstralh JT, Miner ED (eds) (1991) Uranus. University of Arizona Press, Tucson Melosh HJ (ed) (2011) Planetary surface processes. Cambridge University Press, Cambridge Miner E (1998) Uranus: the planet, rings, and satellites. Wiley, New York Morbidelli A, Levison HF (2014) Kuiper belt: dynamics. In: Spohn T, Johnson T, Breuer D (eds) Encyclopedia of the solar system. Elsevier, Saint Louis Muller-Wodarg I, Griffith CA, Lellouch E, Cravens TE (eds) (2014) Titan: interior, surface, atmosphere, and space environment. Cambridge University Press, Cambridge Nimmo F, McKenzie D (1998) Volcanism and tectonics on Venus. Annu Rev Earth Planet Sci 26:23–53 Pappalardo RT, McKinnon WB, Khurana K (eds) (2009) Europa. University of Arizona Press, Tucson Parker A, Ivezi´c Ž, Juri´c M et al (2008) The size distribution of asteroid families in the SDSS moving object catalog 4. Icarus 198:138–155 Reipurth B, Jewitt D, Keil K (eds) (2007) Protostars and planets V. University of Arizona Press, Tucson Schenk PM, Clark RN, Howett CJA, Verbiscer AJ, Waite JH (2017) Enceladus and the icy moons of Saturn. University of Arizona Press, Tucson Spencer JR, Nimmo F (2013) Enceladus: an active ice world in the Saturn system. AREPS 41:693–717 Spohn T, Johnson T, Breuer D (eds) (2014) Encyclopedia of the solar system, 3rd edn. Elsevier, Saint Louis Taylor SR, McLennan S (eds) (2008) Planetary crusts: their composition, origin and evolution. Cambridge University Press, Cambridge Tiscareno MS, Murray CD (eds) (2017) Planetary ring systems. Cambridge University Press, Cambridge, UK. (in press) Wordsworth RD (2016) The climate of early Mars. AREPS 44:381–408

9

Interiors and Surfaces of Terrestrial Planets and Major Satellites Alberto G. Fairén

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mercury . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Venus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Earth-Moon System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Io . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Europa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ganymede . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Callisto . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Titan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Triton . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

142 144 146 150 152 154 155 157 157 158 159 162

Abstract

To reliably advance in studies of extrasolar planets, or exoplanets, we need previously a solid understanding of our own planetary neighborhood. In this chapter, we will give a general overview of the internal structure, composition, surface features, and geologic units of the four terrestrial planets of the solar system: Mercury, Venus, Earth, and Mars. We will include the Earth’s Moon as part of a large Earth-Moon system. We will also revise the current knowledge of the surfaces and interiors of six most representative major satellites orbiting the giant gas planets of the solar system: Io, Europa, Ganymede, Callisto (Jupiter), Titan (Saturn), and Triton (Neptune). A. G. Fairén () Department of Planetology and Habitability, Centro de Astrobiologia (CSIC-INTA), Madrid, Spain Department of Astronomy, Cornell University, Ithaca, NY, USA e-mail: [email protected]; [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_43

141

142

A.G. Fairén

Introduction A terrestrial planet is a celestial body that is composed primarily of silicate rocks and metals and has a solid surface. The terrestrial planets of the solar system include Mercury, Venus, Earth, and Mars. All of the terrestrial planets, also known as telluric or rocky planets, are inner planets and have a similar size range: compared to the four gas giant planets that make up the outer solar system, the inner planets all have diminutive sizes (Fig. 1). The terrestrial planets in the solar system are

Fig. 1 Mass-radius relationships for solid planets. The solid lines are homogeneous planets. From top to bottom, the homogeneous planets are hydrogen (cyan solid line), a hydrogen-helium mixture with 25% helium by mass (cyan dotted line), water ice (blue solid line), silicate (MgSiO3 perovskite, red solid line), and iron (Fe, green solid line). The nonsolid lines are differentiated planets. The red dashed line is for silicate planets with 32.5% by mass iron cores and 67.5% silicate mantles (similar to Earth), and the red dotted line is for silicate planets with 70% by mass iron core and 30% silicate mantles (similar to Mercury). The blue dashed line is for water planets with 75% water ice, a 22% silicate shell, and a 3% iron core; the blue dot dashed line is for water planets with 45% water ice, a 48.5% silicate shell, and a 6.5% iron core (similar to Ganymede); the blue dotted line is for water planets with 25% water ice, a 52.5% silicate shell, and a 22.5% iron core. The blue triangles are solar system planets, from left to right: Mars, Venus, Earth, Uranus, Neptune, Saturn, and Jupiter. The magenta squares denote the transiting exoplanets (From Seager et al. 2007)

9 Interiors and Surfaces of Terrestrial Planets and Major Satellites

143

Fig. 2 Relative sizes and positions of the planets and dwarf planets in the solar system (NASA/Planets2008/Wikimedia Commons)

Fig. 3 Interior structures of the terrestrial planets and the Earth’s Moon (NASA)

characterized for orbiting relatively close to the Sun, having few or no moons, and lacking ring systems (Fig. 2). All the terrestrial planets have similar inner structures: each has a dense ferrous core, a silicate mantle, and a crust (Fig. 3). The temperatures of the inner planets are low enough that rock exists mostly as a solid at the surface, and these rocky surfaces feature mountains, volcanoes, plains, valleys, impact craters, and other formations. When they have atmospheres, these are secondary atmospheres, generated through volcanism and/or cometary and meteoritic impacts. The Moon is also a terrestrial body, though it is not a planet. It has the same basic internal structure as the terrestrial planets, but with a much smaller metallic core. The Moon is unique among the other solar system satellites, because its mass and orbital radius are significantly larger relative to the mass and radius of the planet around which it orbits than those of the other satellites. The major satellites of the giant gaseous planets are Europa, Io, Ganymede, and Callisto orbiting Jupiter; Mimas, Enceladus, Tethys, Dione, Rhea, Iapetus, and Titan

144

A.G. Fairén

orbiting Saturn; Miranda, Umbriel, Ariel, Titania, and Oberon orbiting Uranus; and Triton orbiting Neptune. These bodies are all large enough for self-gravity to make them round (Titan and Ganymede are even larger than the planet Mercury) (Fig. 4). They are similar to terrestrial planets because they do have a solid surface (Fig. 5). Here we will review the main traits of the six most representative of the outer solar system major satellites: Io, Europa, Ganymede, Callisto, Titan, and Triton.

Mercury Mercury is the innermost planet in the solar system, and it rotates exactly three times around its axis for every two orbits around the Sun. As a result, Mercury has the greatest range (day to night) of surface temperatures of any planet or satellite in the solar system because of its proximity to the Sun and its long solar day, ranging from about 723 K at the equator when it is closest to the Sun to about 90 K at night just before dawn. In addition, Mercury lack seasons because the axis of Mercury has the smallest tilt of all other planets. Mercury contains a much larger fraction of iron than any other planet or satellite in the solar system. Thus, the iron-nickel alloy forming the core must be about 75% of the planet diameter, or some 42% of its volume. Its rocky outer region is only about 600 km thick, with an average crustal thickness of 35 ˙ 18 km (Padovan et al. 2015). In general, the surfaceTerrestrial planets of Mercury can be divided into three major terrains: (1) heavily cratered regions, (2) intercrater plains, and (3) smooth plains. The heavily cratered uplands record the period of heavy meteoroid bombardment, a catastrophic event that peaked about 3.9 billion years ago and ended about 3.8 billion years ago (Fig. 6). Therefore, the age of this surface is probably about 3.9 Ga. The two plain units of Mercury have been interpreted to be volcanic (Spudis and Guest 1988). The older intercrater plains are the most extensive terrain on Mercury. They were emplaced over a range of ages contemporaneous with the period of late heavy bombardment. They are thought to be volcanic plains erupted through a fractured crust during the first 800 million years of Mercury’s evolution. The younger smooth plains are primarily on the interior and exterior of large impact basins. They are similar to the lunar maria and are therefore interpreted to be lava flows erupted relatively late in the geologic history of Mercury (Strom and Sprague 2003). They may have an average age of about 3.8 billion years as indicated by crater abundances. In addition, the tectonic framework of Mercury is unique in the solar system. It consists of a system of thrust faults called lobate scarps (Fig. 7). They are more-or-less uniformly distributed, suggesting that Mercury was subjected to global compressive stresses. Stratigraphic evidence indicates that the system of thrust faults formed after the emplacement of intercrater plains materials, relatively late in the geologic history of Mercury. The faults were probably caused by a decrease in the size of the planet due to cooling. Mercury may have always been void of an atmosphere, as numerous fresh impact craters with bright radial aureoles, which provide a relative age estimated to date

Fig. 4 Selected moons of the solar system, with Earth for scale. Nineteen moons are large enough to be round, and Titan has a substantial atmosphere (NASA/Wikipedia)

9 Interiors and Surfaces of Terrestrial Planets and Major Satellites 145

146

A.G. Fairén

Fig. 5 Interior structures of the largest moons in the solar system (NASA/Brian Wang)

back as far as 4 Gy (Strom and Neukum 1988), would otherwise have been modified or erased even under atmospheric pressures as low as those recorded for Mars. However, near its poles, frozen water ice has been discovered within permanently shadowed impact craters (Slade 1992; Harmon et al. 2001), where volatiles that include water are stable for long periods of time at temperatures of about 112 K (Chabot et al. 2014). The area covered by the ice deposits is estimated to be about 10,000 km2 (Slade et al. 1992). Though the ice probably originated from relatively recent comet impacts, it is possible that the remnant waters could archive early solar system information if the impacts and partial water infilling occurred relatively early in the development of the planet (e.g., shortly after the formation and hardening of the crust).

Venus All planets in the solar system rotate counterclockwise on their axis, except Venus and Uranus which rotate clockwise. The 93 bars atmosphere of Venus consists of 97% CO2 , and sulfuric acid as a significant component of observed cloud particles, making it extremely thick and acidic. The upper parts of the clouds extend 80 km

Fig. 6 Size comparison of large craters of the solar system (Ohio State University)

9 Interiors and Surfaces of Terrestrial Planets and Major Satellites 147

148

A.G. Fairén

Fig. 7 The surface expression of thrust faulting on the Moon (left, LROC data) or Mercury (right, MESSENGER) is an arcuate scarp. The lunar scarps rise only about 100 m above the surrounding terrain and thus are smaller than their Mercurian counterparts, often rising a kilometer above the surrounding terrain (NASA)

above the surface, and the upper atmosphere is in general super rotating, with wind speeds up to 100 ms1 and circling the planet as fast as in four Earth days (Limaye et al. 2009), while other parts of the deep atmosphere seem to move much more slowly (Fukuhara et al. 2017). The acid components remain in the atmosphere because they evaporate at the base of the cloud cover, about 50 km above the surface. Accordingly, the climate on Venus is controlled by the radiative properties of its global cloud cover and a carbon dioxide-water greenhouse effect (Bullock and Grinspoon 2001). Surface temperatures exceed 460 ı C, and they are nearly identical on the planet’s daytime and nighttime and from equator to pole, because of (1) the high heat conductance of the very dense atmosphere and (2) the absence of seasons due to the very small tilt of only 3.39ı with respect to the Sun. As a result, the crust is extremely desiccated. It is impossible to observe the extremely dry surface of Venus from Earth or orbiting spacecrafts without the use of radar because of the cloud envelope. Venus has experienced at least one, and probably multiple, episodes of global resurfacing, as inferred from the impact cratering record (Strom et al. 1994): out of a total of only 967 impact craters on the whole surface of Venus, 84% are in pristine condition, 11% are disrupted by fractures, and only about 3% have been flooded with lava. In addition, these craters are randomly distributed across the surface of the planet and with respect to elevation and size. Most of the geologic units on Venus have been formed since the end of the last global resurfacing event, between 500 ma. and 1 Ga. ago. Geologic mapping of the surface of Venus reveals at least five major stages of geologic activity, each of indefinite duration (Tanaka et al. 1997; Ivanov and Head

9 Interiors and Surfaces of Terrestrial Planets and Major Satellites

149

2001). These stages can be characterized by the following dominant materials, from youngest to oldest: 1. Flow materials from near 100.000 small shield volcanoes (Fig. 8), local flows, and smooth and rough plain units 2. Younger regional plain units 3. Widespread intermediate plain units 4. Concentrically fractured coronae and older plain units (Fig. 9) 5. Tessera material, including rafted and highly deformed crustal materials recording earlier evolutional phases of planetary development Stages 1–4 record a significant areal resurfacing. The Venusian lithosphere is about 30 km thick on average, and the crust is basaltic (Nimmo and McKenzie 1998). The underlying asthenosphere is less ductile than that of Earth (Huang et al. 2013) and thus not likely to support plate tectonics. In fact, SO2 is a million times more abundant in the Venusian atmosphere than in the Earth’s atmosphere, pointing to the existence of major volcanic hot spots spewing out gases such as SO2 , perhaps for billions of years, and it is very likely that Venus

Fig. 8 Volcanoes in Venus, Earth, Mars, and Io (satellite of Jupiter). Volcanoes pepper the entire surface of Venus, and maybe some are still active. Mars has two big volcanic complexes, both inactive, and many minor volcanoes. At least eight active volcanoes have been mapped on Io, with almost perfect plume arcs due to less gravity and no air resistance (NASA)

150

A.G. Fairén

Fig. 9 Magellan radar image showing the near-circular trough of Artemis Chasma, the largest corona on Venus. Image spans 3,500 km from east to west (NASA/JPL)

is still volcanically active today (Shalygin et al. 2015) (Fig. 8). The mantle reaches to a depth of roughly 3,000 km, and the core comprises a liquid iron-nickel alloy.

The Earth-Moon System A gigantic impact of a Mars-sized body with the early Earth (Hartmann et al. 1986), or maybe by a series of different smaller collisions (Rufu et al. 2017), formed the Moon about 4.5 billion years ago. This mode of origin would explain the inclined lunar orbit, the high spin of the Earth-Moon system, the composition of the Moon similar to the Earth’s mantle and its low density compared to Earth’s, and the very low volatile and high refractory elemental abundances. The resulting center of mass of the pair is offset from the center by about 1.8 km toward the Earth, because the outer part of the Moon was molten and differentiated into an anorthositic crust and basic to ultrabasic mantle, and the anorthositic crust is thinner on the frontside than the farside. If the Moon has an iron core, it is very small, no more than about 400 km diameter. Meanwhile, the Earth arrived at a structural, compositional, and dynamic state quite similar to that of today within only a few hundred million years since the planet’s formation (Carlson et al. 2014). There are two primary modes of heat transport inside the Earth today: plate tectonism and mantle convection. The Moon’s heavily cratered highlands, similar to the surfaces of Mercury and Mars (Fig. 6), inform of a period of heavy bombardment that ended about 3.8 Ga ago and that affected all of the inner solar system bodies (Zellner 2017). In addition

9 Interiors and Surfaces of Terrestrial Planets and Major Satellites

151

Fig. 10 River networks on the Earth, Mars, and Titan. Rivers on Earth are constantly adapting to changes in topography derived from plate tectonics; on the contrary, Titan and Mars have not undergone any active plate tectonics in its recent past, and river flow is controlled by other processes such as ice thickness (Titan) and impact cratering (Mars) (B. Black/NASA/Visible Earth/JPL/Cassini RADAR team)

to the bombardment, the surface of Earth is and has been continuously modified by high erosion rates, magmatic-driven activity, and plate tectonism (Black et al. 2017) (Fig. 10). Based on the significant amount of liquid water still present, water on the early Earth is usually assumed as a given (Fig. 10). Zircons, minerals as old as 4.3–4.4 Ga and thought to be formed by magma encountering surface water, are another clear indicator of water on the early Earth (Reimink et al. 2016). The origin of the water on Earth could be both endogenous outgassing and exogenous inputs from meteorites and comets (Hartogh et al. 2011). Contrary to the Earth, the Moon cannot retain an atmosphere because the surface gravity is too low and the daylight temperatures are too high. The maximum and minimum surface temperatures are about 390 K and 104 K. However, the Moon has polar ice deposits in the permanently shadowed areas of craters in the polar regions (Ingersoll et al. 1992; Feldman et al. 1998), although the deposits in the Moon show an ice concentration of only about 1.5% weight fraction, considerably smaller than Mercury’s (Feldman et al. 1998). After the period of heavy bombardment, the Moon experienced a period of basaltic volcanism. Because of the thicker crust on the farside, the equipotential surface is closer to the surface on the frontside, and therefore magmas that originate at the same level can reach the surface more easily on the frontside, explaining why

152

A.G. Fairén

the mare basalts are concentrated on the frontside. As the Moon cooled, compression within the crust occurred, which can be taken up by compacting loose surface material, reducing pore space and triggering thrust faulting (Fig. 7). All internal activity on the Moon ceased about 2–3 Ga. ago. The evolution of the Earth was quite different. Due to a 30% less luminous Sun more than 4 Gy ago, mean surface temperatures during the Archaean should have been below the freezing point of water (Sagan and Chyba 1997) if the atmospheric greenhouse effect was the same as today. However, the earliest atmosphere of Earth contained large amounts of carbon dioxide (the key greenhouse gas in the prebiotic Earth), water vapor, and methane (although methane was probably scarce prior to the origin of life), derived mostly from mantle degassing (Fig. 8), and these gases provided the required greenhouse effect to warm the surface (Kasting 1988). Since cyanobacteria invented oxygenic photosynthesis about 2.4–2.1 Ga ago, a dramatic increase of global oxygen content occurred and has continued to increase to a current level of about 21% (Holland 1999; Farquhar et al. 2000; Canfield et al. 2000; Bekker et al. 2004) with carbon dioxide and methane being atmospheric trace compounds only.

Mars The surface of Mars is divided into two major topographical areas: the ancient southern highland province, which includes extremely ancient geologic terrains marked by pervasive impact craters (Fig. 6) and magnetic anomalies (Acuña et al. 1999), and the more recent northern lowlands, which highlight the Vastitas Borealis Formation that was emplaced during the Early Amazonian (Tanaka et al. 2003). The surface geology of Mars is vastly diverse, including 1. 2. 3. 4. 5. 6. 7.

Polar ice caps (Fig. 11) Gigantic canyons (e.g., Valles Marineris) Volcanoes of diverse sizes and shapes (Fig. 8) Ice deposits within impact basins Mountain ranges (e.g., Thaumasia Highlands) Putative ancient accreted terrains and associated volcanism Regions indicating potential hydrologic activity (such as spring-fed seeps, Squyres et al. 2008) 8. Chaotic terrains (e.g., broken terrain comprised of kilometer-scale mesas marking the source regions of the circum-Chryse outflow channel system) Since erosional features of liquid water are clearly visible on the surface of Mars (Baker et al. 2015) (Fig. 10), there has been much speculation that the Martian atmosphere was significantly thicker early in the history of the solar system. Recent results from the rovers Spirit, Opportunity, and Curiosity also support the presence of ancient water on the surface of Mars (e.g., Grotzinger et al. 2015). During this early epoch, liquid water was stable on the surface, and a magnetosphere shielding the planet from harsh radiation exposure is assumed (Terada et al. 2009). Even today,

9 Interiors and Surfaces of Terrestrial Planets and Major Satellites

153

Fig. 11 Orbital view of the north polar region of Mars. The ice-rich polar cap is approximately 1,000 km across and is riven with spiral-shaped deep troughs. (NASA/JPL-Caltech/MSSS)

liquid water can transiently exist in rock pores near the surface when ice is melted by the Sun. Brines can even exist longer at temperatures well below the freezing point of water (Fairén et al. 2009). Interestingly, Mars has seasons twice as long those of Earth, because Mars is tilted on its axis by about 25.19ı , similar to the axial tilt of the Earth (22.5ı ); however, Mars’ seasons are more extreme than Earth’s because its elliptical, oval-shaped orbit around the Sun is more elongated than that of any of the other major planets. Weather patterns include winds, dust storms, frost, and fog. For decades, the prevailing paradigm was that the surface of early Mars was kept “warm and wet” for extended periods by the incorporation of different greenhouse gases into the atmosphere. These greenhouse gases would have compensated for the lesser solar heat received by the planet because of the reduced solar intensity in early solar system history, about a third less than today (Gough 1981). But stateof-the-art, 3D global simulations of the early Martian climate are confirming today that previous models were substantially overestimating the amount of warmth that a greenhouse atmosphere on early Mars could have produced. The latest published results confirm that only special amounts or combinations of greenhouse gases would have been able to produce a “warm and wet” climate on early Mars, raising the mean temperature on the surface of the planet above 273 K, only very transiently (not a “steady-state”; see Wordsworth et al. 2017). As global geomorphological and mineralogical evidence has confirmed that early Mars was indeed “wet” for long periods, it is necessary to find an explanation to the presence of substantial amounts of liquid water on the surface of early Mars under near-freezing conditions. Advanced geochemical refinements are confirming the possibility that a vigorous hydrological cycle was active during the Noachian even under mean freezing temperatures (Fairén 2010), thus completing a new perspective to the old problem of water on early Mars. But a complete understanding of the early Martian environment is still a work in progress, and novel ideas are continuing to develop on this issue. Ultimately, liquid water on early Mars most likely was the result of the combination of diverse astronomical, geochemical, and geological factors, such as the existence

154

A.G. Fairén

of briny waters in a planet occasionally heated by obliquity transitions, together with warming mechanisms like volcanism and clathrate destabilization providing greenhouse gases such as CO2 , CH4 , and H2 O. It is unlikely that the presence of liquid water on early Mars responds to just one single and unique process. How did Mars transition from its origin as a rocky planet endowed with standing reservoirs of liquid water into the very cold and hyperarid planet it is today? Too small to hold a significant atmosphere, and too far from the Sun to receive enough heat, the oceans of cold water froze and sublimated and were lost to space or became sequestered beneath the surface. The global drastic change is thought to have occurred about 3.2–3.0 Ga ago, at the end of the Hesperian, after the endogenic activity steadily decreased, the magnetosphere collapsed, and the atmosphere/hydrosphere were mostly lost to space (Clifford and Parker 2001). Also, around that time or even before, whatever plate tectonic activity (if any) had been occurring probably ceased (Black et al. 2017) (Fig. 10). All these phenomena seem to support the notion that the Martian center has cooled. Geological structure is mainly rock and metal, with a mantle below the crust comprising iron oxide-rich silicate and a core composed of an iron-nickel alloy and iron sulfide.

Io Io is the innermost of the four major moons of Jupiter and orbits the gaseous giant closer than the Moon does around the Earth. Composed primarily of silicate rock and iron, Io is internally differentiated, and its inner structure includes a dense molten iron-sulfur core making up 20% of the Moon’s mass, a partially melted silicate rock mantle rich in magnesium minerals, and a thin rock crust composed of basalt and sulfur deposited by volcanism (McGovern et al. 2016). Io is the denser moon in the solar system. Being so close to Jupiter, and also because of Io’s orbital resonance with Europa and Ganymede, the continuous warming produced by the tidal forces results in extensive and continuous volcanism on the surface. The surface of Io shows 500– 700 volcanic centers, with over 400 active volcanic calderas (Hamilton et al. 2013); some of them are as high as 16 km, and lava flows and lakes are also common. Some geysers expel sulfur gases up to 500 km high. As such, Io is the most volcanically active object known in the universe (Fig. 8) and hosts the most powerful persistently active volcano in the solar system, Loki Patera (de Kleer et al. 2017). Though the surface temperatures are very low, there is no water ice on the surface of Io, which is the driest planet/satellite in the solar system; sulfur dioxide frost is common across the satellite’s surface (Dundas 2017). The constant rejuvenation because of volcanic activity explains why Io’s surface is almost completely lacking in impact craters.

9 Interiors and Surfaces of Terrestrial Planets and Major Satellites

155

Europa The internal structure of Europa would include a dense core of metal or metal sulfides, a rocky mantle, and a low-density ice crust or ice-crusted ocean with a thickness of 80–170 km. The presence of a subsurface liquid water ocean in Europa has been inferred from the observations of: 1. Surface fracture features consistent with traveling ice pieces mobilized by subsurface liquid (Carr et al. 1998; Hoppa et al. 1999) 2. Magnetic fields induced by eddy currents in an inner mobile and conducting medium (Khurana et al. 1998) 3. The asynchronous rotation of Europa implying a friction generated by subsurface material (Geissler et al. 1998) 4. Doppler tracking of the Galileo spacecraft, suggesting a differentiated internal structure (Anderson et al. 1998) 5. Models for the origin of the larger outer satellites of Jupiter (Consolmagno and Lewis 1976) Recently, ongoing and varying eruptions of plumes of water vapor from Europa’s surface have been spotted with the Hubble Space Telescope (Roth et al. 2014) (Fig. 12). The ocean bottom environment on Europa may resemble that on Earth where hydrothermal discharge areas on the aphotic ocean bottom support hydrothermal vent communities (Vinogradov et al. 1996; Amend and Shock 1998), as Europa’s metallic core can be assumed to provide internal heat through radioactive decay, subjecting the ocean floor to volcanic eruptions. This geothermal heat served McKinnon (1999) to interpret the chaos regions on Europa as resulting from convection: large convection cells, generated from the thermal gradient existing between the icy surface and the volcanic ocean floor, would circulate water in the inner ocean, from the warm bottom areas to beneath the outer icy crust (Fig. 13). Jupiter’s strong gravitational attraction may also contribute to generate tidal channels both on the ocean floor and in the ice ceiling (McKay 2008). Hydrated materials, including sulfate salts, have been revealed using nearinfrared remote sensing of the surface of Europa (Dalton 2003). Surface coloration on Europa’s ice suggests that the saline content of its ocean is high, and different pH systems have been suggested, all with sulfate as the most common anion: a neutral pH system consisting of Na-Mg-Ca-SO4 -Cl-H2 O (Kargel et al. 2000), an alkaline system of Na-K-Cl-SO4 -CO3 -H2 O (Marion 2001), and an acidic pH system consisting of Na-H-Mg-SO4 -H2 O (Marion 2002). Actually, the inner ocean may have a redox balance similar to Earth’s (Vance et al. 2016).

156

A.G. Fairén

Fig. 12 Composite image showing plumes of water vapor erupting at the 7 o’clock position off the limb of Europa and rising over 160 km above its icy surface. The plumes were seen in silhouette as the moon passed in front of Jupiter. The water is believed to come from a subsurface ocean on Europa. Similar features have also been identified in Saturn’s moon Enceladus (NASA/ESA/W. Sparks (STScI)/USGS Astrogeology Science Center)

Fig. 13 Comparison of the surfaces of the three icy Galilean satellites, Europa, Ganymede, and Callisto, scaled to a common resolution of 150 m per pixel. Europa has a sparsely cratered surface and ridged plains, indicating that geologic activity took place recently. Ganymede’s landscape is widely formed by impacts and includes much tectonic deformation. Callisto is saturated with impact craters but is also covered by a dark material layer of so far unknown origin that erodes or covers small craters (NASA/JPL)

9 Interiors and Surfaces of Terrestrial Planets and Major Satellites

157

Ganymede Ganymede is the largest satellite in the solar system, comparable in size to Mercury. The surface of Ganymede shows two differentiated halves, recording a complex geologic history (Figs. 4 and 13): about 34% of its surface is a densely cratered and dark ancient terrain, composed by silicates and ice; and the rest is a much younger bright terrain, flooded by now frozen low-viscosity aqueous lavas (Schenk et al. 2001), and probably reshaped by tectonic and volcanic activity later on (Collins et al. 2013). Ganymede is composed by silicates and water in the same proportions, and its inner structure is differentiated into inner core, convecting mantle and outer rigid crust, as suggested by its mean density of 1.94 g/cm3 which increases with increasing depth (Showman and Malhotra 1999). The rocky core is probably differentiated, with a very big inner metallic core. The presence of a magnetic field in Ganymede, confirmed by data from Galileo spacecraft (Kivelson et al. 1998), strongly argues for an inner ocean to exist, as the heat released from the Moon’s interior is huge enough to create an inner molten core sustaining a core dynamo. Alternatively, the magnetic field may be generated by the interaction of the dynamo with the salty ocean. Ganymede may harbor an inner ocean because the combination of four factors: (1) the heat released from the core, which can melt part of the ice inside Ganymede; (2) the tidal heating induced by the proximity of Jupiter, which will reduce ice viscosity (McKinnon 1999); (3) the stability against convection (Ruiz 2001); and (4) the addition of antifreeze substances that can reduce the freezing point of water ice, such as ammonia hydrates and/or salts (see Cassen et al. 1982; Stevenson 1998; Kargel et al. 2000). Actually, Ganymede is covered with water ice in a proportion of up to 90% ice by mass (McKinnon and Parmentier 1986), with the addition of hydrated salt minerals, mostly MgSO4 likely derived from a subsurface brine-laden and layered ocean (Vance et al. 2014). The inner ocean may be a compelling explanation for the young appearance of the brighter terrains.

Callisto Callisto may also harbor an ancient liquid water internal ocean some 20 km thick, hidden 100 km below the surface (Lindkvist et al. 2015). There is indirect geological evidence for such an ocean. On one hand, magnetometric data have been explained by a dynamo effect resulting from the electronic conductivity of salty water (Khurana et al. 1998; Liuzzo et al. 2015). On the other hand, the Voyager and Galileo missions provided images demonstrating that Callisto’s crust age dates back as far as 4 Gyr, making this moon the most densely cratered

158

A.G. Fairén

object in the solar system, with no volcanic or tectonic landforms. While in similar-sized bodies (Mercury, the Moon) big impacts create significant tectonic deformations and hummocky terrains in the antipodal surface crust, the antipodal zone of Callisto’s greatest impact crater, Valhalla, is not different from the rest of the surface: this suggests the absorption of the impact energy and the dissipation of the seismic waves by an internal ocean (Watts et al. 1991; Hendrix and Johnson 2008). The ocean in Callisto is generated and kept from freezing by the heat provided by the decay of radioactive elements in the core, assuming that the crust of Callisto can be resistant to internal heat loss by convection (Ruiz 2001). This ocean stays comprised between two thick ice layers (solid convecting regions, Khurana et al. 1998) that preclude heat circulation. Callisto’s surface contains an average of 50% ice and shows low albedo features formed by contamination with darker materials (Showman and Malhotra 1999) (Fig. 13). Up to 50% of the Moon’s radius in size may be occupied by the central silicate core (Anderson et al. 2001).

Titan Titan is the largest Saturn’s moon, and its 1.5 bar atmosphere is thicker than Earth’s atmosphere, extending high up to 600 km over the surface. The reducing atmosphere, consisting primarily of nitrogen (95%) and methane (5%), is similar to the one which is believed to have existed on early Earth. Methane clouds have been detected at Titan (Roe et al. 2002), and methane rain is consistent with modeling results (Tokano et al. 2001; Chanover et al. 2003). Surface temperatures reach only about 95 K. The geothermal heat flow on Titan has been estimated as 4 mWm2 (Lorenz and Shandera 2001), and the relative lack of heavy elements generate minor radioactivity, suggesting the presence of only meager volcanic activity. However, geothermal activity could exist in some regions on Titan (e.g., Lorenz 2002). Volcanism, together with meteorite impacts, are the most likely energy sources providing heat to the surface. Episodes of fluid chemistry may have been triggered the appearance of hydrocarbon lakes on Titan’s surface by both mechanisms, and some of these episodes lasted perhaps thousands of years before freezing over (Mitri et al. 2007). Methane lakes are mainly concentrated near the southern pole (Fig. 14). Ethane and methane are available as nonpolar solvents on Titan’s surface, together with traces of dissolved N2 (Jakosky 1998), and a methane cycle with some similarities to the Earth’s hydrological cycle (including rain, rivers, and lakes; see Figs. 10 and 14) exists on Titan. HCN-based polymers, relevant for possible prebiotic reactions, may have formed on the surface from products of atmospheric chemistry (Rahm et al. 2016).

9 Interiors and Surfaces of Terrestrial Planets and Major Satellites

159

Fig. 14 False-color image of Ligeia Mare, the second largest known body of liquid on Titan. Titan’s maria (seas) are abundant on Titan’s north polar region and are filled with liquid hydrocarbons, such as ethane and methane. In the solar system, only Titan and the Earth have liquid deposits on their surfaces (NASA)

An ammonia-water mixture can be forming a subsurface ocean (Fortes 2000), staying liquid at very low temperatures because ammonia mixed with water would act as an “antifreeze,” and the ammonia-water mixture would be a polar solvent. This large subsurface ocean, 30–40 km deep, may extend up to south latitudes of about 50ı (Iess et al. 2014).

Triton It has been suggested that Triton originated from the family of the Kuiper Belt objects, i.e., it was an independent heliocentric body, eventually gravitationally captured by Neptune thousands of millions of years ago (Masters et al. 2014). This hypothesis is supported by two facts: Triton is the only big satellite in the solar system rotating retrograde, and it has a big rocky nucleus as suggested by its density (>2 g/cm3 ), much higher than those of the rest of the icy satellites of Saturn and Uranus. Hidden at a depth of a few hundreds of kilometers, Triton may harbor an inner ocean. The surface layer (composed of water, nitrogen, methane, carbon monoxide, and carbon dioxide ices) acts as a thermal insulator because of its very low thermal conductivity, lower than water ice (McKinnon et al. 1995), triggering an important temperature elevation in the near subsurface. The maintaining of the inner heat favors the stability of the subsurface ocean at depths probably not deeper than

160

A.G. Fairén

Fig. 15 Triton’s south polar terrain. About 50 dark plumes mark what may be ice volcanoes (NASA/JPL)

200–300 km (Ruiz and Fairén 2005) or 350–400 km (McKinnon et al. 1995). If the ice also contains antifreeze substances such as ammonia, capable to lower the water ice fusion point to 176 K, the stability of the inner ocean may be reached at depths of only 30–40 km. Obliquity tidal dissipation in the subsurface ocean heats Triton, causing convective yielding and the observed young surface age (Nimmo and Spencer 2015). Craterization rates also suggest very recent crustal recycling processes. Triton’s crust is composed of iced nitrogen with traces of methane, carbon monoxide, and carbon dioxide (Cruikshank et al. 1993). The sublimation and condensation of nitrogen on the surface makes Triton the coldest body in the solar system, with a constant mean temperature of 38 K (Elliot et al. 1998). Long-lasting active geysers on the surface belch out liquid nitrogen and dust particles (organic polymers, hydrocarbons), resulting from gas decompression (Fig. 15). Triton, Titan, and Earth are the only bodies in the solar system which atmospheres are constituted basically by nitrogen. Triton has a very tenuous N2 atmosphere (16 bar), peppered with thin clouds of nitrogen ice crystals (Soderblom et al. 1990) (Table 1).

Diameter (km)

4,878 12,103 12,675 3,474 6,778 3,642 3,120 5,268 4,820 5,150 2,706

Body

Mercury Venus Earth Moon Mars Io Europa Ganymede Callisto Titan Triton

Density (g/cm3 )

5.44 5.25 5.52 3.34 3.93 3.53 3.01 1.94 1.84 1.88 2.06

Mass (kg)

3.301  1023 4.868  1024 5.972  1024 7.342  1022 6.417  1023 8.931  1022 4.799  1022 1.482  1023 1.076  1023 1.345  1023 2.140  1022 0.38 0.90 1.00 0.16 0.38 0.18 0.13 0.15 0.17 0.14 0.08

Surface gravity (g)

58.65 243 1 28 1 (C 400 ) 1.77 3.5 7.1 16.7 15.9 5.8

Rotation period (days)

87.97 225 365 28 687 1.77 3.5 7.1 16.7 15.9 5.8

Orbit (days)

Table 1 Main geophysical parameters of the terrestrial planets and major satellites of the solar system

Nearly (2/3) No No Yes (Earth) No Yes (Jupiter) Yes (Jupiter) Yes (Jupiter) Yes (Jupiter) Yes (Saturn) Yes (Neptune)

Rotation-orbit synchronized?

Mean surface temperature (ı C) 167 464 15 20 65 163 171 163 139 180 235

0 92 1 0 0.01 0 0 0 0 1.45 0

Surface pressure (bars)

9 Interiors and Surfaces of Terrestrial Planets and Major Satellites 161

162

A.G. Fairén

References Acuña MH, Connerney JE, Ness NF, Lin RP, Mitchell D, Carlson CW, McFadden J, Anderson KA, Reme H, Mazelle C, Vignes D, Wasilewski P, Cloutier P (1999) Global distribution of crustal magnetization discovered by the Mars Global Surveyor MAG/ER experiment. Science 284:790 Amend JP, Shock EL (1998) Energetics of amino acid synthesis in hydrothermal ecosystems. Science 281:1659–1662 Anderson JD, Schubert G, Jacobson RA, Lau EL, Moore WB, Sjogren WL (1998) Europa’s differentiated internal structure: inferences from four Galileo encounters. Science 281:2019– 2022 Anderson JD, Jacobsen RA, McElrath TP, Schubert G, Moore WB, Thomas PC (2001) Shape, mean radius, gravity field, and interior structure of Callisto. Icarus 153:157–161 Baker VR, Hamilton CW, Burr DM, Gulick VC, Komatsu G, Luo W, Rice JW, Rodriguez JAP (2015) Fluvial geomorphology on Earth-like planetary surfaces: a review. Geomorphology 245:149–182 Bekker A, Holland HD, Wang P-L, Rumble D III, Stein HJ, Hannah JL, Coetzee LL, Beukes (2004) Dating the rise of atmospheric oxygen. Nature 427:117–120 Black BA, Perron JT, Hemingway D, Bailey E, Nimmo F, Zebker H (2017) Global drainage patterns and the origins of topographic relief on Earth, Mars, and Titan. Science 356(6339): 727–731 Bullock MA, Grinspoon DH (2001) The recent evolution of climate on Venus. Icarus 150:19–37 Canfield DE, Habicht KS, Thamdrup B (2000) The Archaean sulfur cycle and the early history of atmospheric oxygen. Science 288:658–661 Carlson RW, Garnero E, Harrison TM, Li J, Manga M, McDonough WF, Mukhopadhyay S, Romanowicz B, Rubie D, Williams Q, Zhong S (2014) How did early Earth become our modern world? Annu Rev Earth Planet Sci 42:151–178 Carr MH et al (1998) Evidence for a subsurface ocean on Europa. Nature 391:363–365 Cassen PM et al (1982) Structure and thermal evolution of the Galilean satellites. In: Morrison D (ed) Satellites of Jupiter. University of Arizona Press, Tucson, pp 93–128 Chabot NL, Ernst CM, Denevi BW, Nair H, Deutsch AN, Blewett DT, Murchie SL, Neumann GA, Mazarico E, Paige DA, Harmon JK (2014) Images of surface volatiles in Mercury’s polar craters acquired by the MESSENGER spacecraft. Geology 42(12):1051–1054 Chanover NJ, Anderson CM, McKay CP, Rannou P, Glenar DA, Hillman JJ, Blass WE (2003) Probing Titan’s lower atmosphere with acousto-optic tuning. Icarus 163:150–163 Clifford SM, Parker TJ (2001) The evolution of the Martian Hydrosphere: implications for the fate of a primordial ocean and the current state of the northern plains. Icarus 154(1):40–79 Collins GC, Patterson GW, Head JW, Pappalardo RT, Prockter LM, Lucchitta BK, Kay JP, (2013) Global geologic map of Ganymede. USGS Scientific Investigations Map, 3237 Consolmagno GJ, Lewis J (1976) Structural and thermal models of icy Galilean satellites. In: Gehrels T (ed) Jupiter. University of Arizona Press, Tucson, pp 1035–1051 Cruikshank DP, Roush TL, Owen TC, Geballe TR, de Bergh C, Schmitt B, Brown RH, Bartholomew MJ (1993) Ices on the surface of Triton. Science 261:742–745 Dalton JB (2003) Spectral behavior of hydrated sulfate salts: implications for Europa mission spectrometer design. Astrobiology 3:771–784 Dundas CM (2017) Effects of lava heating on volatile-rich slopes on Io. J Geophys Res Planets 122(3):546–559 Elliot JL, Hammel HB, Wasserman LH, Franz OG, McDonald SW, Person MJ, Olkin CB, Dunham EW, Spencer JR, Stansberry JA, Buie MW, Pasachoff JM, Babcock BA, McConnochie TH (1998) Global warming on Triton. Nature 393:765–767 Fairén AG (2010) A cold and wet Mars. Icarus 208:165–175 Fairén AG, Davila AF, Duport LG, Amils R, Mckay C (2009) Stability against freezing of aqueous solutions on early Mars. Nature 459:401–404

9 Interiors and Surfaces of Terrestrial Planets and Major Satellites

163

Farquhar J, Bao M, Thiemens M (2000) Atmospheric influence of Earth’s earliest sulfur cycle. Science 289:756–758 Feldman WC, Maurice S, Binder AB, Barraclough BL, Elphic RC, Lawrence DJ (1998) Fluxes of fast and epithermal neutrons from Lunar Prospector: evidence for water ice at the lunar poles. Science 281:1496–1500 Fortes AD (2000) Exobiological implications of a possible ammonia-water ocean inside Titan. Icarus 146:444–452 Fukuhara T, Futaguchi M, Hashimoto GL, Horinouchi T, Imamura T, Iwagaimi N, Kouyama T, Murakami SY, Nakamura M, Ogohara K, Sato M (2017) Large stationary gravity wave in the atmosphere of Venus. Nat Geosci 10(2):85–88 Geissler PE et al (1998) Evidence for non-synchronous rotation of Europa. Nature 391:368–370 Gough DO (1981) Solar interior structure and luminosity variations. Sol Phys 74(1): 21–34 Grotzinger JP, Gupta S, Malin MC, Rubin DM, Schieber J, Siebach K, Sumner DY, Stack KM, Vasavada AR, Arvidson RE, Calef F (2015) Deposition, exhumation, and paleoclimate of an ancient lake deposit, Gale crater, Mars. Science 350(6257):aac7575 Hamilton CW, Beggan CD, Still S, Beuthe M, Lopes RM, Williams DA, Radebaugh J, Wright W (2013) Spatial distribution of volcanoes on Io: implications for tidal heating and magma ascent. Earth Planet Sci Lett 361:272–286 Harmon JK, Perillat PJ, Slade MA (2001) High-resolution radar imaging of Mercury’s north pole. Icarus 149:1–15 Hartmann WK, Phillips RJ, Taylor GJ (eds) (1986) Origin of the moon. Lunar and Planetary Institute, Houston Hartogh P, Lis DC, Bockelée-Morvan D, de Val-Borro M, Biver N, Küppers M, Emprechtinger M, Bergin EA, Crovisier J, Rengel M, Moreno R (2011) Ocean-like water in the Jupiter-family comet 103P/Hartley 2. Nature 478(7368):218–220 Hendrix AR, Johnson RE (2008) Callisto: new insights from Galileo disk-resolved UV measurements. Astrophys J 687(1):706 Holland HD (1999) When did the Earth’s atmosphere become oxic? A reply. Geochem News 100:20–23 Hoppa GV, Tufts BR, Greenberg R, Geissler PE (1999) Formation of cycloidal features on Europa. Science 285:1899–1902 Huang J, Yang A, Zhong S (2013) Constraints of the topography, gravity and volcanism on Venusian mantle dynamics and generation of plate tectonics. Earth Planet Sci Lett 362:207– 214 Iess L, Stevenson DJ, Parisi M, Hemingway D, Jacobson RA, Lunine JI, Nimmo F, Armstrong JW, Asmar SW, Ducci M, Tortora P (2014) The gravity field and interior structure of Enceladus. Science 344(6179):78–80 Ingersoll AP, Svitek T, Murray BC (1992) Stability of polar frosts in spherical bowl-shaped craters on the Moon, Mercury, and Mars. Icarus 100:40–47 Ivanov MA, Head JW (2001) Geology of Venus: mapping of a global traverse at 30ı N latitude. J Geophys Res 106:17,515–17,566 Jakosky B (1998) The search for life on other planets. Cambridge University Press, Cambridge, UK Kargel J, Kaye JZ, Head JW, Marion GM, Sassen R, Crowley JK, Prieto Ballesteros O, Grant SA, Hogenboom DL (2000) Europa’s crust and ocean: origin, composition and prospects for life. Icarus 148:226–265 Kasting JF (1988) Runaway and moist greenhouse atmospheres and the evolution of Earth and Venus. Icarus 74:472–494 Khurana KK, Kivelson MG, Stevenson DJ, Schubert G, Russell CT, Walker RJ, Polanskey C (1998) Induced magnetic fields as evidence for subsurface oceans in Europa and Callistro. Nature 395:777–780

164

A.G. Fairén

Kivelson MG, Warnecke J, Bennett L, Joy S, Khurana KK, Linker JA, Russell CT, Walker RJ, Polanskey C (1998) Ganymede’s magnetosphere: Magnetometer overview. J Geophys Res 103:19,963 de Kleer K, Skrutskie M, Leisenring J, Davies AG, Conrad A, de Pater I, Resnick A, Bailey V, Defrère D, Hinz P, Skemer A (2017) Multi-phase volcanic resurfacing at Loki Patera on Io. Nature 545(7653):199–202 Limaye SS, Kossin JP, Rozoff C, Piccioni G, Titov DV, Markiewicz WJ (2009) Vortex circulation on Venus: dynamical similarities with terrestrial hurricanes. Geophys Res Lett 36(4):L04204 Lindkvist J, Holmström M, Khurana KK, Fatemi S, Barabash S (2015) Callisto plasma interactions: hybrid modeling including induction by a subsurface ocean. J Geophys Res Space Phys 120(6):4877–4889 Liuzzo L, Feyerabend M, Simon S, Motschmann U (2015) The impact of Callisto’s atmosphere on its plasma interaction with the Jovian magnetosphere. J Geophys Res Space Phys 120(11):9401– 9427 Lorenz RD (2002) Thermodynamics of geysers: application to Titan. Icarus 156:176–183 Lorenz RD, Shandera SE (2001) Physical properties of ammonia-rich ice: application to Titan. Geophys Res Lett 28:215–218 Marion GM (2001) Carbonate mineral solubility at low temperatures in the Na-K-Mg-Ca-H-ClSO4-OH-HCO3-CO3-CO2-H2O system. Geochim Cosmochim Acta 65:1883–1896 Marion GM (2002) A molal-based model for strong acid chemistry at low temperatures (80 m s1 ) circumpolar jet stream dominating the circulation at  D 1=16. Broadly consistent results were also found by Potter et al. (2014). Prograde flow is found on the equator at upper levels at almost all values of  < 1=2. This gradually gives way with increasing  to equatorial subrotation, which is consistent with what is observed for Earth itself (cf. Fig. 1a), except during the westerly phase of the quasi-biennial oscillation (see, e.g., Andrews et al. 1987 or Vallis 2017). The presence of prograde flow on the equator is a

15 Atmospheric Dynamics of Terrestrial Planets

301

clear indication of local super-rotation, defined as an excess of specific angular momentum (AM) compared to what the flow would have in solid body corotation with the underlying surface (Read 1986; Read and Lebonnois 2018) and represented by the dimensionless ratio, s, of that AM excess to the corotating value. This excess is typically very small for Earth, though some modeling studies have indicated the possibility of a bifurcation to an alternative circulation regime (Suarez and Duffy 1992; Saravanan 1993; Williams 2003; Kraucunas and Hartmann 2005) characterized by significant equatorial super-rotation at upper levels. Some models (Suarez and Duffy 1992; Saravanan 1993; Kraucunas and Hartmann 2005) suggest that this may be produced by non-axisymmetric heating in the tropics, although the model of Williams (2003) seems to generate super-rotation through low-latitude barotropic instability. As  is increased beyond 1, however, there is a trend toward increasing numbers of parallel, zonal jet streams across the hemisphere which alternate in sign toward the pole, somewhat similar to the pattern of jets found on the rapidly rotating gas giant planets. The jets tend to increase in strength with altitude, peaking in the lower stratosphere, and are almost certainly driven by interactions with baroclinic eddies. The corresponding temperature fields in Fig. 8 also show some interesting trends that follow a somewhat similar pattern to what is found in laboratory analogues. Thus, decreasing  below 1 initially leads to a reduction in the slope of the isotherms in the troposphere, though this increases again as  is changed from 1/2 to 1/4. The reduction in isotherm slope between  D 1=4 and 1/2 is associated with the onset of geostrophic baroclinic instability around   1=2 (Williams 1988a), roughly consistent with other studies (e.g., by Geisler et al. 1983, Del Genio and Suozzo 1987, and Navarra and Boccaletti 2002 who found an onset of baroclinic instability around  D 1=4). Further reductions in  seem then to lead to a monotonic reduction of isotherm slope until, at  D 1=16, the isotherms are almost entirely horizontal except close to the pole > 60ı . As  is increased beyond 1, however, the tropospheric isotherm slope tends to increase and develop step-like features at various latitudes that correlate with the appearance of alternating baroclinic zonal flows. This appears to be associated with the formation of parallel baroclinic zones which become separately unstable and evolve as a highly anisotropic form of geostrophic turbulence, with many similarities to the baroclinically unstable flows found in recent laboratory studies with sloping endwall boundaries (cf. Bastin and Read 1998; Wordsworth et al. 2008). Such similarities would strongly suggest a common role for strong and complex nonlinear wavezonal flow interactions, modified by the ˇ-effect, in producing the strongly zonal organization of the flow. Figure 9 presents a series of instantaneous snapshots of the zonal velocity, u, at the top of the troposphere for each value of  , and provides an impression of the typical form of the dominant eddies in each regime. This shows the principal eddies at low rotation rates  . 1=4 to manifest as an irregular meander of the main circumpolar jet stream with a zonal wavenumber m D 1  2. These meanders take a qualitatively different form around  D 1=2, becoming much more symmetrical and regular in both space and time. By  D 1, the meanders have increased in

302

P. L. Read et al.

(a)

(b)

(c)

(d)

(e)

(f )

Fig. 9 Snapshots of upper level (200 hPa) zonal wind, projected onto a spherical surface at  D 1=16 [(a)], 1/8 [(b)], 1/4 [(c)], 1/2 [(d)], 1 [(e)], and 2 [(f)], illustrating the basic organization of winds across the planet

15 Atmospheric Dynamics of Terrestrial Planets

303

zonal wavenumber to m  5  7, much as observed on Earth. A further bifurcation evidently occurs between  D 1 and 2, generating a second parallel meandering jet at a latitude around 60ı , resulting in a complex zonally banded circulation.

Dimensionless Parameters Although we can clearly order the sequence of atmospheric circulation regimes encountered in simplified model simulations according to physical parameters such as  , it is hard to make comparisons between these simulations and those, e.g., from models of planets of differing sizes, densities, or distances from their parent star. It may be more useful, therefore, to identify appropriate dimensionless parameters that play a major role in determining the form and intensity of atmospheric circulation regimes. These entail forming dimensionless combinations of physical parameters, such as those related to the size of the planet and its atmosphere, its rotation rate, and other physical attributes affecting processes such as buoyancy, radiative forcing, and friction. The planetary radius, a, is the obvious length scale to take for the horizontal scale. Also, it is conventional to take as the vertical length scale the pressure scale height, H D RT0 =g, where R is the specific gas constant and T0 a characteristic temperature. This might be justified theoretically on the grounds (Held 1978) that the scale height represents the maximum vertical scale over which baroclinic instability transports heat in a compressible atmosphere. A further aspect is that the horizontal and vertical temperature contrasts are not typically imposed directly by the applied solar heating or boundary conditions, but rather the typical scales for these quantities, h and v (where  is potential temperature), are internal parameters, determined by the heat transports within the circulation itself. While it would be desirable in principle to be able to predict these quantities from the imposed insolation and planetary parameters, this problem is still incompletely understood (e.g., see Stone 1978; Schneider 2006; Zurita-Gotor and Lindzen 2007). So for the present we take the appropriate temperature scales as given from model simulations. Since the scale for the horizontal thermal contrast is taken along a near-isobaric surface or at the ground, it can be measured either in terms of absolute or potential temperature to within a factor of order unity (given D  H ). With these modifications, we can define the following parameters, similar to those in Mitchell and Vallis (2010) and approximately equivalent to those found for laboratory systems (Geisler et al. 1983; Read et al. 2015), Rh Thermal Rossby number 2 a2 Rv Burger number Bu D 22 a2

RoT D

based on a thermal wind scale

(1) (2)

304

P. L. Read et al.

UT D

Rh : a

(3)

The thermal Rossby number is a convenient nondimensional measure of rotation, with low rotation rates corresponding to high thermal Rossby numbers. For example, in a sequence of experiments with an Earth-like model planet, Potter et al. (2014) found a strong dependence of super-rotation on RoT , with super-rotation consistently arising at thermal Rossby numbers of one and higher. The Burger number is a measure of the deformation radius, which we discuss more below. Another characteristic horizontal length scale is the so-called Rhines scale, LR , which represents the scale over which advection of relative and planetary vorticity is in balance (Rhines 1975; Vallis and Maltrud 1993). On the beta-plane, the pRhines scale is commonly expressed in terms of a wavenumber kR D 2=R  ˇ=2U , where U is a turbulent velocity scale. If we use the above thermal wind scale to give U (thus making the strong, and potentially doubtful, assumption that the mean and eddy velocities are of the same order of magnitude) and take ˇ  =a, then we obtain kR ' .2Rh =2 /1=2 . This implies a length scale LR corresponding to R =2, equivalent to LR '

 D kR



2Rh 2

1=2 (4)

which we take here to be an estimate of the Rhines length scale. This scale is often found to be a reasonable estimate of the width or spacing, Ljet , of eddy-driven jets on a spherical planet or ˇ-plane (e.g., Danilov and Gurarie 2002). The square of the ratio of the planetary radius to the Rhines scale is then given by Rh D

a2 2 a2 1 ' D Ro1 T ; 2 2 2R 2 LR h

(5)

i.e., Rh is proportional to the inverse of the thermal Rossby number. The smaller the thermal Rossby number, the more the number of zonal jets and/or baroclinic zones that might be expected, with the number of jets in each hemisphere scaling roughly as p 1=2 RoT a  Rh : D p NJ ' ' 2LR 2 2 2

(6)

The significance of RoT is further reinforced by the observation that the lateral scale, YH , predicted for the width of the Hadley circulation in the conceptual zonally symmetric model of Held and Hou (1980), is given by  YH D

5Rh 32

r

1=2 Da

5 RoT ; 3

(7)

15 Atmospheric Dynamics of Terrestrial Planets

305

which assumes an angular momentum-conserving pattern of zonal wind in geostrophic balance. For slow rotation in which RoT 1, the width of the Hadley circulation approaches the entire hemisphere of the planet itself, and one finds (Hou 1984) YH

  3 a 1 ; ! 2 8RoT

(8)

again exhibiting a significant dependence on RoT . The other major length scale that can be identified, and which is nominally independent of LR , is the Rossby deformation radius, LD . This is determined by the stratification v rather than h . Bu again measures the (squared) ratio of a representative average scale for LD (taken here for  45ı ) to the planetary radius and may be expected to provide an indication of the likely importance of baroclinic instability in the circulation (applicable when Bu is significantly less than 1). Other parameters to be considered concern, e.g., the respective roles of friction and radiative damping, for which characteristic timescales, f and R , can be identified. Appropriate dimensionless measures of these quantities, analogous to the Taylor number, Ta , in the laboratory (e.g., Read et al. 2015) and measuring the ratio of these timescales to that of planetary rotation, can be defined as Fr D 4.r /4

Radiative damping

(9)

Ff D 4.f /

Frictional damping

(10)

4

where the dominant of these parameters is likely to be the one with the lowest value. The planetary obliquity angle is another important parameter that determines the respective roles of seasonal changes and the geographical distribution of radiative forcing.

Circulation Regimes in Parameter Space The trends in circulation regime presented in section “Circulation Regimes with Varying Rotation Rate” can now also be interpreted in the context of the dimensionless parameters introduced above. From the zonal mean temperature fields in Fig. 8, we can estimate typical values for h and v , allowing a crude determination of RoT and Bu. This also enables an estimate for NJ for comparison with the zonal wind fields in Fig. 8. Finally, given an estimate of the radiative time constant in the troposphere of around 10 Earth days and a surface drag timescale of around 2.5 days, we can also provide a rough indication of the smaller of the main dissipation parameters, Ff . These estimates are listed for each of the cases shown in Figs. 8 and 9 in Table 3. The onset of baroclinic instability for  > 1=4 is consistent with an effective suppression of the instability beyond RoT  1:0 and Bu > 0:6, which is broadly

306

P. L. Read et al.

Table 3 Key dimensionless parameters for the baseline set of numerical simulations with h D 60 K, f t D 5 Earth days, and R D 25:9 Earth days, as defined by Eqs. (1), (2), (10), (9), and (6), respectively  1=16 1=8 1=4 1=2 1 2

Bu 17:1 4:28 1:18 0:29 0:066 0:017

RoT 20:5 5:14 1:28 0:32 0:080 0:020

Ff 238 3.80 103 6.09 104 9.74 105 1.56 107 2.49 108

Fr 1.71 105 2.73 106 4.67 107 6.98 108 1.12 1010 1.79 1011

NJ 0:078 0:16 0:31 0:62 1:25 2:50

consistent with what is found in laboratory experiments (cf. Hide and Mason 1975; Read et al. 2015). However, one should be wary of overinterpreting such quantitative comparisons with systems that differ significantly from a planetary atmosphere in geometrical configuration, aspect ratio, and a variety of other parameters. In laboratory systems, for example, almost pure axisymmetric flow is found for RoT > 1, whereas in atmospheric simulations (e.g., Geisler et al. 1983; Del Genio and Suozzo 1987; Williams 1988a; Kaspi and Showman 2015; Wang et al. 2018), baroclinic instability may give way to forms of barotropic instability at low rotation rates, which seem to be much less efficient at transporting heat than their baroclinic counterparts. But this does at least indicate some broad, semiquantitative parallels that can be drawn with the laboratory analogues. Such parallels extend further into the high rotation regimes, with a clear tendency for the formation of multiple zonal banded structures when LR a and Rh 1. The corresponding estimate of numbers of zonal jets, given by NJ in Table 3, also seems to provide a fair guide as to the number of eastward and westward baroclinic jets appearing in each hemisphere, especially for  & 1. For the Earth-like case  D 1, NJ  1  2, suggesting that Earth is in an intermediate regime just beyond the situation where only a single jet in each hemisphere is favored. Finally, we note that the frictional and radiative parameter estimates are all 1, though not by such a large margin at low values of  . This may mean that the low rotation regimes represented in this sequence of simulations are relatively strongly damped, which should be borne in mind when comparing model results with real planetary atmospheres.

Discussion We see from a range of model studies (Williams and Holloway 1982; Geisler et al. 1983; Del Genio and Suozzo 1987; Williams 1988a,b; Navarra and Boccaletti 2002; Mitchell and Vallis 2010; Kaspi and Showman 2015) that a parameter space can at least be constructed for hypothetical, somewhat idealized, circulations in Earthlike atmospheric systems, within which the style of circulation changes with  in ways that parallel those found in laboratory analogues quite closely. It is of interest,

15 Atmospheric Dynamics of Terrestrial Planets Table 4 Estimates of the main dimensionless parameters for Earth, Mars, Pluto, Triton, Venus, and Titan, including the dimensionless super-rotation parameter smax

307 Body Earth Mars Pluto Triton Titan Venus

Bu 0:02 0:04 163 22:4 11:8 140

RoT 0:06 0:20 4:8 1:05 18 370

NJ 1:26 0:92 0:16 0:35 0:06 0:009

smax .0:04 0.16 0.75–1.1 0.3–0.6 8.5–15 55–65

therefore, to look further into where the terrestrial planets amenable to study in the solar system might lie within the parameter space we have constructed. Table 4 presents some approximate estimates of the key dimensionless parameters discussed above for Earth, Mars, Pluto, Triton, Venus, and Titan, for comparison with the model simulations discussed in the previous section. From this, we see that these six terrestrial planetary bodies fall into three main groups. The first contains Mars and Earth, both with values of RoT < 1 and forming a group of relatively rapid rotators. Given the relatively small values of RoT , we would anticipate relatively narrow tropical Hadley circulations (Held and Hou 1980) and baroclinically active midlatitudes, much as observed. Moreover, the relative values of RoT for Earth and Mars in Table 4 would suggest (a) that Mars’s tropical Hadley p circulation would likely extend to higher latitudes than on Earth, by a factor  3, and (b) baroclinic instability on Earth is likely to be more strongly supercritical and on a smaller scale relative to the planet than on Mars. From a comparison of Figs. 1 and 2, it is clear that (a) is borne out quite accurately, with Hadley cells extending to around 30–40ı on Earth and up to 60ı on Mars. For (b) also, baroclinic instability on Mars is known to favor relatively low planetary wavenumber features (m D 1–3 typically; Read and Lewis 2004) with baroclinic storms disappearing altogether during part of the seasonal cycle in each hemisphere, whereas on Earth, wavenumbers around m D 4–8 are more typical and persist throughout the year. The second group contains Venus and Titan as extreme members, for which RoT 1, consistent with slowly rotating atmospheres in predominantly cyclostrophic or gradient wind balance that are likely to support barotropic instabilities in preference to baroclinic eddy processes. This group also seems consistently to exhibit strong values of super-rotation s, broadly consistent with the model simulations in section “Circulation Regimes with Varying Rotation Rate” though the quantitative agreement between models and observations may be less close than this comparison would suggest. The Hadley circulations in the slowly rotating model simulations in the previous section were also found to extend almost to the poles which, again, seem consistent with the Venus-like simulation illustrated in Fig. 3 and other evidence, e.g., from Venus observations and model simulations of Titan (Sánchez-Lavega et al. 2017; Lebonnois et al. 2012). The third group contains Pluto and Triton, for which RoT  1, indicative also of non-geostrophic flow and broad Hadley cells in an extended tropical zone. The available GCM simulations to date suggest relatively weak winds in both

308

P. L. Read et al.

atmospheres. This may reflect the strongly dissipative states of both atmospheres, for which molecular viscosity may play a much stronger role than on other terrestrial planets in the solar system because of the very low surface pressure, comparable to thermospheric pressures on Earth and other terrestrial planets. Another factor peculiar to both of these atmospheres is their tendency to condense and evaporate the main atmospheric constituent (N2 ) on seasonal timescales, leading to a strong circulation transferring mass from the evaporating summer polar ice cap toward the condensing cap over the winter pole associated with zonal flows through deflection of such pole-pole circulations by Coriolis forces. Such effects are likely to be fairly extreme for Pluto, whose obliquity is nearly 120ı and whose orbital period is extremely long (248 Earth years), allowing time for a large fraction of Pluto’s tenuous atmosphere to condense out during winter. The radiative damping parameter, Fr , is another factor where some of the solar system planets lie some significant distance away from the model simulations in our parameter space. Mars in particular is likely to be significantly more strongly dissipative than both the corresponding model cases ( D 1=21=4; with roughly 2–3 times the values of RoT and Bu as for Earth itself). By analogy with the laboratory systems discussed above (Hide and Mason 1975; Read et al. 2015), this would suggest the likelihood that baroclinic instabilities on Mars would be more strongly dissipative and hence less strongly supercritical and nonlinear. Such a condition is consistent with a tendency to favor smaller planetary wavenumber structures that are relatively coherent and long-lived, with comparatively simple time variations. This is consistent with a lot of evidence on Mars from both models and observations that baroclinic activity on Mars is relatively coherent and periodic (Barnes 1980, 1981; Collins et al. 1996). The location of planets in this basic parameter space thus seems to provide a number of useful insights that can account for a variety of different features found on the terrestrial planets within the solar system, including the size of the Hadley circulation and the nature of large-scale waves and instabilities that may develop within the circulation. It also indicates the validity and possible utility of the concept of dynamical similarity, widely used in fluid mechanics, which anticipates that circulating fluid systems of differing sizes, with different rotation rates, thermal properties, etc., will organize themselves in similar ways if a set of key dimensionless parameters are of similar magnitudes. In the present problem, this would seem to imply similarity provided RoT is matched between two different planetary circulations, together with some other parameters, e.g., related to the dissipation and radiative timescales. This was clearly illustrated in a recent study by Pinto and Mitchell (2014) for a slowly rotating planetary circulation, in which they compared simple model simulations obtained either for an Earth-sized planet rotating with a period of 20 Earth days or for a planet 1/20 the radius of Earth rotating with a period of 1 day. The results are illustrated in Fig. 10, in which (a) and (b) are shown for  D 1=20 and (c) and (d) are shown for a D aEarth =20, RoT D 1:3 for both cases, however, and the resulting fields of zonal mean velocity and temperature are very similar in both structure and amplitude. Some caution is necessary in making this comparison, however, since this degree of similarity

15 Atmospheric Dynamics of Terrestrial Planets

309

Fig. 10 An illustration of the potential for dynamical similarity between the atmospheric circulations of two planets with differing radii and rotation rates but with common values of RoT D 1:3 and dissipation parameters. Panels a and b are for an Earth-sized planet rotating at 1/20 of the Earth’s rotation rate, while panels c and d are for a planet 1/20th the size of Earth rotating at the same speed as Earth. Panels a and c show the zonal wind velocity (in m s1 ) and panels b and d show the corresponding zonal mean temperature field (color shaded) and meridional mass stream function (line contours). (Figures taken from Pinto and Mitchell 2014 with permission from Elsevier)

also required the models to match dissipation and radiative damping parameters, Ff and Fr (represented in Pinto and Mitchell 2014 by Ek D .4Ff /1=4 and a D .4Fr /1=4 ) as well as RoT . Similarity of obliquity would also be necessary in practice, and several other parameters, e.g., involving the thermodynamic properties of the atmosphere and its composition, though similarity of these four main parameters would be likely to capture most of the gross properties of the circulation.

Prediction From Dynamical Similarity This potential for dynamical similarity therefore suggests the possibility of predicting certain aspects of planets that are yet to be discovered, provided their position in dimensionless parameter space can be established. As an example, consider a

310

P. L. Read et al.

hypothetical “super-Earth” with a mass around 8–10 M˚ , rotating about its axis at about the same angular velocity as Earth itself and in orbit around a star within its “habitable zone.” Provided the obliquity of such a planet were not too large, we would anticipate the circulation to be largely determined by the value of RoT , which scales as .a/2 , where a would be around twice Earth’s radius. For a value of  comparable to what occurs on Earth, therefore, this would indicate a circulation pattern roughly equivalent to the  D 2 case presented in the previous section. This would suggest a more complicated circulation than on Earth with at least two distinct, parallel baroclinic storm zones at midlatitudes and a relatively narrow tropical Hadley circulation extending only to around 15ı in latitude from the equator. Westward winds would be expected to dominate near the equator and around 60ı , with eastward zonal winds elsewhere. If moisture is also taken into account in this circulation, as potentially appropriate for a “waterworld” super-Earth (Selsis et al. 2007), these models could even be used to predict climatological regions of high and low rainfall (e.g., with low rainfall at the poleward edge of the Hadley cell).

Tidally Locked Planets Although such ideas of dynamical similarity should apply generically to wide classes of non-synchronously rotating planets, many recently discovered exoplanets are located in orbits very close to their parent stars. The strong gravitational tidal interactions between star and planet would then be likely to force the rotation of the planet into a synchronous state, in which the rotation and orbital periods are in the ratio of simple integers, such as for Mercury in our solar system, for which the ratio is 3:2 (Pettengill and Dyce 1965; McGovern et al. 1965), with a relatively small obliquity. For closer-in planets, a 1:1 ratio is more likely with an obliquity close to zero, analogous to the Earth-Moon system, so that the planet permanently presents the same face to the star. Under these conditions, the atmospheric circulation would be rather different in form to the non-synchronously rotating cases discussed above (e.g., see Joshi et al. 1997; Joshi 2003), with a dominant circulation transporting air in longitude from the substellar point to the anti-stellar point, as well as from the equator to poles. Recent studies (e.g., Showman et al. 2010; Penn and Vallis 2017) have shown that the circulation under these circumstances may be dominated by a stationary pattern of dynamically coupled planetary waves (mainly equatorial Kelvin and Rossby waves, possibly resembling the Gill pattern; Gill (1980); Vallis (2017) in association with a strongly prograde equatorial zonal jet stream. However, tidal locking may be prevented by thermal tides in sufficiently massive atmospheres (Ingersoll and Dobrovolskis 1978; Leconte et al. 2015), and relative strengths of the stationary wave pattern and zonal jet also depend upon factors such as the actual rotation rate, the stratification (and hence the radius of deformation), and the overall mass of the atmosphere. As with solar system planets, idealized simulations can give insight and two examples are shown in Fig. 11. These show simulations with a primitive equation model in a tidally locked state (left panel) and a state in which the substellar

15 Atmospheric Dynamics of Terrestrial Planets

311

Fig. 11 Primitive equation simulations of tidally locked (left) and nontidally locked (right) exoplanets, with forcing via thermal relaxation to a specified field. Color shading shows the temperature at 700 hPa and white contours show the location of the thermal forcing. For the nontidally locked case, the substellar point is shown with a small white arrow denoting its direction of passage, which is to the left, here with a velocity of 25 m s1 . (Figure from Vallis et al. 2018)

point traverses the planet from right to left (right panel). In the tidally locked state, Rossby lobes can be seen to the west and poleward of the heating, as in the Gill pattern. Interestingly, the hotspot in the nontidally locked case is not synchronous with the maximum of the thermal forcing (the substellar point), and the investigation of this continues. Understanding and quantifying the balance between stationary wave patterns and zonal flows is important for interpreting observations of close-in extrasolar planets, since it affects the thermal contrast between the substellar and antistellar hemispheres which can be deduced from observed phase curves (e.g., Crossfield 2015). A number of studies have begun to explore the sensitivity of the synchronously locked circulation of close-in planetary atmospheres to various parameters (e.g., Merlis and Schneider 2010; Edson et al. 2011; Carone et al. 2015, 2016; Showman et al. 2015; Penn and Vallis 2017; Noda et al. 2017), although a complete picture of the relevant parameter space is yet to emerge. But progress so far indicates that such an approach using idealized numerical simulations, combining basic dynamical theory with the identification of the key dimensionless parameters, is well worthwhile. Acknowledgements Thanks are due to the many colleagues and students who have contributed to our understanding of this subject and carried out some of the research described herein, particularly Raymond Hide, Sebastien Lebonnois, James Penn, Fachreddin Tabataba-Vakili, Yixiong Wang, and Gareth Williams. GKV also acknowledges support from the Leverhulme Trust.

References Andrews DG, Holton JR, Leovy CB (1987) Middle atmosphere dynamics. Academic, New York Barnes JR (1980) Time spectral analysis of mid-latitude disturbances in the Martian atmosphere. J Atmos Sci 37:2002–2015

312

P. L. Read et al.

Barnes JR (1981) Midlatitude disturbances in the Martian atmosphere: a second Mars year. J Atmos Sci 38:225–234 Bastin M, Read PL (1998) Experiments on the structure of baroclinic waves and zonal jets in an internally heated rotating cylinder of fluid. Phys Fluids 10:374–389 Beaulieu JP et al (2006) Discovery of a cool planet of 5.5 Earth masses through gravitational microlensing. Nature 439:437–440 Bourke W (1974) A multilevel spectral model. I. Formulation and hemispheric integrations. Mon Weather Rev 102:687–701 Carone L, Keppens R, Decin L (2015) Connecting the dots – II: phase changes in the climate dynamics of tidally locked exoplanets. Mon Not R Astr Soc 453:2412–2437 Carone L, Keppens R, Decin L (2016) Connecting the dots – III: nightside cooling and surface friction affect climates of tidally-locked terrestrial planets. Mon Not R astr Soc 461:1981–2002 Collins M, Lewis SR, Read PL, Hourdin F (1996) Baroclinic wave transitions in the martian atmosphere. Icarus 120:344–357 Crossfield IJM (2015) Observations of exoplanet atmospheres. Pub Astron Soc Pac 127:941–960 Danilov S, Gurarie D (2002) Rhines scale and the spectra of ˇ-plane turbulence with bottom drag. Phys Rev E 65:067301 Dee DP, Uppala SM, Simmons AJ et al (2011) The ERA-interim reanalysis: configuration and performance of the data assimilation system. Quarterly Q J R Meteorol Soc 137(656):553–597 Del Genio A, Suozzo RJ (1987) A comparative study of rapidly and slowly rotating circulation regimes in a terrestrial general circulation model. J Atmos Sci 44:973–986 Edson A, Lee S, Bannon P, Kasting JF, Pollard D (2011) Atmospheric circulation of terrestrial planets orbiting low mass stars. Icarus 212:1–13 Forget F, Bertrand T, Vangvichith M, Leconte J, EM Lellouch E (2017) A post-New Horizons global climate model of Pluto including the N2 , CH4 and CO cycles. Icarus 287:54–71 Fraedrich K, Kirk E, Lunkeit F (1998) Puma: Portable University Model of the Atmosphere. Deutsches Klimarechenzentrum Technical Report 16 Frisius T, Lunkeit F, Fraedrich K, James IN (1998) Storm-track organization and variability in a simplified global circulation model. Quart J R Meteor Soc 124:1019–1043 Fultz D, Long RR, Owens GV et al. (1959) Studies of thermal convection in a rotating cylinder with some implications for large-scale atmospheric motions. Meteorological monographs, vol 4. American Meteorological Society, Boston Geisler JE, Pitcher EJ, Malone RC (1983) Rotating-fluid experiments with an atmospheric general circulation model. J Geophys Res 88:9706–9716 Ghil M, Childress S (1987) Topics in geophysical fluid dynamics: atmospheric dynamics, dynamo theory and climate dynamics. Springer, New York Gill AE (1980) Some simple solutions for heat induced tropical circulation. Quart J R Meteorol Soc 106:447–462 Held IM (1978) The vertical scale of an unstable baroclinic wave and its importance for eddy heat flux parameterizations. J Atmos Sci 35:572–576 Held IM, Hou AY (1980) Nonlinear axially symmetric circulations in a nearly inviscid atmosphere. J Atmos Sci 37:515–533 Hide R (1953) Some experiments on thermal convection in a rotating liquid. Quart J R Meteorol Soc 79:161 Hide R, Mason PJ (1975) Sloping convection in a rotating fluid. Adv Phys 24:47–99 Hoskins BJ, Simmons AJ (1975) A multi-layer spectral model and the semi-implicit method. Quart J R Meteor Soc 101:637–655 Hou AY (1984) Axisymmetric circulations forced by heat and momentum sources: a simple model applicable to the Venus atmosphere. J Atmos Sci 41:3437–3455 Hubbard WB, Yelle RV, Lunine JL (1990) Nonisothermal Pluto atmosphere models. Icarus 84:1–11 Ingersoll AP, Dobrovolskis AR (1978) Venus’ rotation and atmospheric tides. Nature Jenkins GS (1996) A sensitivity study of changes in Earth’s rotation rate with an atmospheric general circulation model. Glob Plan Change 11:141–154

15 Atmospheric Dynamics of Terrestrial Planets

313

Joshi MM (2003) Climate model studies of synchronously rotating planets. Astrobiology 3:415–427 Joshi MM, Haberle RM, Reynolds RT (1997) Simulations of the atmospheres of synchronously rotating terrestrial planets orbiting M Dwarfs: conditions for atmospheric collapse and the implications for habitability. Icarus 129:450–465 Kållberg P, Berrisford P, Hoskins B et al. (2004) ERA-40 atlas. ERA40 Project Reports Series 19, ECMWF Kaspi Y, Showman AP (2015) Atmospheric dynamics of terrestrial exoplanets over a wide range of orbital and atmospheric parameters. Astrophys J 804:60 Kraucunas I, Hartmann DL (2005) Equatorial superrotation and the factors controlling the zonalmean zonal winds in the tropical upper troposphere. J Atmos Sci 62:371–389 Lebonnois S, Hourdin F, Eymet V et al (2010) Superrotation of Venus’ atmosphere analyzed with a full general circulation model. J Geophys Res 115:E06006 Lebonnois S, Burgalat J, Rannou P, Charnay B (2012) Titan global climate model: a new 3-dimensional version of the IPSL Titan GCM. Icarus 218(1):707–722 Lebonnois S, Sugimoto N, Gilli G (2016) Wave analysis in the Venus atmosphere below 100-km, simulated by the LMD Venus GCM. Icarus 278:38–51 Leconte J, Wu H, Menou K, Murray N (2015) Asynchronous rotation of Earth-mass planets in the habitable zone of lower-mass stars. Science 347:632–635 Leovy CB (1973) Rotation of the upper atmosphere of Venus. J Atmos Sci 30:1218–1220 Lewis S, Read P (2003) Equatorial jets in the dusty Martian atmosphere. J Geophys Res E: Planets 108(4):15 Lewis SR, Collins M, Read PL et al (1999) A climate database for Mars. J Geophys Res 104:24177–24194 Lorenz EN (1969) The nature and theory of the general circulation of the atmosphere. T. P. 115, No. 218, W. M. O., Geneva Lorenz EN (1993) The essence of chaos. UCL Press, London Mayor M, Udry S, Lovis C et al (2009) The HARPS search for southern extra-solar planets XIII. A planetary system with 3 super-Earths (4.2, 6.9, and 9.2 M˚ ). Astron Astrophys 493: 639–644 McGovern WE, Gross SH, Rasool SI (1965) Rotation period of the planet Mercury. Nature 208:375 Merlis TM, Schneider T (2010) Atmospheric dynamics of Earth-like tidally locked aquaplanets. J Adv Model Earth Syst 2:13 Mitchell J, Vallis GK (2010) The transition to superrotation in terrestrial atmospheres. J Geophys Res (Planets) 115:E12,008, 17 Navarra A, Boccaletti C (2002) Numerical general circulation experiments of sensitivity to Earth rotation rate. Climate Dyn 19:467–483 Noda S, Ishiwatari M, Nakajima K et al (2017) The circulation and day-night heat transport in the atmosphere of a synchronously rotating aquaplanet: dependence on planetary rotation rate. Icarus 282:1–18 Penn J, Vallis GK (2017) The thermal phase curve offset on tidally and nontidally locked exoplanets: A shallow water model. Astrophys J 842(2):101. http://stacks.iop.org/0004-637X/ 842/i=2/a=101 Pettengill GH, Dyce RB (1965) A radar determination of the rotation of the planet Mercury. Nature 206:1240 Pinto JRD, Mitchell JL (2014) Atmospheric superrotation in an idealized GCM: parameter dependence of the eddy response. Icarus 238:93–109 Potter SF, Vallis GK, Mitchell JL (2014) Spontaneous superrotation and the role of Kelvin waves in an idealized dry GCM. J Atmos Sci 71(2):596–614 Read PL (1986) Super-rotation and diffusion of axial angular momentum: II. A review of quasi-axisymmetric models of planetary atmospheres. Quart J R Meteorol Soc 112: 253–272 Read PL (2001) Transition to geostrophic turbulence in the laboratory, and as a paradigm in atmospheres and oceans. Surv Geophys 22:265–317

314

P. L. Read et al.

Read PL, Lebonnois S (2018, in press) Super-rotation on Venus, Titan and elsewhere. Ann Rev Earth Plan Sci 46 Read PL, Lewis SR (2004) The Martian climate revisited: atmosphere and environment of a desert planet. Springer-Praxis, Berlin/Heidelberg/New York Read PL, Pérez EP, Moroz IM, Young RMB (2015) General circulation of planetary atmospheres: insights from rotating annulus and related experiments. In: von Larcher T, Williams PD (eds) Modelling atmospheric and oceanic flows: insights from laboratory experiments and numerical simulations. American Geophysical Union/Wiley, Hoboken, pp 9–44 Rhines PB (1975) Waves and turbulence on a ˇ-plane. J Fluid Mech 69:417–443 Sánchez-Lavega A, Lebonnois S, Imamura T, Read P, Luz D (2017) The atmospheric dynamics of Venus. Space Sci Rev 212:1541–1616 Saravanan R (1993) Equatorial superrotation and maintenance of the general circulation in twolevel models. J Atmos Sci 50:1221–1227 Schneider T (2006) The general circulation of the atmosphere. Ann Rev Earth Plan Sci 34:655–688 Selsis F, Kasting JF, Levrard B et al (2007) Habitable planets around the star Gliese 581? Astron Astrophys 476:1373–1387 Showman AP, Cho J, Menou K (2010) Atmospheric circulation of extrasolar planets. In: Seager S (ed) Exoplanets. University of Arizona Press, Arizona, pp 471–516 Showman AP, Lewis NK, Fortney JJ (2015) Three-dimensional atmospheric circulation of warm and hot Jupiters: effects of orbital distance, rotation period and nonsynchronous rotation. Astrophys J 801:95 Stone P (1978) Baroclinic adjustment. J Atmos Sci 35:561–571 Suarez MJ, Duffy DG (1992) Terrestrial superrotation: a bifurcation of the general circulation. J Atmos Sci 49:1541–1554 Udry S, Bonfils X, Delfosse X et al (2007) The HARPS search for southern extra-solar planets XI. Super-Earths (5 and 8 M˚ ) in a 3-planet system. Astron Astrophys 469:L43–L47 Vallis GK (2016) Geophysical fluid dynamics: whence, whither and why? Proc Roy Soc A 472: 1–23 Vallis GK (2017) Atmospheric and oceanic fluid dynamics: fundamentals and large-scale circulation, 2nd edn. Cambridge University Press, Cambridge Vallis GK, Maltrud ME (1993) Generation of mean flows and jets on a beta plane and over topography. J Phys Oceanog 23:1346–1362 Vallis GK, Colyer G, Geen R et al (2018) Isca, v1.0: a framework for the global modeling of the atmospheres of Earth and other planets at varying levels of complexity. Geosci Model Dev 11:843–859 https://doi.org/10.5194/gmd-11-843-2018 von Hardenberg J, Fraedrich K, Lunkeit F, Provenzale A (2000) Transient chaotic mixing during a baroclinic life cycle. Chaos 10:122–134 Wang Y, Read PL, Tabataba-Vakili F, Young RMB (2018) Comparative terrestrial atmospheric circulation regimes in simplified global circulation models: I. From cyclostrophic super-rotation to geostrophic turbulence. Quart J R Meteorol Soc submitted Williams GP (1988a) The dynamical range of global circulations – I. Clim Dyn 2:205–260 Williams GP (1988b) The dynamical range of global circulations – II. Clim Dyn 3:45–84 Williams GP (2003) Barotropic instability and equatorial superrotation. J Atmos Sci 62:2136–2152 Williams GP, Holloway JL (1982) The range and unity of planetary circulations. Nature 297: 295–299 Wordsworth RD, Read PL, Yamazaki YH (2008) Turbulence, waves and jets in a differentially heated rotating annulus experiment. Phys Fluids 20:126602. https://doi.org/10.1063/1.2990,042 Yelle RV, Lunine JL, Pollack JB, Brown RH (1995) Lower atmospheric structure and surfaceatmosphere interactions on Triton. In: Cruikshank DP, Matthews MS Schumann AM (eds) Neptune and triton. University of Arizona Press, Tucson, pp 1031–1105 Zalucha AM (2016) An atmospheric general circulation model for Pluto with predictions for New Horizons temperature profiles. Mon Not R astr Soc 459:902–923

15 Atmospheric Dynamics of Terrestrial Planets

315

Zalucha AM, Michaels TI (2013) A 3D general circulation model for Pluto and Triton with fixed volatile abundance and simplified surface forcing. Icarus 223:819–831 Zurita-Gotor P, Lindzen R (2007) Theories of baroclinic adjustment and eddy equilibration. In: Schneider T, Sobel A (eds) The global circulation of the atmosphere: phenomena, theory, challenges. Princeton University Press, Princeton

Atmospheric Dynamics of Giants and Icy Planets

16

A. Sánchez-Lavega and M. Heimpel

Contents Bulk Properties of the Fluid Planets in the Solar System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Wind System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Meteorology of the “Weather” Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cloud Bands: Belts and Zones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Anticyclones and Cyclones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Convective Storms and Planetary-Scale Disturbances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Polar Vortices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Theories and Models for the Zonal Jets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Magnetic Fields and Dynamo Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Deep Convection Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Shallow Forcing and Hybrid Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Laboratory Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions: The Perspective for Fluid Exoplanets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross-References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

318 321 325 327 328 329 332 334 337 337 338 342 343 344 344 345

Abstract

We present the state of knowledge of the dynamics of the atmospheres of the gas giants (Jupiter and Saturn) and ice giants (Uranus and Neptune), describing their general circulation, the most relevant atmospheric phenomena, and the models

A. Sánchez-Lavega () Departamento Física Aplicada I, Escuela de Ingeniería de Bilbao, Universidad del País Vasco UPV/EHU, Bilbao, Spain e-mail: [email protected] M. Heimpel Department of Physics, University of Alberta, Edmonton, AB, Canada e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_51

317

318

A. Sánchez-Lavega and M. Heimpel

developed so far to explain their atmospheric dynamics. Observations show that these two types of fluid and cold planets differ in their general circulation at cloud level. Jupiter and Saturn are dominated by a jet system that alternates in their direction with latitude, and both possess an intense eastward equatorial jet. On the other hand, Uranus and Neptune show a dominating intense and wide in latitude westward jet symmetric with respect to the equator. In spite of this difference, the four planets present similar atmospheric dynamical phenomena (large-scale vortices, storms, and long waves, among others). Deep convection models have shown that turbulent convection resulting in angular momentum mixing may explain the westward (retrograde) equatorial flow on the ice giants. The jet systems of Jupiter and Saturn have been successfully reproduced using deep convection and shallow forcing models. However, the prograde equatorial flow of the gas giants is more naturally reproduced with deep models or hybrid shallow models incorporating aspects of deeper forcing.

Bulk Properties of the Fluid Planets in the Solar System At a mean distance of 778 to 4496 million of km from the Sun, we find the giant (Jupiter and Saturn) and icy planets (Uranus and Neptune) (Fig. 1). These bodies are totally different in size, mass, rotation, and composition to the inner telluric of terrestrial planets, having low mean density and being structurally mainly fluid planets. Their atmospheres are essentially made of hydrogen and helium with traces of other gasses. Telescopic and spacecraft images show the clouds and hazes that populate their upper atmospheres and represent the only part accessible to remote sensing, and these occupy only a small fraction of their radius. Table 1 gives the basic orbital data. All the data in this section are taken from Sánchez-Lavega (2011) and Sánchez-Lavega et al. (2017). Their orbits are nearly circular, but the tilt of the rotation axis of these planets is large, except for Jupiter, and therefore the solar heating rates vary seasonally along the large year in these planets. An extreme case is Uranus whose axis is tilted close to 98ı implying extreme insolation variation between the pole and equator. Table 2 summarizes the basic physical properties of these planets. Because of their size, we clearly see that we can group them in two, the giants (Jupiter and Saturn) and the icy, (Uranus and Neptune) about 10 and 4 times the radius of the Earth. Their masses are large compared to that of the terrestrial planets, but their mean density is low, close to that of water. The acceleration of gravity for Saturn, Uranus, and Neptune is similar to that of the Earth but that of Jupiter is about 2.2 times greater. These planets are fast rotators with short rotation periods and essentially fluid, and therefore they are oblate bodies (spheroids) with significant differences between the equatorial and polar radius, particularly in the case of Saturn. In Table 3 we give the energetic properties for these four fluid planets. At their distances from the Sun, the equilibrium temperature of these bodies ranges from 113 K for Jupiter to 48 K for Neptune. Their geometric albedo is very similar in

16 Atmospheric Dynamics of Giants and Icy Planets

319

Fig. 1 The giant and icy planets in the visible. (a) Jupiter on April 21, 2014 (Hubble Space Telescope, HST); (b) Saturn on June 30, 2015 (Hubble Space Telescope, HST); (c) Uranus on July 25, 2012 (W. M. Keck Observatory): (d) Neptune on August 20, 1989 (Voyager 2). Credits: (a) NASA/ESA/A. Simon; (b) NASA/ESA/Grupo Ciencias Planetarias UPV/EHU; (c) L. Sromovsky, University of Wisconsin–Madison/W. M. Keck Observatory; (d) NASA/JPL. Not to scale Table 1 Orbital data for giant and icy planets Planet Jupiter Saturn Uranus Neptune

Mean distance to the Sun (x108 km) 7.78 14.27 28.69 44.96

Orbital eccentricity 0.0483 0.0560 0.0461 0.0097

Orbital tilt (degrees) 3.08 26.7 97.9 28.8

Orbital period (years) 11.86 29.5 84.01 164.79

all cases (about 0.5) telling us that the scattering and absorption properties of their clouds and gasses are very similar in all of them. However, their infrared spectrum shows an excess of radiation emission relative to what it is received in all cases

320

A. Sánchez-Lavega and M. Heimpel

Table 2 Physical properties for giant and icy planets Planet Jupiter Saturn Uranus Neptune

RP (km) 71,300 60,100 25,500 24,800

Mp (kg) 1.90  1027 5.68  1026 8.68  1025 1.02  1026

(g cm3 ) 1.33 0.69 1.32 1.64

g (cm s2 ) 2288 950 869 1100

Rotation period 9. 84 h 10.6 h 17.9 h 19.2 h

Oblateness 0.061 0.09 0.03 0.03

Table 3 Energy properties of the giant and icy planets Object Jupiter Saturn Uranus Neptune

Geometric albedo 0.52 0.47 0.51 0.41

Insolation (W/m2 ) 50.6 15.1 3.72 1.52

Internal energy (Wm2 ) 5.44 2.01 0.042 0.433

Teff (K) 124.4 95 59.1 59.3

Teq (K) 113 83 60 48

except for Uranus (Fig. 2). This implies that an internal energy source exists on those planets, and their effective temperature is above that expected from equilibrium with solar radiation (Table 3). The gas giants’ bulk composition is mostly hydrogen according to their mass and size, with their atmospheres dominated by H2 and He (Table 4). Current models suggest that the neutral deep atmosphere of Jupiter extends down to about 15,000 km and in Saturn down to 30,000 km. The hydrogen is mostly in molecular form (H2 ) up to pressures of 1 Mbar. Uranus and Neptune models predict that below the molecular atmosphere of hydrogen and helium, at pressures around 0.1 Mbar (T 2000 K), a transition occurs where a mixture of ices (water, ammonia, and methane) in ionic state is predicted to exist (they dominate their bulk composition), perhaps mixed with some heavier material. Accordingly, the neutral atmosphere of Uranus extends down to about 5500 km and in Neptune down to 3500 km. Interior models are described in full detail in  Chap. 10, “Internal Structure of Giant and Icy Planets: Importance of Heavy Elements and Mixing”. Thus we see that the gas and icy giants have deep atmospheres when compared to the terrestrial ones, i.e., their atmospheric thickness is a significant fraction of the planetary radius. In the upper part of these deep atmospheres at a pressure of 1 bar where the temperatures are 75 K for Uranus and Neptune, 135 K for Saturn, and 165 K for Jupiter, clouds and hazes form according to thermochemical models (Fig. 3) (see  Chap. 14, “Temperature, Clouds, and Aerosols in Giant and Icy Planets”). Clouds made of NH3 , NH4 SH, and H2 O ices (in Jupiter and Saturn) with CH4 ices as the top cloud in Uranus and Neptune extend vertically across a layer with a thickness of 200–500 km, commonly known as the “weather layer” (Atreya and Wong 2005). These clouds are mixed with different chromophore agents that are in part responsible of the colors we see in the images of these planets (Fig. 1). The dynamical phenomena we describe in next sections occur in the weather layer

16 Atmospheric Dynamics of Giants and Icy Planets

321

Fig. 2 Jupiter at different wavelengths and spectra. Top row, Jupiter at different wavelengths. From left to right, ultraviolet showing the aurora emission (HST with WFC2), visible (HST with WFC3), infrared at 5 m (VLT with VISIR at ESO), infrared at 8.7 m (NASA-IRTF), and radio (VLA). Bottom, Jupiter spectrum adapted from Hanel et al. (1981). Credits, J. T. Clarke and G. E. Ballester (University of Michigan), J. Trauger and R. Evans (JPL/NASA); NASA/ESA/M. H. Wong (University of California, Berkeley), H. B. Hammel (Space Science Institute, Boulder, Colo.), I. de Pater (University of California, Berkeley), and the Jupiter Impact Team; ESO/L.N. Fletcher; NASA/JPL/IRTF; I. de Pater, M. H. Wong (UC Berkeley), R. J. Sault (Univ. Melbourne) Table 4 Main atmospheric composition of Jupiter, Saturn, Uranus, and Neptune

Species H2 He

Jupiter 0.864 0.136

Saturn 0.881 0.11–0.16

Uranus 0.85 0.18

Neptune 0.85 0.18

and manifest at optical wavelengths with different patterns that draw the cloud morphologies we see in these planets. At some wavelengths in the thermal part of the spectrum, the clouds and the absorbing gasses act as opacity sources providing the radiance contrasts we see.

The Wind System Atmospheric motions are measured using discrete clouds as tracers of the flow (Sánchez-Lavega et al. 2017 and references therein). Most cloud features are passive

322

A. Sánchez-Lavega and M. Heimpel

0.1

JUPITER

SATURN

P (bar)

P (bar)

NH3

1 NH4SH

NH3

1

NH4SH H2O

H2O

10 10

10-5 10-4 10-3 10-2 10-1

10-5

Particle density (g/l)

URANUS

10-1

NEPTUNE

CH4

1 P (bar)

P (bar)

10-2

0.1

SH2

10

10-3

Particle density (g/l)

0.1

1

10-4

NH3

SH2

10

NH4SH

10-5 10-4 10-3 10-2 10-1 1 Particle density (g/l)

NH3 NH4SH

H2O

100

CH4

10

H2O

100 10-5 10-4 10-3 10-2 10-1 1 Particle density (g/l)

10

Fig. 3 Vertical cloud structure. The four plots show the vertical density particle distribution of clouds predicted by thermochemical models. See  Chap. 14, “Temperature, Clouds, and Aerosols in Giant and Icy Planets” (Courtesy of R. Hueso (Grupo Ciencias Planetarias UPV/EHU))

tracers, so the phase speed of the waves and active clouds suffering rapid changes are not considered as to retrieve the mean flow. The global wind velocity has been determined in these planets using images taken with ground-based telescopes, the Hubble Space Telescope, and from a spacecraft in a flyby trajectory or in orbit (case of Jupiter and Saturn). In flyby mode, winds were measured by Voyager 1 and 2 in Jupiter (1979) and Saturn (1980 and 1981), Cassini in Jupiter (2000), and Voyager 2 in Uranus (1986) and Neptune (1989). In orbit only the Cassini mission has so far obtained global wind profiles for Saturn (2005–2016). The Galileo orbiter obtained high-resolution wind measurements in small areas of Jupiter. Combining all these data, the picture that emerges is that at upper cloud level (for all the cases at pressure

16 Atmospheric Dynamics of Giants and Icy Planets

323

level of P 1 bar), the wind system has similar patterns in Jupiter and Saturn on one hand (multiple east–west jets with a strong eastward equatorial jet) and in Uranus and Neptune in the other hand (a broad westward equatorial jet and two eastward mid-latitude or latitude jets) (Figs. 4 and 5). Above the clouds, the winds are determined using temperature measurements and the use of thermal wind equation under geostrophic balance in the momentum equation (Sánchez-Lavega 2011). This method is valid for all latitudes except for a narrow band in the equatorial region ( from 5ı N to 5ı S in latitude). The vertical wind structure above clouds has been mapped with good resolution using for Jupiter and Saturn the Voyager and Cassini temperature retrievals (Fig. 6). The data extend from the cloud level (where the winds determined from cloud tracking are used as a reference) to the upper stratosphere at about P 1 mbar. The data show that globally the winds decrease with altitude although more complex is the wind behavior in the equatorial regions of Jupiter and Saturn. The equatorial stratospheres of both planets exhibit vertical temperature and wind oscillations. In Jupiter the phenomenon is known as the quasi-quadrennial oscillation (QQO) due to its 4-year period and occurs at pressure altitudes 10–20 mbar with a temperature amplitude of 2 K (Orton et al. 1991). In Saturn, the phenomenon is called semiannual oscillation (SAO) and has a period close to 15 years and amplitude oscillation in the range 26 K in the altitude range from 0.01 to 20 mbar (Fig. 6). The corresponding oscillations of the wind speed are in the range  ˙ 100 ms1 . A narrow equatorial jet ( ˙ 2ı in latitude) has been reported in the tropopause (60 mbar altitude level) with speeds 100 ms1 above the broad equatorial jet (Fig. 4). Beneath the upper clouds, we have few inferences of the wind structure. In Jupiter, the Galileo probe penetrated in July 1995 at latitude 7.5ı N in a region known as a “hot spot” because of the low cloud opacity and escaping radiation (and higher temperatures) from deeper levels. The wind measurements indicate that the velocity increased rapidly with depth from 100 ms1 to 180 ms1 down to 5 bar and then was nearly constant downward up to 22 bar (about 400 km in altitude) (Fig. 7). In Saturn, wind profile measurements using cloud tracking at a wavelength of 5 m, sensitive to altitude levels 2–4 bar, indicate the persistence of the wind profile structure but with an apparent increase in speed in the equatorial area. There is indirect evidence from models of planetary-scale disturbances in both planets that in the weather layer (1–5 bar in Jupiter and 1–10 bar in Saturn) the winds remain steady or increase slightly with depth (Sánchez-Lavega et al. 2008, 2011). The temperature maps at cloud level do not show large anomalies related to the daily or seasonal insolation cycles. The rotation axis tilt relative to the orbital plane varies from 3ı to 97ı in these planets (Table 1). Since the orbits are nearly circular, tilt is the main parameter influencing the insolation at the top of the atmosphere and its seasonal cyclic change. However, the motions are steady in the zonal direction in the four planets without a measurable component in the direction of the heating– cooling areas. This is due in part to the large value of the radiative time constant at cloud level, above the day or year duration (Sánchez-Lavega 2011), and to the presence of the internal heat source in Jupiter, Saturn, and Neptune.

324

A. Sánchez-Lavega and M. Heimpel

Fig. 4 The wind profiles at upper cloud level in Jupiter and Saturn. (a) Jupiter, red curve is from Voyager 1 and 2 images in 1979 (Limaye 1986), black (from E. García-Melendo) and blue curves (Porco et al. 2003) are from Cassini images in 2000, and green is from HST images in 1995–2000 (García-Melendo and Sánchez-Lavega 2001); (b) Saturn, green curve is from Voyager data (Sánchez-Lavega et al. 2000), and black and red are from Cassini images sensing the main cloud (altitude level 350–500 mbar) and the upper haze layer (altitude level 60–250 mbar) (GarcíaMelendo et al. 2011)

16 Atmospheric Dynamics of Giants and Icy Planets

325

Fig. 5 The wind profiles at upper cloud level in Uranus and Neptune. The solid curves are fits to the available wind velocity data. (a) Uranus from Sromovsky et al. (2015); (b) Neptune from L. A. Sromovsky compilation of wind measurements as published in Sánchez-Lavega et al. (2017) where details on the profiles can be found

Meteorology of the “Weather” Layer The “weather layer” is defined in these planets by the thickness of the region where the upper clouds and hazes form. Thermochemical models predict this layer to extend from 0.1 to 5 bar (90 km) in Jupiter, 0.1 to 12 bar (250 km) in Saturn, and 0.1 to 100 bar (360 km) in Uranus and Neptune (Fig. 3). In the upper clouds made of ammonia ice in Jupiter and Saturn and made of methane ice in Uranus and Neptune, a rich meteorology is observed, dominated by a marked banded aspect in Jupiter and Saturn that is less contrasted and conspicuous in Uranus and Neptune. Imbedded in the banding and jet system, these planets exhibit a variety of meteorological phenomena in an ample range of spatial and temporal scales: vortices, waves, convective storms, planetary-scale disturbances, etc. In what follows we present the basic data about them (see reviews by Allison et al. 1991; Ingersoll et al. 1995, 2004; del Genio et al. 2009).

Fig. 6 Thermal windsin Jupiter and Saturn. Jupiter thermal winds: (a) from Cassini CIRS retrievals in 2000 and (b) from temperature data with TEXES– IRTF in 2013–2016 (Fletcher et al. 2016a); (c) Saturn thermal winds in 2009 (Fletcher et al. 2016b); (d) Saturn semiannual oscillation (SAO) at equatorial latitudes (Fouchet et al. 2008)

326 A. Sánchez-Lavega and M. Heimpel

16 Atmospheric Dynamics of Giants and Icy Planets

a

327

200

160 Wind speed (m s )

140

-1

Wind speed (m s-1)

180

120 100 80

210 200 190 180 170 160 150 0

5

10 15 Pressure (bar)

20

60 0

Latitude [planetographic]

c

5

10 15 Pressure (bar)

20

80 Voyager Cassini (ISS) Cassini (VIMS)

60 40 20 0 -20 -40 -60 -200

0

200 Velocity (m/s)

400

600

Fig. 7 Jupiter and Saturn deeper winds profiles. (a) Jupiter’s vertical wind profile measured from Doppler tracking of the Galileo probe in 1995 (Atkinson et al. 1998) during its descent in a “hot spot” area (marked by a circle) shown in the 5 m image of Jupiter in (b) (G. Orton, NASAIRTF); (c) Saturn wind profile at 2–4 bar altitude level from cloud tracking using Cassini VIMS images at 5 m (Choi et al. 2009) as shown on (d) November 1, 2008, and (e) February 24, 2007). Image credits: (d) NASA/JPL/U. Arizona/U. Leicester and (e) NASA/JPL/U. Arizona

Cloud Bands: Belts and Zones The aspect of Jupiter when observed with a small telescope is that of a banded planet, the bands traditionally known as “belts” for those with low albedo (dark bands) and as “zones” for the high albedo (bright bands) (Fig. 1). Chromophore agents mixed with the white ices forming the clouds provide the colors and reflectivity contrast. The edges of the belts and zones are in general aligned with the latitude circles where the zonal wind has velocity extremes. According to the ambient meridional wind shear, the vorticity is cyclonic in the belts and anticyclonic in the zones. The belts have thinner and lower clouds than the zones and represent regions where the global motions are dominated by subsidence and descent (the

328

A. Sánchez-Lavega and M. Heimpel

belts) and ascent and divergence (the zones) (Ingersoll et al. 2004). The belts and zones in Jupiter suffer cyclic albedo and morphology changes, the major ones related to planetary-scale disturbances as described below. In Saturn, the banded aspect is less contrasted because of the presence of a thick haze above the upper cloud that blurs at short wavelengths the pattern contrast (West et al. 2009). In Uranus and Neptune, the bands show low contrast in the visible due to their intrinsic low reflectivity and presence of the hazes above clouds (Allison et al. 1991; Baines et al. 1995). Only at the wavelengths of the methane absorption the banding aspect shows a large contrast mainly due to differences in altitude within the upper cloud.

Anticyclones and Cyclones The four planets show discrete oval spots visible due to their different contrast and color with background clouds. Dynamically they represent vorticity areas where the atmospheric parcels rotate around their centers following approximately concentric elliptical paths. Jupiter has the largest number and longevity record for some of these ovals. Saturn has fewer in number but very similar to those in Jupiter. Neptune has few of these spots, and in Uranus just one has been observed, but in none of these cases in the icy giants has it been possible to measure the rotation speed. Table 5 summarizes the main properties of selected representative vortices. Jupiter has the best known and best studied vortex in the solar system: the Great Red Spot (GRS) (Rogers 1995; Ingersoll et al. 2004). The GRS has been observed for more than 300 years. It sits at latitude 22.5ı S where the wind profile is close to zero velocity, but its zonal velocity has varied along its history from 2 to C1 ms1 . Its size has shrunk from 40,000 km (east–west) times 13,000 km (north– south) during the last century to its current size of 16,100 km times 11,200 km, respectively. The GRS is an anticyclone rotating with maximum tangential velocity of 100 ms1 at its periphery and has its cloud tops above the level of the

Table 5 Properties of representative vortices in the giant planets Planet Jupiter

Saturn Uranus Neptune

Vortex name GRS BA Barge NPS BS UDS GDS DS2

Latitude 22.5ı S 33ı S 15ı N 75ı N 42ı N 28ı N 20ı S 52.5ı S

Lx –Ly (km) 24,000–11,000 10,000–6000 12,000–3000 10,000–5000 6000–4000 2700–1300 15,000–6000 5200–2600

VT (ms1 ) 100–160 100–120 55 – 55 – – –

U (ms1 ) 2 /C1 C2 2.5 0.2 C5 43 350 0

Vorticity (s1 ) 6.1  105 A 6.4  105 A 4  105 C A C4  105 A A A A

Notes: (1) Lx –Ly are the zonal and meridional oval axes; (2) VT is the tangential (maximum velocity at periphery); (3) U is the zonal velocity of the vortex; (4) vorticity is the circulation divided by the oval area with A for anticyclone and C for cyclone Adapted from Sánchez-Lavega (2011)

16 Atmospheric Dynamics of Giants and Icy Planets

329

surrounding clouds. In the GRS interior, the clouds are distributed in filaments with turbulent patterns at the center. The agents producing the red color of this spot are still a mystery. The second most enduring anticyclone in Jupiter is called BA, located at 33ı S; it resulted from the successive merger in 1998 and 2000 of three ovals that formed around 1940. This oval is currently about half the size of the GRS and is very similar to it in all its dynamical properties, except in color. Typically it is white but a reddish ring in the middle of the oval. On the other hand, cyclones are more elongated ovals than anticyclones and show chaotic and turbulent interiors, and their lifetime is shorter (typically 1–2 years at maximum). Classic representatives of this family of features are the “barges” located at 16ı N. Saturn also has anticyclones and cyclones distributed at most latitudes, but they are smaller in size (typically 2000–4000 km) and shorter lived than those in Jupiter (del Genio et al. 2009). Their dynamical properties are very similar to those of Jupiter although the cloud morphology tends to be more laminar and less turbulent. Neptune ovals are placed in regions of anticyclonic shear, north and south of the equator (they have been observed from latitudes 15ı to 55ı ), and their east–west length ranges from 5000 to 15,000 km. They have low albedo at short wavelengths where they are contrast enough to be identified against background clouds. Their cloud tops are at 2–3 bar level, i.e., at a lower altitude than surrounding clouds. A distinctive property of these features is that they are accompanied by patches of irregular clouds, placed at higher altitude than the vortex, so they detach by their high brightness at the wavelength of the methane absorption bands. Some of them have been observed to migrate in latitude, and their lifetime is shorter than those in the giants (Figs. 8 and 9).

Convective Storms and Planetary-Scale Disturbances Convective storms with sizes above 1000 km are regularly observed in Jupiter and Saturn. These storms manifest by the sudden outbreak of a bright and rapidly growing spot whose cloud tops elevate above the main cloud deck. Because of their vigorous development, moist convection is the mechanism that has been proposed as driving the storm (Hueso and Sánchez-Lavega 2001, 2004; Hueso et al. 2002; Ingersoll et al. 2004; del Genio et al. 2009; Sánchez-Lavega et al. 2017 and references therein). In Jupiter and Saturn, moist convection is due to latent heat release by ammonia and water (more important) on the ascending convective parcels. However, on the giant and icy planets, moist parcels weigh more than the bulk hydrogen atmosphere, as opposed to Earth where the moist parcels (water) weigh less than the ambient air. Therefore, convective motions in the giants are favored when the relative humidity of the environment is high and when preexisting vertical motions help to trigger the updrafts. Model predictions indicate that for Jupiter and Saturn, vertical velocities inside the upwelling parcels can reach 100 ms1 .

330

A. Sánchez-Lavega and M. Heimpel

d

100 m/s:

latitude

–15

–25

120

110 longitude

100

90

Fig. 8 The Great Red Spot (GRS) of Jupiter. (a) Image obtained during the Pioneer 11 flyby (1974); (b) this image was taken on July 6, 1979, by Voyager 2; (c) Galileo image on June 26, 1996; (d) wind velocity field in the GRS relative to its vorticity center from Voyager 2 images (Dowling and Ingersoll 1988). Image credit: NASA/JPL

Convective storms in Jupiter and Saturn grow until their cloud tops reach a horizontal size of 5000 km from when they interact with the meridional wind velocity shear of the background zonal flow. The convective outbreak (bright spots) can trigger large-scale planetary disturbances when, depending on the wind shear structure, a turbulent wake is continuously generated forward and/or backward of the storm source. The disturbance grows as a cloud pattern that moves with a velocity different to that of the source, encircling the planet along a latitude band where the storm source is placed. Examples are the disturbances that occur at 16ı S in the South Equatorial Belt of Jupiter (SEBD) and at 23.5ı N in the peak of the highest speed Jovian jet, known as the North Temperate Belt Disturbance or NTBD (Sánchez-Lavega et al. 2008). In such cases a band with a width of about 6ı –8ı in latitude is affected by the disturbance and transforms from a “zone-like” region to a dark belt. Single or multiple convective sources, known as “plumes” because of their aspect in infrared images, can act simultaneously in the disturbance development. In Saturn, moist convective storms have been observed at mid-latitudes and are usually irregular in shape with sizes of 3000 km. However, on rare occasions, a

16 Atmospheric Dynamics of Giants and Icy Planets

331

Fig. 9 Vortices in Jupiter, Saturn, and Uranus. (a) Jupiter’s anticyclone “white ovals” and a cyclone “brown” cell between them imaged by Galileo Orbiter on February 19, 1997; (b) Jupiter’s cyclone “barge” on July 6, 1979, by Voyager 2; (c) “Brown Spot” anticyclone in Saturn from Voyager 2 on August 23 and 24, 1981; (d) The Great Dark Spot of Neptune, an anticyclone, observed by Voyager 2 in 1989. Image credit: NASA/JPL

phenomenon known as the Great White Spot (GWS) occurs on Saturn. This is a major storm that evolves from a single and large convective source to a planetaryscale disturbance (Sánchez-Lavega et al. 2016 and references therein). It has been observed on six occasions (years 1876, 1903, 1933, 1960, 1990, and 2010), three at the equator, two at mid-latitudes, and one in a subpolar area, always in the northern hemisphere. As in Jupiter, the GWS disturbance evolution depends on the latitude, i.e., on the ambient zonal wind where the outbreak occurs. In 2010, the best studied case thanks to Cassini observations, the head of the storm (the initial plume) acquired a front-like shape, and the cloud morphology in its wake was dominated by periodic patterns of spots propagating at different speeds than the source. A large and long-lived anticyclonic vortex formed in this wake (Sayanagi et al. 2013). The outbreak originates from moist convection at the water cloud level (at pressure level

332

A. Sánchez-Lavega and M. Heimpel

10 bar), but the updrafts reach the 0.1 bar level, about 250 km altitude above the water clouds. The effects of the storm propagated into the stratosphere forming two warm areas at the 0.5–5 mbar pressure level altitude that merged into a so-called beacon because of its high infrared emission due to its 80 K temperature excess relative its surroundings. Uranus and Neptune display large and irregular bright spots that detach in particular when imaged in the methane band filters (0.89, 1.6, and 2.1–2.3 m,) sensitive to high-altitude cloud tops in the atmosphere. In the case of Uranus, it seems that this activity is related to the long-term seasonal insolation cycle. The brightest features in Uranus have east–west sizes of 5000 km with cloud tops at altitudes 300 mbar, but they distribute in altitude as occurred with the socalled Bright Northern Complex (de Pater et al. 2015). They are termed as “storms” (extremely bright features relative to background, often large and long lived), but their dynamical nature has not been specified. In Neptune bright spots with sizes up to 10,000 km E–W times 5000 km are observed regularly in the methane band filters at different latitudes (they are abundant in the tropical belt between latitudes 30ı and 40ı north and south). These are clouds made of methane ice that has high cloud tops at 200 mbar. Some of them are related to the dark ovals (i.e., to vortices, as presented in section 3b), but others form and evolve rapidly into filamentary streaks oriented along latitude circles surrounding both poles at latitudes 70ı N and 67ı –75ı S (Fig. 10).

Waves Waves are abundant in the upper troposphere and stratospheres of the giant and icy planets. They are observed at cloud level as regularly spaced periodic patterns and in the temperature field (as oscillations in the horizontal and vertical directions). Table 6 gives a compilation of the properties of some representative large-scale waves observed in Jupiter and Saturn, and in Fig. 11 we show some examples. Jupiter polar tropospheric clouds are overcast by a permanent haze that detaches in the ultraviolet and methane band filters. The edge of this haze is formed by a pattern of waves with wave numbers 12–14 that encircle the poles of Jupiter at different latitudes and move westward relative to the background flow, so they are probably a manifestation of Rossby waves. In the equatorial region of this planet at latitude 7ı N, there is a permanent family of “hot spots” (areas of high thermal emission at 5 m) that manifest as “dark projections” at visual wavelengths (in occasions with active bright plumes). The pattern has a zonal wave number of 10–12 and has been proposed to form part of an equatorial Rossby wave. Saturn has two unique and characteristic waves at cloud level, both located in the northern hemisphere: “the ribbon wave” at mid-latitudes and the “hexagon wave” surrounding the pole (del Genio et al. 2009 and references therein). The ribbon wave moves at the high speed of the eastward jet where it sits (velocity 120 ms1 ) and has a large wave number (45–60). The hexagon wave forms a stationary relative to the planet rotation although a high-speed eastward jet (velocity 120 ms1 ) flows

16 Atmospheric Dynamics of Giants and Icy Planets

333

Fig. 10 Storms and major disturbances in the giant and icy planets. (a) Jupiter North Temperate Belt Disturbance (NTBD, marked with an arrow) and “plumes” outbreak (disk) in March–May 2007. Credit: NASA/ESA/GCP-UPV/EHU and NASA/IRTF; (b) Jupiter’s South Equatorial Disturbance (SED, upper arrow) and South Tropical Disturbance (STrZD, lower arrow) from HST. Credit NASA/ESA/GCP-UPV/EHU; (c) Jupiter’s North Temperate Belt Disturbance (NTBD, upper arrow) and South Equatorial Belt Disturbance (SEBD, lower arrow) from HST. Credit NASA/ESA/GCP-UPV/EHU; (d) Saturn’s Great White Spot (GWS) 1990 – HST image in November 1990 and initial stages in October 1990 from Pic du Midi Observatory (inset). Credits: NASA/ESA and Pic du Midi Observatory; (e) and (f) Saturn’s GWS 2010 on December 24, 2010, and February 25, 2011, from Cassini ISS. Credit NASA/JPL; (g) Uranus active spots in the near infrared on July 12, 2004. Credit: L. Sromovsky, U. of Wisconsin–Madison/W. M. Keck Observatory; (h) Uranus brightest active storm in near infrared on August 6, 2014. Credit: I. de Pater (UC Berkeley)/W. M. Keck Observatory; (i) Neptune active storms at temperate latitudes in both hemispheres on July 26, 2007. Credit – W. M. Keck Observatory

334

A. Sánchez-Lavega and M. Heimpel

Table 6 Properties of representative large-scale waves in the giant planets Planet and wave Jupiter “polar methane” Jupiter – “polar UV” Jupiter “hot spots” Jupiter temperature Saturn “ribbon” Saturn “hexagon” Saturn temperature

Latitude 67ı S 55ı S 7ı N 20ı N-20ı S 47ı N 78ı N 20ı N-40ı N

Altitude (mbar) 100 100 700 20–250 500 500 250

n (zonal) 12–14 18–23 10–12 6–11 45–60 6 1–4

cx  u¯ (ms1 ) 10 to 30 30 – 95 0 0 100

Notes: (1) wave identification; (2) zonal wavenumber n; (3) phase velocity relative to the background wind. Adapted from a compilation in Sánchez-Lavega (2011) where references can be found

in its interior. It is not clear if these are Rossby waves or the result of a baroclinic instability mechanism. In addition to these large-scale waves, small-scale gravity waves (wavelength 300 km) have been observed in some occasions at cloud level in Jupiter’s equator. In the vertical direction, the signature of gravity waves has been recorded in all the giant and icy planets as temperature fluctuations (typically ˙ 1 K relative to the mean) at stratospheric levels (Ingersoll et al. 2004).

Polar Vortices Prominent polar vortices are present in the upper clouds of Saturn but not in Jupiter. This difference is intriguing in view of the dynamical similarities between the atmospheres of both planets. Saturn polar vortices have a radius of 1500 km and intense cyclonic vorticity with peak tangential velocities of 140 to 180 ms1 . The velocity decreases rapidly toward the vortex center. Both polar vortices have a rich and changing cloud morphology that distributes in rings and spiral shapes. The polar vortices are subsidence regions, and their cloud tops are at a lower altitude level than (Fig. 12, Sayanagi et al. 2016). In the case of Jupiter, images obtained by Juno spacecraft in its polar orbit since August 2016 (Bolton et al. 2007) have shown that no polar vortices exist on this planet at the upper cloud level but that chains of turbulent patterns of clouds and closed vortices encircle and move slowly around the poles. In the case of Uranus and Neptune, there is no evidence for the presence of polar vortices at cloud level. Uranus southern pole has been explored during the Voyager 2 flyby in 1986 (Allison et al. 1991; Karkoschka 2015) and later by large ground-based telescopes and the Hubble Space Telescope reaching the northern pole (Sromovsky et al. 2015). No polar vortices have been observed in agreement with the global wind profile shown in Fig. 5a. The southern polar region of Uranus has

0 0 -4 -2 0

60

65

70

75

80

85

90

0 0 0 0 0 20 40 60 80 10 12 14 16 18 u (m/s)

Fig. 11 Large-scale planetary waves in Jupiter and Saturn. (a) Jupiter dark projections in the North Equatorial Belt (these are hot spots at 5 m in the infrared) marked by arrows on December 11–12, 2000, from Cassini ISS. Credit: NASA/JPL; (b) Jupiter’s south polar wave imaged in the 890 nm methane absorption band from Cassini ISS. Grupo Ciencias Planetarias UPV/EHU; (c) and (d) Saturn’s “ribbon” wave from Voyager 1 images in November 1980. Credit NASA/JPL; (e) Saturn’s “hexagon” wave imaged by Cassini ISS on December 10, 2012. Credit NASA/JPL; (f) meridional profile of the wind speed in the north polar region of Saturn. The jet stream at latitude 75ı N is embedded and flows along the hexagonal wave (Antuñano et al. 2015)

Planetocentric Latitude (Degrees)

f

16 Atmospheric Dynamics of Giants and Icy Planets 335

336

A. Sánchez-Lavega and M. Heimpel

Fig. 12 Polar regions of Jupiter and Saturn. (a) Jupiter’s south pole on February 2, 2017 observed by JunoCam onboard Juno spacecraft. Credit NASA/SwRI/MSSS/D. Peach; (b) Saturn’s north polar cyclonic vortex on June 14, 2013, observed by Cassini ISS. Credit NASA/JPL/SSI; (c) and (d) visions of the south polar cyclonic vortex on Saturn imaged on July 14 and 15, 2008. Credit NASA/JPL/SSI

plenty of bright spots with sizes 500–1000 km enclosed by the jet at 60ı S that cluster around the pole. For Neptune, the wind profile shows a peak in the velocity of 275 ms1 at 75ı S (probably with a symmetric counterpart in the north, although not measured, Fig. 5b). No polar vortex has been reported so far.

16 Atmospheric Dynamics of Giants and Icy Planets

337

Theories and Models for the Zonal Jets The rich dynamics observed in giant planet atmospheres are driven primarily by two sources of energy. One energy source is solar radiation (insolation) deposited in the shallow atmosphere. The other is heat from the deep interior (section 1, Table 3, and Fig. 2). The ratio Q of emitted thermal energy to absorbed solar energy can be used to infer the relative forcing of planetary global circulation. The internal heat flux from Jupiter has been estimated from Voyager infrared observations to be roughly 5.4 W m2 and the energy ratio for Jupiter Q ' 1.7 (Hanel et al. 1981). Saturn (Q ' 1.8), Uranus (Q ' 1.1), and Neptune (Q ' 2.6) all have values of Q greater than unity (Aurnou et al. 2007). In addition, latitudinal thermal emission profiles of Jupiter and Saturn are relatively flat (Ingersoll 1976; Pirraglia 1984). This may imply efficient meridional mixing of the weather layer or enhanced internal heat flow toward the poles (Aurnou et al. 2007). Given the nearly equal contributions of deep and shallow driving, it is perhaps not surprising that attempts to model global atmospheric dynamics of the giant planets have fallen into these two rather separate categories. In this section we describe theoretical and computational modeling of the giant planets. We begin by describing models that include generation of magnetic fields in the deep metallic interiors of Jupiter and Saturn and the deep ionic oceans of Uranus and Neptune. Then, focusing on Jupiter and Saturn, we describe models of the electrically insulating molecular envelopes and the weather layer, where cloud motions are directly observed.

Magnetic Fields and Dynamo Models The four solar system giant planets, Jupiter, Saturn, Uranus, and Neptune, all exhibit broadly similar global dynamic features as shown in the previous section. Each has an intrinsic magnetic field, generated by dynamo action in an electrically conducting and convecting deep interior. All four planets also exhibit strong zonal flows that alternate in east–west directions with latitude. However, whereas the magnetic fields of Jupiter and Saturn are dominantly dipolar (Connerney 2007), those of Uranus and Neptune are multipolar, with strong contributions of dipolar, quadrupolar, and octupolar magnetic field components (Holme and Bloxham 1996). Also, while Jupiter and Saturn show similar large-scale atmospheric features, with prograde equatorial flow and multiple high-latitude jets, Uranus and Neptune each have a strong retrograde equatorial jet and only one high-latitude jet in each hemisphere. Each planet has a broad and powerful prograde (eastward) equatorial jet and higher latitude jets that are relatively narrow and weak. Compared to Jupiter, Saturn’s jet is broader and faster. In order to understand the origin of these large-scale features, we must ask how deeply the zonal flows are seated (Fig. 13). Whereas the tropospheres of the gas giants are primarily composed of molecular hydrogen (section 1), both laboratory experiments (Nellis et al. 1996) and theoretical models have shown that hydrogen dissociates gradually with increasing tempera-

338

A. Sánchez-Lavega and M. Heimpel

Fig. 13 Radial magnetic field on the outer surface and the left cut. Vertical and horizontal cuts show the azimuthal magnetic field. The thickness of the magnetic field lines is scaled with the third root of the local magnetic field strength (Gastine et al. 2014b)

ture and pressure. Thus, the electrical conductivity increases super-exponentially through the semiconducting molecular envelope, undergoing a transition to metallic hydrogen at a radius of roughly 0.9 RJ in Jupiter and roughly 0.7 RS in Saturn (French et al. 2012; Nettelmann et al. 2008).

Deep Convection Models Zonal winds may be driven by convection in the deep interior. Indirect evidence of deep convection is the existence of the global magnetic fields of all four solar system giant planets, which are sustained by the dynamo process. The dipolar magnetic fields of Jupiter and Saturn originate in the fluid metallic hydrogen interior. The multipolar magnetic fields of ice giants, Uranus and Neptune, originate in electrolytic oceans, beneath their thick atmospheres. The velocity of fluid flow in the deep interiors of Jupiter and Saturn is limited by the electrical conductivity (Guillot et al. 2004; French et al. 2012), and associated Lorentz forces, to the order of millimeters per second. This is an order of magnitude less than the typically 100 m/s velocities of giant planet zonal flows. Based on an estimate of the associated Ohmic dissipation, Liu et al. (2008) concluded that the zonal flows must therefore be confined to approximately 3000 km depth in Jupiter (corresponding

16 Atmospheric Dynamics of Giants and Icy Planets

339

to a scaled radius of 0.96 RJ, Jovian radius) and about 9000 km depth in Saturn (roughly 0.86 RS, Saturn radius). This magnetic exclusion of fast zonal flow to an outer region has been demonstrated in recent dynamo models that include radially variable electrical conductivity (Heimpel and Gómez Pérez 2011; Duarte et al. 2013). Figure 13 shows the resulting flow and magnetic fields of an anelastic dynamo model with variable electrical conductivity used to model Jupiter (Gastine et al. 2014b). It is not yet clear whether the electrical conductivity of the ionic oceans in Uranus and Neptune is sufficient to drastically decrease deep interior zonal flow velocities. However, deep convection models have been able to reproduce the large-scale zonal flows of Uranus and Neptune with highly turbulent convection in relatively thick spherical shells (Aurnou et al. 2007; Soderlund et al. 2013, Gastine et al. 2013). Uranus and Neptune each have a strong retrograde equatorial jet (opposite the flow direction of the equatorial jets of Jupiter and Saturn) and a single prograde jet in each hemisphere. Deep convection models reproduce these large-scale flow features when convection is sufficiently turbulent to effectively mix angular momentum in the fluid shell (Aurnou et al. 2007). This turbulent convective state is also associated with multipolar dynamo action, as strong three-dimensional mixing inhibits the dominance of the magnetic dipole component (Soderlund et al. 2013). Figure 14 shows results from a nonmagnetic numerical simulation of flow in a moderately deep spherical shell that produces zonal flows with the main features of those on the ice giants.

Fig. 14 Simulated motions in icy giants. Axial vorticity at a snapshot in time (a) and timeaveraged zonal (left) and meridional flow (b), for a deep convection model of the ice giants. Cyclonic (anticyclonic) vorticity is shown as red (blue) isosurfaces (a). Zonal (left) and meridional flow are shown in (b)

340

A. Sánchez-Lavega and M. Heimpel

The deep convection hypothesis has been tested by 3D numerical models of turbulent convection in rapidly rotating spherical shells (e.g., Christensen 2001; Heimpel et al. 2005; Kaspi et al. 2009). Rapid rotation causes convection to develop as axially oriented quasigeostrophic columns (Busse 1970). This columnar flow gives rise to Reynolds stresses, a statistical correlation between the convective flow components that feeds energy into zonal flows (e.g., Busse 1976). Prograde tilt of the convective columns and positive flux of angular momentum away from the rotation axis yield eastward equatorial flow (e.g., Zhang 1992). While such models can easily reproduce the correct direction and amplitude of the equatorial jets observed on Jupiter and Saturn, they have, until relatively recently, failed to produce multiple high-latitude jets (Christensen 2007). Numerical models in relatively deep layers and moderately small Ekman numbers (i.e., inner-to-outer radius ratio ri /ro D 0.6 and Ekman number E D 104  105 ) typically produce only a pair of jets in each hemisphere (Christensen 2001; Jones and Kuzanyan 2009; Gastine and Wicht 2012). Compared to these thick shell models, Boussinesq and anelastic models that use relatively thinner shells, which are more representative of the depth of the weakly magnetic molecular envelopes of Jupiter and Saturn, display relatively more numerous high-latitude jets (Heimpel et al. 2005; Jones and Kuzanyan 2009; Gastine et al. 2014a). These model jets are cylindrical flows that flow prograde and retrograde with respect to the average rotation rate. The equatorial jet forms outside the tangent cylinder (TC), the imaginary axial cylinder tangent to the inner boundary of the model spherical shell. Thus, prograde equatorial zonal flow spans the northern and southern hemispheres, whereas high-latitude jets form in either hemisphere and are truncated by the inner boundary (see Fig. 15). This difference between equatorial and high-latitude jets is manifest in differences in the jet speed and latitudinal wavelength, which is predicted quite well by a Rhines scale based on the topographic “-effect (Heimpel et al. 2005; Heimpel and Aurnou 2007). Deep convection models thus imply that giant planet zonal flows are the cloud level manifestation of cylindrical flows limited in depth where electrical conductivity and associated magnetic (Lorentz) forces inhibit fast zonal flows. More recently anelastic models with relatively strong density contrast and shallow stable stratification (see Fig. 16) have been shown to develop shallow vortices that coexist with deep, convectively driven jets (Heimpel et al. 2016). The zonal velocity and dynamical characteristics are dependent on the maximum zonal flow depth. Boussinesq dynamo models with radially variable electrical conductivity have been used to study the effect of varying the depth of the semiconducting molecular envelope (Heimpel and Gómez Pérez 2011). It was shown that for models with similar forcing, the equatorial jet zonal flow velocity, measured as a Rossby number based on the sphere radius, is proportional to the maximum depth of fast zonal flow. However, the zonal flow velocity scaled by a Rossby number based on the zonal flow depth is roughly independent of that depth. Thus, when comparing the zonal dynamics of different planets, it may be

16 Atmospheric Dynamics of Giants and Icy Planets

341

Fig. 15 Illustration of rapidly rotating deep convection in a spherical shell. Flow occurs between the outer radius ro and inner radius ri . The latitudes ™TC D ˙ cos1 (ri /ro ) mark the intersection of the tangent cylinder with the outer boundary. The shaded and white areas in the northern hemisphere correspond (on the outer surface) to the visible Jovian belts and zones (Heimpel et al. 2005).

Fig. 16 Results of a deep convection model with shallow stable stratification. (a) Zonal velocity on outer and inner boundaries and a meridional slice. Color saturation, in Rossby number units, is 0.009 (dark blue, westward) and C0.009 (dark red, eastward). (b) Axial vorticity for the same model. Color saturation: 0.6 (dark blue, anticyclonic) and C0.6 (dark red, cyclonic) in planetary rotation rate units. This shows vorticity of the zonal shear, which is much weaker than that of vortices. (Heimpel et al. 2016)

342

A. Sánchez-Lavega and M. Heimpel

preferable to estimate a zonal flow depth based on the Rossby number. Jupiter has peak equatorial wind speeds of roughly 100 m/s, at the cloud level. Saturn has peak equatorial winds of roughly 300–400 m/s, depending on the altitude of observation (see Figures 4, 6 and 7; Pérez-Hoyos and Sánchez-Lavega 2006). Using the estimated maximum zonal flow depths of d D 0.96 RJ and d D 0.86 RS , Liu et al. (2008) give Rossby numbers Ro D v/d of roughly 0.2, where v is the kinematic viscosity, for both Jupiter and Saturn. In addition to the maximum depth of fast zonal flows, the variation of zonal flow velocity with depth (vertical shear) is also important for the dynamics. However, there is at present no clear consensus on the precise form of vertical shear. For uniform density Boussinesq deep convection models, the vertical shear is minor. Zonal flow is rather strictly cylindrical and quasigeostrophic, with little variation from the outer radius to the inner radius. Indeed most Boussinesq and anelastic models employ free-slip boundary conditions at the outer and inner boundaries. This is clearly not as realistic as using outer free-slip and inner no-slip velocity boundary conditions, since we have argued that Lorentz forces slow the zonal flow at depth. However, given limitations on model resolution, most studies use the inner free-slip boundary condition because it favors the establishment of strong, coherent high-latitude jet structures (Aurnou and Heimpel 2004). Anelastic models, even with relatively strong density contrasts, have, in some studies, been consistent with Boussinesq models, yielding deep cylindrical flows with the equatorial jet outside the TC, high-latitude jets inside the TC, and minor variation of zonal velocity throughout the spherical shell depth (Jones and Kuzanyan 2009; Gastine and Wicht 2012; Gastine et al. 2014a). However, by using different internal heat sources and sinks, Kaspi et al. (2009) showed that anelastic models can also produce zonal jets with strong vertical shear, such that velocity decreases strongly with depth near the outer boundary. In this case, in addition to Reynolds stresses, a main driver of the zonal flow is the thermal wind.

Shallow Forcing and Hybrid Models These models are typically based on the hydrostatic approximation of fluid dynamics. Flow is modeled in the troposphere (e.g., Vallis 2006; Schneider and Liu 2009), and zonal winds are maintained by turbulent motions originating from several possible physical forcings that occur at the stably stratified cloud level, such as latent heat release and latitudinally variable insolation. Shallow flow models produce jets and vortices from 2D turbulence in a very thin spherical layer. A criticism of these models has been that they generically produce equatorial zonal flow in the westward direction, opposite the eastward equatorial jets on Jupiter and Saturn (e.g., Williams 1978; Cho and Polvani 1996; Showman 2007). However, shallow forcing models with additional forcing mechanisms, such as water vapor condensation (Lian and Showman 2010) or enhanced radiative damping (e.g., Scott and Polvani 2008), can reproduce observed equatorial superrotation. Figure 17 shows a result from a

16 Atmospheric Dynamics of Giants and Icy Planets

343

Fig. 17 Jupiter simulation with a shallow forcing global circulation model. Zonal velocity is shown at 0.65 bar at an instant in time (Schneider and Liu 2009)

shallow forcing model motivated by the fully three-dimensional quasigeostrophic situation, where axially cylindrical flows reach a maximum depth corresponding to an inner radius of 0.96 RJ (Liu et al. 2008). Here the tangent cylinder intersects the outer radius at a latitude of 16.3ı . By assuming that truncation of cylindrical flows is analogous to higher friction, the equatorial region has diminished dissipation (Schneider and Liu 2009). Thus the decreased equatorial friction is based on the deep convection hypothesis, and this model may be thought of as a hybrid of shallow and deep models.

Laboratory Experiments Laboratory experiments that yield zonal flows analogous to those on the giant planets have been based on the concept of conservation of potential vorticity in an apparatus of variable bottom topography (Read 2004; Cabanes et al. 2017). Whereas Read (2004) used a very large tank with linearly sloped bottom topography (“plane), Cabanes et al. (2017) experiment features curved topography at the top free boundary. Figure 18 shows the experimental setup and resulting zonal flows of Cabanes et al. (2017). The zonal flows are driven by Reynolds stress and the “-effect in a rotating tank of water. The depth of the layer varies parabolically with radius due to the effect of the centrifugal force on the upper free surface. This mimics the variable depth of axially directed fluid columns in the nearly spherical shell of the outer nonmagnetic part of a giant planet.

344

A. Sánchez-Lavega and M. Heimpel

a

b

Ω

ce. h

(r )

R

Free

Surf a

h(r )

h0 hmin Pump System

Rotary Table

Fig. 18 Laboratory experiment to simulate the flow in a giant planet. (a) Scheme of the experimental device. The tank is 1.37 m high by 0.5 m in radius with roughly 400 L of water. The tank is spun up to 75 revolutions per minute; (b) image of a simulated flow

Conclusions: The Perspective for Fluid Exoplanets The advance in the knowledge of the giants and icy planets will come on one hand from observations with current and future space missions (Juno/NASA, JUISE/ESA), space telescopes (James Web Space Telescope, JWSR, Norwood et al. 2016), and ground-based telescopes. On the other hand, progress in interpretation of these data is expected from advanced general circulation models (GCM) that incorporate detailed heating–cooling rates to new improved analysis of their internal structure, with support from experimental work on high-pressure and hightemperature physics. Detailed observations of thermal structure have been obtained for the tidally locked, so-called hot Jupiters (gas giants orbiting close to their parent stars). These observations and recent modeling efforts [Dobbs-Dixon and Lin 2008, Showman and Polvani 2011] have shown that the observed longitudinal thermal structure implies prograde equatorial flow, reminiscent of Jovian and Saturn zonal flows. Future observations with high spatial and spectral resolution from the different planned space missions and large ground-based telescopes (as the ESO E-ELT) will obtain data that will allow a direct comparison with the more advanced deep, shallow, and mixed dynamical models.

Cross-References  Internal Structure of Giant and Icy Planets: Importance of Heavy Elements and

Mixing  Temperature, Clouds, and Aerosols in Giant and Icy Planets

16 Atmospheric Dynamics of Giants and Icy Planets

345

Acknowledgments A.S.-L. research is supported by the Spanish project AYA2015-65041-P with FEDER support, Grupos Gobierno Vasco IT-765-13.

References Allison M, Beebe RF, Conrath BJ, Hinson DP, Ingersoll AP (1991) Uranus atmospheric dynamics and circulation. In: Bergstralh J, Miner E (eds) Uranus. University of Arizona Press, Tucson, pp 253–295 Antuñano A, del Rio-Gaztelurrutia T, Sánchez-Lavega A, Hueso R (2015) Dynamics of Saturn’s polar regions. J Geophys Res-Planets 120:155–176 Atkinson DH, Pollack JB, Seiff A (1998) The Galileo probe Doppler wind experiment: measurement of the deep zonal winds on Jupiter. J Geophys Res 103:22911–22928 Atreya AK, Wong AH (2005) Coupled clouds and chemistry of the Giant planets – a case for multiprobes. Space Sci Rev 116:121–136. https://doi.org/10.1007/s11214-005-1951-5 Aurnou JM, Heimpel M, Wicht J (2007) The effects of vigorous mixing in a convective model of zonal flow on the ice giants. Icarus 190(1):110–126 Aurnou JM, Heimpel M (2004) Zonal jets in rotating convection with mixed mechanical boundary conditions. Icarus 169:492–498 Baines KH, Hammel HB, Rages KA, Romani PN, Samuelson RE (1995) Clouds and hazes in the atmosphere of Neptune. In: Cruikshank DP (ed) Neptune and triton. University of Arizona Press, Tucson, pp 613–682 Bolton SJ, Adriani A, Adumitroaie V, Anderson J, Atreya S, Bloxham J, Brown S, Connerney JEP, DeJong E, Folkner W, Gautier D, Gulkis S, Guillot T, Hansen C, Hubbard WB, Iess L, Ingersoll A, Janssen M, Jorgensen J, Kaspi Y, Levin SM, Li C, Lunine J, Miguel Y, Orton G, Owen T, Ravine M, Smith E, Steffes P, Stone E, Stevenson D, Thorne R, Waite J (2017) Jupiter’s interior and deep atmosphere: the first close polar pass with the Juno spacecraft. Science 356:821–825 Busse FH (1970) Thermal instabilities in rapidly rotating systems. J Fluid Mech 44(03):441–460 Busse FH (1976) A simple model of convection in the jovian atmosphere. Icarus 20:255–260 Cabanes S, Aurnou J, Favier B, Le Bars M (2017) A laboratory model for deep-seated jets on the gas giants. Nat Phys 13:387–390 Cho JYK, Polvani LM (1996) The morphogenesis of bands and zonal winds in the atmospheres on the giant outer planets. Science 273:335–337 Choi DS, Showman AP, Brown RH (2009) Cloud features and zonal wind measurements of Saturn’s atmosphere as observed by Cassini/VIMS. Journal of Geophysical Research (Planets) 114:4007 Christensen UR (2001) Zonal flow driven by deep convection on the major planets. Geophys Res Lett 28:2553–2556 Christensen UR (2007) Zonal flow driven by strongly supercritical convection in rotating spherical shells. J Fluid Mech 470(115–133):2002 Connerney JEP (2007) Planetary magnetism. In: Schubert G (ed) Treatise on geophysics, vol 20. Elsevier, Amsterdam, pp 243–280 de Pater I, Sromovsky L, Fry PM, Hammel HB, Baranec C, Sayanagi K (2015) Record-breaking storm activity on Uranus in 2014. Icarus 252:121–128 del Genio AD, Achterberg RK, Baines KH, Flasar FM, Read PL, Sánchez-Lavega A, Showman AP (2009) Chapter 6: Saturn atmospheric structure and dynamics. In: Dougherty M, Esposito L, Krimigis T (eds) Saturn after Cassini-Huygens. Springer, Dordrecht, pp 113–159 Dobbs-Dixon I, Lin D (2008) Atmospheric dynamics of short-period extrasolar gas giant planets. i. Dependence of nightside temperature on opacity. Astrophys J 673(1):513 Duarte LD, Gastine T, Wicht J (2013) Anelastic dynamo models with variable electrical conductivity: an application to gas giants. Phys Earth Planet Inter 222:22–34 Dowling TE, Ingersoll AP (1988) Potential vorticity and layer thickness variations in the flow around Jupiter’s great red spot and white oval BC. J Atmos Sci 45:1380–1396

346

A. Sánchez-Lavega and M. Heimpel

Fletcher LN, Irwin PGJ, Achterberg RK, Orton GS, Flasar FM (2016a) Seasonal variability of Saturn’s tropospheric temperatures, winds and para-H2 from Cassini far-IR spectroscopy. Icarus 264:137–159 Fletcher LN, Greathouse TK, Orton GS, Sinclair JA, Giles RS, Irwin PGJ (2016b) Midinfrared mapping of Jupiter’s temperatures, aerosol opacity and chemical distributions with IRTF/TEXES. Icarus 278:128–161 Fouchet T et al (2008) An equatorial oscillation in Saturn’s middle atmosphere. Nature 453:200– 202 French M, Becker A, Lorenzen W, Nettelmann N, Bethkenhagen M, Wicht J, Redmer R (2012) Ab initio simulations for material properties along the jupiter adiabat. Astrophysical J Suppl Series 202(1):5–15 García-Melendo E, Sánchez-Lavega A (2001) A study of the stability of Jovian zonal winds from HST images: 1995–2000. Icarus 152:316–330 García-Melendo E, Pérez-Hoyos S, Sánchez-Lavega A, Hueso R (2011) Saturn’s zonal wind profile in 2004 - 2009 from Cassini ISS images and its long-term variability. Icarus 215: 62–74 Gastine T, Wicht J (2012) Effects of compressibility on driving zonal flow in gas giants. Icarus 219(1):428–442 Gastine T, Wicht J, Aurnou JM (2013) Zonal flow regimes in rotating anelastic spherical shells: an application to giant planets. Icarus 225:156–172 Gastine T, Heimpel M, Wicht J (2014a) Zonal flow scaling in rapidly-rotating compressible convection. Phys Earth Planet Inter 232:36–50 Gastine T, Wicht J, Duarte L, Heimpel M, Becker A (2014b) Explaining jupiter’s mag netic field and equatorial jet dynamics. Geophys Res Lett 41(15):5410–5419 Guillot T, Stevenson DJ, Hubbard W, Saumon D (2004) The interior of jupiter. In: Bagenal F, Dowling T, McKinnon W (eds) Jupiter, the planet, satellites and magnetosphere. Cambridge University Press, Cambridge, pp 35–67 Hanel RA, Conrath BJ, Herath LW, Kunde VG, Pirraglia JA (1981) Journal Geophysical Research 86:8705–8712 Heimpel M, Aurnou J (2007) Turbulent convection in rapidly rotating spherical shells: a model for equatorial and high latitude jets on jupiter and saturn. Icarus 187(2):540–557 Heimpel M, Gómez Pérez N (2011) On the relationship between zonal jets and dynamo action in giant planets. Geophys Res Lett 38(14):14201–14206 Heimpel M, Aurnou J, Wicht J (2005) Simulation of equatorial and high-latitude jets on Jupiter in a deep convection model. Nature 438:193–196 Heimpel M, Gastine T, Wicht J (2016) Simulation of deep-seated zonal jets and shallow vortices in gas giant atmospheres. Nat Geosci 9:19–23 Hueso R, Sánchez-Lavega A (2001) A three-dimensional model of moist convection for the giant planets: the Jupiter case. Icarus 151:257–274 Hueso R, Sánchez Lavega A, Guillot T (2002) A model for large scale convective storms in Júpiter. Journal Geophsical Research-Planets 107(10):5/1–5/11 Hueso R, Sánchez-Lavega A (2004) A three – dimensional model of moist convection for the Giant planets II: Saturn’s water and ammonia moist convective storms. Icarus 172:255–271 Holme R, Bloxham J (1996) The magnetic fields of uranus and neptune: methods and models. Journal of Geophysical Research: Planets 101(E1):2177–2200 Ingersoll AP, Barnet CD, Beebe RF, Flasar FM, Hinson DP, Limaye SS, Sromovsky LA, Suomi VE (1995) Dynamic meteorology of Neptune. In: Cruikshank DP (ed) Neptune and triton. University of Arizona Press, Tucson, pp 613–682 Ingersoll AP, Dowling TE, Gierasch PJ, Orton GS, Read PL, Sanchez-Lavega A, Showman AP, Simon-Miller AA, Vasavada AR (2004) Chapter 6: dynamics of Jupiter’s atmosphere. In: Bagenal F, McKinnon W, Dowling T (eds) Jupiter: the planet, satellites & magnetosphere. Cambridge University Press, Cambridge, pp 105–128 Ingersoll A (1976) Pioneer 10 and 11 observations and the dynamics of jupiter’s atmosphere. Icarus 29(2):245–253

16 Atmospheric Dynamics of Giants and Icy Planets

347

Jones CA, Kuzanyan KM (2009) Compressible convection in the deep atmospheres of giant planets. Icarus 204:227–238 Karkoschka E (2015) Uranus’ southern circulation revealed by voyager 2: unique characteristics. Icarus 250:294–307 Kaspi Y, Flierl GR, Showman AP (2009) The deep wind structure of the giant planets: results from an anelastic general circulation model. Icarus 202:525–542 Lian Y, Showman AP (2010) Generation of equatorial jets by large-scale latent heating on the giant planets. Icarus 207(1):373–393 Limaye SS (1986) Jupiter - new estimates of the mean zonal flow at the cloud level. Icarus 65: 335–352 Liu J, Goldreich PM, Stevenson DJ (2008) Constraints on deep-seated zonal winds inside Jupiter and Saturn. Icarus 196(2):653–664 Nellis WJ, Weir ST, Mitchell AC (1996) Metallization and electrical conductivity of hydrogen in jupiter. Science 273:936–938 Nettelmann N, Holst B, Kietzmann A, French M, Redmer R, Blaschke D (2008) Ab initio equation of state data for hydrogen, helium, and water and the internal structure of jupiter. Astrophys J 683(2):1217–1228 Norwood P, Moses J, Fletcher LN, Orton G, Irwin PGJ, Atreya S, Rages K, Cavalié T, SánchezLavega A, Hueso R (2016) Giant planet observations with the James Webb space telescope. Pub Astron Soc Pacific 128:018005 Orton GS, Friedson AJ, Caldwell J, Hammel HB, Baines KH, Bergstrahl JT et al (1991) Thermal maps of Jupiter: spatial organisation and time-dependence of stratospheric temperature, 1980 to 1990. Science 252:537–542 Pérez-Hoyos S, Sánchez-Lavega A (2006) On the vertical wind shear of Saturn’s equatorial jet at cloud level. Icarus 180(1):161–175 Pirraglia J (1984) Meridional energy balance of jupiter. Icarus 59(2):169–176 Porco CC, West RA, McEwen A, Del Genio AD, Ingersoll AP, Thomas P, Squyres S, Dones L, Murray CD, Johnson TV, Burns JA, Brahic A, Neukum G, Veverka J, Barbara JM, Denk T, Evans M, Ferrier JJ, Geissler P, Helfenstein P, Roatsch T, Throop H, Tiscareno M, Vasavada AR (2003) Cassini imaging of Jupiter’s atmosphere, satellites, and rings. Science 299: 1541–1547 Read PL (2004) Jupiter’s and saturn’s convectively driven banded jets in the laboratory. Geophys Res Lett 31. https://doi.org/10.1029/GL020106 Rogers JH (1995) The giant planet Jupiter. Cambridge Univ Press, Cambridge, UK Sánchez-Lavega A, Rojas JF, Sada PV (2000) Saturn’s zonal winds at cloud level. Icarus 147: 405–420 Sánchez-Lavega A, Orton GS, Hueso R, García-Melendo E, Pérez-Hoyos S, Simon-Miller A, Rojas JF, Gómez JM, Yanamandra-Fisher P, Fletcher L, Joels J, Kemerer J, Hora J, Karkoschka E, de Pater I, Wong MH, Marcus PS, Pinilla-Alonso N, Carvalho F, Go C, Parker D, Salway M, Valimberti M, Wesley A, Pujiv Z (2008) Depth of a strong jovian jet from a planetary-scale disturbance driven by storms. Nature 451:437–440 Sánchez-Lavega A, del Río-Gaztelurrutia T, Hueso R, Gómez-Forrellad JM, Sanz-Requena JF, Legarreta J, García-Melendo E, Colas F, Lecacheux J, Fletcher LN, Barrado-Navascués D, Parker D, the International Outer PlanetWatch Team (2011) Deep winds beneath Saturn’s upper clouds from a seasonal long-lived planetary-scale storm. Nature 475:71–74. https://doi.org/10.1038/nature10203 Sánchez-Lavega A (2011) An introduction to planetary atmospheres. Taylor & Francis/CRC Press, Boca Raton. 696 pp Sánchez-Lavega A, Fisher G, Fletcher LN, Garcia-Melendo E, Hesman B, Perez-Hoyos S, Sayanagi K, Sromovsky L (2016) Chapter 13: The Great Storm of 2010–2011. In: Baines KH, Flasar FM, Krupp N, Stallard TS (eds) Saturn in the 21st Century. Cambridge University Press, Cambridge. (in the press) https://arxiv.org/abs/1611.07669 Sánchez-Lavega A, Sromovsky L, Showman A, Del Genio A, Young R, Hueso R, García Melendo E, Kaspi Y, Orton GS, Barrado-Izagirre N, Choi D, Barbara J (2017) Zonal Jets in Gas Giants.

348

A. Sánchez-Lavega and M. Heimpel

In: Galperin B, Read P, ISSI (eds) Zonal Jets. Cambridge University Press. (in the press), Cambridge Sayanagi K, Dyudina UA, Ewald SP, Fisher G, Ingersoll AP, Kurth WS, Muro GD, Porco CC, West RA (2013) Dynamics of Saturn’s great storm of 2010–2011 from Cassini ISS and RPWS. Icarus 223:460–478 Sayanagi K, Baines KH, Dyudina U, Fletcher LN, Sánchez-Lavega A, West RA (2016) Chapter12. Saturn’s Polar Atmosphere. In: Baines KH, Flasar FM, Krupp N, Stallard TS (eds) Saturn in the 21st Century. Cambridge University Press, Cambridge. (in the press).arXiv_1609.09626v2 Schneider T, Liu J (2009) Formation of jets and equatorial superrotation on Jupiter. J Atmos Sci 66:579–601 Scott R, Polvani LM (2008) Equatorial superrotation in shallow atmospheres. Geophys Res Lett 35(24):24202–24206 Showman AP (2007) Numerical simulations of forced shallow-water turbulence: effects of moist convection on the large-scale circulation of Jupiter and Saturn. J Atmos Sci 64(9):3132–3157 Showman AP, Polvani LM (2011) Equatorial superrotation on tidally locked exoplanets. Astrophys J 738(1):71 Soderlund K, Heimpel M, King E, Aurnou J (2013) Turbulent models of ice giant internal dynamics: dynamos, heat transfer, and zonal flows. Icarus 224(1):97–113 Sromovsky LA, de Pater I, Fry PM, Hammel HB, Marcus P (2015) High S/N keck and Gemini AO imaging of Uranus during 2012–2014: new cloud patterns, increasing activity, and improved wind measurements. Icarus 258:192–223 Vallis GK (2006) Atmospheric and oceanic fluid dynamics: fundamentals and large-scale circulation. Cambridge University Press, Cambridge West RA, Baines KH, Karkoschka E, Sánchez-Lavega A (2009) Chapter 7: clouds and aerosols in Saturn’s atmosphere. In: Dougherty M, Esposito L, Krimigis T (eds) Saturn after CassiniHuygens. Springer, pp 161–179 Williams GP (1978) Planetary circulations: 1. Barotropic representation of jovian and terrestrial turbulence. J Atmos Sci 35(8):1399–1426 Zhang K (1992) Spiralling columnar convection in rapidly rotating spherical fluid shells. J Fluid Mech 236:535–556

Upper Atmospheres and Ionospheres of Planets and Satellites

17

Antonio García Muñoz, Tommi T. Koskinen, and Panayotis Lavvas

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Earth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Venus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ion Exospheres . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Upper Atmospheres of Jupiter, Saturn, Uranus, and Neptune . . . . . . . . . . . . . . . . . . . . . . Observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Thermospheres . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ionospheres . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Titan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Photochemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Aerosol Formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Energetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

350 355 357 359 360 361 362 363 365 366 366 367 368 369

Abstract

The upper atmospheres of the planets and their satellites are more directly exposed to sunlight and solar-wind particles than the surface or the deeper

A. García Muñoz () Zentrum für Astronomie und Astrophysik, Berlin, TU, Germany Technische Universität Berlin, Berlin, Germany e-mail: [email protected] T. T. Koskinen Lunar and Planetary Laboratory, University of Arizona, Tucson, AZ, USA e-mail: [email protected] P. Lavvas GSMA, UMR 7331, CNRS, Université de Reims, Champagne-Ardenne, Reims, France e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_52

349

350

A. García Muñoz et al.

atmospheric layers. At the altitudes where the associated energy is deposited, the atmospheres may become ionized and are referred to as ionospheres. The details of the photon and particle interactions with the upper atmosphere depend strongly on whether the object has an intrinsic magnetic field that may channel the precipitating particles into the atmosphere or drive the atmospheric gas out to space. Important implications of these interactions include atmospheric loss over diverse timescales, photochemistry, and the formation of aerosols, which affect the evolution, composition, and remote sensing of the planets (satellites). The upper atmosphere connects the planet (satellite) bulk composition to the nearplanet (-satellite) environment. Understanding the relevant physics and chemistry provides insight to the past and future conditions of these objects, which is critical for understanding their evolution. This chapter introduces the basic concepts of upper atmospheres and ionospheres in our solar system and discusses aspects of their neutral and ion composition, wind dynamics, and energy budget. This knowledge is key to putting in context the observations of upper atmospheres and haze on exoplanets and to devise a theory that explains exoplanet demographics.

Introduction The fundamentals of upper atmospheres and ionospheres have been established in numerous works, including Bauer (1973), Rees (1989), and Schunk and Nagy (2000). Up-to-date views are presented in, e.g., Mendillo et al. (2002) and Nagy et al. (2008). The material covered in this chapter draws from these references as well as from others more specific. The scope of the chapter is that of aeronomy, which refers to the investigation of upper atmospheres and ionospheres as a subfield within the atmospheric sciences. For convenience, the text often refers to planets although it would be appropriate to refer to planets and their satellites. Planetary atmospheres are vertically stratified, and it is common to differentiate various regions according to temperature or composition. Based on thermal structure, here the term upper atmosphere is used to encompass the thermosphere and exosphere. The thermosphere is typically heated by EUV and X-ray solar photons and shows a positive gradient of temperature with altitude up to an asymptotic value known as the exospheric temperature, T1 . Atop the thermosphere, in the exosphere gas particle collisions become rare, and particles with velocities larger than the gravitational escape velocity v1 may leave to space. The exobase is the exospheric lower boundary and, by convention, occurs where H/1, with H being the atmospheric pressure scale height and  the gas mean free path. Gas particles leave the exosphere in various ways. Thermal Jeans escape involves the dimensionless parameter Xexo DGMm/Rexo kT1 , formed as the ratio of the atom (or molecule) gravitational and thermal energies. G is the gravitational constant, M the planet mass, m the particle mass, and Rexo the distance from the planet center to the exobase. Jeans escape occurs for large Xexo values. In this regime, the velocity distribution of gas particles remains essentially Maxwellian. Massive hydrodynamic escape occurs for Xexo 2–3 (Volkov et al. 2011), a condition easier to attain for light gases (H, H2 , He) at high temperatures and on low-mass planets.

17 Upper Atmospheres and Ionospheres of Planets and Satellites

351

Nonthermal escape processes include charge exchange, loss through open magnetic field lines, photoionization and dissociative recombination, and solar-wind pickup and sputtering. They may drive the escape of both neutrals and ions. The existence of a planetary magnetic field affects some of these processes, whose relative significance may change over the atmosphere’s lifetime. Understanding the current escape processes is key to inferring a planet’s history and predicting its evolution. Based on neutral gas composition, the heterosphere refers to the altitudes for which molecular diffusion is more efficient in the transport of gases than eddy diffusion by large-scale winds and turbulent motions. Typically, the upper atmosphere overlaps with the heterosphere. The base of the heterosphere is at the homopause. Above it, the density of each long-lived gas drops according to its own scale height, which is inversely proportional to their corresponding mass. This separation by mass means that only the lighter gases reach the exosphere from where they can escape. Hydrodynamic escape conditions tend to facilitate the access of the heavier gases to the exosphere, thereby attenuating their separation by mass in the heterosphere. The upper atmosphere is exposed to EUV (10–121 nm) and X-ray (0.01–10 nm) solar photons, as well as to cosmic rays and particles of auroral or solar-wind origin, all of which ionize the neutral gas and produce a weakly ionized plasma. Neutrals, ions, and electrons of planet origin coexist in the ionosphere and interact to some extent with the incoming solar wind. On planets without an intrinsic magnetic field, the ionopause sets the boundary between the dayside ionosphere and the solar wind and is perceived as an abrupt drop in the planet plasma density. The ionopause occurs as a balance between the solar-wind dynamic pressure and the planet plasma pressure and acts as an obstacle deflecting the incident solar wind. In the absence of an intrinsic magnetic field, the nightside ionosphere may extend as a tail on the planet shadow. For planets having a magnetic field, the ionosphere is contained inside the magnetosphere, and the planet plasma is confined by the field lines. Ions and electrons escape through open magnetic field lines, a process that is known as polar wind. Meteoritic material ablated as it enters the atmosphere may produce sporadic ionization layers. Airglow and aurora are photoemission phenomena that offer unique opportunities for the remote sensing of upper atmospheres. They result from excited atoms and molecules radiating away their excess energy, thereby providing insight to the emitting gas (identity, abundance, production rate) and into the background atmosphere (density, temperature, velocity, energetic particle fluxes). The aurora is excited by precipitating electrons and ions from outside the atmosphere. Airglow is divided into day and night airglow (dayglow and nightglow, respectively). Sunlight is the ultimate excitation mechanism for airglow emissions, although the connection with solar photons is more direct for the dayglow than for the nightglow. What follows reviews the aeronomy of the terrestrial planets (Earth, Venus, Mars), the gas giants (Jupiter, Saturn, Uranus, Neptune), and Saturn’s moon Titan (see also Table 1). The topic of ion exospheres is briefly mentioned in its application to Mercury and the Moon. We specifically acknowledge the authors of many seminal papers that have contributed greatly to the present knowledge of solar system aeronomy but that could not be cited here due to space limitations.

Intrinsic magnetic field

500

Exobase [altitude, km]

500–1000

T1 [K]

11.2

CO2 , N2 , O, CO

O2 C , CO2 C , OC , NOC

Neutrals Ions N2 , O2 , NOC , O2 C , O, N, OC He

v1 Thermosphere: [km/s]a main gases

Venus 135/7  106 No – 220–350 300 (day) 10.4 ionopause /100 at 225– (night) 375 km

Earth 100/3  107 Yes

Homopause [altitude, km/pressure, bar] Night O(1 S); 557 nm

Some airglow emissions

Diffuse, O(3 S0 ) and O(5 S0 ); 130.4 and 135.6 nm

O2 (b); 762 nm O2 (a); 1270 nm CO(a); 190–260 nm

O(1 S); 557 nm

O(1 D); 630 nm

Day N2 (a); 140–180 nm

O2 (a); 1270 nm OH(X); Vis-IR NO(C, A); 180–300, 1220 nm

O(1 S); 557 nm O2 (a); 1270 nm

O2 (c); 400–700 nm CO2 C (B); 289 nm

O(1 S); 557 nm

O2 (A,A’,c); UV-NIR H, O, N, OC , NC , O2 (a); 1270 nm H2 , N2 ; 90–200 nm OH(X); Vis-IR

OC ; < 90 nm

” rays to radio frequencies

Aurora

Table 1 Summary of properties of the upper atmospheres and ionospheres of solar system planets and Titan

352 A. García Muñoz et al.

Saturn

108 – 107

Yes

Yes Yes

2850

1600

400–600

900–1000

35.5

4.3 59.5

130/3  1010 Crustal 180–210 200–350 5.0 (low/high solar activity)

Mercury Jupiter 106

Mars

H2 , He

H2 , He

CO2 , N2 , O, CO

HC , H3 C , HeC

NaC , OC HC , H3 C , HeC

O2 C , CO2 C , OC , NOC

O(1 S); 297 nm

O2 (a) 1270 nm

(continued)

O2 (a); 1270 nm Na; 589 nm H Lyman-’, H2 UV bands (He 58.4 nm) H Lyman-’ H2 UV bands He 58.4 nm

CO2 C (B); 289 nm

OH (X); 1500–3000 nm

Main: H Lyman-’, H Lyman ’ H2 UV bands, H3 C IR bands Main: H Lyman-’ H Lyman-’ (He 58.4 nm) H2 UV bands H3 C IR bands

Discrete CO(A), CO(a), COC (B), CO2 C (B); 130–300 nm Diffuse: CO(a), CO2 C (B), O(1 S)

NO(C, A); 180–300 CO(a); 190–260 nm

17 Upper Atmospheres and Ionospheres of Planets and Satellites 353

No

1500

2200

4700

150

600

850

T1 [K]

2.6

23.5

21.3

N2 , CH4 , H2

H2 , He

HC , H3 C , HeC Main: HCNHC , C2 H5 C C 100 s of other C/N/H/O composition ions

Neutrals Ions H2 , He HC , H3 C , HeC

v1 Thermosphere: [km/s]a main gases

Additional refs.: a https://nssdc.gsfc.nasa.gov/planetary/factsheet/index.html

850

Yes

Neptune 106

Titan

Yes

Intrinsic Exobase mag[altitude, netic km] field

Uranus 104

Homopause [altitude, km/pressure, bar]

Table 1 (continued)

H Lyman-’

Night H Lyman-’

Some airglow emissions

Day H Lyman-’ H2 UV bands

H Lyman-’, H2 UV bands From Saturn’s From Saturn’s N2 : CY(0,0), magnetosphere, magnetosphere, LBH, VK NI: similar emissions as similar emissions as 120, 124.3, 149.3 from solar photons from solar photons nm NII: 108.5 nm

H Lyman-’ (?) H2 UV bands H3 C IR bands H2 UV bands (?)

Aurora

354 A. García Muñoz et al.

17 Upper Atmospheres and Ionospheres of Planets and Satellites

355

Earth The Earth has been investigated much more thoroughly than the other planets. Only a brief overview of its aeronomy is given here, to serve as background for Venus and Mars. The bibliography quoted above provides valuable references for more extensive treatments. The Earth thermosphere extends from 80 to 500 km. Solar activity causes significant variability (on the order of hundreds of km) in the exobase level: higher activity implies additional energy deposited into the atmosphere, which expands as a consequence, and vice versa (Fig. 1). Smaller variations in the exobase level occur on diurnal timescales. For low (high) solar activity, usual exospheric temperatures are 500 (1000) K. The low abundance of CO2 or other efficient IR radiators (e.g., NO) above the homopause (100 km or 3  107 bar) minimizes the thermostat effect that occurs on Venus and Mars and that prevents extreme exospheric temperature variations on these planets. Photodissociation of N2 and O2 (main neutrals in the bulk atmosphere) and molecular diffusion result in O and N

Fig. 1 Thermospheric temperatures for the Earth, Venus, and Mars. Black: Earth profiles for dayside conditions, as obtained from the NRLMSISE-00 model (http://ccmc.gsfc.nasa.gov/modelweb/ models/nrlmsise00.php). Green: Venus profiles for day- and nightside conditions (VIRA model, moderate solar activity; Keating et al. 1985) and for morning terminator conditions (solar occultation measurements; Mahieux et al. 2015). Red: Mars profiles for dayside (Bougher et al. 2015b) and nightside conditions (stellar occultation measurements; Forget et al. 2009). Error bars are omitted

356

A. García Muñoz et al.

Fig. 2 Neutral gas composition of the Earth (left, NRLMSISE-00 model at high solar activity, dayside), Venus (VIRA model at moderate solar activity; Keating et al. 1985), and Mars (MAVEN/NGIMS measurements; Bougher et al. 2015b). Error bars are omitted

Fig. 3 Dayside ionosphere of the Earth (left, IRI-2007 model; http://ccmc.gsfc.nasa.gov/ modelweb/models/iri_vitmo.php), Venus (photochemical model; Fox and Sung 2001), and Mars (Bougher et al. 2015b). Error bars are omitted

atoms (with He) becoming abundant in the thermosphere (Fig. 2). In the exosphere, the lighter gases H and He prevail. Earth’s ionosphere is traditionally split into D, E, F1 , and F2 layers, ordered from lower to higher altitude. This denomination has guided the naming of other ionospheres. Molecular ions (NOC , O2 C ) dominate in the D and E layers, whereas OC , the product of O photoionization, tends to dominate in the F layers (Fig. 3). Peak electron densities occur in the F2 layer at and above which ion diffusion becomes important.

17 Upper Atmospheres and Ionospheres of Planets and Satellites

357

Hydrodynamic escape is currently ineffectual on Earth but likely to have occurred in the past, especially if exospheric hydrogen was abundant and the young Sun’s EUV output stronger (Hunten 1993; Catling and Zahnle 2009). The impact of nonthermal escape over the planet lifetime is sensitive to the existence of a magnetic field in the early Earth, a possibility that adds uncertainty to the evolution of the terrestrial atmosphere (Lammer et al. 2008). Earth shows a variety of airglow and auroral emissions sharing both commonalities and differences with the Venus and Mars emissions (Meier 1991; Galand and Chakrabarti; Slanger and Wolven, both in Mendillo et al. 2002). Some similarities arise from the fact that O and N atoms are produced on all three planets by photodissociation of O2 /N2 (Earth) and CO2 /N2 (Venus, Mars). The lack of an intrinsic magnetic field on Venus and Mars causes discrepancies in their aurora excitation with respect to Earth.

Venus The Venus homopause lies at 135 km altitude (7  106 bar), the exact level being lower for light gases (H, H2 , He) and higher for the heavy ones. The thermosphere extends from 120 to 220–350 km, and its neutral composition is dominated by CO2 and N2 up to 140–160 km (Fig. 2). Above, CO2 photodissociation and diffusion cause O and CO to become dominant. In the exosphere, H and He are major constituents. Venus’ ionosphere is also structured in layers that reveal changes in electron densities and ionizing processes (Pätzold et al. 2007). The secondary ionization layer rests at the base of the ionosphere at 120 km. Above it, the main ionization layer reaches electron densities of a few times 105 cm3 at its 140 km peak. These layers are caused by soft X-ray (secondary) and EUV (main) photon photoionization, respectively, and tend to form CO2 C . Rapid reaction of CO2 C with O leads to O2 C as the main ion up to 180 km (Fig. 3). The O2 C ion is lost in the dissociative reaction with electrons, forming two O atoms. A third ionization layer dominated by OC occurs above 180 km. A sporadic ionization layer attributed to ablating meteoroids occurs near 115 km, which may consist of MgC and FeC ions (Withers 2012). Electron densities on Venus are highly variable in the topside ionosphere and decay abruptly near the ionopause. For solar minimum conditions, the ionopause level fluctuates between 225 and 375 km within a few days. This variability reflects the ever-changing interaction between the solar-wind and planet plasma. On the nightside, a weak ionosphere occurs sustained by OC transported from the dayside and precipitating electrons. At times, the nightside ionosphere nearly vanishes (Cravens et al. 1982). Fast thermospheric winds participate in a subsolar-to-antisolar (SS-AS) circulation above 100 km driven by dayside solar heating that produces a day-to-night pressure gradient (Fox and Bougher 1991; Clancy et al. 2015). Upwelling and downwelling occur near the subsolar and antisolar points, respectively. Below

358

A. García Muñoz et al.

100 to 120 km, the SS-AS circulation connects with the super-rotating flow that dominates Venus’ lower-atmosphere circulation. It takes a few days for the O and N atoms formed by CO2 and N2 photodissociation to be transported to the nightside and recombine, resulting in O2 , NO, and OH nightglow at UV-to-NIR wavelengths and 90–130 km altitudes (García Muñoz et al. 2009). These emissions are variable and asymmetric with respect to the antisolar point, which suggests complex wind interactions and competing quenching-vs-radiation effects. Venus’ dayglow includes emissions from He, HeC , H, O, OC , N, C, CC , CO, and CO2 C , whose interpretation requires elaborate modeling of different excitation processes (Fox and Sung 2001). A diffuse oxygen aurora exists on the nightside, possibly excited by precipitating energetic electrons. It is unclear what drives the precipitation in the absence of an intrinsic magnetic field, although magnetic reconnection might provide such a mechanism (Zhang et al. 2012). The intermittent oxygen green line nightglow, potentially correlating with solar activity, may prove key to resolving some of these uncertainties (Slanger et al. 2001). X-ray emission, whether the result of fluorescent scattering of solar X-rays or charge exchange interactions with the solar wind, provides a complementary and as-of-yet little explored view of the Venus thermosphere-exosphere and their interaction with the Sun (Dennerl 2008). The ion and electron temperatures reach thousands of degrees in the exosphere, thus exceeding the neutral temperature over most of the ionosphere (Miller et al. 1980). The neutral thermosphere is relatively cold with temperatures of 300 K at the dayside exobase, much less than at Earth (Fig. 1). On the nightside, thermospheric temperatures of 100 K are the lowest on Venus, and this region is often referred to as the cryosphere. Radiation at 15 m from CO2 (a trace gas in the terrestrial atmosphere but abundant on Venus and Mars) in nonlocal thermodynamical equilibrium efficiently cools the Venus thermosphere and attenuates the potentially larger impact of solar activity at an inner planet. The day-night thermal contrast is coupled with the SS-AS circulation. Waves associated with density modulations (Müller-Wodarg et al. 2016), possibly originating from the lower atmosphere, have been reported in the thermosphere. Escape from the Venus upper atmosphere is presently dominated by nonthermal processes (Lammer et al. 2008; Catling and Zahnle 2009), although thermal (hydrodynamic) loss may have been significant in the past. Indeed, the ancient Venus may have experienced a runaway greenhouse effect that resulted in abundant upper-atmosphere steam. Exposed to EUV sunlight, the water would dissociate, with the H atoms thermally escaping more easily than the heavier O atoms. A mass-based fractionation would also occur for the D isotope, which might explain the D/H ratio of (1.6–2.2)  102 , two orders of magnitude larger than at Earth. Hydrodynamic escape may have removed a water ocean (if it existed) in less than 100,000 years. This example highlights the importance of the upper atmosphere for planetary evolution.

17 Upper Atmospheres and Ionospheres of Planets and Satellites

359

Mars Both Mars and Venus lack intrinsic magnetic fields, although Mars does have a leftover crustal field after its presumed original intrinsic field vanished 4 gigayears ago (Mangold et al. 2016). This remnant field is apparent over localized near-surface areas of Mars and affects aspects of its aeronomy such as the ionospheric structure, interaction with the solar wind, and aurora. The Martian homopause lies at 130 km (3  1010 bar). Lower down, CO2 and N2 are well mixed and dominate the background composition (Fig. 2). Above, CO and O are in diffusive equilibrium and become locally abundant. O takes over as the main atmospheric gas for altitudes larger than 200 km, near the exobase (Bougher et al. 2015a, b; Bhardwaj et al. 2016). The bottom boundary of the Mars dayside thermosphere lies at 100 km, where the temperature is 120 K (Fig. 1). The temperature rises to exospheric values of 200 and 350 K for low and high solar activity conditions, respectively (MuellerWodarg et al. 2008; Bougher et al. 2015a). Thermal conduction and CO2 cooling at 15 m, excited in collisions with O, balance the heating by EUV solar energy deposition. The fact that the temperature fluctuations in the Martian thermosphere are larger than for Venus suggests that the CO2 thermostat is less efficient on Mars, probably because the Martian O/CO2 density ratio is smaller. The magnitude of the day-night temperature variations in the thermosphere can be up to 200 K depending on solar activity. Thermospheric winds transport heat from the dayside to polar latitudes and on to the nightside (Bougher et al. 2015a; González-Galindo et al. 2015). They also transport the O and N products of CO2 and N2 photodissociation that recombine on the nightside to produce nightglow (e.g., Clancy et al. 2013). The lower and upper atmospheres of Mars are strongly coupled. The thermosphere is forced by upward-propagating tides and gravity waves originating near the surface, possibly affected by topography. Dust storms cause transient heat deposition that modifies these wave interactions and result in thermospheric density changes by factors of a few (Withers and Pratt 2013). Electron densities in the dayside ionosphere have been measured from 80 to 500 km. Peak densities are a few times 105 cm3 at 120–140 km within the main ionization layer, which is ionized by EUV solar photons (Fig. 3). A secondary layer is formed at 100 km by soft X-ray solar photons. Similar to Venus, O2 C is the dominant ion in the main ionization layer, and NOC is predicted to contribute to the secondary ionization layer (Krasnopolsky 2002; Fox 2009). The O2 C and OC densities become comparable at 250–300 km. Peak electron densities on the nightside are typically two orders of magnitude smaller than on the dayside, consistent with an origin due to plasma transport from the dayside and electron precipitation (Withers et al. 2012). Abrupt drops in dayside electron density characteristic of ionopause-like configurations have been reported (Vogt et al. 2015) and seem to be sensitive to solar-wind conditions and to the local crustal magnetic field.

360

A. García Muñoz et al.

Mars dayglow includes emissions from N2 , CO, CO2 C , C, O, H2 , and H that originate from altitudes up to 200 km. The gases are excited either directly (e.g., CO2 Cphoton leading to excited states of O, CO2 C , CO, COC ) or indirectly via neutral and ion chemistry (e.g., O2 C Ce➔O(1 S)CO). The scale heights for each emission profile reflect the details of the primary (and secondary) excitation mechanisms and are used to infer neutral densities and exospheric temperatures (Huestis et al. 2010). Precipitating electrons of energies 300–1000 eV are accelerated by the crustal field and produced upon collision with the neutral atmosphere sporadic aurorae at 130 km that resemble Earth’s discrete aurora (Bertaux et al. 2005). Mars also exhibits a planetwide diffuse aurora that reaches down to 60 km and is likely excited by solar particles of hundreds of keV (Schneider et al. 2015). The moderately elevated D/H ratio in the Martian atmosphere (a few times that of Earth) and the evidence for past liquid on its surface suggest that the planet may have lost a substantial amount of its initial water (Hunten 1993). The possibility for surface water shows the contrast between Mars’ early (wet, warm) and current (dry, cold) climates. On current Mars, thermal (Jeans) escape is relevant for the loss of hydrogen, whereas nonthermal escape contributes to the loss of heavier atoms (Lammer et al. 2008; Catling and Zahnle 2009). Hydrodynamic escape is not operating now but is likely to have operated in the past. Bursts of solar activity, as during coronal mass ejections, were more frequent for the early Sun than they are now and may have significantly enhanced the past escape rates (Jakosky et al. 2015). The recent detection of a transient extended brightness feature of uncertain physical origin at the Mars morning terminator and 250 km altitude (Sánchez-Lavega et al. 2015) shows that there remain significant uncertainties in our understanding of the Mars upper atmosphere. Three space missions (ExoMars, MAVEN, MOM) have recently entered into Mars orbit and are contributing to a better understanding of Martian aeronomy. Some mission highlights include the discovery of metal ion layers of NaC , MgC , and FeC (Grebowsky et al. 2017) or a better constraint on the atmospheric Argon fractionation and, in turn, the prediction that about two-thirds of this noble gas (and a sizeable amount of the bulk CO2 ) may have been lost to space over the planet history (Jakosky et al. 2017).

Ion Exospheres Ionospheres occur also on bodies with tenuous atmospheres such as Mercury or the Moon. These ion exospheres contain the signature of the surface and interior material that is being released. Mercury’s ionosphere contains NaC and OC concentrated preferentially near the magnetic poles, which points to these regions as sources of the heavy ions probably through solar-wind sputtering. The lighter ion HeC is observed more uniformly

17 Upper Atmospheres and Ionospheres of Planets and Satellites

361

around the planet, and its distribution is consistent with planetwide evaporation from the surface (Zurbuchen et al. 2011). There is evidence that the near-Moon environment is partly ionized and that electron densities can reach values of 103 cm3 (Choudhary et al. 2016). Modeling suggests that the measured plasma is consistent with molecular ions of H2 OC , CO2 C , and H3 OC rather than inert ions (ArC , NeC , HeC ). Other interpretations suggest that the Moon’s ion exosphere is caused by electron emission from dust (Stubbs et al. 2011).

The Upper Atmospheres of Jupiter, Saturn, Uranus, and Neptune Giant planet thermospheres are composed of H2 and He with some H and traces of carbon and oxygen species (Fig. 4). Methane is the dominant carbon-bearing species, and its abundance falls off rapidly with altitude above the homopause. The abundance of He also decreases above the homopause. Atomic H is mostly released by photochemistry below the thermosphere, and, being lighter than H2 , its abundance increases with altitude in the thermosphere. An external flux of water group species has been inferred for all of the giant planets (Feuchtgruber et al. 1997), and on Saturn, water “raining” down from the magnetosphere and rings affects the ionosphere (e.g., Connerney and Waite 1984; O’Donoghue et al. 2013). The abundance of water is roughly constant in the thermosphere and decreases with pressure in the lower atmosphere due to condensation (Moses and Bass 2000; Müller-Wodarg et al. 2012). The dominant ions in the main ionospheric peak are HC and H3 C .

Fig. 4 Mixing ratios in Saturn’s atmosphere illustrate the basic composition of giant planet upper atmospheres (Strobel et al. 2016). The data points (diamonds) were retrieved from a Cassini/UVIS stellar occultation

362

A. García Muñoz et al.

Observations Our understanding of giant planet upper atmospheres is mostly based on the Pioneer, Voyager, Galileo, and Cassini space missions, although observations by Earth-orbiting space telescopes and ground-based telescopes in the near-IR have also contributed. The density and temperature profiles in Jupiter’s equatorial thermosphere were measured by the Galileo probe (Seiff et al. 1998). In situ measurements of Saturn’s thermosphere are also planned for the Cassini Grand Finale in 2017 (Edgington and Spilker 2016). All other information comes from remote sensing. In particular, UV solar and stellar occultations observed by visiting spacecraft are an important tool for retrieving densities of H2 and hydrocarbons as well as temperatures in the upper atmosphere. Occultations by the Voyager Ultraviolet Spectrometer (UVS) have been analyzed on all giant planets (e.g., Yelle et al. 1993; Stevens et al. 1993; Yelle and Miller 2004; Vervack and Moses 2015), the Cassini Ultraviolet Imaging Spectrograph (UVIS) has spent a decade observing them on Saturn (Koskinen et al. 2013, 2015, 2016), and the New Horizons (NH) ALICE instrument also observed stellar occultations by Jupiter (Greathouse et al. 2010). Ultraviolet aurora and airglow, including H Lyman-’ (H Ly’), H2 electronic band, and He 584 Å emissions, are also used to study the upper atmosphere. The aurora is excited by electron precipitation along magnetic field lines connecting the polar ionosphere to the solar-wind interaction region and the magnetosphere. Voyager/UVS obtained the first unambiguous detections of the UV aurora on Jupiter, Saturn, and Uranus and found evidence of the aurora on Neptune. Subsequent observations by the HST and Cassini have been used to study the morphology and physical origin of giant planet aurora (e.g., Grodent 2015). The aurora on Jupiter and Saturn has also been detected at visual wavelengths, and X-ray emissions from the aurora and disk have been detected on Jupiter (e.g., Badman et al. 2015). The primary origin of the H Ly’ and He 584 Å airglow is resonant scattering of sunlight (e.g., Ben-Jaffel et al. 1995; Parkinson et al. 1998). The H2 band emissions are probably produced by a combination of photoelectron excitation and solar fluorescence (e.g., Liu and Dalgarno 1996), although some authors argue that additional excitation by suprathermal electrons or “electroglow” is required to explain these emissions (e.g., Herbert and Sandel 1999). Near-IR emissions from the upper atmosphere consist of 3.3 m CH4 emissions that originate near the homopause (Drossart et al. 1999) and emissions at 2–4 m from the thermosphere (Drossart et al. 1989). Observations of H3 C emissions probe the aurora, the state of the ionosphere, temperature, and dynamics (Miller et al. 2006, 2010). Emissions from the aurora and disk are observed on Jupiter and Uranus, while on Saturn, auroral emissions are observed regularly, and disk emissions appear intermittently (O’Donoghue et al. 2013). Unfortunately, there are at present no means to detect any other ions. The only other information on the ionosphere are electron densities retrieved from spacecraft radio occultations. They have been retrieved for Jupiter and Saturn from Pioneer data, for all giant planets

17 Upper Atmospheres and Ionospheres of Planets and Satellites

363

from Voyager data (Kliore et al. 1980; Lindal et al. 1985, 1987; Lindal 1992; Yelle and Miller 2004), for Jupiter from Galileo data (Hinson et al. 1997), and for Saturn from Cassini data (Kliore et al. 2009).

Thermospheres The temperature in the stratosphere and mesosphere is controlled by solar nearIR heating in CH4 bands and IR emissions by CH4 and photochemical products C2 H6 and C2 H2 (Yelle et al. 2001). As the abundance of CH4 decreases above the homopause (Fig. 4), the lack of radiative cooling allows for a hot thermosphere. Unlike on Earth, on the giant planets, the thermospheres are much hotter than expected from solar heating (Fig. 5), and the solution to this “energy crisis” remains elusive (see below). The upper atmosphere of Neptune is slightly warmer than on Saturn, although generally the temperatures on these two planets appear comparable. The temperatures on Jupiter and Uranus, on the other hand, are much higher than on Saturn and Neptune. These trends do not correlate with distance from the Sun, and, in the absence of a definite solution to the energy crisis, there is no generally accepted explanation for these differences. The location of the base of the thermosphere should coincide roughly with the homopause, i.e., the region where the abundance of CH4 begins to fall rapidly with altitude. On Jupiter, the stratospheric mixing ratio of methane is 1.8  103 , and

Fig. 5 Low-latitude temperature-pressure (T-P) profiles for Jupiter from the Galileo probe (Seiff et al. 1998) and Uranus from the Voyager 2/UVS solar occultation (Stevens et al. 1993). The Saturn T-P profile is an average of 28 low to mid-latitude Cassini stellar occultations combined with Cassini/CIRS data (Koskinen et al. 2015), with error bars reflecting the variability of the observations. The T-P profile for Neptune is based on the Voyager 2/UVS occultations (MüllerWodarg et al. 2008). The highest temperature on Jupiter expected from solar XUV heating is only 230 K

364

A. García Muñoz et al.

the homopause is near the 1 bar level, close to the base of the thermosphere (Seiff et al. 1998). The stratospheric CH4 mixing ratio of 4.8  103 on Saturn (Fletcher et al. 2009) is the highest among the giant planets. At low latitudes, the homopause on Saturn is near 0.01–0.1 bar, in rough agreement with the base of the thermosphere (Koskinen et al. 2015). In contrast to Jupiter and Saturn, temperatures on Uranus and Neptune are cold enough for methane to condense in the troposphere. On Uranus, the stratospheric mixing ratio of CH4 is only 105 –104 , and the abundance decreases further above the 0.1 mbar level (e.g., Lellouch et al. 2015), possibly explaining the relatively deep base of the thermosphere on Uranus. On Neptune, the stratospheric CH4 mixing ratio of about 103 is several times higher than allowed by the mean tropospheric cold trap. This is either because of a leakage through a warm tropopause at high southern latitudes or upwelling and convective overshooting (e.g., Lellouch et al. 2015). The homopause is near the 1 bar level, roughly in line with the base of the thermosphere (Yelle et al. 1993; Moses et al. 1995). Most of the work on the energy crisis has focused on Jupiter and Saturn where more observations are available. The commonly proposed solutions are the deposition of energy by breaking gravity and acoustic waves or resistive (Joule) heating by auroral electrodynamics followed by redistribution of energy by global circulation (e.g., Müller-Wodarg et al. 2006). There are, however, problems associated with both of these solutions. While wave heating has been proposed as a plausible mechanism on Jupiter (e.g., Young et al. 1997; Schubert et al. 2003), the calculations to date are idealized and ignore, for example, momentum deposition by waves, which would considerably decrease their energy flux (Yelle and Miller 2004). Similarly, wave heating has been found to be insufficient on Saturn. In order to be significant, wave heating would also have to be globally distributed and continuously active, which may be unlikely (Strobel et al. 2016). Magnetosphere-ionosphere coupling generates auroral electric currents that power resistive heating of the polar upper atmosphere and ionosphere. The derived resistive heating rates of about 100 TW on Jupiter and about 10 TW on Saturn are in principle sufficient to solve the energy crisis, but the heating is limited to a narrow band of latitudes near the poles (e.g., Müller-Wodarg et al. 2006). Majeed et al. (2005) used a circulation model to argue that meridional winds could transport energy to low latitudes on Jupiter and explain the equatorial temperatures. More recent modeling on Jupiter and Saturn, however, demonstrates that westward ion drag and a “Coriolis barrier” imposed by rapid rotation turn meridional winds into a strong high latitude zonal jet, thus preventing the redistribution of energy to the equator (Smith et al. 2007; Smith and Aylward 2009). In contrast, new data from Cassini show that on Saturn the poles are generally warmer than the equator (Koskinen et al. 2015), indicating that polar heating and redistribution to lower latitudes may in fact be operating. No definite mechanism to facilitate the redistribution of energy, however, has yet been identified and the possibility that some other heating mechanism operates at low to mid-latitudes cannot be ruled out.

17 Upper Atmospheres and Ionospheres of Planets and Satellites

365

Ionospheres In principle, the ionospheres of the giant planets should be simple because the atmospheres are dominated by H2 and He. According to the basic theory, solar UV radiation and electron precipitation ionize H2 , producing H2 C that rapidly reacts with H2 to form short-lived H3 C , which recombines dissociatively with electrons to release H2 and H. Ionization of H and dissociative ionization of H2 form the longlived HC , while ionization of He produces HeC , which can also react with H2 to produce small amounts of HeHC (e.g., Yelle and Miller 2004). Ionization of CH4 near the homopause leads to the production of complex, short-lived hydrocarbon ions and heavier neutral molecules, including C6 H6 , that can act as a stepping stone to ring polyaromatic hydrocarbons and eventually stratospheric haze (Kim and Fox 1994; Friedson et al. 2002; Wong et al. 2003; Kim et al. 2014; Koskinen et al. 2016). This basic theory is undoubtedly correct, and yet models have struggled to match the observed electron densities. Figure 6 compares electron density profiles retrieved from radio occultations. The results indicate strong variability in electron densities on Jupiter and Saturn that is not clearly understood (e.g., Yelle and Miller 2004; Kliore et al. 2009). Similar variability may occur on Uranus and Neptune, but observations are more limited. The observed profiles also include sharp, dense layers that can be driven by waves (Matcheva et al. 2001). Assuming that photoionization dominates at non-auroral latitudes, the electron densities should decrease with distance from the Sun. Figure 6 confirms that the overall electron density decreases

Fig. 6 Electron density profiles in giant planet ionospheres retrieved from radio occultations. The results for Jupiter are from Galileo (Hinson et al. 1997), available through the Planetary Data System. The Saturn low-latitude results are an average of 17 occultations within 30ı latitude from the equator, and the high latitude results are an average of 12 occultations at absolute latitudes higher than 40ı (Kliore et al. 2009). The Voyager ingress and egress results for Uranus and Neptune were taken from Lindal et al. (1987) and Lindal (1992), respectively

366

A. García Muñoz et al.

from Jupiter through Saturn to Neptune. The situation on Uranus appears more complicated, given the sharp high-density peaks in the lower ionosphere and high altitude electron densities that exceed equatorial densities on Saturn. According to models, which focus mostly on Jupiter and Saturn, the lower ionosphere above the hydrocarbon layer is dominated by H3 C , while HC dominates at higher altitudes. Simple models predict that the transition from H3 C to HC takes place below the main ionospheric peak and underestimate the altitude of the peak while significantly overestimating the electron densities. Several possible solutions to these problems have been proposed. On Jupiter, models that include vertical plasma drifts and vibrational excitation of H2 have been used to match the observed electron densities (e.g., Majeed et al. 1999; Yelle and Miller 2004). Plasma drifts move the ionospheric peak to higher altitudes, while reactions of vibrationally excited H2 effectively convert HC to H3 C , allowing for lower electron densities. On Saturn, models that invoke an influx of water from the magnetosphere and rings have been used to achieve conversion of HC to H3 C and an agreement with the observed electron densities (e.g., Müller-Wodarg et al. 2012). In both cases, solar photoionization is sufficient to explain the non-auroral electron densities. Impact ionization in the aurora dominates over photoionization at high latitudes. Indeed, Cassini radio occultations by Saturn point to a clear trend of increasing electron density with latitude (Kliore et al. 2009) that agrees with three-dimensional model calculations including auroral precipitation (Müller-Wodarg et al. 2012). In addition to meridional trends, there is evidence for diurnal variations, although this evidence is less clear (Kliore et al. 2009). The observed electron densities point to a surprising complexity in giant planet ionospheres that will continue to provide interesting problems for future studies. Such efforts would be greatly advanced by any observations of relative ion composition, which may in fact be obtained for the first time during the Cassini Grand Finale tour.

Titan Titan is Saturn’s largest moon and the most characteristic example of a hazy environment in our solar system. Titan can serve as a reference for hazy exoplanets (Robinson et al. 2014); thus we focus here on the mechanisms responsible for the formation of hazes in this atmosphere, as revealed by the latest observations from the Cassini-Huygens mission. Titan is far too complex to be described in detail here, and interested readers are referred to recent reviews of this atmosphere (Müller-Wodarg et al. 2014).

Photochemistry Titan’s atmosphere is dominantly composed of molecular nitrogen (N2 ) with trace amounts of methane (CH4 ) and carbon monoxide (CO) (Niemann et al. 2010; Yelle et al. 2008). Energy deposition in the upper atmosphere is driven mainly by high

17 Upper Atmospheres and Ionospheres of Planets and Satellites

367

energy insolation; photons with œ < 1000 A are responsible for the photolysis of N2 close to 1100 km, while Lyman-’ photons break up CH4 with a peak photolysis rate close to 800 km (Lavvas et al. 2011a). Titan does not have an intrinsic magnetic field, but as it orbits it is subjected to Saturn’s variable magnetosphere. Energetic particles accelerated along the magnetic field lines are deposited in Titan’s upper atmosphere and provide a secondary contribution to the N2 destruction, at altitudes close to 1200 km (see Galand et al. 2014 for more details). The primary products of N2 and CH4 photolysis provide the building blocks for the formation of larger molecules in Titan’s atmosphere. For example, methyl radicals (CH3 ) produced in the photolysis of methane can recombine to form ethane (C2 H6 ) molecules, while excited nitrogen atoms (N2 D) formed in the photolysis of nitrogen react with methane leading eventually to the production of hydrogen cyanide (HCN) molecules. These are just the first steps of photochemistry in Titan’s atmosphere, since the produced molecules are subsequently dissociated by solar radiation and the new products form other molecules. This mechanism allows for the formation of perpetually larger structures in Titan’s atmosphere and terminates with the formation of photochemical aerosols, i.e., hazes. The above “schematic” picture of atmospheric photochemistry can be separated into two different modes: the neutral and the ion contribution. Neutral chemistry driven mainly by Ly-’ and lower energy photons is active in the whole atmosphere and is responsible for the bulk of the main photochemical products observed (see review by Vuitton et al. 2014). Ion chemistry, although limited to the ionosphere, is characterized by much faster reaction rates than the neutral chemistry, while it also allows for chemical pathways that are not possible through neutral reactions (Vuitton et al. 2007, 2008). These two characteristics lead to the rapid formation of macromolecules in Titan’s thermosphere (Waite et al. 2007). The role of ion chemistry became apparent when the mass spectrometers of the Cassini orbiter discovered more than 50 positive ions in the mass range between 1 and 100 Dalton/charge (Da/q) (Vuitton et al. 2007), while the picture was further completed with the detection of multiple negative ions in the same mass range (Coates et al. 2007). Detailed chemical networks are required to identify the intricate pathways leading to the formation of these species, and state-of-the-art models are able to reproduce the composition constraints from the observations of both neutral and ion species (Vuitton et al. 2014).

Aerosol Formation Cassini observations had more surprises to reveal. Measurements at larger masses show an even more significant population of positive and negative ions (Fig.7). At the deepest altitudes probed (900 km), positive ions grow up to a few hundred Da/q (Crary et al. 2009), while negative ions masses up to 10,000 Da/q were detected (Coates et al. 2007). Such large molecules are unprecedented in planetary thermospheres and are a clear demonstration of efficient molecular growth taking

368

A. García Muñoz et al.

Fig. 7 Left: density profiles of different components of Titan’s atmosphere retrieved from CassiniHuygens observations. The thick solid black and red lines present the total density (dominated by N2 ) and temperature measured by Huygens/HASI (Fulchignoni et al. 2005). Colored lines with squares present retrievals of CH4 and other neutral components from Cassini/UVIS occultations (Koskinen et al. 2011), while the thin black line with crosses presents the total positive ion density from Cassini/INMS measurements (Vuitton et al. 2009). Right: mass spectra of positive and negative ions observed by Cassini mass spectrometers (Coates et al. 2007; Vuitton et al. 2009; Crary et al. 2009). The INMS observations (blue line) have a higher mass resolution relative to the CAPS/IBS measurements (black line, positive ions) but are limited up to 100 Da (see Lavvas et al. 2013 for more details)

place in this atmosphere. Theoretical studies for the formation of these large ions demonstrate that they are the first steps of aerosol formation (Lavvas et al. 2013). Yet, aerosol growth does not terminate in the ionosphere. The aerosol mass flux produced in the atmosphere is only a tenth of the flux observed in the lower atmosphere (Wahlund et al. 2009). Further growth of the particles’ bulk mass is only possible through neutral chemical reactions on their surface (heterogeneous chemistry). The large abundance of radicals generated from the photochemistry in the upper atmosphere, along with the extended residence time of the particles in the atmosphere due to Titan’s low gravity, outbalances the lower reaction rates of neutral relative to ion reaction rates and allows for an efficient increase of the aerosol mass flux below the ionosphere (Lavvas et al. 2011b). Thus, both ion and neutral chemistry contributions are important for the formation of aerosols in Titan’s atmosphere, the first for initiating the aerosol formation and the second for defining their final mass flux.

Energetics Heating by solar EUV radiation and energetic particles from Saturn’s magnetosphere and radiative cooling by HCN emissions are the dominant factors controlling the thermal structure of Titan’s upper atmosphere (Yelle et al. 1991). These two main processes generate an average temperature of 150 K, which is consistent with the Cassini observations (Snowden and Yelle 2014). However, the observations

17 Upper Atmospheres and Ionospheres of Planets and Satellites

369

also reveal significant altitude structure in the temperature profile of the upper atmosphere (Fulchignoni et al. 2005), as well as a strong temporal variability of the order of 60 K (Snowden et al. 2013). Theoretical studies demonstrate that this variability is not related to the temporally variable energy input from Saturn’s magnetosphere, but could be affected by the dissipation of waves in this part of the atmosphere (Snowden and Yelle 2014; Cui et al. 2014). Aerosols can interact strongly with the radiation field, therefore affecting the thermal structure of the atmosphere. This is well established for Titan’s lower atmosphere where the particles have an effective size of the order of 2–3 m and are responsible for heating the stratosphere and cooling the surface (see West et al. 2014 and references therein). Occultations of Titan’s atmosphere at UV wavelengths reveal the presence of hazes in the upper atmosphere as well (Liang et al. 2007; Koskinen et al. 2011), verifying that indeed aerosol formation starts there. However, the role of these nascent aerosol particles in the thermal structure of the upper atmosphere is still under investigation. At those altitudes aerosol particles have smaller sizes, but higher populations than in the lower atmosphere, while their optical properties are unknown. Further analyses of Cassini observations and modeling are required to decipher the optical properties of the aerosols and their role in the energy balance of the Titan’s upper atmosphere.

References Badman SV, Branduardi-Raymont G, Galand M et al (2015) Auroral processes at the giant planets: energy deposition, emission mechanisms, morphology and spectra. Space Sci Rev 187:99–179. https://doi.org/10.1007/s11214-014-0042-x Bauer SJ (1973) Physics of planetary ionospheres. Springer, Berlin Ben-Jaffel L, Prange R, Sandel BR et al (1995) New analysis of the Voyager UVS H Lyman-’ emission of Saturn. Icarus 113:91–102. https://doi.org/10.1006/icar.1995.1007 Bertaux JL, Leblanc F, Witasse O, Quemerais E, Lilensten J et al (2005) Discovery of an aurora on Mars. Nature 435:790–794. https://doi.org/10.1038/nature03603 Bhardwaj A, Thampi SV, Das TP et al (2016) On the evening time exosphere of Mars: result from MENCA aboard Mars orbiter mission. GRL 43:1862–1867. https://doi.org/10.1002/2016GL067707 Bougher SW, Cravens TE, Grebowsky J, Luhmann J (2015a) The aeronomy of Mars: characterization by MAVEN of the upper atmosphere reservoir that regulates volatile escape. Space Sci Rev 195:423–456. https://doi.org/10.1007/s11214-014-0053-7 Bougher SW, Jakosky B, Halekas J, Grebowsky J, Luhmann J (2015b) Early MAVEN deep dip campaign reveals thermosphere and ionosphere variability. Science 350:0459. https://doi.org/10.1126/science.aad0459 Catling DC, Zahnle KJ (2009) The planetary air lead. Sci Am 300:36–43. https://doi.org/10.1038/scientificamerican0509-36 Choudhary RK, Ambili KM, Choudhury S, Dhanya MB, Bhardwaj A (2016) On the origin of the ionosphere at the moon using results from Chandrayaan-1 S band radio occultation experiment and a photochemical model. Geophys Res Lett 43:10025–10033. https://doi.org/10.1002/2016GL070612 Clancy RT, Sandor BJ, García Muñoz A, Lefèvre F, Smith MD, Wolff MJ, Montmessin F, Murchie SL, Nair H (2013) First detection of Mars atmospheric hydroxyl: CRISM near-IR

370

A. García Muñoz et al.

measurement versus LMD GCM simulation of OH Meinel band emission in the Mars polar winter atmosphere. Icarus 226:272–281. https://doi.org/10.1016/j.icarus.2013.05.035 Clancy RT, Sandor BJ, Hoge J (2015) Doppler winds mapped around the lower thermospheric terminator of Venus: 2012 solar transit observations from the James Clerk Maxwell Telescope. Icarus 254:233–258. https://doi.org/10.1016/j.icarus.2015.03.031 Coates AJ, Crary FJ, Lewis GR et al (2007) Discovery of heavy negative ions in Titan’s ionosphere. GRL 34:L22103. https://doi.org/10.1029/2007GRL030978 Connerney JEP, Waite JH (1984) New model of Saturn’s ionosphere with an influx of water from the rings. Nature 312:136–138. https://doi.org/10.1038/312136a0 Crary FJ, Magee BA, Mandt K et al (2009) Heavy ions, temperatures and winds in Titan’s ionosphere: combined Cassini CAPS and INMS observations. PSS 57:1847–1856. https://doi.org/10.1016/j.pss.2009.09.006 Cravens TE, Brace LH, Taylor HA Jr, Quenon SJ, Russell CT et al (1982) Disappearing ionospheres on the nightside of Venus. Icarus 51:271–282. https://doi.org/10.1016/0019-1035(82)90083-5 Cui J, Yelle RV, Li T et al (2014) Density waves in Titan’s upper atmosphere. J Geophys R 119:490–518. https://doi.org/10.1002/2013JA019113 Dennerl K (2008) X-rays from Venus observed with Chandra. Planet Space Sci 56:1414–1423. https://doi.org/10.1016/j.pss.2008.03.008 Drossart P, Maillard J-P, Caldwell J et al (1989) Detection of H3 C on Jupiter. Nature 340:539–541. https://doi.org/10.1038/340539a0 Drossart P, Fouchet Th, Crovisier J et al (1999) Fluorescence in the 3 micron bands of methane on Jupiter and Saturn from ISO/SWS observations. In: Cox P, Kessler MF (eds) The Universe as seen by ISO. ESA special publication, vol 427. pp 169–172 Edgington SG, Spilker LJ (2016) Cassini’s grand finale. Nat Geosci 9:472–473 Feuchtgruber H, Lellouch E, de Graauw T et al (1997) External supply of oxygen to the atmospheres of the giant planets. Nature 389:159–162. https://doi.org/10.1038/38236 Fletcher LN, Orton GS, Teanby NA, Irwin PGJ, Bjoraker GL (2009) Methane and its isotopologues on Saturn from Cassini/CIRS observations. Icarus 199:351–367. https://doi.org/10.1016/j.icarus.2008.09.019 Forget F, Montmessin F, Bertaux J-L, González-Galindo F, Lebonnois S et al (2009) Density and temperatures of the upper Martian atmosphere measured by stellar occultations with Mars Express SPICAM. JGR:114, E01004. https://doi.org/10.1029/2008JE003086 Fox JL (2009) Morphology of the dayside ionosphere of Mars: implications for ion outflows. JGR 114:E12005. https://doi.org/10.1029/2009JE003432 Fox JL, Bougher SW (1991) Structure, luminosity and dynamics of the Venus thermosphere. Space Sci Rev 55:357–489. https://doi.org/10.1007/BF00177141 Fox JL, Sung KY (2001) Solar activity variations of the Venus thermosphere/ionosphere. J Geophys Res 106:21305–21335. https://doi.org/10.1029/2001JA000069 Friedson AJ, Wong A-S, Yung YL (2002) Models for polar haze formation in Jupiter’s stratosphere. Icarus 158:389–400. https://doi.org/10.1006/icar.2002.6885 Fulchignoni M, Ferri F, Angrilli F et al (2005) In situ measurements of the physical characteristics of Titan’s environment. Nature 438:785–791. https://doi.org/10.1038/nature043114 Galand M, Coates AJ, Cravens TE, Wahlund J-E (2014) Titan’s ionosphere. In: Titan. Interior, surface, atmosphere, and space environment. Cambridge Planetary Science. Cambridge University Press, Cambridge, UK ISBN 9780521199926 García Muñoz A, Mills FP, Piccioni G, Drossart P (2009) The near-infrared nitric oxide nightglow in the upper atmosphere of Venus. Proc Natl Acad Sci 106:985–988. https://doi.org/10.1073/pnas.0808091106 González-Galindo F, López-Valverde MA, Forget F, García-Comas M, Millour E, Montabone L (2015) Variability of the Martian thermosphere during eight Martian years as simulated by a ground-to-exosphere global circulation model. JGR Planet 120:2020–2035. https://doi.org/10.1002/2015JE004925

17 Upper Atmospheres and Ionospheres of Planets and Satellites

371

Greathouse TK, Gladstone GR, Moses JI et al (2010) New horizons Alice ultraviolet observations of a stellar occultation by Jupiter’s atmosphere. Icarus 208:293–305. https://doi.org/10.1016/j.icarus.2010.02.002 Grebowsky JM, Benna M, Planet JMC, Collinson GA, Mahaffy PR, Jakosky BM (2017) Unique, non-earthlike, meteoritic ion behavior in upper atmosphere of Mars. Geophys Res Lett. https://doi.org/10.1002/2017GL072635 Grodent D (2015) A brief review of ultraviolet auroral emissions on giant planets. Space Sci Rev 187:23–50. https://doi.org/10.1007/s11214-014-0052-8 Herbert F, Sandel BR (1999) Ultraviolet observations of Uranus and Neptune. Planet Space Sci 47:1119–1139. https://doi.org/10.1016/S0032-0633(98)00142-1 Hinson DP, Flasar FM, Kliore AJ, Schinder PJ, Twicken JD, Herrera RG (1997) Jupiter’s ionosphere: results from the first Galileo radio occultation experiment. Geophys Res Lett 24:2107–2110. https://doi.org/10.1029/97GL01608 Huestis DL, Slanger TG, Sharpee BD, Fox JL (2010) Chemical origins of the Mars ultraviolet dayglow. Faraday Discuss 147:307–322. https://doi.org/10.1039/C003456H Hunten DM (1993) Atmospheric evolution of the terrestrial planets. Science 259:915–920. https://doi.org/10.1126/science.259.5097.915 Jakosky BM, Grebowsky JM, Luhmann JG, Connerney J, Eparvier F et al (2015) MAVEN observations of the response of Mars to an interplanetary coronal mass ejection. Science 350:0210. https://doi.org/10.1126/science.aad0210 Jakosky BM, Slipski M, Benna M, Mahaffy P, Elrod M, Yelle R, Stone S, Alsaeed N (2017) Mars’ atmospheric history derived from upper-atmosphere measurements of 38 Ar/36 Ar. Science 355:1408–1410. https://doi.org/10.1126/science.aai7721 Keating GM, Bertaux JL, Bougher SW, Dickinson RE, Cravens TE, Hedin AE (1985) Models of Venus neutral upper atmosphere – structure and composition. Adv Space Res 5:117–171. https://doi.org/10.1016/0273-1177(85)90200-5 Kim YH, Fox JL (1994) The chemistry of hydrocarbon ions in the Jovian ionosphere. Icarus 112:310–325. https://doi.org/10.1006/icar.1994.1186 Kim YH, Fox JL, Black JH, Moses JI (2014) Hydrocarbon ions in the lower ionosphere of Saturn. J Geophys Res 119:384–395. https://doi.org/10.1002/2013JA019022 Kliore AJ, Patel IR, Lindal GF et al (1980) Structure of the ionosphere and atmosphere of Saturn from Pioneer 11 Saturn radio occultation. J Geophys Res 85:5857–5870. https://doi.org/10.1029/JA085iA11p05857 Kliore AJ, Nagy AF, Marouf EA et al (2009) Midlatitude and high-latitude electron density profiles in the ionosphere of Saturn obtained by Cassini radio occultation observations. J Geophys Res 114:A04315. https://doi.org/10.1029/2008JA013900 Koskinen T, Yelle RV, Snowden D et al (2011) The mesosphere and thermosphere of Titan revealed by Cassini/UVIS stellar occultations. Icarus 216:507–534. https://doi.org/10.1016/j.icarus.2011.09.022 Koskinen TT, Sandel BR, Yelle RV et al (2013) The density and temperature structure near the exobase of Saturn from Cassini UVIS solar occultations. Icarus 226:1318–1330. https://doi.org/10.1016/j.icarus.2013.07.037 Koskinen TT, Sandel BR, Yelle RV et al (2015) Saturn’s variable thermosphere from Cassini/UVIS occultations. Icarus 260:174–189. https://doi.org/10.1016/j.icarus.2015.07.008 Koskinen TT, Moses JI, West RA, Guerlet S, Jouchoux A (2016) The detection of benzene in Saturn’s upper atmosphere. Geophys Res Lett 43:7895–7901. https://doi.org/10.1002/2016GL070000 Krasnopolsky VA (2002) Mars’ upper atmosphere and ionosphere at low, medium, and high solar activities: implications for evolution of water. J Geophys Res (Planets) 107:5128. https://doi.org/10.1029/2001JE001809 Lammer H, Kasting JF, Chassefière E, Johnson RE, Kulikov YN, Tiang F (2008) Atmospheric escape and evolution of terrestrial planets and satellites. Space Sci Rev 139(1–4):399–436. https://doi.org/10.1007/s11214-008-9413-5

372

A. García Muñoz et al.

Lavvas P, Galand M, Yelle RV et al (2011a) Energy deposition and primary chemical products in Titan’s upper atmosphere. Icarus 213:233–251. https://doi.org/10.1016/j.icarus. 2011.03.001 Lavvas P, Sander M, Kraft M et al (2011b) Surface chemistry and particle shape: processes for the evolution of aerosols in Titan’s atmosphere. Astrophys J 728:80. https://doi.org/10.1088/0004-637X/728/2/80. (11p) Lavvas P, Yelle RV, Koskinen T et al (2013) Aerosol growth in Titan’s ionosphere. PNAS 110:2729–2734. https://doi.org/10.1073/pnas.1217059110 Lellouch E, Moreno R, Orton GS et al (2015) New constraints on the CH4 vertical profile in Uranus and Neptune from Herschel observations. A&A 579:A121. https://doi.org/10.1051/0004-6361/201526518 Liang MC, Yung YL, Shemansky DE (2007) Photolytically generated aerosols in the mesosphere and thermosphere of Titan. Astrophys J 661:199–202. https://doi.org/10.1086/518785 Lindal GF (1992) The atmosphere of Neptune: an analysis of radio occultation data acquired with Voyager 2. AJ 103:967–982. https://doi.org/10.1086/116119 Lindal GF, Sweetnam DN, Eshleman VR (1985) The atmosphere of Saturn: an analysis of the Voyager radio occultation measurements. ApJ 90:1136–1146. https://doi.org/10.1086/113820 Lindal GF, Lyons JR, Sweetnam DN, Eshleman VR, Hinson DP, Tyler GL (1987) The atmosphere of Uranus: results of radio occultation measurements with Voyager 2. J Geophys Res 92:14987– 15001. https://doi.org/10.1029/JA092iA13p14987 Liu W, Dalgarno A (1996) The ultraviolet spectra of the Jovian dayglow. ApJ 462:502–518. https://doi.org/10.1086/177168 Mahieux A, Vandaele AC, Robert S, Wilquet V, Drummond R et al (2015) Rotational temperatures of Venus upper atmosphere as measured by SOIR on board Venus Express. Planet Space Sci 113:347–358. https://doi.org/10.1016/j.pss.2014.12.020 Majeed T, McConnell JC, Gladstone GR (1999) A model analysis of Galileo electron densities on Jupiter. Geophys Res Lett 26:2335–2338. https://doi.org/10.1029/1999GL900530 Majeed T, Waite JH Jr, Bougher SW et al (2005) Process of equatorial thermal structure at Jupiter: an analysis of the Galileo temperature profile with a three-dimensional model. J Geophys Res 110:E12007. https://doi.org/10.1029/2004JE002351 Mangold N, Baratoux D, Witasse O, Encrenaz T, Sotin C (2016) Mars: a small terrestrial planet. Astron Astrophys Rev 24:15. https://doi.org/10.1007/s00159-016-0099-5 Matcheva KI, Strobel DF, Flasar FM (2001) Interaction of gravity waves with ionospheric plasma: implications for Jupiter’s ionosphere. Icarus 152:347–365. https://doi.org/10.1006/icar.2001.6631 Meier RR (1991) Ultraviolet spectroscopy and remote sensing of the upper atmosphere. Space Sci Rev 58:1–185. https://doi.org/10.1007/BF01206000 Mendillo M, Nagy A, Waite JH (eds) (2002) Atmospheres in the solar system. American Geophysical Union, Geophysical Monograph, Washington, DC, p 130 Miller KL, Knudsen WC, Spenner K, Whitten RC, Novak V (1980) Solar zenith angle dependence of ionospheric ion and electron temperatures and density on Venus. J Geophys Res Space Phys 85:7759–7764. https://doi.org/10.1029/JA085iA13p07759 Miller S, Stallard T, Smith C et al (2006) H3 C : the driver of giant planet atmospheres. Phil Trans R Soc A 364:3121–3137. https://doi.org/10.1098/rsta.2006.1877 Miller S, Stallard T, Melin H, Tennyson J (2010) H3 C cooling in planetary atmospheres. Faraday Discuss 147:283–291. https://doi.org/10.1039/c004152c Moses JI, Bass SF (2000) The effects of external material on the chemistry and structure of Saturn’s ionosphere. J Geophys Res 105:7013–7052. https://doi.org/10.1029/1999JE001172 Moses JI, Rages K, Pollack JB (1995) An analysis of Neptune’s stratospheric haze using high phase-angle Voyager images. Icarus 113:232–266. https://doi.org/10.1006/icar.1995.1022 Mueller-Wodarg ICF, Strobel DF, Moses JI, Waie JH, Crovisier J, Yelle RV, Bougher SW, Roble RG (2008) Neutral atmospheres. Space Sci Rev 139:191–234. https://doi.org/10.1007/s11214-008-9404-6 Müller-Wodarg ICF, Mendillo M, Yelle RV et al (2006) A global circulation model of Saturn’s thermosphere. Icarus 180:147–160. https://doi.org/10.1016/j.icarus.2005.09.002

17 Upper Atmospheres and Ionospheres of Planets and Satellites

373

Müller-Wodarg ICF, Moore L, Galand M, Miller S, Mendillo M (2012) Magnetosphere-atmosphere coupling at Saturn: 1 – response of thermosphere and ionosphere to steady state polar forcing. Icarus 221:481–494. https://doi.org/10.1016/j.icarus.2012.08.034 Müller-Wodarg I, Griffith CA, Lellouch E, Cravens TE (2014) Titan. Interior, surface, atmosphere, and space environment. Cambridge Planetary Science. Cambridge University Press, Cambridge, UK ISBN 9780521199926 Müller-Wodarg ICF, Bruinsma S, Marty J-C, Svedhem H (2016) In situ observations of waves in Venus’s polar lower thermosphere with Venus Express aerobraking. Nat Phys 12:767–771. https://doi.org/10.1038/nphys3733 Nagy AF, Balogh A, Cravens TE, Mendillo M, Mueller-Wodarg I (eds) (2008) Comparative aeronomy. Springer, New York Niemann HB, Atreya SK, Demick JE et al (2010) Composition of Titan’s lower atmosphere and simple surface volatiles as measured by the Cassini-Huygens probe gas chromatograph mass spectrometer experiment. J Geophys Res 115(E):12006. https://doi.org/10.1029/2010JE003659 O’Donoghue J, Stallard TS, Melin H et al (2013) The domination of Saturn’s low-latitude ionosphere by ring rain. Nature 496:193–195. https://doi.org/10.1038/nature12049 Parkinson CD, Griffioen E, McConnell JC, Gladstone GR, Sandel BR (1998) He 584 Å dayglow at Saturn: a reassessment. Icarus 133:210–220. https://doi.org/10.1006/icar.1998.5926 Pätzold A, Häusler B, Bird MK, Tellmann S, Mattei R, Asmar SW, Dehant V, Eidel W, Imamura T, Simpson RA, Tyler GL (2007) The structure of Venus’ middle atmosphere and ionosphere. Nature 450:657–660. https://doi.org/10.1038/nature06239 Rees MH (1989) Physics and chemistry of the upper atmosphere. Cambridge University Press, London Robinson TD, Maltagliati L, Marley M et al (2014) Titan solar occultation observations reveal transit spectra of a hazy world. PNAS 111:9042–9047. https://doi.org/10.1073/pnas.1403473111 Sánchez-Lavega A, García Muñoz A, García-Melendo E, Pérez-Hoyos S, Gómez-Forrelad et al (2015) An extremely high-altitude plume seen at Mars’ morning terminator. Nature 518:525– 528. https://doi.org/10.1038/nature14162 Schneider NM, Deighan JI, Jain SK, Stiepen A, Stewart AIF et al (2015) Discovery of diffuse aurora on Mars. Science 350:0313. https://doi.org/10.1126/science.aad0313 Schubert GMP, Hickey MP, Walterscheid RL (2003) Heating of Jupiter’s thermosphere by the dissipation of upward propagating acoustic waves. Icarus 163:398–413. https://doi.org/10.1016/S0019-1035(03)00078-2 Schunk RW, Nagy AF (2000) Ionospheres. Physics, plasma physics, and chemistry. Cambridge University Press, Cambridge Seiff A, Kirk DB, Knight TCD et al (1998) Thermal structure of Jupiter’s atmosphere in the North Equatorial Belt, near the edge of a 5 m hot spot. J Geophys Res 103:22857–22890. https://doi.org/10.1029/98JE01766 Slanger TG, Cosby PC, Huestis DL, Bida TA (2001) Discovery of the atomic oxygen green line in the Venus night airglow. Science 291:463–465. https://doi.org/10.1126/science.291.5503.463 Smith CGA, Aylward AD (2009) Coupled rotational dynamics of Jupiter’s thermosphere and ionosphere. Ann Geophys 27:199–230. https://doi.org/10.5194/angeo-27-199-2009 Smith CGA, Aylward AD, Millward GH et al (2007) An unexpected cooling effect in Saturn’s upper atmosphere. Nature 445:399–401. https://doi.org/10.1038/nature05518 Snowden D, Yelle RV (2014) The thermal structure of Titan’s upper atmosphere, II: energetics. Icarus 228:64–77. https://doi.org/10.1016/j.icarus.2013.08.027 Snowden D, Yelle RV, Cui J et al (2013) The thermal structure of Titan’s upper atmosphere, I: temperature profiles from Cassini INMS observations. Icarus 226:552–582. https://doi.org/10.1016/j.icarus.2013.06.006 Stevens MH, Strobel DF, Herbert F (1993) An analysis of the Voyager 2 ultraviolet spectrometer occultation data at Uranus: inferring heat sources and model atmospheres. Icarus 100:45–63. https://doi.org/10.1006/icar.1993.1005 Strobel DF, Koskinen T, Müller-Wodarg ICF (2016) Saturn’s variable thermosphere. Astro-ph arXiv:1610.07669v1

374

A. García Muñoz et al.

Stubbs TJ, Glenar DA, Farrell WM, Vondrak RR, Collier MR et al (2011) On the role of dust in the lunar ionosphere. Planet Space Sci 59:1659–1664. https://doi.org/10.1016/j.pss.2011.05.011 Vervack RJ Jr, Moses JI (2015) Saturn’s upper atmosphere during the Voyager era: reanalysis and modeling of the UVS occultations. Icarus 258:135–163. https://doi.org/10.1016/j.icarus.2015.06.007 Vogt MF, Withers P, Mahaffy PR, Benna M, Elrod MK et al (2015) Ionopause-like density gradients in the Martian ionosphere: a first look with MAVEN. Geophys Res Lett 42:8885– 8893. https://doi.org/10.1002/2015GL065269 Volkov AN, Johnson RE, Tucker OJ, Erwin JT (2011) Thermally driven atmospheric escape: transition from hydrodynamic to Jeans escape. Astrophys J 729:L24. https://doi.org/10.1088/2041-8205/729/2/L24 Vuitton V, Yelle RV, McEwan MJ (2007) Ion chemistry and N-containing molecules in Titan’s upper atmosphere. Icarus 191:722–742. https://doi.org/10.1016/j.icarus.2007.06.023 Vuitton V, Yelle RV, Cui J (2008) Formation and distribution of benzene on Titan. J Geophys Res 113(E):05007. https://doi.org/10.1029/2007JE002997 Vuitton V, Lavvas P, Yelle RV et al (2009) Negative ion chemistry in Titan’s upper atmosphere. PSS 57:1558–1572. https://doi.org/10.1016/j.pss.2009.04.004 Vuitton V, Dutuit O, Smith MA et al (2014) Chemistry of Titan’s atmosphere in Titan. Interior, surface, atmosphere, and space environment. Cambridge Planetary Science. Cambridge University Press, Cambridge, UK ISBN 9780521199926 Wahlund J-E, Galand M, Muller-Wodard I et al (2009) On the amount of heavy molecular ions in Titan’s ionosphere. PSS 57:1857–1865. https://doi.org/10.1016/j.pss.2009.07.014 Waite JH Jr, Young DT, Cravens TE et al (2007) The process of Tholin formation in Titan’s upper atmosphere. Science 316:870–875. https://doi.org/10.1126/science.1139727 West RA, Lavvas P, Anderson C et al (2014) Titan’s Haze. In: Titan. Interior, surface, atmosphere, and space environment. Cambridge Planetary Science. Cambridge University Press, Cambridge, UK ISBN 9780521199926 Withers P (2012) How do meteoroids affect Venus’s and Mars’s ionospheres? EOS Trans Am Geophys Union 93:337–338. https://doi.org/10.1029/2012EO350002 Withers P, Pratt R (2013) An observational study of the response of the upper atmosphere of Mars to lower atmospheric dust storms. Icarus 225:378–389. https://doi.org/10.1016/j.icarus.2013.02.032 Withers P, Fillingim MO, Lillis RJ, Häusler B, Hinson DP et al (2012) Observations of the nightside ionosphere of Mars by the Mars Express radio Science experiment (MaRS). J Geophys Res Space Phys 117:A12307. https://doi.org/10.1029/2012JA018185 Wong A-S, Yung YL, Friedson AJ (2003) Benzene and haze formation in the polar atmosphere of Jupiter. Geophys Res Lett 30:1447. https://doi.org/10.1029/2002GL016661 Yelle RV (1991) Non-LTE models of Titan’s upper atmosphere. Astrophys J 383:380–400. https://doi.org/10.1086/170796 Yelle RV, Miller S (2004) In: Bagenal F, Dowling TE, McKinnon WB (eds) Jupiter’s thermosphere and ionosphere. Cambridge University Press, Cambridge, pp 185–218 Yelle RV, Herbert F, Sandel BR, Vervack RJ Jr, Wentzel TM (1993) The distribution of hydrocarbons in Neptune’s upper atmosphere. Icarus 104:38–59. https://doi.org/10.1006/icar.1993.1081 Yelle RV, Griffith CA, Young LA (2001) Structure of Jovian stratosphere at the Galileo probe entry site. Icarus 152:331–346. https://doi.org/10.1006/icar.2001.6640 Yelle RV, Cui J, Muller-Wodarg ICF (2008) Methane escape from Titan’s atmosphere. J Geophys Res 113(E):10003. https://doi.org/10.1029/2007JE003031 Young LA, Yelle RV, Young RE et al (1997) Gravity waves in Jupiter’s thermosphere. Science 276:108–111. https://doi.org/10.1126/science.276.5309.108 Zhang TL, Lu QM, Baumjohann W et al (2012) Magnetic reconnection in the near Venusian magnetotail. Science 336:567–570. https://doi.org/10.1126/science.1217013 Zurbuchen TH, Raines JM, Slavin JA, Gershman DJ, Gilbert JA et al (2011) MESSENGER observations of the spatial distribution of planetary ions near Mercury. Science 333:1862–1865. https://doi.org/10.1126/science.1211302

Rings in the Solar System: A Short Review

18

Sébastien Charnoz, Aurélien Crida, and Ryuki Hyodo

Contents Overview of Solar System Ring Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jupiter Rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Saturn’s Rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Uranus’ Rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Neptune Rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Key Process in Rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Viscous Evolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ring-Satellite Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . External Processes: Ballistic Transport and Various Drags . . . . . . . . . . . . . . . . . . . . . . . . . Rings as Parent Bodies of Moons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Possible Origins of Planetary Rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rings Around Small Bodies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Observational Facts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Some Ideas on Their Origin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

377 378 379 383 383 385 386 387 388 389 389 390 391 392 393 393

S. Charnoz () Université Paris Diderot/Institut de Physique du Globe, Paris, France e-mail: [email protected] A. Crida Université Côte d’Azur/Observatoire de la Côte d’Azur, Lagrange, Nice, France Institut Universitaire de France, Paris, France e-mail: [email protected] R. Hyodo Earth-Life Science Institute/Tokyo Institute of Technology, Tokyo, Japan e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_54

375

376

S. Charnoz et al.

Abstract

Rings are ubiquitous around giant planets in our Solar System. They evolve jointly with the nearby satellite system. They could form either during the giant planet formation process or much later as a result of large-scale dynamical instabilities either in the local satellite system or at the planetary scale. We review here the main characteristics of rings in our Solar System and discuss their main evolution processes and possible origin. We also discuss the recent discovery of rings around small bodies. All giant planets of the Solar System have rings. The brightest and most massive, those of Saturn, have been discovered by Galileo Galilei himself (whereas he did not interpreted them as rings, the first one to interpret them correctly was the Dutch astronomer Christian Huygens), but it is really with the space missions Voyagers 1 and 2, during the 1970s and 1980s that it was realized that all four giant planets harbor rings. Conversely, rings seem absent around terrestrial planets, despite many attempts to find dusty rings (especially around Mars). Rings are still some very mysterious structures that are ubiquitous to all giant planets, either gas or ice giants. However, the four ring systems that we know are all very different, and within a ring system, rings could be either dense and made of large particles (like Saturn and Uranus’ rings) or dusty (like Jupiter or Saturn’s E or G rings). See Fig. 1 for a comparative sketch of the four ring systems found in our Solar System. Unlike a very common thought, rings are not necessarily confined to a planet’s Roche limit (see section “Key Process in Rings”), and dusty rings are found commonly beyond the Roche limit. The generality and the diversity of ring systems make their origin difficult to understand. They could form either concomitant with their host planets (inside the circumplanetary disk), or they could be the result of late dynamical evolution either of their satellite systems (through collisions) or results from comet showers due to large-scale instabilities of the planetary system. Thanks to the Cassini and Galileo missions, Saturn and Jupiter rings were investigated in great details (see, e.g., Cuzzi et al. 2010 for a review of Saturn’s rings as seen by Cassini). Today, due to the increased detection capacities, Uranus and Neptune rings are better understood thanks to observations from Earth (see, e.g., Showalter et al. Science, 311, 973–977, 2008). Rings are rapidly evolving structures under the action of gravity, viscosity, and radiative forces. We understand now that there is a close association between rings and satellites, through dynamical interactions (moons sculpt rings) or material exchange (rings can give birth to moons, see, e.g., Charnoz et al. 2011; Crida and Charnoz IAU#310, 182–189 2012). As they represent a huge surface-to-volume ratio, rings are also very efficient to capture meteoritic bombardment that, over long timescales, may modify their average composition. We have no clear idea of the ring’s age. They could be as old as the Solar System itself (like in the formation scenario proposed by Canup (2010) invoking a migrating satellite that is tidally disrupted) or as young as a few 100 Myrs according to ´ et al. 2016). But no model is consistent with all data, different theories (like in Cuk and the planetary ring origin is still heavily debated. Finally, unexpectedly, rings

18 Rings in the Solar System: A Short Review

377

Fig. 1 Jupiter, Saturn, Uranus, and Neptune rings and inner moons scaled by the planet radius. The shaded regions designate the dusty rings

were found recently about small bodies, like centaurs, making a general theory of their formation even more difficult to build. In this chapter, we will briefly review our current knowledge about rings in our Solar System, emphasizing their diversity, as well as the key fundamental processes that govern their evolution. We also discuss the recent ring detection about small bodies.

Overview of Solar System Ring Systems Main ring systems are those that are found inside the planet’s Roche limit (see section “Key Process in Rings”), i.e., typically inside 2.5 planetary radii. Of course, the definition of the Roche limit depends on the ring material density that

378

S. Charnoz et al.

could be diverse and changing and that is often not very well constrained. However, the Roche limit only scales with the density to the exponent 1/3, so it does not vary greatly from one system to another and is found in general around 2–3 planetary radii. Outside the Roche limit, dusty ring systems are also found, but they need to be replenished because the material is removed by the Poynting-Robertson force. In addition, the collisional timescale of dusty ring system is so long, that they are never flat and are in general vertically extended either due to radiative forces or simply by keeping the memory of the orbit of the moon that is the dust source. We now briefly present the main characteristics of the ring systems around the four giant planets.

Jupiter Rings Jupiter rings are extremely tenuous and are mostly dusty (Fig. 2). They are closely associated with the four small moons, Methis, Adrastea, Amalthea, and Thebe, that

Standard image sensitivity

10 times sensitivity

Main Ring Radius 1.81 Rj

20 times sensitivity

Amalthea Gossamer Ring 2.55 Rj

390 times sensitivity

260 times sensitivity

Thebe Gossamer Ring 3.15 Rj

Jupiter

Metis 1.79 Rj

Adrastea 1.81 Rj

Amalthea 2.54 Rj

Thebe 3.11 Rj

Gossamer Rings Halo Main Ring = Earth for scale Amalthea

Adrastea Metis

Thebe

Fig. 2 Jupiter’s ring structure. RjDJupiter’ radius

18 Rings in the Solar System: A Short Review

379

release dust from their orbit. This dust adopts a 3D structure with finite thickness due to the randomization of their longitude of nodes and pericenters. The two outermost rings (the “Amalthea” and the “Thebe” gossamer rings) extend up to about 2.5 and 3.15 Jovian radii, respectively, and have an optical depth about 107 and 108 . They are vertically extended (unlike Saturn’s main rings) with thickness about 2300 and 8400 km, respectively, that reflects the inclination of the two parent moons. The volume density of these two rings is not maximum in the midplane, but rather at the top and the bottom, because of the simple geometric effect: a particle on an inclined orbit (like all dust grains in these rings) spends more time at the top and bottom of its orbit because the vertical velocity there is zero. This creates a “sandwich”like structure whose origin is purely kinematic. Closer to the planet, the main rings, bounded by Adrastea, are much flatter, about 30 km, and are composed mainly of big particles as testified by a weak transmission in forward-scattered light. A vertical extension appears, up a to a few 100 km, that may be composed of micrometer-sized dust. Then comes the innermost ring, called the “halo” that is very puffed up (up to a few 10,000 km) and is dusty. The vertical extension of the halo is maybe due to the coupling of the charged dusty grains with the magnetic field of Jupiter that excites the particle inclinations (Hamilton and Krüger 2008). Dust circulates efficiently in this ring system due to the Poynting-Robertson force (Burns 1979, see also section “Key Process in Rings”), forcing the dust grains to spiral toward the central planet. Particle sizes in Jupiter’s ring system show a steep size distribution and are typically in the range 0.1–30 m. Only the main ring seems to contain a significant population of cm-sized particles. The structure of the Jovian rings is currently the best understood four ring systems. It shows well that big particles that are not subject to radiative forces and lose energy through collisions gather in a thin sheet (the main rings), whereas dusty particles, subject to radiative forces and that are in a low collisional environment (owing to the lower optical depth in the dusty rings), in general do not flatten into a thin ring system but keep a torus-like structure in 3D. For a detailed review of Jupiter rings and physics, we recommend the Burns et al.’s (1984) review paper.

Saturn’s Rings They are the most famous, the richest, and the most massive ring system of the four giant planets. Very broadly speaking, they are composed of a dense and vertically thin ring system inside the Saturn’s Roche limit (about 2.5 Saturn’s radii), as well as several of dusty rings, sometimes associated with satellites, beyond the Roche limit. Hundreds of dynamical structures have been identified in Saturn’s main ring system, with a number of them clearly associated with ringmoons’ gravitational interactions, but the origin of numerous structures is not clearly associated with moons. The study of Saturn’s rings is a very dynamical subject, fueled by the Cassini mission, and several review papers can be found in the Saturn from Cassini-Huygens book (2009, M. Dougherty, L. Esposito, S. Krimigis Eds. Springer) and the forthcoming Planetary Rings book (C.D. Murray, M. Tiscareno

380

S. Charnoz et al.

Eds, Univ. of Arizona Press) that should be edited in 2018. For a review of the most surprising structures, we recommend the Cuzzi et al.’s (2010) paper or the (excellent) Tiscareno’s (2013) review paper about rings in the Solar System. The main ring systems consist of four regions with different optical depths and different types of structures (Fig. 3). By increasing order of distance to Saturn, we find first the D and C rings, deeply embedded in Saturn’s Roche limit, that consist of ringlets and plateaus of material with varying optical depths about 104 and 103 , respectively, and then comes the B ring, the densest ring, and probably the one that contains most of Saturn’s rings’ mass, with optical depth much higher than 1. Outside the B ring, we find the Cassini division, a low optical depth regions rich in ringlets and dusty plateaus, and then finally, we find the translucent A ring, with optical depth about 1 and that extends up to about 136,000 km. A thin and very dynamical ringlet, the F ring, lays precisely at Saturn’s Roche limit (about 140,000 km). It is bounded inside and outside by two moons, Prometheus and Pandora, whose complex dynamical interactions trigger a wide variety of structures in the F ring. Prometheus and Pandora are called “Shepherd” moons, because they were thought, once, to gravitationally confine the F ring. Transient accretion structures (like clumps) have been identified but seem very ephemeral (Beurle et al. 2010). Whereas no moons have been found in the D, C, and B and the Cassini division, the A ring contains several moonlets that create gap and density waves (Pan, Daphnis, Atlas), but more distant moons also trigger waves (Prometheus, Pandora, Janus, Epimetheus, etc.). Making a list of all dynamical structures found in Saturn’s main rings is out of the scope of the present chapter, but among the most notable structures, we find numerous spiral density waves in the A ring and some in the B ring (see, e.g., Tiscareno et al. 2007). The outer edge of Saturn’s B ring shows complex vertical structures and seems to be maintained by the exchange of angular momentum with the Mimas moon, which is in 2:1 resonance with the B ring outer edge (Cuzzi et al. 2010; Spitale and Porco 2010). Inclined moons create also vertical waves in the main rings (called “bending waves”) visible in the A ring. The main ring’s thickness has been estimated using different techniques (numerical simulations, study of heat conductions) and is estimated about 10 m (see, e.g., Reffet et al. 2015). Particles in the main rings are composed of water ice, with a slight contamination (less than 1% in mass) by some unknown red material that could be silicate or iron (Cuzzi et al. 2009). This peculiar composition (>99% of water ice) is very different from the average composition of outer Solar System objects. It is thought to be a very constraining clue to their origin (see, e.g., Charnoz et al. 2009; Canup 2010), as well as a challenge to be explained. The mass of Saturn’s main ring is a matter of debate. Study of density waves in the A ring gives an idea of the local surface density. Extrapolation to the entire ring system gives a total mass about 1019 –1020 kg, consistent with other independent estimates (see, e.g., Esposito 2010; Reffet et al. 2015). Interestingly, with this mass, and assuming a thickness of a few meters, this implies that Saturn’s ring is a self-gravitating disk, i.e., it may develop spiral structure patterns. Indeed, spiral patterns have been found, but at the km scales called “wakes” which is explained by the strong Keplerian shear in

Fig. 3 Top: Saturn’s main ring optical depth as a function of distance. Bottom: Saturn main rings seen in visible wavelength. Cassini ISS/JPL/NASA. (Figure taken from Colwell et al. 2009)

18 Rings in the Solar System: A Short Review 381

382

S. Charnoz et al.

Saturn’s rings, see Schmidt et al. (2009) and Cuzzi et al. (2010). This may have important implication for the ring long-term evolution, as it is known (see, e.g., Salmon et al. 2010) that self-gravitating astrophysical disks are self-regulated by a balance between their surface density and velocity dispersion and maintain their surface density close to a Toomre coefficient close to 1. Star occultations allowed to derive size distribution in different regions of the ring system. Saturn’s main rings do not contain dust, and particle sizes are in the range 1 cm to a few 10 m. The size distribution shows variations depending on the region in the rings, with the largest size in general increasing with distance to Saturn that may be a sign of more efficient accretion at large distance (Cuzzi et al. 2009). Dusts (i.e., micrometer-sized grains) are in general absent from the main rings, except in a handful of very narrow ringlet like in the Encke gap or in the Laplace gap. Beyond the main rings, dusty rings are also found and have been detected in forward-scattered light thanks to Voyager and Cassini missions. The G ring presents long arc structures (i.e., incomplete rings) maybe due to the erosion of an unseen population of km-sized moonlets but still undetected. Saturn’s E ring is a dusty ring and results from the ejection of icy dust in Enceladus’ geysers. Small moons have dusty rings associated (Methone, Pallene, Anthe) that may result from their erosion. More recently was discovered a very distant and huge ring, the Phoebe ring, that has a torus-like structure and is composed of dust thought to be lost from the Phoebe satellite that is on an inclined retrograde orbit and that spirals slowly inward the Saturn’s system due to the Poynting-Robertson drag (Verbiscer et al. 2009). The Saturn’s system shows that there is a complex interplay between moons and rings, and studies suggest that there is a kind of cycle of material in ring-satellite systems: rings may give birth to satellites at the Roche limit, and satellite may release dusty material that moves inward due to radiation forces (see, e.g., Charnoz et al. 2011; Crida and Charnoz 2012). Is there a full recycling of material? For the moment, this is an open question (Esposito 2010). Another intensive discussion about Saturn’s rings is their age. If the rings are as old as the age of the Solar System, rings would have been bombarded by micro-meteoroid, polluting ring materials through time. However, current rings look cleaner than it is expected assuming a specific pollution rate (Zhang et al. 2017), and thus rings might be young (100 million years old). However, the flux of the micro-meteoroid bombardment through the evolution of the Solar System is not well constrained, and question remains how to dynamically form rings of such young (see section “Possible Origins of Planetary Rings”). In addition, the viscous spreading of rings of arbitrary initial mass leads in 4.5 Gyrs to rings of about half the mass of Mimas, which is the mass observed today (see below). This mass coincidence is a strong argument in favor of old rings; in this frame, the pollution could have been diluted in the initially massive rings and stored, for instance, in satellites (Charnoz et al. 2011). So, the age of Saturn’s rings is still an open question. Further studies and detailed data analysis of the Cassini spacecraft are required to understand more about age of Saturn’s rings.

18 Rings in the Solar System: A Short Review

383

Uranus’ Rings Uranus’ rings are very different from those of Jupiter and Saturn. They consist mainly of a discrete collection of about ten ringlets and two dust bands, with low albedo (0.35 Msol and M M9.5 nir M9.5 nir >M9.5 nir >M9.5 nir M9.5 nir >M9.5 nir >M9.5 nir >M9.5 nir M9.5 nir >M9.5 nir >M9.5 nir >M9.5 nir L1 nir

Lucas et al. (2001) Lucas et al. (2001) Lucas et al. (2001) Weights et al. (2005) Weights et al. (2005) Weights et al. (2005) Weights et al. (2005) Weights et al. (2005) Weights et al. (2005) Weights et al. (2005) Weights et al. (2005) Weights et al. (2005) Weights et al. (2005) Weights et al. (2005) Weights et al. (2005) Weights et al. (2005) Weights et al. (2005) Weights et al. (2005) Weights et al. (2005) Weights et al. (2005) Lodieu et al. (2008)

L1 nir

Lodieu et al. (2008)

L1 nir

Lodieu et al. (2008)

L1 nir

Lodieu et al. (2008)

L1 nir

Lodieu et al. (2008)

L1 nir

Lodieu et al. (2008)

L1 nir

Lodieu et al. (2008)

L2 nir

Lodieu et al. (2008) (continued)

22 Brown Dwarfs and Free-Floating Planets in Young Stellar Clusters

487

Table 3 (continued) Cluster or association

¡ Ophiuchus

Name USco J155150.2-213457 15593638-2214159 16130231-2124285 16142256-2331178 16114437-2215446 16091868-2229239 16073161-2146544 16041304-2241034 16140756-2211522 16053909-2403328 16042042-2134530 16013692-2212027 16151270-2229492 4450

CFHTWIR-Oph 9 CFHTWIR-Oph 18 CFHTWIR-Oph 90 CFHTWIR-Oph 100 CFHTWIR-Oph 103 CFHTWIR-Oph 33 Chamaleon Cha I J11070768-7626326 Lupus 3 SONYC-ChaI-1 Taurus 2MASS J04373705 C 2,331,080 Pleiades Calar 25

Spectral type (opt or nir) L4-L6 nir

References Peña-Ramírez et al. (2016)

L1 nir L1.5 nir L2 nir L5 nir L3 nir L4.5 nir L6 nir L4.5 nir L4.5 nir L6 nir L7 nir L7 nir T nir

Lodieu et al. (2017) Lodieu et al. (2017) Lodieu et al. (2017) Lodieu et al. (2017) Lodieu et al. (2017) Lodieu et al. (2017) Lodieu et al. (2017) Lodieu et al. (2017) Lodieu et al. (2017) Lodieu et al. (2017) Lodieu et al. (2017) Lodieu et al. (2017) Marsh et al. (2010a)

L0 nir L0 nir L0 nir L0 nir L0 nir L4 nir L0 opt

Alves de Oliveira et al. (2012) (2012) Alves de Oliveira et al. (2012) Alves de Oliveira et al. (2012) Alves de Oliveira et al. (2012) Alves de Oliveira et al. (2012) Luhman et al. (2008)

L3 nir L0 opt

Mužic et al. (2015) Luhman et al. (2009a)

L/T nir

Zapatero Osorio et al. (2014a)

Acknowledgments ELM and VJSB are supported by grants AyA2015-69350-C3-1-P and AyA2015-69350-C3-2-P from the Spanish Ministry of Economy and Competitiveness (MINECO/FEDER).

References Aller KM, Liu MC, Magnier EA et al (2016) Brown dwarfs in young moving groups from panSTARRS1. I. AB Doradus. Astrophys J 821:120. https://doi.org/10.3847/0004-637X/821/2/120, 1604.04284 Allers KN, Liu MC (2013) A near-infrared spectroscopic study of young field ultracool dwarfs. Astrophys J 772:79. https://doi.org/10.1088/0004-637X/772/2/79, 1305.4418 Alves de Oliveira C, Moraux E, Bouvier J et al (2010) The low-mass population of the ¡ Ophiuchi molecular cloud. Astron Astrophys 515:A75. https://doi.org/10.1051/0004-6361/200913900, 1003.2205

488

V. J. S. Béjar and E. L. Martín

Alves de Oliveira C, Moraux E, Bouvier J, Bouy H (2012) Spectroscopy of new brown dwarf members of ¡ Ophiuchi and an updated initial mass function. Astron Astrophys 539:A151. https://doi.org/10.1051/0004-6361/201118230, 1201.1912 Alves de Oliveira C, Moraux E, Bouvier J et al (2013) Spectroscopy of brown dwarf candidates in IC 348 and the determination of its substellar IMF down to planetary masses. Astron Astrophys 549:A123. https://doi.org/10.1051/0004-6361/201220229, 1211.4029 Ardila D, Martín E, Basri G (2000) A survey for low-mass stars and Brown dwarfs in the Upper Scorpius OB association. Astron J 120:479–487. https://doi.org/10.1086/301443astro-ph/0003316 Baker DEA, Jameson RF, Casewell SL et al (2010) Low-mass stars and brown dwarfs in Praesepe. Mon Not R Astron Soc 408:2457–2475. https://doi.org/10.1111/j.1365-2966.2010.17302.x1007.0751 Baraffe I, Chabrier G, Allard F, Hauschildt PH (1998) Evolutionary models for solar metallicity low-mass stars: mass-magnitude relationships and color-magnitude diagrams. Astron Astrophys 337:403–412, astro-ph/9805009 Baraffe I, Chabrier G, Barman TS, Allard F, Hauschildt PH (2003) Evolutionary models for cool brown dwarfs and extrasolar giant planets. The case of HD 209458. Astron Astrophys 402: 701–712. https://doi.org/10.1051/0004-6361:20030252, astro-ph/0302293 Barman TS, Macintosh B, Konopacky QM, Marois C (2011) Clouds and chemistry in the atmosphere of extrasolar planet HR8799b. Astrophys J 733:65. https://doi.org/10.1088/0004-637X/733/1/65, 1103.3895 Barrado y Navascués D, Martín EL (2003) An empirical criterion to classify T Tauri stars and substellar analogs using low-resolution optical spectroscopy. Astron J 126:2997–3006. https://doi.org/10.1086/379673, astro-ph/0309284 Barrado y Navascués D, Zapatero Osorio MR, VJS B et al (2001) Optical spectroscopy of isolated planetary mass objects in the ¢ Orionis cluster. Astron Astrophys 377:L9–L13. https://doi.org/10.1051/0004-6361:20011152, astro-ph/0108249 Barrado y Navascués D, Bouvier J, Stauffer JR, Lodieu N, McCaughrean MJ (2002a) A substellar mass function for Alpha Persei. Astron Astrophys 395:813–821. https://doi.org/10.1051/0004-6361:20021262, astro- ph/0209032 Barrado y Navascués D, Zapatero Osorio MR, Martín EL et al (2002b) Discovery of a very cool object with extraordinarily strong H’ emission. Astron Astrophys 393:L85–L88. https://doi.org/10.1051/0004-6361:20021251, astro-ph/0208497 Barrado y Navascués D, VJS B, Mundt R et al (2003) The sigma Orionis substellar population. VLT/FORS spectroscopy and 2MASS photometry. Astron Astrophys 404:171–185. https://doi.org/10.1051/0004-6361:20030407 Barrado y Navascués D, Stauffer JR, Bouvier J, Jayawardhana R, Cuillandre JC (2004a) The substellar population of the young cluster œ orionis. Astrophys J 610:1064–1078. https://doi.org/10.1086/421762, astro- ph/0404072 Barrado y Navascués D, Stauffer JR, Jayawardhana R (2004b) Spectroscopy of very low mass stars and Brown dwarfs in IC 2391: lithium depletion and H’ emission. Astrophys J 614:386–397. https://doi.org/10.1086/423485, astro-ph/0406436 Barrado D, Morales-Calderón M, Palau A et al (2009) A proto brown dwarf candidate in Taurus. Astron Astrophys 508:859–867. https://doi.org/10.1051/0004-6361/200912276 Basri G, Martín EL (1999) PPL 15: the first Brown dwarf spectroscopic binary. Astron J 118:2460– 2465. https://doi.org/10.1086/301079, astro-ph/9908015 Basri G, Marcy GW, Graham JR (1996) Lithium in Brown dwarf candidates: the mass and age of the faintest Pleiades stars. Astrophys J 458:600. https://doi.org/10.1086/176842 Bayo A, Barrado D, Stauffer J et al (2011) Spectroscopy of very low mass stars and brown dwarfs in the lambda Orionis star forming region. I. Enlarging the census down to the planetary mass domain in Collinder 69. Astron Astrophys 536:A63. https://doi.org/10.1051/0004-6361/201116617, 1109.4917 Béjar VJS, Osorio MRZ, Rebolo R (1999) A search for very low mass stars and Brown dwarfs in the young ¢ Orionis cluster. Astrophys J 521:671–681. https://doi.org/10.1086/307583, astroph/9903217

22 Brown Dwarfs and Free-Floating Planets in Young Stellar Clusters

489

Béjar VJS, Martín EL, Zapatero Osorio MR et al (2001) The substellar mass function in ¢ Orionis. Astrophys J 556:830–836. https://doi.org/10.1086/321621, astro-ph/0104097 Béjar VJS, Zapatero Osorio MR, Rebolo R (2004) Optical and infrared photometry of new very low-mass stars and brown dwarfs in the\sigma Orionis cluster. Astronomische Nachr 325: 705–713. https://doi.org/10.1002/asna.200310247, astro-ph/0404125 Béjar VJS, Zapatero Osorio MR, Pérez-Garrido A et al (2008) Discovery of a wide companion near the deuterium-burning mass limit in the Upper Scorpius association. Astrophys J 673:L185. https://doi.org/10.1086/527557, 0712.3482 Béjar VJS, Zapatero Osorio MR, Rebolo R et al (2011) The substellar population of ¢ Orionis: a deep wide survey. Astrophys J 743:64. https://doi.org/10.1088/0004-637X/743/1/64, 1109.1210 Best WMJ, Liu MC, Magnier EA et al (2015) A search for L/T transition dwarfs with panSTARRS1 and WISE. II. L/T transition atmospheres and young discoveries. Astrophys J 814:118. https://doi.org/10.1088/0004-637X/814/2/118, 1612.02824 Bihain G, Rebolo R, Béjar VJS et al (2006) Pleiades low-mass brown dwarfs: the cluster L dwarf sequence. Astron Astrophys 458:805–816. https://doi.org/10.1051/0004-6361:20065124, astroph/0608255 Bihain G, Rebolo R, Zapatero Osorio MR et al (2009) Candidate free-floating superJupiters in the young ¢ Orionis open cluster. Astron Astrophys 506:1169–1182. https://doi.org/10.1051/0004-6361/200912210, 0909.0802 Bihain G, Rebolo R, Zapatero Osorio MR, Béjar VJS, Caballero JA (2010) Near-infrared low- resolution spectroscopy of Pleiades L-type brown dwarfs. Astron Astrophys 519:A93. https://doi.org/10.1051/0004-6361/200913676, 1005.3249 Binks AS, Jeffries RD (2014) A lithium depletion boundary age of 21 Myr for the Beta Pictoris moving group. Mon Not R Astron Soc 438:L11–L15. https://doi.org/10.1093/mnrasl/slt141, 1310.2613 Boss AP (2011) Formation of Giant planets by disk instability on wide orbits around Protostars with varied masses. Astrophys J 731:74. https://doi.org/10.1088/0004-637X/731/1/74, 1102.4555 Boss AP, Basri G, Kumar SS et al (2003) Nomenclature: brown dwarfs, gas giant planets, and ? In: Martín E (ed) Brown dwarfs, IAU Symposium, vol 211, p 529 Boudreault S, Lodieu N (2013) The first spectroscopically identified L dwarf in Praesepe. Mon Not R Astron Soc 434:142–147. https://doi.org/10.1093/mnras/stt1009, 1306.1065 Boudreault S, Bailer-Jones CAL (2009) A constraint on brown dwarf formation via ejection: radial variation of the stellar and substellar mass function of the young open cluster IC 2391. ApJ 706:1484–1503. https://doi.org/10.1088/0004-637X/706/2/1484, 0909.0842 Boudreault S, Bailer-Jones CAL, Goldman B, Henning T, Caballero JA (2010) Brown dwarfs and very low mass stars in the Praesepe open cluster: a dynamically unevolved mass function? Astron Astrophys 510:A27. https://doi.org/10.1051/0004-6361/200913011, 0910.4529 Boudreault S, Lodieu N, Deacon NR, Hambly NC (2012) Astrometric and photometric initial mass functions from the UKIDSS galactic clusters survey – III. Praesepe. Mon Not R Astron Soc 426:3419–3434. https://doi.org/10.1111/j.1365-2966.2012.21854.x, 1208.0466 Bouvier J, Stauffer JR, Martín EL et al (1998) Brown dwarfs and very low-mass stars in the Pleiades cluster: a deep wide-field imaging survey. Astron Astrophys 336:490–502 Bouvier J, Kendall T, Meeus G et al (2008) Brown dwarfs and very low mass stars in the Hyades cluster: a dynamically evolved mass function. Astron Astrophys 481:661–672. https://doi.org/10.1051/0004-6361:20079303, 0801.0670 Bouy H, Huélamo N, Pinte C et al (2008) Structural and compositional properties of brown dwarf disks: the case of 2MASS J04442713C2512164. Astron Astrophys 486:877–890. https://doi.org/10.1051/0004-6361:20078866, 0803.2051 Bowler BP, Liu MC, Shkolnik EL, Dupuy TJ (2013) Planets around low-mass stars. III. A young dusty L dwarf companion at the deuterium-burning limit. Astrophys J 774:55. https://doi.org/10.1088/0004-637X/774/1/55, 1307.2237 Bowler BP, Liu MC, Mawet D et al (2017) Planets around low-mass stars (PALMS). VI. Discovery of a remarkably red planetary-mass companion to the AB dor moving group candidate

490

V. J. S. Béjar and E. L. Martín

2MASS J22362452C4751425*. Astron J 153:18. https://doi.org/10.3847/1538-3881/153/1/18, 1611.00364 Briceño C, Hartmann L, Stauffer J, Martín E (1998) A search for very low mass pre-main-sequence stars in Taurus. Astron J 115:2074–2091. https://doi.org/10.1086/300338 Briceño C, Luhman KL, Hartmann L, Stauffer JR Kirkpatrick JD (2002) The initial mass function in the Taurus star-forming region. ApJ 580:317–335. https://doi.org/10.1086/343127 Bryja C, Jones TJ, Humphreys RM et al (1992) Candidate brown dwarfs in the Hyades. Astrophys J 388:L23–L26. https://doi.org/10.1086/186321 Bryja C, Humphreys RM, Jones TJ (1994) The lowest mass stars in the Hyades. Astron J 107: 246–253. https://doi.org/10.1086/116848 Burgess ASM, Moraux E, Bouvier J et al (2009) Young T-dwarf candidates in IC 348. Astron Astrophys 508:823–831. https://doi.org/10.1051/0004-6361/200912444, 0909.0917 Burke CJ, Pinsonneault MH, Sills A (2004) Theoretical examination of the lithium depletion boundary. Astrophys J 604:272–283. https://doi.org/10.1086/381242, astro-ph/0309461 Burningham B, Naylor T, Littlefair SP, Jeffries RD (2005) Contamination and exclusion in the ¢ Orionis young group. Mon Not R Astron Soc 356:1583–1591. https://doi.org/10.1111/j.1365-2966.2005.08631.x, astro- ph/0411418 Caballero JA (2008) Stars and brown dwarfs in the ¢ Orionis cluster: the Mayrit catalogue. Astron Astrophys 478:667–674. https://doi.org/10.1051/0004-6361:20077885, 0710.5882 Caballero JA (2017) A TGAS/Gaia DR1 parallactic distance to the ¢ Orionis cluster. Astronomische Nachr 338:629–634. https://doi.org/10.1002/asna.201713312, 1702.06046 Caballero JA, Béjar VJS, Rebolo R et al (2007) The substellar mass function in ¢ Orionis. II. Optical, near-infrared and IRAC/Spitzer photometry of young cluster brown dwarfs and planetarymass objects. Astron Astrophys 470:903–918. https://doi.org/10.1051/0004-6361:20066993, 0705.0922 Cargile PA, James DJ, Jeffries RD (2010) Identification of the lithium depletion boundary and age of the southern open cluster Blanco 1. Astrophys J 725:L111–L116. https://doi.org/10.1088/2041-8205/725/2/L111, 1010.6100 Carson J, Thalmann C, Janson M et al (2013) Direct imaging discovery of a “Super-Jupiter” around the late B-type star › and. Astrophys J 763:L32. https://doi.org/10.1088/2041-8205/763/2/L32, 1211.3744 Casewell SL, Dobbie PD, Hodgkin ST et al (2007) Proper motion L and T dwarf candidate members of the Pleiades. Mon Not R Astron Soc 378:1131–1140. https://doi.org/10.1111/j.1365-2966.2007.11848.x, 0704.1578 Casewell SL, Jameson RF, Burleigh MR et al (2011) Methane band and Spitzer mid-IR imaging of L and T dwarf candidates in the Pleiades. Mon Not R Astron Soc 412:2071–2078. https://doi.org/10.1111/j.1365-2966.2010.18044.x, 1011.4218 Chabrier G (2001) The galactic disk mass budget. I. Stellar mass function and density. Astrophys J 554:1274–1281. https://doi.org/10.1086/321401, astro-ph/0107018 Chabrier G, Baraffe I (2000) Theory of low-mass stars and substellar objects. Annu Rev Astron Astrophys 38:337–377. https://doi.org/10.1146/annurev.astro.38.1.337, astro-ph/0006383 Chabrier G, Baraffe I, Allard F, Hauschildt P (2000) Evolutionary models for very low-mass stars and Brown dwarfs with dusty atmospheres. Astrophys J 542:464–472. https://doi.org/10.1086/309513, astro- ph/0005557 Chappelle RJ, Pinfield DJ, Steele IA, Dobbie PD, Magazù A (2005) Crossing into the substellar regime in Praesepe. Mon Not R Astron Soc 361:1323–1336. https://doi.org/10.1111/j.1365-2966.2005.09285.x, astro- ph/0506493 Chauvin G, Lagrange AM, Dumas C et al (2004) A giant planet candidate near a young brown dwarf. Direct VLT/NACO observations using IR wavefront sensing. Astron Astrophys 425:L29–L32. https://doi.org/10.1051/0004-6361:200400056, astro-ph/0409323 Chauvin G, Lagrange AM, Zuckerman B et al (2005) A companion to AB pic at the planet/brown dwarf boundary. Astron Astrophys 438:L29–L32. https://doi.org/10.1051/0004-6361:200500111, astro-ph/0504658

22 Brown Dwarfs and Free-Floating Planets in Young Stellar Clusters

491

Chiang P, Chen WP, Albert L, Liu M, Magnier EA (2015) Searching for T dwarfs in the ¡ Oph dark cloud L 1688. Mon Not R Astron Soc 448:522–540. https://doi.org/10.1093/mnras/ stu2701 Comeron F, Rieke GH, Burrows A, Rieke MJ (1993) The stellar population in the rho Ophiuchi cluster. Astrophys J 416:185. https://doi.org/10.1086/173225 Comerón F, Rieke GH, Neuhaüser R (1999) Faint members of the Chamaeleon I cloud. Astron Astrophys 343:477–495 Cossburn MR, Hodgkin ST, Jameson RF, Pinfield DJ (1997) Discovery of the lowest mass brown dwarf in the Pleiades. Mon Not R Astron Soc 288:L23–L27. https://doi.org/10.1093/mnras/288.1.L23 Cruz KL, Kirkpatrick JD, Burgasser AJ (2009) Young L dwarfs identified in the field: a preliminary low-gravity, optical spectral sequence from L0 to L5. Astron J 137:3345–3357. https://doi.org/10.1088/0004-6256/137/2/3345, 0812.0364 Cummings JD, Deliyannis CP, Maderak RM, Steinhauer A (2017) WIYN open cluster study. LXXV. Testing the metallicity dependence of stellar lithium depletion using Hyades-aged clusters. I. Hyades and Praesepe. Astron J 153:128. https://doi.org/10.3847/1538-3881/aa5b86, 1702.03936 Da Rio N, Robberto M, Hillenbrand LA, Henning T, Stassun KG (2012) The initial mass function of the Orion nebula cluster across the H-burning limit. Astrophys J 748:14. https://doi.org/10.1088/0004-637X/748/1/14, 1112.2711 Dahm SE (2015) Reexamining the lithium depletion boundary in the Pleiades and the inferred age of the cluster. Astrophys J 813:108. https://doi.org/10.1088/0004-637X/813/2/108 Dawson P, Scholz A, Ray TP (2011) New brown dwarfs in the south part of the Upper Scorpius association. Mon Not R Astron Soc 418:1231–1237. https://doi.org/10.1111/j.1365-2966.2011.19573.x, 1108.1309 Dawson P, Scholz A, Ray TP et al (2013) New brown dwarf discs in Upper Scorpius observed with WISE. Mon Not R Astron Soc 429:903–914. https://doi.org/10.1093/mnras/sts386 Dawson P, Scholz A, Ray TP et al (2014) Near-infrared spectroscopy of young brown dwarfs in Upper Scorpius. Mon Not R Astron Soc 442:1586–1596. https://doi.org/10.1093/mnras/stu973, 1405.3842 de Bruijne JHJ (1999) Structure and colour-magnitude diagrams of Scorpius OB2 based on kinematic modelling of Hipparcos data. Mon Not R Astron Soc 310:585–617. https://doi.org/10.1046/j.1365-8711.1999.02953.x de Wit WJ, Bouvier J, Palla F et al (2006) Exploring the lower mass function in the young open cluster IC 4665. Astron Astrophys 448:189–202. https://doi.org/10.1051/0004-6361:20054102, astro-ph/0511175 de Zeeuw PT, Hoogerwerf R, de Bruijne JHJ, Brown AGA, Blaauw A (1999) A HIPPARCOS census of the nearby OB associations. Astron J 117:354–399. https://doi.org/10.1086/300682, astro-ph/9809227 Deacon NR, Hambly NC (2004) Proper motion surveys of the young open clusters alpha Persei and the Pleiades. Astron Astrophys 416:125–136. https://doi.org/10.1051/0004-6361:20034238, astro-ph/0311565 Delorme P, Gagné J, Malo L et al (2012) CFBDSIR2149-0403: a 4-7 Jupiter-mass freefloating planet in the young moving group AB Doradus? Astron Astrophys 548:A26. https://doi.org/10.1051/0004-6361/201219984, 1210.0305 Delorme P, Gagné J, Girard JH et al (2013) Direct-imaging discovery of a 12-14 Jupiter-mass object orbiting a young binary system of very low-mass stars. Astron Astrophys 553:L5. https://doi.org/10.1051/0004-6361/201321169, 1303.4525 Delorme P, Dupuy T, Gagné J et al (2017) CFBDSIR 2149-0403: young isolated planetary-mass object or high-metallicity low-mass brown dwarf?? ArXiv e-prints 1703.00843 Dobbie PD, Kenyon F, Jameson RF et al (2002a) A deep large-area search for very low-mass members of the Hyades open cluster. Mon Not R Astron Soc 329:543–555. https://doi.org/10.1046/j.1365-8711.2002.05002.x

492

V. J. S. Béjar and E. L. Martín

Dobbie PD, Kenyon F, Jameson RF et al (2002b) A deep IZ survey of 1.1 deg2 of the Pleiades cluster: three candidate members with M < 0.04 Mˇ . Mon Not R Astron Soc 335:687–697. https://doi.org/10.1046/j.1365-8711.2002.05650.x Dobbie PD, Lodieu N, Sharp RG (2010) IC 2602: a lithium depletion boundary age and new can- didate low-mass stellar members. Mon Not R Astron Soc 409:1002–1012. https://doi.org/10.1111/j.1365-2966.2010.17355.x Faherty JK, Burgasser AJ, Walter FM et al (2012) The Brown dwarf kinematics project (BDKP). III. Parallaxes for 70 ultracool dwarfs. Astrophys J 752:56. https://doi.org/10.1088/0004-637X/752/1/56, 1203.5543 Faherty JK, Riedel AR, Cruz KL et al (2016) Population properties of Brown dwarf analogs to exoplanets. Astrophys J Suppl Ser 225:10. https://doi.org/10.3847/0067-0049/225/1/10, 1605.07927 Fan X, Knapp GR, Strauss MA et al (2000) L dwarfs found in Sloan digital sky survey commissioning imaging data. Astron J 119:928–935. https://doi.org/10.1086/301224, astroph/9909263 Fernández M, Comerón F (2001) Intense accretion and mass loss of a very low mass young stellar object. Astron Astrophys 380:264–276. https://doi.org/10.1051/0004-6361:20011443 Festin L (1997) Brown dwarfs in the Pleiades. A deep IJK survey. Astron Astrophys 322:455–459, astro- ph/9701205 Festin L (1998) Brown dwarfs in the Pleiades. II. A deep optical and near infrared survey. Astron Astrophys 333:497–504, astro-ph/9802292 Gagné J, Lafrenière D, Doyon R, Malo L, Artigau É (2014) BANYAN. II. Very low mass and substellar candidate members to nearby, young kinematic groups with previously known signs of youth. Astrophys J 783:121. https://doi.org/10.1088/0004-637X/783/2/121, 1312.5864 Gagné J, Burgasser AJ, Faherty JK et al (2015a) SDSS J111010.01C011613.1: a new planetary- mass T dwarf member of the AB Doradus moving group. Astrophys J 808:L20. https://doi.org/10.1088/2041-8205/808/1/L20, 1506.04195 Gagné J, Lafreniére D, Doyon R, Malo L, Artigau É (2015b) BANYAN. V. A systematic all-sky survey for new very late-type low-mass stars and Brown dwarfs in nearby young moving groups. Astrophys J 798:73. https://doi.org/10.1088/0004-637X/798/2/73, 1410.4864 Garrison RF (1967) The ¢ Orionis clustering. Publ Astron Soc Pac 79:433. https://doi.org/10.1086/128517 Gauza B, Béjar VJS, Pérez-Garrido A et al (2015) Discovery of a young planetary mass Companion to the nearby M dwarf VHS J125601.92-125723.9. Astrophys J 804:96. https://doi.org/10.1088/0004-637X/804/2/96, 1505.00806 Gauza (2016) A direct imaging search and characterization of brown dwarfs and massive planets around stars. PhD thesis, Univ. de La Laguna, Tenerife Geers V, Scholz A, Jayawardhana R et al (2011) Substellar objects in nearby young clusters (SONYC). II. The Brown dwarf population of ¡ Ophiuchi. Astrophys J 726:23. https://doi.org/10.1088/0004-637X/726/1/23, 1010.5801 Gizis JE, Reid IN, Monet DG (1999) A 2MASS survey for Brown dwarfs toward the Hyades. Astron J 118:997–1004. https://doi.org/10.1086/300982, astro-ph/9905168 Goldman B, Marsat S, Henning T, Clemens C, Greiner J (2010) A new benchmark T8-9 brown dwarf and a couple of new mid-T dwarfs from the UKIDSS DR5C LAS. Mon Not R Astron Soc 405:1140–1152. https://doi.org/10.1111/j.1365-2966.2010.16524.x, 1002.2637 González Hernández JI, Caballero JA, Rebolo R et al (2008) Chemical abundances of latetype pre-main sequence stars in the ¢ Orionis cluster. Astron Astrophys 490:1135–1142. https://doi.org/10.1051/0004-6361:200810398, 0809.0090 González-García BM, Zapatero Osorio MR, Béjar VJS et al (2006) A search for substellar members in the Praesepe and ¢ Orionis clusters. Astron Astrophys 460:799–810. https://doi.org/10.1051/0004-6361:20065909, astro-ph/0609283 Gorlova NI, Meyer MR, Rieke GH, Liebert J (2003) Gravity indicators in the near-infrared spectra of Brown dwarfs. Astrophys J 593:1074–1092. https://doi.org/10.1086/376730, astroph/0305147

22 Brown Dwarfs and Free-Floating Planets in Young Stellar Clusters

493

Guieu S, Dougados C, Monin JL, Magnier E, Martín EL (2006) Seventeen new very lowmass members in Taurus. The brown dwarf deficit revisited. Astron Astrophys 446:485–500. https://doi.org/10.1051/0004-6361:20053493, astro-ph/0509317 Haisch KE Jr, Barsony M, Tinney C (2010) A methane imaging survey for T dwarf candidates in ¡ Ophiuchi. Astrophys J 719:L90–L94. https://doi.org/10.1088/2041-8205/719/1/L90, 1007.2406 Hambly NC, Jameson RF, Hawkins MRS (1991) A deep proper motion survey of the Pleiades for low-mass stars and brown dwarfs. Mon Not R Astron Soc 253:1–8. https://doi.org/10.1093/mnras/253.1.1 Hayashi C, Nakano T (1963) Evolution of stars of small masses in the pre-main-sequence stages. Prog Theor Phys 30:460–474. https://doi.org/10.1143/PTP.30.460 Hillenbrand LA (1997) On the stellar population and star-forming history of the Orion nebula cluster. Astron J 113:1733–1768. https://doi.org/10.1086/118389 Hillenbrand LA, Carpenter JM (2000) Constraints on the stellar/substellar mass function in the inner Orion nebula cluster. Astrophys J 540:236–254. https://doi.org/10.1086/309309, astroph/0003293 Hogan E, Jameson RF, Casewell SL, Osbourne SL, Hambly NC (2008) L dwarfs in the Hyades. Mon Not R Astron Soc 388:495–499. https://doi.org/10.1111/j.1365-2966.2008.13437.x, 0805.1189 Ireland MJ, Kraus A, Martinache F, Law N, Hillenbrand LA (2011) Two wide planetarymass companions to solar-type stars in Upper Scorpius. Astrophys J 726:113. https://doi.org/10.1088/0004-637X/726/2/113, 1011.2201 Itoh Y, Hayashi M, Tamura M et al (2005) A young Brown dwarf companion to DH Tauri. Astrophys J 620:984–993. https://doi.org/10.1086/427086, astro-ph/0411177 Jameson RF, Skillen I (1989) A search for low-mass stars and brown dwarfs in the Pleiades. Mon Not R Astron Soc 239:247–253. https://doi.org/10.1093/mnras/239.1.247 Jeffries RD, Oliveira JM (2005) The lithium depletion boundary in NGC 2547 as a test of pre-main- sequence evolutionary models. Mon Not R Astron Soc 358:13–29. https://doi.org/10.1111/j.1365-2966.2005.08820.x, astro- ph/0411112 Jeffries RD, Naylor T, Mayne NJ, Bell CPM, Littlefair SP (2013) A lithium depletion boundary age of 22 Myr for NGC 1960. Mon Not R Astron Soc 434:2438–2450. https://doi.org/10.1093/mnras/stt1180, 1306.6339 Kalas P, Graham JR, Chiang E et al (2008) Optical images of an exosolar planet 25 light-years from earth. Science 322:1345. https://doi.org/10.1126/science.1166609, 0811.1994 Kenyon MJ, Jeffries RD, Naylor T, Oliveira JM, Maxted PFL (2005) Membership, binarity and accretion among very low-mass stars and brown dwarfs of the ¢ Orionis cluster. Mon Not R Astron Soc 356:89–106. https://doi.org/10.1111/j.1365-2966.2004.08455.x, astro-ph/0409749 Kirkpatrick JD (2005) New spectral types L and T. Annu Rev Astron Astrophys 43:195–245. https://doi.org/10.1146/annurev.astro.42.053102.134017 Kirkpatrick JD, Barman TS, Burgasser AJ et al (2006) Discovery of a very young field L dwarf, 2MASS J01415823-4633574. Astrophys J 639:1120–1128. https://doi.org/10.1086/499622, astro- ph/0511462 Kraus AL, Hillenbrand LA (2007) The stellar populations of Praesepe and Coma Berenices. Astron J 134:2340–2352. https://doi.org/10.1086/522831, 0708.2719 Kraus AL, Shkolnik EL, Allers KN, Liu MC (2014) A stellar census of the Tucana-Horologium moving group. Astron J 147:146. https://doi.org/10.1088/0004-6256/147/6/146, 1403.0050 Kroupa P (2001) On the variation of the initial mass function. Mon Not R Astron Soc 322:231–246. https://doi.org/10.1046/j.1365-8711.2001.04022.x, astro-ph/0009005 Kumar SS (1963) The structure of stars of very low mass. Astrophys J 137:1121. https://doi.org/10.1086/147589 Kuzuhara M, Tamura M, Ishii M et al (2011) The widest-separation substellar companion candidate to a binary T Tauri star. Astron J 141:119. https://doi.org/10.1088/0004-6256/141/4/119 Kuzuhara M, Tamura M, Kudo T et al (2013) Direct imaging of a cold Jovian exoplanet in orbit around the sun-like star GJ 504. Astrophys J 774:11. https://doi.org/10.1088/0004-637X/774/1/11, 1307.2886

494

V. J. S. Béjar and E. L. Martín

Lafrenière D, Jayawardhana R, van Kerkwijk MH (2008) Direct imaging and spectroscopy of a planetary-mass candidate companion to a young solar analog. Astrophys J 689:L153. https://doi.org/10.1086/595870, 0809.1424 Lagrange AM, Bonnefoy M, Chauvin G et al (2010) A giant planet imaged in the disk of the young star “ Pictoris. Science 329:57. https://doi.org/10.1126/science.1187187, 1006.3314 Lee TA (1968) Interstellar extinction in the Orion association. Astrophys J 152:913. https://doi.org/10.1086/149607 Liu MC, Magnier EA, Deacon NR et al (2013) The extremely red, young L dwarf PSO J318.533822.8603: a free-floating planetary-mass analog to directly imaged young gas-giant planets. Astrophys J 777:L20. https://doi.org/10.1088/2041-8205/777/2/L20, 1310.0457 Liu MC, Dupuy TJ, Allers KN (2016) The Hawaii infrared parallax program. II. Young ultracool field dwarfs. Astrophys J 833:96. https://doi.org/10.3847/1538-4357/833/1/96, 1612.02426 Lodieu N (2013) Astrometric and photometric initial mass functions from the UKIDSS galactic clusters survey – IV. Upper Sco. Mon Not R Astron Soc 431:3222–3235. https://doi.org/10.1093/mnras/stt402, 1303.1351 Lodieu N, McCaughrean MJ, Barrado Y, Navascués D, Bouvier J, Stauffer JR (2005) A near- infrared survey for new low-mass members in ’ Per. Astron Astrophys 436:853–865. https://doi.org/10.1051/0004-6361:20042048, astro-ph/0502446 Lodieu N, Hambly NC, Jameson RF (2006) New members in the Upper Scorpius association from the UKIRT infrared deep sky survey early data release. Mon Not R Astron Soc 373:95–104. https://doi.org/10.1111/j.1365-2966.2006.10958.x, astro-ph/0609532 Lodieu N, Dobbie PD, Deacon NR et al (2007a) A wide deep infrared look at the Pleiades with UKIDSS: new constraints on the substellar binary fraction and the low-mass initial mass function. Mon Not R Astron Soc 380:712–732. https://doi.org/10.1111/j.1365-2966.2007.12106.x, 0706.2234 Lodieu N, Hambly NC, Jameson RF et al (2007b) New brown dwarfs in Upper Sco using UKIDSS galactic cluster survey science verification data. Mon Not R Astron Soc 374:372–384. https://doi.org/10.1111/j.1365-2966.2006.11151.x, astro-ph/0610140 Lodieu N, Hambly NC, Jameson RF, Hodgkin ST (2008) Near-infrared cross-dispersed spectroscopy of brown dwarf candidates in the UpperSco association. Mon Not R Astron Soc 383:1385–1396. https://doi.org/10.1111/j.1365-2966.2007.12676.x, 0711.1109 Lodieu N, Zapatero Osorio MR, Rebolo R, Martín EL, Hambly NC (2009) A census of verylow- mass stars and brown dwarfs in the ¢ Orionis cluster. Astron Astrophys 505:1115–1127. https://doi.org/10.1051/0004-6361/200911966, 0907.2185 Lodieu N, de Wit WJ, Carraro G et al (2011a) The mass function of IC 4665 revisited by the UKIDSS galactic clusters survey. Astron Astrophys 532:A103. https://doi.org/10.1051/0004-6361/201116883, 1106.3957 Lodieu N, Dobbie PD, Hambly NC (2011b) Multi-fibre optical spectroscopy of lowmass stars and brown dwarfs in Upper Scorpius. Astron Astrophys 527:A24. https://doi.org/10.1051/0004-6361/201014992, 1101.0919 Lodieu N, Hambly NC, Dobbie PD et al (2011c) Testing the fragmentation limit in the Upper Sco association. Mon Not R Astron Soc 418:2604–2617. https://doi.org/10.1111/j.1365-2966.2011.19651.x, 1108.4783 Lodieu N, Deacon NR, Hambly NC (2012a) Astrometric and photometric initial mass functions from the UKIDSS galactic clusters survey – I. The Pleiades. Mon Not R Astron Soc 422:1495– 1511. https://doi.org/10.1111/j.1365-2966.2012.20723.x, 1204.2659 Lodieu N, Deacon NR, Hambly NC, Boudreault S (2012b) Astrometric and photometric initial mass functions from the UKIDSS Galactic Clusters Survey – II. The Alpha Persei open cluster. Mon Not R Astron Soc 426:3403–3418. https://doi.org/10.1111/j.1365-2966.2012.21811.x, 1207.6978 Lodieu N, Dobbie PD, Cross NJG et al (2013a) Probing the Upper Scorpius mass function in the planetary-mass regime. Mon Not R Astron Soc 435:2474–2482. https://doi.org/10.1093/mnras/stt1460, 1308.1310

22 Brown Dwarfs and Free-Floating Planets in Young Stellar Clusters

495

Lodieu N, Ivanov VD, Dobbie PD (2013b) Proper motions of USco T-type candidates. Mon Not R Astron Soc 430:1784–1789. https://doi.org/10.1093/mnras/sts726, 1303.1985 Lodieu N, Boudreault S, Béjar VJS (2014) Spectroscopy of Hyades L dwarf candidates. Mon Not R Astron Soc 445:3908–3918. https://doi.org/10.1093/mnras/stu2059, 1410.0192 Lodieu N, Zapatero Osorio MR, Béjar VJS, Peña-Ramírez K (2018) The opticalCinfrared L dwarf spectral sequence of young planetary-mass objects in the Upper Scorpius association. ArXiv eprints 1709.02139 López Martí B, Eislöffel J, Scholz A, Mundt R (2004) The brown dwarf population in the Chamaeleon I cloud. Astron Astrophys 416:555–576. https://doi.org/ 10.1051/0004-6361:20031720, astro-ph/0312026 Low C, Lynden-Bell D (1976) The minimum Jeans mass or when fragmentation must stop. Mon Not R Astron Soc 176:367–390. https://doi.org/10.1093/mnras/176.2.367 Lucas PW, Roche PF (2000) A population of very young brown dwarfs and free-floating planets in Orion. Mon Not R Astron Soc 314:858–864., https://doi.org/10.1046/j.1365-8711.2000. 03515.x astro-ph/0003061 Lucas PW, Roche PF, Allard F, Hauschildt PH (2001) Infrared spectroscopy of substellar objects in Orion. Mon Not R Astron Soc 326:695–721. https://doi.org/10.1046/j.1365-8711.2001. 04666.x, astro-ph/0105154 Lucas PW, Roche PF, Tamura M (2005) A deep survey of brown dwarfs in Orion with Gemini. Mon Not R Astron Soc 361:211–232. https://doi.org/10.1111/j.1365-2966.2005.09156.x, astroph/0504570 Lucas PW, Weights DJ, Roche PF, Riddick FC (2006) Spectroscopy of planetary mass brown dwarfs in Orion. MNRAS 373:L60–L64. https://doi.org/10.1111/j.1745-3933.2006.00244.x, astroph/0609086 Luhman KL (1999) Young low-mass stars and Brown dwarfs in IC 348. Astrophys J 525:466–481. https://doi.org/10.1086/307902, astro-ph/9905287 Luhman KL (2000) The initial mass function of low-mass stars and Brown dwarfs in Taurus. Astrophys J 544:1044–1055. https://doi.org/10.1086/317232 Luhman KL (2004a) A survey for low-mass stars and Brown dwarfs in the ˜ Chamaeleon- tis and " Chamaeleontis young associations. Astrophys J 616:1033–1041. https://doi.org/10.1086/424963, astro- ph/0411444 Luhman KL (2004b) New Brown dwarfs and an updated initial mass function in Taurus. Astrophys J 617:1216–1232. https://doi.org/10.1086/425647, astro-ph/0411447 Luhman KL (2007) The stellar population of the Chamaeleon I star-forming region. Astrophys J Suppl Ser 173:104–136. https://doi.org/10.1086/520114, 0710.3037 Luhman KL, Muench AA (2008) New low-mass stars and Brown dwarfs with disks in the Chamaeleon I star-forming region. Astrophys J 684:654–662. https://doi.org/10.1086/590364, 0805.3722 Luhman KL, Liebert J, Rieke GH (1997) Spectroscopy of a young Brown dwarf in the rho Ophiuchi cluster 1. Astrophys J 489:L165. https://doi.org/10.1086/316784 Luhman KL, Rieke GH, Young ET et al (2000) The initial mass function of low-mass stars and Brown dwarfs in young clusters. Astrophys J 540:1016–1040. https://doi.org/10.1086/309365, astro-ph/0004386 Luhman KL, Briceño C, Stauffer JR et al (2003a) New low-mass members of the Taurus starforming region. Astrophys J 590:348–356. https://doi.org/10.1086/374983, astro-ph/f\breakg 0304414 Luhman KL, Stauffer JR, Muench AA et al (2003b) A census of the young cluster IC 348. Astrophys J 593:1093–1115. https://doi.org/10.1086/376594, astro-ph/0304409 Luhman KL, Whitney BA, Meade MR et al (2006a) A survey for new members of Taurus with the Spitzer space telescope. Astrophys J 647:1180–1191. https://doi.org/10.1086/505572, astroph/0608548 Luhman KL, Wilson JC, Brandner W et al (2006b) Discovery of a young substellar companion in Chamaeleon. Astrophys J 649:894–899. https://doi.org/10.1086/506517, astro-ph/0609187

496

V. J. S. Béjar and E. L. Martín

Luhman KL, Allen LE, Allen PR et al (2008) The disk population of the Chamaeleon I star- forming region. Astrophys J 675:1375–1406. https://doi.org/10.1086/527347, 0803.1019 Luhman KL, Mamajek EE, Allen PR, Cruz KL (2009a) An infrared/X-ray survey for new mem- bers of the Taurus star-forming region. Astrophys J 703:399–419. https://doi.org/10.1088/0004-637X/703/1/399, 0911.5451 Luhman KL, Mamajek EE, Allen PR, Muench AA, Finkbeiner DP (2009b) Discovery of a wide binary Brown dwarf born in isolation. Astrophys J 691:1265–1275. https://doi.org/10.1088/0004-637X/691/2/1265, 0902.0425 Luhman KL, Burgasser AJ, Bochanski JJ (2011) Discovery of a candidate for the coolest known Brown dwarf. Astrophys J 730:L9. https://doi.org/10.1088/2041-8205/730/1/L9, 1102.5411 Lynga G (1982) Open clusters in our galaxy. Astron Astrophys 109:213–222 Lynga G (1983) The Lund catalogue of open cluster parameters. In: Ruprecht J Palous J (eds) Star clusters and associations and their relation to the evolution of the galaxy, Ceskoslovenska Akademie Ved Lyo AR, Song I, Lawson WA, Bessell MS, Zuckerman B (2006) A deep photometric survey of the ˜ Chamaeleontis cluster down to the brown dwarf – planet boundary. Mon Not R Astron Soc 368:1451–1455. https://doi.org/10.1111/j.1365-2966.2006.10232.x, astro-ph/0604054 Magazzù A, Martín EL, Rebolo R (1993) A spectroscopic test for substellar objects. Astrophys J 404:L17–L20. https://doi.org/10.1086/186733 Magazzù A, Rebolo R, Zapatero Osorio MR, Martín EL, Hodgkin ST (1998) A Brown dwarf candidate in the Praesepe open cluster. Astrophys J 497:L47–L50. https://doi.org/10.1086/311273, astro-ph/9802178 Mainzer AK, McLean IS (2003) Using narrowband photometry to detect young Brown dwarfs in IC 348. Astrophys J 597:555–565. https://doi.org/10.1086/378197, astro-ph/0306631 Malo L, Doyon R, Lafrenière D et al (2013) Bayesian analysis to identify new star candidates in nearby young stellar kinematic groups. Astrophys J 762:88. https://doi.org/10.1088/0004-637X/762/2/88, 1209.2077 Manzi S, Randich S, de Wit WJ, Palla F (2008) Detection of the lithium depletion boundary in the young open cluster IC 4665. Astron Astrophys 479:141–148. https://doi.org/10.1051/0004-6361:20078226, 0712.0226 Marley MS, Saumon D, Cushing M et al (2012) Masses, radii, and cloud properties of the HR 8799 planets. Astrophys J 754:135. https://doi.org/10.1088/0004-637X/754/2/135, 1205.6488 Marois C, Macintosh B, Barman T et al (2008) Direct imaging of multiple planets orbiting the star HR 8799. Science 322:1348. https://doi.org/10.1126/science.1166585, 0811.2606 Marsh KA, Kirkpatrick JD, Plavchan P (2010a) A young planetary-mass object in the ¡ Oph cloud Core. Astrophys J 709:L158–L162. https://doi.org/10.1088/2041-8205/709/2/L158,f\breakg 0912.3774 Marsh KA, Plavchan P, Kirkpatrick JD et al (2010b) Deep near-infrared imaging of the ¡ Oph cloud core: clues to the origin of the lowest-mass Brown dwarfs. Astrophys J 719:550–560. https://doi.org/10.1088/0004-637X/719/1/550, 1006.2506 Martín EL, Rebolo R, Zapatero-Osorio MR (1996) Spectroscopy of new substellar candidates in the Pleiades: toward a spectral sequence for young Brown dwarfs. Astrophys J 469:706. https://doi.org/10.1086/177817, astro-ph/9604080 Martín EL, Basri G, Gallegos JE et al (1998a) A new Pleiades member at the lithium substellar boundary. Astrophys J 499:L61–L64. https://doi.org/10.1086/311350, physics/9803026 Martín EL, Basri G, Zapatero-Osorio MR, Rebolo R, López RJG (1998b) The first L-type Brown dwarf in the Pleiades. Astrophys J 507:L41–L44. https://doi.org/10.1086/311675, astroph/9809031 Martín EL, Dougados C, Magnier E et al (2001a) Four Brown dwarfs in the Taurus star-forming region. Astrophys J 561:L195–L198. https://doi.org/10.1086/324754, astro-ph/0110100 Martín EL, Zapatero Osorio MR, Barradoy Navascués D, VJS B, Rebolo R (2001b) Keck NIRC observations of planetary-mass candidate members in the ¢ Orionis open cluster. Astrophys J 558:L117–L121. https://doi.org/10.1086/323633, astro-ph/0108066

22 Brown Dwarfs and Free-Floating Planets in Young Stellar Clusters

497

Martín EL, Delfosse X, Guieu S (2004) Spectroscopic identification of DENIS-selected Brown dwarf candidates in the Upper Scorpius OB association. Astron J 127:449–454. https://doi.org/10.1086/380226r, astro-ph/0310819 Martín EL, Phan-Bao N, Bessell M et al (2010) Spectroscopic characterization of 78 DENIS ultracool dwarf candidates in the solar neighborhood and the upper Scorpii OB association. Astron Astrophys 517:A53. https://doi.org/10.1051/0004-6361/201014202r, 1004.1775 Mayne NJ, Harries TJ, Rowe J, Acreman DM (2012) Bayesian fitting of Taurus brown dwarf spectral energy distributions. Mon Not R Astron Soc 423:1775–1804. https://doi.org/10.1111/j.1365-2966.2012.20999.x, 1203.6657 McCaughrean M, Zinnecker H, Rayner J, Stauffer J (1995) Low-mass stars and Brown dwarfs in the trapezium cluster. In: Tinney CG (ed) The bottom of the main sequence – and beyond, Springer-Verlag, Berlin/Heidelberg, p 209 Melis C, Reid MJ, Mioduszewski AJ, Stauffer JR, Bower GC (2014) A VLBI resolution of the Pleiades distance controversy. Science 345:1029–1032. https://doi.org/10.1126/science.1256101, 1408.6544 Miller GE, Scalo JM (1979) The initial mass function and stellar birthrate in the solar neighborhood. Astrophys J Suppl Ser 41:513–547. https://doi.org/10.1086/190629 Moraux E, Bouvier J, Stauffer JR, Cuillandre JC (2003) Brown dwarfs in the Pleiades cluster: clues to the substellar mass function. Astron Astrophys 400:891–902. https://doi.org/10.1051/0004-6361:20021903, astro- ph/0212571 Moraux E, Bouvier J, Stauffer JR, Barradoy Navascués D, Cuillandre JC (2007) The lower mass function of the young open cluster Blanco 1: from 30 MJup to 3 Mˇ . Astron Astrophys 471:499– 513. https://doi.org/10.1051/0004-6361:20066308, 0706.2102 Muench AA, Lada EA, Lada CJ (2000) Modeling the near-infrared luminosity functions of young stellar clusters. Astrophys J 533:358–371. https://doi.org/10.1086/308638, astro-ph/9912384 Muzˇic K, Scholz A, Geers VC, Jayawardhana R, López Martí B (2014) Substellar objects in nearby young clusters (SONYC). VIII. Substellar population in lupus 3. Astrophys J 785:159. https://doi.org/10.1088/0004-637X/785/2/159, 1403.0813 Muzˇic K, Scholz A, Geers VC, Jayawardhana R (2015) Substellar objects in nearby young clusters (SONYC) IX: the planetary-mass domain of Chamaeleon-I and updated mass function in Lupus-3. Astrophys J 810:159. https://doi.org/10.1088/0004-637X/810/2/159, 1507.07780 Najita JR, Tiede GP, Carr JS (2000) From stars to Superplanets: the low-mass initial mass function in the young cluster IC 348. Astrophys J 541:977–1003. https://doi.org/10.1086/309477, astroph/0005290 Nakajima T, Oppenheimer BR, Kulkarni SR et al (1995) Discovery of a cool brown dwarf. Nature 378:463–465. https://doi.org/10.1038/378463a0 Naud ME, Artigau É, Malo L et al (2014) Discovery of a wide planetary-mass companion to the young M3 star GU Psc. Astrophys J 787:5. https://doi.org/10.1088/0004-637X/787/1/5, 1405.2932 Oliveira JM, Jeffries RD, van Loon JT (2009) The low-mass initial mass function in the young cluster NGC6611. Mon Not R Astron Soc 392:1034–1050. https://doi.org/10.1111/j.1365-2966.2008.14140.x, 0810.4444 Pascucci I, Apai D, Luhman K et al (2009) The different evolution of gas and dust in disks around sun-like and cool stars. Astrophys J 696:143–159. https://doi.org/10.1088/0004-637X/696/1/143, 0810.2552 Pecaut MJ, Mamajek EE, Bubar EJ (2012) A revised age for Upper Scorpius and the star formation history among the F-type members of the Scorpius-Centaurus OB association. Astrophys J 746:154. https://doi.org/10.1088/0004-637X/746/2/154, 1112.1695 Peña-Ramírez K, Zapatero Osorio MR, Béjar VJS, Rebolo R, Bihain G (2011) Search and characterization of T-type planetary mass candidates in the ¢ Orionis cluster. Astron Astrophys 532:A42. https://doi.org/10.1051/0004-6361/201116812, 1105.4043 Peña-Ramírez K, Béjar VJS, Zapatero Osorio MR, Petr-Gotzens MG Martín EL (2012) New isolated planetary-mass objects and the stellar and substellar mass function of the ¢ Orionis cluster. ApJ 754:30. https://doi.org/10.1088/0004-637X/754/1/30, 1205.4950

498

V. J. S. Béjar and E. L. Martín

Peña-Ramírez K, Zapatero Osorio MR, Béjar VJS (2015) Characterization of the known T-type dwarfs towards the ¢ Orionis cluster. Astron Astrophys 574:A118. https://doi.org/10.1051/0004-6361/201424816, 1411.3370 Peña-Ramírez K, Béjar VJS, Zapatero Osorio MR (2016) A new free-floating planet in the Upper Scorpius association. Astron Astrophys 586:A157. https://doi.org/10.1051/0004-6361/201527425, 1511.05586 Pérez-Garrido A, Lodieu N, Rebolo R (2017) A new L5 brown dwarf member of the Hyades cluster with chromospheric activity. Astron Astrophys 599:A78. https://doi.org/10.1051/0004-6361/201628778, 1701.03398 Perryman MAC, Lindegren L, Kovalevsky J et al (1997) The HIPPARCOS catalogue. Astron Astrophys 323:L49–L52 Pinfield DJ, Hodgkin ST, Jameson RF, Cossburn MR, von Hippel T (1997) Brown dwarf candidates in Praesepe. Mon Not R Astron Soc 287:180–188. https://doi.org/10.1093/mnras/287.1.180 Pinfield DJ, Hodgkin ST, Jameson RF et al (2000) A six-square-degree survey for Pleiades low- mass stars and brown dwarfs. Mon Not R Astron Soc 313:347–363. https://doi.org/10.1046/j.1365-8711.2000.03238.x Preibisch T, Zinnecker H (1999) The history of low-mass star formation in the Upper Scorpius OB association. Astron J 117:2381–2397. https://doi.org/10.1086/300842 Preibisch T, Guenther E, Zinnecker H et al (1998) A lithium-survey for pre-main sequence stars in the Upper Scorpius OB association. Astron Astrophys 333:619–628 Preibisch T, Brown AGA, Bridges T, Guenther E, Zinnecker H (2002) Exploring the full stellar population of the Upper Scorpius OB association. Astron J 124:404–416. https://doi.org/10.1086/341174 Quanz SP, Goldman B, Henning T et al (2010) Search for very low-mass Brown dwarfs and free-floating planetary-mass objects in Taurus. Astrophys J 708:770–784. https://doi.org/10.1088/0004-637X/708/1/770, 0911.1925 Rameau J, Chauvin G, Lagrange AM et al (2013) Discovery of a probable 45 Jupiter-mass exoplanet to HD 95086 by direct imaging. Astrophys J 772:L15. https://doi.org/10.1088/2041-8205/772/2/L15, 1305.7428 Rebolo R, Martín EL, Magazzù A (1992) Spectroscopy of a brown dwarf candidate in the alpha Persei open cluster. Astrophys J 389:L83–L86. https://doi.org/10.1086/186354 Rebolo R, Zapatero Osorio MR, Martín EL (1995) Discovery of a brown dwarf in the Pleiades star cluster. Nature 377:129–131. https://doi.org/10.1038/377129a0 Rebolo R, Martín EL, Basri G, Marcy GW, Zapatero-Osorio MR (1996) Brown dwarfs in the Pleiades cluster confirmed by the lithium test. Astrophys J 469:L53. https://doi.org/10.1086/310263, astro- ph/9607002 Rebolo R, Zapatero Osorio MR, Madruga S et al (1998) Discovery of a low-mass Brown dwarf companion of the young nearby star G 196-3. Science 282:1309. https://doi.org/10.1126/science.282.5392.1309, astro-ph/9811413 Reid IN, Kirkpatrick JD, Liebert J et al (1999) L dwarfs and the substellar mass function. Astrophys J 521:613–629. https://doi.org/10.1086/307589, astro-ph/9905170 Reipurth B (2002) The formation of Brown dwarfs. In: Alves JF McCaughrean MJ (eds) The origin of stars and planets: the VLT view, Springer-Verlag, p 114, https://doi.org/10.1007/ 1085651814 Riaz B, Honda M, Campins H et al (2012) The radial distribution of dust species in young brown dwarf discs. Mon Not R Astron Soc 420:2603–2624. https://doi.org/10.1111/j.1365-2966.2011.20233.x, 1111.4480 Rieke GH, Rieke MJ (1990) Possible substellar objects in the rho Ophiuchi cloud. Astrophys J 362:L21–L24. https://doi.org/10.1086/185838 Rodriguez DR, Bessell MS, Zuckerman B, Kastner JH (2011) A new method to identify nearby, young, low-mass stars. Astrophys J 727:62. https://doi.org/10.1088/0004-637X/727/2/62, 1010.2493 Salpeter EE (1955) The luminosity function and stellar evolution. Astrophys J 121:161. https://doi.org/10.1086/145971

22 Brown Dwarfs and Free-Floating Planets in Young Stellar Clusters

499

Saumon D, Hubbard WB, Burrows A et al (1996) A theory of extrasolar Giant planets. Astrophys J 460:993. https://doi.org/10.1086/177027, astro-ph/9510046 Scalo JM (1986) The stellar initial mass function. Fund Cosm Phys 11:1–278 Schlieder JE, Lépine S, Simon M (2012) Cool young stars in the northern hemisphere: “ Pictoris and AB Doradus moving group candidates. Astron J 143:80. https://doi.org/10.1088/0004-6256/143/4/80, 1201.4047 Schneider AC, Cushing MC, Kirkpatrick JD et al (2014) Discovery of the young L dwarf WISE J174102.78-464225.5. Astrophys J 147:34. https://doi.org/10.1088/0004-6256/147/2/34, 1311.5941 Schneider AC, Windsor J, Cushing MC, Kirkpatrick JD, Shkolnik EL (2017) A 2MASS/AllWISE search for extremely red L dwarfs: the discovery of several likely L type members of “ pic, AB dor, Tuc-hor, Argus, and the Hyades. Astron J 153:196. https://doi.org/10.3847/1538-3881/aa6624, 1703.03774 Scholz A, Geers V, Jayawardhana R et al (2009) Substellar objects in nearby young clusters (SONYC): the bottom of the initial mass function in NGC 1333. Astrophys J 702:805–822. https://doi.org/10.1088/0004-637X/702/1/805, 0907.2243 Sherry WH, Walter FM, Wolk SJ (2004) Photometric identification of the low-mass population of Orion OB1b. I. The ¢ Orionis cluster. Astron J 128:2316–2330. https://doi.org/10.1086/424863, astro-ph/0410244 Shkolnik E, Liu MC, Reid IN (2009) Identifying the young low-mass stars within 25 pc. I. Spectroscopic observations. Astrophys J 699:649–666. https://doi.org/10.1088/0004-637X/699/1/649, 0904.3323 Shkolnik EL, Liu MC, Reid IN, Dupuy T, Weinberger AJ (2011) Searching for young M dwarfs with GALEX. Astrophys J 727:6. https://doi.org/10.1088/0004-637X/727/1/6, 1011.2708 Shkolnik EL, Anglada-Escudé G, Liu MC et al (2012) Identifying the young low-mass stars within 25 pc. II. Distances, kinematics, and group membership. Astrophys J 758:56. https://doi.org/10.1088/0004-637X/758/1/56, 1207.5074 Siess L, Dufour E, Forestini M (2000) An internet server for pre-main sequence tracks of low- and intermediate-mass stars. Astron Astrophys 358:593–599, astro-ph/0003477 Silk J (1977) On the fragmentation of cosmic gas clouds. II – opacity-limited star formation. Astrophys J 214:152–160. https://doi.org/10.1086/155240 Simón-Díaz S, Caballero JA, Lorenzo J et al (2015) Orbital and physical properties of the ¢ Ori Aa, Ab, B triple system. Astrophys J 799:169. https://doi.org/10.1088/0004-637X/799/2/169, 1412.3469 Slesnick CL, Carpenter JM, Hillenbrand LA (2006) A large-area search for low-mass objects in Upper Scorpius. I. The photometric campaign and new Brown dwarfs. Astron J 131:3016–3027. https://doi.org/10.1086/503560, astro-ph/0602298 Slesnick CL, Hillenbrand LA, Carpenter JM (2008) A large-area search for low-mass objects in Upper Scorpius. II. Age and mass distributions. Astrophys J 688:377–397. https://doi.org/10.1086/592265, 0809.1436 Soderblom DR, Nelan E, Benedict GF et al (2005) Confirmation of errors in Hipparcos parallaxes from Hubble space telescope fine guidance sensor astrometry of the Pleiades. Astron J 129:1616–1624. https://doi.org/10.1086/427860, astro-ph/0412093\ Spezzi L, Alves de Oliveira C, Moraux E et al (2012) Searching for planetary-mass T-dwarfs in the core of Serpens. Astron Astrophys 545:A105. https://doi.org/10.1051/0004-6361/201219559, 1208.0702 Stamatellos D, Hubber DA, Whitworth AP (2007) Brown dwarf formation by gravitational fragmentation of massive, extended protostellar discs. Mon Not R Astron Soc 382:L30–L34. https://doi.org/10.1111/j.1745-3933.2007.00383.x, 0708.2827 Stauffer J, Hamilton D, Probst R, Rieke G, Mateo M (1989) Possible Pleiades members with M of about 0.07 solar mass – identification of brown dwarf candidates of known age, distance, and metallicity. Astrophys J 344:L21–L24. https://doi.org/10.1086/185521 Stauffer JR, Hamilton D Probst RG (1994) A CCD-based search for very low mass members of the Pleiades cluster. AJ108:155–159, https://doi.org/10.1086/117053

500

V. J. S. Béjar and E. L. Martín

Stauffer JR, Balachandran SC, Krishnamurthi A et al (1997) Rotational velocities and chromospheric activity of M dwarfs in the Hyades. Astrophys J 475:604–622. https://doi.org/10.1086/303567 Stauffer JR, Schultz G, Kirkpatrick JD (1998) Keck spectra of Pleiades Brown dwarf candidates and a precise determination of the lithium depletion edge in the Pleiades. Astrophys J 499:L199– L203. https://doi.org/10.1086/311379, astro-ph/9804005 Stauffer JR, Barrado y Navascués D, Bouvier J et al (1999) Keck spectra of Brown dwarf candidates and a precise determination of the lithium depletion boundary in the ’ Persei open cluster. Astrophys J 527:219–229. https://doi.org/10.1086/308069, astro-ph/9909207 Steele IA, Jameson RF (1995) Optical spectroscopy of low-mass stars and brown dwarfs in the Pleiades. Mon Not R Astron Soc 272:630–646. https://doi.org/10.1093/mnras/272.3.630 Strom KM, Kepner J, Strom SE (1995) The evolutionary status of the stellar population in the rho Ophiuchi cloud core. Astrophys J 438:813–829. https://doi.org/10.1086/175125 Thalmann C, Carson J, Janson M et al (2009) Discovery of the coldest imaged companion of a sun-like star. Astrophys J 707:L123–L127. https://doi.org/10.1088/0004-637X/707/2/L123, 0911.1127 Todorov K, Luhman KL, McLeod KK (2010) Discovery of a planetarymass companion to a Brown dwarf in Taurus. Astrophys J 714:L84–L88. https://doi.org/10.1088/2041-8205/714/1/L84, 1004.0539 Tognelli E, Prada Moroni PG, Degl’ Innocenti S (2015) Cumulative theoretical uncertainties in lithium depletion boundary age. Mon Not R Astron Soc 449:3741–3754. https://doi.org/10.1093/mnras/stv577, 1504.02698 van der Plas G, Ménard F, Ward-Duong K et al (2016) Dust masses of disks around 8 Brown dwarfs and very low-mass stars in upper Sco OB1 and Ophiuchus. Astrophys J 819:102. https://doi.org/10.3847/0004-637X/819/2/102, 1602.01724 van Leeuwen F (2009) Parallaxes and proper motions for 20 open clusters as based on the new Hipparcos catalogue. Astron Astrophys 497:209–242. https://doi.org/10.1051/0004-6361/200811382, 0902.1039 Walter FM, Vrba FJ, Mathieu RD, Brown A, Myers PC (1994) X-ray sources in regions of star formation. 5: the low mass stars of the Upper Scorpius association. Astron J 107:692–719. https://doi.org/10.1086/116889 Walter FM, Wolk SJ, Freyberg M, Schmitt JHMM (1997) Discovery of the ¢ Orionis cluster. Mem Soc Astron Ital 68:1081 Wang W, Boudreault S, Goldman B et al (2011) The substellar mass function in the central region of the open cluster Praesepe from deep LBT observations. Astron Astrophys 531:A164. https://doi.org/10.1051/0004-6361/201015598, 1105.5682 Wagner K, Apai D, Kasper M et al (2016) Direct imaging discovery of a Jovian exoplanet within a triple-star system. Science 353:673–678. https://doi.org/10.1126/science.aaf9671, 1607.02525 Weights DJ, Lucas PW, Roche PF, Pinfield DJ, Riddick F (2009) Infrared spectroscopy and analysis of brown dwarf and planetary mass objects in the Orion nebula cluster. Mon Not R Astron Soc 392:817–846. https://doi.org/10.1111/j.1365-2966.2008.14096.x, 0810.3584 Whelan ET, Ray TP, Bacciotti F et al (2005) A resolved outflow of matter from a brown dwarf. Nature 435:652–654. https://doi.org/10.1038/nature03598, astro-ph/0506485 White RJ, Basri G (2003) Very low mass stars and Brown dwarfs in Taurus-Auriga. Astrophys J 582:1109–1122. https://doi.org/10.1086/344673, astro-ph/0209164 Whitworth AP, Stamatellos D (2006) The minimum mass for star formation, and the origin of binary brown dwarfs. Astron Astrophys 458:817–829. https://doi.org/10.1051/0004-6361:20065806, astro-ph/0610039 Wolk SJ, Walter FM (2000) Very low mass stars and brown dwarfs in the belt of orion. In: Rebolo R Zapatero-Osorio MR (eds) Very low-mass stars and Brown dwarfs, Cambridge University Press, p 38 Zakhozhay OV, Zapatero Osorio MR, Béjar VJS, Boehler Y (2017) Spectral energy distribution simulations of a possible ring structure around the young, red brown dwarf G 196-3 B. Mon Not R Astron Soc 464:1108–1118. https://doi.org/10.1093/mnras/stw2308, 1609.02859

22 Brown Dwarfs and Free-Floating Planets in Young Stellar Clusters

501

Zapatero Osorio MR, Rebolo R, Martín EL (1997a) Brown dwarfs in the Pleiades cluster: a CCDbased R, I survey. Astron Astrophys 317:164–170, astro-ph/9604079 Zapatero Osorio MR, Rebolo R, Martín EL et al (1997b) New Brown dwarfs in the Pleiades cluster. Astrophys J 491:L81–L84. https://doi.org/10.1086/311073, astro-ph/9710300 Zapatero Osorio MR, Rebolo R, Martín EL et al (1999) Brown dwarfs in the Pleiades cluster. III. A deep IZ survey. A&AS 134:537–543. https://doi.org/10.1051/aas:1999443, astro-ph/9810051 Zapatero Osorio MR, Béjar VJS, Martín EL et al (2000) Discovery of young, isolated planetary mass objects in the ¢ Orionis star cluster. Science 290:103–107. https://doi.org/10.1126/science.290.5489.103 Zapatero Osorio MR, Béjar VJS, Martín EL, Barradoy Navascués D, Rebolo R (2002a) Activity at the deuterium-burning mass limit in Orion. Astrophys J 569:L99–L102. https://doi.org/10.1086/340690, astro-ph/0203283 Zapatero Osorio MR, Béjar VJS, Martín EL et al (2002b) A methane, isolated, planetary-mass object in Orion. Astrophys J 578:536–542. https://doi.org/10.1086/342474, astro-ph/0206353 Zapatero Osorio MR, Béjar VJS, Pavlenko Y et al (2002c) Lithium and H’ in stars and brown dwarfs of sigma Orionis. Astron Astrophys 384:937–953. https://doi.org/10.1051/0004-6361:20020046, astro- ph/0202147 Zapatero Osorio MR, Béjar VJS, Martín EL et al (2014a) Spectroscopic follow-up of Land T-type proper-motion member candidates in the Pleiades. Astron Astrophys 572:A67. https://doi.org/10.1051/0004-6361/201424634, 1410.2383 Zapatero Osorio MR, Béjar VJS, Miles-Páez PA et al (2014b) Trigonometric parallaxes of young field L dwarfs. Astron Astrophys 568:A6. https://doi.org/10.1051/0004-6361/201321340, 1406.1345 Zapatero Osorio MR, Gálvez Ortiz MC, Bihain G et al (2014c) Search for freefloating planetary- mass objects in the Pleiades. Astron Astrophys 568:A77. https://doi.org/10.1051/0004-6361/201423848, 1407.2849 Zapatero Osorio MR, Béjar VJS, Peña-Ramírez K (2017) Optical and nearinfrared spectra of ¢ Orionis isolated planetary-mass objects. Astrophys J 842:65. https://doi.org/10.3847/1538-4357/aa70ec, 1705.01336 Zuckerman B, Becklin EE (1987) A search for brown dwarfs and late M dwarfs in the Hyades and the Pleiades. Astrophys J 319:L99–L102. https://doi.org/10.1086/184962

Large-Scale Searches for Brown Dwarfs and Free-Floating Planets

23

Ben Burningham

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Significant Large Area Surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Deep Near-Infrared Survey of the Southern Sky (DENIS) . . . . . . . . . . . . . . . . . . . . . . The Two Micron All-Sky Survey (2MASS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Sloan Digital Sky Survey (SDSS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The UKIRT Infrared Deep Sky Survey (UKIDSS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Canada-France Brown Dwarf Survey CFBDS(IR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Wide-field Infrared Survey Explorer (WISE) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Visible and Infrared Survey Telescope for Astronomy (VISTA) . . . . . . . . . . . . . . . . . The Panoramic Survey Telescope and Rapid Response System (Pan-STARRS) . . . . . . . . Science Drivers for Large-Scale Brown Dwarf Searches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Initial Mass Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Extending the Spectral Sequence to Ever Lower Temperatures . . . . . . . . . . . . . . . . . . . . . . The Bottom of the IMF: Planetary Mass Objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Near Future . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross-References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

504 509 509 509 509 511 512 512 514 515 515 515 518 519 521 521 522

Abstract

Searches of large-scale surveys have resulted in the discovery of over 1000 brown dwarfs in the Solar neighborhood. In this chapter we review the progress in finding brown dwarfs in large datasets, highlighting the key science goals and summarizing the surveys that have contributed most significantly to the current sample.

B. Burningham () Centre for Astrophysics Research, School of Physics, Astronomy and Mathematics, University of Hertfordshire, Hatfield, UK e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_118

503

504

B. Burningham

Introduction The first two confirmed brown dwarf discoveries were published in 1995 following many years of intense searching (Nakajima et al. 1995; Rebolo et al. 1995). One of these, G229B, was found via high-contrast imaging as a faint companion to a nearby M dwarf (Nakajima et al. 1995). The other, Tiede 1, was found from a deep imaging search (with follow-up spectroscopy) of the Pleiades cluster (Rebolo et al. 1995). While both these discoveries represent the result of a large effort to identify brown dwarfs, their discovery routes turned out to be relatively minor contributors to the growing catalogue of brown dwarfs. In the following 20 years, the main discovery route for brown dwarfs was via large-scale surveys. That the first discoveries did not arise from large-scale sky surveys reflects the faintness of the targets in comparison to the depths of photographic surveys available in the early 1990s and the difficulty of identifying targets with hitherto unknown photometric properties from large catalogues. However, despite the challenge of searching for such objects in largescale surveys, it remains the fundamental pathway for characterizing the substellar population. The first wide field searches for brown dwarfs used the all-sky photographic surveys carried out during the second half of the twentieth century. These searches were limited by the available photometric bands (typically BRI ) and initial ignorance of the spectral energy distributions of brown dwarfs in the Solar neighborhood. In addition, the depth of these surveys was insufficient to detect more than a handful of the nearest targets. Large-scale discovery of brown dwarfs would need to wait for the large-scale digital surveys with sensitivity in the near infrared that came online at the turn of the twenty-first century. The discovery methods largely followed those applied to finding late-type M dwarfs in photographic surveys, which fell into two categories: searches that relied on the motion of nearby and faint objects and searches that distinguish the targets via their photometric colors. Of these, the dominant search route for brown dwarfs from wide field surveys has been photometric selection. The efficacy of the method depends on the difference between the spectral morphology of the targets of interest and the background population. This allows the definition of a region of color space that effectively isolates the target population. In practice, the separation is rarely perfect, and photometric selections are typically contaminated with one or more type of interloper. As a result, spectroscopic confirmation is usually required to define reliable samples. Searches for brown dwarfs in large-scale surveys followed on from the ongoing process of extending the stellar spectral sequence to ever lower temperatures and later spectral types. Spectral types of M7 and later are collectively known as ultracool dwarfs (UCDs). As such, all but the youngest brown dwarfs are also UCDs. The search for brown dwarfs in large-scale surveys is thus also the search for UCDs, and it both follows and drives the definition the UCD spectral sequence. Rapid progress was made extending the UCD sequence in the late 1990s and early 2000s, resulting in the definition of the L and T spectral classes (Kirkpatrick et al.

23 Large-Scale Searches for Brown Dwarfs and Free-Floating Planets

505

1999; Burgasser et al. 2006). A detailed review of these spectral classes is beyond the scope of this chapter, but may be found in Kirkpatrick (2005). Extending the spectral sequence beyond the T sequence proved impossible using ground-based surveys and will be discussed later in this chapter and in detail elsewhere in this Handbook. Figure 1 shows the spectral sequence from M7 to T6. The basis for key color selection criteria is immediately apparent. Both L and T dwarfs may be distinguished from M dwarfs via the red slope of their spectra as we move from optical to the near-infrared J band. Color cuts such as i 0  z0 & 1:8 colors or a z0  J & 2:5 cut are often used as initial selectors for LT dwarfs. Often such selections require the combination of optical with near-infrared surveys, which may or may not have complementary depths. The faintness of the targets in the submicron region means that optical surveys often probe significantly smaller volumes for LT dwarfs than their near-infrared contemporaries. In these cases, the full depth of a near-infrared survey can only be searched by allowing non-detections in the optical survey to place limits on the optical to near-infrared colors. Such search methods are known as “dropout” methods: candidates are required to be undetected at certain wavelengths. Typical contaminants in such selections are late-M dwarfs that have been scattered into the selection by photometric error. Such contaminants can be weeded out of searches for T dwarfs via further selection based on blue J  H colors. Alternatively, selecting mid- to late-L dwarfs is facilitated by requiring very red J  H colors. Selections targeting LT transition objects and early-T dwarfs suffer the greatest contamination due to J  H  0 colors that overlap with those of M dwarfs. As a result they have been generally underrepresented in the samples selected from near-infrared surveys. The nature of the photometric selection methods for LT dwarfs means that, with very few exceptions (e.g., Folkes et al. 2012), searches for these objects have been focused away from the Galactic plane. This is for two main reasons. Firstly, matching survey catalogues of different wavelengths and epochs is extremely problematic in the crowded fields when the targets of interest often have large proper motions. Secondly, selecting targets based on red optical to near-infrared color is prone to significant contamination from reddened background stars. As a result, sight lines near the Galactic plane represent the greatest source of incompleteness in the census of UCDs in the Solar neighborhood. Most of the dedicated large-scale searches for brown dwarfs have been led by teams within, or closely associated with, survey science teams. The details of the survey design, which may or may not have been devised with brown dwarf science in mind, drive the search strategies that these teams employ. While discovery science has undoubtedly been pursued by scientists beyond the survey consortia, the realities of winning telescope time for extensive follow-up mean that the head start given to the survey teams has often been decisive. Consequently, discoveries of brown dwarfs are commonly tied to one particular survey, even though multiple survey datasets are typically used to select candidates from the many millions of detected sources.

506

B. Burningham

Fig. 1 Near-infrared spectra of M7–T6 spectral standards from the SpeX Prism Library (Burgasser 2014). Key absorption features are indicated, and the approximate bandpasses for commonly used photometric filters are shaded gray

23 Large-Scale Searches for Brown Dwarfs and Free-Floating Planets

507

Table 1 summarizes the tally of L and T dwarf discoveries from wide field surveys over the past two decades, along with an estimate total number of detectable LT dwarfs if the full survey depth and area was exploited in each case. Each of these surveys is discussed in detail later in this chapter. What is immediately apparent is that the numbers of spectroscopically confirmed LT dwarfs fall well short of the maximum detectable number. This is for several reasons. Firstly, even low-resolution spectroscopic confirmation of LT dwarfs requires significant resources in the form 4 m-class telescope time for targets with J . 17:0 or 8 m-class telescope time for fainter targets. This means that the number of confirmed LT dwarfs in each survey often depends on how its volume is accessed: via wider coverage or deeper imaging over smaller areas. For example, the UKIDSS Large Area Survey (LAS) has confirmed fewer L dwarfs than 2MASS, despite probing nearly ten times the volume. This reflects the fact that the extra volume probed by the UKIDSS LAS was at much fainter magnitudes. As a result, spectroscopic follow-up was largely restricted to 8 m-class facilities and focused on tightly defined science goals such as understanding the historic birthrate of brown dwarfs (e.g., Day-Jones et al. 2013; Marocco et al. 2015) or constraining the substellar initial mass function (e.g., Burningham et al. 2013). The changing nature of the science that drives the search for brown dwarfs in large surveys is another big factor. Current community effort appears to be focused on exploring the extremes of brown dwarf parameter space such as low temperatures (e.g., Luhman 2014; Skemer et al. 2016), young ages and planetary masses (e.g., Gagné et al. 2017; Faherty et al. 2013), and low metallicity (e.g., Lodieu et al. 2010; Mace et al. 2013b). Studying such extremes typically involves detailed follow-up of particularly interesting targets. The sheer expense of detailed characterization of very faint brown dwarfs precludes their study unless faint targets are only the examples of their kind, as in the case of the coolest brown dwarfs. This limits the exploitation of deeper surveys, so all-sky surveys continue to provide the dominant resource for ongoing studies of brown dwarfs in the Solar neighborhood. Although significant questions remain regarding the properties of the brown dwarf population, particularly in a Galactic context, there is currently little appetite within the community for addressing these questions via large-scale spectroscopic follow-up of faint brown dwarfs selected from current or future surveys. The work of Skrzypek et al. (2015, 2016) is worth noting here. They combined eight filter bandpasses (i zYJHK W 1W 2) across three surveys (SDSS, UKIDSS, and WISE) to estimate photometric spectral types for some 1361 LT dwarfs. Limited spectroscopic follow-up suggests that their method is sound and their photo-typing method achieves ˙1 subtype accuracy across the LT range. Such approaches provide the opportunity to extend current studies of brown dwarfs to much larger samples and likely represent a key future methodology for large-scale searches for brown dwarfs. In the remainder of this chapter, we summarize the details of the surveys which contributed most significantly to the current brown dwarf census and highlight some of the key science results of large-scale searches for brown dwarfs.

Survey DENIS 2MASS SDSS (I & II) UKIDSS LAS CFDBS(IR) WISE

Bands iJH JHKs ugri z YJHK i z.J / W 1W 2W 3W 4

Area/deg2 20,000 All sky 15,000 3600 1000 (355) All sky Depth J D 16:5 J D 16:5 z0 D 20:5 J D 19:6 z0AB D 24:0I J D 20:0 W 2 D 15:6

Npub (L) 49 403 381 142 170 10

Npub (T) 1 57 55 263 45 176

Refs 1–7 8–39 40–50 50–59 60–64 64–67

Ndet (T) 56 110 1100 180 1200

Ndet (L) 1400 2800 22,000 3800 19,000

Table 1 The numbers of published L and T dwarfs by survey, discovery references, and the potential numbers detectable based on the surveys’ depths and current estimates of the L and T dwarf space densities. Space densities were compiled from data in Cruz et al. (2007), Day-Jones et al. (2013), and Kirkpatrick et al. (2012). SDSS potential yields are not projected due to poor availability of mean z0 magnitudes for LT dwarfs. The CFBDS(IR) yields are based on the region with J band overlap. References: (1) Delfosse et al. (1997); (2) Martín et al. (1999); (3) Martín et al. (2010); (4) Bouy et al. (2003); (5) Kendall et al. (2004); (6) Phan-Bao et al. (2008); (7) Artigau et al. (2010); (8) Kirkpatrick et al. (1999); (9)Kirkpatrick et al. (2000); (10) Kirkpatrick et al. (2008); (11) Kirkpatrick et al. (2010a); (12) Burgasser et al. (1999); (13) Burgasser et al. (2000); (14) Burgasser et al. (2002); (15) Burgasser et al. (2003a); (16) Burgasser et al. (2003b); (17) Burgasser et al. (2003c); (18) Burgasser et al. (2004); (19) Burgasser (2004a); (20) Kirkpatrick et al. (2010b); (21) Reid et al. (2008); (22) Gizis (2002); (23) Gizis et al. (2000); (24) Gizis et al. (2003); (25) Kendall et al. (2003); (26) Kendall et al. (2007); (28) Cruz et al. (2003); (29) Cruz et al. (2004); (30) Cruz et al. (2007); (31) Wilson et al. (2003); (32) Folkes et al. (2007); (33) Metchev et al. (2008); (34) Looper et al. (2007); (35) Looper et al. (2008); (36) Sheppard and Cushing (2009); (37) Scholz et al. (2009); (38) Geißler et al. (2011); (39) Tinney et al. (2005); (40) Fan et al. (2000); (41) Hawley et al. (2002); (42) Geballe et al. (2002); (43) Schneider et al. (2002); (44) Knapp et al. (2004); (45) Chiu et al. (2006); (46) Zhang et al. (2009); (47) Scholz et al. (2009); (48) Schmidt et al. (2010); (49) Leggett et al. (2000); (50) Lodieu et al. (2007b); (51) Pinfield et al. (2008); (52) Burningham et al. (2008); (53) Burningham et al. (2009); (54) Burningham et al. (2010a); (55) Burningham et al. (2010b); (56) Burningham et al. (2013); (57) Cardoso et al. (2015); (58) Day-Jones et al. (2013); (59) Marocco et al. (2015); (60) Delorme et al. (2008b) (61) Reylé et al. (2010); (62) Delorme et al. (2010); (63) Albert et al. (2011); (64) Kirkpatrick et al. (2011); (65) Kirkpatrick et al. (2012); (66) Mace et al. (2013a); (67) Pinfield et al. (2014); (68) Lodieu et al. (2012)

508 B. Burningham

23 Large-Scale Searches for Brown Dwarfs and Free-Floating Planets

509

Significant Large Area Surveys The Deep Near-Infrared Survey of the Southern Sky (DENIS) Carried out between 1996 and 2001, the 1 m-ESO telescope at La Silla (Chile), the Deep Near-Infrared Survey of the Southern Sky (DENIS), was one of the first substantial surveys in the near infrared. It covered 20,000 square degrees in I (  0:8 m), J (  1:25 m), and Ks  2:15 m) photometric bandpasses (Epchtein et al. 1997) and was a significant contributor to the early progress of brown dwarf science in the Solar neighborhood. Its choice of filters provided leverage on the extremely red 0:8  1:2 m SED of L and T dwarfs, and so brown dwarfs could be directly selected from the survey catalogue without the need to cross-reference other surveys sensitive to other wavelengths. This resulted in the discovery of 50 brown dwarfs (see Table 1).

The Two Micron All-Sky Survey (2MASS) The Two Micron All-Sky Survey (2MASS ; Skrutskie et al. 2006) was the first, and to date only, all-sky survey covering the 1  2:5 m near-infrared region. A transformational contribution to the study of low-mass stars and brown dwarfs, it provided discovery images for over 400 L dwarfs and nearly 60 T dwarfs (Table 1) and continues to feature as a key dataset in many ongoing studies of brown dwarfs in the Solar neighborhood. The survey was completed between 1997 and 2001 using one 1.3 m telescope in each hemisphere: one at the Fred Lawrence Whipple Observatory, on Mount Hopkins, Arizona (USA), and one at the Cerro Tololo InterAmerican Observatory (Chile). The survey imaged the whole sky in three filters: J (1:235 m), H (1:662 m), and Ks (2:159 m). Alone, the wave bands covered by 2MASS would not allow photometric selection of brown dwarfs against a background of stars with similar JHKs colors. However, its depth in JHKs was well matched to the depth of the various photographic surveys that were digitized during the 1990s, allowing the selection of candidate brown dwarfs using the dropout technique. The principal search strategy for brown dwarfs in 2MASS was photometric and relied on the red optical-to-NIR colors of L and T dwarfs, combined with their respective red and blue J Ks colors. Initial candidate selections required non-detections in the red photographic plates, which combined with 2MASS detection limits to set a color limit of R  Ks & 5:5 (e.g., Kirkpatrick et al. 1999). The NIR colors of candidates were then used to distinguish T dwarfs and L dwarfs. The many discoveries made in 2MASS are summarized and referenced in Table 1.

The Sloan Digital Sky Survey (SDSS) The SDSS is widely regarded as being the most successful astronomical survey of all time. The SDSS really represents a set survey that continues through the

510

B. Burningham

time of printing, targeting diverse science goals. Here, we consider the original SDSS obtained as part of SDSS I and II, now known as the SDSS Legacy Survey. Data collection started in 2000 (York et al. 2000), and the final data release took place in 2011 (Adelman-McCarthy et al. 2011), incorporating data taken as part of the original SDSS observing plan up until 2008. The facility has diversified its survey portfolio since 2008, pursuing spectroscopic surveys targeting wide-ranging science goals including exoplanets, Galactic structure, and the large-scale structure of the universe across optical to near-infrared wavelengths (e.g., Eisenstein et al. 2011; Blanton et al. 2017). The SDSS Legacy Survey consisted of two principal components: a photometric survey and a spectroscopic survey. Both were obtained using the 2.5-m wide-angle optical telescope at Apache Point Observatory in New Mexico (USA). The SDSS photometric survey imaged some 8,000 square degrees of sky in ugri z filters. The survey region targeted the northern Galactic cap and three stripes in covering the southern Galactic cap. This strategy minimized contamination from Milky Way foreground gas, dust, and stars that would interfere with the survey’s principal goal of constructing a three-dimensional map of the distribution of galaxies. This strategy did not hinder searches for brown dwarfs in the Solar neighborhood, which would generally avoid the Galactic plane in any case. The spectroscopic survey was predominantly targeted at determining redshifts for galaxies, and it obtained spectra of some 1.8 million targets Brown dwarfs were discovered within the SDSS photometric catalogues from the outset, with seven L dwarfs and two T dwarfs identified in commissioning data (Strauss et al. 1999; Tsvetanov et al. 2000; Fan et al. 2000). Candidates were selected via .i 0  z0 / vs .r 0  i 0 / color-color diagrams and via r 0 and i 0 band dropout searches. Searches of SDSS photometric catalogues were also complemented with 2MASS photometry to further constrain the colors of the targets (e.g., Chiu et al. 2006). SDSS provided the discovery data for some 381 L dwarfs and 55 T dwarfs to date (see Table 1). In addition to photometric selections of candidates, the work carried out by Schmidt et al. (2010) is of note for selecting L dwarfs based on their spectra, rather than broadband colors. This unique selection method is largely free of color biases that can be introduced by photometric methods and was made possible thanks to the spectroscopic survey carried out as part of the SDSS. Although the vast majority of targets within the SDSS spectroscopic survey were extragalactic in nature, approximately 5% of its spectra were of objects with late-M spectral type and cooler (Schmidt et al. 2010). The resulting sample of spectroscopically selected L dwarfs had a median J  Ks color 0.1 magnitudes bluer than previous photometric selections for spectral types on the L0–L4 range. This example highlights how color-based selections can introduce bias, particularly when color cuts are aimed at distinguishing objects in spectral-type transition regions.

23 Large-Scale Searches for Brown Dwarfs and Free-Floating Planets

511

The UKIRT Infrared Deep Sky Survey (UKIDSS) The UKIRT Infrared Deep Sky Survey (UKIDSS) was carried out using the purpose built Wide Field CAMera (WFCAM; Casali et al. 2007) between 2005 and 2012. UKIDSS employed what is commonly known as a wedding cake strategy: the survey as a whole consists of a set of sub-surveys of decreasing sky area and increasing depth (Lawrence et al. 2007). The original plan for UKIDSS aimed to image some 7500 square degrees, but the largest planned area for single sub-survey was just over half that area. This approach is seen frequently in modern sky surveys which tension various science goals with differing requirements in terms of depth and coverage against finite observing time on dedicated survey instruments. In order of sky coverage, the surveys that comprise UKIDSS are as follows: 1. The Large Area Survey (LAS): 3700 sq. degs. in YJHK principally covering overlap sky with SDSS outside the Galactic plane to a typical 5 depth of K < 18:4 2. The Galactic Plane Survey (GPS): 1800 sq. degs. at JHK covering the Galactic plane within b ˙ 5 deg to K < 19:0 3. The Galactic Clusters Survey (GCS): 1400 sq. degs. covering 10 star clusters in JHK to a depth of K < 18:7) 4. The Deep Extragalactic Survey (DXS): 35 sq. degs. in JHK to K < 21 5. The Ultra Deep Survey (UDS): 0.8 sq. degs. in JHK to K < 25:3 The most prolific of the UKIDSS for brown dwarf detections was the LAS. The top two LAS headline science goals were discovering the coolest brown dwarfs in the Solar neighborhood and identifying the highest redshift quasars (z > 6). Both of these target populations are heavily contaminated by Galactic M dwarfs in NIR color-color diagrams, so the bulk of the LAS footprint was placed to coincide with the SDSS. This allowed the reddest bands of the SDSS to be used to exclude the populous M dwarfs to nearly the full depth of the J band survey. To further aid in the photometric selection of its key science targets, the LAS was the first wide field survey to employ the MKO Y band filter, centered at 1:02 m. This filter was designed to allow effective discrimination between highredshift quasars and T(+) dwarfs, which otherwise share similar colors in JHK. This discrimination relied on the fact that the L and T dwarfs discovered up to this time had Y  J > 1:0, with cooler brown dwarfs expected to be even redder, while high-redshift quasars were expected to remain bluer than this limit up to z  7 (Warren and Hewett 2002). Early searches of the LAS for extremely cool T dwarfs, and the preemptively classified Y dwarfs, were guided by this expectation. However, the first brown dwarf to be identified with spectral type later than T8, ULAS J00340052, was excluded from early searches and was instead identified as part of a search for quasars, displaying Y  J D 0:75 ˙ 0:1.

512

B. Burningham

As the survey progressed it became clear that late-T dwarf colors frequently overlapped with those of high-redshift quasars, with the latest type objects generally found with Y  J < 1 (e.g., Burningham et al. 2010b, 2013). This trend to bluer Y  J colors is now understood in terms of rainout chemistry removing K I from the gas phase in the coldest objects and weakening the strong, pressure-broadened, potassium line that dominates the T dwarf spectral morphology in the Y band (e.g., Line et al. 2017). The small number of T7+ dwarfs discovered in previous surveys did not initially show this trend, and the experience serves to illustrate the potential pitfalls of using previously justified photometric selection criteria to explore new parameter space for ultracool dwarfs. The final photometric selection methods for the principal UKIDSS late-T dwarf follow-up program are outlined in detail in Burningham et al. (2013), which also provides copies of the SQL queries used to perform the selections in the WFCAM Science Archive (WSA). These selections employed a relatively weak Y  J > 0:5 requirement. This weak criterion was necessary to avoid excluding late-T dwarfs with blue Y  J colors, but let many M dwarfs pass the selection. A J  H < 0:1 cut removed the bulk of L and M dwarfs, while a final z0  J > 2:5 excluded further M dwarfs. For the bulk of the volume searched, this final cut was achieved by requiring candidates to be undetected in the SDSS. This search confirmed some 200 T dwarfs, making it the most prolific source of confirmed T dwarfs to date.

Canada-France Brown Dwarf Survey CFBDS(IR) The Canada-France Brown Dwarf Survey (CFBDS; Delorme et al. 2008b) covered some 1000 square degrees in i z filters, with J band coverage in 355 square degrees. The i z survey drew data from two existing surveys carried out using MegaCam on the Canada-France-Hawaii Telescope (CFHT) on Mauna Kea, Hawaii (USA): the CFHT Legacy Survey (CFHTLS; Cuillandre and Bertin 2006) and the Red Sequence Cluster Survey (RCS-2; Yee et al. 2007). Candidate brown dwarfs were selected on the basis of red i  z colors and followed up with J band imaging. Despite its modest coverage this survey made a significant contribution to determining the local space density of brown dwarfs (Reylé et al. 2010) and discovering extremely cool T dwarfs (Delorme et al. 2008a).

The Wide-field Infrared Survey Explorer (WISE) Launched in 2009, the WISE mission surveyed the entire sky in four wave bands that are largely inaccessible from the ground, centered at 3.4 m (W1), 4.6 m (W2), 12 m (W3), and 22 m (W4) (Wright et al. 2010). In many ways it can be viewed as a successor to 2MASS in terms of brown dwarf science. Early exploitation and follow-up for brown dwarf science was led by the same team at the Infrared Processing and Analysis Center in Pasadena that pursued early brown

23 Large-Scale Searches for Brown Dwarfs and Free-Floating Planets

513

dwarf science with 2MASS. As with 2MASS, early science drivers were concerned with completing a census of brown dwarfs in the Solar neighborhood and defining a new spectral type, in this case the Y spectral type. The initial survey design was based around a cryogenic mission lifetime of 6 months, during which the full sky was imaged in all four passbands. The cryogen lasted after the initial pass was completed, allowing 20% of the sky to be imaged a second time in all four bands. The shortest two wave bands (W1 & W2) do not require additional active cooling of the spacecraft by venting cryogen, and post-cryogen operations were planned to take advantage of this. Following the exhaustion of cryogen, the mission was renamed NEOWISE (for Near-Earth Object WISE) and completed the second pass of the whole sky in W1 and W2 with the aim of detecting potentially hazardous NEOs (Mainzer et al. 2011). The point source catalogue for the full cryogenic mission was released as the WISE all-sky data release in 2012. In 2013, data from the cryogenic mission and subsequent post-cryogenic NEOWISE mission were published as the ALLWISE data release. The WISE spacecraft performed the survey by scanning the sky as it orbited the Earth above the terminator region, building up depth via multiple passes over each region. The regions near the ecliptic poles thus received the greatest number of images (1000s of images at the poles), whereas regions on the ecliptic place typically received 12 to 13 passes in the original cryogenic mission. Moon avoidance led to some regions receiving considerably fewer passes. The WISE all-sky W 1 and W 2 depths probe a similar volume for L dwarfs to that probed by 2MASS; however as one moves to cooler temperatures, the probed volume soon overtakes that probed by 2MASS for T dwarfs. As a result, the brown dwarf discoveries by WISE are dominated by objects with late-T type and beyond (e.g., Kirkpatrick et al. 2011; Mace et al. 2013a). Although its probed volume for late-T dwarfs is similar to that of the UKIDSS LAS, WISE provides a more accessible sample due to its all-sky coverage giving a greater volume at smaller distances. Nonetheless, the comparable discovery number of T dwarfs in UKIDSS LAS and WISE largely reflects shifting science priorities, and the lack of clear motivation for spectroscopic follow-up of large numbers of faint mid- to late-T dwarfs. The early searches of the WISE dataset were also successful in discovering the long sought Y dwarfs (Cushing et al. 2011). The presence of multiple passes and two full sky surveys notably facilitated NEO detection, but also allows for efficient searches for high proper motion objects beyond the Solar System. Of particular note are the efforts of Kevin Luhman at the Pennsylvania State University Center for Exoplanets and Habitable Worlds, whose proper motion searches using WISE’s multi-epoch imaging have resulted in the discovery of some of the Sun’s closest substellar neighbors (Luhman 2013, 2014). An L7.5 + T0.5 binary (Burgasser et al. 2013) at a distance of less than 2 pc (Sahlmann and Lazorenko 2015), Luhman 16AB was visible in previous surveys DSS, 2MASS, and DENIS but was not identified as a nearby brown dwarf system due to confusion with other nearby sources. By contrast, WISE J085510.83071442.5, at a distance of 2.2 pc (Luhman 2014; Luhman and Esplin 2016), lacked a

514

B. Burningham

ground-based detection for nearly 2 years following its discovery (e.g., Skemer et al. 2016). This reflects the tiny flux emitted at wavelengths easily accessible from the ground by this extremely cool brown dwarf, with an estimated Teff  200  250 K (Schneider et al. 2016). Whereas Luhman’s strategy relied on fairly cumbersome searching the singleexposure catalogues, followed by labor-intensive vetting of candidates by eye, the ALLWISE catalogue provides a more convenient route to leveraging the WISE extended dataset to search for high motion objects. The ALLWISE catalogue forms the basis for a number of motion-based surveys for brown dwarfs in the Solar neighborhood (e.g., Kirkpatrick et al. 2014). In December 2013, the WISE spacecraft was reactivated and NEOWISE survey operations continued through Spring 2018. The increasing depth and motion data raise the likelihood of further exciting discoveries from the WISE spacecraft. Ongoing efforts to mine these data for brown dwarfs include the citizen science project “Backyard Worlds,” which has found at least one cool brown dwarf to date (Kuchner et al. 2017).

The Visible and Infrared Survey Telescope for Astronomy (VISTA) The European Southern Observatory’s (ESO) Visible and Infrared Survey Telescope for Astronomy (Dalton et al. 2006) has facilitated a number of public surveys, the first tranche of which follow a similar wedding cake design to that seen for the UKIDSS. Three of the public surveys hold particular potential for brown dwarf science: the VISTA Hemisphere Survey (VHS; McMahon et al. 2013), the VISTA Kilo-degree Infrared Galaxy Survey (VIKING; Edge et al. 2013), and the VISTA Variables in the Via Lactea (VVV; Minniti et al. 2010). The first of these has delivered a handful of brown dwarfs found following similar photometric methods to those applied to exploitation of UKIDSS (e.g., Lodieu et al. 2012). One of the reddest known L dwarfs, VHS J1256601.92-125723.9, was identified in the VHS as a companion to a brown dwarf binary with a probable age of 300 Myr (Gauza et al. 2015; Stone et al. 2016). However, although the nearly half-sky coverage of VHS in the near infrared provides excellent opportunities for large-scale searches for brown dwarfs, it initially lacked the complimentary optical coverage that is so important for selecting LT dwarfs. The 1500-square degree ZYJHK VIKING survey provides opportunity to select LT dwarfs in a self-sufficient way, but the limited coverage means that the candidates will be faint and require expensive follow-up. The VVV has provided one of the best opportunities to date to search within the Galactic plane for nearby brown dwarfs. This survey covers the southern Galactic plane in ZYJHKs , but with around 100 epochs in the Ks filter over a 7-year period (Minniti et al. 2010). This multi-epoch survey has been optimized for studying variable stars; however, it also allows astrometric selection of fast-moving and nearby brown dwarfs that are otherwise hard to spot in the crowded Galactic plane (e.g., Beamín et al. 2013).

23 Large-Scale Searches for Brown Dwarfs and Free-Floating Planets

515

The Panoramic Survey Telescope and Rapid Response System (Pan-STARRS) The Panoramic Survey Telescope and Rapid Response System (Pan-STARRS; Kaiser et al. 2010) is a wide-field imaging facility based around a 1.8-m telescope in Haleakala, Maui (USA). The design of the facility is based around rapid scanning of the sky to allow identification of fast-moving sources, transients, and Solar System objects. Several synoptic surveys have been carried out with Pan-STARRS, but the survey of most interest for brown dwarf searches is the Pan-STARRS1 3 survey (PS1: Chambers et al. 2016). Named for its coverage of the three quarters of the sky that it can access, it incorporates around 12 visits in each of its 5 filters (gri zy) over the 4 years it took to complete (2010–2014). It’s typical 5 detection threshold is zps1  22:3. The optical coverage of PS1 has been combined with WISE data to target LT transition objects that have been otherwise difficult to identify in near-infrared surveys, and this is one the most significant contributions that PS1 has made to the census of local LT dwarfs (Best et al. 2015, 2018). The multiple epochs also provide accurate proper motions, and many brown dwarfs have also been discovered as wide common proper motion companions to stars using PS1 (e.g., Deacon et al. 2014, 2017). PS1 is also noteworthy for the discovery of the planetary mass brown dwarf PSO J318.5-22 (Liu et al. 2013).

Science Drivers for Large-Scale Brown Dwarf Searches The science drivers behind large-scale searches for brown dwarfs have evolved over the years, from identifying first examples of new classes of objects to growing statistically useful samples for population studies, targeting outliers and expanding parameter space. Here we explore some of the key science goals that have driven searches for brown dwarfs via large-scale surveys in the past few decades.

The Initial Mass Function Much of the popular excitement surrounding the search for brown dwarfs in the 1990s was thanks to idea that brown dwarfs might account for a significant component of dark matter, under the umbrella of massive compact halo objects (MACHOs). However, there were already good reasons to suspect that brown dwarfs at most accounted for a small proportion of dark matter. This did not significantly reduce the impetus for confirming the existence of the hitherto unseen population, since (for many) the real motivation for this search was an understanding of the star formation process via a full accounting of its products. The initial mass function (IMF; Salpeter 1955) describes the rate of star formation as a function of mass and is often thought of as the distribution of mass in a coeval stellar population as a function of stellar mass. It has long been regarded as

516

B. Burningham

one of the fundamental diagnostics for testing models of star formation, despite the impossibility of measuring it directly over the full stellar and substellar mass range. This measurement is challenging for many reasons. Prime of these is the fact masses are not directly observed, so the IMF is inferred from the luminosity function (LF). Although the mass-luminosity relationship is generally well characterized for main sequence stars in hydrostatic equilibrium, it is not possible to observe coeval populations of main sequence stars over the full mass range: the most massive stars have exploded as supernovae before the lowest-mass stars have reached the main sequence. Determining masses for pre-main sequence (PMS) stars in young coeval populations depends on PMS evolutionary models and correctly determining the age of the population. Both of these hurdles introduce significant uncertainty to the resulting IMF. These issues can be avoided to by considering the field population, as Salpeter did in his seminal paper that first defined the IMF (Salpeter 1955). Instead, one must account for rate of stars evolving off the main sequence and the star formation history of the field population. In the case of the substellar population, which never reaches the main sequence and thus lacks a unique mass-luminosity relationship in a mixed age population, reference must also still be made to evolutionary models to determine object masses. Salpeter was untroubled by this last point: his luminosity function corresponded to a mass range of roughly 0:410Mˇ . His mass distribution was well fit by a power law, .m/ / m˛ , with ˛ D 2:35, which is now known as the Salpeter mass function. The human story behind this first derivation of the IMF is wonderfully described in Salpeter (2005). The essential quality of the Salpeter mass distribution is that the number of stars increases steeply with decreasing mass. Moreover, since ˛ > 2:0, it reflects more mass being sequestered in lower-mass stars than higher-mass stars when integrated over equal logarithmic mass bins. Extrapolating this to ever lower masses leads to the prediction that brown dwarfs (and planetary mass objects) might represent a dominant constituent of baryonic matter in the Galaxy. However, in the following years it became clear that the mass distribution flattened below 1Mˇ , and by the 1990s it was clear that a significant upturn in the substellar regime would be required for brown dwarfs to be numerous enough to account for dark matter (Sandage 1957; Schmidt 1959; Miller and Scalo 1979; Scalo 1986). An in-depth review of progress in constraining the stellar IMF is beyond the scope of this chapter, so the reader is directed to an excellent review by Bastian et al. (2010). Continued studies of the IMF in young clusters, globular clusters, other galaxies and the local field support the idea that the IMF is apparently universal across much of the stellar mass range, regardless of environment (e.g., Scalo 1986; Bastian et al. 2010). There is consensus that the Salpeter IMF holds for masses greater than about 1Mˇ . For masses below 1Mˇ , however, the IMF can be fit by shallower power law (e.g., ˛  1:0  1:3 Reid et al. 2002) or log-normal form for low-mass stars (Chabrier 2003). The first extension of the luminosity function to ultracool temperatures, and substellar masses, was facilitated by the Two Micron All-Sky Survey (2MASS; see Section ; Skrutskie et al. 2006), with key papers by Cruz et al. (2003, 2007) which robustly explored the LF across the M7–L8 spectral-type range. However, studies

23 Large-Scale Searches for Brown Dwarfs and Free-Floating Planets

517

of this spectral-type range had limited value for constraining the IMF. The lack of unique mass-luminosity relationship for brown dwarfs significantly complicates the estimation of the form of IMF in the mixed age field population. Since the age, and hence the mass, for isolated brown dwarfs cannot be reliably determined, estimates for the IMF depend on comparisons of observed LFs or spectral-type distributions to simulations based on different assumed IMFs and historic formation rates. Figure 2 shows simulated Teff distributions under a range of assumed IMF power laws and a log-normal IMF from Burgasser (2004b). The luminosity function in the late-M and early-L spectral-type range shows relatively weak dependence on the form of the underlying mass function. As such, the well-measured space densities in this regime placed weak constraints on the form of the substellar IMF, with Allen et al. (2005) finding the LF consistent with ˛ D 0:0 ˙ 0:5. From Fig. 2 it is clear that the sub-1000 K late-T dwarf Teff distribution carries the greatest potential for constraining the slope of the substellar IMF in the field. Although the 2MASS and SDSS were responsible for defining the T dwarf spectral sequence, they lacked the depth to detect a sufficient number of T6+ dwarfs to constrain the IMF with any statistical power. For this reason, late-T dwarfs were preferentially targeted for large-scale searches as soon as the probed volume allowed for the selection of useful samples. The UKIDSS was the first to achieve this, and

Fig. 2 From Burgasser (2004b, Fig. 5). Simulated Teff distributions for the local field population under different assumed IMF forms for historically constant formation rate

518

B. Burningham

it was found that the space density of T6T8 dwarfs was most consistent with a steeply declining substellar IMF, with ˛ < 0:0 (Pinfield et al. 2008; Burningham et al. 2010b, 2013). This was subsequently confirmed in the WISE census of the Solar neighborhood (Kirkpatrick et al. 2012). This is in contrast to a number of determinations for the substellar IMF in young clusters and associations which have generally found ˛ > 0:0, e.g., Upper Sco, 0.3–0.01Mˇ , ˛ D 0:6 ˙ 0:1 (Lodieu et al. 2007a); Pleiades, 0.48–0.03Mˇ , ˛ D 0:60 ˙ 0:11 (Moraux et al. 2003); ˛ Per, 0.2–0.04Mˇ , ˛ D 0:59 ˙ 0:05 (Barrado y Navascués et al. 2002);  Orionis, 0.5–0.01Mˇ , ˛ D 0:5 ˙ 0:2 (Lodieu et al. 2009); and  Orionis, 0.25–0.004Mˇ , ˛ D 0:6 ˙ 0:2 (Peña Ramírez et al. 2012). The reason for the discrepancy between the substellar IMF estimated in the field and that estimated in young clusters is not clear. Trivial incompleteness in the field studies is an unlikely origin of the discrepancy since the surveys would need to miss more late-T dwarfs than they found to account for the difference. Another possibility is incorrect treatment of the historic substellar formation rate when simulating the IMF, which has generally assumed a flat formation rate (e.g., Burningham et al. 2013). For example, a low historic formation rate might give rise to an underabundance of late-T dwarfs in the Solar neighborhood, despite sharing the young cluster IMF. However, studies of the kinematics of the late-T population suggest that it is of a similar age to the stellar population on the Solar neighborhood (Smith et al. 2013). Similarly, work by Dupuy and Liu (2017) supports the assumption of a relatively flat formation history for brown dwarfs. Another possible cause is some issue with the evolutionary models used to transform between mass and temperature at young ages or over Gyr timescales. Alternatively, the form of the IMF may deviate significantly from a power law or log normal form below the masses probed in young clusters. In that case, simulations of the mixed age field population based on such assumed forms may be expected to disagree with observed space densities in the field due to influence from the mass population below the sensitivity of previous cluster studies.

Extending the Spectral Sequence to Ever Lower Temperatures One of the headline science goals for large area surveys in the period following 2MASS and SDSS was the discovery of objects cooler than the T8 (Teff  700 K) low-temperature extent of the T spectral sequence defined in Burgasser et al. (2006). Speculation was divided over the question as to whether another spectral type would be required beyond the T sequence or if the T dwarfs would be final entry in the stellar spectral classification scheme. Comparisons of the coolest T dwarfs with Saturn and Jupiter suggested that ammonia absorption should be become increasingly important with decreasing temperature and that new features in the Y and J bands might distinguish a new spectral sequence (e.g., Leggett et al. 2007). Another speculated driver for a shift in the spectral sequence was the impact of water clouds condensing at Teff & 400  500 K. Regardless of the ultimate rationale

23 Large-Scale Searches for Brown Dwarfs and Free-Floating Planets

519

for adopting a new scheme, the class beyond the T dwarfs was preemptively named the Y dwarfs (Kirkpatrick et al. 1999). Two surveys in the mid-2000s were targeted at identifying objects beyond the T sequence: UKIDSS and CFBDS(IR). Both surveys were successful at identifying cooler objects than had been found previously, e.g.: CFBDS J005910.90-011401.3 at Teff  620 K (Delorme et al. 2008a) and UGPS J072227.51-054031.2 at Teff  520K (Lucas et al. 2010). However, even at Teff  500 K, the objects’ spectra continued to appear as a continuation of the T spectral sequence. Although the methane bands continued to strengthen with approximately similar relative changes between subtypes, the absolute changes were small and strongly argued for the continuation of the T sequence for these new objects (e.g., Burningham et al. 2008, 2010b; Lucas et al. 2010). More recent analysis has highlighted the measurable impact of ammonia on the near-infrared spectra on objects with spectral types T8 and later (e.g., Line et al. 2015; Canty et al. 2015), but its effect at these temperatures does not cause a qualitative deviation from the T sequence. The launch of the WISE spacecraft provided the necessary sensitivity to identify even cooler objects, which would justify the adoption of a new spectral type. Selected by their brightness at W2 above all else, the Y dwarfs are differentiated from the T sequence in the near infrared by the similar comparative heights of their Y and J band peaks and a narrowing of the J band flux peak (Fig. 3 and Cushing et al. 2011; Kirkpatrick et al. 2012). It is reasonable to note that the differences between the near-infrared spectra of late-T and Y dwarfs show more subtle differences than seen across the LT transition. However, they also display significantly redder J  W 2 colors and much fainter near-infrared magnitudes than late-T dwarfs. These differences suggest that the adoption of a new spectral type is appropriate. A review of the progress in characterizing the Y dwarf population is provided in Chap. 25, “Y Dwarfs: The Challenge of Discovering the Coldest Substellar Population in the Solar Neighborhood”, by S. Leggett. Given that the bulk of their emission escapes at wavelengths that are particularly challenging from the ground, detailed study of the Y dwarfs will have to wait for the successful commissioning of JWST. However, initial characterization based on parallaxes and the available limited near-infrared spectroscopy and multi-wavelength photometry suggest the Y dwarfs have Teff ranging from 500 K down to 300 K (e.g., Leggett et al. 2017). Evolutionary models suggest that at typical thin disk ages of a few Gyr, these temperatures correspond to masses near, and below, the deuterium burning limit (e.g., Baraffe et al. 2003). As such, a significant proportion of the Y dwarf population can also be classified as isolated planetary mass objects.

The Bottom of the IMF: Planetary Mass Objects Finding and studying the lowest-mass brown dwarfs is compelling for a variety of reasons. As we’ve already discussed, determining the form of the IMF and the presence or otherwise of a low-mass cutoff is considered a key observable of the

520

B. Burningham

UGPS 0722−05

8

WISEPC J0148−7202

WISEPC J2056+1459

Normalized fλ + Constant

6 WISEP J0410+1502

WISEP J1541−2250

4

WISEP J1738+2732

2 WISEPC J1405+5534

WISEP J1828+2650

0 1.0

1.5 Wavelength (μm)

2.0

Fig. 3 From Cushing et al. (2011, Fig. 2). Spectra of the first Y dwarfs found in the WISE survey. Note the comparable heights of the flux peaks in the Y and J bands compared to the T9 standard UGPS J0722-05. Also apparent is the narrower J band peak in the Y dwarfs

star formation process. This motivation has driven many searches for planetary mass brown dwarfs. In recent years, however, emphasis has shifted to look at how brown dwarfs can provide insights to understand the atmospheres of giant exoplanets (e.g., Burgasser 2011). Although LT dwarfs span the same temperature range as giant exoplanets, they typically have larger masses, higher gravity, and thus higher pressure photospheres. However, at ages of &500 Myr, planetary mass brown dwarfs occupy a wide range of spectral types on the LT sequence and display similarly low surface gravity to that expected for giant exoplanets. To leverage this shared parameter space, a number of large-scale searches are ongoing to identify planetary mass brown dwarfs as members of young moving groups in the Solar neighborhood (e.g., Aller et al. 2016; Gagné et al. 2015b).

23 Large-Scale Searches for Brown Dwarfs and Free-Floating Planets

521

Planetary mass brown dwarfs have most often been identified initially by anomalously red J  Ks colors (e.g., Faherty et al. 2013; Liu et al. 2013). Spectroscopic signatures of low gravity also highlight young, low-mass brown dwarfs (Allers and Liu 2013). However, careful kinematic characterization is then required to establish their membership of a young moving group to independently constrain their age (e.g., Gagné et al. 2015b). Over 150 low-gravity L dwarfs are now known, and their population property as a red and faint extension of the field L dwarf sequence is well established (Faherty et al. 2016). However, discoveries of T dwarf members of moving groups are few, and their observed spectral properties and colors do not obviously distinguish them from apparently older objects of similar type in the field (Naud et al. 2014; Gagné et al. 2015a). Large-scale kinematic searches are thus necessary to uncover the lowest mass contingent of young associations in the Solar neighborhood.

The Near Future The new generation of large-scale surveys has leveraged advances in optical design and imaging capabilities to achieve rapid coverage of the sky. This has opened the door to wide field synoptic surveys in the past few years such as VISTA VVV, Pan-STARRS, and the under-construction Large Synoptic Survey Telescope (LSST; LSST Science Collaboration et al. 2009). The LSST will survey the entire visible sky from Cerro Pachón (Chile) every few nights. This observing strategy, aimed at discovering transients, will also provide proper motions and, more crucially, parallaxes for all the sources with measurable motions. This will open the door for a new unbiased method for searching for brown dwarfs through parallax selections. By selecting candidates based solely on their parallax and apparent luminosity, biases due to assumptions about color and motion can be avoided. Such searches will likely find numerous brown dwarfs in the Solar neighborhood that have been missed previously in regions such as the Galactic plane. Also of note is the European Space Agency’s Euclid mission (Refregier et al. 2010) which will provide deep optical and YJH imaging along with slitless spectroscopy of the 1 W 1  2 W 0m region with R  250 over large areas of sky. Although targeted at extragalactic science, this mission will also provide the opportunity to study the brown dwarf population on sufficient scale to place them in a Galactic context.

Cross-References  Brown Dwarfs and Free-Floating Planets in Young Stellar Clusters  Spectral Properties of Brown Dwarfs and Unbound Planetary Mass Objects  Y Dwarfs: The Challenge of Discovering the Coldest Substellar Population in the

Solar Neighborhood

522

B. Burningham

References Adelman-McCarthy JK et al (2011) VizieR online data catalog: the SDSS photometric catalog, Release 8 (Adelman-McCarthy+, 2011). VizieR Online Data Catalog 2306 Albert L, Artigau É, Delorme P et al (2011) 37 new T-type brown dwarfs in the Canada-France brown dwarfs survey. AJ 141:203. https://doi.org/10.1088/0004-6256/141/6/203 Allen PR, Koerner DW, Reid IN, Trilling DE (2005) The substellar mass function: a Bayesian approach. ApJ 625:385–397. https://doi.org/10.1086/429548, astro-ph/0502189 Aller KM, Liu MC, Magnier EA et al (2016) Brown dwarfs in young moving groups from Pan-STARRS1. I. AB Doradus. ApJ 821:120. https://doi.org/10.3847/0004-637X/821/2/120, 1604.04284 Allers KN, Liu MC (2013) A near-infrared spectroscopic study of young field ultracool dwarfs. ApJ 772:79. https://doi.org/10.1088/0004-637X/772/2/79, 1305.4418 Artigau É, Radigan J, Folkes S et al (2010) DENIS J081730.0-615520: An overlooked mid-T dwarf in the solar neighborhood. ApJ 718:L38–L42. https://doi.org/10.1088/2041-8205/718/1/ L38, 1006.3577 Baraffe I, Chabrier G, Barman TS, Allard F, Hauschildt PH (2003) Evolutionary models for cool brown dwarfs and extrasolar giant planets. The case of HD 209458. A&A 402:701–712. https:// doi.org/10.1051/0004-6361:20030252, arXiv:astro-ph/0302293 Barrado y Navascués D, Bouvier J, Stauffer JR, Lodieu N, McCaughrean MJ (2002) A substellar mass function for Alpha Persei. A&A 395:813–821. https://doi.org/10.1051/0004-6361: 20021262, arXiv:astro-ph/0209032 Bastian N, Covey KR, Meyer MR (2010) A universal stellar initial mass function? A critical look at variations. ARA&A 48:339–389. https://doi.org/10.1146/annurev-astro-082708101642, 1001.2965 Beamín JC, Minniti D, Gromadzki M et al (2013) One more neighbor: the first brown dwarf in the VVV survey. A&A 557:L8. https://doi.org/10.1051/0004-6361/201322190, 1308.3216 Best WMJ, Liu MC, Magnier EA et al (2015) A search for L/T transition dwarfs with PanSTARRS1 and WISE. II. L/T transition atmospheres and young discoveries. ApJ 814:118. https://doi.org/10.1088/0004-637X/814/2/118, 1612.02824 Best WMJ, Magnier EA, Liu MC et al (2018) Photometry and proper motions of M, L, and T dwarfs from the Pan-STARRS1 3 survey. ApJS 234:1. https://doi.org/10.3847/1538-4365/ aa9982, 1701.00490 Blanton MR, Bershady MA, Abolfathi B et al (2017) Sloan digital sky survey IV: mapping the milky way, nearby galaxies, and the distant universe. AJ 154:28. https://doi.org/10.3847/15383881/aa7567, 1703.00052 Bouy H, Brandner W, Martín EL et al (2003) Multiplicity of nearby free-floating ultracool dwarfs: a hubble space telescope WFPC2 search for companions. AJ 126:1526–1554. https://doi.org/ 10.1086/377343, astro-ph/0305484 Burgasser AJ (2004a) Discovery of a second L subdwarf in the two micron all sky survey. ApJ 614:L73–L76. https://doi.org/10.1086/425418, astro-ph/0409179 Burgasser AJ (2004b) T dwarfs and the substellar mass function. I. Monte carlo simulations. ApJS 155:191–207. https://doi.org/10.1086/424386, arXiv:astro-ph/0407624 Burgasser AJ (2011) The brown Dwarf-exoplanet connection. In: Beaulieu JP, Dieters S, Tinetti G (eds) Molecules in the atmospheres of extrasolar planets. Astronomical society of the Pacific conference series, vol 450, p 113 Burgasser AJ (2014) The SpeX Prism Library: 1000+ low-resolution, near-infrared spectra of ultracool M, L, T and Y dwarfs. In: Astronomical Society of India conference series, vol 11. 1406.4887 Burgasser AJ, Kirkpatrick JD, Brown ME et al (1999) Discovery of four field methane (T-type) dwarfs with the two micron all-sky survey. ApJ 522:L65–L68. https://doi.org/10.1086/312221 Burgasser AJ, Kirkpatrick JD, Cutri RM et al (2000) Discovery of a brown dwarf companion to Gliese 570ABC: a 2MASS T dwarf significantly cooler than Gliese 229B. ApJ 531:L57–L60. https://doi.org/10.1086/312522, arXiv:astro-ph/0001194

23 Large-Scale Searches for Brown Dwarfs and Free-Floating Planets

523

Burgasser AJ, Kirkpatrick JD, Brown ME et al (2002) The spectra of T dwarfs. I. Nearinfrared data and spectral classification. ApJ 564:421–451. https://doi.org/10.1086/324033, astro-ph/0108452 Burgasser AJ, Kirkpatrick JD, Burrows A et al (2003a) The first substellar subdwarf? Discovery of a metal-poor L dwarf with Halo kinematics. ApJ 592:1186–1192. https://doi.org/10.1086/ 375813, arXiv:astro-ph/0304174 Burgasser AJ, Kirkpatrick JD, McElwain MW et al (2003b) The 2Mass wide-field T dwarf search. I. Discovery of a bright T dwarf within 10 Parsecs of the Sun. AJ 125:850–857. https://doi.org/ 10.1086/345975, astro-ph/0211117 Burgasser AJ, McElwain MW, Kirkpatrick JD (2003c) The 2MASS wide-field T dwarf search. II. Discovery of three T dwarfs in the Southern Hemisphere. AJ 126:2487–2494. https://doi.org/ 10.1086/378608, astro-ph/0307374 Burgasser AJ, McElwain MW, Kirkpatrick JD et al (2004) The 2MASS wide-field T dwarf search. III. Seven new T dwarfs and other cool dwarf discoveries. AJ 127:2856–2870. https://doi.org/ 10.1086/383549, arXiv:astro-ph/0402325 Burgasser AJ, Geballe TR, Leggett SK, Kirkpatrick JD, Golimowski DA (2006) A unified nearinfrared spectral classification scheme for T dwarfs. ApJ 637:1067–1093. https://doi.org/10. 1086/498563, arXiv:astro-ph/0510090 Burgasser AJ, Sheppard SS, Luhman KL (2013) Resolved near-infrared spectroscopy of WISE J104915.57-531906.1AB: a flux-reversal binary at the L dwarf/T dwarf transition. ApJ 772:129. https://doi.org/10.1088/0004-637X/772/2/129, 1303.7283 Burningham B, Pinfield DJ, Leggett SK et al (2008) Exploring the substellar temperature regime down to ˜550 K. MNRAS 391:320–333. https://doi.org/10.1111/j.1365-2966.2008.13885.x, 0806.0067 Burningham B, Pinfield DJ, Leggett SK et al (2009) The discovery of an M4+T8.5 binary system. MNRAS 395:1237–1248. https://doi.org/10.1111/j.1365-2966.2009.14620.x, 0902.1812 Burningham B, Leggett SK, Lucas PW et al (2010a) The discovery of a very cool binary system. MNRAS 404:1952–1961. https://doi.org/10.1111/j.1365-2966.2010.16411.x, 1001.4393 Burningham B, Pinfield DJ, Lucas PW et al (2010b) 47 new T dwarfs from the UKIDSS large area survey. MNRAS 406:1885–1906. https://doi.org/10.1111/j.1365-2966.2010.16800.x, 1004.1912 Burningham B, Cardoso CV, Smith L et al (2013) 76 T dwarfs from the UKIDSS LAS: benchmarks, kinematics and an updated space density. MNRAS 433:457–497. https://doi.org/ 10.1093/mnras/stt740, 1304.7246 Canty JI, Lucas PW, Yurchenko SN et al (2015) Methane and ammonia in the near-infrared spectra of late-T dwarfs. MNRAS 450:454–480. https://doi.org/10.1093/mnras/stv586, 1503.04715 Cardoso CV, Burningham B, Smart RL et al (2015) 49 new T dwarfs identified using methane imaging. MNRAS 450:2486–2499. https://doi.org/10.1093/mnras/stv380, 1502.06503 Casali M, Adamson A, Alves de Oliveira C et al (2007) The UKIRT wide-field camera. A&A 467:777–784. https://doi.org/10.1051/0004-6361:20066514 Chabrier G (2003) Galactic stellar and substellar initial mass function. PASP 115:763–795. https:// doi.org/10.1086/376392, astro-ph/0304382 Chambers KC, Magnier EA, Metcalfe N et al (2016) The Pan-STARRS1 Surveys. ArXiv e-prints 1612.05560 Chiu K, Fan X, Leggett SK et al (2006) Seventy-one new L and T dwarfs from the Sloan digital sky survey. AJ 131:2722–2736. https://doi.org/10.1086/501431, arXiv:astro-ph/0601089 Cruz KL, Reid IN, Liebert J, Kirkpatrick JD, Lowrance PJ (2003) Meeting the cool neighbors. V. A 2MASS-selected sample of ultracool dwarfs. AJ 126:2421–2448. https://doi.org/10.1086/ 378607, astro-ph/0307429 Cruz KL, Burgasser AJ, Reid IN, Liebert J (2004) 2MASS J05185995-2828372: discovery of an unresolved L/T binary. ApJ 604:L61–L64. https://doi.org/10.1086/383415, astro-ph/0402172 Cruz KL, Reid IN, Kirkpatrick JD et al (2007) Meeting the cool neighbors. IX. The luminosity function of M7-L8 ultracool dwarfs in the field. AJ 133:439–467. https://doi.org/10.1086/ 510132, arXiv:astro-ph/0609648

524

B. Burningham

Cuillandre JC, Bertin E (2006) CFHT legacy survey (CFHTLS): a rich data set. In: Barret D, Casoli F, Lagache G, Lecavelier A, Pagani L (eds) SF2A-2006: semaine de l’Astrophysique Francaise. p 265 Cushing MC, Kirkpatrick JD, Gelino CR et al (2011) The discovery of Y dwarfs using data from the wide-field infrared survey explorer (WISE). ApJ 743:50. https://doi.org/10.1088/0004637X/743/1/50, 1108.4678 Dalton GB, Caldwell M, Ward AK et al (2006) The VISTA infrared camera. In: Society of photo-optical instrumentation engineers (SPIE) conference series, vol 6269. https://doi.org/10. 1117/12.670018 Day-Jones AC, Marocco F, Pinfield DJ et al (2013) The sub-stellar birth rate from UKIDSS. MNRAS 430:1171–1187. https://doi.org/10.1093/mnras/sts685, 1301.4996 Deacon NR, Liu MC, Magnier EA et al (2014) Wide, cool and ultracool companions to nearby stars from Pan-STARRS1. ArXiv e-prints 1407.2938 Deacon NR, Magnier EA, Liu MC et al (2017) 2MASS 0213+3648 C: a wide T3 benchmark companion to an active, old M dwarf binary. MNRAS 467:1126–1139. https://doi.org/10.1093/ mnras/stx065, 1701.03104 Delfosse X, Tinney CG, Forveille T et al (1997) Field brown dwarfs found by DENIS. A&A 327:L25–L28 Delorme P, Delfosse X, Albert L et al (2008a) CFBDS J005910.90-011401.3: reaching the T-Y brown dwarf transition? A&A 482:961–971. https://doi.org/10.1051/0004-6361:20079317, arXiv:0802.4387 Delorme P, Willott CJ, Forveille T et al (2008b) Finding ultracool brown dwarfs with MegaCam on CFHT: method and first results. A&A 484:469–478. https://doi.org/10.1051/0004-6361: 20078843, 0804.1477 Delorme P, Albert L, Forveille T et al (2010) Extending the Canada-France brown dwarfs survey to the near-infrared: first ultracool brown dwarfs from CFBDSIR. A&A 518:A39+. https://doi. org/10.1051/0004-6361/201014277, 1004.3876 Dupuy TJ, Liu MC (2017) Individual dynamical masses of ultracool dwarfs. ApJS 231:15. https:// doi.org/10.3847/1538-4365/aa5e4c, 1703.05775 Edge A, Sutherland W, Kuijken K et al (2013) The VISTA kilo-degree infrared galaxy (VIKING) survey: bridging the gap between low and high redshift. Messenger 154:32–34 Eisenstein DJ, Weinberg DH, Agol E et al (2011) SDSS-III: massive spectroscopic surveys of the distant Universe, the milky way, and extra-solar planetary systems. AJ 142:72. https://doi.org/ 10.1088/0004-6256/142/3/72, 1101.1529 Epchtein N, de Batz B, Capoani L et al (1997) The deep near-infrared southern sky survey (DENIS). Messenger 87:27–34 Faherty JK, Rice EL, Cruz KL, Mamajek EE, Núñez A (2013) 2MASS J035523.37+113343.7: a young, dusty, nearby, isolated brown dwarf resembling a giant exoplanet. AJ 145:2. https://doi. org/10.1088/0004-6256/145/1/2, 1206.5519 Faherty JK, Riedel AR, Cruz KL et al (2016) Population properties of brown dwarf analogs to exoplanets. ApJS 225:10. https://doi.org/10.3847/0067-0049/225/1/10, 1605.07927 Fan X, Knapp GR, Strauss MA et al (2000) L dwarfs found in Sloan digital sky survey commissioning imaging data. AJ 119:928–935. https://doi.org/10.1086/301224, astro-ph/9909263 Folkes SL, Pinfield DJ, Kendall TR, Jones HRA (2007) Discovery of a nearby L-T transition object in the Southern Galactic plane. MNRAS 378:901–909. https://doi.org/10.1111/j.13652966.2007.11789.x, arXiv:astro-ph/0703808 Folkes SL, Pinfield DJ, Jones HRA et al (2012) Identifying ultra-cool dwarfs at low Galactic latitudes: a southern candidate catalogue. MNRAS 427:3280–3319. https://doi.org/10.1111/j. 1365-2966.2012.21132.x, 1204.4477 Gagné J, Burgasser AJ, Faherty JK et al (2015a) SDSS J111010.01+011613.1: a new Planetarymass T dwarf member of the AB Doradus moving group. ApJ 808:L20. https://doi.org/10. 1088/2041-8205/808/1/L20, 1506.04195 Gagné J, Faherty JK, Cruz KL et al (2015b) BANYAN. VII. A new population of young substellar candidate members of nearby moving groups from the BASS survey. ApJS 219:33. https://doi. org/10.1088/0067-0049/219/2/33, 1506.07712

23 Large-Scale Searches for Brown Dwarfs and Free-Floating Planets

525

Gagné J, Faherty JK, Burgasser AJ et al (2017) SIMP J013656.5+093347 Is likely a planetarymass object in the carina-near moving group. ApJ 841:L1. https://doi.org/10.3847/2041-8213/ aa70e2, 1705.01625 Gauza B, Béjar VJS, Pérez-Garrido A et al (2015) Discovery of a young planetary mass companion to the nearby M dwarf VHS J125601.92-125723.9. ApJ 804:96. https://doi.org/10.1088/0004637X/804/2/96, 1505.00806 Geballe TR, Knapp GR, Leggett SK et al (2002) Toward spectral classification of L and T dwarfs: infrared and optical spectroscopy and analysis. ApJ 564:466–481. https://doi.org/10.1086/ 324078, arXiv:astro-ph/0108443 Geißler K, Metchev S, Kirkpatrick JD, Berriman GB, Looper D (2011) A cross-match of 2MASS and SDSS. II. Peculiar L dwarfs, unresolved binaries, and the space density of T dwarf secondaries. ApJ 732:56. https://doi.org/10.1088/0004-637X/732/1/56, 1103.1160 Gizis JE (2002) Brown dwarfs and the TW hydrae association. ApJ 575:484–492. https://doi.org/ 10.1086/341259, astro-ph/0204342 Gizis JE, Monet DG, Reid IN et al (2000) New neighbors from 2MASS: activity and kinematics at the bottom of the main sequence. AJ 120:1085–1099. https://doi.org/10.1086/301456, astro-ph/0004361 Gizis JE, Reid IN, Knapp GR et al (2003) Hubble space telescope observations of binary very low mass stars and brown dwarfs. AJ 125:3302–3310. https://doi.org/10.1086/374991, astro-ph/0302526 Hawley SL, Covey KR, Knapp GR et al (2002) Characterization of M, L, and T dwarfs in the Sloan digital sky survey. AJ 123:3409–3427. https://doi.org/10.1086/340697, astro-ph/0204065 Kaiser N, Burgett W, Chambers K et al (2010) The Pan-STARRS wide-field optical/NIR imaging survey. In: Ground-based and airborne telescopes III, Proceedings of SPIE, vol 7733, p 77330E. https://doi.org/10.1117/12.859188 Kendall TR, Mauron N, Azzopardi M, Gigoyan K (2003) Serendipitous discovery of seven new southern L-dwarfs. A&A 403:929–936. https://doi.org/10.1051/0004-6361:20030218, astro-ph/0302344 Kendall TR, Delfosse X, Martín EL, Forveille T (2004) Discovery of very nearby ultracool dwarfs from DENIS. A&A 416:L17–L20. https://doi.org/10.1051/0004-6361:20040046, astro-ph/0402171 Kendall TR, Tamura M, Tinney CG et al (2007) Two T dwarfs from the UKIDSS early data release. A&A 466:1059–1064. https://doi.org/10.1051/0004-6361:20066403, arXiv:astro-ph/0702534 Kirkpatrick JD (2005) New spectral types L and T. ARA&A 43:195–245. https://doi.org/10.1146/ annurev.astro.42.053102.134017 Kirkpatrick JD, Reid IN, Liebert J et al (1999) Dwarfs cooler than “M”: the definition of spectral type “L” using discoveries from the 2 micron all-sky survey (2MASS). ApJ 519:802–833. https://doi.org/10.1086/307414 Kirkpatrick JD, Reid IN, Liebert J et al (2000) 67 additional L dwarfs discovered by the two micron all sky survey. AJ 120:447–472. https://doi.org/10.1086/301427, arXiv:astro-ph/0003317 Kirkpatrick JD, Cruz KL, Barman TS et al (2008) A sample of very young field L dwarfs and implications for the brown dwarf “Lithium Test” at early ages. ApJ 689:1295–1326. https://doi. org/10.1086/592768, 0808.3153 Kirkpatrick JD, Looper DL, Burgasser AJ et al (2010a) Discoveries from a near-infrared proper motion survey using multi-epoch two micron all-sky survey data. ApJS 190:100–146. https:// doi.org/10.1088/0067-0049/190/1/100, 1008.3591 Kirkpatrick JD, Looper DL, Burgasser AJ et al (2010b) Discoveries from a near-infrared proper motion survey using multi-epoch two micron all-sky survey data. ApJS 190:100–146. https:// doi.org/10.1088/0067-0049/190/1/100, 1008.3591 Kirkpatrick JD, Cushing MC, Gelino CR et al (2011) The first hundred brown dwarfs discovered by the wide-field infrared survey explorer (WISE). ApJS 197:19. https://doi.org/10.1088/00670049/197/2/19, 1108.4677 Kirkpatrick JD, Gelino CR, Cushing MC et al (2012) Further defining spectral type “Y” and exploring the low-mass end of the field brown dwarf mass function. ApJ 753:156. https://doi. org/10.1088/0004-637X/753/2/156, 1205.2122

526

B. Burningham

Kirkpatrick JD, Schneider A, Fajardo-Acosta S et al (2014) The AllWISE motion survey and the quest for cold subdwarfs. ApJ 783:122. https://doi.org/10.1088/0004-637X/783/2/122, 1402.0661 Knapp GR, Leggett SK, Fan X et al (2004) Near-infrared photometry and spectroscopy of L and T dwarfs: the effects of temperature, clouds, and gravity. AJ 127:3553–3578. https://doi.org/ 10.1086/420707, arXiv:astro-ph/0402451 Kuchner MJ, Faherty JK, Schneider AC et al (2017) The first brown dwarf discovered by the backyard worlds: planet 9 citizen science project. ApJ 841:L19. https://doi.org/10.3847/20418213/aa7200, 1705.02919 Lawrence A, Warren SJ, Almaini O et al (2007) The UKIRT infrared deep sky survey (UKIDSS). MNRAS 379:1599–1617. https://doi.org/10.1111/j.1365-2966.2007.12040.x, arXiv:astroph/0604426 Leggett SK, Geballe TR, Fan X et al (2000) The missing link: early methane (“T”) dwarfs in the Sloan digital sky survey. ApJ 536:L35–L38. https://doi.org/10.1086/312728, astro-ph/0004408 Leggett SK, Marley MS, Freedman R et al (2007) Physical and spectral characteristics of the T8 and later type dwarfs. ApJ 667:537–548. https://doi.org/10.1086/519948, arXiv:0705.2602 Leggett SK, Tremblin P, Esplin TL, Luhman KL, Morley CV (2017) The Y-type brown dwarfs: estimates of mass and age from new astrometry, homogenized photometry, and near-infrared spectroscopy. ApJ 842:118. https://doi.org/10.3847/1538-4357/aa6fb5, 1704.03573 Line MR, Teske J, Burningham B, Fortney J, Marley M (2015) Uniform atmospheric retrieval analysis of ultracool dwarfs I: characterizing benchmarks, Gl570D and HD3651B. ArXiv e-prints 1504.06670 Line MR, Marley MS, Liu MC et al (2017) Uniform atmospheric retrieval analysis of ultracool dwarfs. II. Properties of 11 T dwarfs. ApJ 848:83. https://doi.org/10.3847/1538-4357/aa7ff0, 1612.02809 Liu MC, Magnier EA, Deacon NR et al (2013) The extremely red, Young L Dwarf PSO J318.5338-22.8603: a free-floating planetary-mass analog to directly imaged young gas-giant planets. ApJ 777:L20. https://doi.org/10.1088/2041-8205/777/2/L20, 1310.0457 Lodieu N, Dobbie PD, Deacon NR et al (2007a) A wide deep infrared look at the Pleiades with UKIDSS: new constraints on the substellar binary fraction and the low-mass initial mass function. MNRAS 380:712–732. https://doi.org/10.1111/j.1365-2966.2007.12106.x, 0706.2234 Lodieu N, Pinfield DJ, Leggett SK et al (2007b) Eight new T4.5-T7.5 dwarfs discovered in the UKIDSS large area survey data release 1. MNRAS 379:1423–1430. https://doi.org/10.1111/j. 1365-2966.2007.12023.x, arXiv:0705.3727 Lodieu N, Zapatero Osorio MR, Rebolo R, Martín EL, Hambly NC (2009) A census of very-lowmass stars and brown dwarfs in the  Orionis cluster. A&A 505:1115–1127. https://doi.org/10. 1051/0004-6361/200911966, 0907.2185 Lodieu N, Zapatero Osorio MR, Martín EL, Solano E, Aberasturi M (2010) GTC/OSIRIS spectroscopic identification of a faint L subdwarf in the UKIRT infrared deep sky survey. ApJ 708:L107–L111. https://doi.org/10.1088/2041-8205/708/2/L107, 0912.3364 Lodieu N, Burningham B, Day-Jones A et al (2012) First T dwarfs in the VISTA hemisphere survey. A&A 548:A53. https://doi.org/10.1051/0004-6361/201220182, 1210.5148 Looper DL, Kirkpatrick JD, Burgasser AJ (2007) Discovery of 11 new T dwarfs in the two micron all sky survey, including a possible L/T transition binary. AJ 134:1162–1182. https://doi.org/ 10.1086/520645, arXiv:0706.1601 Looper DL, Gelino CR, Burgasser AJ, Kirkpatrick JD (2008) Discovery of a T dwarf binary with the largest known J-band flux reversal. ArXiv e-prints 803, 0803.0544 LSST Science Collaboration, Abell PA, Allison J et al (2009) LSST science book, Version 2.0. ArXiv e-prints 0912.0201 Lucas PW, Tinney CG, Burningham B et al (2010) The discovery of a very cool, very nearby brown dwarf in the Galactic plane. MNRAS 408:L56–L60. https://doi.org/10.1111/j.17453933.2010.00927.x, 1004.0317 Luhman KL (2013) Discovery of a binary brown dwarf at 2 pc from the Sun. ApJ 767:L1. https:// doi.org/10.1088/2041-8205/767/1/L1, 1303.2401

23 Large-Scale Searches for Brown Dwarfs and Free-Floating Planets

527

Luhman KL (2014) Discovery of a ˜250 K brown dwarf at 2 pc from the Sun. ApJ 786:L18. https:// doi.org/10.1088/2041-8205/786/2/L18, 1404.6501 Luhman KL, Esplin TL (2016) The spectral energy distribution of the coldest known brown dwarf. AJ 152:78. https://doi.org/10.3847/0004-6256/152/3/78, 1605.06655 Mace GN, Kirkpatrick JD, Cushing MC et al (2013a) A study of the diverse T dwarf population revealed by WISE. ApJS 205:6. https://doi.org/10.1088/0067-0049/205/1/6, 1301.3913 Mace GN, Kirkpatrick JD, Cushing MC et al (2013b) The Exemplar T8 subdwarf companion of wolf 1130. ApJ 777:36. https://doi.org/10.1088/0004-637X/777/1/36, 1309.1500 Mainzer A, Bauer J, Grav T et al (2011) Preliminary results from NEOWISE: an enhancement to the wide-field infrared survey explorer for solar system science. ApJ 731:53. https://doi.org/10. 1088/0004-637X/731/1/53, 1102.1996 Marocco F, Jones HRA, Day-Jones AC et al (2015) A large spectroscopic sample of L and T dwarfs from UKIDSS LAS: peculiar objects, binaries, and space density. MNRAS 449:3651–3692. https://doi.org/10.1093/mnras/stv530, 1503.05082 Martín EL, Delfosse X, Basri G et al (1999) Spectroscopic classification of late-M and L field dwarfs. AJ 118:2466–2482. https://doi.org/10.1086/301107 Martín EL, Phan-Bao N, Bessell M et al (2010) Spectroscopic characterization of 78 DENIS ultracool dwarf candidates in the solar neighborhood and the Upper Scorpii OB association. A&A 517:A53. https://doi.org/10.1051/0004-6361/201014202, 1004.1775 McMahon RG, Banerji M, Gonzalez E et al (2013) First scientific results from the VISTA hemisphere survey (VHS). Messenger 154:35–37 Metchev SA, Kirkpatrick JD, Berriman GB, Looper D (2008) A cross-match of 2MASS and SDSS: newly found L and T dwarfs and an estimate of the space density of T dwarfs. ApJ 676:1281–1306. https://doi.org/10.1086/524721, 0710.4157 Miller GE, Scalo JM (1979) The initial mass function and stellar birthrate in the solar neighborhood. ApJS 41:513–547. https://doi.org/10.1086/190629 Minniti D, Lucas PW, Emerson JP et al (2010) VISTA variables in the Via Lactea (VVV): the public ESO near-IR variability survey of the milky way. New A 15:433–443. https://doi.org/ 10.1016/j.newast.2009.12.002, 0912.1056 Moraux E, Bouvier J, Stauffer JR, Cuillandre J (2003) Brown dwarfs in the Pleiades cluster: clues to the substellar mass function. A&A 400:891–902. https://doi.org/10.1051/0004-6361: 20021903, arXiv:astro-ph/0212571 Nakajima T, Oppenheimer BR, Kulkarni SR et al (1995) Discovery of a cool brown dwarf. Nature 378:463–+. https://doi.org/10.1038/378463a0 Naud ME, Artigau É, Malo L et al (2014) Discovery of a wide planetary-mass companion to the young M3 star GU Psc. ApJ 787:5. https://doi.org/10.1088/0004-637X/787/1/5, 1405.2932 Peña Ramírez K, Béjar VJS, Zapatero Osorio MR, Petr-Gotzens MG, Martín EL (2012) New isolated planetary-mass objects and the stellar and substellar mass function of the  orionis cluster. ApJ 754:30. https://doi.org/10.1088/0004-637X/754/1/30, 1205.4950 Phan-Bao N, Bessell MS, Martín EL et al (2008) Discovery of new nearby L and late-M dwarfs at low Galactic latitude from the DENIS data base. MNRAS 383:831–844. https://doi.org/10. 1111/j.1365-2966.2007.12564.x, 0708.4169 Pinfield DJ, Burningham B, Tamura M et al (2008) Fifteen new T dwarfs discovered in the UKIDSS large area survey. MNRAS 390:304–322. https://doi.org/10.1111/j.1365-2966.2008. 13729.x, 0806.0294 Pinfield DJ, Gomes J, Day-Jones AC et al (2014) A deep WISE search for very late type objects and the discovery of two halo/thick-disc T dwarfs: WISE 0013+0634 and WISE 0833+0052. MNRAS 437:1009–1026. https://doi.org/10.1093/mnras/stt1437 Rebolo R, Zapatero Osorio MR, Martín EL (1995) Discovery of a brown dwarf in the Pleiades star cluster. Nature 377:129–131. https://doi.org/10.1038/377129a0 Refregier A, Amara A, Kitching TD et al (2010) Euclid imaging consortium science book. ArXiv e-prints 1001.0061

528

B. Burningham

Reid IN, Gizis JE, Hawley SL (2002) The Palomar/MSU nearby star spectroscopic survey. IV. The luminosity function in the solar neighborhood and M dwarf kinematics. AJ 124:2721–2738. https://doi.org/10.1086/343777 Reid IN, Cruz KL, Kirkpatrick JD et al (2008) Meeting the cool neighbors. X. Ultracool dwarfs from the 2MASS all-sky data release. AJ 136:1290–1311. https://doi.org/10.1088/0004-6256/ 136/3/1290 Reylé C, Delorme P, Willott CJ et al (2010) The ultracool-field dwarf luminosity-function and space density from the Canada-France Brown Dwarf Survey. A&A 522:A112. https://doi.org/ 10.1051/0004-6361/200913234, 1008.2301 Sahlmann J, Lazorenko PF (2015) Mass ratio of the 2 pc binary brown dwarf LUH 16 and limits on planetary companions from astrometry. MNRAS 453:L103–L107. https://doi.org/10.1093/ mnrasl/slv113, 1506.07994 Salpeter EE (1955) The luminosity function and stellar evolution. ApJ 121:161–+. https://doi.org/ 10.1086/145971 Salpeter EE (2005) Introduction to IMF@50. In: Corbelli E, Palla F Zinnecker H (eds) The initial mass function 50 years later. Astrophysics and space science library, vol 327, p 3. https://doi. org/10.1007/978-1-4020-3407-7_1 Sandage A (1957) Observational approach to evolution. I. Luminosity functions. ApJ 125:422. https://doi.org/10.1086/146318 Scalo JM (1986) The stellar initial mass function. Fund Cosmic Phys 11:1–278 Schmidt M (1959) The rate of star formation. ApJ 129:243. https://doi.org/10.1086/146614 Schmidt SJ, West AA, Hawley SL, Pineda JS (2010) Colors and kinematics of L dwarfs from the Sloan digital sky survey. AJ 139:1808–1821. https://doi.org/10.1088/0004-6256/139/5/1808, 1001.3402 Schneider DP, Knapp GR, Hawley SL et al (2002) L dwarfs found in Sloan digital sky survey commissioning data. II. Hobby-Eberly telescope observations. AJ 123:458–465. https://doi. org/10.1086/338095, astro-ph/0110273 Schneider AC, Cushing MC, Kirkpatrick JD, Gelino CR (2016) The collapse of the Wien Tail in the coldest brown dwarf? Hubble space telescope near-infrared photometry of WISE J085510.83071442.5. ApJ 823:L35. https://doi.org/10.3847/2041-8205/823/2/L35, 1605.05618 Scholz RD, Storm J, Knapp GR, Zinnecker H (2009) Extremely faint high proper motion objects from SDSS stripe 82. Optical classification spectroscopy of about 40 new objects. A&A 494:949–967. https://doi.org/10.1051/0004-6361:200811053, 0812.1495 Sheppard SS, Cushing MC (2009) An infrared high proper motion survey using the 2MASS and SDSS: discovery of M, L, and T dwarfs. AJ 137:304–314. https://doi.org/10.1088/0004-6256/ 137/1/304, 0809.0697 Skemer AJ, Morley CV, Allers KN et al (2016) The first spectrum of the coldest brown dwarf. ApJ 826:L17. https://doi.org/10.3847/2041-8205/826/2/L17, 1605.04902 Skrutskie MF, Cutri RM, Stiening R et al (2006) The Two micron all sky survey (2MASS). AJ 131:1163–1183. https://doi.org/10.1086/498708 Skrzypek N, Warren SJ, Faherty JK et al (2015) Photometric brown-dwarf classification. I. A method to identify and accurately classify large samples of brown dwarfs without spectroscopy. A&A 574:A78. https://doi.org/10.1051/0004-6361/201424570, 1411.7578 Skrzypek N, Warren SJ, Faherty JK (2016) Photometric brown-dwarf classification. II. A homogeneous sample of 1361 L and T dwarfs brighter than J = 17.5 with accurate spectral types. A&A 589:A49. https://doi.org/10.1051/0004-6361/201527359, 1602.08582 Smith L, Lucas P, Burningham B et al (2013) The kinematic age of the coolest T dwarfs. ArXiv e-prints 1303.5288 Stone JM, Skemer AJ, Kratter KM et al (2016) Adaptive optics imaging of VHS 1256-1257: A low mass companion to a brown dwarf binary system. ApJ 818:L12. https://doi.org/10.3847/ 2041-8205/818/1/L12, 1601.03377 Strauss MA, Fan X, Gunn JE et al (1999) The discovery of a field methane dwarf from Sloan digital sky survey commissioning data. ApJ 522:L61–L64. https://doi.org/10.1086/312218, astro-ph/9905391

23 Large-Scale Searches for Brown Dwarfs and Free-Floating Planets

529

Tinney CG, Burgasser AJ, Kirkpatrick JD, McElwain MW (2005) The 2MASS Wide-Field T dwarf search. IV. Hunting out T dwarfs with methane imaging. AJ 130:2326–2346. https://doi. org/10.1086/491734, arXiv:astro-ph/0508150 Tsvetanov ZI, Golimowski DA, Zheng W et al (2000) The discovery of a second field methane brown dwarf from Sloan digital sky survey commissioning data. ApJ 531:L61–L65. https://doi. org/10.1086/312515, arXiv:astro-ph/0001062 Warren S, Hewett P (2002) WFCAM, UKIDSS, and z = 7 Quasars. In: Metcalfe N, Shanks T (eds) A new Era in cosmology. Astronomical society of the Pacific conference series, vol 283, p 369. astro-ph/0201216 Wilson JC, Miller NA, Gizis JE et al (2003) New M and L dwarfs confirmed with CorMASS. In: Martín E (ed) Brown Dwarfs, IAU Symposium, vol 211, p 197 Wright EL, Eisenhardt PRM, Mainzer AK et al (2010) The wide-field infrared survey explorer (WISE): mission description and initial on-orbit performance. AJ 140:1868. https://doi.org/10. 1088/0004-6256/140/6/1868, 1008.0031 Yee HKC, Gladders MD, Gilbank DG et al (2007) The red-sequence cluster surveys. In: Metcalfe N, Shanks T (eds) Cosmic frontiers. Astronomical society of the Pacific conference series, vol 379, p 103 York DG, Adelman J, Anderson JE Jr et al (2000) The Sloan digital sky survey: technical summary. AJ 120:1579–1587. https://doi.org/10.1086/301513, arXiv:astro-ph/0006396 Zhang ZH, Pokorny RS, Jones HRA et al (2009) Ultra-cool dwarfs: new discoveries, proper motions, and improved spectral typing from SDSS and 2MASS photometric colors. A&A 497:619–633. https://doi.org/10.1051/0004-6361/200810314, 0902.2798

Spectral Properties of Brown Dwarfs and Unbound Planetary Mass Objects

24

Jacqueline K. Faherty

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Overlap of Brown Dwarfs and Giant Planets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Identifying Young, Isolated Brown Dwarfs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . An Age-Calibrated Exoplanet Analog Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Clouds on Brown Dwarfs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Interplay of Age and Clouds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Diversity in the Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . What Lies Ahead . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

532 532 533 535 537 537 538 539 540

Abstract

Brown dwarfs overlap in effective temperature, luminosity, and mass with both the lowest mass stars and the highest mass planets. Specifically, young brown dwarfs and directly imaged exoplanets have enticingly similar photometric and spectroscopic characteristics, indicating that their cool, low gravity atmospheres should be studied in concert. Similarities between the peculiar-shaped H band, near and mid-IR photometry, and location on color magnitude diagrams provide important clues about how to extract physical properties of planets from current brown dwarf observations. In this chapter, objects newly assigned to 10–150 Myr nearby moving groups – many of which are unbound planetary mass objects – will be discussed, the diversity of this uniform age-calibrated brown dwarf sample will be highlighted, and the implication for understanding current and future planetary data will be reflected upon.

J. K. Faherty () Department of Astrophysics, American Museum of Natural History, New York, NY, USA e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_188

531

532

J. K. Faherty

Introduction Studying exoplanet atmospheres is a daunting task. The high contrast ratio of exoplanets to host stars precludes many detailed direct imaging or spectroscopic techniques which can yield a wealth of knowledge about the composition and patterns within an atmosphere. Emerging as a bridge population to giant exoplanets are young (10–150 Myr), isolated brown dwarfs that are vastly easier to study (e.g., Faherty et al. 2016, 2013; Allers and Liu 2013; Gagné et al. 2014, 2015a,b,c, 2017, 2018; Cruz et al. 2009; Liu et al. 2013, 2016). They are isolated, bright, and can be studied with a variety of small to large facilities from either the ground or space. Despite different formation mechanisms, brown dwarfs and giant exoplanets share many physical properties, including overlapping temperature regimes and condensate clouds in their atmospheres. Recent studies have revealed a striking resemblance between observations of directly imaged giant exoplanets and young low-temperature brown dwarfs (e.g., Faherty et al. 2016). Both populations deviate significantly from older, equivalent temperature objects, and it has been proposed that thick clouds present in the young objects but not in the old ones could explain anomalous observables (Barman et al. 2011; Currie et al. 2011; Madhusudhan et al. 2011). While only a handful of planetary systems can be directly studied with current technology, young brown dwarfs are relatively numerous, bright, and isolated in the field. They were largely discovered serendipitously while conducting an all-sky or proper motion search for nearby brown dwarfs (e.g., Kirkpatrick et al. 2010; Cruz et al. 2009; Gizis et al. 2012; Thompson et al. 2013). The current collection lends itself to low, medium, and high resolution optical and/or NIR spectroscopy, parallax programs, as well as precise photometric follow-ups. As such, they are excellent candidates for extensive studies not currently possible for exoplanets (Cruz et al. 2007, 2009; Rice et al. 2010a,b; Faherty et al. 2012, 2013; Allers and Liu 2013).

The Overlap of Brown Dwarfs and Giant Planets At masses 15% near-infrared variability (2MASS J21391365-3529507, hereafter 2M2139; Radigan et al. 2012). The surveys by Koen et al. (2004, 2005) provided the first near-continuous monitoring of a relatively large sample of objects in J, H , and K bands. The uninterrupted monitoring was sufficiently long to cover the expected median rotation period of brown dwarfs and led to a better handling of systematics. While no clear detection was reported, the 2 % variables as seen from an equatorial viewpoint raises to 80 %. These results led to the conclusion that high-amplitude variability is indeed more common at the L/T transition. Wilson et al. (2014) presented a similar dataset obtained at the ESO 3.6 m NTT telescope, albeit with typically shorter (2–4 h) monitoring. The results suggested that the fraction of variable objects within the L/T transition is similar to that of objects outside the transition, with a nearly uniform distribution of variable objects through the early-L to late-T sequence. This result was called into question by Radigan (2014) in which a reanalysis of the dataset was performed. Most early and mid-L variability detections were found to be constant at the sub-% level in this new analysis. A statistical assessment of combined datasets, with 82 individual objects, indicated a fraction of high-amplitude variables of 24C11 9 % within the L/T transition, much larger than the 3 % fraction of variables outside of it. The Metchev et al. (2015) sample does not display a significant increase in the fraction of high-amplitude variables close to the L/T transition compared to earlier or later objects. The very high level of stability of Spitzer allowed the detection of very low-level variability in the brighter – and typically earlier spectral types – objects. These observations showed that nearly two-thirds of L dwarfs display 0.2–2% variability, and about half of these variables are irregular, showing an evolution of surface features on time scales of a few hours. The steady increase in the maximum amplitude detected with spectral type at both 3.6 and 4.5 m is also notable. No L dwarf was found to vary by more than 2 %, while some T dwarfs vary by up to 4.6 %. The absence of a clear relation between the 3.6 or 4.5 m amplitudes is also noteworthy. If a single type of cloud was contrasting against a typical background, one would expect the contrast ratio between the two wavelength regions to be constant. Among both L and T variables, the 4.5 to 3.6 m amplitude ratio varies from 4%. The subsequent detection by Lew et al. (2016) of a 8% variability in the planetary-mass L dwarf WISEP J004701:06 C 680352:1 (W0047) with WFC3 grism observations seems to further confirm the link between high-amplitude variability and low gravity. Figure 3 compiles all J , 3.6, and 4.5 m variability detections in a spectral-type versus color diagram. The detections to date suggest a higher fraction of highamplitude variables among very red Ls, but one must bear in mind that some of these discoveries were from surveys explicitly targeting very red low-gravity objects. While surface gravity may be the link between color and variability amplitude, other physical parameters may explain this correlation. Rotation-induced variability is maximal for inclination close to 90ı ; if brown dwarfs have colors that differ at the equator relative to the poles, then the unresolved color of an object will correlate with its inclination, providing an alternative explanation for the correlation described earlier. Through high-resolution spectroscopy, Vos et al. (2017) measured the projected rotational velocity of a sample of early-L to early-T dwarfs with known rotation periods, thus constraining their inclination to the line of sight. The sample showed a significant correlation between the J K color anomaly (an object’s color relative to the mean color of objects of a similar spectral type) and its inclination: redder objects having higher inclination (i.e., seen by the equator). This results implies that on average, brown dwarf equators, at least in this spectral type range, are on average redder than their poles. This is a first hint at the latitudinal variations in brown dwarf properties, complementary to the longitudinal inhomogeneities probed through rotation-induced variability. Furthermore, as expected from geometric arguments, the authors found a correlation between variability amplitude and correlation, which also translates into a correlation between variability amplitude and color. To what extent low surface gravity and viewing angle, respectively, contribute to the observed correlation between color and variability amplitude remains an open question.

26 Variability of Brown Dwarfs

563

Fig. 3 Compilation of variable L, T, and Y dwarfs in the literature in a J  K versus spectraltype diagram (open color circles), with field objects (black dots). The symbol size is proportional to the variability amplitude, the most variable of all being the T1.5 dwarf 2M2139 (26% in J ). The largest reported amplitude is shown for objects with more than one detection in a given bandpass. Y dwarfs that do not yet have a published K-band magnitude have been arbitrarily set at J  K D 0 for display purpose. The blue, orange, and red circles, respectively, denote variability detections in J and the 3.6 and 4:5 m Spitzer IRAC bandpasses. A third-order polynomial fit to the color/spectral type relation is also shown. Most high-amplitude variables fall within the L/T transition region, where variability is expected as cloud decks sink below the photosphere. Interestingly, no high-amplitude variable has been found among L dwarfs bluer than the average for their spectral type. This exclusion region suggests that only redder, and generally lower-gravity, L dwarfs can display >2% variability in the near and mid-IR. (Data from Artigau et al. (2009), Radigan et al. (2014), Yang et al. (2015), Metchev et al. (2015), Radigan (2014), Biller et al. (2015), Buenzli et al. (2015), Clarke et al. (2008), Lew et al. (2016), Cushing et al. (2016), and Leggett et al. (2016). The field brown dwarf photometric data is from Gagné et al. (2015) and publicly available at www.astro.umontreal.ca/ gagne/listLTYs.php)

Spectroscopic Variability of Brown Dwarfs While time-resolved photometry provides constraints on the presence of weatherlike patterns on brown dwarfs, time-resolved spectroscopy provides information on the physical nature of the processes at play. From the ground, spectrophotometry with sub-% accuracy is very challenging, with variable slit losses, telluric absorption, and instrument flexures all masking low-level variability. The Wide Field Camera 3 (WFC3; MacKenty et al. 2010) slitless grism mode aboard the Hubble Space Telescope (HST) avoids these issues and provides the most compelling

564

É. Artigau

spectroscopic detections of variability in the literature. This mode covers the 1.1– 1.7 m domain at a =  100 resolving power. It covers methane absorption longward of 1.6 m as well as a deep water feature centered at 1.4 m, a wavelength range that is largely inaccessible from the ground due to the Earth’s atmospheric absorption. This WFC3 mode is also central to exoplanet study, either through phase curve (Stevenson et al. 2014) or transit spectroscopy. Buenzli et al. (2012) reports the detection of variability in the T6.5 2MASS J22282889–4310262 (2M2228) in WFC3 spectroscopy obtained simultaneously with Spitzer 4:5 m photometry. 2M2228 was a known photometric variable with a relatively rapid 1.4 h rotation period (Clarke et al. 2008; Radigan et al. 2014). The resulting light curves display sinusoidal modulation at various wavelengths with significant relative phase lags, by up to 180ı . These phase lags correlate with the effective pressure probed by each wavelength range (see Fig. 3 in Buenzli et al. 2012). A single spot on the surface of the BD would lead to a photometric modulation with a common phase at all wavelengths. A flux reversal, for example, a redder spot on a bluer surface, can lead to an anticorrelation between wavelengths or a phase lag of 180ı . The phase lag at values other than 0ı or 180ı would imply that the weather patterns in cause span a significant fraction of the circumference of the object. The typical scale height of a BD is on the order of a few km, while the radius is about 7  104 km. It would therefore seem unphysical for a single atmospheric feature that spans a few scale heights to be stretched half across the disk of the BD while preserving its integrity. The authors suggest the presence of large-scale temperature and/or opacity gradients across the surface as a plausible explanation, but much detailed dynamical simulations will be required to draw any firm conclusion. Yang et al. (2016) presented a second set of observations of this mid-T, with simultaneous HST and Spitzer observations obtained 2 years later. Their observations show phase shifts similar to those reported by Buenzli et al. (2012), except for the 4.5 m bandpass. Light curves in that dataset show phase lags clustering either around 0ı or 180ı (see Figs. 14 and 19 in Yang et al. 2016). This may not require significant extent of surface features and may be explained by a flux reversal within a single spot. This difference between two visits of the same objects highlights the fact that a better understanding of the range of behaviors seen on a single BD is needed before drawing a firm conclusion regarding the differences between objects. Apai et al. (2013) present a dataset similar to the HST observations of Buenzli et al. (2012) for two of the highest amplitude variable T dwarfs: SIMP0136 and 2M2139. Their variability was detected at a high significance, but no significant phase lag between spectral features probing different pressures in the atmosphere was observed. As shown in Fig. 4, both objects show a decreased variability in the 1.4 m water absorption feature. This feature typically probes pressures of 3 bar, while the middle of J band probes pressures of 10 bar (Yang et al. 2015). This

26 Variability of Brown Dwarfs

565

Fig. 4 (From Yang et al. 2015. Reproduced by permission of the AAS.) Spectral variability for two mid-L dwarfs (2M1821, 2M1507) and two early-T dwarfs (SIMP0136, 2M2139). The earlyTs, on the right, display a lower variability in their water bands compared to other wavelengths, suggesting that the cloud decks involved in the photometric modulation lay at a depth intermediate between the altitude at which the J - and H -band flux is emitted and that of water absorption. This behavior is not seen in L dwarfs, which is indicative of high-altitude clouds, above the depth of water absorption (4 bars). The upper plot shows the spectra corresponding to the maximum and minimum total fluxes of all four objects (respectively, red and blue curves). The lower plots show the ratio of the maximum and minimum spectra; values close to unity indicate no variability

566

É. Artigau

Fig. 5 Schematic representation of the cloud structure suggested by the results of Apai et al. (2013) and Yang et al. (2016). In early T dwarfs, clouds of varying thicknesses are present on the surface, and the fractional coverage of thin and thick clouds changes through time. With thinner, low-lying cloud decks, the flux at 1.2 and 1.6 m (i.e., the peak of J and H bands) probes deeper into the atmosphere, while water bands at 1.4 m probe the cooler layers of the atmosphere. In the presence of thicker clouds, fluxes at 1.2 and 1.6 m sample cooler, high-altitude cloud layers, while the depth probed at 1.4 m does not change significantly. This leads to a higher variability at 1.2 and 1.6 m and lower variability at 1.4 m

lower variability in the deep water bands implies that cloud decks that lead to photometric modulation in these two objects predominantly lay between these two pressures (see Fig. 5). This behavior is also seen in the slightly warmer T0.5 dwarf Luhman 16B (Buenzli et al. 2015). This consistent behavior between three early-Ts differs from the two L5 described in Yang et al. (2015), 2MASS J18212815 C 1414010 and 2MASS J15074769  1627386 (2M1821, 2M1507), and the very dusty L6 dwarf WISE0047 (Lew et al. 2016), where the variability within the water bands is similar to that of the J - and H -band peaks (see Fig. 4). This indicates that variability is due to hazes occurring above an optical depth of  D 1 for water absorption in mid-L dwarfs. Interestingly, the effective pressure at the peak of J and in the 1.4 m water feature differs less for mid-L dwarfs (4.3 versus 6.5 bar) than they do in early-T dwarfs (4.1 versus 8.1 bar; see Table 6 in Apai et al. 2013), which leads to a more wavelength-independent variability in L dwarfs. A noteworthy characteristic of the variability spectrum of 2M1821 (Fig. 4) and in WISE0047 (See Fig. 4 in Lew et al. 2016) is the slope in the variability spectrum. Variability at 1.1 m is larger than at 1.6 m, with a linear trend in between. This behavior is best explained by a wavelength-dependent extinction within the high-altitude clouds, and the slope provides information on the typical grain size within the clouds (0:4 m; Lew et al. 2016).

26 Variability of Brown Dwarfs

567

Photometric Variability of Y Dwarfs The WISE mission (Wright et al. 2010) is an all-sky survey satellite that operated between 3:4 and 22 m and allowed for the first time to identify a sample of BDs well below 600 K, corresponding to the Y spectral class (Cushing et al. 2011; Kirkpatrick et al. 2012). At such low temperatures, Y dwarfs are not expected to host silicate-bearing dust grains close to or above their photosphere such as what is seen in L and early T dwarfs. A number of chemical species are nevertheless expected to form clouds in cool atmosphere, such as sulfides (Morley et al. 2012), or for the coolest objects, ammonia, and water (Skemer et al. 2016; ?). The coolest of these objects, such as WISE J085510.83–071442.5 (W0855; Luhman and Esplin 2016; Luhman 2014), may well host weather patterns that include the rain and snow that are familiar to earthlings, albeit in a completely different physical setting. From the ground, the faintness of Y dwarfs in the near-infrared and the overwhelming thermal background in the mid-infrared strongly limit the possibilities to study their variability. Only Spitzer is sufficiently sensitive beyond 3 m to allow Y dwarf studies at high photometric accuracy. Cushing et al. (2016) and Leggett et al. (2016) reported preliminary results of a Y dwarf variability survey, with the detection of similar variability amplitudes (3% at 4.5 m) and periods (6–8.5 h) in WISEP J140518:40 C 553421:4.W 1405/ and WISEP J173835:53 C 273258:9. The very red 3.6 to 4.5 m color of Y dwarfs makes variability detection at 3.6 m challenging, and only one epoch of the W1405 observations shows an unambiguous detection at 3.6 m, leading to a 3.6 to 4.5 m amplitude ratio close to unity. The variability of these Y dwarfs, both in terms of amplitude and time scale, is comparable to that of T dwarf as measured by Metchev et al. (2015). However, the important difference in temperature suggests that different cloud species are most likely at play. The Y dwarf W0855 has a temperature of only 250 K and an estimated mass below the deuterium burning limit (Luhman 2014). It is a free-floating analog to the cold-evolved giant planets found by radial velocity (RV) surveys and provides a unique opportunity to understand their atmospheres. This brown dwarf was monitored by Esplin et al. (2016) on two epochs with Spitzer, and clear photometric variability was detected at 3.6 and 4:5 m (See Fig. 6). While the variability is unambiguous, no accurate period can be measured because the evolution of the light curve masks any clear periodicity. The best estimates suggest a 9–14 h rotation, comparable to solar system gas giants. The variability amplitude ratio between the two bandpasses is close to unity, similar to the two warmer Y dwarfs mentioned above. While it could be tempting to attribute this variability to water ice clouds, the data in hand for W0855 falls short from confirming this hypothesis, and only time-resolved spectroscopy could establish whether we are witnessing our first BD snowstorm. With a collection area 50 times larger and a much more diverse suite of observing modes, the James Webb Space Telescope (JWST) will provide vastly improved constraints on the nature of Y dwarf variability compared to what is currently possible with Spitzer data. The predicted sensitivities of JWST’s Near-Infrared

568

É. Artigau

Fig. 6 (From Esplin et al. 2016. Reproduced by permission of the AAS.) Spitzer IRAC timeseries of W0855 at 4.5 and 3.6 m (respectively red and blue dots). The 3.6 m filter probes a deep methane absorption band and W0855 is much fainter at this wavelength than at 4.5 m, hence the lower photometric precision

Spectrograph (NIRSpec) should allow spectroscopy at a resolving power of 1000 and a signal-to-noise ratio above 100 per resolution element at 4 m for an hourlong integration on W0855. This will lead to a detection of spectroscopic variability within each resolution element, assuming a variability level comparable to that reported by Esplin et al. (2016) and will allow to perform detailed modelling of cloud dynamics and chemistry.

Brown Dwarf Variability as a Limitation to Radial Velocity Surveys Brown dwarf variability provides a wealth of information on weather patterns that would be difficult or impossible to obtain otherwise, but these patterns will also pose significant challenges to other aspects of BD study. BDs are hosts to planetarymass companions, which are often referred to as planets even if the formal IAU definition states that “planets” orbit stars. The first directly imaged planet was found around the young BD 2MASSW J1207334393254 (Chauvin et al. 2005). Disks are also common around young BDs (e.g., Luhman et al. 2005), suggesting that planet formation is commonplace around these objects. BDs are enticing targets for radial velocity (RV) searches as planets will induce a much larger signal than for Sun-like stars due to the lower mass of the host (0:2% variability in the Spitzer bandpasses. With a majority of L dwarfs rotating faster than 20 km/s, and typically having at least 0:2 % variability, one should expect a typical variabilityinduced jitter of >40 m/s. This is analogous to the activity jitter encountered with M dwarfs, a problem that has received significant attention as RV surveys extend to ever cooler targets (Boisse et al. 2011; Reiners et al. 2010). While hampering future RV planet searches around L and T dwarfs, weather-induced RV jitter opens the door to Doppler imaging (Vogt and Penrod 1983) as a new technique for exploring the atmosphere of BDs.

Doppler Imaging of Brown Dwarfs The main motivation behind the study of BD variability is to have a glance at the diversity of weather patterns on their surfaces. Doppler imaging is a wellestablished technique that has been used to resolved stellar features for decades. As a star rotates, brightness variations on its surface translate into time-varying signatures in its mean spectral line profile. This technique has been extended to active M dwarfs (Barnes and Collier Cameron 2001), and BDs are promising targets for Doppler imaging studies. The presence of cloud patterns is well established through photometric variability. and their rotation profiles can be resolved by stateof-the-art RV spectrographs operating in the near-infrared. The rich molecular bands are amenable to least-squares deconvolution, which partially offsets the loss in signal-to-noise due to their relative faintness. Crossfield et al. (2014) presented the first demonstration of Doppler imaging on the T0.5 dwarf Luhman 16B. This objects is one of the best possible cases for this type of study as it is much brighter than any other T dwarf due to its proximity (2.0 pc; Luhman 2013) and displays one of the largest known photometric variabilities among T dwarfs (Buenzli et al. 2015; Biller et al. 2013). The map was reconstructed using only a relatively short wavelength interval centered on the 2.29 m CO bandhead, with CRIRES at the VLT (Kaeufl et al. 2004). The recovered map (see Fig. 7) shows largescale inhomogeneities in surface brightness; whether these represent differences in temperature or composition remains to be seen. These results represent an exciting proof of concept as a tool to probe BD atmosphere. Obtaining simultaneous maps of various chemical species with strong near-infrared signatures (e.g., methane or water) is possible, as well as a multi-epoch monitoring of cloud maps. These will yield strong constraints on atmosphere dynamics of BDs. A few brighter M/L transition dwarfs and a handful L dwarfs will be amenable to Doppler imaging with 4–8 m class telescopes equipped with broadband precision radial velocity

570

É. Artigau

Fig. 7 (Reprinted by permission from Springer Nature: A global map of the nearest known brown dwarf, Crossfield et al. 2014.) Global surface brightness map of Luhman 16B derived from Doppler imaging. Each of the six maps shows the observer-facing hemisphere through the 4.9 h rotation period. A dark region close to the equator (2.4 h) and a bright region close to the pole (0.0 h) are recovered at high significance. The contrast between the darkest and brightest regions is ˙10%

spectrographs in the near-infrared. The advent of similar instruments on 30 m-class telescopes will pave the way to surface mapping of dozens of BDs and possibly the brighter imaged exoplanets (Crossfield 2014).

Cross-References  Large-Scale Searches for Brown Dwarfs and Free-Floating Planets  Spectral Properties of Brown Dwarfs and Unbound Planetary Mass Objects  Y Dwarfs: The Challenge of Discovering the Coldest Substellar Population in the

Solar Neighborhood

References Apai D, Radigan J, Buenzli E, Burrows A, Reid IN, Jayawardhana R (2013) HST spectral mapping of L/T transition brown dwarfs reveals cloud thickness variations. ApJ 768:121 Apai D et al (2017) Zones, spots, and planetary-scale waves beating in brown dwarf atmospheres. Science 357:683 Artigau É, Doyon R, Lafrenière D, Nadeau D, Robert J, Albert L (2006) Discovery of the brightest T dwarf in the northern hemisphere. ApJ 651:L57

26 Variability of Brown Dwarfs

571

Artigau E, Bouchard S, Doyon R, Lafrenière D (2009) Photometric variability of the T2.5 brown dwarf SIMP J013656.5+093347: evidence for evolving weather patterns. ApJ 701:1534 Artigau, É et al (2014) SPIRou: the near-infrared spectropolarimeter/high-precision velocimeter for the Canada-France-Hawaii telescope. In: Society of photo-optical instrumentation engineers (SPIE) conference series, vol 9147, p 15 Bailer-Jones CAL, Mundt R (1999) A search for variability in brown dwarfs and L dwarfs. A&A 348:800 Bailer-Jones CAL, Mundt R (2001) Variability in ultra cool dwarfs: evidence for the evolution of surface features. A&A 367:218 Barnes JR, Collier Cameron A (2001) Starspot patterns on the M dwarfs HK Aqr and RE 1816 +541. MNRAS 326:950 Belu AR et al (2013) Habitable planets eclipsing brown dwarfs: strategies for detection and characterization. ApJ 768:125 Biller BA et al (2013) Weather on the nearest brown dwarfs: resolved simultaneous multiwavelength variability monitoring of WISE J104915.57-531906.1AB. ApJ 778:L10 Biller BA et al (2015) Variability in a young, L/T transition planetary-mass object. ApJ 813:L23 Boisse I, Bouchy F, Hébrard G, Bonfils X, Santos N, Vauclair S (2011) Disentangling between stellar activity and planetary signals. In: IAU symposium, vol 273, pp 281–285 Bouchy F et al (2017) Near-infrared planet searcher to join HARPS on the ESO 3.6-metre telescope. The messenger 169:21 Boyajian TS et al (2012) Stellar diameters and temperatures. II. Main-sequence K- and M-stars. ApJ 757:112 Buenzli E et al (2012) Vertical atmospheric structure in a variable brown dwarf: pressure-dependent phase shifts in simultaneous Hubble Space Telescope-Spitzer light curves. ApJ 760:L31 Buenzli E, Marley MS, Apai D, Saumon D, Biller BA, Crossfield IJM, Radigan J (2015) Cloud structure of the nearest brown dwarfs. II. High-amplitude variability for Luhman 16 A and B in and out of the 0.99 m FeH feature. ApJ 812:163 Burgasser AJ et al (1999) Discovery of four field methane (T-type) dwarfs with the two micron all-sky survey. ApJ 522:L65 Burgasser AJ, Liebert J, Kirkpatrick JD, Gizis JE (2002) A search for variability in the active T dwarf 2MASS 1237+6526. AJ 123:2744 Chabrier G, Baraffe I (2000) Theory of low-mass stars and substellar objects. ARA&A 38:337 Chauvin G, Lagrange A-M, Dumas C, Zuckerman B, Mouillet D, Song I, Beuzit J-L, Lowrance P (2005) Giant planet companion to 2MASSW J1207334-393254. A&A 438:L25 Clarke FJ, Tinney CG, Covey KR (2002) Periodic photometric variability of the brown dwarf Kelu-1. MNRAS 332:361 Clarke FJ, Tinney CG, Hodgkin ST (2003) Time-resolved spectroscopy of the variable brown dwarf Kelu-1 . MNRAS 341:239 Clarke FJ, Hodgkin ST, Oppenheimer BR, Robertson J, Haubois X (2008) A search for J-band variability from late-L and T brown dwarfs. MNRAS 386:2009 Crossfield IJM (2014) Doppler imaging of exoplanets and brown dwarfs. A&A 566:A130 Crossfield IJM et al (2014) A global cloud map of the nearest known brown dwarf. Nature 505:654 Cushing MC et al (2011) The discovery of Y dwarfs using data from the wide-field infrared survey explorer (WISE). ApJ 743:50 Cushing MC et al (2016) The first detection of photometric variability in a Y dwarf: WISE J140518.39+553421.3. ApJ 823:152 Enoch ML, Brown ME, Burgasser AJ (2003) Photometric variability at the L/T dwarf boundary. AJ 126:1006 Esplin TL, Luhman KL, Cushing MC, Hardegree-Ullman KK, Trucks JL, Burgasser AJ, Schneider AC (2016) Photometric monitoring of the coldest known brown dwarf with the Spitzer space telescope. ApJ 832:58 Gagné J et al (2015) BANYAN. VII. A new population of young substellar candidate members of nearby moving groups from the BASS survey. ApJS 219:33

572

É. Artigau

Gelino C, Marley M (2000) Variability in an unresolved Jupiter. In: Griffith CA, Marley MS (eds) From giant planets to cool stars. Astronomical society of the Pacific conference series, vol 212, p 322 Gelino CR, Marley MS, Holtzman JA, Ackerman AS, Lodders K (2002) L dwarf variability: I-band observations. ApJ 577:433 Girardin F, Artigau É, Doyon R (2013) In search of dust clouds: photometric monitoring of a sample of late L and T dwarfs. ApJ 767:61 Kaeufl H-U et al (2004) CRIRES: a high-resolution infrared spectrograph for ESO’s VLT. In: Moorwood AFM, Iye M (eds) Ground-based instrumentation for astronomy. Proceedings of the SPIE, vol 5492, pp 1218–1227 Kirkpatrick JD et al (2012) Further defining spectral type “Y” and exploring the low-mass end of the field brown dwarf mass function. ApJ 753:156 Koen C, Matsunaga N, Menzies J (2004) A search for short time-scale JHK variability in ultracool dwarfs. MNRAS 354:466 Koen C, Tanabé T, Tamura M, Kusakabe N (2005) JHK time-series observations of a few ultracool dwarfs. MNRAS 362:727 Leggett SK et al (2000) The missing link: early methane (“T”) dwarfs in the Sloan digital sky survey. ApJ 536:L35 Leggett SK, Morley CV, Marley MS, Saumon D (2015) Near-infrared photometry of Y dwarfs: low ammonia abundance and the onset of water clouds. ApJ 799:37 Leggett SK et al (2016) Observed variability at 1 and 4 m in the Y0 brown dwarf WISEP J173835.52+273258.9. ApJ 830:141 Lew BWP et al (2016) Cloud atlas: discovery of patchy clouds and high-amplitude rotational modulations in a young, extremely red L-type brown dwarf. ApJ 829:L32 Liu MC, Leggett SK (2005) Kelu-1 is a binary L dwarf: first brown dwarf science from laser guide star adaptive optics. ApJ 634:616 Luhman KL (2013) Discovery of a binary brown dwarf at 2 pc from the sun. ApJ 767:L1 Luhman KL (2014) Discovery of a 250 K brown dwarf at 2 pc from the sun. ApJ 786:L18 Luhman KL, Esplin TL (2016) Photometric monitoring of the coldest known brown dwarf with the Spitzer space telescope. AJ 152:78 Luhman KL et al (2005) The disk fractions of brown dwarfs in IC 348 and Chamaeleon I. ApJ 631:L69 MacKenty JW, Kimble RA, O’Connell RW, Townsend JA (2010) On-orbit performance of HST wide field camera 3. In: Space telescopes and instrumentation 2010: optical, infrared, and millimeter wave. Proceedings of the SPIE, vol 7731, p 77310Z Marley MS, Ackerman AS, Burgasser AJ, Saumon D, Lodders K, Freedman R (2003) Clouds and clearings in the atmospheres of the L and T dwarfs. In: Martín E (ed) Brown dwarfs. IAU symposium, vol 211, p 333 Metchev S et al (2013) Clouds in brown dwarfs and giant planets. Astron Nachr 334:40 Metchev SA et al (2015) Weather on other worlds. II. Survey results: spots are ubiquitous on L and T dwarfs. ApJ 799:154 Mohanty S, Basri G (2003) Rotation and activity in mid-M to L field dwarfs. ApJ 583:451 Morales-Calderón M et al (2006) A sensitive search for variability in late L dwarfs: the quest for weather. ApJ 653:1454 Morley CV, Fortney JJ, Marley MS, Visscher C, Saumon D, Leggett SK (2012) Neglected clouds in T and Y dwarf atmospheres. ApJ 756:172 Nakajima T, Oppenheimer BR, Kulkarni SR, Golimowski DA, Matthews K, Durrance ST (1995) Discovery of a cool brown dwarf. Nature 378:463 Quirrenbach A et al (2014) CARMENES: blue planets orbiting red dwarfs. In: IAU symposium, vol 299, pp 395–396 Racine R (1978) The Mont Megantic astronomical observatory – a new major observatory for Canada. JRASC 72:324 Radigan J (2014) Strong brightness variations signal cloudy-to-clear transition of brown dwarfs. ApJ 797:120

26 Variability of Brown Dwarfs

573

Radigan J, Jayawardhana R, Lafrenière D, Artigau É, Marley M, Saumon D (2012) Largeamplitude variations of an L/T transition brown dwarf: multi-wavelength observations of patchy, high-contrast cloud features. ApJ 750:105 Radigan J, Lafrenière D, Jayawardhana R, Artigau E (2014) Strong brightness variations signal cloudy-to-clear transition of brown dwarfs. ApJ 793:75 Rebolo R, Zapatero Osorio MR, Martín EL (1995) Discovery of a brown dwarf in the Pleiades star cluster. Nature 377:129 Reiners A, Bean JL, Huber KF, Dreizler S, Seifahrt A, Czesla S (2010) Detecting planets around very low mass stars with the radial velocity method. ApJ 710:432 Simon AA et al (2016) Neptune’s dynamic atmosphere from Kepler K2 observations: implications for brown dwarf light curve analyses. ApJ 817:162 Skemer A et al (2016) The first spectrum of the coldest brown dwarf. ApJ 826:L17 Stevenson KB et al (2014) Thermal structure of an exoplanet atmosphere from phase-resolved emission spectroscopy. Science 346:838 Tamura M et al (2012) Infrared Doppler instrument for the Subaru telescope (IRD). In: Proceedings of the SPIE, vol 8446. SPIE Tinney CG, Tolley AJ (1999) Searching for weather in brown dwarfs. MNRAS 304:119 Vogt SS, Penrod GD (1983) Doppler imaging of spotted stars – application to the RS Canum Venaticorum star HR 1099. PASP 95:565 Vos JM, Allers KN, Biller BA (2017) The viewing geometry of brown dwarfs influences their observed colours and variability properties. ApJ 842:78 Wilson PA, Rajan A, Patience J (2014) The brown dwarf atmosphere monitoring (BAM) project. I. The largest near-IR monitoring survey of L and T dwarfs. A&A 566:A111 Wright EL et al (2010) The wide-field infrared survey explorer (WISE): mission description and initial on-orbit performance. AJ 140:1868 Yang H et al (2015) HST rotational spectral mapping of two L-type brown dwarfs: variability in and out of water bands indicates high-altitude haze layers. ApJ 798:L13 Yang H et al (2016) Extrasolar storms: pressure-dependent changes in light-curve phase in brown dwarfs from simultaneous HST and Spitzer observations. ApJ 826:8

Metal-Depleted Brown Dwarfs

27

Nicolas Lodieu

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Census of Ultracool Subdwarfs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Colors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Spectral Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Spectral Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Multiplicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Physical Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

576 577 577 579 580 582 583 584 585

Abstract

This chapter reviews our current knowledge of metal-poor ultracool dwarfs with spectral types later than M7. The current census of M, L, and T subdwarfs is explored. The main color trends of subdwarfs from the optical to the mid-infrared are described and their spectral features presented, which led to a preliminary and tentative spectral classification subject to important changes in the future when more of these metal-poor objects are discovered. Their multiplicity and the determination of their physical parameters (effective temperature, gravity, metallicity, and mass) are discussed. Finally, some suggestions and future guidelines are proposed to foster our knowledge on the oldest and coolest members of our Galaxy.

N. Lodieu () Instituto de Astrofísica de Canarias, La Laguna, Spain e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_173

575

576

N. Lodieu

Keywords

Metal-poor dwarfs · Low-mass stars and brown dwarfs · Techniques: photometry · spectroscopy

Introduction M dwarfs represent the majority of stars in the solar neighborhood (Kirkpatrick et al. 2012) and in our Galaxy where the mass function peaks (e.g., Chabrier 2003). At lower masses, three new classes have defined during the past two decades: the L dwarfs whose atmospheres are affected by dust (Kirkpatrick et al. 1999; Martín et al. 1999), the T dwarfs shaped by methane and water absorption bands (Leggett et al. 2000; Geballe et al. 2002; Burgasser et al. 2002, 2006), and the Y dwarfs with the potential presence of ammonia at infrared wavelengths (Cushing et al. 2011; Kirkpatrick et al. 2012). The classification of L dwarfs is mostly morphological, but the large variety of sources discovered in optical and infrared large-scale surveys triggered a preliminary spectral scheme incorporating a new parameter: gravity (i.e., ages) as proposed by two independent teams (Cruz et al. 2009; Allers et al. 2007; Allers and Liu 2013). However, the spectral classification of metal-poor L dwarfs remains in its infancy due to the small sample existing in the literature. Nonetheless, recent discoveries offered new hints to elaborate a tentative spectral sequence. Metal-poor dwarfs belong to the spectral class VI in the Morgan-Keenan scheme (Morgan et al. 1943). They are also known as subdwarfs and often abbreviated “sd” (Joy 1947; Gizis 1997; Lépine et al. 2007). They usually exhibit bluer optical and infrared colors than their solar-like analogues (Lodieu et al. 2017) and show distinct spectral features such as the weakening of TiO bands (i.e., less TiO opacity implying more flux radiated from deeper and hotter layers of the atmosphere), strengthening of CaH bands, and strong collision-induced hydrogen absorption beyond 1 m (Gizis 1997; Lépine et al. 2007). They typically have high proper motions and large radial velocities translating into space motions compatible with membership to the thick disk and halo (Schmidt 1975). This population of metal-poor dwarfs is important for several reasons. Firstly, they represent key tracers of the history of our Galaxy because they are very old. Secondly, the knowledge of their physical parameters will impact on the study of globular clusters whose main populations are metal-poor and old. Thirdly, the census of metal-poor stars and brown dwarfs helps the determination of the luminosity and mass functions early on in the formation of our Galaxy to gauge the impact of metallicity in star formation processes. Unfortunately, metal-poor stars are not so numerous compared to their solar-like counterparts with only three subdwarfs of the 250 systems located within 10 pc ( Cas AB; Kapteyn’s star; CF Uma). This review will focus mainly on ultracool subdwarfs (UCSDs) with spectral types later than M7 and metallicities (Fe/H) equal or less than 0:5 dex unless otherwise stated. For more massive subdwarfs, readers are referred to one of the section dedicated to M subdwarfs in the book of Reid and Hawley (2005). The coolest L-type subdwarfs might be brown dwarfs, but none of them have been unambiguously proven to be substellar at the time of writing (Lodieu et al. 2015).

27 Metal-Depleted Brown Dwarfs

577

A few T-type metal-poor dwarfs have been announced as companions to bright stars with well-determined metallicities (Pinfield et al. 2012; Burningham et al. 2013), but only one has a metallicity below 0:5 dex, WISE J20052038C5424339 (Mace et al. 2013). This review summarizes the techniques employed over the past decades to identify metal-poor low-mass stars and brown dwarfs and describes the main spectral features leading to a preliminary and tentative classification scheme for UCSDs. The colors of UCSDs are mentioned and compared to those of nearby field M and L dwarfs. Our current knowledge on the multiplicity of UCSDs and the actual estimates of the physical parameters of the lowest mass metal-poor stars and brown dwarfs are also presented here. Finally, future needs are highlighted to characterize in more detail the population of UCSDs and their physical parameters.

Census of Ultracool Subdwarfs Dedicated searches for metal-poor stars and brown dwarfs usually focus on proper motion surveys to bias their final sample toward high-velocity objects, thus halo stars (Schmidt 1975). Most of the late-M subdwarfs have been identified in photographic plates from the Digital Sky Survey and the SuperCOSMOS Sky Survey (Gizis 1997; Gizis and Reid 1997; Gizis et al. 1997; Schweitzer et al. 1999; Lépine et al. 2003, 2007; Scholz et al. 2004a,b; Lodieu et al. 2005) and more recent all-sky or large-scale surveys such as the Two Micron All-Sky Survey (2MASS; Burgasser and Kirkpatrick 2006; Cushing et al. 2009), the Sloan Digital Sky Survey (SDSS; Lépine and Scholz 2008; Sivarani et al. 2009; Zhang et al. 2013), the UKIRT Infrared Deep Sky Survey (UKIDSS; Lodieu et al. 2012, 2017), and the Wide-field Infrared Survey Explorer (WISE; Kirkpatrick et al. 2014, 2016). Over the past years, the number of L subdwarfs has grown rapidly and is now just slightly over 30. The first one was identified in 2MASS (Burgasser et al. 2003) followed by other discoveries in the same database (Burgasser 2004; Cushing et al. 2009), SDSS (Sivarani et al. 2009; Schmidt et al. 2010; Bowler et al. 2010), WISE (Kirkpatrick et al. 2014, 2016), and cross-correlations of various surveys (Lodieu et al. 2010, 2012, 2017; Zhang et al. 2017a,b). In the T dwarf regime, a few examples of metal-poor T dwarfs have been reported as companions to brighter stars with well-determined metallicities both in UKIDSS (Pinfield et al. 2012; Burningham et al. 2013) and WISE (Mace et al. 2013). Nonetheless, their metallicities generally do not exceed 0:5 dex, except in the case of WISE J200520C542433 with Fe/H = 0:64 dex (Mace et al. 2013).

Colors Due to the dearth of metals in their atmospheres, UCSDs tend to exhibit on average bluer colors than their solar-type analogues (Fig. 1). Moreover, they lie below solarmetallicity dwarfs in reduced proper motions (left panel in Fig. 1) which constitute powerful tools to identify UCSDs (Lépine et al. 2007; Lépine and Scholz 2008;

Fig. 1 Reduced proper motion (left) and (J  w2, J  Ks ) color-color diagrams for main-sequence stars, M, L, and T dwarfs of different metallicities (Figures taken Kirkpatrick et al. 2016)

578 N. Lodieu

27 Metal-Depleted Brown Dwarfs

579

Kirkpatrick et al. 2016; Lodieu et al. 2017). This section summarizes differences reported in the literature going from blue to red wavelengths. These differences helped out increasing the numbers of UCSDs over the past years and represent key indices for upcoming surveys like the Large Synoptic Survey Telescope (LSST; Ivezic et al. 2008) and the Euclid mission (Mellier 2016). • Their locus in a (g  r, r  i ) color-color diagram is distinct from their solar-type counterparts. Selecting sources with g  r > 2 mag and g  i > 3 mag will strongly bias the photometric selection toward metal-poor M dwarfs (Fig. 2 in Lépine and Scholz 2008). • Their r z colors are bluer than field M dwarfs by at least 1 mag (Fig. 2 in Lépine and Scholz 2008). • The optical-to-infrared colors (e.g., i  J ) of UCSDs are bluer as metallicity decreases, yielding a flattening of the far-red part of their optical spectra (Gizis 1997; Lépine et al. 2007; Lodieu et al. 2017; Zhang et al. 2017b). • Their infrared colors (e.g., J  K < 0:7 mag) are much bluer than solar-type M dwarfs due to the strong pressure-induced H2 opacity beyond 1 micron (Lodieu et al. 2017; Zhang et al. 2017b). • Subdwarfs of types M and L typically fall blueward in near-infrared to midinfrared colors (e.g., J  w1, J  w2, H  w2) of their solar-metallicity counterparts due to the increasing influence of collision-induced hydrogen absorption (Kirkpatrick et al. 2016; Lodieu et al. 2017). • Their mid-infrared colors look similar to those of their solar-type analogues although opposite trends have been reported in the literature: red in Kirkpatrick et al. (2016) and blue in Lodieu et al. (2017).

Spectral Features The main spectral features indicative of low metallicity are the strengthening of the CaH bands and weakening of the TiO bands around 620–740 nm and the effect of the collision-induced absorption beyond 1000 nm. However, there are other features which can be employed to distinguish UCSDs from their solar-type cousins over the optical-infrared wavelength range (Gizis 1997; Lépine et al. 2007; Burgasser et al. 2007; Kirkpatrick et al. 2016; Zhang et al. 2017b), as enumerated below: 1. The CaH bands around 640–700 nm is stronger with lower metallicity but is also dependent on temperature in the 3600–3200 K range. 2. The TiO bands at 720, 780, and 850 nm are weaker with lower metallicity. The bluest of these bands, however, becomes more sensitive to temperature than metallicity for late-type M subdwarfs. Below 3200 K, the strength of the TiO band at 720 nm is not monotonic anymore with decreasing metallicity, making the classification of late-M and early-L subdwarfs based on spectral indices more unreliable. 3. The VO band at 800 nm is a strong indicator of metallicity in L dwarfs: it weakens as Fe/H decreases.

580

N. Lodieu

4. The CO band at 2300 nm weakens as metallicity decreases and eventually disappears in extreme and ultra-subdwarfs. 5. The near-infrared flux beyond 1000 nm becomes more depressed with lower metallicity due to the strong collision-induced H2 absorption.

Spectral Types The first spectral classification of M subdwarfs has been proposed by Gizis (1997), dividing M dwarfs into three main classes: solar-type M dwarfs with metallicity Fe/H around 0, M subdwarfs (sdM) with FeH  1:2 ˙ 0:3 dex, and extreme M subdwarfs (esdM) with approximate Fe/H of 2:0 ˙ 0:5 dex. This classification scheme is based on the strength of the TiO and CaH bands in the 620–740 nm optical range. Spectral indices have been defined to infer both metal class and spectral type. Ten years later, Lépine et al. (2007) revised the boundaries of the original metal classes based on a larger sample (factor of five bigger) of known metalpoor single and multiple systems. They introduced a new parameter, T iO=C aH , as well as an additional metal class with metallicities even lower than the esdM, the ultrasubdwarfs (usdM), to take into account the positions of known binaries sharing the same metallicity in the (early-)M dwarf regime. They also proposed spectral standards for each metal class and spectral types ranging from M0 to M8. The quality of the optical spectra of the sdM, esdM, and usdM templates has been improved by Savcheva et al. (2014) by stacking all SDSS spectra available per subtype and per metal class (Fig. 2). Jao et al. (2008) presented new thoughts about the naming and spectral classification of subdwarfs based on a sample of 88 K3–M6 sources. They propose to use the class VI of the Morgan-Keenan scheme instead of the terminology “sd” to name subdwarfs, a prefix that can be confused with the hotter sdO/sdB-type star which also appear sub-luminous in the HR diagram. And they showed the influence of gravity in the spectra of M subdwarfs and suggested to avoid classification based solely on spectral indices. In the L dwarf, the current classification is only tentative due to the small number of L subdwarfs announced to date and the narrow range of physical parameters. Nonetheless, several groups attempted to extend the M dwarf classification into the L regime. Zhang et al. (2017b) built on the extensions proposed by Burgasser et al. (2007) and Kirkpatrick et al. (2016), keeping the concept of the three metal classes proposed by Lépine et al. (2007) and applying it to the L subdwarfs (sdL, esdL, and usdL). Their spectral classification is based on the comparison of spectra of about 30 L0–L8 subdwarfs with those of solar-type L dwarf standards defined in the literature (e.g., Kirkpatrick et al. 2000). They focused on key features sensitive to metallicity and temperatures (their Table 3 and previous section) to evaluate the differences in the optical (CaH and VO bands around 700 nm, VO band at 800 nm, strength of the KI doublet at 770 nm, TiO band at 850 nm) and in the infrared (FeH band at 990 nm, CO band at 2300 nm, and the depression in the H and K passbands due to the collision-induced H2 absorption), yielding a revised classification of known metal-poor L dwarfs that can be used for future discoveries (Fig. 3).

b

c

Fig. 2 Sequence of M-type subdwarfs (sdM; left), extreme subdwarfs (esdM, middle), and ultrasubdwarfs (usdM; right) (Figure taken from Savcheva et al. (2014) with an earlier version shown in Lépine et al. 2007)

a

27 Metal-Depleted Brown Dwarfs 581

582

N. Lodieu

Fig. 3 Optical and near-infrared spectra of L4, L6, and L7 dwarfs/subdwarfs with different subclasses. Spectra have been normalized at 0.89 m. The missing wavelength region in the spectrum of 2M0532 (1.008–1.153 m) has been replaced by the best BT-Settl model fit in magenta (Teff D 1600 K, [Fe/H] = 1.6 dex, and log(g) = 5.25 dex) (Figure from Zhang et al. 2017b)

Multiplicity The multiplicity fraction and binary properties of UCSDs is poorly constrained for two main reasons. On the one hand, most of the known UCSDs have been identified very recently, and, on the other hand, they are faint both at optical and infrared wavelengths. High-resolution imaging is feasible for a limited subsample because bright reference stars are needed to close the loop as in the case of adaptive optics, for example. Only a limited number of surveys have been conducted to look at the multiplicity of metal-poor M dwarfs over a wide range of separations. Jao et al. (2009) found that the multiplicity rate of K and M subdwarfs is slightly lower than their solar-

27 Metal-Depleted Brown Dwarfs

583

type counterparts (26 ˙ 6% vs 36 ˙ 5%) from an optical speckle survey. The total multiplicity fraction of M subdwarfs can be divided up as follows: 3% have companions within 10 au, another 3% within the 10–100 au range, 14% beyond 100 au, and the remaining 6% are spectroscopic binaries. Combining the outcome of a Hubble survey of 28 metal-poor M dwarfs of Riaz et al. (2008) with high spatial lucky imaging observations of 24 M subdwarfs, Lodieu et al. (2009) resolved only one system with a projected separation of 0.7 arcsec (LHS 182), deriving a binary frequency of 3:7 ˙ 2:6 (1 confidence limit) for M subdwarfs (mainly M0–M5). This result is in line with the two companions exhibiting H˛ in emission among a sample of 68 LHS objects (2.9%; Gizis 1998). Finally, only three M subdwarfs with metallicities below 0:5 dex and masses less than 0.5 Mˇ have dynamical mass measurements. They belong to two doublelined eclipsing binaries with resolved orbits: the secondary of the  Cas AB system with a mass of 0.17 Mˇ (Drummond et al. 1995) and both components of the G 006– 026 BC system whose masses span 0.43–0.47 Mˇ (Jao et al. 2016).

Physical Parameters At the time of writing, no mass estimate independent of evolutionary models exists for UCSDs because they are either too faint or no eclipsing/spectroscopic binary exists for direct dynamical mass measurement. Currently, the range of physical parameters for UCSDs originates from the direct comparison of observed optical and/or near-infrared spectra with state-ofthe-art evolutionary models. Burgasser et al. (2008) identified a subdwarf with the latest spectral type reported to date (2MASS J0532C8246; sdL7), lying at the stellar/substellar boundary. These authors fitted the full spectral energy distribution (SED) of 2MASS J0532C8246 with the NextGen models (Baraffe et al. 1998), deriving an effective temperature (Teff ) of 1730 ˙ 90 K and a mass in the range 0.0744–0.0835 Mˇ for metallicities between 0 and 2:0 dex and ages of 10–15 Gyr. However, the lithium feature at 6707.8 Å has not been detected in higher-resolution spectra (Lodieu et al. 2015), suggesting a minimum mass of 0.06 Mˇ (Magazzu et al. 1992; Rebolo et al. 1996). Burgasser et al. (2009) repeated a similar process for a warmer L subdwarf classified as sdL4 (Sivarani et al. 2009) and derived a mean Teff of 2300 ˙ 200 K, log(g) = 5.0–5.5 dex, and metallicity between 1:0 and 1:5 dex using the Drift-Phoenix models (Helling et al. 2008). Zhang et al. (2017b) extended such a procedure to six new L subdwarfs identified in the cross-match of SDSS and UKIDSS as well as all previously known L subdwarfs as of 2017. They determined Teff and metallicities for 22 L subdwarfs, extreme subdwarfs, and ultrasubdwarfs with spectral types in the L0– L7 range by direct comparison with the BT-Settl models (Allard et al. 2012), yielding temperatures of 1500–2700 K for metallicities between 1:0 and 2:0 dex. On average, the Teff of subdwarfs are 100–400 K higher than solar-metallicity L dwarfs depending on the spectral subtype and metal class. Some of these L subdwarfs might be brown dwarfs rather than very low-mass stellar members of

584

N. Lodieu

the halo based on the latest BT-Settl models (Zhang et al. 2017a), but the exact location of the stellar/substellar boundary at low metallicity still requires dynamical measurements.

Future Work The past two decades have witness the existence of M, L, and T subdwarfs and the first spectral classification of this metal-poor population. The advent of large-scale surveys at both optical and infrared wavelengths (2MASS, SDSS, UKIDSS, WISE) have increased the numbers of UCSDs and extended the sequence to cooler metalpoor L and T dwarfs. However, our knowledge of UCSDs still remains in its infancy. A number of improvements and discoveries are required to take up this field to the next level. • The improvement in the determination of physical parameters of subdwarfs requires the fitting of optical and infrared spectral energy distributions of several hundreds of spectra to complement the extensive work on the properties and kinematics of subdwarfs by Savcheva et al. (2014). • The saturation of the T iO=C aH index around M8–M9 suggests that a revision of the current spectral classification scheme is needed. Moreover, a near-infrared spectral classification is also necessary to classify future discoveries in largescale infrared surveys. • The progress in the accuracy of metallicity scale for ultracool subdwarfs calls for searches of close-in and/or wide companions to brighter subdwarfs with welldetermined metallicities. • Searches for M, L, and T subdwarfs should be enhanced to improve the determination of the object density as a function of metallicity and allow for a determination of the luminosity and mass functions in a distance or magnitudelimited volume. • The discovery of short-period binaries and eclipsing binaries is heavily needed to infer model-independent masses over a wide range of masses and metallicities. The future of the field sounds bright with upcoming deep photometric surveys such as the Large Synoptic Survey Telescope (LSST; Ivezic et al. 2008) and the Euclid mission (Mellier 2016) and large-scale spectroscopic campaigns planned with WHT/WEAVE (Dalton et al. 2012) and VISTA/4MOST (de Jong et al. 2014). Let’s look forward to the first substellar subdwarfs and long-waited mass determinations! Acknowledgements NL is supported by programme AYA2015-69350-C3-2-P from Spanish Ministry of Economy and Competitiveness (MINECO). NL thanks ZengHua Zhang for his input on the review.

27 Metal-Depleted Brown Dwarfs

585

References Allard F, Homeier D, Freytag B (2012) Models of very-low-mass stars, brown dwarfs and exoplanets. R Soc Lond Philos Trans Ser A 370:2765–2777 Allers KN, Liu MC (2013) A near-infrared spectroscopic study of young field ultracool dwarfs. ApJ 772:79 Allers KN, Jaffe DT, Luhman KL et al (2007) Characterizing young brown dwarfs using lowresolution near-infrared spectra. ApJ 657:511–520 Baraffe I, Chabrier G, Allard F, Hauschildt PH (1998) Evolutionary models for solar metallicity low-mass stars: mass-magnitude relationship and color-magnitude diagrams 337:403–412. http://adsabs.harvard.edu/cgi-bin/nph-bib_query?bibcode=1998A%26A...337.. 403B&db_key=AST&high=3d1b390e2520250 Bowler BP, Liu MC, Dupuy TJ (2010) SDSS J141624.08+134826.7: a nearby blue L dwarf from the sloan digital sky survey. ApJ 710:45–50 Burgasser AJ (2004) Discovery of a second L subdwarf in the two micron all sky survey. ApJ 614:L73–L76 Burgasser AJ, Kirkpatrick JD (2006) Discovery of the coolest extreme subdwarf. ApJ 645: 1485–1497 Burgasser AJ, Kirkpatrick JD, Brown ME et al (2002) The spectra of T dwarfs. I. Near-infrared data and spectral classification. ApJ 564:421–451 Burgasser AJ, Kirkpatrick JD, Burrows A et al (2003) The first substellar subdwarf? Discovery of a metal-poor L dwarf with halo kinematics. ApJ 592:1186–1192 Burgasser AJ, Geballe TR, Leggett SK, Kirkpatrick JD, Golimowski DA (2006) A unified nearinfrared spectral classification scheme for T dwarfs. ApJ 637:1067–1093 Burgasser AJ, Cruz KL, Kirkpatrick JD (2007) Optical spectroscopy of 2MASS color-selected ultracool subdwarfs. ApJ 657:494–510 Burgasser AJ, Vrba FJ, Lépine S et al (2008) Parallax and luminosity measurements of an L subdwarf. ApJ 672:1159–1166 Burgasser AJ, Witte S, Helling C et al (2009) Optical and near-infrared spectroscopy of the L subdwarf SDSS J125637.13-022452.4. ApJ 697:148–159 Burningham B, Cardoso CV, Smith L et al (2013) 76 T dwarfs from the UKIDSS LAS: benchmarks, kinematics and an updated space density. MNRAS 433:457–497 Chabrier G (2003) Galactic stellar and substellar initial mass function. PASP 115:763–795 Cruz KL, Kirkpatrick JD, Burgasser AJ (2009) Young L dwarfs identified in the field: a preliminary low-gravity, optical spectral sequence from L0 to L5. AJ 137:3345–3357 Cushing MC, Looper D, Burgasser AJ et al (2009) 2MASS J06164006–6407194: the first outer halo L subdwarf. ApJ 696:986–993 Cushing MC, Kirkpatrick JD, Gelino CR et al (2011) The discovery of Y dwarfs using data from the wide-field infrared survey explorer (WISE). ApJ 743:50 Dalton G, Trager SC, Abrams DC et al (2012) WEAVE: the next generation wide-field spectroscopy facility for the William Herschel telescope. In: Ground-based and airborne instrumentation for astronomy IV. Proceedings of SPIE, vol 8446, p 84460P. https://doi.org/ 10.1117/12.925950 de Jong RS, Barden S, Bellido-Tirado O et al (2014) 4MOST: 4-metre multi-object spectroscopic telescope. In: Ground-based and airborne instrumentation for astronomy V. Proceedings of SPIE, vol 9147, p 91470M. https://doi.org/10.1117/12.2055826 Drummond JD, Christou JC, Fugate RQ (1995) Full adaptive optics images of ADS 9731 and MU Cassiopeiae: orbits and masses. ApJ 450:380 Geballe TR, Knapp GR, Leggett SK et al (2002) Toward spectral classification of L and T dwarfs: infrared and optical spectroscopy and analysis. ApJ 564:466–481 Gizis JE (1997) M-Subdwarfs: spectroscopic classification and the metallicity scale. AJ 113: 806–822

586

N. Lodieu

Gizis JE (1998) High chromospheric activity in M subdwarfs. AJ 115:2053–2058 Gizis JE, Reid IN (1997) Probing the LHS catalog. I. New nearby stars and the coolest subdwarf. PASP 109:849–856 Gizis JE, Scholz RD, Irwin M, Jahreiss H (1997) APMPM J1523-0245 – a new high proper motion cool subdwarf. MNRAS 292:L41–L43 Helling C, Ackerman A, Allard F et al (2008) A comparison of chemistry and dust cloud formation in ultracool dwarf model atmospheres. MNRAS 391:1854–1873 Ivezic Z, Axelrod T, Brandt WN et al (2008) Large synoptic survey telescope: from science drivers to reference design. Serbian Astron J 176:1–13 Jao WC, Henry TJ, Beaulieu TD, Subasavage JP (2008) Cool subdwarf investigations. I. New thoughts on the spectral types of K and M subdwarfs. AJ 136:840–880 Jao WC, Mason BD, Hartkopf WI, Henry TJ, Ramos SN (2009) Cool subdwarf investigations. II. Multiplicity. AJ 137:3800–3808 Jao WC, Nelan EP, Henry TJ, Franz OG, Wasserman LH (2016) Cool subdwarf investigations. III. Dynamical masses of low-metallicity subdwarfs. AJ 152:153 Joy AH (1947) Radial velocities and spectral types of 181 dwarf stars. ApJ 105:96 Kirkpatrick JD, Reid IN, Liebert J et al (1999) Dwarfs cooler than “M”: the definition of spectral type “L” using discoveries from the 2 micron all-sky survey (2MASS). ApJ 519:802–833. http:// adsabs.harvard.edu/cgi-bin/nph-bib_query?bibcode=1999ApJ...519..802K&db_key=AST Kirkpatrick JD, Reid IN, Liebert J et al (2000) 67 additional L dwarfs discovered by the two micron all sky survey. AJ 120:447–472. http://adsabs.harvard.edu/cgi-bin/nph-bib_query?bibcode= 2000AJ....120..447K&db_key=AST Kirkpatrick JD, Gelino CR, Cushing MC et al (2012) Further defining spectral type “Y” and exploring the low-mass end of the field brown dwarf mass function. ApJ 753:156 Kirkpatrick JD, Schneider A, FAJardo-Acosta S et al (2014) The AllWISE motion survey and the quest for cold subdwarfs. ApJ 783:122 Kirkpatrick JD, Kellogg K, Schneider AC et al (2016) The AllWISE motion survey, part 2. ApJS 224:36 Leggett SK, Geballe TR, Fan X et al (2000) The missing link: early methane (“T”) dwarfs in the sloan digital sky survey. ApJ 536:L35–L38 Lépine S, Scholz RD (2008) Twenty-three new ultracool subdwarfs from the sloan digital sky survey. ApJ 681:L33–L36 Lépine S, Shara MM, Rich RM (2003) Discovery of an ultracool subdwarf: LSR 1425+7102, the first star with spectral type sdM8.0. ApJ 585:L69–L72 Lépine S, Rich RM, Shara MM (2007) Revised metallicity classes for low-mass stars: dwarfs (dM), subdwarfs (sdM), extreme subdwarfs (esdM), and ultrasubdwarfs (usdM). ApJ 669:1235–1247 Lodieu N, Scholz RD, McCaughrean MJ et al (2005) Spectroscopic classification of red high proper motion objects in the southern sky. A&A 440:1061–1078 Lodieu N, Zapatero Osorio MR, Martín EL (2009) Lucky imaging of M subdwarfs. A&A 499: 729–736 Lodieu N, Zapatero Osorio MR, Martín EL, Solano E, Aberasturi M (2010) GTC/OSIRIS spectroscopic identification of a faint L subdwarf in the UKIRT infrared deep sky survey. ApJ 708:L107–L111 Lodieu N, Espinoza Contreras M, Zapatero Osorio MR et al (2012) New ultracool subdwarfs identified in large-scale surveys using virtual observatory tools. I. UKIDSS LAS DR5 vs. SDSS DR7. A&A 542:A105 Lodieu N, Burgasser AJ, Pavlenko Y, Rebolo R (2015) A search for lithium in metal-poor L dwarfs. A&A 579:A58 Lodieu N, Espinoza Contreras M, Zapatero Osorio MR et al (2017) New ultracool subdwarfs identified in large-scale surveys using virtual observatory tools. A&A 598:A92 Mace GN, Kirkpatrick JD, Cushing MC et al (2013) The exemplar T8 subdwarf companion of wolf 1130. ApJ 777:36 Magazzu A, Rebolo R, Pavlenko IV (1992) Lithium abundances in classical and weak T Tauri stars. ApJ 392:159–171

27 Metal-Depleted Brown Dwarfs

587

Martín EL, Delfosse X, Basri G et al (1999) Spectroscopic classification of late-M and L field dwarfs. AJ 118:2466–2482. http://adsabs.harvard.edu/cgi-bin/nph-bib_query?bibcode= 1999AJ....118.2466M&db_key=AST Mellier Y (2016) Euclid and the dark universe. In: 41st COSPAR scientific assembly, abstracts from the meeting that was to be held 30 July–7 Aug at the Istanbul Congress Center (ICC), Turkey, but was cancelled. See http://cospar2016.tubitak.gov.tr/en/, Abstract H0.2-1-16., COSPAR Meeting, vol 41 Morgan WW, Keenan PC, Kellman E (1943) An atlas of stellar spectra, with an outline of spectral classification. The University of Chicago press, Chicago Pinfield DJ, Burningham B, Lodieu N et al (2012) Discovery of the benchmark metal-poor T8 dwarf BD +01deg2920B. MNRAS 422:1922–1932 Rebolo R, Martín EL, Basri G, Marcy GW, Zapatero-Osorio MR (1996) Brown dwarfs in the Pleiades cluster confirmed by the lithium test. ApJ 469:L53 Reid IN, Hawley SL (2005) New light on dark stars: red dwarfs, low-mass stars, brown dwarfs. https://doi.org/10.1007/3-540-27610-6 Riaz B, Gizis JE, Samaddar D (2008) Hubble space telescope search for M subdwarf binaries. ApJ 672:1153–1158 Savcheva AS, West AA, Bochanski JJ (2014) A new sample of cool subdwarfs from SDSS: properties and kinematics. ApJ 794:145 Schmidt M (1975) The mass of the galactic halo derived from the luminosity function of highvelocity stars. ApJ 202:22–29 Schmidt SJ, West AA, Burgasser AJ, Bochanski JJ, Hawley SL (2010) Discovery of an unusually blue L dwarf within 10 pc of the sun. AJ 139:1045–1050 Scholz RD, Lehmann I, Matute I, Zinnecker H (2004a) The nearest cool white dwarf (d4 pc), the coolest M-type subdwarf (sdM9.5), and other high proper motion discoveries. A&A 425: 519–527 Scholz RD, Lodieu N, McCaughrean MJ (2004b) SSSPM J1444–2019: an extremely high proper motion, ultracool subdwarf. A&A 428:L25–L28 Schweitzer A, Scholz RD, Stauffer J, Irwin M, McCaughrean MJ (1999) APMPM J0559–2903: the coolest extreme subdwarf known. A&A 350:L62–L64 Sivarani T, Lépine S, Kembhavi AK, Gupchup J (2009) SDSS J125637–022452: a high proper motion L subdwarf. ApJ 694:L140–L143 Zhang ZH, Pinfield DJ, Burningham B et al (2013) A spectroscopic and proper motion search of sloan digital sky survey: red subdwarfs in binary systems. MNRAS 434:1005–1027 Zhang ZH, Homeier D, Pinfield DJ et al (2017a) Primeval very low-mass stars and brown dwarfs – II. The most metal-poor substellar object. MNRAS 468:261–271 Zhang ZH, Pinfield DJ, Gálvez-Ortiz MC et al (2017b) Primeval very low-mass stars and brown dwarfs – I. Six new L subdwarfs, classification and atmospheric properties. MNRAS 464:3040– 3059

Radio Emission from Ultracool Dwarfs

28

Peter K. G. Williams

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phenomenology of the Radio Emission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bright, Polarized Bursts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Non-bursting Emission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Intermediate Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . UCD Radio Emission in Context . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Prevalence of Radio Activity in UCDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Multiwavelength Correlations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interpretation of the Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Auroral Radio Emission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gyrosynchrotron Radio Emission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Emergence of “Planet-Like” Magnetism in UCDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Exoplanetary Connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Future Directions of Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

590 591 591 594 595 596 596 598 600 600 602 603 604 605 606

Abstract

The 2001 discovery of radio emission from ultracool dwarfs (UCDs), the very low-mass stars and brown dwarfs with spectral types of M7 and later, revealed that these objects can generate and dissipate powerful magnetic fields. Radio observations provide unparalleled insight into UCD magnetism: detections extend to brown dwarfs with temperatures .1000 K, where no other observational probes are effective. The data reveal that UCDs can generate strong (kG) fields, sometimes with a stable dipolar structure; that they can produce and retain nonthermal plasmas with electron acceleration extending to MeV

P. K. G. Williams () Harvard-Smithsonian Center for Astrophysics, Cambridge, MA, USA e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_171

589

590

P. K. G. Williams

energies; and that they can drive auroral current systems resulting in significant atmospheric energy deposition and powerful, coherent radio bursts. Still to be understood are the underlying dynamo processes, the precise means by which particles are accelerated around these objects, the observed diversity of magnetic phenomenologies, and how all of these factors change as the mass of the central object approaches that of Jupiter. The answers to these questions are doubly important because UCDs are both potential exoplanet hosts, as in the TRAPPIST1 system, and analogues of extrasolar giant planets themselves. Keywords

Brown dwarfs · Ultracool dwarfs · Radio emission · Magnetic activity · Dynamo

Introduction The process that generates the solar magnetic field is called the dynamo. It is widely believed to depend on the tachocline, the shearing layer between the Sun’s radiative inner core and its convective outer envelope (e.g., Charbonneau 2014). As stellar masses drop below 0:35 Mˇ (spectral types M3.5 and later), the tachocline disappears (Limber 1958; Chabrier and Baraffe 2000), which made it challenging to explain how mid-M dwarf stars can in fact generate strong magnetic fields (Saar and Linsky 1985). The surprising magnetic properties of fully convective M dwarfs raised the question of what dynamo action would be like in the coolest, lowest-mass objects: the ultracool dwarfs (UCDs), stars and brown dwarfs with spectral types M7 and later (Kirkpatrick et al. 1999; Martín et al. 1999). (The very youngest and most massive brown dwarfs have spectral types M7; the very lowest-mass stars have spectral types L4. Objects with spectral types between these limits can be of either category.) But it was not until the CCD revolution that it became possible to study UCDs systematically. The first results suggested that magnetic activity faded out in the UCDs (e.g., Drake et al. 1996; Basri and Marcy 1995). The consensus model was that magnetic field generation became ineffective in the lowest-mass objects due to the loss of the Sun-like “shell” dynamo and the transition to cool outer atmospheres, expected to be largely neutral and therefore unable to couple the energy of their convective motions into any fields generated below the surface (Mohanty et al. 2002). This picture was muddied, however, by reports of flares from very late M dwarfs in the ultraviolet (UV; Linsky et al. 1995), H˛ (Reid et al. 1999; Liebert et al. 1999), and X-ray (Fleming et al. 2000). These results suggested that UCDs could generate and dissipate magnetic fields at least intermittently. A breakthrough occurred in 2001 with the detection of an X-ray flare from LP 944–20, a bona fide brown dwarf (M9.5; Rutledge et al. 2000), which was shortly followed by the detection of both bursting and quiescent radio emission from the same object by a team of summer students using the NRAO Very Large Array (Berger et al. 2001). Radio detections of UCDs were thought to be

28 Radio Emission from Ultracool Dwarfs

591

impossible: scaling arguments had led to radio flux density predictions of .0:1 Jy, not achievable even with present-day observatories. But Berger et al. (2001) detected LP 944–20 at a flux density 104 times brighter than these predictions, demonstrating that UCD magnetism is – at least sometimes – vigorous and of a fundamentally different nature than observed in higher-mass objects. The detection of quiescent emission further demonstrated that not only can UCDs generate stable magnetic fields but that they can also sustainably source the highly energetic, nonthermal electrons needed to produce observable radio emission. Radio observations have since proved to be the best available probe of magnetism in the UCD regime, with a major leap in capabilities coming with VLA upgrade project (Perley et al. 2011). In the rest of this chapter, we describe the phenomenology of UCD radio emission, place it in a broader astrophysical context, and deduce the implications of the data for the magnetic properties of UCDs. We close by presenting the unique contribution that studies of UCD magnetism can make to exoplanetary science and probable future directions of research in the field.

Phenomenology of the Radio Emission Radio observations of UCDs have revealed a complex phenomenology that can broadly be divided into “bursting” and “non-bursting” components. The nonbursting components can also be variable and evolve significantly over long timescales (large compared to the rotation period Prot ) so we prefer to use this terminology rather than refer to such emission as “quiescent.” Table 1 presents the list of all known radio-active UCDs at the time of writing.

Bright, Polarized Bursts UCDs emit bright, circularly polarized radio bursts at GHz frequencies that have durations 1–100 min. In the initial discovery by Berger et al. (2001), the radio bursts of LP 944–20 had a brightness temperature TB  1010 K and a fractional circular polarization fC  30%, consistent with synchrotron emission mechanisms (Dulk 1985). (Brightness temperature is a proxy for specific intensity often used by radio astronomers: I  2 2 kTB =c 2 .) Subsequent observations have, however, revealed cases with brightness temperatures and fractional polarizations too large to be explained by synchrotron emission. In two early examples, Burgasser and Putman (2005) detected two bursts from DENIS J104814:7395606 (M8), one with flux density S  20 mJy,   5 min, TB  1013 K, and fC  100%. Hallinan et al. (2007) detected repeated bursts from TVLM 513–46546 (M9) with S  3 mJy,   5 min, TB & 1011 K, and fC  100% with both left- and right-handed helicities observed. In many cases, these radio bursts have been observed to occur periodically, and in all such cases where the rotation period Prot is measured through independent means, the periodicity of the bursts matches Prot . Figure 1 shows a classic example of this

592

P. K. G. Williams

Table 1 The 23 radio-detected UCDs as of mid-2017. “Other name” entries ending in “J: : :” indicate position-based survey names that are nearly identical to the canonical source name; for instance, 2MASS J104814633956062 is also known as DENIS J104814:7395606. “SpT” shows a spectral type from SIMBAD; UCD spectral typing is challenging and subtle (e.g., Kirkpatrick et al. 2012), but to conserve space we omit details and references. Spectral types with asterisks ( ) are known to come from the blended spectra of more than one object. “Var?” indicates whether the source has been confirmed to have radio emission that varies on short (.1 h) timescales. This is the case for all well-studied UCDs except LP 349–25 AB (Osten et al. 2009) Source name Other name 2MASS J095221881924319 AB 2MASS J13142039C1320011 B NLTT 33370 B 2MASS J145638312809473

SpT M7 M7 M7

Var? First radio detection McLean et al. (2012) Y McLean et al. (2011) Burgasser and Putman (2005) 2MASS J00275592C2219328AB LP 349–25 AB M8 N Phan-Bao et al. (2007) 2MASS J15010818C2250020 TVLM 513–46546 M8.5 Y Berger (2002) 2MASS J18353790C3259545 LSR J1835C3259 M8.5 Y Berger (2006) 2MASS J104814633956062 DENIS J: : : M9 Y Burgasser and Putman (2005) 2MASS J002424630158201 BRI B00210214 M9.5 Y Berger (2002) 2MASS J033935213525440 LP 944–20 M9.5 Y Berger et al. (2001) 2MASS J072003250846499 AB M9.5 + T5 Y Burgasser et al. (2015) 2MASS J07464256C2000321 B L1.5 Y Berger et al. (2009) 2MASS J19064801C4011089 WISE J: : : L1 Gizis et al. (2013) 2MASS J052338221403022 L2.5 Berger (2006) 2MASS J00361617C1821104 L3.5 Y Berger (2002) 2MASS J131530942649513 AB L3.5 + T7 Burgasser et al. (2013) 2MASS J000434844044058 AB L5 + L5 Lynch et al. (2016) 2MASS J042348580414035 SDSS J: : : L7.5 Y Kao et al. (2016) 2MASS J10430758C2225236 L8 Y Kao et al. (2016) 2MASS J06073908C2429574 WISE J: : : L9 Gizis et al. (2016) 2MASS J01365662C0933473 SIMP J: : : T2.5 Y Kao et al. (2016) WISEP J112254:73C255021:5 T6 Y Route and Wolszczan (2016) 2MASS J10475385C2124234 T6.5 Y Route and Wolszczan (2012) 2MASS J12373919C6526148 T6.5 Y Kao et al. (2016)

phenomenology from Berger et al. (2009). In the objects with such measurements, 2 . Prot . 4 h, but there are likely significant selection effects at play that make it difficult to infer the true distribution of Prot of the radio-active UCDs. In objects with repeated observations, the periodic bursts are sometimes present and sometimes not (e.g., LSR J1835C3259; Berger et al. 2008a; Hallinan et al. 2008). TVLM 513–46546 is the best-studied member of this class, with burst observations spanning years that enable claims of extremely precise (millisecond) determinations

28 Radio Emission from Ultracool Dwarfs

593

Flux Density (mJy)

15

4.86 GHz Stokes I

10 5 0 0 –5

–10 Stokes V

–15

Flux Density (mJy)

2

3

4

5

7 6 UT Time (hr)

8

9

6 4 2 0

10

8.46 GHz

Stokes I 6 4 2 0 Stokes V 2

3

4

5

6 7 UT Time (hr)

8

9

10

Fig. 1 [From Berger et al. (2009). Reproduced by permission of the AAS.] Radio light curve of 2MASS J07464256C2000321 B showing periodic, highly polarized, rapid, bright bursts. The black and red points show the data averaged into 5- and 60-sec bins, respectively. The negative Stokes V values, jV j  I , indicate 100% left circular polarization in the bursts. The burst spectra do not extend to the VLA’s 8.46 GHz band (lower panels)

of the rotation period (Doyle et al. 2010; Harding et al. 2013a; Wolszczan and Route 2014). These bursts have been generally been detected at frequencies between 1 and 10 GHz. Once again, selection effects make it difficult to draw conclusions about the fundamental character of the burst spectra given the observational results: the vast majority of searches for UCD radio emission of have been conducted in the 1– 10 GHz frequency window. This window is where the VLA’s sensitivity peaks, but it is challenging to quantify how important intrinsic effects are as well (we observe in this window because there truly are more bursts to be seen in it). The spectral shapes of the bursts are not fully understood. Both high- and low-frequency cutoffs have been observed in different bursts (Lynch et al. 2015; Williams et al. 2015a), but in no burst has there been definitive evidence that the flux density peak has

594

P. K. G. Williams

been identified. Later in this chapter we will argue that the bursts are probably of moderate bandwidth, =  1. The total energy contained in the bursts is not large, which is commonly the case for radio processes. Using the properties of the bursts from TVLM 513–46546 quoted above (Hallinan et al. 2007), the energy content of an individual burst is 1027 erg, assuming isotropic emission. For coherent emission processes, the emission is unlikely to be isotropic, reducing the energy budget further. The burst luminosities are typically  106 of the bolometric (sub)stellar radiative output.

Non-bursting Emission UCDs also produce non-bursting radio emission that is generally steady over the timescales of individual observations. Repeated observations of numerous UCDs have revealed, however, that this emission often varies at the order-of-magnitude level on longer (week and above) timescales (e.g., Antonova et al. 2007; McLean et al. 2012). Several UCDs have been detected once in the radio and not detected in deeper follow-up observations (e.g., McLean et al. 2012). On the other hand, archival detections show that the hyperactive M7 star NLTT 33370 B has sustained a broadly consistent level of radio emission for at least a decade (McLean et al. 2011). Figure 2 shows that this object, the most radio-bright UCD, nonetheless displays both periodic (at Prot ) and long-term variability in its radio emission. Radio-detected UCDs typically have non-bursting spectral luminosities of L;R 1012 –1014 erg s1 Hz1 , usually about an order-of-magnitude fainter than the peak observed burst luminosity when both phenomena have been observed. Selection effects are important here, too: the lower bound of this range corresponds to the sensitivity that is achieved in typical VLA reconnaissance observations (1 h duration) of nearby (10 pc) UCDs. The deepest upper limit on a UCD is 1011 erg s1 Hz1 , obtained in observations of the nearby binary Luhman 16 AB (Osten et al. 2015). The brightest UCD radio emitter, NLTT 33370 B, reaches 1014:7 erg s1 Hz1 (McLean et al. 2011; Williams et al. 2014). The non-bursting emissions generally have low or moderate circular polarization. Linear polarization has not been detected. As shown in Fig. 2, 0 < fC < 20% in the case of NLTT 33370 B, with periodic variability at Prot indicating that the apparent circular polarization depends on orientation. The recently discovered radio-active T6.5 dwarf WISEP J112254:73C255021:5 presents a new, unusual case: unlike the other UCDs, WISEP J112254:73C255021:5 produces highly polarized emission that is not clearly confined to rapid bursts (Williams et al. 2017). The only published observations of this object are too brief, however, to allow a firm interpretation. The non-bursting spectra are broadband. They peak around 1–10 GHz and generally have shallow spectral indices on both the low- and high-frequency sides of the peak. Only a few UCDs have been observed at a wide range of radio frequencies, however. TVLM 513–46546 has been detected at frequencies ranging from 1.4 GHz all the way to 98 GHz; the latter detection was achieved with ALMA and represents the first demonstration that UCDs can be detected at

28 Radio Emission from Ultracool Dwarfs

595

Fig. 2 [From Williams et al. (2015a). Reproduced by permission of the AAS.] Radio light curve of NLTT 33370 B showing periodic variation and moderate polarization in the non-bursting radio emission. In the upper panels, filled and empty points show Stokes I and V components, respectively. The lower panels show the fractional circular polarization derived from these values. The leftmost and center panel show two observations separated by 24 h; the rightmost panel shows observations made 1 year later. Vertical black lines indicate times that the dwarf’s periodically modulated optical emission reaches maximum. Rapid, 100% circular polarized radio bursts have been excised from these data

millimeter wavelengths (Williams et al. 2015b). NLTT 33370 B has been detected from 1–40 GHz (McLean et al. 2011; Williams et al. 2015a) and has an extremely flat spectrum, with significant circular polarization at all observed frequencies. DENIS J104814:7395606 has been detected from 5–18 GHz with a negative spectral index ˛ D 1:71 ˙ 0:09 (S /  ˛ ; Ravi et al. 2011). Searches for emission from UCDs at frequencies below 1 GHz have thus far been unsuccessful (Jaeger et al. 2011; Burningham et al. 2016) although the famous low-mass flare star UV Cet (M6) was recently detected at 154 MHz using the Murchison Widefield Array (Lynch et al. 2017).

Intermediate Cases It is not always possible to cleanly separate UCD radio emission into bursting and non-bursting components. Figure 3 shows an example from TVLM 513–46546 in which variability is observed with both circular polarization helicities and null polarization (Hallinan et al. 2006). 2MASS J00361617C1821104 and NLTT 33370 B have shown similarly ambiguous phenomenologies (Berger et al. 2005; Hallinan et al. 2008; Williams et al. 2015a).

596

P. K. G. Williams

Fig. 3 [From Hallinan et al. (2006). Reproduced by permission of the AAS.] Radio light curve of TVLM 513–46546 showing periodic behavior that is not cleanly separable into burst and non-burst components. The data are phased to a period of 2 h and shown binned at 6, 7, and 8 min, with each binned light curve being plotted twice. The observing frequency was 4.88 GHz.

UCD Radio Emission in Context The previous section focused narrowly on the properties of the radio emission detected from UCDs. In this section, we place this emission in a broader astrophysical context.

The Prevalence of Radio Activity in UCDs Volume-limited radio surveys of UCDs achieve a detection rate of approximately 10% (Berger 2006; McLean et al. 2012; Antonova et al. 2013; Lynch et al. 2016). However, recent work by Kao et al. (2016) demonstrates that biased surveys can achieve a substantially higher detection rate: in a sample of five late-L and T dwarfs selected to have prior detections of H˛ emission or optical variability, four of the

28 Radio Emission from Ultracool Dwarfs

597

targets were detected. These findings are consistent because the H˛ detection rate of L and T dwarfs is also about 10% on average, with a noticeably higher detection rate for objects warmer than L5 (Pineda et al. 2016). This “headline number” comes with three important caveats. First, it derives from an observer-dependent binary classification (“did the object’s apparent radio flux density have sufficiently high S/N?”) rather than a fundamental physical measurement (“what is the object’s radio spectral luminosity?”). Second, the radio detectability of individual objects varies over time in ways that are not well understood. Third, the reported number averages across a wide variety of objects, while studies of FGKM dwarfs lead us to expect that activity strength should depend strongly on fundamental (sub)stellar parameters. In particular, mass, rotation, and age are generally believed to be the most important for setting stellar activity levels (e.g., Barnes 2003; Wright et al. 2011). Correlations between fundamental parameters are pervasive, however, so it is challenging to determine causation (e.g., Reiners et al. 2014). Below we consider how UCD radio emission scales with some of these physical parameters, considering only the radio-detected objects. A proper analysis of the entire radio-observed UCD sample that takes into account nondetections has yet to be performed. Numerous UCDs have upper limits on their radio emission that are inconsistent with the trends described.

Mass, Spectral Type, and Effective Temperature Because brown dwarfs do not evolve to a stable main sequence and direct mass measurements of astronomical objects are difficult to obtain, spectral type (SpT) is widely used as a proxy for mass in UCD activity studies. The magnetic activity levels of FGKM stars are often quantified with the ratio of the stellar X-ray luminosity to bolometric luminosity (LX =Lbol ; e.g., Wright et al. 2011). This ratio decreases as SpT increases (that is, moves toward cooler Teff ) even though Lbol on its own scales strongly with Teff , implying a significant drop in the un-normalized LX (Stelzer et al. 2006; Berger et al. 2010; Williams et al. 2014). It is therefore striking that in UCDs, L;R shows only a mild decrease with SpT, with typical values of 1013:5 erg s1 Hz1 at M7 and 1012:5 erg s1 Hz1 in the T dwarfs (Gizis et al. 2016, their Figure 8). Over this range of SpTs, L;R =Lbol increases from typical values of 1017 to 1016 Hz1 . Rotation Magnetically active FGKM stars follow a “rotation/activity relation” in which the level of magnetic activity increases with increasing rotation rate up until a “saturation point,” past which further increases in rotation rate do not affect the level of magnetic activity (e.g., Wright et al. 2011). Here the level of magnetic activity is most commonly quantified with LX =Lbol , but analogous trends are observed in most other measurements that trace activity. The nature of the radio rotation/activity relationship in UCDs is more ambiguous. Plots of L;R =Lbol against rotation show a scaling relationship that has no sign of a saturation point (McLean et al. 2012). However, the fastest rotators tend to be

598

P. K. G. Williams

the objects with the latest spectral types, introducing a covariance with the mass trend described above. Cook et al. (2014) studied a subset of UCDs at a relatively narrow range of SpT, M6.5–M9.5, and found weak evidence that LX =Lbol is in fact anticorrelated with rotation rate.

Age Sun-like stars become less active as they age since they shed angular momentum through their winds (Skumanich 1972). This process becomes much less efficient as stellar mass decreases, with the average activity lifetime of M dwarfs going from 1 Gyr for M0–M2 stars to 8 Gyr for M5–M7 stars (West et al. 2008). The data suggest that brown dwarfs rotate rapidly for their entire lives (Bouvier et al. 2014). The relation between age and radio activity has not been studied systematically in UCDs. However, several noteworthy radio-active UCDs have age constraints, including LP 944–20 (500 Myr; Tinney and Reid 1998), NLTT 33370 B (80 Myr; Dupuy et al. 2016), and SIMP J01365662C0933473 (200 Myr; Gagné et al. 2017). Very young UCDs can also have radio emission associated with youngstar phenomena such as accretion, jets, and disks (e.g., Rodriguez et al. 2017).

Multiwavelength Correlations Stellar magnetism is associated with emission across the electromagnetic spectrum, and different bands probe different physical regions or processes. In Sun-like stars, H˛ emission probes the chromosphere; UV, the transition region; X-rays, hot dense coronal plasma; and radio/millimeter emission, particle acceleration. Multiwavelength observations, especially simultaneous ones, therefore yield insights that cannot be obtained through single-band studies. The radio and X-ray luminosities of active stars are nearly linearly correlated, a phenomenon known as the “Güdel-Benz relation” (Güdel and Benz 1993; Benz and Güdel 1994). A single power law can fit observations spanning ten orders of magnitude in L;R , in systems ranging in size from individual solar flares to active binaries (Fig. 4, gray points). As shown in Fig. 4, however, the Güdel-Benz relation breaks down dramatically in the UCD regime (Figure 4, colored points; Berger et al. 2001; Williams et al. 2014). Correlations between the luminosities of UCDs in radio and other bands (e.g., H˛) have not yet been investigated in the literature. Simultaneous multiwavelength observations can illuminate the physics of stellar and substellar flares, although extensive observations of flare stars demonstrate that very few general statements can be made: individual events may or may not be associated with emission in each of the bands that trace magnetic activity, and the relative ordering and magnitude of the emission in these bands are variable (e.g., Osten et al. 2004). The UCD with the best simultaneous multiwavelength observational coverage is NLTT 33370 B, and the data show a similar variety of phenomenologies (Williams et al. 2015a). A detailed understanding of the underlying physics remains elusive.

28 Radio Emission from Ultracool Dwarfs

599

Fig. 4 [From Williams et al. (2014). Reproduced by permission of the AAS.] Radio and X-ray emission for active stars and brown dwarfs. Gray points and the red line show the “Güdel-Benz relation” defined for active stars and solar flares. Green, red, and blue points show data for M3–M6, M6.5–M9.5, and L0 dwarfs, respectively. While some UCDs may obey the Güdel-Benz relation, there is a substantial population of outliers with radio emission that far exceeds what would be predicted from their X-ray emission

The evidence that optical/IR variability is a useful indicator of UCD radio activity (Kao et al. 2016) suggests that the two are correlated. Only a handful of UCDs have data sets that allow the optical and radio variability (either bursts or non-bursting periodic variations) to be phased. While the radio and optical maxima of TVLM 513–46546 are significantly out of phase (Wolszczan and Route 2014; Miles-Páez et al. 2015), there is a hint that the millimeter and optical maxima may occur at the same phase (Williams et al. 2015b). The radio and optical maxima of NLTT 33370 B are also significantly out of phase. Intriguingly, long-term monitoring of this object suggests that its non-bursting polarized radio emission remains in phase with its optical variability, but the total radio intensity does not (Williams et al. 2015a). Keck spectroscopic monitoring of LSR J1835C3259 revealed periodic variations in the optical emission that

600

P. K. G. Williams

were argued to originate in a high-altitude opaque blackbody with T 2200 K (Hallinan et al. 2015). Radio and H˛ variability can also be correlated. In 2MASS J07464256 C 2000321 B, the radio bursts are 90ı out of phase with the maxima of periodic changes in the H˛ equivalent width (Berger et al. 2009). Recent observations of LSR J1835C3259 showed radio and H˛ variations that were approximately in phase (Hallinan et al. 2015), but other observations of the same object have shown aperiodic H˛ variability with no clear connection to the radio emission (Berger et al. 2008a). Simultaneous multiwavelength monitoring of TVLM 513–46546 revealed periodic H˛ variability with no clear connection to emission in other bands, although there is some evidence for radio bursts at the times of the H˛ minima (Berger et al. 2008b).

Interpretation of the Data We now turn to the astrophysical interpretation of the observations presented in the previous sections.

Auroral Radio Emission The periodic, bright, highly polarized radio bursts observed in radio-active UCDs are consistent with the auroral radio bursts observed in Solar System planets (Zarka et al. 2001), which are generally agreed to originate from the electron cyclotron maser instability (ECMI; Wu and Lee 1979; Treumann 2006). The ECMI converts the free energy of a magnetized plasma into electromagnetic waves through resonant interactions between the waves and the particles’ cyclotron motion. The ECMI is relatively easy to trigger in physical systems involving beams of mildly relativistic electrons that are accelerated along magnetic field lines by the presence of a coaligned electric field, if the ambient medium is of sufficiently low density. This happens at the Earth when energetic solar wind particles funnel down its magnetic field lines toward the poles. Observable ECMI emission is expected to be dominated by a narrow-band signal at the electron cyclotron frequency of the local magnetic field, ce D

  B eB 2:8 MHz: 2me c 1 G

(1)

Observations of ECMI bursts from UCDs therefore measure the strengths of their magnetic fields. In practice, the ECMI occurs in regions that span a variety of field strengths, so the observed emission has a moderate bandwidth, =1 (Zarka et al. 2001), with a cutoff at high frequencies because the body’s magnetic field reaches some peak value at its surface. ECMI emission is beamed and likely refracts through the plasmasphere that evidently envelops the radio-active UCDs,

28 Radio Emission from Ultracool Dwarfs

601

necessitating detailed simulations to predict its observed properties (e.g., Kuznetsov et al. 2012; Yu et al. 2012). At a typical VLA observing frequency of 5 GHz, the inferred strength is 2 kG, comparable to the strongest surface field strengths observed on active M dwarfs (Kochukhov et al. 2017). A polarized pulse at 10 GHz from the T6.5 dwarf 2MASS J10475385C2124234 implies a field strength of at least 3.6 kG (Williams and Berger 2015), demonstrating that the fully convective dynamo can generate strong fields even in extremely low-mass, cool (900 K) objects. Observations of multiple consecutive ECMI bursts at the rotation period imply the presence of a relatively stable “electrodynamic engine” that accelerates the beams of electrons responsible for the emission. Understanding the nature of this engine is one of the great tasks in the field of UCD magnetism. In the Solar System planets, the engine is often powered by the solar wind (e.g., Dungey 1961; Axford 1969), but this driver is not available for solivagant UCDs. The only persuasive explanation is that the engine is ultimately powered by the body’s rotation (Schrijver 2009). This is largely the case for Jupiter (McComas and Bagenal 2007), raising the exciting possibility that sophisticated models developed in the context of the Solar System gas giants can be brought to bear on the UCD case. For instance, studies of Jupiter inform a model in which rotational energy is converted into nonthermal particle acceleration through shear-induced currents at the corotation breakdown radius (Nichols et al. 2012). Rapid rotation and the stable operation of the electrodynamic engine imply that the magnetospheres of radio-active UCDs likely have a dipole-dominated topology. This inference is supported by observations that probe the topology of the magnetic fields of cool stars. Studies using Zeeman Doppler imaging (ZDI; Semel 1989) show that strong, axisymmetric, dipolar fields emerge in the coolest M dwarfs currently accessible to the technique (Morin et al. 2010) and that such a topology may be associated with enhanced radio activity and variability (Kochukhov and Lavail 2017). Auroral electron beams do not only produce radio emission. First, auroral processes are associated with emission across the electromagnetic spectrum, with the highest luminosities concentrated at FUV and IR wavelengths (Bhardwaj and Gladstone 2000). However, the emission at these wavelengths is not nearly as bright as it is in the radio, such that the auroral fluxes in other bands inferred for known active UCDs are beyond the capabilities of present-day instruments. Second, the energetic auroral electrons eventually precipitate into the upper atmosphere, where they can drive chemical processes like haze production (e.g., Wong et al. 2003). Hallinan et al. (2015) interpreted their simultaneous radio and optical observations in this framework, arguing that an electron beam delivering 1024 –1026 erg s1 of kinetic power drove both the radio emission of LSR J1835C3259 and its optical variability by creating a compact, high-altitude layer of H upon precipitation. This model also motivated the targeted survey of Kao et al. (2016), under the assumption that auroral electron beams cause detectable H˛ and/or optical variability. A recent study, however, does not find a correlation between H˛ and high-amplitude optical variability in a large sample of L/T dwarfs (Miles-Páez et al. 2017).

602

P. K. G. Williams

Gyrosynchrotron Radio Emission Non-bursting UCD radio emission bears the hallmarks of gyrosynchrotron emission, the same process that is believed to be responsible for the bulk of the radio emission observed from active stars (Dulk 1985; Güdel 2002). Gyrosynchrotron emission is produced by mildly relativistic electrons spiraling in an ambient magnetic field, resulting in a broadband spectrum with low to moderate circular polarization. Analysis of the spectral properties can constrain the ambient magnetic field strength, the total number and volume density of energetic particles, and their energy distribution. It has been argued that the non-bursting UCD radio emission may instead represent an unusual form of ECMI emission (Hallinan et al. 2006, 2008), but several lines of evidence, most notably the millimeter-wavelength detection of TVLM 513–46546, discourage this interpretation (Williams et al. 2015a,b). The standard equations for gyrosynchrotron emission are derived for spatially homogeneous field and particle properties (Dulk 1985). A robust result of this analysis is that the optically thick (low-frequency) side of the spectrum should have a spectral index ˛ D 5=2, much steeper than that observed for sources like NLTT 33370 B and TVLM 513–46546 (Osten et al. 2006; McLean et al. 2011). While the flat observed spectra can be reproduced qualitatively with more realistic inhomogeneous models (e.g., White et al. 1989; Trigilio et al. 2004), homogeneous models should still give a sense of the average properties of the emitting region. Spectral fits with both kinds of model suggest that the ambient field strength in the synchrotron-emitting region is 10–100 G, typical of flare stars (Berger 2006; Osten et al. 2006; Metodieva et al. 2017). Assuming standard energetic electron densities and brightness temperatures, the typical source size is a few R (Berger 2006; Williams et al. 2014). The fact that R evolves only slowly with mass in the UCD regime may help explain why L;R appears to settle at a typical value of 1013 erg s1 Hz1 in the radio-active UCDs, if the other factors that set the synchrotron radio luminosity (B and ne ) are also mass-insensitive. Radio emission is energetically insignificant, so if the particle acceleration process saturates in some way, this value of L;R could be achieved in UCDs with widely varying bolometric and spindown luminosities. Analyses of the non-bursting radio emission of UCDs have not yet begun to leverage the detailed models that have been developed for analogous systems. Magnetic chemically peculiar (MCP) stars have high masses but also possess strong, dipole-dominated magnetospheres with persistent and periodically variable radio emission. Numerical modeling of MCP particle populations can constrain the magnetospheric structure in detail (Trigilio et al. 2004; Leto et al. 2017). Even more excitingly, Jupiter’s radiation (van Allen) belts have been studied in exquisite detail and produce centimeter-wavelength emission with variability, spectra, and polarization that are highly reminiscent of the UCD observations (de Pater 1981; de Pater et al. 2003). The application of Jovian models to UCD data has the potential to yield a treasure trove of insight. For instance, the presence of Jupiter’s moons

28 Radio Emission from Ultracool Dwarfs

603

can be determined from the spectrum of its radiation belts alone (Santos-Costa and Bolton 2008), and observations made at different orientations of the planet can be combined to reconstruct the full three-dimensional structure of the belts (Sault et al. 1997).

The Emergence of “Planet-Like” Magnetism in UCDs The data show that UCDs can generate strong magnetic fields and dissipate their energy vigorously but that they do so in processes that are fundamentally different than the typical flare star phenomenology. This is demonstrated most clearly by the substantial drop in UCD X-ray emission (both LX and LX =Lbol ), violation of the Güdel-Benz relation, and the emergence of periodic, bright, highly polarized radio bursts. This can be understood as the emergence of “planet-like” magnetism in UCDs, characterized by processes that occur in a large-scale, stable, rotation-dominated magnetosphere (Schrijver 2009). These include the operation of an electrodynamic engine that accelerates auroral electron beams and sustains a population of mildly relativistic gyrosynchrotron-emitting electrons. The lack of X-ray emission indicates that Sun-like coronal heating does not occur. Historically, this has been explained as being due to the outer atmosphere becoming electrically neutral and therefore unable to couple the energy of convective motions into magnetic flux tubes (Mohanty et al. 2002). More recent work has argued that UCD atmospheres should in fact still couple to the magnetic field efficiently (Rodríguez-Barrera et al. 2015), suggesting that more detailed analysis is needed. One of the fundamental questions about this picture is why only 10% of UCDs are detected in the radio. While early thinking focused on the possible roles of inclination and rotation rate (e.g., Harding et al. 2013b), current data suggest that planet-like magnetism is only sometimes present in UCDs and that the presence or absence of planet-like behavior is not linked to any particular fundamental parameter. The most compelling evidence for this is the NLTT 33370 AB system: while NLTT 33370 B is the most radio-luminous UCD known, its binary companion is at least 30 times fainter than it, despite being nearly identical in mass, age, rotation rate, and composition (Williams et al. 2015a; Dupuy et al. 2016; Forbrich et al. 2016). Population studies show evidence for bimodality when considering the Güdel-Benz relation (Stelzer et al. 2012; Williams et al. 2014), the rotation/activity relation (Cook et al. 2014), and ZDI-derived magnetic field topologies (Morin et al. 2010). The large-scale topology of the magnetic field may be the key factor that determines whether planet-like magnetic behavior arises in a given UCD. This hypothesis is tenable because geodynamo simulations indicate that the fully convective dynamo may be bistable in the conditions encountered in the UCD regime, with identical objects sustaining either dipole-dominated or multipolar fields depending on initial conditions (Gastine et al. 2013). Recent observations provide the first

604

P. K. G. Williams

direct evidence in favor of this model: Zeeman Doppler imaging reveals that UV Cet (M6) has an axisymmetric, dipole-dominated magnetic field, while the field of its nearly identical binary companion BL Cet is weaker and non-axisymmetric (Kochukhov and Lavail 2017). Consistent with the proposed model, UV Cet is more luminous and variable in the radio than BL Cet. Detectable radio emission requires the presence of both a magnetic field and nonthermal electrons. The difference between the radio-active and radio-inactive UCDs may therefore hinge not on the field topology but on the presence of a source of plasma that can eventually produce the gyrosynchrotron and ECMI emission. In analogy with Jupiter, the 10% of UCDs that are radio-active might be the ones possessing volcanic planets resembling Io. This scenario can potentially be tested by searching for ECMI bursts that repeat periodically not at Prot but at the synodic period of the planetary orbit. No evidence of such a non-rotational periodicity has yet been reported.

The Exoplanetary Connection Radio studies of UCDs make a unique contribution to exoplanetary science because they are the only effective way to observe the magnetic properties of cool, extrasolar bodies. One reason that this is important is that UCDs may host large numbers of observationally accessible small planets, as demonstrated by the TRAPPIST-1 system (Gillon et al. 2016, 2017). Understanding UCD activity is therefore important for the same reasons that it is important for any exoplanet host star: magnetic phenomena make planet discovery more challenging (e.g., Robertson et al. 2014), and they can have a significant impact on atmospheric retention and the broader question of habitability (e.g., Jakosky et al. 2015; Shields et al. 2016). Because UCD magnetism can be so different from that of Sun-like stars and M dwarfs, its impact on habitability may differ substantially from the cases that have been investigated thus far in the literature. For instance, the detection of millimeterwavelength radiation from TVLM 513–46546 points to a surprisingly high-energy radiation environment of MeV electrons, which can produce -ray emission when they precipitate into the stellar atmosphere (Williams et al. 2015b). The moons of the Solar System gas giants should serve as useful reference points in this domain (e.g., Paty et al. 2008). UCD magnetic fields can strongly resemble those of the Solar System gas giant planets. Radio observations therefore provide insight into the magnetospheres of exoplanets themselves, which observers have been struggling to probe since well before the first confirmed exoplanet discovery (Yantis et al. 1977). Currently, exoplanetary magnetospheres can only be investigated using indirect and modeldependent means (e.g., Ekenbäck et al. 2010). Direct observations of exoplanetary magnetospheres would not only shed light on the question of habitability, but also internal structure; for instance, magnetic field generation in rocky planets may require the presence of plate tectonics (Breuer et al. 2010). The first direct

28 Radio Emission from Ultracool Dwarfs

605

measurement of the magnetic field of a planetary-mass object may already have occurred, because SIMP J01365662C0933473, detected in the radio by Kao et al. (2016), was recently argued to be a member of the 200-Myr-old Carina-Near moving group, which would give it a mass of 12:7 ˙ 1:0 MJ according to standard evolutionary models (Gagné et al. 2017).

Future Directions of Research One of the top priorities in the field of UCD radio studies is the extension of its techniques to genuine exoplanets. By analogy with Solar System examples, exoplanets are expected to have magnetic fields that are much weaker than those of UCDs, which leads to the expectation that their radio emission will occur at lower radio frequencies, .300 MHz. Fortunately the past decade has witnessed a dramatic investment in low-frequency radio arrays such as the Low Frequency Array (LOFAR), the Murchison Widefield Array (MWA), the Long-Wavelength Array (LWA), the Giant Metrewave Radio Telescope (GMRT), and the Hydrogen Epoch of Reionization Array (HERA). While the first generation of these instruments has not yielded any detections of genuine UCDs, the first positive results are starting to emerge (Lynch et al. 2017), and virtually all of these observatories are undergoing upgrades that are expected to yield significant sensitivity improvements. While many of the nearest UCDs have been surveyed by the Very Large Array, the results of Kao et al. (2016) suggest that targeted searches may be able to yield detections beyond the typical detection horizon (30 pc) for blind searches thus far. Furthermore, radio studies of southern UCDs have historically been hampered by the lack of an instrument as powerful as the VLA (latitude C34ı ). The commissioning of the MeerKAT radio telescope in South Africa (Jonas 2009), with science operations slated to begin in late 2018, will introduce a powerful new observatory in the south. MeerKAT should be especially valuable in surveys for radio emission from young, directly imaged exoplanets, which are promising targets because they are as warm, or even warmer, than the coolest UCDs with confirmed radio detections and convect vigorously. Most of these young planets are in the southern hemisphere, however, and have not been the subject of sensitive radio observations. Surveys for radio-active UCDs in both hemispheres will be transformed by the deeper insight into the natures of the stars and brown dwarfs in the solar neighborhood afforded by upcoming surveys from observatories such as Gaia, the Transiting Exoplanet Survey Satellite (TESS), and Spektr-RG, the spacecraft bearing the e-ROSITA instrument. Finally, a great deal of theoretical work remains to be done. More detailed models of the bursting and non-bursting radio emission will strengthen the astrophysical inferences that can be drawn from the radio data. The population statistics of radioactive UCDs should be understood better by a more rigorous treatment of the many nondetections and a more careful characterization of the long-term variability of their radio emission. This sort of work will lay the foundations upon which models

606

P. K. G. Williams

can be constructed that explain fundamental puzzles such as the source of the radioemitting plasma, the possible existence of a bistable dynamo, and the relationship between rotation and magnetic activity in the ultracool regime.

Cross-References  Future Exoplanet Research: Radio Detection and Characterization  Habitability in Brown Dwarf Systems  Planetary Interiors, Magnetic Fields, and Habitability

References Antonova A, Doyle JG, Hallinan G, Golden A, Koen C (2007) A&A 472(1):257. https://doi.org/ 10.1051/0004-6361%3A20077231 Antonova A, Hallinan G, Doyle JG et al (2013) A&A 549:A131. https://doi.org/10.1051/00046361/201118583 Axford WI (1969) Rev Geophys Space Phys 7:421. https://doi.org/10.1029/RG007i001p00421 Barnes SA (2003) ApJ 586:464. https://doi.org/10.1086/367639 Basri G, Marcy GW (1995) AJ 109:762. https://doi.org/10.1086/117319 Benz AO, Güdel M (1994) A&A 285:621. http://adsabs.harvard.edu/abs/1994A%26A...285..621B Berger E (2002) ApJ 572:503. https://doi.org/10.1086/340301 Berger E (2006) ApJ 648:629. https://doi.org/10.1086/505787 Berger E, Ball S, Becker KM et al (2001) Natur 410:338. https://doi.org/10.1038/35066514 Berger E, Rutledge RE, Reid IN et al (2005) ApJ 627(2):960. https://doi.org/10.1086/430343 Berger E, Basri G, Gizis JE et al (2008a) ApJ 676:1307. https://doi.org/10.1086/529131 Berger E, Gizis JE, Giampapa MS et al (2008b) ApJ 673:1080. https://doi.org/10.1086/524769 Berger E, Rutledge RE, Phan-Bao N et al (2009) ApJ 695:310. https://doi.org/10.1088/0004-637x/ 695/1/310 Berger E, Basri G, Fleming TA et al (2010) ApJ 709:332. https://doi.org/10.1088/0004-637x/709/ 1/332 Bhardwaj A, Gladstone GR (2000) RvGeo 38(3):295. https://doi.org/10.1029/1998RG000046 Bouvier J, Matt SP, Mohanty S et al (2014) In: Beuther H, Klessen RS, Dullemond CP, Henning T (eds) Protostars and planets VI. University of Arizona Press, Tucson, p 433. https://doi.org/10. 2458/azu_uapress_9780816531240-ch019 Breuer D, Labrosse S, Spohn T (2010) SSRv 152:449. https://doi.org/10.1007/s11214-009-9587-5 Burgasser AJ, Putman ME (2005) ApJ 626(1):486. https://doi.org/10.1086/429788 Burgasser AJ, Melis C, Zauderer BA, Berger E (2013) ApJL 762:L3. https://doi.org/10.1088/ 2041-8205/762/1/L3 Burgasser AJ, Melis C, Todd J et al (2015) AJ 150(6):180. https://doi.org/10.1088/0004-6256/ 150/6/180 Burningham B, Hardcastle M, Nichols JD et al (2016) MNRAS 463(2):2202. https://doi.org/10. 1093/mnras/stw2065 Chabrier G, Baraffe I (2000) ARA&A 38:337. https://doi.org/10.1146/annurev.astro.38.1.337 Charbonneau P (2014) ARA&A 52(1):251. https://doi.org/10.1146/annurev-astro-081913-040012 Cook BA, Williams PKG, Berger E (2014) ApJ 785(1):10. https://doi.org/10.1088/0004-637X/ 785/1/10 de Pater I (1981) JGR 86(A5):3397. https://doi.org/10.1029/JA086iA05p03397 de Pater I, Butler BJ, Green DA et al (2003) Icarus 163:434. https://doi.org/10.1016/S0019-1035 %2803%2900067-8

28 Radio Emission from Ultracool Dwarfs

607

Doyle JG, Antonova A, Marsh MS et al (2010) A&A 524:A15. https://doi.org/10.1051/00046361/201015274 Drake JJ, Stern RA, Stringfellow G et al (1996) ApJ 469:828. https://doi.org/10.1086/177830 Dulk GA (1985) ARA&A 23(1):169. https://doi.org/10.1146/annurev.aa.23.090185.001125 Dungey JW (1961) Phys Rev Lett 6(2):47. https://doi.org/10.1103/PhysRevLett.6.47 Dupuy TJ, Forbrich J, Rizzuto A et al (2016) ApJ 827(1):23. https://doi.org/10.3847/0004-637X/ 827/1/23 Ekenbäck A, Holmström M, Wurz P et al (2010) ApJ 709(2):670. https://doi.org/10.1088/0004637X/709/2/670 Fleming TA, Giampapa MS, Schmitt JHMM (2000) ApJ 553:372. https://doi.org/10.1086/308657 Forbrich J, Dupuy TJ, Reid MJ et al (2016) ApJ 827(1):22. https://doi.org/10.3847/0004-637X/ 827/1/22 Gagné J, Faherty JK, Burgasser AJ et al (2017) ApJL 841:1. https://doi.org/10.3847/2041-8213/ aa70e2 Gastine T, Morin J, Duarte L et al (2013) A&A 549:L5. https://doi.org/10.1051/0004-6361/ 201220317 Gillon M, Jehin E, Lederer SM et al (2016) Nature 533:221. https://doi.org/10.1038/nature17448 Gillon M, Triaud AHMJ, Demory BO et al (2017) Nature 542:7642. https://doi.org/10.1038/ nature21360 Gizis JE, Burgasser AJ, Berger E et al (2013) ApJ 779(2):172. https://doi.org/10.1088/0004-637x/ 779/2/172 Gizis JE, Williams PKG, Burgasser AJ et al (2016) AJ 152(5):123. https://doi.org/10.3847/00046256/152/5/123 Güdel M (2002) ARA&A 40:217. https://doi.org/10.1146/annurev.astro.40.060401.093806 Güdel M, Benz AO (1993) ApJL 405:L63. https://doi.org/10.1086/186766 Hallinan G, Antonova A, Doyle JG et al (2006) ApJ 653:690. https://doi.org/10.1086/508678 Hallinan G, Bourke S, Lane C et al (2007) ApJL 663:L25. https://doi.org/10.1086/519790 Hallinan G, Antonova A, Doyle JG et al (2008) ApJ 684(1):644. https://doi.org/10.1086/590360 Hallinan G, Littlefair S, Cotter G et al (2015) Nature 523:568. https://doi.org/10.1038/nature14619 Harding LK, Hallinan G, Boyle RP et al (2013a) ApJ 779(2):101. https://doi.org/10.1088/0004637X/779/2/101 Harding LK, Hallinan G, Konopacky QM et al (2013b) A&A 554:A113. https://doi.org/10.1051/ 0004-6361/201220865 Jaeger TR, Osten RA, Lazio TJ, Kassim N, Mutel RL (2011) AJ 142:189. https://doi.org/10.1088/ 0004-6256/142/6/189 Jakosky BM, Grebowsky JM, Luhmann JG, Brain DA (2015) GeoRL 42:8791. https://doi.org/10. 1002/2015GL065271 Jonas JL (2009) IEEE Proc 97(8):1522. https://doi.org/10.1109/jproc.2009.2020713 Kao MM, Hallinan G, Pineda JS et al (2016) ApJ 818(1):24. https://doi.org/10.3847/0004-637X/ 818/1/24 Kirkpatrick JD, Reid IN, Liebert J et al (1999) ApJ 519(2):802. https://doi.org/10.1086/307414 Kirkpatrick JD, Gelino CR, Cushing MC et al (2012) ApJ 753(2):156. https://doi.org/10.1088/ 0004-637X/753/2/156 Kochukhov O, Lavail A (2017) ApJL 835(1):L4. https://doi.org/10.3847/2041-8213/835/1/L4 Kochukhov O, Petit P, Strassmeier KG et al (2017) AN 338:428. https://doi.org/10.1002/asna. 201713310 Kuznetsov AA, Doyle JG, Yu S et al (2012) ApJ 746(1):99. https://doi.org/10.1088/0004-637x/ 746/1/99 Leto P, Trigilio C, Oskinova L et al (2017) MNRAS 467:2820. https://doi.org/10.1093/mnras/ stx267 Liebert J, Kirkpatrick JD, Reid IN, Fisher MD (1999) ApJ 519:345. https://doi.org/10.1086/307349 Limber DN (1958) ApJ 127:363. https://doi.org/10.1086/146468 Linsky JL, Wood BE, Brown A, Giampapa MS, Ambruster C (1995) ApJ 499:670. https://doi.org/ 10.1086/176614

608

P. K. G. Williams

Lynch C, Mutel RL, Güdel M (2015) ApJ 802:106. https://doi.org/10.1088/0004-637X/802/2/106 Lynch C, Murphy T, Ravi V et al (2016) MNRAS 457(2):1224. https://doi.org/10.1093/mnras/ stw050 Lynch CR, Lenc E, Kaplan DL, Murphy T, Anderson GE (2017) ApJL 835:30. https://doi.org/10. 3847/2041-8213/aa5ffd Martín EL, Delfosse X, Basri G et al (1999) AJ 118(5):2466. https://doi.org/10.1086/301107 McComas DJ, Bagenal F (2007) GeoRL 34(20):L20,106. https://doi.org/10.1029/2007GL031078 McLean M, Berger E, Irwin J, Forbrich J, Reiners A (2011) ApJ 741(1):27. https://doi.org/10. 1088/0004-637x/741/1/27 McLean M, Berger E, Reiners A (2012) ApJ 746(1):23. https://doi.org/10.1088/0004-637x/746/ 1/23 Metodieva YT, Kuznetsov AA, Antonova AE et al (2017) MNRAS 465(2):1995. https://doi.org/ 10.1093/mnras/stw2597 Miles-Páez PA, Zapatero Osorio MR, Pallé E (2015) A&A 580:L12. https://doi.org/10.1051/ 0004-6361/201424626 Miles-Páez PA, Metchev SA, Heinze A, Apai D (2017) ApJ 840:83. https://doi.org/10.3847/15384357/aa6f11 Mohanty S, Basri G, Shu F, Allard F, Chabrier G (2002) ApJ 571:469. https://doi.org/10.1086/ 339911 Morin J, Donati JF, Petit P et al (2010) MNRAS 407(4):2269. https://doi.org/10.1111/j.13652966.2010.17101.x Nichols JD, Burleigh MR, Casewell SL et al (2012) ApJ 760(1):59. https://doi.org/10.1088/0004637x/760/1/59 Osten RA, Brown A, Ayres TR et al (2004) ApJS 153(1):317. https://doi.org/10.1086/420770 Osten RA, Hawley SL, Bastian TS, Reid IN (2006) ApJ 637(1):518. https://doi.org/10.1086/ 498345 Osten RA, Phan-Bao N, Hawley SL, Reid IN, Ojha R (2009) ApJ 700:1750. https://doi.org/10. 1088/0004-637x/700/2/1750 Osten RA, Melis C, Stelzer B et al (2015) ApJL 805:L3. https://doi.org/10.1088/2041-8205/805/ 1/L3 Paty C, Paterson W, Winglee R (2008) JGR 113:A06,211. https://doi.org/10.1029/2007JA012848 Perley RA, Chandler CJ, Butler BJ, Wrobel JM (2011) ApJL 739(1):L1. https://doi.org/10.1088/ 2041-8205/739/1/l1 Phan-Bao N, Osten RA, Lim L, Martín EL, Ho PTP (2007) ApJ 658:553. https://doi.org/10.1086/ 511061 Pineda JS, Hallinan G, Kirkpatrick JD et al (2016) ApJ 826(1):73. https://doi.org/10.3847/0004637X/826/1/73 Ravi V, Hallinan G, Hobbs G, Champion DJ (2011) ApJL 735(1):L2. https://doi.org/10.1088/ 2041-8205/735/1/l2 Reid IN, Kirkpatrick JD, Gizis JE, Liebert J (1999) ApJL 527(2):L105. https://doi.org/10.1086/ 312409 Reiners A, Schuessler M, Passegger VM (2014) ApJ 794(2):144. https://doi.org/10.1088/0004637X/794/2/144 Robertson P, Mahadevan S, Endl M, Roy A (2014) Sci 345(6915):440. https://doi.org/10.1126/ science.1253253 Rodriguez LF, Zapata L, Palau A (2017) Astron J 153:209. https://doi.org/10.3847/1538-3881/ aa6681 Rodríguez-Barrera MI, Helling C, Stark CR, Rice AM (2015) MNRAS 454:3977. https://doi.org/ 10.1093/mnras/stv2090 Route M, Wolszczan A (2012) ApJL 747(2):L22. https://doi.org/10.1088/2041-8205/747/2/l22 Route M, Wolszczan A (2016) ApJL 821(2):L21. https://doi.org/10.3847/2041-8205/821/2/L21

28 Radio Emission from Ultracool Dwarfs

609

Rutledge RE, Basri G, Martín EL, Bildsten L (2000) ApJL 538:L141. https://doi.org/10.1086/ 312817 Saar SH, Linsky JL (1985) ApJ 299:47. https://doi.org/10.1086/184578 Santos-Costa D, Bolton SJ (2008) P&SS 56:326. https://doi.org/10.1016/j.pss.2007.09.008 Sault RJ, Oosterloo T, Dulk GA, Leblanc Y (1997) A&A 324:1190. http://adsabs.harvard.edu/abs/ 1997A%26A...324.1190S Schrijver CJ (2009) ApJL 699(2):L148. https://doi.org/10.1088/0004-637X/699/2/L148 Semel M (1989) A&A 225(2):456. http://adsabs.harvard.edu/abs/1989A%26A...225..456S Shields AL, Ballard S, Johnson JA (2016) Phys Rep 663:1. https://doi.org/10.1016/j.physrep. 2016.10.003 Skumanich A (1972) ApJ 171:565. https://doi.org/10.1086/151310 Stelzer B, Micela G, Flaccomio E, Neuhäuser R, Jayawardhana R (2006) A&A 448(1):293. https://doi.org/10.1051/0004-6361%3A20053677 Stelzer B, Alcalá J, Biazzo K et al (2012) A&A 537:A94. https://doi.org/10.1051/0004-6361/ 201118097 Tinney CG, Reid IN (1998) Mon Not R Astron Soc 301:1031. https://doi.org/10.1046/j.13658711.1998.02079.x Treumann R (2006) A&ARv 13(4):229. https://doi.org/10.1007/s00159-006-0001-y Trigilio C, Leto P, Umana G, Leone F, Buemi CS (2004) A&A 418:593. https://doi.org/10.1051/ 0004-6361%3A20040060 West AA, Hawley SL, Bochanski JJ et al (2008) AJ 135:785. https://doi.org/10.1088/0004-6256/ 135/3/785 White SM, Kundu MR, Jackson PD (1989) A&A 225(1):112. http://adsabs.harvard.edu/abs/ 1989A%26A...225..112W Williams PKG, Berger E (2015) ApJ 808(2):189. https://doi.org/10.1088/0004-637X/808/2/189 Williams PKG, Cook BA, Berger E (2014) ApJ 785(1):9. https://doi.org/10.1088/0004-637X/785/ 1/9 Williams PKG, Berger E, Irwin J, Berta-Thompson ZK, Charbonneau D (2015a) ApJ 799(2):192. https://doi.org/10.1088/0004-637X/799/2/192 Williams PKG, Casewell SL, Stark CR et al (2015b) ApJ 815:64. https://doi.org/10.1088/0004637X/815/1/64 Williams PKG, Gizis JE, Berger E (2017) ApJ 834(2):117. https://doi.org/10.3847/1538-4357/ 834/2/117 Wolszczan A, Route M, (2014) ApJ 788(1):23. https://doi.org/10.1088/0004-637X/788/1/23 Wong AS, Yung YL, Friedson AJ (2003) GeoRL 30(8):1447. https://doi.org/10.1029/ 2002GL016661 Wright NJ, Drake JJ, Mamajek EE, Henry GW (2011) ApJ 743(1):48. https://doi.org/10.1088/ 0004-637x/743/1/48 Wu CS, Lee LC (1979) ApJ 230:621. https://doi.org/10.1086/157120 Yantis WF, Sullivan WT III, Erickson WC (1977) BAAS 9:453. http://adsabs.harvard.edu/abs/ 1977BAAS....9..453Y Yu S, Doyle JG, Kuznetsov A et al (2012) ApJ 752(1):60. https://doi.org/10.1088/0004-637x/752/ 1/60 Zarka P, Treumann RA, Ryabov BP, Ryabov VB (2001) Ap&SS 277:293. https://doi.org/10.1023/ a%3A1012221527425

Definition of Exoplanets and Brown Dwarfs

29

Jean Schneider

Contents What Is a Planet? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . From the Heaven of Concepts to the Hell of Observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

612 613 616 616

Abstract

This chapter reviews the definition of exoplanets and of brown dwarfs. Emphasis is given to the separation of these two populations, whose masses may present some overlap. Keywords

Brown Dwarf · Planet · Definition

Thousands of substellar objects (that is, brown dwarfs and planets) have been detected since 1989 (see, for instance, the following web sites: exoplanet.eu and https://exoplanetarchive.ipac.caltech.edu). Their masses run from 0.02 times the mass of the Earth (PSR 1257 C 20 b) to about 63 Jupiter masses (CoRoT-15 b), about five orders of magnitude in mass. More than 2700 confirmed planets have been detected by the photometric transit technique (e.g., see  Chap. 4, “Discovery of the First Transiting Planets” by Dunham in this Handbook of Exoplanets), which led to an accurate determination of their size. Their radii run from 0.32 Earth radius (Kepler-37 b) to 2.1 Jupiter radius (HAT-P-67 b), about two orders of magnitude in

J. Schneider () LUTh, UMR 8102, Observatoire de Paris, 5 place Jules Janssen, F-92195 Meudon Cedex, France e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_119

611

612

J. Schneider

3e+0 exoplanet.eu, 2017-02-20

PADC

2e+0

Planetary Radius (Rjup)

1e+0

5e-1 4e-1 3e-1 2e-1

1e-1

5e-2 4e-2 3e-2 1e-4

1e-3

1e-2

1e-1 Planetary Mass (Mjup)

1e+0

1e+1

Fig. 1 Mass radius relation for substellar objects (24 February 2017) from exoplanet.eu

size. The corresponding mass-radius diagram is represented in Fig. 1. One could be tempted to think that the more massive the object is, the larger it is in size and that there is some limit in mass and/or radius that distinguishes planets from everything else, even if this mass limit is below the well accepted substellar borderline at 72 Jupiter masses (minimum mass required for stable nuclear fusion of hydrogen in the interiors of solar metallicity stars). The objects between planets and very low-mass stars have been named brown dwarfs since the paper by Jill Tarter in 1973. But Fig. 1 shows that beyond 0.2 Jupiter mass, there is a degeneracy in the objects’ radii. One is then facing two problems: terminology (what is a planet? what is a brown dwarf?) and classification (how to decide if a given object is a planet or a brown dwarf according to a given definition?). Let us discuss these two aspects.

What Is a Planet? The debate, strongly motivated by the discovery of the first planets orbiting stars at the end of the last century (see  Chap. 1, “The Discovery of the First Exoplanets”) and the finding of planetary-mass objects in isolation at the beginning of this century (see  Chap. 22, “Brown Dwarfs and Free-Floating Planets in Young Stellar Clusters”) and encouraged by several authors (see, e.g., Baraffe et al. 2010; Schneider et al. 2011; Hatzes and Rauer 2015; Schlaufman 2018), is still ongoing and will not be closed by the present contribution. Names are arbitrary conventions, but the natural trend is to classify objects using sufficiently elaborated concepts.

29 Definition of Exoplanets and Brown Dwarfs

613

Derived from the solar system analogy, one is tempted to call “exoplanets” (in short planets) to small bodies that are orbiting around other stars and were formed by condensation in a circumstellar dust disk. A first question is of course how small or massive the body has to be for being a planet. The problem here is that there actually exist small bodies orbiting stars that probably did not form like planets but like stars and brown dwarfs; this is from the collapse and fragmentation of a (possibly dusty) gas cloud.

From the Heaven of Concepts to the Hell of Observations Formation provides a clear conceptual discrimination between planets and brown dwarfs (keeping in mind that it is a convention). But it is based on a criterion involving an inobservable concept, namely, the formation scenario, because we do not have brown dwarfs and planets birth movie at hand (see also  Chap. 21, “Brown Dwarf Formation: Theory” by Whitworth and the section on “Formation and Evolution of Planets and Planetary Systems” in this Handbook of Exoplanets). We can only rely on actual observables. Standard basic observables are the object mass, radius, and temperature. An ideal situation would be that, at least for one of these observables, there exist two domains Dplanet and Dbrown dwarf of values which do not intersect. It is unfortunately not the case since there are, according to formation models, objects formed by condensation of dust (i.e., planets according to the proposed convention) (smaller or larger, heavier or lighter, cooler or hotter) than objects formed by collapse (i.e., brown dwarfs). Even worse, there are a few pulsar companions with masses lower than 30 Jupiter mass. They are probably the relic of stellar companions eroded by the pulsar strong wind (Ray and Loeb 2015). One can argue that as such they are not planets nor brown dwarfs, their formation process being very different. But one cannot exclude that such erosion mechanism happened also for low-mass companions of main sequence stars with strong winds (see e.g., Sanz-Forcada et al. 2010). Consequently, the choice between “planet” and “brown dwarf” classification for a given object can only be arbitrary. The choice made by the Extrasolar Planets Encyclopaedia at exoplanet.eu is to consider all objects below 60 Jupiter mass as “planets,” based on the results by Hatzes and Rauer (2015). Hatzes and Rauer (2015) argument is that the mass-radius and the mass-density relations present a well-defined increasing trend for objects more massive than Saturn (giant planetary régime) up to a certain mass value where the slope of the trend changes dramatically. This happens at 60 Jupiter mass (Fig. 2). Unfortunately, the number (statistics) of objects in the 30–60 Jupiter mass region is very low (the so-called brown dwarf desert); for producing Fig. 2, Hatzes and Rauer (2015) used only transiting planets and brown dwarfs and did not correct the relation shown in Fig. 2 from the mass distribution (or mass histograms) at different mass intervals. Earlier data suggested a dip around 40 Jupiter mass (Sahlman et al. 2011, Udry et al. 2010 – see also Figs. 3 and 4) in the mass histogram. More statistics will become available in the near future thanks to radial velocity surveys from ground facilities and astrometric data from Gaia that will allow to investigate whether the feature around 40 Jupiter mass in the mass-radius diagram exists or not.

614

J. Schneider 1000 Low Mass Planets

Giant Planets

Stars

Density (gm cm–3)

100

10

1

.1

.01 .001

.01

.1

1 10 Mass (Jupiter)

100

1000

10000

Fig. 2 Empirical mass-density relation (Hatzes and Rauer 2015) Fig. 3 Mass histogram for low-mass objects (Udry et al. 2010)

35 Two different mechanisms for binary and planet formation 30

25 Brown-dwarf desert N

20

15 Planets? 10

5 0.001 Exoplanets

0.1 Binaries 1 0.01 log(m2 sin i) [M . ] Halbwachs et al. 2003

A future improvement to separate the planet and brown dwarf populations will likely be possible from advanced observables, like the spectral type and species composition. They will help to constrain the formation mechanism of the object

29 Definition of Exoplanets and Brown Dwarfs

615

4.5 4 3.5

Number of stars

3 2.5 2 1.5 1 0.5 0

20

30

40

50

60

70

80

M2 sin i (Mj)

Fig. 4 Low-mass object histogram in the 20–75 Jupiter mass region (Sahlman et al. 2011)

(accretion in a dust disk or collapse and fragmentation from a gas cloud). At least one conclusion is clear: the former criterion to discriminate planets and brown dwarfs based on the mass boundary at 13 Jupiter masses, corresponding to the triggering of nuclear burning of deuterium at the interiors of the small bodies, is not relevant since, theoretically speaking, objects can be grown by dust accretion from a circumstellar disk (planet-like formation) and acquire a final mass larger than 13 Jupiter masses. There is a second, more factual problem: the value of some observables, particularly the mass, can be very uncertain. This is especially the case for companion objects detected by direct imaging where the mass cannot be inferred from radial velocity measurements but only from spectra, photometry, and models. A typical example is the object 2M1207b (Chauvin et al. 2004), which is located at a distance of >55 AU from its parent brown dwarf and has a mass of 4 ˙ 1 Jupiter mass. Indeed, in these cases the planet-star or planet-brown dwarf separation is so wide q  that the semi-amplitude K D GM star =aplanet of the parent object radial velocity variation induced by the planet motion is too low to be measurable with current technology. Even more, when the mass determination is as precise as a few percent (in case of radial velocity or astrometric measurements), one faces the absurd situation of a sharp mass limit. For example, how should we classify objects like CoRoT-15 b that has a mass at the borderline, M D 63.3 ˙ 4 MJup ?

616

J. Schneider

An additional problem, which will not be addressed here, is the existence of the “interstellar wanderers,” i.e., planets that are free-floating (they do not orbit around any more massive object) either because they were formed in an isolated way or because they were expelled by dynamical interactions within multiple planetary systems during the early stages of the formation of the system. For more information regarding this population, see the section “Between Planets and Stars: Substellar Objects in this Handbook of Exoplanets”.

Conclusion Assuming that the definition of planets and brown dwarfs is adopted according to their formation mechanism, to separate the two populations is not an easy task. Any catalog contains necessarily a mixture of both populations. Since catalogs are useful not only to list the objects properties but also to make statistics regarding these properties, this author recommends to take low constrains (a mass limit as high as 60 Jupiter mass is adopted at exoplanet.eu) on the properties used to define a sample, in order not to miss interesting objects. Modern software used to read electronic catalogs allows to eliminate easily objects from a catalog that do not fulfil the criteria of each user, who is free to impose his/her own criteria.

References Baraffe I et al (2010) The physical properties of extrasolar planets. Rep Progr Phys 73:016901 Chauvin G et al. (2004) A giant planet candidate near a young brown dwarf. Astron & Astrophys 425:L29 Hatzes A, Rauer H In: (2015) A Definition for Giant Planets Based on the Mass-Density Relationship. Astrophys J Letters 810:L25 Ray A, Loeb A (2017) Inferring the composition of super-Jupiter mass companions of pulsars with radio line spectroscopy. Astrophys J 836:135 Sahlman J et al (2011) Search for brown-dwarf companions of stars. Astron Astrophys 525:A95 Sanz-Forcada J et al (2010) A scenario of planet erosion by coronal radiation. Astron Astrophys 511:L8 Schlaufman K (2018) Evidence of an Upper Bound on the Masses of Planets and its Implications for Giant Planet Formation. Astrophys J 853:37 Schneider J et al (2011) Defining and cataloging exoplanets: The exoplanet.eu database. Astron Astrophys 532:A79 Udry S et al (2010) Detection and characterization of exoplanets: from gaseous giants to superearths. In: Proceedings of “in the spirit of Lyot,” Paris, Oct 2010

Section IV Planet Discovery Methods Alexander Wolszczan

Alexander Wolszczan is the Evan Pugh Professor of Astronomy and Astrophysics at Pennsylvania State University. His research interests focus on astronomy of planets beyond the Solar System. He has also worked on topics in relativistic gravitation, pulsars, brown dwarfs, and the physics of the interstellar medium. He is best known for his discovery, in 1992, and the subsequent confirmation, of the first planets orbiting a star other than the Sun. He is also a discoverer and co-discoverer of many pulsars and giant planets around evolved stars.

Radial Velocities as an Exoplanet Discovery Method

30

Jason T. Wright

Contents Radial Velocities in Astronomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Binary Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Redshift Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Barycentric Motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Measuring Precise Radial Velocities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stable Spectrographs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Absorption Cell Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Radial Velocity Jitter: Spurious Doppler Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stellar Magnetic Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Photospheric Motions and Global Oscillations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Identifying Jitter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Target Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Optical Vs. Infrared . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Future Challenges and Opportunities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

620 620 622 623 623 624 624 625 625 626 626 628 629 629 630

Abstract

The precise radial velocity technique is a cornerstone of exoplanetary astronomy. Astronomers measure Doppler shifts in the star’s spectral features, which track the line-of-sight gravitational accelerations of a star caused by the planets orbiting it. The method has its roots in binary star astronomy, and exoplanet detection represents the low-companion-mass limit of that application. This limit requires control of several effects of much greater magnitude than the signal sought: the motion of the telescope must be subtracted, the instrument must be

J. T. Wright () Department of Astronomy and Astrophysics, Center for Exoplanets and Habitable Worlds, The Pennsylvania State University, University Park, PA, USA e-mail: [email protected]; [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_4

619

620

J. T. Wright

calibrated, and spurious Doppler shift “jitter” must be mitigated or corrected. Two primary forms of instrumental calibration are the stable spectrograph and absorption cell methods, the former being the path taken for the next generation of spectrographs. Spurious, apparent Doppler shifts due to non-center-of-mass motion (jitter) can be the result of stellar magnetic activity or photospheric motions and granulation. Several avoidance, mitigation, and correction strategies exist, including careful analysis of line shapes and radial velocity wavelength dependence.

Radial Velocities in Astronomy Most of what we know about heavens comes from information encoded in various forms of light. While many important advances have come from studies of meteorites, in situ measurements of the solar system by space probes, detection of high-energy particles, and more recently observations of astronomical neutrinos and gravitational waves, the vast majority of astronomical observations come from eking as much information as possible from photons. Because of this restriction, certain properties of stars and galaxies are much more easily determined than others. Positions on the sky can be measured to great accuracy, but distances can often only be approximated. For the time derivatives of these quantities, the situation is usually reversed; motions of objects in the plane of the sky can be imperceptible due to the vast distances involved, but thanks to the Doppler shift, motions along the line of sight (radial velocities) can often be measured with great precision. Much of astronomy is concerned with the motions of objects under the influence of their mutual gravity, and so radial velocity measurements naturally form a cornerstone of astronomical research. They tell us the masses of everything from moons in the solar system to stars to galaxies; they have revealed the existence of unseen planets, dark matter, and the expansion and acceleration of the universe itself.

Binary Systems Binary stars provide much of our foundational knowledge of stellar structure. The relative masses of the stars in a binary system can be determined from the stars’ orbits about their common center of mass from the ratio of the amplitudes of their radial velocity variations. Further knowledge of the nonradial component of their motion – either from astrometric measurements of their orbits or from the fact that they eclipse each other – yields absolute masses for the individual stars from Newton’s laws. In some cases, only a single star in a binary system is bright enough to have its spectrum measured (a single-lined spectroscopic binary system, or SB1). In this case, the semi-amplitude of the radial velocity variations K over the course of an

30 Radial Velocities as an Exoplanet Discovery Method

621

orbit with period P is related to the properties of the system via the so-called mass function, f : 3

f 

3 sin3 i PK 3 .1  e 2 / 2 Munseen D 2G .Munseen C Mseen /2

(1)

Here, e represents the eccentricity of the orbit, i represents the inclination of the plane of the orbit with respect to the plane of the sky (i D 0 represents a face-on orbit with no radial component), G is Newton’s constant, and Mseen and Munseen are the masses of the seen and unseen stars. The phase and orientation of the elliptical orbit with respect to the line of sight are represented by an additional pair of parameters, T0 and !, and the overall radial velocity of the center of mass of the system is usually given by . The orientation of the orbit on the plane of the sky is given by ˝, but, as with i , its measurement requires additional information not found in the radial velocities. (For a more detailed discussion of single-lined spectroscopic orbits in general, see Wright and Gaudi (2013), and for a more detailed treatment of the mechanics of translating measurements into orbital parameters, see Wright and Howard 2009.) The orbital parameters measured in this way represent the orbit of the observed star about the center of mass of the system; in single-lined system, the existence of the unseen object is often inferred from this motion, and its (unmeasured and usually unreported) orbital elements are identical except that K is scaled by a factor of Munseen =Mseen and ! differs by . This application of radial velocities is not restricted to binary stars; it is used to detect unseen binary companions ranging in mass from black holes to planets. For the case of exoplanets, we can approximate the mass function in the large massratio limit Munseen =Mseen D Mplanet =M 1, yielding the more familiar equation relating the amplitude of a Doppler shift to the mass and orbital properties of the planet:  K

2G PM2

 13

Mplanet sin i p 1  e2

(2)

The ability to detect exoplanets with precise radial velocimetry thus depends on the planet’s mass, inclination, and orbital period (there is also a dependence on orbital eccentricity, which is complex but weak for low eccentricities). The period dependence is weak, and for giant planets, detectability is often limited more by the duration of the observations than their RV amplitude. The strongest dependence is on the quantity Mplanet sin i (pronounced “em-sineeye”). The true mass of the planet is larger by a geometric factor 1= sin i , and so this is often referred to as the “minimum mass” of the planet (a precise term which can be properly computed from the more exact Equation 1). For scale, we can express the quantities in more familiar units:

622

J. T. Wright 1

1

2

 12

1 3

 23

1

1

2

1

1

2

K D 28 m=s.1  e 2 / 2 .P =yr / 3 .M =Mˇ / 3 .Mplanet =MJupiter / sin i D 200 m=s.1  e 2 /

.P =day/ .M =Mˇ /

.Mplanet =MJupiter / sin i

D0:09 m=s.1  e 2 / 2 .P =yr / 3 .M =Mˇ / 3 .Mplanet =MEarth / sin i D0:64 m=s.1  e 2 / 2 .P =day/ 3 .M =Mˇ / 3 .Mplanet =MEarth / sin i

(3) (4) (5) (6)

We see here why the first strong exoplanet detections (Latham et al. 1989; Mayor and Queloz 1995) were of Jovian planets in short-period orbits: 51 Peg b, for instance, has minimum mass of 0:4 MJupiter and a 4.2 -day orbit, so its RV amplitude is a relatively large 60 m/s. Jupiter induces a 12 m/s amplitude motion on the Sun; the Earth’s motion is a (currently) undetectable 9 cm/s. Finally, the dependence on stellar mass means that exoplanet detection is in principle most sensitive around the lowest mass stars, a point we shall revisit later.

Redshift Measurements Starlight is imprinted with many absorption lines by ions, atoms, and molecules in stellar atmospheres, and atomic physics allows us to calculate or measure the rest wavelengths rest of these lines. A star’s radial motion causes these lines to be Doppler shifted to their observed wavelengths obs . The Doppler formula then allows astronomers to calculate the relative radial speed between the star and the observatory that measured the light vr via the redshift z : z

1 obs  rest 1 D rest .1 C vr =c/

(7)

where here (unlike above) is the relativistic factor 1=.1  .v=c/2 /, c is the speed of light, and v is the scalar relative speed between the frame of the star and the observatory (which is not necessarily in the radial direction). Radial velocity measurements made with respect to the “laboratory” in this manner are called absolute radial velocities and form the basis of our understanding of the dynamics of the Galaxy and the expansion of the universe. Such measurements have accuracy limited by the wavelength calibration of the spectrograph and understanding of complicating factors such as the internal motions of the emitting material and redshifts from general relativity. Typical accuracies of absolute radial velocities of stars are of order 100 m/s (Chubak et al. 2012). More precise measurements can be made by measuring differential radial velocities, that is, the change in the redshift between two epochs. Differential measurements have the advantage that some uncertainties (such as those from systematic effects, imperfectly known rest wavelengths, or the model of the emitting material) will effect all measurements made with a certain instrument or technique equally, and so differences between measurements do not suffer from them. Thus,

30 Radial Velocities as an Exoplanet Discovery Method

623

measurements of the change in the redshift of a star’s spectral features can be made to more than two orders of magnitude better precision than the accuracy of the absolute redshift.

Barycentric Motion The observatory measuring a radial velocity is on a moving platform: the Earth. The Earth rotates at 300 m/s and orbits the Sun at 30 km/s. As a result, the measured radial velocity of a perfectly stable star will appear to vary on diurnal and annual timescales by some fraction of these amounts. Correcting measurements for this motion is known as a barycentric correction, where the term barycenter refers to the center of mass of the solar system. The idea is that one must correct measurements made on the Earth to measurements that would have been made in the (inertial) frame comoving with the solar system’s center of mass. These motions are often much larger than the orbital motions of the stars, and in the case of stars moving under the influence of planets, the barycentric correction can be four or five orders of magnitude larger. Thus, practically speaking, the problem of radial velocity precision is not one of measuring very small redshifts, but one of measuring modest redshifts very precisely, often to one part in 104 or better. Fortunately, the motion of the Earth is well studied and well measured for purposes that require much more precision than exoplanet detection. Ephemerides (tables describing the location of celestial bodies as a function of time, pronounced eff-em-AIR-i-deez, singular ephemeris, pronounced eh-FEM-er-iss) for the Earth in the barycentric frame are maintained by the Jet Propulsion Laboratory and others, and the orientation of the Earth can be predicted into the future with high accuracy (and measurements of that orientation are available through, for instance, the International Earth Rotation Service). Finally, all modern observatories provide precise timekeeping, so observations can be tagged with an appropriate time stamp for barycentric correction algorithms later. A detailed description of the barycentric correction process in the context of exoplanet detection can be found in Wright and Eastman (2014).

Measuring Precise Radial Velocities Two primary methods of precise radial velocimetry have been used to discover and characterize exoplanets via the reflex velocities of their host stars: absorption cell spectroscopy and spectrograph stabilization. They differ primarily in the method used to calibrate the spectrograph. The chief limit to the precision of differential redshift measurements is the wavelength calibration of the spectrograph. Typical RV spectrographs resolve starlight with a power of R D =  50,000–100,000, meaning that a shift of a single pixel corresponds to a change in radial velocity of 1 km/s. Since giant planets

624

J. T. Wright

change their host stars’s velocities by of order tens of m/s, one must measure shifts to a precision of 102 pixel for giant planets and two or three orders of magnitude better than that to detect Earth-mass planets. Since typical pixels on astronomical detectors are of order 15 m across, one is measuring the “motions” of stellar lines to a fraction of a nanometer. Astronomical spectrographs are not this stable. Most are general-purpose instruments with many moving parts that are actuated so the spectrograph can be used for many purposes. A given wavelength of light cannot typically be expected to land to its usual position by much better than a pixel from night to night (or year to year). The solution then is to employ some combination of stabilization and calibration and employ differential techniques. Wright and Gaudi (2013) provide a discussion of the historical development and first applications of both techniques, and Fischer et al. (2016) provide an overview of the state of the field.

Stable Spectrographs The first strong exoplanet detection (recognized as such only after the fact) was that of Latham et al. (1989), who stabilized a spectrograph in two ways: by removing it from the telescope (to prevent its orientation with respect to the local gravity vector from changing, minimizing flexure) and coupling it to the starlight via an optical fiber (which served to “scramble” the starlight, presenting the spectrograph with a uniform image of the star despite variations in guiding and seeing). Remaining variations in the wavelength solution of the spectrograph (from thermal changes to the spectrograph or small variations in the fiber illumination) were tracked with a thorium-argon emission lamp, which provided a stable set of reference lines. This combination of stabilization and calibration allowed for precise differential measurements over the course of several nights. Since then, this technique has been taken to extreme lengths. State- of-the-art stable radial velocimeters today control the vibration, temperature, and pressure of spectrographs with exquisite precision using cryostats and vacuum chambers. The remaining, unavoidable changes in the spectrograph (from, for instance, slow changes in the crystalline structure of the metals involved or irregular thermal outputs from the detector electronics) are tracked via emission sources such as laser frequency combs, which are locked to atomic clocks and provide essentially perfect wavelength references. Today, the state of the art is represented by the HARPS (Queloz et al. 2001b) and ESPRESSO (Pepe et al. 2010) spectrographs of ESO, which are stable below the 1 m/s level (the latter aspires to 10 cm/s precision.)

Absorption Cell Spectroscopy A more widely applicable method, and the one responsible for most of the first several dozen exoplanet discoveries, is that of absorption cell calibration. A cell

30 Radial Velocities as an Exoplanet Discovery Method

625

of gas is placed in the path of the starlight, imprinting it with a set of spectral absorption features. These features follow the starlight through the spectrograph and so suffer all of the same instrumental shifts as the starlight itself. They thus provide an opportunity to track and calibrate all changes in a spectrograph, no matter how unstable. Early work by Campbell and Walker (1979) used HF gas and achieved precision near 10 m/s. Later, Marcy and Butler (1992) found success with molecular iodine (I2 ), and Butler et al. (1996) demonstrated precision near 3 m/s through careful modeling of the iodine spectrum and the spectrograph line spread function. Many high-resolution spectrographs have since been retrofitted with absorption cells, turning them into Doppler velocimeters capable of 1–10 m/s precision.

Radial Velocity Jitter: Spurious Doppler Signals Observed stellar spectroscopic absorption features are the product of absorption, emission, scattering, and motions of gasses throughout stellar atmospheres, across the differentially rotating stellar disk. The lines will change their shape and centroid positions due to many effects besides a true center-of-mass movement of the star itself. This effect is called jitter, and it operates on a variety of timescales, amplitudes, and with a range of noise distributions. The problem is most severe for spotty, rapidly rotating, and low-gravity stars. This not only makes detection sensitivity progressively worse for younger and more evolved stars but also introduces the more insidious problem of false detections around these stars, a problem that has plagued the field since its inception (e.g., Queloz et al. 2001a). The solution to the problem of detecting planets with Doppler amplitudes near or below the level of the jitter requires a variety of avoidance and mitigation techniques.

Stellar Magnetic Activity Cool stars – those with convective envelopes and enough spectral features that precise Doppler work is possible – have dynamos that generate surface magnetic fields. These fields interact strongly with the stellar atmosphere and wind, which are (at least partly) ionized, and affect their motions. The wind in particular is heavily influenced by the field, which is anchored to the rotating star, and this coupling causes the wind to carry angular momentum away from the star, which thus spins down. This spin-down weakens the dynamo, lessening the effects of activity-based jitter. Young stars thus spin quickly and have a lot of activity-based jitter, and the problem is not as severe for older stars. On the surface of stars, magnetic fields cause surface brightness inhomogeneities, including bright plage and faculae, and dark spots. This breaks the symmetry of the rotational broadening kernel of the emergent intensity from the stellar atmosphere,

626

J. T. Wright

producing asymmetric line profiles. That is, a bright spot on the approaching limb of the star will produce a line with a slightly blueshifted centroid with respect to a homogeneous stellar surface, and a spot there will produce a redshifted line. This effect is proportional to the spot contrast and surface coverage, and also the projected rotational velocity of the star, both of which are worse for younger stars.

Photospheric Motions and Global Oscillations Cool stars have surfaces characterized by granulation – a network of cells with centers of hot, rising material and edges of cooler, sinking material. The rising material (moving toward the observer) is hotter, and so contributes more light to the spectrum, resulting in a convective blueshift. The velocities at different heights in the atmosphere, the different angles taken by our line of sight at different parts of the stellar disk, and the turbulent nature of the motions combine to produce an asymmetric line profile, overall. Finally, the stochastic creation, destruction, and reorganization of these granules make this profile time-variable. The amplitudes of these variations are of order meters per second and decrease with increasing surface gravity (Bastien et al. 2013, 2014). These motions are also altered by the strength of the global surface field strength, resulting in RV variations on the timescales of stellar activity cycles (Lovis et al. 2011). Stars also undergo global oscillations excited by the surface granulation and magnetic events. These are probed by the Sun using helioseismology (and on stars via asteroseismology), and the dominant mode is the 5 min p-mode oscillation. The up and down motions visible on different parts of the stellar disk due to these oscillations do not precisely cancel, and the result is precise radial velocity variations on minute-to-hour timescales with amplitudes of order 1 m/s. (Kjeldsen et al. 2005)

Identifying Jitter There are two broad strategies for dealing with jitter: avoidance and mitigation. Avoidance is a viable strategy for surveys that can seek the “quietest” stars, especially those that require no more than 1 m/s precision. In this case, bright, unevolved, old, late-G, and K dwarfs offer the best opportunity to find lowamplitude planets with minimal jitter mitigation (e.g., Wright 2005; Howard et al. 2009). Below 1 m/s, or for other kinds of stars, one must deal with jitter. The different kinds of jitter require different mitigation strategies. The two most common are to correlate radial velocities with activity indicators, such as emission in the cores of deep lines and photometry, and to carefully examine the shapes of lines for evidence of non-center-of-mass line shifts.

30 Radial Velocities as an Exoplanet Discovery Method

627

Magnetic Activity Indicators Stellar magnetic activity results in strong, tangled field lines in stellar atmospheres, which heat the gas via magnetic reconnection and Alfvén waves. This creates the temperature inversion responsible for the existence of stellar chromospheres, which, being optically thin in the continuum, cool predominantly via emission in resonance lines such as CaII H & K, NaI D, and H-˛. This emission appears in stellar spectra as a filling-in or inversion of the cores of these absorption features in cool stars. The amount of emission in the cores of these lines gives an indication of the amount of cooling (and, therefore, heating) in the chromosphere and so serves as a good proxy for the overall level of magnetic activity on the star (at least, on the hemisphere facing the Earth). Perhaps surprisingly, given the somewhat tenuous connection between the overall strength of the global field and the mechanisms by which that field creates spurious RV signatures, the measured strength of these emission features often has a simple relationship to the Doppler anomaly. In the case of rotationally modulated spots, the Doppler signature is out of phase with the activity measurements by /2, because the spots are most prominent at the center of the disk, where they have zero rotational motion in the radial direction. In this case, photometry may also show a signal in-phase with the activity measurements (Queloz et al. 2001a). In the case of magnetic activity cycles and other variations in the global field strength, the effect is a simple correlation, with stronger fields yielding more redshifted lines. Neglecting to check for such a correlation can lead to spurious planet detections (Wright et al. 2008; Robertson et al. 2014). In some cases, this simple correlation is probably due to the suppression of convective blueshift as the field lines restrict the motion of the (slightly) ionized atmosphere (e.g., as in the case of ˛ Cen B Dumusque et al. 2012). Even in the case where a direct correlation between activity indicators and RVs is not clear, the effects of stellar magnetic activity can be diagnosed via their rotational modulation. For sufficiently densely sampled activity time series, a power spectrum will often reveal the rotation period (and harmonics thereof). Any RV variation at these periods is suspect. Of course, real planets may have orbital periods that match these periods (as in the case of  Boo b Butler et al. 1997), whether simply by coincidence or due to tidal locking, but in general the burden of proof for a claimed exoplanet discovery grows higher in the presence of such coincidences, especially for low-amplitude planets (Robertson and Mahadevan 2014).

Line Shape Analysis Sources of spurious RV changes will, in general, not have exactly the same spectral signature as a Doppler shift due to true center-of-mass motion. While true Doppler shifts will leave line shapes unchanged, rotationally modulated spots or changes to the convective blueshift pattern will alter those shapes, if only slightly. The most commonly seen measures of line shape are the line bisector – which traces the center of a line as a function of depth below the continuum – and the

628

J. T. Wright

width of the lines. True Doppler shifts will preserve these shapes, while changes in a line’s profile should alter them (Hatzes 1996). The “span” of the bisector is the difference in the bisector position near the top and bottom of the line, measured in units of velocity. This is equivalent to the inverse of the mean slope of the bisector (for continuum normalized spectra) and is usually designated BIS (Queloz et al. 2001a; Boisse et al. 2011). Correlations between measured radial velocities and BIS or line widths are thus red flags that the signal is not due to the reflex motion of an orbiting companion. It may indicate a spectrum blended with light from other stars, stellar pulsations, or other sources of jitter (see Wright et al. 2013, for several examples.) Bisector and other line shape variations are measured most easily in stable spectrographs, where the shape of the cross-correlation function (CCF) can serve as a proxy for the mean shape of all lines used to derive velocities. They are not routinely measured in absorption cell spectroscopy.

Wavelength Dependence True center-of-mass motion shifts all line wavelengths by the same fraction, while spurious effects will generally have different effects on lines of different wavelengths, depths, formation heights, and ionization states. For instance, rotationally modulated spot-induced RV anomalies should be less pronounced in the infrared, where brightness temperature contrasts are lower. In principle, RVs measured from various lines should be checked for consistency, but in practice this is difficult. In absorption cell velocimetry, the spectrum is not decomposed into individual lines, and only a relatively narrow range of wavelengths is examined, so measured RVs cannot be examined as a function of wavelength or line properties. The problem is more tractable in stable spectrograph velocimetry, where individual lines are typically chosen for the analysis and can, in principle, be compared with each other for consistency.

Target Selection The first targets of precise Doppler surveys were generally very bright (naked-eye) stars because early practitioners were generally using small telescopes and because of the large number of photons needed to make a precise Doppler measurement. The stars must also have a rich set of absorption features to measure, making hot stars unsuitable targets, and must have narrow features, favoring stars with low projected rotational velocities (v sin i < 10 km/s). Young stars have many surface features and flares that make them unsuitable for the highest precision work, in addition to their high rotational velocities. Giant stars also turn out to be poorer precise Doppler targets, because their atmospheres exhibit large variability in their Doppler motions (e.g., Hekker et al. 2006). The highest precision is thus achieved on bright, old, dwarf, cool stars, which typically show intrinsic variations only at the 1 m/s level (Wright 2005).

30 Radial Velocities as an Exoplanet Discovery Method

629

There are many reasons to push the technique outside of this optimal region. One is exploration of the dependence of planet occurrence on stellar mass (e.g., Johnson et al. 2010). Another is to measure the masses of planets known to transit, such as the faint, sometimes young or evolved targets of the Kepler and K2 missions (Borucki et al. 2010).

Optical Vs. Infrared In general, precise Doppler work is most easily performed in the optical, where absorption cells are well calibrated, detectors are better behaved and more easily calibrated, and telluric absorption and emission is less difficult to work with. There are compelling reasons to pursue infrared precise Doppler work, however. One is if one wishes to optimize not precision, but sensitivity to planets in the habitable zone (Kasting et al. 1993), where surface liquid water is most likely to be found. Very cool late-M dwarfs have close-in habitable zones, and so these planets have shorter periods than around G and K dwarfs. This, combined with the more favorable planet-to-star mass ratios for a given planet mass, makes the necessary precision for discovering their habitable-zone planets an order of magnitude less stringent. Since these very cool stars have very little optical luminosity, the only practical solution is to measure their spectra in the near infrared where they have most of their energy output. Recently, these and other concerns have created a strong push to extend precise Doppler velocimetry to the near infrared, with stable instruments such as the Habitable Zone Planet Finder and CARMENES (Mahadevan et al. 2012; Quirrenbach et al. 2010) and gas cell work as well (Bean et al. 2010; Gao et al. 2016).

Future Challenges and Opportunities Precise radial velocimetry will continue to be a cornerstone of exoplanetary research into the foreseeable future. Space-based transit surveys are often limited by the availability of radial velocities to rule out many common sources of false positives and to measure the masses of the planets discovered. Discovery of Jupiter analogs and very long-period companions requires decades of observation, making archival radial velocities relevant for years to come. Advances in both instrumental precision and the understanding of the sources of and methods for mitigating stellar jitter will push radial velocity sensitivity down to lower and lower mass planets, with a common goal being the detections of an Earth analog (requiring detection of a 9 cm/s signal). The most ambitious plans call for stability of 1 cm/s with spectrographs that can access enough collecting area to ensure the measurements are not photon limited (e.g., CODEX, Pepe and Lovis 2008).

630

J. T. Wright

References Bastien FA, Stassun KG, Basri G, Pepper J (2013) An observational correlation between stellar brightness variations and surface gravity. Nature 500:427–430. https://doi.org/10.1038/ nature12419, 1308.4728 Bastien FA, Stassun KG, Pepper J et al (2014) Radial velocity variations of photometrically quiet, chromospherically inactive Kepler stars: a link between RV jitter and photometric flicker. AJ 147:29. https://doi.org/10.1088/0004-6256/147/2/29, 1310.7152 Bean JL, Seifahrt A, Hartman H et al (2010) The proposed giant planet orbiting VB 10 does not exist. ApJ 711:L19–L23. https://doi.org/10.1088/2041-8205/711/1/L19 Boisse I, Bouchy F, Hébrard G et al (2011) Disentangling between stellar activity and planetary signals. A&A 528:A4. https://doi.org/10.1051/0004-6361/201014354 Borucki WJ, Koch D, Basri G et al (2010) Kepler planet-detection mission: introduction and first results. Science 327:977–980 https://doi.org/10.1126/science.1185402 Butler RP, Marcy GW, Williams E et al (1996) Attaining Doppler precision of 3 M s-1. PASP 108:500 Butler RP, Marcy GW, Williams E, Hauser H, Shirts P (1997) Three new “51 Pegasi–Type” planets. ApJ 474:L115 Campbell B, Walker GAH (1979) Precision radial velocities with an absorption cell. PASP 91:540– 545. https://doi.org/10.1086/130535 Chubak C, Marcy G, Fischer DA et al (2012) Precise radial velocities of 2046 nearby FGKM stars and 131 standards. ArXiv e-prints 1207.6212 Dumusque X, Pepe F, Lovis C et al (2012) An earth-mass planet orbiting ˛ centauri B. Nature 491:207–211. https://doi.org/10.1038/nature11572 Fischer DA, Anglada-Escude G, Arriagada P, et al (2016) State of the field: extreme precision radial velocities. PASP 128:066001 Gao P, Plavchan P, Gagné J et al (2016) Retrieval of precise radial velocities from near-infrared high-resolution spectra of low-mass stars. PASP 128(10):104,501. https://doi.org/10.1088/ 1538-3873/128/968/104501, 1603.05997 Hatzes AP (1996) Simulations of stellar radial velocity and spectral line bisector variations: I. nonradial pulsations. PASP 108:839. https://doi.org/10.1086/133805 Hekker S, Reffert S, Quirrenbach A et al (2006) Precise radial velocities of giant stars. I. Stable stars. A&A 454:943–949. https://doi.org/10.1051/0004-6361:20064946, astro-ph/0604502 Howard AW, Johnson JA, Marcy GW et al (2009) The NASA-UC Eta-Earth program. I. A super-earth orbiting HD 7924. ApJ 696:75–83. https://doi.org/10.1088/0004-637X/696/1/75, 0901.4394 Johnson JA, Aller KM, Howard AW, Crepp JR (2010) Giant planet occurrence in the stellar massmetallicity plane. PASP 122:905–915. https://doi.org/10.1086/655775, 1005.3084 Kasting JF, Whitmire DP, Reynolds RT (1993) Habitable zones around main sequence stars. Icarus 101:108–128. https://doi.org/10.1006/icar.1993.1010 Kjeldsen H, Bedding TR, Butler RP et al (2005) Solar-like oscillations in ˛ centauri B. ApJ 635:1281–1290. https://doi.org/10.1086/497530 Latham DW, Stefanik RP, Mazeh T, Mayor M, Burki G (1989) The unseen companion of HD114762 – a probable brown dwarf. Nature 339:38–40. https://doi.org/10.1038/339038a0 Lovis C, Dumusque X, Santos NC et al (2011, submitted) The HARPS search for southern extrasolar planets XXXI. Magnetic activity cycles in solar-type stars: statistics and impact on precise radial velocities. A&A arXiv:11075325 arXiv:1107.5325 Mahadevan S, Ramsey L, Bender C et al (2012) The habitable-zone planet finder: a stabilized fiberfed NIR spectrograph for the Hobby-Eberly Telescope. Proc SPIE 8446:84461S. Ground-based and airborne instrumentation for astronomy IV. https://doi.org/10.1117/12.926102, 1209.1686. Marcy GW, Butler RP (1992) Precision radial velocities with an iodine absorption cell. PASP 104:270–277. https://doi.org/10.1086/132989 Mayor M Queloz D (1995) A Jupiter-Mass companion to a solar-type star. Nature 378:355

30 Radial Velocities as an Exoplanet Discovery Method

631

Pepe FA, Lovis C (2008) From HARPS to CODEX: exploring the limits of Doppler measurements. Phys Scr T130(1):014007. https://doi.org/10.1088/0031-8949/2008/T130/014007 Pepe FA, Cristiani S, Rebolo Lopez R et al (2010) ESPRESSO: the Echelle spectrograph for rocky exoplanets and stable spectroscopic observations. In: Ground-based and airborne instrumentation for astronomy III. Proc SPIE 7735:77350F. https://doi.org/10.1117/12.857122 Queloz D, Henry GW, Sivan JP et al (2001a) No planet for HD166435. A&A 379:279–287 Queloz D, Mayor M, Udry S et al (2001b) From CORALIE to HARPS. The way towards 1 m s1 precision Doppler measurements. Messenger 105:1–7 Quirrenbach A, Amado PJ, Mandel H et al (2010) CARMENES: Calar Alto high-resolution search for M dwarfs with exo-earths with a near-infrared Echelle spectrograph. In: Ground-based and airborne instrumentation for astronomy III. Proc SPIE 7735:773513. https://doi.org/10.1117/12. 857777 Robertson P, Mahadevan S (2014) Disentangling planets and stellar activity for Gliese 667C. ApJ 793:L24. https://doi.org/10.1088/2041-8205/793/2/L24, 1409.0021 Robertson P, Mahadevan S, Endl M, Roy A (2014) Stellar activity masquerading as planets in the habitable zone of the M dwarf Gliese 581. Science 345:440–444. https://doi.org/10.1126/ science.1253253, 1407.1049 Wright JT (2005) Radial velocity jitter in stars from the California and Carnegie planet search at keck observatory. PASP 117:657–664. https://doi.org/10.1086/430369 Wright JT, Eastman JD (2014) Barycentric corrections at 1 cm s1 for precise Doppler velocities. PASP 126:838–852. https://doi.org/10.1086/678541, 1409.4774 Wright JT, Gaudi BS (2013) Exoplanet detection methods, p 489. https://doi.org/10.1007/978-94007-5606-9_10 Wright JT, Howard AW (2009) Efficient fitting of multiplanet Keplerian models to radial velocity and astrometry data. ApJS 182:205–215. https://doi.org/10.1088/0067-0049/182/1/ 205, 0904.3725 Wright JT, Marcy GW, Butler RP et al (2008) The Jupiter twin HD 154345b. ApJ 683:L63–L66. https://doi.org/10.1086/587461, arXiv:0802.1731 Wright JT, Roy A, Mahadevan S et al (2013) MARVELS-1: a face-on double-lined binary star masquerading as a resonant planetary system and consideration of rare false positives in radial velocity planet searches. ApJ 770:119. https://doi.org/10.1088/0004-637X/770/2/119, 1305.0280

Transit Photometry as an Exoplanet Discovery Method

31

Hans J. Deeg and Roi Alonso

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fundamentals of the Transit Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Detection Probability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . False Positives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Transit Surveys: Factors Affecting their Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Transit Detection in Light Curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Transit Surveys: Past, Current, and Future Projects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Surveys for Planets of Low-Mass Stars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross-References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

634 637 639 639 640 644 645 651 652 652 653

Abstract

Photometry with the transit method has arguably been the most successful exoplanet discovery method to date. A short overview about the rise of that method to its present status is given. The method’s strength is the rich set of parameters that can be obtained from transiting planets, in particular in combination with radial velocity observations; the basic principles of these parameters are given. The method has however also drawbacks, which are the low probability that transits appear in randomly oriented planet systems and

H. J. Deeg ()R. Alonso Instituto de Astrofísica de Canarias, La Laguna, Tenerife, Spain Departamento de Astrofísica, Universidad de La Laguna, La Laguna, Tenerife, Spain e-mail: [email protected]; [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_117

633

634

H. J. Deeg and R. Alonso

the presence of astrophysical phenomena that may mimic transits and give rise to false detection positives. In the second part, we outline the main factors that determine the design of transit surveys, such as the size of the survey sample, the temporal coverage, the detection precision, the sample brightness and the methods to extract transit events from observed light curves. Lastly, an overview over past, current, and future transit surveys is given. For these surveys we indicate their basic instrument configuration and their planet catch, including the ranges of planet sizes and stellar magnitudes that were encountered. Current and future transit detection experiments concentrate primarily on bright or special targets, and we expect that the transit method remains a principal driver of exoplanet science, through new discoveries to be made and through the development of new generations of instruments.

Introduction Since the discovery of the first transiting exoplanet, HD 209458b (Henry et al. 2000; Charbonneau et al. 2000), the transit method has become the most successful detection method, surpassing the combined detection counts of all other methods (see Fig. 1) and giving rise to the most thoroughly characterized exoplanets at present. The detection of planetary transits is among the oldest planet detection methods; together with the radial velocity (RV) method, it was proposed in 1952 in a brief paper by Otto Struve (see also  Chap. 3, “Prehistory of Transit Searches”). The early years of exoplanet discoveries were however dominated by planets found by RVs, and prior to the 1999 discovery of transits on HD 209458, the transit method was not considered overly promising by the community at large. For example, a 1996 (Elachi et al.) NASA Road Map for the Exploration of Neighboring Planetary Systems (ExNPS) revises in some detail the potential of RV, astrometry, and microlensing detections, with a recommended focusing onto space interferometry, while transits were considered only cursory. Consequently, activities to advance transit detections were rather limited; most notable are early proposals for a spacebased transit search by Borucki, Koch, and collaborators (Borucki et al. 1985; Koch et al. 1996; Borucki et al. 1997; see also  Chap. 56, “Space Missions for Exoplanet Science: Kepler/K2”) and the TEP project, a search for transiting planets around the eclipsing binary CM Draconis that had started in 1994 (Deeg et al. 1998; Doyle et al. 2000; see also  Chap. 5, “The Way to Circumbinary Planets”). The discovery of the first transiting planets, with some of them like HD 209458b already known from RV detections, quickly led to intense activity to more deeply characterize them, mainly from multicolor photometry (e.g., Jha et al. 2000; Deeg et al. 2001) or from spectroscopy during transits (e.g., Queloz et al. 2000; Charbonneau et al. 2002; Snellen 2004); to provide the community with efficient transit fitting routines (e.g., Mandel and Agol 2002; for more see below), or to extract the most useful

31 Transit Photometry as an Exoplanet Discovery Method

635

Fig. 1 The fractions by which various detection methods contributed to the accumulated sample of known planets are shown, for years since 1995. At the end of 1995, only five planets were known, three from pulsar timing and two from radial velocities. Between 1996 and 2013, the sample of known planets was dominated by those discovered with radial velocities, while in 2018, 78% of all known planets had been discovered by transits. Based on data from the NASA Exoplanet Archive in Feb. 2018, and using its classification by discovery methods. ‘Timing’ includes planets found by pulsar timing, eclipse timing, or transit timing. Other detection methods (astrometry, orbital brightness variations) generate only a very small contribution that is barely visible at the bottom of the graph, for years following 2010

set of physical parameters from transit light curves (Seager and Mallén-Ornelas 2003). The first detections of transits also provided a strong motivation toward the setup of dedicated transit searches, which soon led to the first planet discoveries by that method, namely, OGLE-TR-56b (Konacki et al. 2003) and further planets by the OGLE-III survey, followed by TrES-1 (Alonso et al. 2004a), which was the first transit discovery on a bright host star. Transits were therefore established as a valid method to find new planets (see also  Chap. 4, “Discovery of the First Transiting Planets”). Central to the method’s acceptance was also the fact that planets discovered by transits across bright host stars permit the extraction of a wealth of information from further observations. Transiting planets orbiting bright host stars, such as HD 209458b, HD 189733 (Bouchy et al. 2005), WASP-33b (Christian et al. 2006; Collier Cameron et al. 2010), or the terrestrial planet 55 Cnc e (McArthur et al. 2004; Winn et al. 2011), are presently the planets about which we have the most detailed knowledge. Besides RV observations for the mass and orbit determinations, further

636

H. J. Deeg and R. Alonso

characterization may advance with the following techniques: transit photometry with increased precision or in different wavelengths, transit photometry to derive transit timing variations (TTVs), and spectroscopic observations during transits (transmission spectroscopy, line-profile tomography of exoplanet transits, RossiterMcLaughlin effect). Furthermore, the presence of transits – strictly speaking primary transits of a planet in front of its central star – usually implies the presence of secondary eclipses or occultations, when a planet disappears behind its central star. These eclipses, as well as phase curves of a planet’s brightness in dependence of its orbital position, might be observable as well (see corresponding chapters for these techniques). Given the modest instrumental requirements to perform such transit searches on bright star samples – both HD209458b’s transits and the planet TReS-1 were found with a telescope of only 10 cm diameter – the first years of the twenty-first century saw numerous teams attempting to start their own transit surveys. Also, the two space-based surveys that were launched a few years later, CoRoT and Kepler, are unlikely to have received the necessary approvals without the prior ground-based discovery of transiting planets (see  Chap. 54, “Space Missions for Exoplanet Research: Overview and Introduction” and chapters on the individual missions). The enthusiasm for transit search projects at that time is well represented by a paper by Horne (2003) which lists 23 transit surveys that were being prepared or already operating. Its title “Hot Jupiters Galore” also typifies the expectation that significant numbers of transiting exoplanets will be found in the near future: Summing all 23 surveys, Horne predicted a rate of 191 planet detections per month! In reality, advances were much slower, with none of these surveys reaching the predicted productivity. By the end of 2007, before the first space-based discoveries from the CoRoT mission (Barge et al. 2008; Alonso et al. 2008), only 27 planets had been found through transit searches. This slower advance can be traced to two issues that revealed themselves only during the course of the first surveys: The amount of survey time required under real conditions was higher than expected, and the presence of red noises degraded sensitivity to transit-like events (see later in this chapter). Once these issues got understood and accounted for, some of these ground-based surveys became very productive, and both WASP and HAT/HATS have detected over 100 planets to date. The next major advances based on the transit method arrived with the launch of the space missions CoRoT in 2006 and Kepler in 2009. These led to the discoveries of transiting terrestrial-sized planets (CoRoT-7b by Léger et al. 2009, Kepler-10b by Batalha et al. 2011) to planets in the temperate regime (CoRoT-9b, Deeg et al. 2010), to transiting multi-planet systems (Lissauer et al. 2011), and to a huge amount of transiting planets that permit a deeper analysis of planet abundances in a very large part of the radius – period (or Teff) parameter space (see  Chap. 96, “Planet Occurrence: Doppler and Transit Surveys”). In the following, an introduction is given on the methodology of the transit detection and its surveys, as well as an overview about the principal projects that implement these surveys.

31 Transit Photometry as an Exoplanet Discovery Method

637

Fundamentals of the Transit Method A schematic view of a transit event is given in Fig. 2, where the bottom part represents the observed flux of the system. As the planet passes in front of the star, its flux diminishes by a fractional denoted as F . Under the assumptions of negligible flux from the planet and of spherical shapes of the star and planet, F is given by the ratio of the areas of the planet and the star:  F 

Rp Rs

2

D k2

(1)

where Rp is the radius of the planet, Rs the radius of the star, and k is the radius ratio. The total duration of the transit event is represented as tT , and the time of totality, in which the entire planet disk is in front of the stellar disk (the time

Fig. 2 Outline of the transit of an exoplanet, with the main quantities used to describe the orbital configuration, from the observables given in the lower solid curve (the observed light curve) to the model representations from the observer’s point of view (central panel) or other viewpoints (top panels). Note that in the central panel, projected views of a and i are drawn

638

H. J. Deeg and R. Alonso

between second and third contacts, using eclipse terminology), is given by tF , during which the light curve is relatively flat. Using basic geometry, the work by Seager and Mallén-Ornelas (2003) derived analytic expressions that relate these observables to the orbital parameters. In particular, the impact parameter b, defined as the minimal projected distance to the center of the stellar disk during the transit, can be expressed as:

b

n .1  k/2  Œsin2 .t =P /= sin2 .t =P /.1 C k/2 o1=2 a F T cos i D Rs cos2 .tF =P /= cos2 .tT =P /

(2)

where a is the orbital semimajor axis, i the orbital inclination, and P the orbital period. A commonly used quantity that can be obtained from photometric data alone is the so-called scale of the system or the ratio between the semimajor axis and the radius of the star: p a 1 D .1 C k/2  b 2 Rs tan.tT =P /

(3)

which, using Kepler laws of motion and making the reasonable approximations of the mass of the planet being much smaller than the mass of its host star, and assuming a spherical shape for the star, can be transformed into a measurement of the mean stellar density: 3 s D GP 2



a Rs

3 (4)

Consistency between this measurement and a stellar density estimated by other means (through spectroscopy, mass-radius relations, or asteroseismology) has often been used as a way to prioritize the best transit candidates from a survey (Seager and Mallén-Ornelas 2003; Tingley et al. 2011; Kipping et al. 2014). In the previous equations, we have assumed circular orbits for simplicity; a derivation of equivalent equations including the eccentricity terms can be found in Tingley et al. (2011, with a correction in Eq. 15 of Parviainen et al. 2013). While the previous expressions allow quick estimates of the major parameters of an observed transit, more sophisticated methods using the formalisms of Mandel and Agol (2002) or Giménez (2006) are commonly used for their more precise derivation. This is in part due to a subtle effect visible in Fig. 2: the limb-darkening of the star that manifests itself as a nonuniform brightness of the stellar disk, which is described in detail in  Chap. 67, “Stellar Limb Darkening’s Effects on Exoplanet Characterization”. The limb-darkening of the star makes it challenging to determine the moments of second and contact of the transit and to precisely measure F . For more detailed introductions into the parameters that can be measured from transits, we refer to Winn (2010) and Haswell (2010).

31 Transit Photometry as an Exoplanet Discovery Method

639

Detection Probability The geometric probability to observe a planet in transit is given by Winn (2010):  pt ra D

Rs ˙ R p a



1 C e sin ! 1  e2

 (5)

where the C sign is used to include grazing transits and the  sign refers to the probability of full transits that have second and third contacts. For a typical Hot Jupiter with a semimajor axis of 0.05 AU, this probability is on the order of 10%, while for an earth-like planet at 1 AU from a solar-like star, it goes down to 0.5%. These relatively low probabilities are the main handicap of the transit method, since the majority of existing planet systems will not display transits.

False Positives A transit-shaped event in a light curve is not always caused by a transiting planet, as there are a number of astrophysical configurations that can lead to similar signatures. These are the so-called false positives in transit searches, which have been a nuisance of transit surveys since their early days. One example of a false positive would be a stellar eclipsing binary that is so close to a brighter single star that the light of both objects falls within the same photometric aperture of a detector: the deep eclipses of the eclipsing binary are diluted due to the flux of the brighter star, resulting in a light curve with a shallower eclipse, with a shape very similar to a transiting planet. More complete descriptions of the types of false positives that affect transit searches, and their expected frequencies, can be found in Brown (2003), Alonso et al. (2004b), Almenara et al. (2009), and Santerne et al. (2013). To detect false positives and to confirm the planetary nature of a list of candidates provided by a transit survey, a series of follow-up observations are required (e.g., Latham 2003; Alonso et al. 2004b; Latham 2007, 2008; Deeg et al. 2009; Moutou et al. 2013; Günther et al. 2017b), which apply to both ground-based or space-based surveys. Traditionally, the confirmation that a transit signal is caused by a planet takes place when its mass is measured with high-precision RV measurements. In some cases, particularly with planets orbiting faint host stars, or for the confirmation of the smallest planets, the achievable RV precision is insufficient to measure the planet’s mass. As these cases are of high interest, for example, planets with similar sizes as the Earth orbiting inside the habitable zone of its host star, statistical techniques have been developed to estimate the probability of the observed signals being due to planets relative to every other source of false positives we know of. In this case, the planets are known as validated. Current validation procedures use the fact that astrophysical false-positive scenarios have very low probabilities when several transiting signals are seen on the same star (Lissauer et al. 2012), or they use all the available information (observables and knowledge of the galactic

640

H. J. Deeg and R. Alonso

population and stellar evolution) to compare the probabilities of a signal being due to a transiting planet vs. anything else. A few examples of validation studies are Torres et al. (2011), Morton (2012), Lissauer et al. (2014), Rowe et al. (2014), Díaz et al. (2014), Torres et al. (2015), Morton et al. (2016), and Torres et al. (2017), some of which use these current state-of-the-art validation procedures: BLENDER, VESPA, or PASTIS. Finally, some false positives may be due to artifacts from red noise or other instrumental effects, even in the most precise surveys to date (e.g., Coughlin et al. 2014). In a few cases, planets that were previously validated have been disproved after an independent analysis (Cabrera et al. 2017; Shporer et al. 2017), which should generate some caution about the use of results from validations, which are statistical by design.

Transit Surveys: Factors Affecting their Performance The task of surveying a stellar sample for the presence of transiting planets must overcome the inherent inefficiencies of the transit method: The planets need to be aligned correctly (see previous section), and the observations must be made when transits occur. The expected abundances of the desired planet catch must be taken into account, and their transits need to be detectable with sufficient photometric precision. Furthermore, transit-like events (false positives) may arise from other astrophysical as well as instrumental sources, and means to identify them need to be provided. The success of a transit detection experiment must take these factors into account, which are discussed in the following: Sample size The probability pt r for transits to occur in a given random-oriented system is between a few percent for Hot Jupiters and less than 0.1% for cool giant planets. In order to achieve a reasonable probability that N transiting system will be found in a given stellar field, the number of surveyed stars (that is, stars for which light curves with sufficient precision for transit detection are obtained) should be at least Nsurvey  N =.ptra f /, where f is the fractional abundance of the detectable planet population in the stellar sample. For surveys of Hot Jupiters, with f  1% of main-sequence (MS) stars (Wright et al. 2012; Mayor et al. 2011), this leads to minimum samples of 2000 MS stars to expect a single transit discovery. Given that most stars in the bright samples of small-telescope surveys are not on the MS, sample sizes of 5000–10000 targets are however more appropriate. Survey fields that provide sufficient numbers of suitable stars, by brightness and by desired stellar type, need therefore be defined. The size of the sample is then given by the size of the field of view ( fov) and by the spatial density of suitable target stars, which depends on the precision of the detector (primarily depending on the telescope aperture) and on the location of the stellar fields. Also, in most surveys, sample size is increased through successive observations of different fields.

31 Transit Photometry as an Exoplanet Discovery Method

641

Temporal coverage At any given moment, the probability for the observation of a transit of a correctly aligned system is p  tT =P , where tT is the duration of a transit and P the orbital period. This probability goes from 5–8% for Ultrashort periodic planets over 2–3% for typical Hot Jupiter systems to 0.15% for an Earth-Sun alike. For an estimation of the number of transits for a given sample at a given time, we need to multiply this probability and the probability for correct alignment with the abundance of detectable planets. To determine a planet’s period, of course at least two transits need to be observed. The requirement to observe three transit-like events that are periodic has however been habitual in ground-based observations, which are prone to produce transit-like events from meteorologic and other non-astronomic causes. Furthermore, for an increased S/N of transit detections, especially toward the detection of smaller planets, as well as toward a more precise derivation of physical parameters, the rule is “the more transits, the better.” Continuous observational coverage is the most time-efficient way to achieve the observation of a minimum number of transits (e.g., Ntr;min 3) for a given system. However, only space missions are able to observe nearly continuously over a timescale of weeks, which is the only way to ascertain that transiting planets above some size threshold and below some maximum period are being detected with near certainty. Ground-based surveys, with their interruptions from the day/night cycle and from meteorological incidences, can only seek reasonable probabilities (but no certainty) to catch a desired number of transits from a given planet. The principal factor that determines the number of observed transits in a given discontinuous light curve is a planet’s orbital phase (taken at some reference time, such as the beginning of observations) or its epoch (the time when one of its transits occurs); both are of course unknown prior to a planet’s discovery. An example of the effect of phase on the number of observed transits in discontinuous data is shown in Fig. 3. As a rough rule, in order to achieve reasonable detection probabilities (e.g., 70%) for typical Hot Jupiters (P D 34 d) with a requirement of 3 observed transits, ground-based surveys should cover a stellar field for at least 300 h. Transit detection precision A basic version of the S/N of a single transit is given by the ratio .S =N /t r  F =lc

(6)

where F is the fractional flux loss during a transit and lc is the fractional noise of the light curve on the timescale of the transit duration TT . This noise is composed of various sources, most notably photon noise from the target and the surrounding sky background, cosmic ray hits, CCD read noise, and flat-fielding or jitter noises (which arise from variations of the positions or shapes of stellar point spread functions on detectors whose sensitivity is not uniform). For groundbased surveys, we also have to add variations from atmospheric transparency and scintillation noise. Figure 4 shows the scatter over 1 h timescales from the most precise space-based survey, Kepler, and from NGTS, one of the leading groundbased surveys.

642

H. J. Deeg and R. Alonso

Fig. 3 The expected number of transits by a test planet with a period of 5 days that would have been observed in a ground-based light curve covering 617 h (D 25:7 d). The clockwise direction is the planet’s phase at some time of reference, and the radial distance gives the number of transits that would have been observed at each phase. Achieving that the large majority of the potential phases would produce at least three transits required an observational coverage that was much longer than three times the orbital period. (Adapted from Deeg et al. 1998)

Early estimates for planet detection yields assumed commonly a white-noise scaling from the point-to-point scatter of an observed light curve to the usually much longer duration of a transit. In practice, red or correlated noises degrade the precision of nearly all photometric time series over longer timescales, as was first shown by Pont et al. (2006), based on data from the OGLE-III transit survey. Only the space-based data from the Kepler mission uphold a white-noise scaling from their acquisition cycle of 30 min to a transit-like duration of 6 h (Jenkins et al. 2010a; see Gilliland et al. 2011 for more details on Kepler’s noise properties). The CoRoT mission, in contrary to Kepler on a low Earth orbit, produced light curves that on timescales of 2 h were already about twice as noisy as would be the result of a white-noise scaling from their acquisition cycle of 8.5 min (Aigrain et al. 2009). At least as strongly affected are ground-based surveys, with the principal culprit being the nightly air mass variation, which is on a similar timescale as the duration of most transits. Correlated noises have been the principal source for the early overestimations of detection yields. In the case of SuperWASP, recognizing their influence led to a revision of detection yields and to an increase in temporal coverage early in its operational phase (Smith et al. 2006). For surveys that attempt to detect shallow transits, brightness variations due to the sample stars’ activity might also be of concern. The demonstration that this variability does not prevent the detection of terrestrial planets of solar-like stars (Jenkins 2002) was an important advance during the development of the Kepler mission. Aigrain et al. (2004) found

Fractional Precision (ppm)

31 Transit Photometry as an Exoplanet Discovery Method

643

104

103

102

101 7

8

9

10

11

12

13

14

15

16

Magnitude

Fig. 4 Comparison of the precision between Kepler (gray points) and NGTS (violet points) photometry. Given is the light curves’ rms scatter on a 1h timescale. The magnitudes for NGTS are in the I-band; for Kepler they are in its own system. The Kepler noises are p from long-cadence (0.5h cycle time) curves of Q1 targets (Jenkins et al. 2010a) and scaled by 1=2 toward the 1h timescale. The NGTS data are from 695 h of monitoring with a 12 s cadence, rebinned to exposure times of 1 h, from Wheatley et al.p(2018). The difference between NGTS and Kepler precision would be reduced by a factor of 950=200  2:2 if the different telescope aperture sizes are taken into account. The precision for the brightest NGTS targets is limited by scintillation noise, which is independent of the targets’ brightness

then that K stars are the most promising targets for transit surveys, while a survey’s performance drops significantly for stars earlier than G and younger than 2.0 Gyr. For a quantitative discussion of the factors that influence the yield of transit surveys, we refer to Beatty and Gaudi (2008). Algorithms to dampen red noises and other systematic effects have been developed to either “clean” directly a light curve from their influences or as part of a detection algorithm, thereby increasing its sensitivity towards transit-like features. Examples are the pre-whitening employed in the Kepler pipeline (Jenkins et al. 2010b), the cleaning of CoRoT light curves (Guterman et al. 2015), or the widely used SYSREM (Tamuz et al. 2005) and TFA (Kovács et al. 2005) algorithms. Brightness of the sample: Rejection of false positives and characterization of the planet catch As mentioned, a large number of transit surveys were initiated in the first years of the twenty-first century, after the discovery of the first transiting planets. These early efforts were aimed about equally at deep surveys of small fields using larger (1m and more) telescopes and at shallow surveys with small instruments having wide fields of view. The surveys with larger telescopes, including early projects with Hubble Space Telescope (Gilliland et al. 2000, on the 47 Tucanae globular cluster, and the SWEEPS survey by Sahu et al. 2006), were met however with limited success, with the most productive one becoming the OGLE-III (Udalski 2003) survey using a dedicated 1m telescope. Besides the difficulties to get access

644

H. J. Deeg and R. Alonso

to the required large facilities to perform a deep survey over a sufficiently long time, a major drawback of such surveys is the faintness of the sample. RV verifications or further observational refinements of their transit detections are either impossible, or if possible at all, they likely require the largest existing telescope facilities. For example, from the SWEEPS survey that targeted the Sagittarius I window of the Galactic bulge with the Hubble Space Telescope, Sahu et al. (2006) report the detection of transits on 16 targets. Their faintness of V D 18.8–26.2 as well as crowding permitted however only for the two of them (SWEEPS-04 and SWEEPS11) a confirmation as planets, based on RVs taken with the 8 m VLT. All other SWEEPS detections have remained in candidate status until the present. We also note the comparatively small impact (relative to brighter targets) of the very large number of planets found on the fainter end of the Kepler mission’s sample (Rowe et al. 2014 with 815 planets, Morton et al. 2016 with 1284 planets). These planets count only with probabilistic validations, and their principal usefulness are statistical studies on planet abundances across their known parameters (radius, period, central star type, planet multiplicity). The brightness of a target sample is therefore a very valuable parameter toward the science return of a transit survey! The most common follow-up observations of transit detections are RV measurements, which do not only prove (or disprove) a planet’s existence beyond reasonable doubt but also greatly improve our knowledge about them, providing masses, orbital eccentricity, and, occasionally, also the detection of further non-transiting planets in the same system. In practice, from the RV follow-up of numerous candidates for the Kepler, K2, and CoRoT missions, we found that a magnitude of 14.5 is a soft limit for their routinary follow-up. This is due to that brightness being near the limit for RV measurements at several relatively well-accessible midsized telescopes with appropriate instrumentation (e.g., the FIES instrument on the 2.5 m Nordic Optical Telescope or the HARPS instruments on the 3.6 m Telescopio Nazionale Galileo (TNG) and on the ESO 3.6 m telescope). For transiting systems of bright central stars, a host of further possibilities to examine these systems opens up – such as observation of the Rossiter-McLaughlin effect, transit spectroscopy, secondary eclipse measurements, or the detection of phase curves (see this handbook’s part on Exoplanet Characterization). For this reason – increased knowledge about the discovered systems – the upcoming spacebased transit surveys, TESS and PLATO, will focus on samples that are brighter than those of Kepler and CoRoT, while ground-based surveys continue with their efforts to find transiting planets principally on bright or on special types of target stars.

Transit Detection in Light Curves Efficient recognition of transit-like features in light curves is a central part of any transit detection experiment. This task is usually performed in two steps. In the first one, detection statistical values that describe the likelihood of a light curve to contain a transit-like event are assigned. These values might also be expressed

31 Transit Photometry as an Exoplanet Discovery Method

645

as a function of a candidate planet’s size, period, and further parameters. In the second step, these statistical values are evaluated, and those candidates that deserve closer investigation are extracted. We reproduce here a description of this step in the Kepler pipeline, from Jenkins et al. (2010b): “Light curves whose maximum folded detection statistic exceeds 7.1 are designated Threshold Crossing Events (TCEs) and subjected to a suite of diagnostic tests in Data Validation (DV) to fit a planetary model to the data and to establish or break confidence in the planetary nature of the transit-like events.” The threshold value for the extraction of candidates needs to be chosen with care, as it must provide a balance between the number of false positives – which increases to unmanageable levels if the threshold is too low – and the risk to miss detections of true planets if the threshold is too high. As representative transit detection methods and algorithms, we mention here the early work on matched-filter detection algorithms by Jenkins et al. (1996), which provided the basics for the transit detection of the early TEP observing project as well as for the Kepler mission, the widely used box least-squares (BLS) algorithm (Kovács et al. 2002) with derivatives (e.g., Collier Cameron et al. 2006), or algorithms using wavelets (e.g., Régulo et al. 2007). For the second step of a detection procedure, the evaluation of a transit candidate as a planet-like event, usually a more detailed modelling (or fitting) of the light curve of the presumed transit is performed. The Mandel and Agol (2002) algorithm and the analytical eclipsing formulae by Giménez (2006) are widely used basic transit modellers that have also been integrated into several transit fitting packages. For an overview of such tools, we refer to  Chap. 76, “Tools for Transit and Radial Velocity Modeling and Analysis”.

Transit Surveys: Past, Current, and Future Projects The Extrasolar Planets Encyclopaedia (http://www.exoplanet.eu/research/) lists currently websites of 39 planet search projects that indicate “transits” as a principal observing method. These projects include finished ones, currently operating ones, projects that are in various preparation stages, as well as projects or proposals that have never moved beyond some design phase. It includes also some projects that aren’t dedicated to the discovery but to the follow-up of transiting planets, such as ESA’s CHEOPS space mission. In Table 1 and in the following notes, we provide an overview over a selection of well-known transit detection surveys. For most of them, this handbook contains also dedicated chapters, listed in the Cross-References. The columns of Table 1 have the following meaning: years: The years of operation. config: Instrument configuration, with the aperture diameters of the individual optical units (cam = camera). fovsingle : The sky area in deg2 covered by a single optical unit of the detection experiment.

years

2001–2009 2003–2010

2003–2014

Since 2003

Since 2009

Since 2004

Since 2005

Since 2016

Name

OGLE TrES

XO

HATnet

HATSouth

WASP

KELT

NGTS

f ovsingle (deg2 )

Ground-based surveys Single telescope of 1000 mmØ 0.34 3 sites, each with one telescope of 36 100 mmØ Single site (3 sites in 2012–2014) with 49 common mount for 2 cams. 2 sites, one (FLWO) with 5 cams and 64 one (Mauna Kea) with 2 cams, all individually mounted with 110 mmø 3 sites, each with 2 mounts that each 17 hold 4 cameras with 180 mmØ, covering adjacent fields in 2  2 pattern 2 sites, each with single-mount array of 60 8 cams of 111 mmØ covering adjacent fields 2 sites, each with single camera of 676 42 mmØ Single site with 12 telescopes of 8 200 mmØ on independent mounts

config

Table 1 Selected transit surveys

96

480

67

300

f ovinstr (deg2 /

1

1.33

1.11, 1.53, 1.91

0.85, 1.24, 1.71

146c

19

0.87, 1.25, 1.75

0.82, 1.27, 1.79

64c

36

0.97, 1.20, 2.07

1.08, 1.25, 1.61 1.10, 1.21, 1.71

R05 ; R50 ; R95 (Rjup )

8

8 5

Nplanet

15.5

7.6,10.0, 11.5

9.8, 11.5, 12.9

12.1, 13.5, 14.6

10.0, 11.8, 13.6

9.8, 11.1, 12.1

14 0 15 0 15 8b 11.4, 11.8, 13.7

m05 ; m50 ; m95 (Vmag )

646 H. J. Deeg and R. Alonso

2007–2012 2008–2013 2014–2018 Launched 2018

Launch 2026

CoRoT Kepler K2 TESS

PLATO

Space-based surveys Hubble Space Telescope (2.5 mØ) with 0.003 ACS cam Space Telescope of 270 mmØ 3.5/1.7a Space Telescope of 950 mmØ 107/102e Same as Kepler 97/92e 4 cams of 105 mmØ each covering 576 adjacent areas 24 cams of 100 mmØ in 4 groups, with 1100 6 co-aligned cams in each group. Partial overlap of all 24 units. 2 further “fast cams” with color filter and short cycle time 2232

2300

a

Notes Original fov and after March 2009, due to failed DPU b I-band magnitudes c Includes three planets common to HAT and WASP d Average of the two confirmed planets e fov values before/after Jan 2010 (Kepler) and before/after July 2016 (K2), due to detector failures f Kepler bandpass magnitudes

2006

SWEEPS 35 2300 176

2 0.3, 1.05, 1.47 0.083, 0.189, 0.633 0.097, 0.202, 1.05

0.97d 11.9, 14.9, 15.8 11.8, 14.6, 15.9f 11.1, 13.3, 15.7f

19.3d

31 Transit Photometry as an Exoplanet Discovery Method 647

648

H. J. Deeg and R. Alonso

fovinstr : The sky area in deg2 that is covered simultaneously be the experiment at a single site in its usual operating mode. Only given if there are multiple optical units at a site. Npl : Count of planet detections in February 2018, based on the number of planets carrying an instrument’s designation in the name. Planets labeled with other designations, such as HD or GJ numbers, are missed. R05 , R50 , R95 : 5th, 50th (median), and 95th percentiles of the radii of the detected planets. If Npl  20, the smallest and largest planets are given. m05 , m50 , m95 : Similar to the previous but indicating the V-mag brightness of the detected systems. Planet counts, radii, and magnitudes are from the Extrasolar Planets Encyclopaedia and from the NASA Exoplanet Explorer. Below, some notes are provided for the transit surveys listed in Table 1. OGLE-III The Optical Gravitational Lensing Experiment has been implemented in four phases, with the fourth one operational at present (spring 2018). OGLE has been dedicated to the detection of substellar objects from microlensing, except during its third phase (OGLE-III, Udalski 2003), when the observing procedure of the 1m OGLE telescope at Las Campanas Observatory was modified to enable the detection of transits. OGLE-TR-56 was the first planet discovered in a transit search, with a posterior verification from RV follow-up (Konacki et al. 2003). TrES The “Transatlantic Exoplanet Survey” was the first project with instruments that were specifically designed and dedicated for transit surveying. Its first telescope, originally named STARE, was used in the 1999 discovery of the transits of HD 209458b during tests at the High Altitude Observatory at Boulder, Colorado. In 2001, it was relocated to Teide Observatory, Tenerife, where a systematic transit search began. Since 2003, the project operated under the TrES name, after the merger with two other projects using similar instrumentation, namely, PSST at Lowell Observatory and the Sleuth Project at Palomar Observatory (O’Donovan 2008). The principal success of TrES was the detection of the first transiting planets orbiting bright stars (TrES 1, Alonso et al. 2004a; TrES, 2 O’Donovan et al. 2006) by a dedicated survey. TrES was discontinued in 2010. XO This survey started in 2003 at a single site, with a second phase observing from three sites from 2012 to 2014. The CCDs are red in time-delayed integration (TDI): pixels are red continuously, while stars move along the columns on the detector, owing to a slewing motion of the telescope. This setup enlarges the effective field of view and results in stripes of 7ı  43ı that are acquired during each single exposure. HAT This denominator (Hungarian-made Automated Telescope) encompasses two surveys: For one, HATnet operates since 2003 seven CCD cameras with 110 mm apertures on individual mounts, with five of them at Fred Lawrence Whipple Observatory at Mount Hopkins in Arizona and two at Mauna Kea Observatory

31 Transit Photometry as an Exoplanet Discovery Method

649

in Hawaii. For another, HATSouth (Bakos et al. 2013) is a network across three sites in the southern hemisphere that is able to track stars continuously over longer time spans. Since 2009, it operates at Las Campanas Observatory (Chile), at the High Energy Stereoscopic System site (Namibia), and at Siding Spring Observatory (Australia). Each of these sites contains two mounts, with each of them holding four Takahashi astrographs with individual apertures of 180 mm. The HAT consortium is also advancing the HATPI Project of an all-sky camera consisting of 63 optical units on a single mount. WASP (Wide Angle Search for Planets, see Pollacco et al. 2006 for an instrument description; Smith et al. 2014 for a review.) This consortium operates two instruments: SuperWASP-North, since 2004 at Roque de los Muchachos Observatory on the Canary Island of La Palma, and WASP-South, since 2006 at the South African Astronomical Observatory. A predecessor instrument, WASP0, was operated during the year 2000 on La Palma. SuperWASP-North is an array of 8 cameras covering 4802 of sky with each exposure; WASP-South is a close copy of it. WASP is currently the ground-based search that has detected the most planets, among them several exoplanets (such as WASP-3b, 12b, 43b) that stand out for their excellent suitability for deeper characterization work, due to their short orbital period and/or large size. KELT The “Kilodegree Extremely Little Telescope” has to date been the most successful survey using very wide-field detectors (with a fov of 26ı  26ı ) with commercial photographic optics of short focal length. KELT-North operates since 2005 from Winer Observatory, Arizona, and KELT-South since 2009 from Sutherland, South Africa. Both instruments use a CCD camera with an 80 mm/f1.8 Mamya lens. NGTS The Next-Generation Transit Survey (Wheatley et al. 2013, 2018) is operated by a consortium of seven institutions from Chile, Germany, Switzerland, and the United Kingdom. After testing in La Palma and at Geneva Observatory, operations started in 2016 at ESO’s Paranal Observatory. NGTS employs an automated array of twelve 20-centimeter f/2.8 telescopes on independent mounts, sensitive to orange to near-infrared wavelengths (600–900 nm). It is a successor project to WASP that achieves significantly better photometric precision (Fig. 5) but with a focus on late-type stars. Its first planet discovery was the most massive planet known to transit an M-dwarf (Bayliss et al. 2018). Simulations for a 4-year survey predict the discovery of about 240 planets, among them about 20 planets of 4 REarth or less (Günther et al. 2017a). CoRoT Named after “Convection, Rotation, and Transits,” this was the first space mission dedicated to exoplanets. Launched in December 2006 by the French space agency CNES and partners into a low polar orbit for a survey lasting initially 4 years, it surveyed 163,665 targets distributed over 26 stellar fields in two opposite regions in the galactic plane, with survey coverages lasting between 21 and 152 days

650

H. J. Deeg and R. Alonso

Fig. 5 Single transit observations of the Hot Jupiter WASP- 4b with one NGTS telescope unit (top) and WASP (bottom). (From Wheatley et al. 2018, reproduced with permission)

(Deleuil et al. 2018). In May 2009, its first data processing unit failed, and CoRoT’s fov was reduced to half, while the failure of the other unit in Nov. 2012 caused the end of the mission. Its most emblematic discovery was CoRoT-7b, the first transiting terrestrial planet (Léger et al. 2009) Kepler This NASA mission was launched in 2009 into an Earth-trailing orbit, for a mission of 4 years to survey a single field of 170,000 stars, principally for the presence of Earth-sized planets. Kepler has discovered the majority of currently known exoplanets, with discoveries that have revolutionized the field of exoplanets. In contrary to the planets found by any other transit survey, only a small fraction (3%) of Kepler planets are Jupiter-sized ( 0:9Rjup ), while the vast majority are Earth- or super-Earth-sized ones. Science operations under the “Kepler” denomination ended in May 2013 when two of the spacecraft’s reaction wheels failed and its pointing become unreliable. K2 In March 2014, the Kepler spacecraft was returned into service under the K2 name. Its observing mode was adapted to the reduced number of reaction wheels, surveying fields near the ecliptic plane for about 80 days each (Howell et al. 2014).

31 Transit Photometry as an Exoplanet Discovery Method

651

Planets found by K2 have a rather similar size distribution to Kepler, albeit with a somewhat larger fraction of giant planets (8% are larger than 0.9Rjup ). K2 is expected to end around Oct. 2018, when the spacecraft runs out of fuel. TESS The “Transiting Exoplanets Survey Satellite” by NASA aims to scan about 85% of the entire sky for transits across relatively bright stars (Ricker et al. 2015). Most areas will be covered by pointings lasting 28 days. The spacecraft harbors four wide-field telescopes that cover jointly a stripe of the sky of 24ı by 96ı . TESS was launched in April 2018 into an elliptical orbit with a 13.7-day period in a 2:1 resonance with the Moon’s orbit, for a mission with an initial duration of 2 years. PLATO This ESA mission, named after “PLAnetary Transits and Oscillation of stars,” is expected to be launched in 2026 into an orbit around the L2 point, to perform during at least 4 years a survey of several large sky areas (Rauer et al. 2014). The mission’s core sample are 15,000 stars of 8  mV  11, while a secondary “statistical” sample includes 245,000 targets up to mV  16. PLATO will have four groups of detectors, each with six cameras that all point to the same fov. Between the groups there is a partial overlap due to which areas near the center of the common fov will be covered by all 24 cameras, while outer zones will be covered by 6 or 12 cameras only. Two additional “fast” cameras with rapid cycle times and color filters will survey the brightest stars of 4–8 mV .

Surveys for Planets of Low-Mass Stars Several surveys, which are not listed in Table 1, have been designed specifically for the detection of planets around low-mass stars, and in particular, M-stars. Given the difficulties to detect planets in the habitable zone of solar-like stars, planet searches around such stars provide an alternative path for the detection of potentially habitable planets (e.g., Scalo et al. 2007). Their small size permits that terrestrial planets produce transits that are deep enough to be observable from moderate ground-based instruments. Also, the habitable zone around these stars corresponds to orbital periods of a few days to weeks, making habitable planets’ transits shorter, more frequent, and hence easier to detect than for solar-type stars. Disadvantages of low-mass stars as targets are however a flux variability that is exhibited by most of them and the sparsity of such stars with sufficient apparent brightness. As a consequence, these detection projects are not performed as wide-field surveys but as searches that point to selected target stars, which are covered sequentially. As such, these projects cover relatively few targets and have only a small planet catch but may provide discoveries of large impact toward our knowledge of potentially habitable planets. MEarth This project operates since 2008, consisting of eight 40 cm telecopes at Mount Hopkins, Arizona, and since 2014, of a similar setup at Cerro Tololo, Chile (Berta et al. 2012). MEarth has discovered several small planets, among them

652

H. J. Deeg and R. Alonso

LHS1140b, a planet of 1.4 REarth in the habitable zone of an M-dwarf at a distance of 10.5 parsec (Dittmann et al. 2017). TRAPPIST/SPECULOOS The “TRAnsiting Planets and PlanetesImals Small Telescope” survey consists of two 60 cm robotic telescopes, one operating since 2010 at ESO’s La Silla Observatory, Chile, and one since 2016 at Oukaimeden Observatory, Morocco. It has the dual objective of transit detection and the study of comets and other small bodies in the Solar System (Jehin et al. 2014). Its outstanding discovery has been the TRAPPIST-1 system of seven planets, with some of them in the habitable zone, around an ultracool M8 dwarf at a distance of 12 parsec (Gillon et al. 2017). TRAPPIST is also a prototype of the SPECULOOS (Search for habitable Planets EClipsing ULtra-cOOl Stars) project, whose first phase will consist of four 1m robotic telescopes at ESO’s Paranal Observatory.

Conclusion In the year 2003, K. Horne predicted the success of transit surveys in a paper entitled “Hot Jupiters Galore.” It took longer than expected to get to that point and required the understanding and resolution of several subtle issues affecting these surveys, but today the paper’s title has become reality, and the discovery of transiting planets is commonplace. This applies not only to Hot Jupiters but also to planets across the entire size regime and has been a consequence of the continued refinement of observing techniques and of the development of new instruments, both ground and space based. At the time of writing, the transit method is expected to remain the largest contributor toward the discovery of new planets and planet systems, with several ambitious ground- and space-based searches under way. Planet systems found in transit searches will also continue to provide the motivation for the continued development of instruments and observing techniques, which take advantage of the opportunities for deeper insights that transiting systems are offering. In that sense, systems found by transit surveys will continue as a basic nutrient of the field of exoplanet science. For further reading about transits as a tool to detect and characterize exoplanets, we refer to the reviews by Winn (2010) and by Cameron (2016) and to a book dedicated to transiting exoplanets by Haswell (2010).

Cross-References  Characterization of Exoplanets: Secondary Eclipses  CoRoT: The First Space-Based Transit Survey to Explore the Close-in Planet

Population  Discovery of the First Transiting Planets  Exoplanet Phase Curves: Observations and Theory

31 Transit Photometry as an Exoplanet Discovery Method

653

 KELT: The Kilodegree Extremely Little Telescope, a Survey for Exoplanets

Transiting Bright, Hot Stars  Microlensing Surveys for Exoplanet Research (OGLE Survey Perspective)  Planet Occurrence: Doppler and Transit Surveys  Populations of Extrasolar Giant Planets from Transit and Radial Velocity Surveys  Prehistory of Transit Searches  Radial Velocities as an Exoplanet Discovery Method  Small Telescope Exoplanet Transit Surveys: XO  Space Missions for Exoplanet Science: Kepler/K2  Space Missions for Exoplanet Science: PLATO  SPECULOOS Exoplanet Search and Its Prototype on TRAPPIST  The HATNet and HATSouth Exoplanet Surveys  The Rossiter-McLaughlin Effect in Exoplanet Research  Tools for Transit and Radial Velocity Modeling and Analysis  Transit-Timing and Duration Variations for the Discovery and Characterization of

Exoplanets Acknowledgements Financial support by the Spanish Secretary of State for R&D&i (MINECO) is acknowledged by HD under the grant ESP2015-65712-C5-4-R and by RA for the Ramón y Cajal program RYC-2010-06519 and the programs RETOS ESP2014-57495-C2-1-R and ESP201680435-C2-2-R. This contribution has benefited from the use of the NASA Exoplanet Archive and the Extrasolar Planets Encyclopaedia, and the authors acknowledge the people behind these tools.

References Aigrain S, Favata F Gilmore G (2004) Characterising stellar micro-variability for planetary transit searches. A&A 414:1139–1152 Aigrain S, Pont F, Fressin F et al. (2009) Noise properties of the CoRoT data. A planet-finding perspective. A&A 506:425–429 Almenara JM, Deeg HJ, Aigrain S et al. (2009) Rate and nature of false positives in the CoRoT exoplanet search. A&A 506:337–341 Alonso R, Brown TM, Torres G et al. (2004a) TrES-1: The transiting planet of a bright K0 V star. ApJ 613:L153–L156 Alonso R, Deeg HJ, Brown TM Belmonte JA (2004b) Strategies to recognize false alarms in transit experiments: experiences from the STARE project. In: Favata F, Aigrain S Wilson A (eds) Stellar structure and habitable Planet finding, vol 538. ESA Special Publication, Noordwijk, pp 255–259 Alonso R, Auvergne M, Baglin A et al. (2008) Transiting exoplanets from the CoRoT space mission. II. CoRoT-Exo-2b: a transiting planet around an active G star. A&A 482:L21–L24 Bakos GÁ, Csubry Z, Penev K et al. (2013) HATSouth: a global network of fully automated identical wide-field telescopes. PASP 125:154 Barge P, Baglin A, Auvergne M et al. (2008) Transiting exoplanets from the CoRoT space mission. I. CoRoT-Exo-1b: a low-density short-period planet around a G0V star. A&A 482:L17–L20 Batalha NM, Borucki WJ, Bryson ST et al. (2011) Kepler’s first rocky planet: Kepler-10b. ApJ 729:27 Bayliss D, Gillen E, Eigmuller P et al. (2018) NGTS-1b: a hot Jupiter transiting an M-dwarf. MNRAS 475:4467–4475 Beatty TG, Gaudi BS (2008) Predicting the yields of photometric surveys for transiting extrasolar planets. ApJ 686:1302–1330

654

H. J. Deeg and R. Alonso

Berta ZK, Irwin J, Charbonneau D, Burke CJ Falco EE (2012) Transit detection in the MEarth survey of nearby M dwarfs: bridging the clean-first, search-later divide. AJ 144:145 Borucki WJ, Scargle JD Hudson HS (1985) Detectability of extrasolar planetary transits. ApJ 291:852–854 Borucki WJ, Koch DG, Dunham EW Jenkins JM (1997) The Kepler mission: a mission to detennine the frequency of inner planets near the habitable zone for a wide range of stars. In: Soderblom D (ed) Planets beyond the solar system and the next generation of space missions. Astronomical Society of the Pacific conference series, vol 119. p 153 Bouchy F, Udry S, Mayor M et al. (2005) ELODIE metallicity-biased search for transiting hot Jupiters. II. A very hot Jupiter transiting the bright K star HD 189733. A&A 444:L15–L19 Brown TM (2003) Expected detection and false alarm rates for transiting Jovian planets. ApJ 593:L125–L128 Cabrera J, Barros SCC, Armstrong D et al. (2017) Disproving the validated planets K2-78b, K282b, and K2-92b. The importance of independently confirming planetary candidates. A&A 606:A75 Cameron AC (2016) Extrasolar planetary transits. In: Bozza V, Mancini L, Sozzetti A (eds) Methods of detecting exoplanets: 1st advanced school on exoplanetary science. Astrophysics and space science library, vol 428, p 89. https://doi.org/10.1007/978-3-319-27458-4_2 Charbonneau D, Brown TM, Latham DW, Mayor M (2000) Detection of planetary transits across a sun-like star. ApJ 529:L45–L48 Charbonneau D, Brown TM, Noyes RW, Gilliland RL (2002) Detection of an extrasolar planet atmosphere. ApJ 568:377–384 Christian DJ, Pollacco DL, Skillen I et al. (2006) The SuperWASP wide-field exoplanetary transit survey: candidates from fields 23 h < RA < 03 h. MNRAS 372:1117–1128 Collier Cameron A, Pollacco D, Street RA et al. (2006) A fast hybrid algorithm for exoplanetary transit searches. MNRAS 373:799–810 Collier Cameron A, Guenther E, Smalley B et al (2010) Line-profile tomography of exoplanet transits – II. A gas-giant planet transiting a rapidly rotating A5 star. MNRAS 407:507–514 Coughlin JL, Thompson SE, Bryson ST et al. (2014) Contamination in the Kepler field. Identification of 685 KOIs as false positives via ephemeris matching based on Q1–Q12 data. AJ 147:119 Deeg HJ, Doyle LR, Kozhevnikov VP et al (1998) Near-term detectability of terrestrial extrasolar planets: TEP network observations of CM Draconis. A&A 338:479–490 Deeg HJ, Garrido R Claret A (2001) Probing the stellar surface of HD 209458 from multicolor transit observations. New Astronomy 6:51–60 Deeg HJ, Gillon M, Shporer A et al (2009) Ground-based photometry of space-based transit detections: photometric follow-up of the CoRoT mission. A&A 506:343–352 Deeg HJ, Moutou C, Erikson A et al. (2010) A transiting giant planet with a temperature between 250K and 430K. Nature 464:384–387 Deleuil M, Aigrain S, Moutou C et al (2018) Planets, candidates, and binaries from the corot/exoplanet program. A&A, in print Díaz RF, Almenara JM, Santerne A et al (2014) PASTIS: Bayesian extrasolar planet validation – I. General framework, models, and performance. MNRAS 441:983–1004 Dittmann JA, Irwin JM, Charbonneau D et al (2017) A temperate rocky super-earth transiting a nearby cool star. Nature 544:333–336 Doyle LR, Deeg HJ, Kozhevnikov VP et al. (2000) Observational limits on terrestrial-sized inner planets around the CM Draconis system using the photometric transit method with a matchedfilter algorithm. ApJ 535:338–349 Elachi C, Angel R, Beichman CA et al (1996) A road map for the exploration of neighboring planetary systems (ExNPS). Jet Propulsion Laboratory report, NASA Gilliland RL, Brown TM, Guhathakurta P et al (2000) A lack of planets in 47 Tucanae from a hubble space telescope search. ApJ 545:L47–L51 Gilliland RL, Chaplin WJ, Dunham EW et al (2011) Kepler mission stellar and instrument noise properties. ApJS 197:6

31 Transit Photometry as an Exoplanet Discovery Method

655

Gillon M, Triaud AHMJ, Demory BO et al (2017) Seven temperate terrestrial planets around the nearby ultracool dwarf star TRAPPIST-1. Nature 542:456–460 Giménez A (2006) Equations for the analysis of the light curves of extra-solar planetary transits. A&A 450:1231–1237 Günther MN, Queloz D, Demory BO, Bouchy F (2017a) A new yield simulator for transiting planets and false positives: application to the next generation transit survey. MNRAS 465: 3379–3389 Günther MN, Queloz D, Gillen E et al (2017b) Centroid vetting of transiting planet candidates from the next generation transit survey. MNRAS 472:295–307 Guterman P, Mazeh T, Faigler S (2015) Exposure-based algorithm for removing systematics out of the CoRoT light curves. In: Martins F, Boissier S, Buat V, Cambrésy L, Petit P (eds) SF2A2015: proceedings of the annual meeting of the French society of astronomy and astrophysics, pp 277–281 Haswell CA (2010) Transiting exoplanets. Cambridge University Press, Cambridge. ISBN:9780521139380 Henry GW, Marcy GW, Butler RP, Vogt SS (2000) A transiting “51 Peg-like” planet. ApJ 529: L41–L44 Horne K (2003) Status and prospects of planetary transit searches: hot Jupiters galore. In: Deming D, Seager S (eds) Scientific frontiers in research on extrasolar planets. Astronomical Society of the Pacific conference series, vol 294. pp 361–370 Howell SB, Sobeck C, Haas M et al (2014) The K2 mission: characterization and early results. PASP 126:398 Jehin E, Opitom C, Manfroid J, Hutsemékers D, Gillon M (2014) The TRAPPIST comet survey. In: Muinonen K, Penttilä A, Granvik M et al. (eds) Asteroids, comets, meteors 2014: Proceedings of the conference held 30 June – 4 July, 2014 in Helsinki, Finland Jenkins JM (2002) The impact of solar-like variability on the detectability of transiting terrestrial planets. ApJ 575:493–505 Jenkins JM, Doyle LR, Cullers DK (1996) A matched filter method for ground-based sub-noise detection of terrestrial extrasolar planets in eclipsing binaries: application to CM Draconis. Icarus 119:244–260 Jenkins JM, Caldwell DA, Chandrasekaran H et al (2010a) Initial characteristics of Kepler long cadence data for detecting transiting planets. ApJ 713:L120–L125 Jenkins JM, Chandrasekaran H, McCauliff SD et al (2010b) Transiting planet search in the Kepler pipeline. In: Software and cyberinfrastructure for astronomy. Proceeding of SPIE, vol 7740, p 77400D. https://doi.org/10.1117/12.856764 Jha S, Charbonneau D, Garnavich PM et al. (2000) Multicolor observations of a planetary transit of HD 209458. ApJ 540:L45–L48 Kipping DM, Bastien FA, Stassun KG et al (2014) Flicker as a tool for characterizing planets through asterodensity profiling. ApJ 785:L32 Koch D, Borucki W, Cullers K et al (1996) System design of a mission to detect earth-sized planets in the inner orbits of solar-like stars. J Geophys Res 101:9297–9302 Konacki M, Torres G, Jha S, Sasselov DD (2003) An extrasolar planet that transits the disk of its parent star. Nature 421:507–509 Kovács G, Zucker S, Mazeh T (2002) A box-fitting algorithm in the search for periodic transits. A&A 391:369–377 Kovács G, Bakos G, Noyes RW (2005) A trend filtering algorithm for wide-field variability surveys. MNRAS 356:557–567 Latham DW (2003) Spectroscopic follow-up observations of planetary transit candidates identified by project vulcan. In: Deming D Seager S (eds) Scientific frontiers in research on extrasolar planets. Astronomical Society of the Pacific conference series, vol 294. pp 409–412 Latham DW (2007) Spectroscopic and photometric follow-up observations. In: Afonso C, Weldrake D, Henning T (eds) Transiting extrapolar planets workshop. Astronomical Society of the Pacific conference series, vol 366, p 203

656

H. J. Deeg and R. Alonso

Latham DW (2008) Characterization of terrestrial planets identified by the Kepler mission. Physica Scripta 130(1):014034 Léger A, Rouan D, Schneider J et al (2009) Transiting exoplanets from the CoRoT space mission. VIII. CoRoT-7b: the first super-earth with measured radius. A&A 506:287–302 Lissauer JJ, Fabrycky DC, Ford EB et al (2011) A closely packed system of low-mass, low-density planets transiting Kepler-11. Nature 470:53–58 Lissauer JJ, Marcy GW, Rowe JF et al (2012) Almost all of Kepler’s multiple-planet candidates are planets. ApJ 750:112 Lissauer JJ, Marcy GW, Bryson ST et al (2014) Validation of Kepler’s multiple planet candidates. II: refined statistical framework and descriptions of systems of special interest. ApJ 784, 44 Mandel K, Agol E (2002) Analytic light curves for planetary transit searches. ApJ 580:L171–L175 Mayor M, Marmier M, Lovis C et al (2011) The HARPS search for southern extra-solar planets XXXIV. Occurrence, mass distribution and orbital properties of super-earths and Neptune-mass planets. ArXiv:11092497 McArthur BE, Endl M, Cochran WD et al (2004) Detection of a Neptune-Mass planet in the 1 Cancri system using the hobby-eberly telescope. ApJ 614:L81–L84 Morton TD (2012) An efficient automated validation procedure for exoplanet transit candidates. ApJ 761:6 Morton TD, Bryson ST, Coughlin JL et al (2016) False positive probabilities for all Kepler objects of interest: 1284 newly validated planets and 428 likely false positives. ApJ 822:86 Moutou C, Deleuil M, Guillot T et al (2013) CoRoT: harvest of the exoplanet program. Icarus 226:1625–1634 O’Donovan FT (2008) The detection and exploration of planets from the Trans-atlantic Exoplanet Survey. PhD thesis, California Institute of Technology O’Donovan FT, Charbonneau D, Mandushev G et al (2006) TrES-2: the first transiting planet in the Kepler field. ApJ 651:L61–L64 Parviainen H, Deeg HJ, Belmonte JA (2013) Secondary eclipses in the CoRoT light curves. A homogeneous search based on Bayesian model selection. A&A 550:A67 Pollacco DL, Skillen I, Collier Cameron A et al (2006) The WASP project and the SuperWASP cameras. PASP 118:1407–1418 Pont F, Zucker S, Queloz D (2006) The effect of red noise on planetary transit detection. MNRAS 373:231–242 Queloz D, Eggenberger A, Mayor M et al (2000) Detection of a spectroscopic transit by the planet orbiting the star HD209458. A&A 359:L13–L17 Rauer H, Catala C, Aerts C et al (2014) The PLATO 2.0 mission. Exp Astron 38:249–330 Régulo C, Almenara JM, Alonso R, Deeg H, Roca Cortés T (2007) TRUFAS, a wavelet-based algorithm for the rapid detection of planetary transits. A&A 467:1345–1352 Ricker GR, Winn JN, Vanderspek R et al (2015) Transiting exoplanet survey satellite (TESS). J Astron Telesc Instrum Syst 1(1):014003 Rowe JF, Bryson ST, Marcy GW et al (2014) Validation of Kepler’s multiple planet candidates. III. Light curve analysis and announcement of hundreds of new multi-planet systems. ApJ 784:45 Sahu KC, Casertano S, Bond HE et al (2006) Transiting extrasolar planetary candidates in the Galactic bulge. Nature 443:534–540 Santerne A, Fressin F, Díaz RF et al (2013) The contribution of secondary eclipses as astrophysical false positives to exoplanet transit surveys. A&A 557:A139 Scalo J, Kaltenegger L, Segura AG et al (2007) M stars as targets for terrestrial exoplanet searches and biosignature detection. Astrobiology 7:85–166 Seager S Mallén-Ornelas G (2003) A unique solution of planet and star parameters from an extrasolar planet transit light curve. ApJ 585:1038–1055 Shporer A, Zhou G, Vanderburg A et al (2017) Three statistically validated K2 transiting warm Jupiter exoplanets confirmed as low-mass stars. ApJ 847:L18 Smith AMS WASP Consortium (2014) The SuperWASP exoplanet transit survey. Contributions of the Astronomical Observatory Skalnate Pleso 43:500–512

31 Transit Photometry as an Exoplanet Discovery Method

657

Smith AMS, Collier Cameron A, Christian DJ et al (2006) The impact of correlated noise on SuperWASP detection rates for transiting extrasolar planets. MNRAS 373:1151–1158 Snellen IAG (2004) A new method for probing the atmospheres of transiting exoplanets. MNRAS 353:L1–L6 Struve O (1952) Proposal for a project of high-precision stellar radial velocity work. Observatory 72:199–200 Tamuz O, Mazeh T Zucker S (2005) Correcting systematic effects in a large set of photometric light curves. MNRAS 356:1466–1470 Tingley B, Bonomo AS Deeg HJ (2011) Using stellar densities to evaluate transiting exoplanetary candidates. ApJ 726:112 Torres G, Fressin F, Batalha NM et al (2011) Modeling Kepler transit light curves as false positives: rejection of blend scenarios for Kepler-9, and validation of Kepler-9 d, a super-earth-size planet in a multiple system. ApJ 727:24 Torres G, Kipping DM, Fressin F et al (2015) Validation of 12 small Kepler transiting planets in the habitable zone. ApJ 800:99 Torres G, Kane SR, Rowe JF et al (2017) Validation of small Kepler transiting planet candidates in or near the habitable zone. AJ 154:264 Udalski A (2003) The optical gravitational lensing experiment. Real time data analysis systems in the OGLE-III survey. Acta Astron 53:291–305 Wheatley PJ, Pollacco DL, Queloz D et al (2013) The next generation transit survey (NGTS). Eur Phys J Web Conf 47:13002. https://doi.org/10.1051/epjconf/20134713002 Wheatley PJ, West RG, Goad MR et al (2018) The next generation transit survey (NGTS). MNRAS 475, 4476–4493 Winn JN (2010) Exoplanet transits and occultations. In: Seager S (ed) Exoplanets. University of Arizona Press, Tucson, pp 55–77. arXiv:1001.2010 Winn JN, Matthews JM, Dawson RI et al (2011) A super-Earth transiting a naked-eye star. ApJ 737:L18 Wright JT, Marcy GW, Howard AW et al (2012) The frequency of hot Jupiters orbiting nearby solar-type stars. ApJ 753:160

Finding Planets via Gravitational Microlensing

32

Virginie Batista

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Microlensing Basics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Simple Lens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Multiple Lens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Finite Source Effects and Computation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Microlens Parallax, E . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Orbital Motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mass Estimates from Galactic Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Detections Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cold Super-Earths and Neptunes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Triple Lens Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . An Exomoon Around a Free-Floating Planet? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Massive Planets Around Very Low-Mass Stars/BDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stellar Remnants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Exoplanet Mass-Ratio Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Galactic Distribution of Planets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Free-Floating Planets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion and Future Prospects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acronyms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

660 661 661 662 663 666 668 670 671 671 672 673 674 675 675 676 678 679 681 682 682

Abstract

Gravitational microlensing is a technique to probe compact objects toward the center of the galaxy, such as distant stars, planets, white and brown dwarfs,

V. Batista () Institut d’astrophysique de Paris, Paris, France Centre National d’Etudes Spatiales, Paris Cedex 1, France e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_120

659

660

V. Batista

black holes, and neutron stars. Since the first microlensing planet discovered in 2003, more than 40 planets have been detected with this technique, as well as several black hole candidates, and a population of potential free-floating planets. This chapter first provides a presentation of the microlensing theory, including numerical aspects to solve binary and triple lens problems, and a discussion of the microlensing planetary detection efficiency, with a high potential regarding cold planets beyond the snow line. It also explains how the planetary characterization can be facilitated when the microlensing light curves exhibit distortions due to second-order effects, such as parallax, planetary orbital motion, and extended source, and how they can also introduce degeneracies in the models. The chapter then reviews the main discoveries to date and the recent statistical results from high-cadence ground-based surveys and space-based observations, especially on the planet mass function and the distance distribution of the microlensing planetary systems. Finally, future prospects are discussed, with the expected advances from dedicated space missions, extending the planet sensitivity range down to Mercury masses.

Introduction Gravitational microlensing is not the easiest way to discover planets, and the development of this technique over the last 30 years required tremendous efforts and steady determination from the scientific community. Even Einstein, who was the first to really bring attention to this phenomenon Einstein (1936), i.e., the light of a distant source being bent by the gravitational field of a massive body, was pessimistic about the chances of observing such a phenomenon. Indeed, the microlensing optical depth toward the galactic center is of about v 2 =c 2  106 , meaning millions of stars need to be monitored in order to catch a handful of magnified ones. Later on, several people considered experiments of microlensing observations (e.g., Refsdal 1964; Chang and Refsdal 1979), and finally Paczynski (1986) managed to convince the international community to develop the MACHO, EROS, and OGLE experiments to search for compact objects in the galactic halo (see acronyms at the end of the chapter). Hunting for microlensing events toward the galactic center with the intention to discover planets appeared a few years later, under the suggestion of Mao and Paczynski (1991) and Gould and Loeb (1992), who gave a detailed description of microlensing planetary perturbations. Nowadays, thousands of microlensing events are observed per year, leading to the detection of up to ten exoplanets for each season, accounting for more than 40 planetary systems published to date and ten more under analysis. This amount is quite modest compared to the total amount of discovered exoplanets, but this technique offers a considerable advantage in probing a parameter space unaccessible to other methods. Earth mass planets can be detected by microlensing (Bennett and Rhie 1996), within a range of separations to the host star that is complementary to other techniques. Microlensing is indeed mostly sensitive to planets orbiting just outside

32 Finding Planets via Gravitational Microlensing

661

of the snow line, where planetary formation is predicted to be the most efficient according to the core accretion theory (Ida and Lin 2004). Moreover, as the only effect measured by microlensing is that of deflecting light by gravitational distortions of a given field, there is no need to measure the light of the lens system itself. This opens a domain of exploration down to very faint, even invisible objects, such as brown dwarfs, extremely far systems in the galactic bulge, as well as black holes, neutron stars, and free-floating planets. More generally, the formation processes may be influenced by many environmental factors, such as star metallicity (Wang and Fischer 2015; Zhu et al. 2016), as well as the mass of the host star (Johnson et al. 2010), the star multiplicity (Kaib et al. 2013), and larger scale conditions like higher ambient radiations and star density in the galactic bulge (Thompson 2013). Thus, the distant population of planets unveiled by microlensing provides critical tests to the planetary formation theories.

Observations The strategy to hunt for planets via microlensing has evolved over the last 10 years. The microlensing community has made a transition toward a network of wide-field imagers. They now conduct high-cadence surveys from different longitudes of the southern hemisphere. Since 2006, the MOA collaboration (Bond et al. 2001; Sumi et al. 2013) uses a 2.2-deg2 FOV CCD camera (Sako et al. 2008) mounted on a 1.8 m MOA-II telescope in New Zealand. The OGLE collaboration (Udalski 2003) upgraded its camera in 2010 to a 1.4-deg2 FOV on a 1.3-m telescope in Chile. These two groups monitor hundreds of fields in the bulge and alert 2000–3000 microlensing events each year. Additionally, (KMTNet, Kim et al. 2016) became operational in 2015, with three telescopes equipped with a 4 deg2 camera, in Chile, South Africa, and Australia. The Wise observatory in Israel also conducts a survey (Gorbikov et al. 2010), despite its less favorable northern latitude. In parallel to these surveys, several follow-up groups (e.g., PLANET, microFUN, RobotNet, MiNDSTEp) dedicate their telescopes to the most interesting events, with observatories at all longitudes to ensure a round-the-clock coverage of the planetary anomalies. In 2014–2017, the microlensing campaigns from the ground were also combined with space-based observations using the telescopes Spitzer (Werner et al. 2004; Gould and Horne 2013) and Kepler (K2-C9) (Howell et al. 2014), as a pathway toward the future space missions WFIRST (Spergel et al. 2015) and Euclid (Penny et al. 2013).

Microlensing Basics The basic principle of a microlensing effect is illustrated in Fig. 1, showing a light ray coming from a source S being deflected by a point-like mass L, called “lens.” The light is deviated by an angle ˛, creating the illusion for the observer (O) that the

662

V. Batista S’

α

I

S rmin θ O

β

DL

L

P DS

Fig. 1 Deviation of a light ray from a source S due to the gravitational field of a microlens L

source is at the position I (“image”) in the lens plane. Under the assumption of small angles, the three angles of this figure are linked by the lens equation, ˛.DS DL / D  DS ˇDS , where ˛ D 4GM =.c 2 DL  /, with M the lens mass, G the gravitational constant, and c the speed of light. The deviation of light does not affect only one ray but an ensemble of rays coming from the source, so when the source and the lens are perfectly aligned on the observer’s line of sight, ˇ D 0 and the source image in the lens plan appears as a ring, called the “Einstein ring,” whose angular size E is expressed as s E D

M

DS  DL ; DS DL

where  D 4G=c 2 ' 8:14 kpc:M1 ˇ :

(1)

The lens equation is thus commonly expressed as ˇ D   E2 =. The Einstein angular radius E is used as a scale standard in microlensing units, and as long as its value remains unknown, most microlensing parameters will be expressed in units of E .

Simple Lens If we normalize the angles by E and define the parameters u D ˇ=E and z D =E , we can express the lens-source separation u as a function of the separation z between the image of the source and the lens, both being projected in the lens plane, u D z  1=z. If the lens and thepsource are not perfectly aligned (u ¤ 0), this equation has two solutions: z˙ D ˙. u2 C 4 ˙ u/=2. The positive solution creates the major image, and the negative one the minor image, outside and inside of the Einstein ring, respectively.

32 Finding Planets via Gravitational Microlensing

663

The magnification of the source by the lens is then estimated as the area ratio of the images over the source. Since the images are tangentially stretched by a factor z˙ =u compared to the source and radially shrunk by a factor d z˙ =d u, the magnification is then expressed as ˇ ˇ  2 ˇ z˙ d z ˙ ˇ 1 u C2 ˇ ˇ D ˙1 ; A˙ D ˇ p u du ˇ 2 u u2 C 4

u2 C 2 A.u/ D p : u u2 C 4 (2) In its simplest form, the lens-source relative movement is rectilinear, and u can be expressed as a function of the impact parameter u0 (in units of E ), the time of maximum magnification t0 , and the Einstein crossing time tE : " u.t / D u20 C



thus in total;

t  t0 tE

2 #1=2 :

(3)

The observed flux is a function of time, F .t / D A.t /Fs C Fb , where Fs is the source flux and Fb the blended light (any unresolved light) that is not magnified. It reaches its maximum at t0 , when the impact parameter is at its minimum u0 . The smaller u0 is, the higher the maximum of magnification will be. For a point-sourcepoint-lens (PSPL) event, the light curve can be fit by five parameters: t0 , u0 , tE , Fs , and Fb . In practice, each observatory has its specific Fs and Fb , since different observatories may have different filter bandpasses and resolutions.

Multiple Lens We now consider a lens composed of NL objects, which total mass will be M D NL ˙iD1 mi . Let  i be the two-dimensional coordinates of these individual masses in the lens plane. The deflection angle of the source is then expressed by 4G ˛./ D 2 c



1 1  DL DS

X NL

mi

i

  i ; j   i j2

(4)

where  are the angular positions of the images. It is common to consider the lens equation in complex coordinates (Witt 1990; Rhie 1997). The source position u D .; / and the image position z D .x; y/ can then be defined in the complex form, u D  C i  and z D x C iy, respectively. The lens equation thus becomes uDz

NL X mi =M i

zN  zNi

;

(5)

where zN is the complex conjugate of z. Equation (5) can be solved numerically to find the image position zj , where j is the index for the source images and i those of the

664

V. Batista

individual lens masses. The magnification can be greater or less than 1, depending on the images area, and is equal to the inverse of the Jacobian matrix determinant,

Aj D

1 ˇˇ ; ˇ detJ zDzj

detJ 

@u @u @.; / D1 : @.x; y/ @z @z

(6)

The total magnification is the sum of individual ones: A D ˙j jAj j. Singularities occur for positions of the source where detJ D 0, theoretically inducing an infinite magnification in the case of a point source. In practice, the source cannot be considered as a point in most cases, and the singularities appear as very high but finite magnifications.  P From Eq. (5), @u=@Nz D "i =.zNi  zN/2 , where "i D mi =M is the mass fraction of individual masses mi . According to Eq. (6), the image positions for which detJ D 0 are given by ˇ2 ˇN L ˇ ˇX "i ˇ ˇ ˇ D 1: ˇ 2ˇ ˇ . z N  z N / i i

(7)

The solutions of Eq. (7) shape as closed contours in the lens plane, called “critical curves.” To these curves correspond an ensemble of source positions in the source plane, called “caustics.” When the lens is composed of two masses, NL D 2, the lens equation is expressed by uDzC

"1 "2 C : zN1  zN zN2  zN

(8)

Equation (8) can be written as a 5-degree polynomial, which coefficients are given in Witt and Mao (1995). The solutions to this polynomial are not necessarily solutions to Eq. (8), because when the source is outside of the caustics, two of the five solutions yield to an imaginary magnification. Thus, the number of images in the lens plane varies from three to five when the source crosses a caustic. Schneider and Weiss (1986) showed that there are three different topologies of caustics for a given value of mass ratio q between the two components of the lens. These three topologies present either one, two, or three caustics. As shown on Fig. 2, in the case of small separations between the two lenses, d 1 (in Einstein units), there are three caustics, one being the central caustic (with 4 cusps) and two secondary ones (with 3 cusps). The transition toward big separations goes through an intermediary phase that creates a single caustic with six cusps, of bigger size, called “resonant caustic.” This configuration happens when the separation is close to the Einstein radius. Finally, for large separations, this caustic splits into two caustics, with four cusps each. We then naturally call these three domains “close,” “resonant,” and “wide” separations.

32 Finding Planets via Gravitational Microlensing

665

Fig. 2 Topology of caustics for three regimes of separation, “close” (top), “resonant” (middle), and “wide” (bottom)

b 0.40

0.40

0.20

0.20 θ2 /θE

θ2 /θE

a

–0.00

–0.00

–0.20

–0.20

–0.40

–0.40

–0.6

–0.4

–0.2

0.0

0.2

θ1/θE

Fig. 3 Example of d

0.4

–0.2

0.0

0.2

0.4

0.6

0.8

θ1/θE

! d 1 degeneracy

For triple lens modeling, Rhie (2002) gives the lens equation for point-source calculations. It is solved the same way as the binary lens equation, but the polynomial is of tenth order instead of fifth. For mass ratios q 1 and a given separation d , the shape and size of the central caustic are invariant under the d ! d 1 transformation. This duality is often called “close/wide degeneracy” (Griest and Safizadeh 1998; Dominik 1999; Albrow et al. 1999; An 2005). It comes from the fact that the Taylor development of the lens equation to the second degree are identical for d ! d 1 .1 C q/1=2 , i.e., d ! d 1 when q 1. Figure 3 gives an illustration of this degeneracy, where the source passes very close to the central caustic of a system with a q D 0:006 mass ratio. Two (d; 1=d ) models provide the same light curve. This figure also points out the importance of high-magnification events, for which u0 1. Indeed, when the source passes extremely close to the central caustic, it probes the effect of the planetary companion on the caustic shape. Significant efforts have been expended on

666

V. Batista

the observation of high-magnification events (HME), because they probe the central caustic (Griest and Safizadeh 1998; Rhie et al. 2000; Rattenbury et al. 2002). Any planets in the system are likely to affect the central caustic, resulting in potentially high sensitivity to the presence of low-mass planets, whose signature can be detected at the peak of the microlensing event.

Finite Source Effects and Computation As previously mentioned, in reality the source is a disk, and the point-like approximation is only valid when the source is far enough from the lens (projected in the lens plane). This means that when the impact parameter is very small, u0 1, finite source effects have to be taken into account at the peak of the microlensing event and require specific numerical treatments. It is also true when the source approaches or crosses a caustic in the case of multiple lens systems. In the first case, the corresponding effect on the light curve peak is very noticeable, with a rounded shape and a damped magnification. This effect was first analyzed by Gould (1994) and Nemiroff and Wickramasinghe (1994). They showed that the deviation from the standard PSPL light curve provides informations on the source proper motion, under assumptions on its brightness and color. Indeed, one can determine E D  = , where is the source radius in Einstein units and  the physical source size in as. The first parameter is derived from the light curve fit, and the latter from available informations on the source star color (Bensby et al. 2013) and brightness (Nataf et al. 2013), and empirical color/brightness relations (Kervella et al. 2004). Thus, the lens-source relative proper motion is derived by  D E =tE . When the source crosses a caustic, the finite source effects start to be noticeable as the source projected distance from the caustic is of a few source radii (Pejcha and Heyrovsky 2009). Figure 4 gives an illustration of this effect. Similarly, caustic crossings enable us to determine the source radius in Einstein units, as well as its luminosity profile, i.e., limb-darkening effects on its disk. Formally, the source magnification can be calculated by integration over its entire surface, each point having its own magnification. Concretely, this approach is hard to implement due to divergences near the caustics. Several algorithms and techniques have been developed to solve the extended source case (Dominik 1995; Bennett and Rhie 1996; Wambsganss 1997; Gould and Gaucherel 1997; Gould 2008; Pejcha and Heyrovsky 2009; Bennett 2010). The most efficient approach seems to be the one that consists in cutting the light curve in segments as a function of the source distance to the caustics. Two different methods are often applied, one when the source is relatively far from the caustics and the other for close approaches or crossings of caustics. The first method is based on an analytical approximation, such as the “hexadecapole” approximation (Gould 2008; Pejcha and Heyrovsky 2009) that evaluates the magnification of 13 points well distributed on the surface of the source. The second one is a more complex numerical evaluation of the source magnification that integrates the images in the lens plane. One possibility is to compute an

32 Finding Planets via Gravitational Microlensing

667

80 r = 0.0002

Amplification

60

r = 0.0003

0.02 0 –0.02 –0.04 –0.06 –0.08 –0.1 –0.1

–0.05

0

r = 0.0004 r = 0.0005

40 r = 0.0008

r = 0.002

20

5047.25

5047.3

5047.35

5047.4

5047.45

5047.5

Fig. 4 Zoom on microlensing light curves showing a caustic crossing (see the inset) with different source sizes: D 0:0001, 0.0002, 0.0003, 0.0004, 0.0008, 0.002. The lens system here has a mass ratio q D 0:013 and a separation d D 0:914

“inverse ray shooting,” i.e., to throw rays from the lens plane to the source plane, in order to calculate the images/source areas ratio. Indeed, if computing the sourceto-lens path of light rays is difficult, the lens-to-source path is straightforward. One just needs to apply the equations of light deflection for a given lens model that is tested, and then count the amount of rays falling in the source disk. The ratio of images/source areas gives the total magnification. By Liouville’s theorem, the surface brightness is conserved; thus, the density of light rays in the image plane is uniform. This method requires a dense sampling of the image plane, a grid search on a large .d; q/ parameter space, and an efficient algorithm to fit the other parameters .t0 ; u0 ; tE ; ; ˛/, where ˛ is the angle between the source trajectory and the lens system axis. Furthermore, triple lens modeling is often required and has been implemented in several codes, in particular those using a faster ray shooting technique, the image-centered ray shooting (Bennett and Rhie 1996; Bennett 2010). Another method consists in using Green’s theorem to compute a 1-D mapping of the image contours, instead of 2-D (Dominik 1995, 1998; Gould and Gaucherel 1997). Each point requires to solve a fifth-order polynomial, which makes individual computations time-consuming, but this disadvantage is largely compensated by the faster 1-D integration. However, the task of connecting points in the lens plane in order to obtain closed contours is sometimes confusing, in particular when the source passes near a cusp. Such a code has been released to the public by Valerio Bozza in 2017, whose algorithm is described in Bozza (2010) (http://www.fisica. unisa.it/gravitationAstrophysics/VBBinaryLensing.htm).

668

V. Batista

Microlens Parallax, E The parallax effects count between the most remarkable and useful to solve microlensing problems. They appear as a distortion in the light curve, due to a nonuniform motion of the observer. For fairly long microlensing events (tE 1year=2), the velocity of the Earth can no longer be considered as constant and rectilinear, and this induces asymmetries in the light curve called “orbital parallax” (e.g., Gould and Loeb 1992; Alcock et al. 1995; Mao 1999; Smith et al. 2002). Parallax effects can also be measured when the same event is observed from different observatories. Indeed, “terrestrial” parallax can result from simultaneous ground-based observations at different longitudes, for high-magnification events monitored at very high cadence. And “satellite” parallax can be measured when ground-based observations are combined with space-based observations. Detection of such effects yields the Einstein radius determination, projected in the observer plane, rQE  .2RSch Drel /1=2 , where RSch  2GM c 2 is the lens 1 Schwarzschild radius and Drel  DL1  DS1 . If in addition the Einstein angular 1=2 radius E D .2RSch =Drel / is measured from finite source effects, then the lens mass can be deduced (Gould and Loeb 1992) c2 M D rQE E D 0:1227Mˇ 4G



rQE 1 AU



 E : 1 mas

(9)

Parallax effects are generally defined by the vector  E , called microlens parallax, which magnitude gives the Einstein radius projected in the observer plane, E 

1 AU L  S 1 kpc=Drel D W ; E E =1 mas rQE

(10)

where L =1 mas D 1 kpc=DL and S =1 mas D 1 kpc=DS . The orientation of  E is that of the lens-source relative proper motion. Knowing E and E gives a relation between the lens and source distances, DL D .E E C 1=DS /1 . Measuring the microlensing parallax E and the Einstein radius E represents the optimal way to derive physical parameters of the lens system. As orbital and terrestrial parallaxes are only detectable for rare high-magnification or long microlensing events, the idea of measuring a satellite parallax has recently increased importance within the microlensing community, as was first suggested by Refsdal (1966) and Gould (1999). The first space-based microlensing observations were done with Spitzer for a Small Magellanic Cloud event (OGLE-2005-SMC-0001 Dong et al. 2007). Such observations were also conducted for a microlensing event toward the bulge (MOA-2009-BLG-266 Muraki et al. 2013) using the Deep Impact spacecraft. Spitzer appeared to be an excellent microlensing parallax satellite being on a solar orbit at 1 AU from the Earth, and the microlensing community, under a pilot program led by A. Gould (Gould and Horne 2013; Gould et al. 2014a, 2015), carried out Spitzer observations in 2014, 2015, and 2016. These campaigns resulted

32 Finding Planets via Gravitational Microlensing

669

Fig. 5 OGLE-2014-BLG-0124Lb: a 0:5 Mjup planet orbiting 0:7 Mˇ star, at a?  3:1 AU. The planetary anomaly and the peak of the event have been observed by Spitzer simultaneously to ground-based observations by OGLE (Figure from Udalski et al. 2015)

in a fair amount of highly precise Earth-Spitzer microlens parallaxes, e.g., the case of an isolated star (Yee et al. 2015), a binary star (Han et al. 2016a), and several exoplanets (e.g., Udalski et al. 2015; Street et al. 2016). However, a recurrent obstacle encountered with Spitzer microlensing light curves has been the faintness of the observed targets leading to a low S/N and the fact that only fragments are being observed, with limited baseline data. Figure 5 shows an example of a microlensing event, OGLE-2014-BLG-0124Lb, observed simultaneously with the OGLE telescope and Spitzer. The planetary perturbation is due to a 0:5 Mjup planet orbiting a 0:7 Mˇ star. The two distinct light curves provided a parallax measurement, to a precision 7%. Unfortunately, the error on the mass determination turned out to be much higher (30%) due to poor constraints from finite source effects (Udalski et al. 2015). Space-based microlensing observations have also been conducted in 2016 with the Kepler telescope in its K2-C9 campaign (Howell et al. 2014), with the aim of detecting free-floating planets. In the longer term, this mass measurement method will play an even bigger role in the context of microlensing dedicated space missions like WFIRST (Spergel et al. 2015). However, to fully exploit the potential of this effect, further investigations have been carried out to understand an underlying fourfold discrete degeneracy associated with the Earth-space microlensing parallax. This degeneracy was already

670

V. Batista

discussed in the original paper by Refsdal (1966). Since the same microlensing event is seen from two different observatories, the two observers obtain distinct light curves with different t0 and u0 , and the parallax is given by the difference in these quantities (Gould 1994): E D

AU D?



 t0 ; u0 ; tE

(11)

where t0 D t0;sat  t0;˚ , u0 D u0;sat  u0;˚ , and D? is the physical vector between the two observers projected in the lens plane. The direction of D? gives the one of  E . Since u0 is a signed quantity which magnitude can be derived from the single lens light curve, but not the sign, there are actually four solutions u0;˙;˙ D ˙ju0;sat j ˙ ju0;˚ j. Hence, one degeneracy creates four different orientations of the parallax vector, and a second, two different magnitudes of this vector. Several attempts have been made to break these degeneracies under special circumstances (e.g., Calchi Novati et al. 2015; Yee et al. 2015), but it might actually require a revision of the parallax formalism itself. Calchi Novati and Scarpetta (2016) revisited the analysis of the microlensing parallax from a heliocentric reference frame instead of the commonly adopted geocentric frame and extended the Gould (1994) expression of the parallax to observers in motion, as opposed to observers at rest. They showed that this new approach provides a clearer understanding of the fourfold degeneracy and interesting leverages to break it. In a similar way to parallax effects, if the source is animated by a significant acceleration and nonrectilinear motion during the microlensing event, in other words if the source is a binary, the light curve can be affected by additional distortions. These effects are called “xallarap” by symmetry with the parallax phenomenon. Unless the source motion coincidentally mimics the Earth’s motion, these two effects are generally well distinguishable (Poindexter et al. 2005). Such effects have been detected in the light curve of OGLE-2007-BLG-368Lb, a cold Neptune around a K-star (Sumi et al. 2010).

Orbital Motion In the case of multiple lenses, the orbital motion of the lens components can create deviations in the light curve (Dominik 1998) due to variations in the shape and orientation of the caustics. These effects are generally negligible as microlensing events mostly involve companions with long periods. Although a Keplerian orbit would require five parameters to be characterized, microlensing light curve perturbations can only constrain two additional parameters. For a binary system, one reflects the variation of the separation d between the two bodies, called s here for clarity, and the second being the variation ! of the angle ˛ between the source trajectory and the binary lens axis s D s0 C ds=dt .t  t0 /;

˛ D ˛0 C !.t  t0 /:

(12)

32 Finding Planets via Gravitational Microlensing

671

We consider r? D sDL E the projected separation of the lens system and evaluate the instantaneous projected velocity of the lens components as the quadratic sum of r? ? D r? ! and r? jj D r? .ds=dt /=s, perpendicular and parallel to the projected binary axis, respectively. Unfortunately, there is an important degeneracy in the microlensing models between E;? and ? (Batista et al. 2011; Skowron et al. 2011), where E;? is the component of  E that is perpendicular to the instantaneous direction of the Earth’s acceleration. For bright lenses, radial velocity (RV) measurements can be a good test of microlensing orbital models, as it was done by Yee et al. (2016) for OGLE-2009BLG-020, confirming the model published by Skowron et al. (2011).

Mass Estimates from Galactic Models The lens mass and distance have been directly determined in some microlensing events, thanks to microlens parallax measurements combined with Einstein angular radius estimates (e.g., Bennett et al. 2008, 2016; Gaudi et al. 2008; Batista et al. 2009; Muraki et al. 2013; Kains et al. 2013; Furusawa et al. 2013). However, this is not the case for many microlensing events, and in most cases, in addition to informations provided by the light curve, mass and distance estimates require galactic models that generate distributions of disk and bulge stars according to some empirical constraints (e.g., density of stars in the galaxy, proper motion, mass function, binary stars abundances) (Han and Gould 1995; Duquennoy and Mayor 1991; Raghavan et al. 2002; Duchene and Kraus 2013). In the absence of parallax and/or E measurements, additional observations with high angular resolution imaging are strongly advocated, in order to derive constraints on the lens brightness and thus on its mass with mass-luminosity relations. It has been done for many microlensing systems, reducing drastically the uncertainties on their physical properties, using the VLT, Keck, and Subaru telescopes equipped with adaptive optics (AO) or the Hubble Space Telescope (HST) (Bennett et al. 2006, 2014, 2015, 2016; Dong et al. 2009; Janczak et al. 2010; Sumi et al. 2010; Batista et al. 2011, 2014, 2015; Kubas et al. 2012; Beaulieu et al. 2016; Fukui et al. 2015; Koshimoto et al. 2017).

Detections Overview More than 40 exoplanets have been discovered via microlensing since the first discovery 13 years ago (Bond et al. 2004). Although this number is relatively modest compared to transit and RV methods, the microlensing technique explores a population of planets that is hardly accessible, in the medium term, to other techniques. It is sensitive to planets down to Earth mass beyond the snow line, where the temperature is cool enough for ices to form. According to the core accretion theory (e.g., Ida and Lin 2004), planetary formation is predicted to be the most efficient in this region of the protoplanetary disk. The planet sample provided by

672

V. Batista

microlensing offers a range of various planetary systems and shows the potential of the technique to bring important constraints to the planetary formation theories. It also revealed a population of free-floating planet candidates, down to Jupiter-masses (Sumi et al. 2011).

Cold Super-Earths and Neptunes Although the microlensing method is, like any other technique, more sensitive p to massive planets, with a sensitivity decreasing with mass as  mp , the third and fourth microlensing detections were super-Earth and Neptune-like planets, suggesting that these planets are more common than gas giants. OGLE-2005-BLG390Lb was the first super-Earth detected via microlensing (Beaulieu et al. 2006), with light curve shown in Fig. 6. A Bayesian analysis using a galactic model, combined with a E measurement indicated that the planet is likely to be a 5:5 M˚ super-Earth orbiting a M  0:22 Mˇ star, at a projected separation of 2.6 AU. The equilibrium temperature of the planet is 50 K, making it the first cold super-Earth discovered at that time. This discovery was immediately followed by the detection of a Neptune-/Uranuslike planet, OGLE-2005-BLG-169Lb (Gould et al. 2006), confirming that such low-mass planets should be common. The mass and distance estimates of this system were confirmed and refined down to a 10% precision (mp D 13:2˙1:3 M˚ ) with high angular resolution observations, taken 6 and 8 years after the event using HST (Bennett et al. 2015) and the Keck telescope in Hawaii (Batista et al. 2015), respectively. It is the first case of a planetary microlensing event for which the source and the lens are resolved, being separated by 62 mas on the Keck images. It allowed the detection of the lens brightness and the lens-source relative proper motion. The latter provides an additional mass-distance relation of the lens system.

Fig. 6 OGLE-2005-BLG-390Lb light curve (Beaulieu et al. 2006)

32 Finding Planets via Gravitational Microlensing

673

Several other cold super-Earths and Neptunes have been detected (Muraki et al. 2013; Sumi et al. 2010, 2016; Furusawa et al. 2013), as well as two very low-mass planets of mp  3:2 M˚ (Bennett 2008b; Kubas et al. 2012) and mp  1:66 M˚ (Gould et al. 2014b).

Triple Lens Systems The first triple lens system discovered by microlensing is the two-planet one-star system, OGLE-2006-BLG-109Lb,c, observed in 2006 (Gaudi et al. 2008; Bennett et al. 2010). This system presents a remarkable similarity to a scaled version of Jupiter and Saturn. It is also one of the most complex microlensing light curves observed to date, exhibiting five distinct features from the two planets (Fig. 7). Finite source and parallax effects were detected for this event, as well as orbital motion of the Saturn-like planet. This led to a complete solution to the lens system and provided the physical characteristics of the two planets. The next published multiple planet event was OGLE-2012-BLG-0026Lb,c (Han et al. 2013), composed of a Jupiter-like planet and a sub-Saturn orbiting a solar-mass star. This result was confirmed and refined by Beaulieu et al. (2016) who detected the flux coming from the lens using the Keck telescope. The three other triple lens microlensing systems are one-planet two-star systems. OGLE-2007-BLG-349(AB)c is the first circumbinary planet found by microlensing (Bennett et al. 2016). The main features of the light curve are two cusp approaches that reveal the presence of a companion with a Saturn-to-Sun mass ratio, but only a third body could explain the slope between the two bumps. At that point of the analysis, it is not possible to disentangle the scenario of an additional planet, or an Fig. 7 OGLE-2006-BLG109Lb,c: a Jupiter/Saturn analog (Figure from Gaudi et al. 2008)

674

V. Batista

Fig. 8 OGLE-2007-BLG-349(AB)c: the first circumbinary planet found via microlensing compared to known circumbinary planet systems. The black dots show the orbital separation of the stars, and the crosses represent the planet separation from the stars center of mass (except for the microlensing one that shows a black dot with error bars) (Figure from Bennett et al. 2016)

additional star. HST high-resolution observations ruled out the two-planet scenario because the lens flux could not explain the brightness of the host star in such a configuration. The system is then likely to be a Saturn-like planet orbiting two tight M dwarfs, separated by only 0.08 AU, at 2.6 AU. Figure 8 compares this system to the known circumbinary systems, making it the most compact regarding the binary stars, and the largest regarding the planetary orbit. OGLE-2013-BLG-0341Lb,c is a triple lens system with a very small planetary signal, and a large signature from the crossing of a resonant caustic due to a binary star (Gould et al. 2014b). A low-amplitude bump in a very early phase of the event enabled the authors to identify the binary as being wide, separated by 15 AU. The planet orbits one of them at 1 AU. It is also the smallest microlensing planet discovered to date (1:66 M˚ ). A similar system has been published by Poleski et al. (2014), OGLE-2008-BLG-092, i.e., a wide stellar binary in which the primary star is orbited by a Uranus-like planet (4 MUranus ). All these triple lens events (except the last one) exhibited finite source and parallax effects, allowing to derive precise estimates of their physical properties from the best models.

An Exomoon Around a Free-Floating Planet? Bennett et al. (2014) has shown that it is already possible to detect exomoons of terrestrial mass, orbiting either free-floating planets or planets distant from their

32 Finding Planets via Gravitational Microlensing

675

star. MOA-2011-BLG-262 is a system composed of two bodies with a mass ratio of 3:8  104 and a very short duration. The analysis led to two scenarios, a nearby 0:5 M˚ exomoon orbiting a gaseous rogue planet of 5 MJ or a high-velocity system composed of a Neptune orbiting an M dwarf or a brown dwarf in the galactic bulge. Adaptive optics observations did not allow to discriminate these two models. However, if exomoons are abundant, with more systematic space-based observations in the future, it will be possible to break the mass/distance degeneracy and identify such objects.

Massive Planets Around Very Low-Mass Stars/BDs Determining the frequency of planets orbiting low-mass stars is of interest because these systems provide important tests to planet formation theories. In particular, the core accretion theory predicts that gas giants should be less common around low-mass stars (Laughlin et al. 2004; Ida and Lin 2004). Indeed, in smaller protoplanetary disks, hydrogen and helium might dissipate before the 10 M˚ planet cores grow enough to trigger rapid accretion. This lack of efficiency in the accretion process would be responsible for a larger population of failed gas giants around low-mass stars. Numerous microlensing discoveries seem to confirm these expectations, being cold 10–40 M˚ planets around M dwarfs (e.g., Gould et al. 2006; Muraki et al. 2013; Furusawa et al. 2013; Sumi et al. 2016). However, microlensing discoveries also revealed a population of giant planets >1 MJ around very low-mass stars ( 1:7  104 , n D 0:93 ˙ 0:13. Their distribution thus reaches a peak around the Neptune mass ratio, as shown in Fig. 9, with a slightly lower abundance of Earthlike planets than Neptunes, beyond the snow line. Suzuki et al. (2016) compared their microlensing mass ratio function to previous RV results. For low mass ratios, q  3  105 , it is consistent with Mayor et al. (2009) and Howard et al. (2010), the latter being for period orbits 7 kpc. Disk and bulge populations of stars might have distinct properties (e.g., age, metallicity, density of nearby stars, intensity of radiations) that could affect the planet formation process. Therefore, comparing the planet frequencies and system morphologies of these two populations may be of great interest. The actual microlensing planet sample is not ideal for building objective distance distributions because it has involved heterogeneous hunting and modeling

32 Finding Planets via Gravitational Microlensing

679

strategies over the last decade. Some planetary anomalies were covered by followup groups, others only by survey groups. Some of the events benefitted from a dense observation coverage, parallax, and finite source constraints, while others required Bayesian analyses using galactic models to be characterized. Penny, Henderson et al. (2016) compared the 31 planets discovered before 2016 to a simulated lens distance sensitivity distribution. They used simulations made by Henderson et al. (2014) for the KMNet survey that follow the Han and Gould (2003) galactic model. Their result reveals a discrepancy between the distance estimates from data and their model, expecting a larger amount of systems at further distances (6–7 kpc) and less nearby systems (.2 kpc). They examined different factors that could be responsible for this discordance and found several possible explanations, aside from the quick conclusion that the galactic bulge would be devoid of planets. Indeed, their “model vs data” shows a better agreement when they take into account additional dependences of planet detection efficiency on some microlensing parameters, such as the event timescale. Recently, Zhu et al. (2017) analyzed all 41 microlensing events from the 2015 Spitzer campaign that were simultaneously observed by OGLE and KMNet. They computed a planetary detection efficiency on this free-planet sample and combined these calculations with their determination of the microlensing event rate. It amounts to adding planet sensitivity to the expected distance distribution of observed lenses. From this, they were able to build a distribution of planet sensitivity (i.e., no longer host star sensitivity) as a function of the lens distance. Their total distribution divides into a disk and a bulge distribution, plus the contribution of high-magnification events (Amax > 8), computed separately, as such events generally provide much stronger constraints. They predict that if the planet formation is equally efficient in the disk and in the bulge, 1/3 of the planet population unveiled by an automatic high-cadence survey with parallax measurements should be in the bulge. They also find a lens mass distribution peaking at 0:5 Mˇ , as anticipated. The methodology developed by Zhu et al. (2017) could be used on large samples of planets that will be found by dedicated space missions (Euclid and WFIRST), supported by ground-based observations. They should provide much better constraints on the planet distribution as a function of their host’s location within the galaxy.

Free-Floating Planets From the 2006–2007 MOA-II survey, Sumi et al. (2010) (S11 hereafter) identified an overabundance of short-timescale microlensing events (tE < 2 days) that could not be explained by known stellar and brown dwarf populations. They interpreted these light curves as signatures of either free-floating planets (FFP) or wide-orbit Jupiters (separation & 10 AU from their host star). Bennett et al. (2012) estimated that if all of this population was due to planets bound in wide orbits, their semimajor axis would likely be >30 AU.

680

V. Batista

Clanton and Gaudi (2017) generated a population of distant bound planets with distributions of mass and distance consistent with microlensing, RV, and direct imaging surveys (Clanton and Gaudi 2016) and estimated which fraction of them would not show any signature of the host star in a microlensing light curve. They tried to fit the microlensing event timescale distribution measured by S11 with a lens mass function composed of brown dwarfs, main-sequence stars, and remnants, augmented by the distribution of these bound planets with no star signature. They found that this fraction does not entirely explain the excess of short-timescale events from S11 and concluded that 60% of it should be due to FFP. Figure 10 shows their timescale distribution, consistent with S11 data and composed of three contributors including their predicted population of FFP. This would imply that FFP are 1.3 times more abundant than main-sequence stars in the galaxy. K2-C9 was designed by the microlensing community as a program with the main purpose of detecting FFP with Kepler-Earth parallax measurements (Henderson et al. 2016). Penny et al. (2016) expected a handful of FFP detections with Kepler according to the predictions of S11. With the revised numbers of Clanton and Gaudi (2017), this amount should be reduced by a factor of  0:6, and it might be unlikely to derive reliable statistics from this short survey. WFIRST combined with ground-based surveys will surely provide stronger constraints on the abundance and characteristics of free-floating planets.

Fig. 10 Best fit model from Clanton and Gaudi (2017) to the observed timescale distribution of S11 (black histogram), showing the individual contributors: brown dwarfs, main-sequence stars, remnants (thin pink and blue lines centered on long timescales), bound planets at wide separations (thin pink and blue lines centered on short timescales), and FFP (dashed black lines). The total is shown in thick lines. Blue and pink colors correspond to two formation processes, disk instability (hot start) and core accretion (cold start), respectively (Figure from Clanton and Gaudi 2017)

32 Finding Planets via Gravitational Microlensing

681

Conclusion and Future Prospects Microlensing is currently capable of detecting low-mass planets beyond the snow line, as well as Jupiter-mass free-floating planets, with a network of wide-field telescopes strategically located around the world. This technique is almost uniformly sensitive to planets orbiting all types of stars and remnants, while other methods are most sensitive to FGK dwarfs and are now extending to M dwarfs. It is therefore an independent and complementary detection method for aiding a comprehensive understanding of the planet formation process. Ultimately, comprehensive demographics of planets down to below an Earth mass at separations ranging from the habitable zone to infinity will be achieved from a space-based microlensing survey, such as the WFIRST mission (Spergel et al. 2015) and hopefully the Euclid mission (Penny et al. 2013). The concept was initiated with a dedicated mission (Bennett et al. 2002) called the Microlensing Planet Finder (MPF), which has been proposed to NASA’s Discovery program but not selected. The objective is to be able to monitor turnoff stars in the galactic bulge as microlensing sources. Bennett and Rhie (1996) had shown that the detectability of exoplanets via microlensing depends strongly on the size of the source star, lower-

Fig. 11 Sensitivity range of the WFIRST and Euclid missions in a planet-mass versus semimajor axis diagram, complementary to the parameter space probed by the Kepler mission (Figure courtesy of M. Penny (Ohio State University) & J. D. Myers (JPL))

682

V. Batista

mass planets being detectable only with small source stars. The ideal microlensing planet hunter is a 1–2 m class telescope, with a wide-field near-infrared imager (about 0:5 deg2 ) and high angular resolution (0.1–0.3 arcsec/pixel), to observe the galactic bulge with a sampling rate better than 20 min. Euclid is an ESA medium-size mission, scheduled to launch in 2020, that will measure parameters of dark energy using weak gravitational lensing and baryonic acoustic oscillation and test the general relativity and the cold dark matter paradigm for structure formation. Since its original submission in 2007 to ESA, a microlensing planet hunting program has been listed as part of the Legacy science (Beaulieu et al. 2008). The vision adopted by the Europeans of a joint mission with Dark Energy probes and microlensing has been promoted and adopted by the Astro 2010 Decadal Survey when it created and ranked as top priority the WFIRST mission. A baseline for the design of the Euclid microlensing survey is described by Penny et al. (2013) with a detailed simulation demonstrating the capabilities and the expected scientific outcomes. They conducted the same simulations for the WFIRST survey (Spergel et al. 2015) and predicted that such a mission will detect several thousand bound planets, in addition to several thousand free-floating planets. Moreover, for a planet separation of 2–5 AU, i.e., the range of highest microlensing sensitivity, it will be possible to detect masses lower than Mercury, and even down to the mass of Ganymede (see Fig. 11). Combined with the Kepler survey, it will determine how common Earthlike planets are over a wide range of orbital parameters.

Acronyms MACHO: EROS: OGLE: MOA: PLANET: FUN: MiNDSTEp: KMTNet: WFIRST:

MAssive Compact Halo Objects Expé rience de Recherche d’Objets Sombres Optical Gravitational Lens Experiment Microlensing Observations in Astrophysics Probing Lensing Anomalies NETwork lensing Follow-up Network Mincrolensing Network for the Detection of Small Terrestrial Exoplanets Korean Microlensing Telescope Network Wide-Field InfraRed Survey Telescope

References Albrow MD, Beaulieu J-P, Caldwell JAR et al (1999) A complete set of solutions for caustic crossing binary microlensing events. ApJ 522:1022 Alcock C, Allsman RA, Alves D et al (1995) First observation of parallax in a gravitational microlensing event. ApJ 454:L125+ An JH (1999) The binary gravitational lens and its extreme cases. A&A 349:108

32 Finding Planets via Gravitational Microlensing

683

Batista V, Dong S, Gould A et al (2009) Mass measurement of a single unseen star and planetary detection efficiency for OGLE 2007-BLG-050. A&A 508:467 Batista V, Gould A, Dieters S et al (2011) MOA-2009-BLG-387Lb: a massive planet orbiting an M dwarf. A&A 529:102 Batista V, Beaulieu J-P, Gould A et al (2014) MOA-2011-BLG-293Lb: first microlensing planet possibly in the habitable zone. ApJ 780:54 Batista V, Beaulieu J-P, Bennett DP et al (2015) Confirmation of the OGLE-2005-BLG-169 planet signature and its characteristics with lens-source proper motion detection. ApJ 808:170 Beaulieu J-P, Bennett DP, Fouqué P et al (2006) Discovery of a cool planet of 5.5 Earth masses through gravitational microlensing. Nature 439:437 Beaulieu J-P, Kerins E, Mao S et al (2008) Towards a census of Earth-mass exo-planets with gravitational microlensing. arXiv:0808.0005 Beaulieu J-P, Bennett DP, Batista V et al (2016) Revisiting the microlensing event OGLE 2012BLG-0026: a solar mass star with two cold giant planets. ApJ 824:83 Bennett DP (2008) Detection of extrasolar planets by gravitational microlensing. Exoplanets, Springer praxis books. Praxis Publishing Ltd, Chichester, p 47. ISBN:978-3-540-74007-0 Bennett DP (2010) An efficient method for modeling high-magnification planetary microlensing events. ApJ 716:1408 Bennett DP, Rhie SH (1996) Detecting Earth-mass planets with gravitational microlensing. ApJ 472:660 Bennett DP, Becker AC, Calitz JJ et al (2002) The microlensing event MACHO-99-BLG22/OGLE-1999-BUL-32: an intermediate mass black hole, or a lens in the bulge. astro.ph, 7006 Bennett DP, Anderson J, Gaudi BS et al (2006) Characterization of gravitational microlensing planetary host stars. In: DPS meeting 38, id.45.14; Bull Am Astron Soc 38:1105 Bennett DP, Bond IA, Udalski A et al (2008) A low-mass planet with a possible sub-stellar-mass host in microlensing event MOA-2007-BLG-192. ApJ 684:663 Bennett DP, Rhie SH, Nikolaev S et al (2010) Masses and orbital constraints for the OGLE-2006BLG-109Lb,c Jupiter/Saturn analog planetary system. ApJ 713:837 Bennett DP, Sumi T, Bond I et al (2012) Planetary and other short binary microlensing events from the MOA short-event analysis. ApJ 757:119 Bennett DP, Batista V, Bond I et al (2014) MOA-2011-BLG-262Lb: a sub-Earth-mass moon orbiting a gas giant primary or a high velocity planetary system in the galactic bulge. ApJ 785:155 Bennett DP, Bhattacharya A, Anderson J et al (2015) Confirmation of the planetary microlensing signal and star and planet mass determinations for event OGLE-2005-BLG-169. ApJ 808:169 Bennett DP, Rhie SH, Udalski A et al (2016) The first circumbinary planet found by microlensing: OGLE-2007-BLG-349L(AB)c. AJ 152:125 Bensby T, Yee YC, Feltzing S et al (2013) Chemical evolution of the galactic bulge as traced by microlensed dwarf and subgiant stars. V. Evidence for a wide age distribution and a complex MDF. A&A 549, 147 Bond IA, Abe F, Dodd RJ et al (2001) Real-time difference imaging analysis of MOA galactic bulge observations during 2000. MNRAS 327:868 Bond IA, Udalski A, Jaroszynski M et al (2004) OGLE 2003-BLG-235/MOA 2003-BLG-53: a planetary microlensing event. ApJ 606:155 Bonfils X, Delfosse X, Udry S et al (2013) The HARPS search for southern extra-solar planets. XXXI. The M-dwarf sample. A&A 549:109 Boss A (2006) Rapid formation of gas giant planets around M dwarf stars. ApJ 643:501 Bowler BP, Liu MC, Shkolnik E et al (2015) Planets around low-mass stars (PALMS). IV. The outer architecture of M dwarf planetary systems. ApJS 216:7 Bozza V (2010) Microlensing with an advanced contour integration algorithm: green’s theorem to third order, error control, optimal sampling and limb darkening. MNRAS 408:2188 Calchi Novati S, Scarpetta G (2016) Microlensing parallax for observers in heliocentric motion. ApJ 824:109

684

V. Batista

Calchi Novati S, Gould A, Udalski A et al (2015) Pathway to the galactic distribution of planets: combined Spitzer and ground-based microlens parallax measurements of 21 single-lens events. ApJ 804:20 Cassan A, Kubas D, Beaulieu J-P et al (2012) One or more bound planets per Milky Way star from microlensing observations. Nature 481:167 Chang K, Refsdal S (1979) Flux variations of QSO 0957+561 A, B and image splitting by stars near the light path. Nature 282:561 Choi J-Y, Han C, Udalski A et al (2013) Microlensing discovery of a population of very tight, very low mass binary brown dwarfs. ApJ 768:129 Clanton C, Gaudi BS (2016) Synthesizing exoplanet demographics: a single population of longperiod planetary companions to M dwarfs consistent with microlensing, radial velocity, and direct imaging surveys. ApJ 819:125 Clanton C, Gaudi BS (2017) Constraining the frequency of free-floating planets from a synthesis of microlensing, radial velocity, and direct imaging survey results. ApJ 834:46 Cumming A, Butler RP, Marcy GW et al (2008) The Keck planet search: detectability and the minimum mass and orbital period distribution of extrasolar planets. PASP 120:531 Dominik M (1995) Improved routines for the inversion of the gravitational lens equation for a set of source points. A&A 109:597 Dominik M (1998) A robust and efficient method for calculating the magnification of extended sources caused by gravitational lenses. A&A 333:79 Dominik M (1999) The binary gravitational lens and its extreme cases. A&A 349:108 Dominik M, Sahu KC(2000) Astrometric microlensing of stars. ApJ 534:213 Dong S, Udalski A, Gould A et al (2007) First space-based microlens parallax measurement: Spitzer observations of OGLE-2005-SMC-001. ApJ 664:862 Dong S, Gould A, Udalski A et al (2009) OGLE-2005-BLG-071Lb, the most massive M dwarf planetary companion? ApJ 695:970 Duchene G, Kraus A (2013) Stellar multiplicity. A&A 51:269 Duquennoy A, Mayor M (1991) Multiplicity among solar-type stars in the solar neighbourhood. II – distribution of the orbital elements in an unbiased sample. A&A 248:485 Einstein A (1936) Lens-like action of a star by the deviation of light in the gravitational field. Science 84:506 Fukui A, Gould A, Sumi T et al (2015) OGLE-2012-BLG-0563Lb: a saturn-mass planet around an M dwarf with the mass constrained by Subaru AO imaging. ApJ 809:74 Furusawa K, Udalski A, Sumi T et al (2013) MOA-2010-BLG-328Lb: a sub-Neptune orbiting very late M dwarf? ApJ 779:91 Gaudi BS (2012) Microlensing surveys for exoplanets. ARA&A 50:411 Gaudi BS, Albrow MD, An J et al (2002) Microlensing constraints on the frequency of Jupitermass companions: analysis of 5 years of PLANET photometry. ApJ 566:463 Gaudi BS, Bennett DP, Udalski A et al (2008) Discovery of a Jupiter/Saturn analog with gravitational microlensing. Science 319:927 Gorbikov E, Brosch N, Afonso C (2010) A two-color CCD survey of the North celestial cap: I. The method. Ap&SS 326:203 Gould A (1994) MACHO velocities from satellite-based parallaxes. ApJ 421:75 Gould A (1999) Microlens parallaxes with SIRTF. ApJ 514:869 Gould A (2008) Hexadecapole approximation in planetary microlensing. ApJ 681:1593 Gould A, Gaucherel C (1997) Stokes’s theorem applied to microlensing of finite sources. ApJ 477:580 Gould A, Horne K (2013) Kepler-like multi-plexing for mass production of microlens parallaxes. ApJ 779:28 Gould A, Loeb A (1992) Discovering planetary systems through gravitational microlenses. ApJ 396:104 Gould A, Udalski A, An D et al (2006) Microlens OGLE-2005-BLG-169 implies that cool Neptune-like planets are common. ApJ 644:37

32 Finding Planets via Gravitational Microlensing

685

Gould A, Dong S, Gaudi BS (2010) Frequency of solar-like systems and of ice and gas giants beyond the snow line from high-magnification microlensing events in 2005–2008. ApJ 720:1073 Gould A, Carey S, Yee J (2014a) Galactic distribution of planets from Spitzer microlens parallaxes. sptz.prop11006G Gould A, Udalski A, Shin I-G et al (2014b) A terrestrial planet in a 1-AU orbit around one member of a 15-AU binary. Science 345:46 Gould A, Yee J, Carey S (2015) Galactic distribution of planets from high-magnification microlensing events. sptz.prop12013G Griest K, Safizadeh N (1998) The use of high-magnification microlensing events in discovering extrasolar planets. ApJ 500:37 Han C, Gould A (1995) Statistics of microlensing optical depth. ApJ 449:521 Han C, Gould A (2003) Stellar contribution to the galactic bulge microlensing optical depth. ApJ 592:172 Han C, Jung YK, Udalski A et al (2016) Microlensing discovery of a tight, low-mass-ratio planetary-mass object around an old field brown dwarf. ApJ 778:38 Han C, Udalski A, Gould A et al (2016a) OGLE-2015-BLG-0479LA,B: binary gravitational microlens characterized by simultaneous ground-based and space-based observations. ApJ 828:53 Han C, Udalski A, Gould A et al (2016b) OGLE-2015-BLG-0051/KMT-2015-BLG-0048Lb: a giant planet orbiting a low-mass bulge star discovered by high-cadence microlensing surveys. AJ 152:95 Henderson CB, Park H, Sumi T et al (2014) Candidate gravitational microlensing events for future direct lens imaging. ApJ 794:71 Henderson CB, Poleski R, Penny M et al (2016) Campaign 9 of the K2 mission: observational parameters, scientific drivers, and community involvement for a simultaneous space- and ground-based microlensing survey. PASP 128:14401 Hog E, Novikov ID, Polnarev AG et al (1995) MACHO photometry and astrometry. A&A 294:287 Howard AW, Marcy GW, Johnson JA et al (2010) The occurrence and mass distribution of close-in super-Earths, Neptunes, and Jupiters. Science 330:653 Howell SB, Sobeck C, Haas M et al (2014) The K2 mission: characterization and early results. PASP 126:398 Ida S, Lin DNC (2004) Toward a deterministic model of planetary formation. I. A desert in the mass and semimajor axis distributions of extrasolar planets. ApJ 604:388 Janczak J, Fukui A, Dong S et al (2010) Sub-Saturn planet MOA-2008-BLG-310Lb: likely to be in the galactic bulge. ApJ 711:731 Johnson JA, Aller KM, Howard AW et al (2010) Giant planet occurrence in the stellar massmetallicity plane. PASP 122:905 Jung YK, Udalski A, Sumi T et al (2015) OGLE-2013-BLG-0102LA,B: microlensing binary with components at star/brown dwarf and brown dwarf/planet boundaries. ApJ 798:123 Kaib NA, Raymond SN, Duncan M (2013) Planetary system disruption by galactic perturbations to wide binary stars. Nature 493:381 Kains N,Street RA, Choi J-Y et al (2013) A giant planet beyond the snow line in microlensing event OGLE-2011-BLG-0251. A&A 552:70 Kains N, Bramich DM, Sahu KC et al (2016) Searching for intermediate-mass black holes in globular clusters with gravitational microlensing. MNRAS 460:2025 Kervella P, Thévenin F, Di Folco E (2004) The angular sizes of dwarf stars and subgiants. Surface brightness relations calibrated by interferometry. A&A 426:297 Kim S-L, Lee C-U, Park B-G et al (2016) KMTNET: a network of 1.6 m wide-field optical telescopes installed at three Southern observatories. JKAS 49:37 Koshimoto N, Udalski A, Beaulieu J-P et al (2016) OGLE-2012-BLG-0950Lb: the first planet mass measurement from only microlens parallax and lens flux. AJ 153:1 Kubas D, Beaulieu J-P, Bennett DP et al (2012) A frozen super-Earth orbiting a star at the bottom of the main sequence. A&A 540:78

686

V. Batista

Lafrenière D, Doyon R, Marois C et al (2007) The gemini deep planet survey. ApJ 670:1367 Laughlin G, Bodenheimer P, Adams FC (2004) The core accretion model predicts few Jovian-mass planets orbiting red dwarfs. ApJ 612:73 Lodato G, Delgado-Donate E, Clarke CJ (2005) Constraints on the formation mechanism of the planetary mass companion of 2MASS 1207334-393254. MNRAS 364:91 Mao S (1999) An ongoing parallax microlensing event OGLE-1999-CAR-1 toward Carina. A&A 350:L19 Mao S, Paczynski B (1991) Gravitational microlensing by double stars and planetary systems. ApJ 374:37 Mao S, Smith MC, Wazniak P et al (2002) Optical gravitational lensing experiment OGLE-1999BUL-32: the longest ever microlensing event – evidence for a stellar mass black hole? MNRAS 329:349 Mayor M, Bonfils X, Forveille T et al (2009) The HARPS search for southern extra-solar planets. XVIII. An Earth-mass planet in the GJ 581 planetary system. A&A 507:487 Montet BT, Crepp JR, Johnson JA (2014) The TRENDS High-contrast imaging survey. IV. The occurrence rate of giant planets around M dwarfs. ApJ 781:28 Muraki Y, Han C, Bennett DP et al (2011) Discovery and mass measurements of a cold, 10 Earth mass planet and its host star. ApJ 741:22 Nataf DM, Gould A, Fouqué P et al (2013) Reddening and extinction toward the galactic bulge from OGLE-III: the inner milky way’s Rv 2.5 extinction curve. ApJ 769:88 Nemiroff RJ, Wickramasinghe (1994) Finite source sizes and the information content of machotype lens search light curves. ApJ 424:21 Paczynski B (1986) Gravitational microlensing by the galactic halo. ApJ 304:1 Pejcha O, Heyrovsky D (2009) Extended-source effect and chromaticity in two-point-mass microlensing. ApJ 690:1772 Penny MT, Kerins E, Rattenbury N et al (2013) ExELS: an exoplanet legacy science proposal for the ESA Euclid mission – I. Cold exoplanets. MNRAS 434:2 Penny MT, Rattenbury NJ, Gaudi BS et al (2016) Predictions for the detection and characterization of a population of free-floating planets with K2 campaign 9. arXiv:1605.01059 Poindexter S, Afonso C, Bennett DP et al (2005) Systematic analysis of 22 microlensing parallax candidates. ApJ 633:914 Poleski R, Skowron J, Udalski A et al (2014) Triple microlens OGLE-2008-BLG-092L: binary stellar system with a circumprimary Uranus-type planet. ApJ 795:42 Raghavan D, McAlister HA, Henry TJ et al (2002) A survey of stellar families: multiplicity of solar-type stars. ApJ 190:1 Rattenbury NJ, Bond IA, Skuljan J et al (2002) Planetary microlensing at high magnification. MNRAS 335:159 Refsdal S(1964) The gravitational lens effect. MNRAS 128:295 Refsdal S(1966) On the possibility of determining the distances and masses of stars from the gravitational lens effect. MNRAS 134:315 Rhie SH (1997) Infimum Microlensing Amplification of the maximum number of images of nPoint lens systems. ApJ 484:63 Rhie SH (2002) How cumbersome is a tenth order polynomial? The case of gravitational triple lens equation. Astroph 2294R. arXiv:astro-ph/0202294 Rhie SH, Bennett DP, Becker AC et al (2000) On planetary companions to the MACHO 98-BLG35 microlens star. ApJ 533:378 Sako T, Sekiguchi T, Sasaki M et al (2008) MOA-cam3: a wide-field mosaic CCD camera for a gravitational microlensing survey in New Zealand. ExA 22:51 Schneider DP, Weiss A (1986) The two-point-mass lens – detailed investigation of a special asymmetric gravitational lens. A&A 164:237 Shvartzvald Y, Udalski A, Gould A et al (2015) Spitzer microlens measurement of a massive remnant in a well-separated binary. ApJ 814:111 Shvartzvald Y, Maoz D, Udalski A et al (2016) The frequency of snowline-region planets from four years of OGLE-MOA-Wise second-generation microlensing. MNRAS 457:4089

32 Finding Planets via Gravitational Microlensing

687

Skowron, J, Udalski A, Gould A et al (2011) Binary microlensing event OGLE-2009-BLG-020 gives verifiable mass, distance, and orbit predictions. ApJ 738:87 Smith C, Rest A, Hiriart R et al (2002) Real-time time-variability analysis of GB to TB datasets: experience from SuperMACHO and supernova projects at NOAO/CTIO. In: Tyson JA, Woll S (eds) Presented at the Society of Photo-Optical Instrumentation Engineers (SPIE) conference, vol 4836, pp 395–405 Snodgrass C, Horne K, Tsapras Y (2004) The abundance of galactic planets from OGLE-III 2002 microlensing data. MNRAS 351:967 Spergel D, Gehrels N, Baltay C et al (2015) Wide-field infrarRed survey telescope-astrophysics focused telescope assets WFIRST-AFTA 2015 report. eprint arXiv:1503.03757 Street RA, Udalski A, Calchi Novati S et al (2015) Spitzer parallax of OGLE-2015-BLG-0966: a cold Neptune in the galactic disk. ApJ 819:93 Sumi T, Bennett DP, Bond IA et al (2010) A cold Neptune-mass planet OGLE-2007-BLG-368Lb: cold Neptunes are common. ApJ 710:1641 Sumi T, Kamiya K, Bennett DP et al (2011) Unbound or distant planetary mass population detected by gravitational microlensing. Nature 473:349 Sumi T, Bennett DP, Bond IA et al (2013) The microlensing event rate and optical depth toward the galactic bulge from MOA-II. ApJ 778:150 Sumi T, Udalski A, Bennett DP et al (2016) The first Neptune analog or super-Earth with a Neptune-like orbit: MOA-2013-BLG-605Lb. ApJ 825:112 Suzuki D, Bennett DP, Sumi T et al (2016) The exoplanet mass-ratio function from the MOA-II survey: discovery of a break and likely peak at a Neptune mass. ApJ 833:145 Thompson TA (2013) Gas giants in hot water: inhibiting giant planet formation and planet habitability in dense star clusters through cosmic time. MNRAS 431:63 Udalski A (2003) The optical gravitational lensing experiment. Real time data analysis systems in the OGLE-III survey. Acta Astron 53:291 Udalski A, Yee JC, Gould A et al (2015) Spitzer as a microlens parallax satellite: mass measurement for the OGLE-2014-BLG-0124L planet and its host star. ApJ 799:237 Wambsganss J (1997) Discovering galactic planets by gravitational microlensing: magnification patterns and light curves. MNRAS 284:172 Wang J, Fischer DA (2015) Revealing a universal planet-metallicity correlation for planets of different sizes around solar-type stars. AJ 149:14 Werner MW, Roellig TL, Low FJ et al (2004) The Spitzer space telescope mission. ApJ 154:1 Winn JN, Fabrycky DC (2015) The occurrence and architecture of exoplanetary systems. ARA&A 53:409 Witt HJ (1990) Statistical investigations of the amplification near gravitational lens caustics. In: Mellier Y, Fort B, Saucail G (eds) Gravitational lensing. Lecture notes in physics, vol 360. Springer, Berlin, pp 192–+ Witt HJ, Mao S (1995) On the minimum magnification between caustic crossings for microlensing by binary and multiple stars. ApJ 447:L105+ Wyrzykowski L, Kostrzewa-Rutkowska Z, Skowron J et al (2016) Black hole, neutron star and white dwarf candidates from microlensing with OGLE-III. MNRAS 458:3012 Yee JC, Udalski A, Calchi Novati S et al (2015) First space-based microlens parallax measurement of an isolated star: Spitzer observations of OGLE-2014-BLG-0939. ApJ 802:76 Yee JC, Johnson JA, Skowron J et al (2016) Two stars two ways: confirming a microlensing binary lens solution with a spectroscopic measurement of the orbit. ApJ 821:121 Zhu W, Wang J, Huang C et al (2016) Dependence of small planet frequency on stellar metallicity hidden by their prevalence. ApJ 832:196 Zhu W, Udalski A, Calchi Novati S et al (2017) Toward a galactic distribution of planets. I. Methodology & planet sensitivities of the 2015 high-cadence Spitzer microlens sample. arXiv:1701.05191

Astrometry as an Exoplanet Discovery Method

33

Fabien Malbet and Alessandro Sozzetti

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . How Astrometry Can Detect Exoplanets? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Determination of the Orbital Motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Historical Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ground-Based Astrometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hubble Space Telescope/Fine Guidance Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ground-Based Dual Star Interferometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Space-Borne Global Astrometric Survey: Gaia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Other Space-Based Astrometric Observatory Projects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Limitations and Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Influence of Giant Planets on Earth-Like Planets Detection . . . . . . . . . . . . . . . . . . . . . . . . . Impact of Stellar Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Specificity of Astrometry in the Exoplanet Field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross-References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

690 690 694 696 696 698 698 699 699 700 700 701 701 702 702

Abstract

Astrometry consists in measuring the positions and the motions of the astronomical objects on the sky compared to other stars. The increased accuracy of such measurements opens the way to determine not only the proper motions of stars and their parallactic displacements due to Earth motion around the Sun but also the orbital motion caused by the presence of orbiting planets of all nature.

F. Malbet () University of Grenoble Alpes, CNRS, IPAG, Grenoble, France e-mail: [email protected] A. Sozzetti INAF, Osservatorio Astrofisico di Torino, Pino Torinese, Italy e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_196

689

690

F. Malbet and A. Sozzetti

Several techniques have been investigated using different types of instrument with limited impact on exoplanet detection so far, but the technique has not only great potentials but is complementary to other discovery methods. The importance of stability and precision of the astrometric measurements over a long period may explain the relative lack of results, but the advent of a space mission like Gaia will certainly change the impact of astrometry in the exoplanet field.

Introduction Astrometry is the oldest branch of astronomy, developed by the Greeks, who noticed that although the positions of most stars were stable in the sky, a few of them were moving, and these objects became known as planets (plan¯etes asteres means moving star in ancient greek), suggesting a major difference in their nature. Tycho Brahé in the sixteenth century made precise astrometric measurements of the positions of the planets which allowed Johannes Kepler to establish that these objects were orbiting the Sun on elliptical orbits, opening the Copernicus revolution. Astrometry by determining masses has already contributed significantly to several key science objectives for the understanding of exoplanets and will certainly continue with the publication of position measurements by space missions with unprecedented accuracy like Gaia. We report in this article the contribution of astrometry from the ground and in space to the astrophysics of planetary systems. We provide elements on how to actually detect the presence and characterize exoplanets around stars. We then give an historical perspective on past efforts to detect planets with astrometry and address the most relevant challenges for astrometry to be a discovery method and emphasize the complementarity of astrometry with the other techniques. The reader is also invited to read more detailed developments on the subject by several authors (Sozzetti 2010; Sahlmann 2012; Fischer et al. 2014).

How Astrometry Can Detect Exoplanets? Planets in a planetary system are moving according to the law of gravitation. Their orbits follow the Kepler’s laws of planetary motions. Newton demonstrated that, for a pair of bodies, the sizes of the orbits are inversely proportional to their masses and that these bodies revolve about their common center of mass. In the case of a planetary system, the mass of the star is usually much larger than the masses of the planets, and therefore the center of mass is located close to the center of the star. In the case of the solar system, the center of mass is located near the surface of the Sun. Not only the planets orbit around the center of mass of the system but the star also moves around it in a motion which is the linear combination of the reflex motions for the different planetary orbits at a scale proportional to the mass of the planets.

33 Astrometry as an Exoplanet Discovery Method

691

Since planets are usually too faint to be directly detected, measuring the motion of the stars allows to detect if one or several planets are orbiting around it and the planets characteristics to be measured. Indeed, the motion of a star projected onto the plane of sky is the combination of three types of apparent motion: parallax which is the apparent motion due to the change of perspective of the observer (mainly the location of the Earth during 1 year), the proper motion which is the motion of the star and its planetary system in the galaxy, and the reflex motion due to the presence of planets described in the previous paragraph. A precise orbit determination can therefore unravel the presence of planets of different masses but only if the effect of parallax and proper motion can be subtracted. It is generally assumed that a minimal signal-to-noise ratio of 5–6 on the reflex motion of the star is required to detect a planet. In this case, astrometric measurements of the motion of a star of mass M located at a distance d due to the presence of a planet yield the period P and the planetary mass MP but also the six parameters of the orbit described in the next section, including the semimajor axis aP . The contribution of a planet to the reflex motion of its host star is given by the following formula of the apparent semimajor axis of the stellar orbit:

˛ D 0:33

1     a  M M 1 d P P as 1 AU 1 MEarth 1 MSun 10 pc

(1)

Typical numbers corresponding to various types of planetary systems are given in Table 1 and are displayed in Figs. 1 and 2. There are almost 4 orders of magnitude between a Jupiter in a solar-like planetary system located at 10 pc which gives a signal of the order of 500 as and an Earth-like planets in its habitable zone which gives a signal of 0:3 as. Astrometry, unlike some other methods, is best suited when looking to nearby sources.

Table 1 Expected astrometric signal for different types of planetary systems

Type of planet Stellar spectral type MP (MEarth ) MP (MJupiter ) aP (AU) P (yr) P (d) M (MSun ) d (pc) Astrometric signal (as)

Giants planets Classical Young jupiter jupiter G2 G2 300 300 1 1 5 5 11 11 4084 4084 1 1 10 150 495 33

Hot jupiter G2 300 1 0.1 0.03 12 1 10 10

Telluric planets Hot Earth super-Earth in HZ M G2 5 1 0.02 0.003 0.1 1 0.05 1 17 365 0.45 1 2.5 10 1 0.3

Earth in HZ M 1 0.003 0.28 0.2 82 0.45 10 0.2

0.001 0.001 0.002

0.01

0.1

1

10

100

1000

0.005

0.01

0.02

0.05

0.1

0.2

Me

0.5

Ve

1

2

Ma

5

Earth in G2 HZ Ea

Planet Orbital Period (yr)

Earth in M HZ

Hot Super Earth

Hot Jupiter

10

20

Young Jupiter

Ju Sa

Classical Jupiter

50

100

Ur

200

500

Earth limit @ 10 pc

transit radial velocity

microlensing

imaging

Ne

1000

Fig. 1 Estimated astrometric signature for all 1160 confirmed exoplanets as known in 2018 February 15 in function of their orbital period listed in the Extrasolar Planets Encyclopaedia (exoplanet.eu). The planets of a solar system located at 10 pc are displayed in black filled squares as well as the typical planets listed in Table 1 in red filled circles

Estimated Astrometric Signature (µas)

1e4

692 F. Malbet and A. Sozzetti

0.001 0.002

1e-4

0.001

0.01

0.1

1

10

0.005

0.01

0.02

0.05

0.1

0.2

Me 0.5

Ve

1

2

Ma

5

Earth in G2 HZ

Ea

Planet Orbital Period (yr)

Earth in M HZ

Hot Super Earth

Hot Jupiter

10

Ju

20

Young Jupiter

Sa

50

Ne

200

500

0.33 µas [email protected]

0.33 µas G2@10pc

radial velocity

microlensing transit

imaging

100

Ur

1000

Fig. 2 Mass-period diagram of all exoplanets displayed in Fig. 1, with the astrometry detection limits of 0:33 as and planetary system located at 2.5 pc around a M-type star and located at 10 pc around a solar-type star. The planets of a solar system located at 10 pc are displayed in black filled squares as well as the typical planets listed in Table 1 in red filled circles

Planet mass (MJupiter)

100

33 Astrometry as an Exoplanet Discovery Method 693

694

F. Malbet and A. Sozzetti

Figure 1 shows also that a very small fraction of system have an astrometric signal larger than 1 mas with a period shorter than 20 years with planet masses around or greater than one 1 MJupiter with a dozen of targets. This is the main reason why astrometry has not been a major provider for exoplanet discoveries in the last 20 years, but this should change soon with the outcome of Gaia (see updates in this volume Sozzetti and Bruijne 2018). In multiple systems, astrometric measurements can also determine the mutual inclination angle between pairs of planetary orbits: cos irel D cos iin cos iout C sin iin sin iout cos.˝out  ˝in /;

(2)

where iin and iout and ˝in and ˝out are the inclinations and lines of nodes of the inner and outer planet, respectively. Thus, meaningful estimates can be obtained of the full three-dimensional geometry of any planetary system, without restrictions on the orbital alignment with respect to the line of sight. There are a variety of techniques to measure accurately the astrometric motion of stars, i.e., to measure accurately the positions of stars on the sky plane. These techniques can lead to wide-angle or narrow-angle observations between stars, using relative or absolute measurements, with local or global strategy. The atmosphere turbulence is an important limitation to perform astrometry, and therefore there have been both ground-based and space-born astrometric projects.

Determination of the Orbital Motion The barycentric orbital motion of an isolated star due to the presence of a planet depends on the period P of this invisible companion, the eccentricity e of its orbit, the time of periastron passage T0 , its inclination relative to the sky plane i , the longitude of periastron !, the longitude of the ascending node ˝, and the semimajor axis a expressed in angular units. Figure 3 represents the different angles of the orbit orientation. Instead of directly using the parameters a , !, ˝, and i , the Thiele-Innes constants A; B; F; G linearize the orbit terms (Hilditch 2001) and give the coordinate offsets in the equatorial system including the proper motion, the parallactic motion, and the orbital motion: ˛ ? D ˛0? C˛? .t  t0 / C$ ˘˛? C.A X C F Y / ı D ı0 Cı .t  t0 / C$ ˘ı C.B X C G Y /

(3)

where .˛ ? ; ı/ are the coordinates of the star expressed in the equatorial system (resp. right ascension RA projected onto the celestial sphere ˛ ? D ˛ cos ı and declination DEC) at epoch t , .˛0? ; ı0 / are the star coordinates at epoch t0 , .˛? ; ı / are the proper motions in RA and DEC, $ is the parallax, .˘˛? ; ˘ı / are the parallax factors in RA and DEC computed from the rectangular geocentric equatorial coordinates of

33 Astrometry as an Exoplanet Discovery Method

695

Fig. 3 Representation of the main orbit elements of a star moving around the center of mass located at the origin : the angles i , !, ˝, and , the ascending node n, and the periastron position p

the solar system barycenter at the time t for the observer location, and X; Y are the elliptical rectangular coordinates. The Thiele-Innes constants are defined by:

AD BD F D GD

a . a . a . a .

C cos ˝ cos ! C sin ˝ cos !  cos ˝ sin !  sin ˝ sin !

 sin ˝ sin ! cos i / C cos ˝ sin ! cos i /  sin ˝ cos ! cos i / C cos ˝ cos ! cos i /

(4)

and the elliptical rectangular coordinates X and Y are functions of the eccentric anomaly E and the eccentricity e:

X D cos E  e

Y D

p

1  e 2 sin E:

(5)

The coordinate offsets computed from Eq. 3 can be projected along a given direction, either defined by the scan circle of Hipparcos or Gaia or by the baseline of an interferometer, in order to derive the measured signal. At precisions of the order of micro-arcseconds, several secondary effects must be taken into account: refraction of the light in the Earth atmosphere which is wavelength dependent, relativistic effects due to the motion of the observer relative to the target, or gravitational light deflection by massive objects in the line of sight.

696

F. Malbet and A. Sozzetti

Historical Perspective The contribution of astrometry to the study of exoplanets has been relatively limited, but with the advent of space missions with increased accuracy, this contribution might change in nature.

Ground-Based Astrometry The presence of very low-mass companions has been suggested around 61 Cygni and 70 Ophiuchi by Strand (1943) and Reuyl and Holmberg (1943) to interpret the perturbations of astrometric measurements based on long-term time series of photographic plates. For many years starting in the 1960s, a narrow-field astrometry was frequently tried for finding exoplanets. These attempts resulted in various discoveries, but none confirmed (Barnard’s star, Lalande 21185). However, as the evidence for planetary mass companions detected by ground-based astrometry around 61 Cyg and 70 Oph has been proved incorrect (Heintz 1978), the failures caused by insufficient precision and systematic errors of the photographic technique have brought some doubts on the use of the astrometry for exoplanet studies. The narrow-field astrometry was considered to be limited by a 1 mas precision due to the atmospheric image motion which came from the expression "  =D 2=3 derived by Lindegren (1980) for the root mean square of the image motion " in the measurement of a distance  between a pair of stars on a telescope of diameter D. Recently, Lazorenko (2002) has found that for the reference field represented by a grid of stars, the power law is "   11=6 =D 3=2 with an improvement with D. Above expression is however asymptotic and refers to the very narrow-field mode of observations requesting the use of very large telescopes. In practice, this opens a way to a 28;000-K pulsating sdB. With a long baseline of nearly continuous observations, many other binaries have been discovered around pulsating stars observed during the original Kepler mission. Hundreds have been discovered around the numerous ı Scuti stars (Murphy et al. 2013; Balona 2014; Compton et al. 2016), including many of which that have been verified from radial-velocity follow-up (Murphy et al. 2016b). In addition, multiple hot subdwarfs have revealed stellar companions from timing their pulsations (Telting et al. 2012, 2014).

Time delay (s)

Only a handful of exoplanets have been claimed using the pulsation timing method. Most secure is the object likely near the deuterium-burning limit that orbits near the habitable zone of an A star in the original Kepler mission: an m sin i D C22 C0:13 11:8C0:8 0:6 MJup object in an 84020 day, slightly eccentric (e D 0:150:10 ) orbit around the ı Scuti star KIC 7917485 (Murphy et al. 2016a). Figure 2 shows time delays of the two highest-amplitude pulsation modes, which have periods of 1.56 and 1.18 h, along with their weighted averages. Both pulsations show an identical phase modulation with an amplitude of  D 7:1 ˙ 0:5 s that reveals a companion close to the brown-dwarf-planet boundary. Interestingly, the planet has an insolation flux between roughly 1.2–1.4 times that of the Earth, such that any moons around this object could potentially be habitable (Murphy et al. 2016a). The least-massive object claimed using the pulsation timing method was proposed by Silvotti et al. (2007) around the hot subdwarf V391 Pegasi. This sdB shows evidence of a  D 5:3 ˙ 0:6 s phase modulation every 1170 ˙ 44 days in two pulsations (with periods of 349.5 and 354.1 s), which would correspond to the effect of an m sin i D 3:2 ˙ 0:7 MJup planet. Additional substellar companions have been proposed to orbit two other hot subdwarfs with long-term monitoring (Lutz et al. 2012).

25 20 15 10 5 0 -5 -10 -15 -20 -25 54800

f2 f1 av

55000

55200

55400

55600

55800

56000

56200

56400

56600

Time (d)

Fig. 2 Time delays in the two largest-amplitude pulsations of the ı Scuti star KIC 7917485 are C22 consistent with an m sin i D 11:8C0:8 0:6 MJup object in an 84020 day, slightly eccentric orbit (Reproduced from Murphy et al. 2016a by permission of the AAS)

794

J. J. Hermes

A major caveat in the use of pulsation timing to detect exoplanets is the inherent stability of the pulsation modes themselves. Here the white dwarfs, some of the most sensitive objects for detecting substellar companions using pulsations, serve as a useful lesson. In the early 2000s, researchers at McDonald Observatory undertook a program to monitor more than a dozen of the hottest pulsating white dwarfs with hydrogen atmospheres, since these stars have the longest mode lifetimes and thus the most stable pulsations (Winget et al. 2003). Early on, one white dwarf in this sample, GD 66, showed phase modulation that could be caused by a 2 MJup planet in a 4.5-year orbit (Mullally et al. 2008). However, this modulation came from just one pulsation mode. Subsequent observations have shown that different modes in the same star show different phase modulation; this excludes an external cause for the changes, complicating the planetary hypothesis for GD 66 (Hermes 2013; Dalessio 2013). Critically, multiple white dwarfs with otherwise “stable” oscillations have shown unexpected changes in the pulsation arrival times: one shows phase evolution hundreds of times faster than expected from stellar evolution (Hermes et al. 2013), another shows divergent phase modulation in multiple different pulsation modes that cannot be caused by an external companion (Dalessio et al. 2013), and yet another shows phase modulation within the same rotational multiplet that is anticorrelated, which again cannot be caused by an external companion (Zong et al. 2016). The authors of the latter result suggest nonlinear resonant mode coupling could explain the phase changes. Regardless of the physical cause, these three independent empirical results warn that internal effects must be ruled out as the source of the phase modulation when using the pulsation timing method to discover exoplanets. This is best accomplished by observing the phase evolution of as many modes as possible, to guarantee they all respond to a hypothetical external companion in the same way.

Conclusions Exoplanets may be revealed around pulsating stars by carefully monitoring the arrival times of the pulsations, searching for periodic modulation from the lighttravel-time delays caused when the host star is tugged by an unseen companion. The method requires the pulsations be extremely stable and is most sensitive to substellar companions around A stars and compact remnants. Additionally, phase modulation should be observed in more than one mode within the star to exclude effects caused by changes to the stellar interior – external effects from an exoplanet or binary companion identically affect the arrival times of all pulsation modes. Hundreds of binary companions have been revealed using the pulsation timing method, which impart larger signals. But exoplanets are not beyond reach – several substellar companions have been discovered by monitoring the arrival times of stellar oscillations.

36 Timing by Stellar Pulsations as an Exoplanet Discovery Method

795

Cross-References  Circumbinary Planets Around Evolved Stars  Pulsar Timing as an Exoplanet Discovery Method  Transit-Timing and Duration Variations for the Discovery and Characterization of

Exoplanets Acknowledgements The author thanks Simon Murphy and Bart Dunlap for helpful discussions. Support for this work was provided by NASA through Hubble Fellowship grant #HST-HF251357.001-A.

References Balona LA (2014) Binary star detection using the time-delay method: application to 34 Kepler objects of interest. MNRAS 443:1946–1954 Barlow BN, Dunlap BH, Clemens JC (2011a) Radial velocity confirmation of a binary detected from pulse timings. ApJ 737:L2 Barlow BN, Dunlap BH, Clemens JC et al (2011b) Fortnightly fluctuations in the O-C diagram of CS 1246. MNRAS 414:3434–3443 Compton DL, Bedding TR, Murphy SJ, Stello D (2016) Binary star detectability in Kepler data from phase modulation of different types of oscillations. MNRAS 461: 1943–1949 Dalessio JR (2013) Peculiar variations of white dwarf pulsation frequencies and maestro. Ph.D. thesis, University of Delaware Dalessio J, Sullivan DJ, Provencal JL et al (2013) Periodic variations in the O – C diagrams of five pulsation frequencies of the DB white dwarf EC 20058-5234. ApJ 765:5 Eastman J, Siverd R, Gaudi BS (2010) Achieving better than 1 minute accuracy in the heliocentric and barycentric Julian dates. PASP 122:935–946 Hermes JJ (2013) Complications to the Planetary Hypothesis for GD 66. American Astronomical Society, AAS Meeting #221, id.424.04 Hermes JJ, Montgomery MH, Mullally F, Winget DE, Bischoff-Kim A (2013) A new timescale for period change in the pulsating DA white dwarf WD 0111C0018. ApJ 766:42 Kepler SO, Winget DE, Nather RE et al (1991) A detection of the evolutionary time scale of the DA white dwarf G117 – B15A with the whole Earth telescope. ApJ 378:L45–L48 Lutz R, Schuh S, Silvotti R (2012) EXOTIME: searching for planets and measuring dotP in sdB pulsators. Astronomische Nachrichten 333:1099 Mullally F, Winget DE, Degennaro S et al (2008) Limits on planets around pulsating white dwarf stars. ApJ 676:573–583 Murphy SJ, Shibahashi H (2015) Deriving the orbital properties of pulsators in binary systems through their light arrival time delays. MNRAS 450:4475–4485 Murphy SJ, Pigulski A, Kurtz DW et al (2013) Asteroseismology of KIC 11754974: a highamplitude SX Phe pulsator in a 343-d binary system. MNRAS 432:2284–2297 Murphy SJ, Bedding TR, Shibahashi H, Kurtz DW, Kjeldsen H (2014) Finding binaries among Kepler pulsating stars from phase modulation of their pulsations. MNRAS 441: 2515–2527 Murphy SJ, Bedding TR, Shibahashi H (2016a) A Planet in an 840 day orbit around a Kepler main-sequence a star found from phase modulation of its pulsations. ApJ 827:L17 Murphy SJ, Shibahashi H, Bedding TR (2016b) Finding binaries from phase modulation of pulsating stars with Kepler IV. Detection limits and radial velocity verification. MNRAS 461:4215–4226

796

J. J. Hermes

Shibahashi H, Kurtz DW (2012) FM stars: a Fourier view of pulsating binary stars, a new technique for measuring radial velocities photometrically. MNRAS 422:738–752 Silvotti R, Schuh S, Janulis R et al (2007) A giant planet orbiting the ‘extreme horizontal branch’ star V391 Pegasi. Nature 449:189–191 Telting JH, Østensen RH, Baran AS et al (2012) Three ways to solve the orbit of KIC 11558725: a 10-day beaming sdBCWD binary with a pulsating subdwarf. A&A 544:A1 Telting JH, Baran AS, Nemeth P et al (2014) KIC 7668647: a 14 day beaming sdBCWD binary with a pulsating subdwarf. A&A 570:A129 Winget DE, Cochran WD, Endl M et al (2003) The search for planets around pulsating white dwarf stars. In: Deming D, Seager S (eds) Scientific frontiers in research on extrasolar planets. ASP conference series, vol 294. ASP, San Francisco, pp 59–64 Zong W, Charpinet S, Vauclair G, Giammichele N, Van Grootel V (2016) Amplitude and frequency variations of oscillation modes in the pulsating DB white dwarf star KIC 08626021. The likely signature of nonlinear resonant mode coupling. A&A 585:A22

Transit-Timing and Duration Variations for the Discovery and Characterization of Exoplanets

37

Eric Agol and Daniel C. Fabrycky

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Theory and Paradigmatic Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chopping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Transit Duration Variations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Observational Considerations: Timing Precision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Science Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

798 801 803 806 807 808 809 812

Abstract

Transiting exoplanets in multi-planet systems have non-Keplerian orbits which can cause the times and durations of transits to vary. The theory and observations of transit-timing variations (TTVs) and transit duration variations (TDVs) are reviewed. Since the last review, the Kepler spacecraft has detected several hundred perturbed planets. In a few cases, these data have been used to discover additional planets, similar to the historical discovery of Neptune in our own solar system. However, the more impactful aspect of TTV and TDV studies has been characterization of planetary systems in which multiple planets transit. After addressing the equations of motion and parameter scalings, the main dynamical mechanisms for TTV and TDV are described, with citations to the observational literature for real examples. We describe parameter constraints, particularly the

E. Agol () Department of Astronomy, University of Washington, Seattle, WA, USA e-mail: [email protected]; [email protected] D. C. Fabrycky Department of Astronomy and Astrophysics, University of Chicago, Chicago, IL, USA e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_7

797

798

E. Agol and D. C. Fabrycky

origin of the mass/eccentricity degeneracy and how it is overcome by the highfrequency component of the signal. On the observational side, derivation of timing precision and introduction to the timing diagram are given. Science results are reviewed, with an emphasis on mass measurements of transiting subNeptunes and super-Earths, from which bulk compositions may be inferred.

Introduction Transit-timing variations (TTVs) and transit duration variations (TDV) are two of the newest tools in the exoplanetary observer’s toolbox for discovering and characterizing planetary systems. Like most such tools, they rely on indirect inferences, rather than detecting light from the planet directly. However, the amount of dynamical information they encode is extremely rich. To decode this information, let us start with the dynamical concepts. Consider the vector stretching from the star of mass m0 to the planet of mass m to be r D .x; y; z/, with a distance r and direction rO . The Keplerian potential per reduced mass, D GM =r (where M  m0 C m and the planet is replaced with a body of reduced mass   m0 m=M ), gives rise to closed orbits. This means that, in the absence of perturbations, the trajectory is strictly periodic, r.t C P / D r.t /. Moreover, Kepler showed that Tycho Brahe’s excellent data for planetary positions were consistent with Copernicus’ idea of a heliocentric system only if the planets (including the Earth) followed elliptical paths of semimajor axis a and one focus on the sun. Newton was successful at finding the principle underlying such orbits, a force law F D Rr D Gm0 r 2 rO , which results in a period P D 2a3=2 .GM /1=2 (i.e., with the a-scaling, Kepler found the planets actually obeyed). This research program was thrown into some doubt by the “great inequality,” the fact that the orbits of Jupiter and Saturn did not fit the fixed Keplerian ellipse model. This obstacle was overcome by the perturbation theory of Laplace, who used the masses derived via their satellite orbits to explain the deviations of their heliocentric orbits (Wilson 1985). The insight can be calculated by writing an additional force to that of gravity of the sun: F1 D G1 M r12 rO 1 C F12 ;

(1)

where now the forces and distances specifically pertains to planet 1 and a force of planet 2 on planet 1 is added. This latter force consists of two terms: F12 D 1 rR 1 D G1 m2 jr2  r1 j3 .r2  r1 /  G1 m2 r22 rO 2 :

(2)

The first term on the right-hand side is the direct gravitational acceleration of planet 1 due to planet 2. The second is an indirect frame-acceleration effect, due to the acceleration the star feels due to the second planet. Since the sun is fixed at the zero of the frame, this acceleration is modeled by acceleration of planet 1 in the opposite direction.

37 Transit-Timing and Duration Variations for the Discovery and: : :

799

Likewise, Le Verrier and Adams used planet-planet perturbations in the first discovery of a planet by gravitational means (Adams 1847; Le Verrier 1877). In this case, they did not know the zeroth-order solution (i.e., the Keplerian ellipse) for the perturber, Neptune. In its place, they assumed the Titius-Bode rule held and sought only the phase of the orbit. This technique worked because they only wanted to see how the acceleration, then deceleration, of Uranus as it passed Neptune would betray Neptune’s position on the sky to optical observers. The task of discovering planets by TTV is more demanding. We do not have any hints as to what the planet’s orbit might be, i.e., we cannot assume it is on a circular orbit or obeys some spacing law. The observation of a single orbit is insufficient for a detection: times of least three transits are needed to measure a period change. However, due to measurement error, in only a small fraction of cases is the high-frequency “chopping” signal (see Chopping section below) statistically significant after just three transits. Moreover, the sampling of the orbit only at transit phase causes aliasing of the dynamical signals. The times of transit are primarily constrained by the decline of stellar flux during transit ingress, and the rise over egress, which occur on a timescale    1 P .Rp =a/  2:2 min



Rp R˚



M? Msun

1=3 

P 10d

1=3 ;

(3)

assuming a circular orbit, edge on to the line of sight (impact parameter of b D 0), around a star of mass M? ; usually timing precision can be measured to better than this timescale. This timing precision gives a sensitive measure of the variation of the angular position of a planet relative to a Keplerian orbit. In contrast, the other dynamical techniques rely on a signal spread through the orbital timescale P , and thus the precision of the orbital phase is poorly constrained unless the measurements are of high precision or long duration (although these conditions have been achieved by pulsar timing in PSR 1257 +12 which detected a great inequality (Wolszczan 1994) and by radial velocity in GJ 876 which detected resonant orbital precession (Laughlin and Chambers 2001)). Orbital positions or transit times are expressed in a table called an ephemeris. Perturbations cause motions or timing deviations from a Keplerian reference model, especially changes to its instantaneous semimajor axis a, eccentricity e, and longitude of periastron !. The latter angle is between the position of closest approach and a plane perpendicular to the line of sight that contains either the primary body or the center of mass. In the case of transit-timing variations, the Keplerian alternative is simply an ephemeris with a constant transit period, P : C D T0 C P  E;

(4)

where E is the epoch – an integer transit number – and T0 is the time of the transit numbered E D 0; C stands for “calculated” based on a constant-period model. Meanwhile, the observed times of transit are denoted O. This notation leads to

800

E. Agol and D. C. Fabrycky

Observed - Calculated (days) Observed mid-time (days)

80 60 40 20 0 2

0

-2

0

2

4

6 8 10 transit number, E

12

14

Fig. 1 An example of timing data. Top panel: the measured midtimes of exoplanet transits, to which a line is fit by least squares. Bottom panel: the residuals of that fit, which is the conventional observed minus calculated (O  C ) diagram; the original sinusoidal function, to which Gaussian noise was added, is also plotted as a line

an O  C (pronounced “O minus C”; Sterken 2005) diagram, in which only the perturbation part is plotted. An instructive version, modeled after the timing of WASP-47 (Becker et al. 2015) but with a greatly exaggerated perturbation, is shown in Fig. 1. The transit times come earlier than the linear model for transit numbers 0–3 and 11–14 and later than the linear model for transit numbers 4–10. These deviations from a constant transit period are what we call TTVs. The other dynamical effect addressed by this review is TDVs. Like TTVs, the cause can be changes in a, e, or !. The most dramatic effect, however, is due to orbital plane reorientation. The angle the orbital plane’s normal vector makes to the observer’s line of sight – the inclination, i – determines the length of the transit chord. Changes in the inclination will change the length of that chord, which in turn changes the amount of time the planet remains in transit: duration variations. The literature on exoplanets has a history of rediscovering effects that had been well studied in the field of binary and multiple stars. In the current focus, it has long been known to eclipsing binary observers that long-term depth changes can result from the torque of a third star orbiting the pair (Mayer 1971). This effect owes to the secular and tidal dynamics which dominate triple star systems (Borkovits et al. 2003), dictated by their hierarchical configuration which allows them to remain

37 Transit-Timing and Duration Variations for the Discovery and: : :

801

stable. TDV due to perturbing planets is simply its exoplanetary analogue (MiraldaEscudé 2002). The first recognition of the importance of transit-timing and duration variations was at the DPS and AAS meetings two decades ago by Dobrovolskis and Borucki (1996a,b), followed a few years later by Miralda-Escudé (2002) and Schneider (2003, 2004). More detailed studies that included the important effect of meanmotion resonance, in which the ratio of two planets’ orbital periods is close to the ratio of small integers, were independently investigated by Holman and Murray (2005) and Agol et al. (2005). The former paper showed that solar system-like perturbations might be used to find Earthlike planets, should transit times be measured with sufficient accuracy. The latter paper coined the term “transit-timing variations,” with acronym TTVs, and defined TTVs as the observable accumulation of transit period changes (i.e., O  C ). Initial studies of TTVs of hot Jupiters were able to place limits on the presence of Earth-mass planets near mean-motion resonance (Steffen and Agol 2005). Some further studies claimed detection of perturbing planets causing TTVs or TDVs, but each of these was quickly disputed or refuted by additional measurements. The first convincing detection awaited the launch of the Kepler spacecraft and the discovery of Kepler-9 which showed large-amplitude TTVs of two Saturn-sized planets with strong significance (Holman et al. 2010); this discovery was remarkably similar to predictions based upon the GJ 876 system (Agol et al. 2005). The Kepler-9 paper kicked off a series of discoveries of TTVs with the Kepler spacecraft, with now more than 100 systems displaying TTVs and a handful showing TDVs (Holczer et al. 2016).

Preliminaries Since the gravitational interactions between planets occur on the orbital timescale, the amplitude of TTVs is proportional to the orbital period of each planet, times a function of other dimensionless quantities. Thanks to Newton’s second law and Newton’s law of gravity, the acceleration of a body does not depend on its own mass. Thus, the TTVs of each planet scale with the masses of the other bodies in the system. In a two-planet system, then, to lowest order in mass ratio, the O  C formulae are: m2 f12 .˛12 ;  12 /; m0 m1 ıt2 D P2 f21 .˛12 ;  21 /; m0

ıt1 D P1

(5)

where the masses of the star and planets are m0 ; m1 ; and m2 and fij describes the perturbations of planet j on planet i , which is a function of the semimajor axis ratio, ˛ij D min.ai =aj ; aj =ai /, and the angular orbital elements of the planets,  ij D .i ; ei ; !i ; Ii ; ˝i ; j ; ej ; !j ; Ij ; ˝j /. The evaluation of these functions can

802

E. Agol and D. C. Fabrycky

be found in a series of papers on perturbation theory: Nesvorný and Morbidelli (2008); Nesvorný (2009); Nesvorný and Beaugé (2010); Agol and Deck (2016); Deck and Agol (2016). With the addition of multiple perturbing planets, if the mass ratios of the planets to the star are sufficiently small and if none of the pairs of planets are in a mean-motion resonance, then the TTVs may be approximately expressed as linear combinations of the perturbations due to each companion. For N planets, the TTVs become X mj ıti D Pi fij .˛ij ;  ij /; (6) m0 j ¤i

for i D 1; : : : ; N . The largest TTVs are caused by orbital period changes associated with librations of the system about a mean-motion resonance. Energy trades can be used to compute the amplitude of the TTV in each planet (see Agol et al. 2005; Holman et al. 2010). Because of Kepler’s relation a / P 3=2 , a period lengthening of ıP1 P1 is associated with a semimajor axis change of ıa1 D .3=2/a1 ıP1 =P1 . Differentiating the orbital energy equation E1 D GM m1 =.2a1 / shows that such a change results in an energy change of ıE1 D .GM m1 a12 =2/ıa1 . To conserve total energy, the other planet will have an energy change of ıE2 D .GM m1 a12 =2/ıa1 , which can also be expressed as C.GM m2 a22 =2/ıa2 . Using the relation ıa2 D .3=2/a2 ıP2 =P2 , and the Keplerian relation a2 =a1 D .P2 =P1 /2=3 , we obtain: ıP2 D ıP1 .m1 =m2 /.P2 =P1 /5=3 :

(7)

When considering the O C shapes that each planet makes over a fixed time interval (e.g., from a survey that measures transits for both planets), we will have a factor of P2 =P1 more orbital periods for the inner planet than the outer planet. Thus, the accumulated time shift of the signal, ıt , builds up more for the inner planet, by one factor of the period ratio. In consideration of Eq. 7, we are left with: ıt2 D ıt1 .m1 =m2 /.P2 =P1 /2=3 :

(8)

This scaling agrees with analytic work performed in the resonant (Nesvorný and Vokrouhlický 2016) and near-resonant (Lithwick et al. 2012; Hadden and Lithwick 2016) regimes. Hence, the TTV curves of the two planets are anticorrelated, with the ratio of planetary masses determining the ratio of TTV amplitudes. In the case that the masses are equal, the amplitude of the outer planet’s TTV is larger because its orbital size needs to change more for its Keplerian orbital energy to equal the change in the inner planet’s Keplerian orbital energy. In general, transit-timing variations afford a means of measuring the density of exoplanets. The two observables associated with a light curve are the time stamp of each photometric measurement and the number of photons measured. The number of photons is a dimensionless number and thus may only constrain dimensionless

37 Transit-Timing and Duration Variations for the Discovery and: : :

803

quantities, such as radius ratio, impact parameter, or the ratio of the stellar size to the semimajor axis. The quantities that have units of time – the period, transit duration, and ingress duration – can constrain the density of the system since the dynamical time relates to stellar density, , as tdyn  .G /1=2 . Seager and Mallén-Ornelas (2003) showed that a single transiting planet on a well-measured circular orbit may be used to gauge the density of the star; in the case of multiple transiting planets, the circular assumption may be relaxed (Kipping 2014). The transit depth, then, gives the radius ratio of the planet to the star, while if two planets transit and show TTVs, their TTVs give an estimate of the mass ratio of the perturbing planet to the star. Thus, two transiting, interacting planets yield an estimate of the density ratio of the planets to the star, and consequently we can obtain the density of the planets. Note that this is true even if the absolute mass and radius of the star are poorly constrained. A caveat to this technique is that there is an eccentricity dependence that is present in the stellar density estimate. However, multi-transiting planet systems typically require low eccentricities to be stable, and in some cases, the eccentricities can be constrained sufficiently from TTV analysis, from analyzing multiple planets (Kipping 2014), or from statistical analysis of an ensemble of planets (Hadden and Lithwick 2017). So this caveat ends up not impacting the stellar density estimate significantly (the mass-eccentricity degeneracy, however, reduces precision on planet-star mass ratios and hence inflates the planet density uncertainty). Another way to obtain an estimate of stellar density is from asteroseismology: in fact, the time dependence of asteroseismic measurements is what enables density to be constrained in that case as well (Ulrich 1986). If a pair of transiting exoplanets can be detected with both TTVs and RVs, then the absolute dimensions of the system may be obtained (Agol et al. 2005; Montet and Johnson 2013) as RVs have dimensions of velocity, which when combined with time measurements from TTVs give dimensions of distance. In practice, this technique has yet to yield useful constraints upon the properties of planetary systems (Almenara et al. 2015), but it may prove fruitful in the future much as double-lined spectroscopic binaries have used to measuring the properties of binary stars, as hinted at by Almenara et al. (2016). Circumbinary planets (CBP) are an extreme example of this technique: the timing offsets of the transits, combined with the eclipses and radial velocity of the binary, give very precise constraints on the absolute parameters of the Kepler-16 system (Doyle et al. 2011).

Theory and Paradigmatic Examples Here, we discuss the physical models for different types of TTV interactions and point the reader to real systems that exhibit each kind of interaction. Close to resonances, a combination of changes in semimajor axis and eccentricity leads to TTV cycles whose period depends on the separation from the resonance (Steffen 2006; Lithwick et al. 2012); the latter refers to this as the “super-period.” The main TTV variation comes from only one resonance, the one the system is

804

E. Agol and D. C. Fabrycky

closest to, which allows its critical angles to move slowly and thus its effect to build up. If the period ratio P2 =P1 is within a few percent of the ratio j =k, with j and k being integers, then the expected TTV period is PTTV D 1=jj =P2  k=P1 j:

(9)

The order of the resonance is jj  kj, and the strength of the resonance depends on the planetary eccentricities to a power of the order minus 1. Therefore, first-order resonances affect planets with no initial eccentricity, but higher-order resonances have a large effect only in the presence of some eccentricity. Seeing two planets transit the star helps immensely to characterize a nearresonant system, because then the relative transit phase of the two planets can be compared with the phase of the TTV signals (Lithwick et al. 2012). If the eccentricities are maximally damped out, then the resonant terms of the interaction continue forcing a small eccentricity that quickly precesses, causing the TTV. In that case, the phase of the signal is predictable, and the two planets’ eccentricities are anti-aligned, so the TTV signals consist of anticorrelated sinusoids. Also useful in that case is that the amplitudes lead directly to the planetary masses. If socalled free eccentricity remains, however, the phases would usually differ from that prediction, the TTV in the two planets may not be in perfect anti-phase, and only an approximate mass scale rather than a measurement is available, which is referred to as the mass-eccentricity degeneracy. The first real system that showed this pattern convincingly was Kepler-18 (Cochran et al. 2011). The degeneracy between mass and eccentricity results from sampling at the period of the transiting planet, which causes short-period variations to be aliased with PT T V (Lithwick et al. 2012; Deck and Agol 2015). The measurement of TTVs and TDVs has been used for confirmation, detection, and characterization of transiting exoplanets and their companions. The Kepler spacecraft discovered thousands of transiting exoplanet candidates; the classification as “candidate” was cautiously used to allow for other possible explanations, such as a blend of a foreground star and a background eclipsing binary causing an apparent transit-like signal. The presence of multiple transiting planets around the same star gave a means of confirming two planets that display anticorrelated TTVs: due to energy conservation (Eq. 8), the anticorrelation indicates dynamical interactions between the two planets, while such a configuration would not be stable for a triple star system. Many papers used this technique to confirm that Kepler planet candidates were bona fide exoplanets using different techniques to identify the anticorrelation in data (Ford et al. 2012a,b; Fabrycky et al. 2012; Steffen et al. 2012; Xie 2013). The characterization of exoplanets with TTVs also began in earnest with the Kepler spacecraft. In addition to Kepler-9, the Kepler-18 system was characterized by a combination of TTVs and RVs, giving density estimates for the three transiting planets (Cochran et al. 2011) and assuring that the new method for mass characterization gave the same answers as the trusted, older method.

37 Transit-Timing and Duration Variations for the Discovery and: : :

805

When only one planet transits in a near-resonant system, the measured TTVs may simply record a sinusoidal signal, which could result from the other planet being close to many different resonances with the transiting planet (Meschiari and Laughlin 2010). In Kepler-19, Ballard et al. (2011) were able to tell that a planetary companion was the only sensible cause of the TTV, but they were not able to break this finite set of degeneracies. This degeneracy has made it extremely difficult to characterize non-transiting planets via TTV, and hence in many cases, an additional planet is suspected due to TTV, but detailed work has not been pursued to determine its nature. The first case of a non-transiting planet being discovered and completely characterized was Kepler-46 (aka KOI-872; Nesvorný et al. 2012). The authors found that the TTVs of the transiting planet were far from a sinusoidal shape; in fact, they could be Fourier decomposed into at least four significant sinusoids. Each of these sinusoids can be identified as the interaction with the non-transiting planet via a different resonance. Even with all this extra information, TTVs could only narrow down the possible perturbing planets to a degenerate set of two, and below, we describe how TDVs broke this degeneracy. Planets that are truly in resonance with each other have the largest TTV signals. On a medium-baseline timescale like that of Kepler, they can perturb each other’s orbital periods. The resonant interaction traps the planets at a specific period ratio, causing the periods to oscillate near that ratio. The period of the full cycle of that oscillation depends on the ratio of the planet masses to the host star’s mass, to the 2=3 power (Agol et al. 2005; Nesvorný and Vokrouhlický 2016). For instance, the touchstone system GJ876 has a 550-day libration cycle, about ten times the outer planet’s period, due to its relatively massive planets and lowmass star. A system which was characterized by resonant interaction is KOI-142 (Nesvorný et al. 2013), in which a non-transiting planet was discovered. A system with two transiting planets in resonance with large TTVs is Kepler-30 (Fabrycky et al. 2012). A system with smaller libration amplitudes but a surprising four planets in resonance (forming a chain of resonances) is Kepler-223 (Mills et al. 2016). Several other TTV mechanisms have been detected which do not rely on resonances but are relevant for more hierarchical situations (P2 =P1 & 4). If the outer planet transits, and the inner orbiting body is very massive, the dominant effect can be the shifting of the primary star with respect to the barycenter. Then, as the outer planet orbits the barycenter, it arrives at the moving target either early or late. This effect was numbered (i) by Agol et al. (2005), and it is seen clearly in circumbinary planet systems. For instance, the secondary star of Kepler16 (Doyle et al. 2011) moves the primary by many times its own radius, resulting in an 8 day TTV on top of a 225-day orbit. A final mechanism of dynamical TTV is relevant for the inner orbit when a massive body orbits at large distance. The tide that the body exerts on the inner orbit causes its orbital period to differ slightly from what it would be in the absence of that outer body. If the outer body is in the plane of the inner orbit, its tide slows down the inner orbit, lengthening its period. If the outer body is far out of the plane of the inner

806

E. Agol and D. C. Fabrycky

orbit, its tide speeds up the inner orbit, shortening its period. The tide also depends on the third power of the distance to that external body. Hence, when the external body moves on an eccentric and/or inclined orbit, it induces a period variation in the inner orbit, which has the period of the outer orbit. Also important for the timing is how the outer perturber instantaneously torques the inner orbit’s eccentricity. These effects were put together and analyzed by Borkovits et al. (2003) in the context of triple star systems, and the in-plane physics was explained as mechanism (ii) of Agol et al. (2005). An example of these effects was provided by Kepler-419 (Dawson et al. 2014), in which an eccentric massive planet accompanies an inner planet with a period ratio of 9.7.

Chopping When two planets are nearly resonant, the degeneracy between the mass ratios of the planets to the star and the eccentricity vector may be broken by examining additional TTV components present in the data (Deck and Agol 2015). Nonresonant perturbations occur on the time from one conjunction of the planets to the next, which is when their separation is smallest and gravitational attraction is strongest. Conjunctions occur on a period of Psyn D .1=P1  1=P2 /1 , also referred to as the synodic period. TTVs at the synodic period, and its harmonics, have smaller amplitude due to the fact that they do not add coherently and thus require higher signal to noise to detect. These synodic variations are referred to as “chopping” as they commonly show TTVs that alternate early and late, on top of the larger amplitude TTVs with period PT T V . Despite the smaller amplitude, the chopping components can be detected in many cases and can break the masseccentricity degeneracy, leading to a unique measurement of the masses of the exoplanets (Nesvorný and Vokrouhlický 2014; Schmitt et al. 2014; Deck and Agol 2015). As an example, consider a pair of planets with period ratio of P2 =P1 D 1:52. This period ratio is close to 3:2 and thus is affected by this resonant term, giving a TTV period of 38P1 by Eq. 9. Figure 2 compares two planets with this period ratio with zero eccentricity and mass ratios of 106 to a pair of planets with eccentricities of e1 D e2 D 0:04 and mass ratios near 107 . Both pairs of planets give nearly identical amplitudes for the large resonant term due to the masseccentricity degeneracy discussed above, while the larger mass ratio planets show a much stronger chopping variation. In this case, there is a clear difference between the TTVs of the two simulated systems: the inner planet shows a drift over three orbital periods, and a sudden jump every third orbital period, while the outer one shows a similar pattern over two orbital periods. In this example, the phase of the orbital parameters are set such that the TTV amplitudes match; change in the phase can also be indicative of a non-zero eccentricity contributing to the TTVs, and with an ensemble of planets which are believed to have a similar eccentricity distribution, the mass-eccentricity degeneracy may be broken statistically (Lithwick et al. 2012; Hadden and Lithwick 2014).

37 Transit-Timing and Duration Variations for the Discovery and: : :

807

Fig. 2 Transit-timing variations of two low-eccentricity planets with larger mass ratios, m1 D m2 D 106 m (green) compared with two higher eccentricity planets (e1 D e2 D 0:04) with smaller mass ratios m1 D m2 D 107 m . The zigzag chopping component is apparent in the high-mass/low-eccentricity case while less apparent in the low-mass/high-eccentricity case

Transit Duration Variations TDVs have given useful results for characterization of individual systems, though fewer in number than TTVs. Three mechanisms for TDV have been observed in planetary system orbiting a single primary star. The first is torque due to the rotational oblateness of the star. It is a convincing model for the duration changes in Kepler-13 b (KOI 13.01 Szabó et al. 2012) and a controversial explanation for transit shape anomalies in PTFO 8-8695 (Barnes et al. 2013). The second planetary cause of TDVs is eccentricity variations due to a resonant interaction. The length of the chord across the star and the speed at which the planet moves along that chord are changed during the planetary interaction. This effect has been observed in KOI-142 (Nesvorný et al. 2013). Slow, secular precession of the eccentricity is expected by general relativity (Pál and Kocsis 2008), by stellar oblateness (Heyl and Gladman 2007), and by tidal distortion (Ragozzine and Wolf 2009), but these mechanisms have not given rise to observable TDV to date, for planets around single stars. It is likely that very long time-baseline measurements,

808

E. Agol and D. C. Fabrycky

or comparing the measurements of two time-separated space missions like Kepler and Plato, will be able to detect this effect. The third cause of TDVs for planets around single stars is inclination changes due to secular precession of the orbital plane. Torques from other planets were observed in Kepler-117 (Almenara et al. 2015) and Kepler-108 (Mills and Fabrycky 2017), the latter indicating mutual inclination of 15ı in a rather hierarchical pair of planets. Earlier the case of Kepler-46 was described, in which TTV measurements of a transiting planet led to two degenerate possibilities for the identity of an additional, non-transiting, planet. The clever resolution (Nesvorný et al. 2012) was to note that in one of those solutions, to get the relative amplitudes of the component sinusoids correct in the TTV signal, the perturbing planet must be somewhat inclined with respect to the transiting planet. As a consequence, a torque on that planet would drive TDV. No such TDVs were observed, so the unique solution – which is at a different orbital period and planetary mass and closer to coplanar – was found. Extending this inclination mechanism of TDV to two stars and a planet, the precession of circumbinary planets (CBPs) has been so extreme as to cause transits to turn on and off (Martin 2017). This observation is similar to the several known cases of stellar triples with inner sometimes eclipsing binaries, but in this case, it is most observable in the outer orbit. The first case of that phenomenon was Kepler-35 (Welsh et al. 2012), and the most spectacular observed so far is Kepler-413 (Kostov et al. 2014), in which a 4ı mutual inclination caused transits to stop and then start again nearly half a precession cycle later. Additional dramatic TDVs can occur in CBP systems due to the moving-target effect described above for TTVs. If the transit occurs while the star is moving in the same direction as the planet, the transit duration is longer; if in opposite directions, the transit duration is shorter. Matching the prediction from the phase of the binary completely secures the interpretation of the signal that an object is in a circumbinary orbit, as discussed extensively by Kostov et al. (2013) for the cases of Kepler-47 and Kepler-64 (aka PH-1, KIC 4862625b).

Observational Considerations: Timing Precision The steepest portions of a transit are the ingress and egress when the planet crosses onto and off of the disk of the star, causing a dip of depth ı D .Rp =R /2 if limb darkening is ignored. Suppose for the moment that the only source of noise is Poisson noise due to the count rate of the star, NP . The photometric uncertainty over the duration of ingress,  (Eq. 3), scales as .NP  /1=2 . If the time of ingress fit from a model is offset by  , then the difference in counts observed versus the model is  ı NP (the pink region in Fig. 3). Equating this count deficit to the photometric uncertainty gives  D  1=2 NP 1=2 ı 1 , which is the 68.3% confidence timing precision assuming that the exposure time is much shorter than the ingress duration and that   . The same formula applies to egress. A longer transit ingress duration leads to a shallower slope in ingress, which makes it more difficult to measure an offset in time of the model. Higher count rates and deeper transits

37 Transit-Timing and Duration Variations for the Discovery and: : :

809

δ = (Rp/R∗)2

στ Fig. 3 Diagram of the transit ingress of a planet, flux versus time. The precision of the timing of ingress,  , is set by when the area of the ingress (pink) equals the timing precision over the duration of ingress. The same applies to egress, albeit with the time flipped in this plot

improve the precision, as expected. Note that we’ve assumed that the duration of the transit is sufficiently long that the error on ı is small. Suppose the transit duration is T . Then, the uncertainty on the duration is given by p the sum of the uncertainties on the ingress and egress, added in quadrature: T D 2 . The timing precision, t , is set by the mean of the ingress and egress, giving t D p1  . 2 A more complete derivation of these expressions is given by Carter et al. (2008), while an expression which includes the effects of a finite integration time is given by Price and Rogers (2014). The assumptions of no limb darkening and Poisson noise are generally broken by stars; in addition, stellar variability contributes to timing uncertainty, for which there is yet to be a general expression. These effects generally increase the uncertainty on the measurement of transit times and durations, and so the best practice would be to estimate the timing uncertainties from the data, accounting for effects of correlated stellar variability by including the full covariance matrix of the timing uncertainty (Carter and Winn 2009; Gibson et al. 2012; Foreman-Mackey et al. 2017). Crossing of the path of the planet across star spots may also cause some uncertainty on the timing precision (Oshagh et al. 2013; Barros et al. 2013); this can be diagnosed by a larger scatter within transit than outside transit or other signs of significant stellar activity and can be handled best by including the spots in the transit model (Ioannidis et al. 2016). Note that the barycentric light-travel time offset must be corrected for carefully for high-precision TTV (Eastman et al. 2010).

Science Results The best characterized pair of small planets to date using TTV resides in the Kepler-36 system (Carter et al. 2012). As this planet pair is in close proximity, the conjunctions cause a significant kick to each planet resulting a TTV amplitude that is 1% of the orbital periods of the planets. Figure 4 shows a “river plot” for all 17 quarters of long-cadence Kepler data for this pair of planets. After each 7(6) orbits of the inner (outer) planet (or so), there is a conjunction which causes a change in the eccentricity vector and period of each planet. The change in the eccentricity vector causes a sudden change in the subsequent transit time, while the

810

E. Agol and D. C. Fabrycky

Fig. 4 River plot of Kepler-36b (left) and Kepler-36c (right). Each row of each panel shows the intensity of the star scaled with color and centered on the mean ephemeris of each planet

change in period causes a change in slope; these are apparent for Kepler-36c in Fig. 4. The large TTVs enable a precise measurement of the planet-star mass ratios for both planets (using the TTVs of the companion planet), while the star shows asteroseismic variability which gives a precise estimate of the stellar mass. The result are masses with uncertainties of 100 m) in space. Over the next few years, a much more complete census of extrasolar planets in the solar neighborhood is expected, from a variety of both ground- and spacebased efforts. Combined with new capabilities at low radio frequencies, the radio discovery and study of extrasolar planets remain a promising field. Subsequent to the writing of this chapter, the Sun Radio Interferometer Space Experiment (SunRISE) was selected by NASA to move to Phase A. If selected for flight, SunRISE will consist of six small spacecraft, each carrying a dual-polarization dipole operating in the frequency range 0.1 MHz–25 MHz. While extrasolar planets are not part of the SunRISE science mission, if selected for flight, SunRISE would pioneer a future, larger space-based radio telescope.

References Alexander JK, Brown LW, Clark TA, Stone RG, Weber RR (1969) The spectrum of the cosmic radio background between 0.4 and 6.5 MHz. ApJ 157:L163 Alexander JK, Kaiser ML, Novaco JC, Grena FR, Weber RR (1975) Scientific instrumentation of the radio-astronomy-explorer-2 Satellite. A&A 40:365 Banazadeh P, Lazio J, Jones D, Scharf DP, Fowler W, Aladangady C (2013) Feasibility analysis of XSOLANTRA: a mission concept to detect exoplanets with an array of CubeSats. In: Proceedings of 2013 IEEE aerospace conference. https://doi.org/10.1109/AERO.2013.6496864 Bastian TS, Dulk GA, Leblanc Y (2000) A search for radio emission from extrasolar planets. ApJ 545:1058. https://doi.org/10.1086/317864 Baumback MM, Gurnett DA, Calvert W, Shawhan SD (1986) Satellite interferometric measurements of auroral kilometric radiation. Geophys Res Lett 13:1105 Berger E, Ball S, Becker KM et al (2001) Discovery of radio emission from the brown dwarf LP944-20. Nature 410:338 Burke BF, Franklin KL (1955) Observations of a variable radio source associated with the planet Jupiter. J Geophys Res 60:213 Burningham B, Hardcastle M, Nichols JD et al (2016) A LOFAR mini-survey for low-frequency radio emission from the nearest brown dwarfs. MNRAS 463:2202. https://doi.org/10.1093/ mnras/stw2065 Cane HV (1979) Spectra of the non-thermal radio radiation from the galactic polar regions. MNRAS 189:465 Carr TD, Gulkis S (1969) The magnetosphere of Jupiter. Ann Rev Astron Astrophys 7:577 Christensen UR (2010) Dynamo scaling laws and applications to the planets. Space Sci Rev:152:565

38 Radio Observations as an Exoplanet Discovery Method

831

Christensen UR, Holzwarth V, Reiners A (2009) Energy flux determines magnetic field strength of planets and stars. Nature 457:167. https://doi.org/10.1038/nature07626 de Pater I, Butler BJ, Green DA, Strom R, Millan R, Klein MJ, Bird MK, Funke O, Neidhöfer J, Maddalena R, Sault RJ, Kesteven M, Smits DP, Hunstead R (2003) Jupiter’s radio spectrum from 74 MHz up to 8 GHz. Icarus 163:434. https://doi.org/10.1016/S0019-1035(03)00067-8 Driscoll P, Olson P (2011) Optimal dynamos in the cores of terrestrial exoplanets: magnetic field generation and detectability. Icarus 213:12 Fares R, Donati J-F, Moutou C, Jardine MM, Griessmeier J-M, Zarka P, Shkolnik EL, Bohlender D, Catala C, Collier Cameron A (2010) Searching for star-planet interactions within the magnetosphere of HD 189733. MNRAS 406:409 Farrell WM, Desch MD, Zarka P (1999) On the possibility of coherent cyclotron emission from extrasolar planets. J Geophys Res 104:14025 Fennelly AJ, Matloff GL (1974) Radio detection of Jupiter-like extra-solar planets. J Brit Interplanet Soc 27:660 Franklin KL, Burke BF (1956) Radio observations of Jupiter. AJ 61:177 French FW, Huguenin GR, Rodman AK (1967) A synthetic aperture approach to space-based radio telescopes. J Spacecr Rocket 4:1649 Fujii Y, Spiegel DS, Mroczkowski T, Nordhaus J, Zimmerman NT, Parsons A, Mirbabayi M, Madhusudhan N (2015) Radio emission from red-giant hot Jupiters. ApJ 820:122. https://doi. org/10.3847/0004-637X/820/2/122 Gallagher DL, Dangelo N (1981) Correlations between solar wind parameters and auroral kilometric radiation intensity. Geophys Res Lett 8:1087 George SJ, Stevens IR (2007) Giant metrewave radio telescope low-frequency observations of extrasolar planetary systems. MNRAS 382:455. https://doi.org/10.1111/ j.1365-2966.2007.12387.x Griessmeier J-M (2007) Aspects of the magnetosphere stellar wind interaction of close-in extrasolar planets. Planet Space Sci 55:530 Grießmeier J-M, Stadelmann A, Penz T et al (2004) The effect of tidal locking on the magnetospheric and atmospheric evolution of ‘Hot Jupiters’. A&A 425:753 Griessmeier J-M, Motschmann U, Mann G, Rucker HO (2005) The influence of stellar wind conditions on the detectability of planetary radio emissions. A&A 437:717 Griessmeier J-M, Preusse S, Khodachenko M, Motschmann U, Mann G, Rucker HO (2007a) Exoplanetary radio emission under different stellar wind conditions. Planet Space Sci 55:618 Griessmeier J-M, Zarka P, Spreeuw H (2007b) Predicting low-frequency radio fluxes of known extrasolar planets. A&A 475:359 Grießmeier J-M, Khodachenko M, Lammer H, Grenfell JL, Stadelmann A, Motschmann U (2010) Stellar activity and magnetic shielding. In: Kosovid’ev AG, Andrei AH, Rozelot J-P (eds)Solar and stellar variability: impact on earth and planets. Proceedings of international astronomical union. IAU symposium, vol 264. Cambridge University Press, Cambridge, p 385 Gurnett DA, Kurth WS, Hospodarsky GB et al (2002) Control of Jupiter’s radio emission and aurorae by the solar wind. Nature 415:985 Hallinan G, Antonova A, Doyle JG, Bourke S, Lane C, Golden A (2008) Confirmation of the electron cyclotron maser instability as the dominant source of radio emission from very low mass stars and brown dwarfs. ApJ 684:644. https://doi.org/10.1086/590360 Hallinan G, Sirothia SK, Antonova A, Ishwara-Chandra CH, Bourke S, Doyle JG, Hartman J, Golden A (2013) Looking for a pulse: a search for rotationally modulated radio emission from the hot Jupiter,  Boötis b. ApJ 762:34 Ignace R, Giroux ML, Luttermoser DG (2010) Radio emissions from substellar companions of evolved cool stars. MNRAS 402:2609. https://doi.org/10.1111/j.1365-2966.2009.16085.x Jaeger TR, Osten RA, Lazio TJ, Kassim N, Mutel RL (2011) 325 MHz very large array observations of ultracool dwarfs TVLM 513-46546 and 2MASS J0036C1821104. AJ 142:189. https://doi.org/10.1088/0004-6256/142/6/189 Jakosky BM, Grebowsky JM, Luhmann JG et al (2015) MAVEN observations of the response of Mars to an interplanetary coronal mass ejection. Science 350:0210. https://doi.org/10.1126/ science.aad0210

832

T. J. W. Lazio

Jardine M, Collier Cameron A (2008) Radio emission from exoplanets: the role of the stellar coronal density and magnetic field strength. A&A 490:843 Jones DL et al (2000) The ALFA medium explorer mission. Adv Space Res 26:743 Kao MM, Hallinan G, Pineda JS, Escala I, Burgasser A, Bourke S, Stevenson D (2016) Auroral radio emission from late L and T dwarfs: a new constraint on dynamo theory in the substellar regime. ApJ 818:24. https://doi.org/10.3847/0004-637X/818/1/24 Klein MJ, Thompson TJ, Bolton S (1989) Systematic observations and correlation studies of variations in the synchrotron radio emission from Jupiter. In: Belton MJS, West RA, Rahe J, Pereyda M (eds) Time-variable phenomena in the Jovian system. NASA special publication series, NASA-SP-494. NASA, Washington, DC, p 151 Lazio TJW, Farrell WM (2007) Magnetospheric emissions from the planet orbiting  Bootis: a multiepoch search. ApJ 668:1182 Lazio TJW, Farrell WM, Dietrick J, Greenlees E, Hogan E, Jones C, Hennig LA (2004) The radiometric bode’s law and extrasolar planets. ApJ 612:511 Lazio TJW, Carmichael S, Clark J, Elkins E, Gudmundsen P, Mott Z, Szwajkowski M, Hennig LA (2010a) A blind search for magnetospheric emissions from planetary companions to nearby solar-type stars. AJ 139:96 Lazio TJW, Shankland PD, Farrell WM, Blank DL (2010b) Radio observations of HD 80606 near planetary periastron. AJ 140:1929 Lazio TJW, Shkolnik E, Hallinan G et al (2016) Planetary magnetic fields: planetary interiors and habitability. W. M. Keck Institute for Space Studies. http://kiss.caltech.edu/new_website/ programs/magnetic_final_report.pdf Lecavelier Des Etangs A, Sirothia SK, Gopal-Krishna A, Zarka P (2009) GMRT radio observations of the transiting extrasolar planet HD 189733 b at 244 and 614 MHz. A&A 500:L51 Lecavelier Des Etangs A, Sirothia SK, Gopal-Krishna A, Zarka P (2011) GMRT search for 150 MHz radio emission from the transiting extrasolar planets HD 189733 b and HD 209458 b. A&A 533:A50 Lecavelier Des Etangs A, Sirothia SK, Gopal-Krishna A, Zarka P (2013) Hint of 150 MHz radio emission from the Neptune-mass extrasolar transiting planet HAT-P-11b. A&A 552:A65 Lynch C, Murphy T, Ravi V, Hobbs G, Lo K, Ward C (2016) Radio detections of southern ultracool dwarfs. MNRAS 457:1224. https://doi.org/10.1093/mnras/stw050 Murphy T, Bell ME, Kaplan DL et al (2015) Limits on low frequency radio emission from southern exoplanets with the Murchison widefield array. MNRAS 446:2560. https://doi.org/10.1093/ mnras/stu2253 Mutel R, Gurnett DA, Christopher I (2004) Spatial and temporal properties of AKR burst emission derived from cluster WBD VLBI studies. Ann Geophys 22:2625 Nichols JD (2011) Magnetosphere-ionosphere coupling at Jupiter-like exoplanets with internal plasma sources: implications for detectability of auroral radio emissions. MNRAS 414:2125 Nichols JD (2012) Candidates for detecting exoplanetary radio emissions generated by magnetosphere-ionosphere coupling. MNRAS 427:L75 Pérez LM, Carpenter JM, Andrews SM et al (2016) Spiral density waves in a young protoplanetary disk. Science 353:1519. https://doi.org/10.1126/science.aaf8296 Reiners A, Christensen UR (2010) A magnetic field evolution scenario for brown dwarfs and giant planets. A&A 522:A13 Rogers LA, Seager S (2010) A framework for quantifying the degeneracies of exoplanet interior compositions. ApJ 712:974 Route M, Wolszczan A (2016) Radio flaring from the T6 dwarf WISEPC J112254.73C255021.5 with a possible ultra-short periodicity. ApJ 821:L21. http://doi.org/10.3847/2041-8205/821/2/ L21 Schubert G, Soderlund KM (2011) Planetary magnetic fields: observations and models. Phys Earth Plan Inter 187:92 Sirothia SK, Lecavelier des Etangs A, Gopal-Krishna A, Kantharia NG, Ishwar-Chandra CH (2014) Search for 150 MHz radio emission from extrasolar planets in the TIFR GMRT sky survey. A&A 562:A108. https://doi.org/10.1051/0004-6361/201321571

38 Radio Observations as an Exoplanet Discovery Method

833

Smith AMS, Collier Cameron A, Greaves J, Jardine M, Langston G, Backer D (2009) Secondary radio eclipse of the transiting planet HD 189733 b: an upper limit at 307–347 MHz. MNRAS 395:335. https://doi.org/10.1111/j.1365-2966.2009.14510.x Stevens IR (2005) Magnetospheric radio emission from extrasolar giant planets: the role of the host stars. MNRAS 356:1053 Stevenson DJ (2010) Planetary magnetic fields: achievements and prospects. Space Sci Rev 152:651 Tarter J (2001) The search for extraterrestrial intelligence (SETI). ARA&A 39:511. https://doi.org/ 10.1146/annurev.astro.39.1.511 Taylor GB, Ellingson SW, Kassim NE et al (2012) First light for the first station of the long wavelength array. J Astron Instrum 1:1250004. https://doi.org/10.1142/S2251171712500043 Treumann RA (2006) The electron-cyclotron maser for astrophysical application. A&A Rev 13:229 van Haarlem MP et al (2013) LOFAR: the lOw-frequency array. A&A 556:A2 Vanhamäki H (2011) Emission of cyclotron radiation by interstellar planets. Planet Space Sci 59:862. https://doi.org/10.1016/j.pss.2011.04.002 Vidotto AA, Opher M, Jatenco-Pereira V, Gombosi TI (2010) Simulations of winds of weak-lined T Tauri stars. II. The effects of a tilted magnetosphere and planetary interactions. ApJ 720:1262. https://doi.org/10.1088/0004-637X/720/2/1262 Vorgul I, Kellett BJ, Cairns RA, Bingham R, Ronald K, Speirs DC, McConville SL, Gillespie KM, Phelps ADR (2011) Cyclotron maser emission: Stars, planets, and laboratory. Phys Plasmas 18:056501. https://doi.org/10.1063/1.3567420 Willes AJ, Wu K (2005) Radio emissions from terrestrial planets around white dwarfs. A&A 432:1091. https://doi.org/10.1051/0004-6361:20040417 Winglee RM, Dulk GA, Bastian TS (1986) A search for cyclotron maser radiation from substellar and planet-like companions of nearby stars. ApJ 309:L59 Wolf S (2008) Detecting protoplanets with ALMA. Ap&SS 313:109. http://doi.org/10.1007/ s10509-007-9660-z Wood BE, Linsky JL, Müller H-R, Zank GP (2001) Observational estimates for the mass-loss rates of ˛ centauri and proxima centauri using hubble space telescope Ly˛ spectra. ApJ 547:L49 Wood BE, Müller H-R, Zank GP, Linsky JL (2002) Measured mass-loss rates of solar-like stars as a function of age and activity. ApJ 574:412 Wood BE, Müller H-R, Zank GP, Linsky JL, Redfield S (2005) New mass-loss measurements from astrospheric Ly˛ absorption. ApJ 628:L143 Yantis WF, Sullivan WT III, Erickson WC (1977) A search for extra-solar Jovian planets by radio techniques. BAAS 9:453 Zarka P (1992) The auroral radio emissions from planetary magnetospheres – what do we know, what don’t we know, what do we learn from them? Adv Space Res 12:99 Zarka P (2006) Hot Jupiters and magnetized stars: giant analogs of the satellite-Jupiter system. In: Rucker H, Kurth W, Mann G (eds) Planetary radio emissions VI. Austrian Academy of Sciences Press, Vienna, p. 543 Zarka P (2007) Plasma interactions of exoplanets with their parent star and associated radio emissions. Planet Space Sci 55:598 Zarka P, Queinnec J, Ryabov BP et al (1997) Ground-based high sensitivity radio astronomy at decameter wavelengths. in planetary radio emission IV. In: Rucker HO, Bauer SJ, Lecacheux A (eds) Proceedings of the 4th international workshop. Austrian Academy of Sciences Press, Vienna, p 101 Zarka P, Treumann RA, Ryabov BP, Ryabov VB (2001) Magnetically-driven planetary radio emissions and application to extrasolar planets. Ap&SS 277:293

Detecting and Characterizing Exomoons and Exorings

39

René Heller

Contents Introduction: Why Bother About Moons? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Early Searches for Exomoons and Exorings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tentative Detections of Exomoons and Exorings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Detection Methods for Exomoons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dynamical Effects on Planetary Transits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Direct Transit Signatures of Exomoons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Other Methods for Exomoon Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Detection Methods for Exorings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Direct Photometric Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Other Detection Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

836 836 837 840 840 842 843 845 846 846 847 848

Abstract

Since the discovery of a planet transiting its host star in the year 2000, thousands of additional exoplanets and exoplanet candidates have been detected, mostly by NASA’s Kepler space telescope. Some of them are almost as small as the Earth’s moon. As the solar system is teeming with moons, more than a hundred of which are in orbit around the eight local planets, and with all of the local giant planets showing complex ring systems, astronomers have naturally started to search for moons and rings around exoplanets in the past few years. We here discuss the principles of the observational methods that have been proposed to find moons and rings beyond the solar system, and we review the first searches. Though no exomoon or exoring has been unequivocally validated so far, theoretical and technological requirements are now on the verge of being mature for such discoveries. R. Heller () Max Planck Institute for Solar System Research, Göttingen, Germany e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_35

835

836

R. Heller

Introduction: Why Bother About Moons? The moons of the solar system planets have become invaluable tracers of the local planet formation and bombardment, e.g., for the Earth-Moon system (Cameron and Ward 1976; Rufu et al. 2017) and for Mars and its tiny moons Phobos and Deimos (Rosenblatt et al. 2016). The composition of the Galilean moons constrains the temperature distribution in the accretion disk around Jupiter 4.5 billion years ago (Pollack and Reynolds 1974; Canup and Ward 2002; Heller et al. 2015). And while the major moons of Saturn, Uranus, and Neptune might have formed from circumplanetary tidal debris disks (Crida and Charnoz 2012), Neptune’s principal moon Triton has probably been caught during an encounter with a minor planet binary (Agnor and Hamilton 2006). The orbital alignment of the Uranian moon systems suggests a “collisional tilting scenario” (Morbidelli et al. 2012) and implies significant bombardment of the young Uranus. A combination of these observations constrains the migration of the giant planets (Deienno et al. 2011), planet-planet encounters (Deienno et al. 2014), bombardment histories (Levison et al. 2001), and the properties of the early circumsolar disk (Jacobson and Morbidelli 2014). The Pluto-Charon system can be considered a planetary binary rather than a planet-moon system, since its center of mass is outside the radius of the primary, at about 1.7 Pluto radii. A giant impact origin of this system delivers important constraints on the characteristic frequency of large impacts in the Kuiper belt region (Canup 2005). Planetary rings consist of relatively small particles, from sub-grain-sized to bouldersized, and they are indicators of moon formation and moon tidal/geophysical activity; see Enceladus around Saturn (Spahn et al. 2006). Exomoon and exoring discoveries can thus be expected to deliver information on exoplanet formation on a level that is fundamentally new and inaccessible through exoplanet observations alone. Yet, although thousands of planets have been found beyond the solar system, no natural satellite has been detected around any of them. So one obvious question to ask is: Where are they?

Early Searches for Exomoons and Exorings So far, most of the searches for exomoons have been executed as piggyback science on projects with a different primary objective. To give just a few examples, Brown et al. (2001) used the exquisite photometry of the Hubble Space Telescope (HST) to observe four transits of the hot Jupiter HD 209458 b in front of its host star. As the star has a particularly high apparent brightness and therefore delivers very high signal-to-noise transit light curves, these observations would have revealed the direct transits of slightly super-Earth-sized satellites (&1:2 R˚ ; R˚ being the Earth’s radius) around HD 209458 b if such a moon were present. Alternatively, the gravitational pull from any moon that is more massive than about 3 M˚ (M˚ being the Earth’s mass) could have been detected as well. Yet, no evidence for such a

39 Detecting and Characterizing Exomoons and Exorings

837

large moon was found. Brown et al. (2001) also constrained the presence of rings around HD 209458 b, which must be either extremely edge-on (so they barely have an effect on the stellar brightness during transit) or they must be restricted to the inner 1:8 planetary radii around HD 209458 b. In a similar vein, Charbonneau et al. (2006) found no evidence for moons or rings around the hot Saturn HD 149026 b, Pont et al. (2007) found no moons or rings around HD 189733 b, and Santos et al. (2015) rejected the exoring hypothesis for 51 Peg b. Maciejewski et al. (2010) used ground-based observations to search for moons around WASP-3b by studying the planet’s transit timing variations (TTVs) and transit duration variations (TDVs). Yet, as TDVs remained undetectable in that system, an exomoon scenario seems very unlikely to cause the observed TTVs. More recently, Heising et al. (2015) scanned a sample of 21 transiting planets observed with the Kepler space telescope (Borucki et al. 2010) and found no conclusive evidence for ring signatures. Later, Lecavelier des Etangs (2017) published a search for moons and rings around the warm Jupiter-sized exoplanet CoRoT-9 b based on infrared (4.5 m) photometry obtained with the Spitzer space telescope during two transits in 2010 and 2011. Moons larger than 2.5 Earth radii were excluded at the 3 confidence level, and large silicate-rich (alternatively: icy) rings with inclinations 13ı (alternatively: 13ı ) against the line of sight were also excluded. The Hunt for Exomoons with Kepler (HEK) project (Kipping et al. 2012), the first dedicated survey targeting moons around extrasolar planets, is probably the best bet for a near-future exomoon detection. Their analysis combines TTV and TDV measurements of exoplanets with searches for direct photometric transit signatures of exomoons. The most recent summary of their Bayesian photodynamical modeling (Kipping 2011) of exomoon transits around a total of 57 Kepler Objects of Interest has been presented by Kipping et al. (2015). Other teams found unexplained TTVs in many transiting exoplanets from the Kepler mission (Szabó et al. 2013), but without additional TDVs or direct photometric transits, a robust exomoon interpretation is impossible.

Tentative Detections of Exomoons and Exorings While a definite exomoon discovery remains to be announced, some tentative claims have already been presented in the literature. One of the first exomoon claims was put forward by Bennett et al. (2014) based on the microlensing event MOA2011-BLG-262. Their statistical analysis of the microlensing light curve, however, has a degenerate solution with two possible interpretations. It turns out that an C28 interpretation invoking a 0:11C0:21 0:06 Mˇ star with a 1710 M˚ planetary companion C0:53 at 0:950:19 AU is a more reasonable explanation than the hypothetical 3:2 MJup mass free-floating planet with a 0:47 M˚ -mass moon at a separation of 0.13 AU. Sadly, the sources of microlensing events cannot be followed up. As a consequence,

838

R. Heller

no additional data can possibly be collected to confirm or reject the exomoon hypothesis of MOA-2011-BLG-262. In the same year, Ben-Jaffel and Ballester (2014) proposed that the observed asymmetry in the transit light curve of the hot Jupiter HD 189733 b might be caused by an opaque plasma torus around the planet, which could be fed by a tidally active natural companion around the planet (which is not visible in the transit light curve itself, in this scenario). But an independent validation has not been demonstrated. Using a variation of the exoplanet transit method, Hippke (2015) presented the first evidence of an exomoon population in the Kepler data. The author used what he refers to as a superstack, a combination of light curves from thousands of transiting exoplanets and candidates, to create an average transit light curve from Kepler with a very low noise-to-signal level of about 1 part per million. This superstack of a light curve exhibits an additional transit-like signature to both sides of the averaged planetary transit, possibly caused by many exomoons that are hidden in the noise of the individual light curves of each exoplanet. The depth of this additional transit candidate feature corresponds to an effective moon radius of 2120C330 370 km, or about 0.8 Ganymede radii. Interestingly, this signal is much more pronounced in the superstack of planets with orbital periods larger than about 35 d, whereas more close-in planets do not seem to show this exomoon-like feature. This finding is in agreement with considerations of the Hill stability of moons, which states that stellar gravitational perturbations may perturb the orbit of a moon around a close-in planet such that the moon will be ejected (Domingos et al. 2006). Teachey et al. (2017) also used a stacking method for 284 Kepler Objects of Interest with sufficiently high signal-to-noise ratios to constrain the occurrence rate of Galilean-analog moon systems to be  < 0:38 with a 95% confidence and  D 0:16C0:13 0:10 with a 68.3% confidence. If the signal in the stacked and phase-folded Kepler lightcurve is genuinely due to an exomoon population, then these moons would have average diameters roughly half the size of the Earth and orbit their host planets at about 5 to 10 planetary radii. Equally important, Teachey et al. (2017) present a viable exomoon candidate in orbit around Kepler-1625 b. The satellite nature of what has provisionally been dubbed Kepler-1625 b-i still needs to be confirmed (observations with the Hubble Space Telescope have been scheduled for October 2017), but the preliminary analysis suggests that this candidate would have a radius roughly the size of Neptune – thus its informal designation as a “Nept-moon” – and it would orbit a very massive Jupiter-sized planet. If validated, this binary would certainly present a benchmark system to study planet and moon formation because it would be fundamentally different from any planet-moon system found in the solar system. Beyond the exquisite photometric data quality of the Kepler telescope, the COnvection ROtation and planetary Transits (CoRoT; Auvergne et al. 2009) space mission also delivered highly accurate space-based stellar observations. One particularly interesting candidate object is CoRoT SRc01 E2 1066, which shows a peculiar bump near the center of the transit light curve that might be induced by the mutual eclipse of a transiting binary planet system (Lewis et al. 2015), i.e., a giant planet with a very large and massive satellite. However, only one single transit of this

39 Detecting and Characterizing Exomoons and Exorings

839

object (or these two objects) has been observed, and so it is currently impossible to discriminate between a binary planet and a star-spot crossing interpretation of the data. There has also been one supposed observation of a transiting ring system, which has been modeled to explain the curious brightness fluctuation of the 16 Myr young K5 star 1SWASP J140747.93-394542.6 (J1407 for short) observed around 29 April 2007 (Mamajek et al. 2012). The lower panel of Fig. 1 shows the observed stellar brightness variations, and the panel above displays the hypothesized ring system that could explain the data. This visualization nicely illustrates the connection between rings and moons, as the gaps in this proposed ring system could have been cleared by large moons that were caught in a stage of ongoing formation (Kenworthy and Mamajek 2015). The most critical aspect of this interpretation though is in the fact that the hypothesized central object has not been observed in transit. Another issue is that the orbital period of this putative ring system around J1407 can only be constrained to be 10 yr (Kenworthy et al. 2015, and private communication with M. Kenworthy). In other words, the periodic nature of this proposed transit event has not actually been established, and it could take decades to reobserve this phenomenon, if the interpretation were valid in the first place.

Fig. 1 The upper panel shows a computer model of a ring system transiting the star J1407 to explain the 70 d of photometric observations from the Super Wide-Angle Search for Planets (SuperWASP) in the lower panel. The thick green line in the background of the upper panel represents the path of the star relative to the rings. Gray annuli indicate regions of the possible ring system that are not constrained by the data. Gradation of the red colors symbolizes the transmissivity of each ring. Note that the hypothesized central object of the ring system (possibly a giant planet) is not transiting the star (Image credit: Kenworthy and Mamajek 2015)

840

R. Heller

Almost all of the abovementioned studies have been inspired by the advent of space-based high-accuracy stellar photometric observations, mostly driven by the CoRoT and Kepler space telescopes. The newly gained access to this kind of a data quality has sparked huge interest in the possibility of novel detection methods for exomoons and exorings. In the following, we first consider detection methods for exomoons and then we discuss exoring detection methods.

Detection Methods for Exomoons About a dozen different theoretical methods have been proposed to search and characterize exomoons. For the purpose of this review, we will group them into three classes: (1) dynamical effects of the transiting host planet, (2) direct photometric transits of exomoons, and (3) other methods.

Dynamical Effects on Planetary Transits The moons of the solar system are small compared to their planet, and so the natural satellites of exoplanets are expected to be small as well. The depth (d ) of an exomoon’s photometric transit scales with the satellite radius (Rs ) squared: d / Rs2 . Consequently, large exomoons could be relatively easy to detect (if they exist), but small satellites would tend to be hidden in the noise of the data. Alternatively, instead of hunting for the tiny brightness fluctuations caused by the moons themselves, it has been suggested that their presence could be derived indirectly by measuring the TTVs and TDVs of their host planets. The amplitudes of both quantities (TTV and TDV ) are linear in the mass of the satellite: TTV / Ms / TTV (Sartoretti and Schneider 1999; Kipping 2009a). Hence, the dynamical effect of low-mass moons is less suppressed than the photometric effect of small-radius moons.

Transit Timing Variation In a somewhat simplistic picture, neglecting the orbital motion of a planet and its moon around their common center of gravity during their common stellar transit, TTVs are caused by the tangential offset of the planet from the planet-moon barycenter (see upper left illustration in Fig. 2, where “BC” denotes the barycenter). In a sequence of transits, the planet has different offsets during each individual event, assuming that it is not locked in a full-integer orbital resonance with its circumstellar orbit. Hence, its transits will not be precisely periodic but rather show TTVs, approximately on the order of seconds to minutes (compared to orbital periods of days to years). Two flavors of observable TTV effects have been discussed in the literature. One is called the barycentric TTV method (TTVb ; Sartoretti and Schneider 1999; Kipping 2009a), and one is referred to as the photocentric TTV method (TTVp or PTV; Szabó et al. 2006; Simon et al. 2007, 2015). A graphical representation of both methods is shown in Fig. 2.

39 Detecting and Characterizing Exomoons and Exorings

841

Moon

Planet

PC

BC

Planet

Moon

Moon

Planet

BC

Moon

Planet

PC

BC

BC

TTV

PTV

0 100

trailing moon

ppm

200 TTV

PTV

0 100

leading moon

200 -600

-400

-200

0 Time [min]

200

400

600

Fig. 2 Physical explanation of the barycentric TTV (upper left) and the photocentric TTV (upper right). The two light curves at the bottom illustrate both the TTVb and the TTVp (or PTV) for a moon that is trailing the planet (upper panel, green curve) and leading the planet during the stellar transit (lower panel, red curve) (Image credit: Simon et al. (2015). ©The Astronomical Society of the Pacific. Reproduced with permission. All rights reserved)

TTVb measurements refer to the position of the planet relative to the planetmoon barycenter. From the perspective of a light curve analysis, this corresponds to measuring the time differences of the planetary transit only, e.g., of the ingress, center, and/or egress (Sartoretti and Schneider 1999). PTV measurements, on the other hand, take into account the photometric effects of both the planet and its moon, and so the corresponding amplitudes can actually be significantly larger than TTVb amplitudes, details depending on the actual masses and radii of both objects (Simon et al. 2015).

Transit Duration Variation Planetary TDVs can be caused by several effects. First, they can be produced by the change of the planet’s tangential velocity component around the planetmoon barycenter between successive transits (referred to as the TDVV component; Kipping 2009a). When the velocity component in the planet-moon system that is tangential to the observer’s line of sight adds to the circumstellar tangential velocity during the transit, then the event is relatively short. On the other hand, if the transit catches the planet during its reverse motion in the planet-moon system, then the total

842

R. Heller

tangential velocity is lower than that of the barycenter, and so the planetary transit takes somewhat longer. TDV effects can also be introduced if the planet-moon orbital plane is inclined with respect to the circumstellar orbital plane of their mutual center of gravity. In this case, the planet’s apparent minimum distance from the stellar center will be different during successive transits, in more technical terms: its transit Impact parameter will change between transits or, if the moon’s orbital motion around the planet is fast enough, even during the transits. This can induce a TDVTIP component in the transit duration measurements of the planet (Kipping 2009b). It is important to realize that the waveforms of the TTV and TDV curves are offset by an angle of =2 (Kipping 2009a). In a more visual picture, when the TTV is zero, i.e., when the planet is along the line of sight with the planetmoon barycenter, then the corresponding TDV measurement is either largest (for moons on obverse motion) or smallest (for moons on reverse motion), since the planet would have the largest/smallest possible tangential velocity in the planetmoon binary system. This phase difference is key to breaking the degeneracy of simultaneous Ms and as measurements (as being the moon’s semimajor axis around the planet). When plotted in a TTV-TDV diagram (Montalto et al. 2012; Awiphan and Kerins 2013), the resulting ellipse contains predictable dynamical patterns, which can help to discriminate an exomoon interpretation of the data from a planetary perturber, and it may even allow the detection of multiple moons (Heller et al. 2016b).

Direct Transit Signatures of Exomoons Like planets, moons could naturally imprint their own photometric transits into the stellar light curves, if they were large enough (Tusnski and Valio 2011). The lower two panels of Fig. 2 show an exomoon’s contribution to the stellar bright variation in case the moon is trailing (upper light curve) or leading (lower light curve) its planet. Note that if the moon is leading, then its transit starts prior to the planetary transit, and so the exomoon transit affects the right part of the planetary transit in the light curve. As mentioned in section “Dynamical Effects on Planetary Transits”, the key challenge is in the actual detection of this tiny contribution, which has hitherto remained hidden in the noise of exoplanet light curves. As a variation of the transit method, it has been suggested that mutual planetmoon eclipses during their common stellar transit might betray the presence of an exomoon or binary planetary companion (Sato and Asada 2009; Pál 2012). This is a particularly interesting method, since the mutual eclipses of two transiting planets have already been observed (de Wit et al. 2016). Yet, in the latter case, the two planets were known to exist prior to the observation of their common transit, whereas for a detection of an exomoon through mutual eclipses, it would be necessary to test the data against a possible origin from star-spot crossings of the planet (Lewis et al. 2015) and to use an independent method for validation.

39 Detecting and Characterizing Exomoons and Exorings

843

Orbital Sampling Effect One way to generate transit light curves with very high signal-to-noise ratios in order to reveal exomoons is by folding the measurements of several transits of the same object into one phase-folded transit light curve. Figure 3 shows a simulation of this phase-folding technique, which is referred to as the orbital sampling effect (OSE; Heller 2014; Heller et al. 2016a). The derived light curve does not effectively contain “better” data than the combination of the individual transit light curves (in fact it loses any information about the individual TTV and TDV measurements), but it enables astronomers to effectively search for moons in large data sets, as has been done by Hippke (2015) to generate superstack light curves from Kepler (see section “Tentative Detections of Exomoons and Exorings”). Scatter Peak The minimum possible noise level of photometric light curves is given by the shot noise (or Poisson noise, white noise, time-uncorrelated noise), which depends on the number of photons collected and, thus, on the apparent brightness of the star. For the amount of photons typically collected with space-based optical telescopes, the minimum possible signal-to-noise ratio (SNR) of a light curve p can be approximated as the square root of the number of photons (n): SNR / n. And so the SNR of a phase-folded transit light curve of a given planet goes down p with the square root of the number of phase-folded transits (N ): SNROSE / N . In other words, for planets transiting photometrically quiet host stars, the noise-to-signal ratio (1/SNR) of the phase-folded light curve converges to zero for an increasing number of transits. If the planet is accompanied by a moon, however, then the variable position of the moon with respect to the planet induces an additional noise component. As a consequence, and although the average light curve is converging toward analytical models (see Fig. 3), the noise in the planetary transit is actually increasing due to the moon. For large N , once dozens and hundreds of transits can be phase-folded, the OSE becomes visible together with a peak in the noise, the latter of which has been termed the scatter peak (Simon et al. 2012). As an aside, the superstack OSE candidate signal found by Hippke (2015) was not accompanied by any evidence of a scatter peak.

Other Methods for Exomoon Detection In some cases, where the planet and its moon (or multiple moons) are sufficiently far from their host star, it could be possible to optically resolve the planet from the star. This has been achieved more than a dozen times now through a method known as direct imaging (Marois et al. 2008). Though direct imaging cannot, at the current stage of technology, deliver images of a resolved planet with individual moons, it might still be possible to detect the satellites. One could either try and detect the shadows and transits of the moons across their host planet in the integrated

Fig. 3 (continued)

844 R. Heller

39 Detecting and Characterizing Exomoons and Exorings

845

(i.e., unresolved) infrared light curve of the planet-moon system (Cabrera and Schneider 2007; Heller 2016), or one could search for variations in the position of the planet-moon photocenter with respect to some reference object, e.g., another star or nearby exoplanet in the same system (Cabrera and Schneider 2007; Agol et al. 2015). Fluctuations in the infrared light received from the directly imaged planet ˇ Pic b, as an example, could be due to an extremely tidally heated moon (Peters and Turner 2013) that is occasionally seen in transit or (not seen) during the secondary eclipse behind the planet. A related method is in the detection of a variation of the net polarization of light coming from a directly imaged planet, which might be caused by an exomoon transiting a luminous giant planet (Sengupta and Marley 2016). It could also be possible to detect exomoons through spectral analyses, e.g., via excess emission of giant exoplanets in the spectral region between 1 and 4 m (Williams and Knacke 2004), enhanced infrared emission by airless moons around terrestrial planets (Moskovitz et al. 2009; Robinson 2011), and the stellar RossiterMcLaughlin effect of a transiting planet with moons (Simon et al. 2010; Zhuang et al. 2012) or the Rossiter-McLaughlin effect of a moon crossing a directly imaged, luminous giant planet (Heller and Albrecht 2014). Some more exotic exomoon detection methods invoke microlensing (Han and Han 2002; Liebig and Wambsganss 2010; Bennett et al. 2014; Skowron et al. 2014), pulsar timing variations (Lewis et al. 2008), modulations of radio emission from giant planets (Noyola et al. 2014, 2016), or the generation of plasma tori around giant planets by volcanically active moons (Ben-Jaffel and Ballester 2014).

Detection Methods for Exorings Just like moons are very common around the solar system planets, rings appear to be a common feature as well. Naturally, the beautiful ring system around Saturn was the first to be discovered. Less obvious rings have also been detected around all other gas giants in the solar system and even around an asteroid (Braga-Ribas et al. 2014). In the advent of exoplanet detections, astronomers have thus started to develop methods for the detection of rings around planets outside the solar system.

J Fig. 3 Orbital sampling effect (OSE) of a simulated transiting exoplanet with moons. The upper panel shows a model of the phase-folded transit light curve of a Jupiter-sized planet around a 0:64 Rˇ K dwarf star using an arbitrarily large number of transits. The planet is accompanied by three moons of 0:86 R˚ , 0:52 R˚ , and 0:62 R˚ in radial size, but their contribution to the phasefolded light curve is barely visible with the naked eye. The lower row of panels shows a sequence of zooms into the prior-to-ingress part of the planetary transit. The evolution of the OSEs of the three moons is shown for an increasing number of transits (N ) used to generate the phase-folded light curves. In each panel, the solid line shows the simulated phase-folded transit, and the dashed line shows an analytical model, both curves assuming a star without limb darkening (Image credit: Heller 2014)

846

R. Heller

Direct Photometric Detection The detection of rings around exoplanets is closely related to many of the abovementioned methods (see section “Direct Transit Signatures of Exomoons”) of direct photometric transit observations of exomoons. Like moons, rings can cause additional dips in the planetary transit light curve (Tusnski and Valio 2011). But rings do not induce any dynamical effects on the planet and, hence, there will be no TTVs or TDVs. In fact, a ring system can be expected to impose virtually the same pattern on each individual transit of its host planet because rings should look the same during each transit. Moons, however, would have a different position relative to the planet during individual transits (except for the case of full-integer orbital resonances between the circumstellar and the circumplanetary orbits). This static characteristic of the expected ring signals make it susceptible to misinterpretation, e.g., by a standard fit of a planet-only model to a hypothetically observed planetwith-ring light curve: in this case, the planet radius would be slightly overestimated, while the ring could remain undetected. However, the O-C diagram (O for observed, C for calculated) could still indicate the ring signature (Barnes and Fortney 2004; Zuluaga et al. 2015) As a consequence, rings could induce a signal into the phase-folded transit light curve, which is very similar to the OSE (section “Orbital Sampling Effect”), since the latter is equivalent to a smearing of the moon over its circumplanetary orbit – very much like a ring. The absence of dynamical effects like TTVs and TDVs, however, means that there would also be no scatter peak for rings (section “Scatter Peak”). Thus, an OSE-like signal in the phase-folded light curve without an additional scatter peak could indicate a ring rather than a moon. One particular effect that has been predicted for light curves of transiting ring systems is diffraction, or forward-scattering (Barnes and Fortney 2004). Diffraction describes the ability of light to effectively bend around an obstacle, a property that is rooted in the waveform nature of light. In our context, the light of the host star encounters the ring particles along the line of sight, and those of which are m- to 10 m-sized will tend to scatter light into the forward direction, that is, toward the observer. In other words, rings cannot only obscure the stellar light during transit, they can also magnify it temporarily (see Fig. 4). Moreover, rings can also reflect the stellar light to a significant degree, which can cause variations in the out-of-transit phases of the lightcurve (Arnold and Schneider 2004).

Other Detection Methods Beyond those potential ring signals in the photometric transit data, rings might also betray their presence in stellar transit spectroscopy. The crucial effect here is similar to the Rossiter-McLaughlin effect of transiting planets: as a transiting ring proceeds over the stellar disk, it modifies the apparent, disk-integrated radial velocity of its rotating host star. This is because the ring covers varying parts of the disk, all of

39 Detecting and Characterizing Exomoons and Exorings

847

Fig. 4 Transit of Saturn and its ring system in front of the sun as seen by the Cassini spacecraft in September 2016. Note how the rings bend the sun light around the planet, an effect known as diffraction (Image credit: NASA/JPL/Space Science Institute)

which have a very distinct contribution to the rotational broadening of the stellar spectral lines (Ohta et al. 2009). Qualitatively speaking, if the planet with rings transits the star in the same direction as the direction of stellar rotation, then the ring (and planet) will first cover the blue-shifted parts of the star. Hence, the stellar radial velocity will occur redshifted during about the first half of the transit – and vice versa for the second half. Another more indirect effect can be seen in the Fourier space (i.e., in the frequency domain rather than the time domain) of the transit light curve, where the ring can potentially stand out as an additional feature in the curve of the Fourier components as a function of frequency (Samsing 2015).

Conclusions In this chapter, we discussed about a dozen methods that various researchers have worked out over little more than the past decade to search for moons and rings beyond the solar system. Although some of the original studies, in which these methods have been presented, expected that moons and rings could be detectable with the past CoRoT space mission or with the still active Kepler space telescope (Sartoretti and Schneider 1999; Barnes and Fortney 2004; Kipping et al. 2009; Heller 2014), no exomoon or exoring has been unequivocally discovered and confirmed as of today. This is likely not because these extrasolar objects and structures do not exist, but because they are too small to be distinguished from the noise. Alternatively, and this is a more optimistic interpretation of the situation, those features could actually be detectable and present in the available archival data (maybe even in the HST archival

848

R. Heller

data of transiting exoplanets), but they just haven’t been found yet. The absence of numerous, independent surveys for exomoons and exorings lends some credence to this latter interpretation: Out of the several thousands of exoplanets and exoplanet candidates discovered with the Kepler telescope alone, only a few dozen have been examined for moons and rings with statistical scrutiny (Heising et al. 2015; Kipping et al. 2015). It can be expected that the Kepler data will be fully analyzed for moons and rings within the next few years. Hence, a detection might still be possible. Alternatively, an independent, targeted search for moons/rings around planets transiting apparently bright stars – e.g., using the HST, CHEOPS, or a 10 m scale ground-based telescope, might deliver the first discoveries in the next decade. If none of these searches would be proposed or proposed but not granted, then it might take more than a decade for the PLAnetary Transits and Oscillations of stars (PLATO) mission (Rauer et al. 2014) to find an exomoon or exoring in its large space-based survey of bright stars. Either way, it can be expected that exomoon and exoring discoveries will allow us a much deeper understanding of planetary systems than is possibly obtainable by planet observations alone.

References Agnor CB, Hamilton DP (2006) Neptune’s capture of its moon Triton in a binary-planet gravitational encounter. Nature 441:192–194. https://doi.org/10.1038/nature04792 Agol E, Jansen T, Lacy B, Robinson T, Meadows V (2015) The center of light: spectroastrometric detection of exomoons. ApJ 812:5. https://doi.org/10.1088/0004-637X/812/1/5, 1509.01615 Auvergne M, Bodin P, Boisnard L et al (2009) The CoRoT satellite in flight: description and performance. A&A 506:411–424. https://doi.org/10.1051/0004-6361/200810860, 0901.2206 Awiphan S, Kerins E (2013) The detectability of habitable exomoons with Kepler. MNRAS 432:2549–2561. https://doi.org/10.1093/mnras/stt614, 1304.2925 Barnes JW, Fortney JJ (2004) Transit detectability of ring systems around extrasolar giant planets. ApJ 616:1193–1203. https://doi.org/10.1086/425067, astro-ph/0409506 Ben-Jaffel L, Ballester GE (2014) Transit of Exomoon plasma tori: new diagnosis. ApJ 785:L30. https://doi.org/10.1088/2041-8205/785/2/L30, 1404.1084 Bennett DP, Batista V, Bond IA et al (2014) MOA-2011-BLG-262Lb: a sub-earth-mass moon orbiting a gas giant primary or a high velocity planetary system in the galactic bulge. ApJ 785:155. https://doi.org/10.1088/0004-637X/785/2/155, 1312.3951 Borucki WJ, Koch D, Basri G et al (2010) Kepler planet-detection mission: introduction and first results. Science 327:977. https://doi.org/10.1126/science.1185402 Braga-Ribas F, Sicardy B, Ortiz JL et al (2014) A ring system detected around the Centaur (10199) Chariklo. Nature 508:72–75. https://doi.org/10.1038/nature13155, 1409.7259 Brown TM, Charbonneau D, Gilliland RL, Noyes RW, Burrows A (2001) Hubble space telescope time-series photometry of the transiting planet of HD 209458. ApJ 552:699–709. https://doi. org/10.1086/320580, astro-ph/0101336 Cabrera J, Schneider J (2007) Detecting companions to extrasolar planets using mutual events. A&A 464:1133–1138. https://doi.org/10.1051/0004-6361:20066111, astro-ph/0703609 Cameron AGW, Ward WR (1976) The origin of the Moon. In: Lunar and planetary science conference, Houston, vol 7 Canup RM (2005) A giant impact origin of Pluto-Charon. Science 307:546–550. https://doi.org/ 10.1126/science.1106818

39 Detecting and Characterizing Exomoons and Exorings

849

Canup RM, Ward WR (2002) Formation of the Galilean satellites: conditions of accretion. AJ 124:3404–3423. https://doi.org/10.1086/344684 Charbonneau D, Winn JN, Latham DW et al (2006) Transit photometry of the core-dominated planet HD 149026b. ApJ 636:445–452. https://doi.org/10.1086/497959, astro-ph/0508051 Crida A, Charnoz S (2012) Formation of regular satellites from ancient massive rings in the solar system. Science 338:1196. https://doi.org/10.1126/science.1226477, 1301.3808 de Wit J, Wakeford HR, Gillon M et al (2016) A combined transmission spectrum of the Earthsized exoplanets TRAPPIST-1 b and c. Nature 537:69–72. https://doi.org/10.1038/nature18641, 1606.01103 Deienno R, Yokoyama T, Nogueira EC, Callegari N, Santos MT (2011) Effects of the planetary migration on some primordial satellites of the outer planets. I. Uranus’ case. A&A 536:A57. https://doi.org/10.1051/0004-6361/201014862 Deienno R, Nesvorný D, Vokrouhlický D, Yokoyama T (2014) Orbital perturbations of the Galilean satellites during planetary encounters. AJ 148:25. https://doi.org/10.1088/0004-6256/148/2/25, 1405.1880 Domingos RC, Winter OC, Yokoyama T (2006) Stable satellites around extrasolar giant planets. MNRAS 373:1227–1234. https://doi.org/10.1111/j.1365-2966.2006.11104.x Han C, Han W (2002) On the feasibility of detecting satellites of extrasolar planets via microlensing. ApJ 580:490–493. https://doi.org/10.1086/343082, astro-ph/0207372 Heising MZ, Marcy GW, Schlichting HE (2015) A search for ringed exoplanets using Kepler photometry. ApJ 814:81. https://doi.org/10.1088/0004-637X/814/1/81, 1511.01083 Heller R (2014) Detecting extrasolar Moons akin to solar system satellites with an orbital sampling effect. ApJ 787:14. https://doi.org/10.1088/0004-637X/787/1/14, 1403.5839 Heller R (2016) Transits of extrasolar moons around luminous giant planets. A&A 588:A34. https://doi.org/10.1051/0004-6361/201527496, 1603.00174 Heller R, Albrecht S (2014) How to determine an exomoon’s sense of orbital motion. ApJ 796:L1. https://doi.org/10.1088/2041-8205/796/1/L1, 1409.7245 Heller R, Marleau GD, Pudritz RE (2015) The formation of the Galilean moons and Titan in the Grand Tack scenario. A&A 579:L4. https://doi.org/10.1051/0004-6361/201526348, 1506.01024 Heller R, Hippke M, Jackson B (2016a) Modeling the orbital sampling effect of extrasolar moons. ApJ 820:88. https://doi.org/10.3847/0004-637X/820/2/88, 1603.07112 Heller R, Hippke M, Placek B, Angerhausen D, Agol E (2016b) Predictable patterns in planetary transit timing variations and transit duration variations due to exomoons. A&A 591:A67. https:// doi.org/10.1051/0004-6361/201628573, 1604.05094 Hippke M (2015) On the detection of exomoons: a search in Kepler data for the orbital sampling effect and the scatter peak. ApJ 806:51. https://doi.org/10.1088/0004-637X/806/1/ 51, 1502.05033 Jacobson SA, Morbidelli A (2014) Lunar and terrestrial planet formation in the grand tack scenario. Philos Trans R Soc Lond Ser A 372:0174. https://doi.org/10.1098/rsta.2013.0174, 1406.2697 Kenworthy MA, Mamajek EE (2015) Modeling giant extrasolar ring systems in eclipse and the case of J1407b: sculpting by exomoons? ApJ 800:126. https://doi.org/10.1088/0004-637X/800/ 2/126, 1501.05652 Kipping DM (2009a) Transit timing effects due to an exomoon. MNRAS 392:181–189. https://doi. org/10.1111/j.1365-2966.2008.13999.x, 0810.2243 Kipping DM (2009b) Transit timing effects due to an exomoon – II. MNRAS 396:1797–1804. https://doi.org/10.1111/j.1365-2966.2009.14869.x, 0904.2565 Kipping DM (2011) LUNA: an algorithm for generating dynamic planet-moon transits. MNRAS 416:689–709. https://doi.org/10.1111/j.1365-2966.2011.19086.x, 1105.3499 Kipping DM, Fossey SJ, Campanella G (2009) On the detectability of habitable exomoons with Kepler-class photometry. MNRAS 400:398–405. https://doi.org/10.1111/j.1365-2966. 2009.15472.x, 0907.3909 Kipping DM, Bakos GÁ, Buchhave L, Nesvorný D, Schmitt A (2012) The hunt for exomoons with Kepler (HEK). I. Description of a new observational project. ApJ 750:115. https://doi.org/10. 1088/0004-637X/750/2/115, 1201.0752

850

R. Heller

Kipping DM, Schmitt AR, Huang X et al (2015) The hunt for exomoons with Kepler (HEK): V. A survey of 41 planetary candidates for exomoons. ApJ 813:14. https://doi.org/10.1088/0004637X/813/1/14, 1503.05555 Levison HF, Dones L, Chapman CR et al (2001) Could the lunar “Late Heavy Bombardment” have been triggered by the formation of Uranus and Neptune? Icarus 151:286–306. https://doi.org/ 10.1006/icar.2001.6608 Lewis KM, Sackett PD, Mardling RA (2008) Possibility of detecting Moons of pulsar planets through time-of-arrival analysis. ApJ 685:L153–L156. https://doi.org/10.1086/592743, 0805.4263 Lewis KM, Ochiai H, Nagasawa M, Ida S (2015) Extrasolar binary planets II: detectability by transit observations. ApJ 805:27. https://doi.org/10.1088/0004-637X/805/1/27, 1504.06365 Liebig C, Wambsganss J (2010) Detectability of extrasolar moons as gravitational microlenses. Astron Astrophys 520:A68, 13 pp Maciejewski G, Dimitrov D, Neuhäuser R et al (2010) Transit timing variation in exoplanet WASP3b. MNRAS 407:2625–2631. https://doi.org/10.1111/j.1365-2966.2010.17099.x, 1006.1348 Mamajek EE, Quillen AC, Pecaut MJ et al (2012) Planetary construction zones in occultation: discovery of an extrasolar ring system transiting a young sun-like star and future prospects for detecting eclipses by circumsecondary and circumplanetary disks. AJ 143:72. https://doi.org/ 10.1088/0004-6256/143/3/72, 1108.4070 Marois C, Macintosh B, Barman T et al (2008) Direct imaging of multiple planets orbiting the star HR 8799. Science 322:1348. https://doi.org/10.1126/science.1166585, 0811.2606 Montalto M, Gregorio J, Boué G et al (2012) A new analysis of the WASP-3 system: no evidence for an additional companion. Mon Not R Astron Soc 427(4):2757–2771. https://doi.org/10. 1111/j.1365-2966.2012.21926.x, http://mnras.oxfordjournals.org/content/427/4/2757.abstract, http://mnras.oxfordjournals.org/content/427/4/2757.full.pdf+html Morbidelli A, Tsiganis K, Batygin K, Crida A, Gomes R (2012) Explaining why the uranian satellites have equatorial prograde orbits despite the large planetary obliquity. Icarus 219:737– 740. https://doi.org/10.1016/j.icarus.2012.03.025, 1208.4685 Moskovitz NA, Gaidos E, Williams DM (2009) The effect of lunarlike satellites on the orbital infrared light curves of Earth-analog planets. Astrobiology 9:269–277. https://doi.org/10.1089/ ast.2007.0209, 0810.2069 Noyola JP, Satyal S, Musielak ZE (2014) Detection of exomoons through observation of radio emissions. ApJ 791:25. https://doi.org/10.1088/0004-637X/791/1/25, 1308.4184 Noyola JP, Satyal S, Musielak ZE (2016) On the radio detection of multiple-exomoon systems due to plasma torus sharing. ApJ 821:97. https://doi.org/10.3847/0004-637X/821/2/97, 1603.01862 Ohta Y, Taruya A, Suto Y (2009) Predicting photometric and spectroscopic signatures of rings around transiting extrasolar planets. ApJ 690:1–12. https://doi.org/10.1088/0004-637X/690/1/ 1, astro-ph/0611466 Pál A (2012) Light-curve modelling for mutual transits. MNRAS 420:1630–1635. https://doi.org/ 10.1111/j.1365-2966.2011.20151.x, 1111.1741 Peters MA, Turner EL (2013) On the direct imaging of tidally heated exomoons. ApJ 769:98. https://doi.org/10.1088/0004-637X/769/2/98, 1209.4418 Pollack JB, Reynolds RT (1974) Implications of Jupiter’s early contraction history for the composition of the galilean satellites. Icarus 21:248–253. https://doi.org/10.1016/0019-1035(74)900402 Pont F, Gilliland RL, Moutou C et al (2007) Hubble space telescope time-series photometry of the planetary transit of HD 189733: no Moon, no rings, starspots. A&A 476:1347–1355. https:// doi.org/10.1051/0004-6361:20078269, 0707.1940 Rauer H, Catala C, Aerts C et al (2014) The PLATO 2.0 mission. Exp Astron 38:249–330. https:// doi.org/10.1007/s10686-014-9383-4, 1310.0696 Robinson TD (2011) Modeling the infrared spectrum of the earth-Moon system: implications for the detection and characterization of Earthlike extrasolar planets and their Moonlike companions. ApJ 741:51. https://doi.org/10.1088/0004-637X/741/1/51, 1110.3744

39 Detecting and Characterizing Exomoons and Exorings

851

Rosenblatt P, Charnoz S, Dunseath KM et al (2016) Accretion of phobos and Deimos in an extended debris disc stirred by transient moons. Nat Geosci 9(8):581–583. https://doi.org/10. 1038/ngeo2742 Rufu R, Aharonson O, Perets HB (2017) A multiple-impact origin for the moon. Nature Geosci. Advance online publication. https://doi.org/10.1038/ngeo2866 Samsing J (2015) Extracting periodic transit signals from noisy light curves using fourier series. ApJ 807:65. https://doi.org/10.1088/0004-637X/807/1/65, 1503.03504 Santos NC, Martins JHC, Boué G et al (2015) Detecting ring systems around exoplanets using high resolution spectroscopy: the case of 51 Pegasi b. A&A 583:A50. https://doi.org/10.1051/ 0004-6361/201526673, 1509.00723 Sartoretti P, Schneider J (1999) On the detection of satellites of extrasolar planets with the method of transits. A&AS 134:553–560. https://doi.org/10.1051/aas:1999148 Sato M, Asada H (2009) Effects of mutual transits by extrasolar planet-companion systems on light curves. PASJ 61:L29. 0906.2590 Sengupta S, Marley MS (2016) Detecting exomoons around self-luminous giant exoplanets through polarization. ApJ 824:76. https://doi.org/10.3847/0004-637X/824/2/76, 1604.04773 Simon A, Szatmáry K, Szabó GM (2007) Determination of the size, mass, and density of “exomoons” from photometric transit timing variations. A&A 470:727–731. https://doi.org/10. 1051/0004-6361:20066560, 0705.1046 Simon AE, Szabó GM, Szatmáry K, Kiss LL (2010) Methods for exomoon characterization: combining transit photometry and the Rossiter-McLaughlin effect. MNRAS 406:2038–2046. https://doi.org/10.1111/j.1365-2966.2010.16818.x Simon AE, Szabó GM, Kiss LL, Szatmáry K (2012) Signals of exomoons in averaged light curves of exoplanets. MNRAS 419:164–171. https://doi.org/10.1111/j.1365-2966.2011.19682. x, 1108.4557 Simon A, Szabó G, Kiss L, Fortier A, Benz W (2015) CHEOPS performance for exomoons: the detectability of exomoons by using optimal decision algorithm. PASP 127:1084–1095. https:// doi.org/10.1086/683392, 1508.00321 Skowron J, Udalski A, Szyma´nski MK et al (2014) New method to measure proper motions of microlensed sources: application to candidate free-floating-planet event MOA-2011-BLG-262. ApJ 785:156. https://doi.org/10.1088/0004-637X/785/2/156, 1312.7297 Spahn F, Schmidt J, Albers N et al (2006) Cassini dust measurements at Enceladus and implications for the origin of the E ring. Science 311:1416–1418. https://doi.org/10.1126/science.1121375 Szabó GM, Szatmáry K, Divéki Z, Simon A (2006) Possibility of a photometric detection of “exomoons”. A&A 450:395–398. https://doi.org/10.1051/0004-6361:20054555, astro-ph/0601186 Szabó R, Szabó GM, Dálya G et al (2013) Multiple planets or exomoons in Kepler hot Jupiter systems with transit timing variations? A&A 553:A17. http://doi.org/10.1051/ 0004-6361/201220132, 1207.7229 Tusnski LRM, Valio A (2011) Transit model of planets with Moon and ring systems. ApJ 743:97. https://doi.org/10.1088/0004-637X/743/1/97, 1111.5599 Williams DM, Knacke RF (2004) Looking for planetary Moons in the spectra of distant Jupiters. Astrobiology 4:400–403. https://doi.org/10.1089/ast.2004.4.400 Zhuang Q, Gao X, Yu Q (2012) The Rossiter-McLaughlin effect for exomoons or binary planets. ApJ 758:111. https://doi.org/10.1088/0004-637X/758/2/111, 1207.6966 Zuluaga JI, Kipping DM, Sucerquia M, Alvarado JA (2015) A novel method for identifying exoplanetary rings. ApJ 803:L14. https://doi.org/10.1088/2041-8205/803/1/L14, 1502.07818

Section V Ground-Based Instrumental Projects for Exoplanet Research Norio Narita

Norio Narita is an Assistant Professor in the Department of Astronomy at the University of Tokyo and a PRESTO researcher of the Japan Science and Technology Agency. He is an affiliated member of the National Astronomical Observatory of Japan and the Astrobiology Center, Japan. He is dedicated to observational studies of transiting exoplanetary systems. Narita is also PI of the MuSCAT and MuSCAT2 multi-channel CCD imager and photometer for studies of transiting exoplanets.

High-Precision Spectrographs for Exoplanet Research: CORAVEL, ELODIE, CORALIE, SOPHIE, and HARPS

40

Francesco Pepe, François Bouchy, Michel Mayor, and Stéphane Udry

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CORAVEL: The Beginning of Spectrovelocimetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Historical Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results, Precision Limitations, Lessons Learned . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ELODIE and CORALIE: Planet Discovery by Numerical Cross-Correlation . . . . . . . . . . . . . Technical Improvements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The ELODIE and CORALIE Sisters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Science Programmes and Results with ELODIE and CORALIE . . . . . . . . . . . . . . . . . . . . . HARPS: Setting New Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A Strategic Decision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Building on Experience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Achieving 1 ms1 Precision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Design Choices and Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results and Achievements of HARPS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SOPHIE: An Extension to the Northern Hemisphere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rationale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Spectrograph Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Main Science Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

856 857 857 858 859 859 862 864 865 865 866 867 869 872 873 873 874 878

F. Pepe () Département d’Astronomie, Observatoire de l’Université de Genéve, Versoix, GE, Switzerland e-mail: [email protected] F. Bouchy Département d’Astronomie, Université de Genéve, Versoix, GE, Switzerland Observatoire astronomique de l’Université de Genéve, Versoix, Switzerland LAM/OHP, Marseille, France e-mail: [email protected] M. Mayor · S. Udry Département d’Astronomie, Université de Genéve, Versoix, GE, Switzerland e-mail: [email protected]; [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_190

855

856

F. Pepe et al.

Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

878

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

879

Abstract

In this chapter, we will present the concepts, designs, and performances of several generations of simultaneous-reference spectrographs that have made the history of exoplanet discoveries by means of the Doppler technique. It is not possible to understand the strengths of this category of spectrographs without understanding first the revolutionary approach of CORAVEL that set the path for the next generation of instruments using CCDs, i.e., ELODIE and CORALIE. These instruments were extremely successful (e.g., with the discovery of 51 Peg b), it however quickly became clear that higher precision would be necessary in order not to remain stuck with Jupiter- and Saturn-mass planets only. The ELODIE/CORALIE concept was optimized in order to reach best possible performances. This process led to the development of HARPS that has become, through the tens of discoveries of Super-Earths and Neptunes, the new reference for high-precision Doppler velocity measurements. The success of HARPS had to be transposed to the northern hemisphere, an ambition that resulted first in the manufacturing of SOPHIE and later of HARPS-N. This work was finally the baseline for the development of ESPRESSO aiming at the next level of precision. We refer, however, to the chapter dedicated to ESPRESSO in this same handbook for a detailed description.

Introduction The pace of early history of exoplanet discoveries was dictated by the Doppler (or Radial-Velocity) technique. A review of past, present, and future spectrographs dedicated to the search for exoplanets is given by Pepe et al. (2014). The successful spectrographs, although of quite different design, can be split into two major categories: The “self-calibrated” and the “simultaneous reference” spectrographs. In the present chapter, we shall focus on this second category. Its concept is based on the ability of a stable spectrograph to measure in the most direct way a spectral line position on the detector, convert it into wavelength, and determine the radialvelocity of the astrophysical object by comparing the measured wavelength with the rest-frame wavelength in the solar barycenter. When aiming at ms1 precision, the motion of the spectral line in the focal plane of the spectrograph becomes so tiny that it is actually comparable or eventually much smaller than any physical motion of the detector with regard to the spectrum. A method must, therefore, be introduced, in addition to the initial wavelength calibration, to track these instrumental drifts. In the case of the “simultaneous reference” technique, this objective is achieved by projecting, at any time, a laboratory reference spectrum (usually a spectral standard) into the focal/spectral plane simultaneously with the calibration or the astrophysical spectrum. This allows to measure possible motions of the focal plane with regard to the detector and determine instrumental drifts, which can eventually be “subtracted” from the radial-velocity measurement.

40 High-Precision Spectrographs for Exoplanet Research: CORAVEL,: : :

857

In the following sections, we present several generations of “simultaneousreference spectrographs.” They are all based on the original CORAVEL concept, which will be introduced in order to develop the historical understanding of the instrumental evolution. We will then describe ELODIE – by which 51 Peg b (Mayor and Queloz 1995) was discovered – and its twin spectrograph CORALIE. We won’t miss the occasion to describe the HARPS(N) spectrograph, which has been crucial in triggering the era of “super-Earths” discoveries, as well as SOPHIE as its extension to the northern hemisphere. It is important to note that these spectrographs have contributed to a significant fraction of exoplanet discoveries (about 400 in total to date). The various sections will therefore also mention examples of the most important achievements of these spectrographs.

CORAVEL: The Beginning of Spectrovelocimetry Historical Background In a stellar spectrum, the Doppler information is distributed over several thousand of spectral lines. A large number of absorption lines are present over a broad spectral domain for the star of different spectral types, although some differences occur with varying atmospheric parameters. Felgett (1953) was first to suggest an instrumental setup to concentrate all this information into a single “signal” in order to extract the Doppler information: It would be “sufficient” to project the stellar spectrum on a physical binary mask/template to produce a cross-correlation signal that would achieve a minimum whenever the absorption lines would match the “holes” in the binary template. The transmitted signal is recorded as a function of the template position by a single detector. The so-obtained cross-correlation function (CCF) reproduces the “average” absorption line and its minimum will occur for the template position corresponding to the radial velocity of the star. A first Coudé instrument (Griffin 1967) demonstrated the feasibility and huge efficiency of this approach compared to the classical radial-velocity measurements using photographic plates. An evolution of this concept was achieved with CORAVEL (Baranne et al. 1979, Fig. 1, Table 1). It is a cross-correlation spectrometer with two significant improvements: On the one hand, a compact crossdispersed spectrograph using an echelle grating that allowed to cover a much larger spectral domain at high spectral resolution;on the other hand, a fully automatic control of the template scanning and a real-time access to the cross-correlation function and thus to the stellar radial velocity. If the position of the CCF allows for an efficient determination of the stellar radial velocity, the CCF itself permits the estimation of several other observables as well. For example, the area of the CCF (i.e., its equivalent width) at a given stellar temperature is a superb estimator of the metallicity of the observed star (Mayor 1980), a method broadly used with all subsequent cross-correlation spectrometers. The width of the CCF is furthermore a precise proxy for the stellar rotation (Benz and Mayor 1981, 1984).

858

F. Pepe et al.

Fig. 1 The CORAVEL instrument at the Cassegrain focus of the 1-m Swiss Telescope at Observatoire de Haute-Provence (OHP)

Table 1 Main characteristics of CORAVEL Parameters Telescope

Specification 1-m class

Feed Design

Slit spectrograph Cross-dispersed echelle spectrograph Photomultiplier tube C physical mask 360–520 nm 200 000 NA 300 m s1

Detector Spectral domain Resolving power R Efficiency RV precision

Comments CORAVEL-N was installed at the Cassegrain focus of the 1-m Swiss telescope at the Haute-Provence Observatory (1977); CORAVEL-S was installed at the Cassegrain focus of the 1.5 m Danish telescope at the ESO-La Silla Observatory (1981)

Results, Precision Limitations, Lessons Learned CORAVEL has been a revolutionary instrument in terms of stellar physics, kinematics, and the exploration of binary systems. In this chapter, it has been described, however, as the precursor of future spectrographs optimized for and dedicated to the search of extrasolar planets through precise Doppler measurements rather than an instrument that contributed to the development of the exoplanet science in a direct way. Nevertheless, it must be recalled that CORAVEL allowed (in combination with Oak Ridge measurements) for the determination of the orbit of the companion of HD 114762 (Latham et al. 1989), a companion that has a minimum mass of 11 MJ at the interface between giant planets and brown dwarfs. According to the planet-mass distribution known today and depending on its actual mass, HD 114762 b could be considered as the first planetary companion to a solar-type star.

40 High-Precision Spectrographs for Exoplanet Research: CORAVEL,: : :

859

CORAVEL certainly had the capability of detecting high-mass exoplanets on short periods in the tail of the planet-mass distribution known today. However, before the discovery of the surprising properties of the population of giant gaseous planets, high-mass exoplanets were not expected (limited accretion of gas on a forming giant planet) and, moreover, we know now that these objects seem to be actually rare (decreasing distribution of the mass function towards high masses). Adding to this the limit of the radial-velocity observations allowing only for the estimate of a minimum mass for the companion, exoplanets were not the primary targets for CORAVEL. CORAVEL had two main instrumental limitations that prevented radial-velocity measurements to be more precise than 300 ms1 . First, it is a slit spectrometer; sub-arcsec errors on guiding translate in the most direct way into radial-velocity errors at the level of a fraction of km s1 . Second, the built-in physical template and the observed stellar spectrum do not always perfectly match, especially considering varying temperature and air pressure of the spectrograph, that furthermore was not of the stabilized type. With the advent of CCDs and optical fibers, new opportunities appeared to remove these limitations. The era of a new generation of spectrovelocimeters was about to start.

ELODIE and CORALIE: Planet Discovery by Numerical Cross-Correlation Technical Improvements The success of the two CORAVEL instruments has clearly demonstrated that the cross-correlation approach was providing an efficient way of extracting the Doppler information from stellar spectra. Also, as mentioned in the previous section, the limitations of the instruments were fairly well understood and new technology developments allowed for the design of new instruments based on the same basic principles for precise velocity measurements while incorporating these new features. CCD detectors and numerical cross-correlation. With CORAVEL, the crosscorrelation between the stellar spectrum and a binary mask was performed mechanically within the instrument, sweeping at high frequency through a predefined set of mask “positions” (channels), building in this way the CCF in real time by measuring the integrated counts in each of the visited channels using a photomultiplier. A major improvement of this approach was possible thanks to the advent of Charge-Coupled Device (CCD) detectors. It was then possible to record the stellar spectrum on the CCD and then perform the cross-correlation numerically, in a subsequent stage. The so-recorded spectra can also be used for spectroscopic analysis and the estimate of important stellar parameters (effective temperature, gravity, metallicity, individual element abundances, etc.). Concerning the measurement of precise radial velocities, the new approach gives the possibility to use optimal numerical templates for the correlation, matching better the spectra of the stars observed. The single physical template in

860

F. Pepe et al.

CORAVEL was built from the spectrum of Arcturus, a giant K1.5 star. The measurement efficiency on stars of different spectral types (effective temperatures) decreases with increasing mismatch between the stellar spectrum and the numerical template (different sets and of spectral lines and different widths and depths). That was, e.g., a clear limit for M dwarfs, in addition to their intrinsic faintness. The use of CCDs also pushed for the development of spectrographs at higher spectral resolution. At a resolving power of R D 20,000 (CORAVEL), many stellar spectral lines are blended due to instrumental broadening. An increase in resolving power allows for a better separation of the lines and a better determination of their position on the CCD due to better sampling and reduced photonic error (see also the section “HARPS: Setting New Standards”). Optical fibers. Another severe limitation of CORAVEL was the use of a slit at the entrance of the spectrograph. Due to atmospheric turbulence, seeing conditions, and varying telescope-instrument flexures, the image of the star on the entrance slit moves and results in a motion of the photocenter of the monochromatic slit image in the focal plane of the spectrograph. The observed shift directly translates into a shift of the stellar spectrum on the detector, mixing up spatial displacement and spectral shift (e.g., due to the Doppler effect). The shift on the CCD will change from one exposure to the next and introduce so a variation of the measured radial velocity up to hundreds of ms1 . The use of optical fibers to feed the instrument allows, thanks to their scrambling effect, to produce a uniform and stable illumination of the spectrograph slit, increasing so the photocenter stability of the spectral lines on the detector to the level of a few ms1 . However, optical fibers scramble only the near-field to a satisfactory level, while the far-field is left almost unchanged. In order to increase the scrambling effect even further, a double-scrambler device is therefore often introduced (Brown 1990). The double scrambler is composed of two fiber sections linked together by an optical device exchanging near- and far-field, thus ensuring that the second fiber scrambles the original (unscrambled) far-field of the first fiber section. Calibration and stability. In order to achieve precise radial-velocity measurements, an important issue is to develop a calibration system allowing for the determination of the zero velocity point of the instrument for each measurement, reliable over long periods of time. With CORAVEL, this velocity calibration was obtained by the measurements of a hollow cathode iron lamp before and after the observation of a star. The interpolation between the zero velocities of the two measurements of the lamp gives then the position of the zero at the time of the star observation. Importantly, the measurement of the spectrum of the lamp allows as well for the determination of the wavelength calibration, i.e., the correspondence between the pixel and wavelength positions. This approach is not sufficient anymore when aiming at high-precision radial velocities. Changes of the local astroclimatic conditions will induce spectral shifts on the CCD, mainly due to changes of the refraction index in the spectrograph. The local astroclimatic conditions impact as well the instrumental profile (IP) of the spectrograph, i.e., the way the photoelectrons of a line are distributed over the

40 High-Precision Spectrographs for Exoplanet Research: CORAVEL,: : :

861

CCD pixels, and thus the estimated radial velocity. Two different methods have been developed to overcome this difficulty: • Self-calibration: The visible light from the star is passing through a gas cell (historically HF, later I2 ) whose absorption spectrum is then superimposed to the stellar one, providing a built-in reference wavelength calibration and radialvelocity measurement (Campbell and Walker 1979; Brown 1990; Butler et al. 1996). This approach is perfectly adapted to existing (nonstabilized) echelle spectrographs directly mounted to the telescope. Any distortion, drift, thermomechanical and optical effect, etc. will affect the IP of the spectrograph such that both stellar spectrum and spectral reference will “see” the same changes. In principle, it is possible to remove perfectly instrumental effects on the stellar spectrum by measuring the IP on the spectral reference. In practice, however, the limited wavelength range covered by the gas spectrum, its absorption, and the deconvolution process will significantly reduce the optical efficiency of the method. Furthermore, nonperfect knowledge of the high-resolution spectra of the spectral reference and the reference-free stellar spectrum may introduce systematic effects at the ms1 -level.pag \enlargethispage*f-6ptg?> • Simultaneous reference: Building on a stabilized spectrograph, the goal is to avoid as much as possible instrumental drifts and IP variations that may impact the stellar spectrum. Since, however, this can not be guaranteed at the level of millipixels on all timescales, a way is developed to track the residual instrumental drifts. For this purpose, two fibers feed the spectrograph simultaneously, one carrying the light of the observed star and the other the light from the spectral reference, usually a Thorium-Argon lamp (Brown 1990; Baranne et al. 1996). The two fibers at the entrance of the spectrograph are located as close together as possible to ensure almost identical optical path for both light beams, and they are organized such to produce on the CCD detector a series of interlaced spectral orders, the orders n of the lamp located in-between orders n and n-1 of the stellar spectrum, and vice versa (Fig. 2). The previously determined (typically Fig. 2 Raw frame of a simultaneous-reference exposure showing the spectral orders of the stellar spectrum (horizontal lines) interlaced with the thorium-argon lamp spectrum (individual emission lines)

862

F. Pepe et al.

at the start of the night) wavelength solution that links together the wavelength with the pixel coordinate can be corrected, before being applied to the scientific observation, by the drift value determined by the shift of the reference spectrum between the calibration exposure and the scientific exposure. The advantages of this solution are that the optical throughput is conserved over the entire wavelength range and that complex deconvolution techniques are avoided. The disadvantage consists in the fact that the spectrograph must be conceived, right from the beginning, to be fiber-fed and as stable as possible, in order to avoid second-order instrumental effects. The method can therefore not be applied to any spectrograph.

The ELODIE and CORALIE Sisters The ELODIE spectrometer (also known as “Super-Coravel”) was considered in 1988 as a CORAVEL-derived instrument that would use available time of the 1.93-m telescope of Observatoire de Haute-Provence (OHP) during a full moon. ELODIE rapidly became one of the centerpieces of OHP in the 1990s. Designed by the OHP technical services in collaboration with André Baranne of the Observatoire de Marseille and with Michel Mayor of the Geneva Observatory, it was assembled in a laboratory and then tested on the sky in 1993. ELODIE was a fiber-fed crossdispersed echelle spectrograph and was located in a temperature-controlled room in the first floor of the 1.93-m telescope building and covered with a double protection to improve its thermal stability. It was connected to the telescope by 20 m of a pair of optical fibers, from the front-end attached to the Cassegrain focus of the 1.93 m telescope. Two focal-plane apertures were available (both 2 arc-sec wide), one of which was used for starlight and the other for either the sky background or the wavelength calibration lamp, but could also be masked. The spectrograph used a 420  100 mm R4 echelle grating. The optical design was based on an asymmetric white-pupil mount, which allowed controlling the collimated beam size. In particular, the asymmetric design reduced the cross-disperser and camera-lens dimensions, therefore keeping the instrument compact and relatively cheap. The spectra covered a 3000 Å wavelength range (3850–6800 Å) with a spectral resolution of R D 42,000. The instrument was entirely computer-controlled and a standard data-reduction pipeline automatically processed the data from CCD readout, through extraction and wavelength calibration, to numerical crosscorrelation and radial-velocity determination. The instrument is described in details by Baranne et al. (1996) and its characteristics are summarized in Table 2. The first ELODIE spectrum was taken on June 1, 1993, during the period of testing which lasted until the end of 1993. The instrument was opened to the community in May 1994. It was with ELODIE that the first planet around a star other than the Sun was discovered around 51 Pegasi (Mayor and Queloz 1995). ELODIE worked until August 10, 2006, when its detector controller electronics broke down. It was quickly replaced by SOPHIE, then in its final phase of integration. ELODIE is today exposed and visible to visitors on the ground floor of the 1.93-m building at OHP (Fig. 3).

40 High-Precision Spectrographs for Exoplanet Research: CORAVEL,: : :

863

Table 2 Main characteristics of ELODIE/CORALIE Parameters Telescope

Specification 1–2 m class

Feed

Fiber-fed spectrograph

Design

Cross-dispersed echelle spectrograph 1k1/2k2 CCD 2/3.3 383–690 nm 42,000/50,000(60,000) 1.5%/1.5%(6%) 15/7(3) m s1

Detector Sampling Spectral domain Resolving power R Total efficiency RV precision

Comments ELODIE was installed on the 1.93-m telescope at the Haute-Provence Observatory (1994); CORALIE was installed on the 1.2-m Swiss Euler telescope at the ESO-La Silla Observatory (1998) Equipped with a double scrambler later on. CORALIE was equipped with octagonal fibers in 2014

ELODIE/CORALIE Pixels per FWHM of unresolved line ELODIE/CORALIE (after upgrade 2007) ELODIE/CORALIE (after upgrade 2007) ELODIE/CORALIE (after upgrade 2014)

Fig. 3 Left: Picture of the ELODIE spectrograph. Right: Prisms-based cross-disperser of CORALIE after the 2007 upgrade

Soon after the discovery of 51 Peg b, an improved copy of the instrument, named CORALIE, was built for the 1.2-m Euler Swiss Telescope of the ESOLa Silla Observatory (Chile). The CORALIE front-end adaptor is located at the Nasmyth focus of the Euler telescope. A set of two fibers feeds the spectrograph that is located in an isolated and temperature controlled room. Thanks to a slightly

864

F. Pepe et al.

different optical combination at the entrance of the spectrograph and the use of a 2 k by 2 k CCD detector with smaller pixels (15 microns), CORALIE achieved a larger resolving power (R D 50,000) than ELODIE. The many improvements carried out in the thermal control and the resolution of the instrument, as well as in the reduction software, yielded a factor two improved radial-velocity precision compared to ELODIE. Operations of CORALIE started in 1998. Several upgrades were performed on the instrument since then: The original cross-disperser made of a combination of prism and grism was replaced by a single prism in 2007. At the same occasion, the input focal ratio of the beam entering the fiber was adapted and a higher resolving power obtained. In 2014, the fiber link (including the scrambler) was entirely replaced using octagonal fibers. All these upgrades helped considerably improving the optical efficiency (by a factor of 4!) and the instrumental precision (from typically 6 to 3 ms1 ). In 2015, CORALIE was eventually equipped with a Fabry-Pérot calibrator that allows replacing the thorium lamp in the simultaneousreference process.

Science Programmes and Results with ELODIE and CORALIE A radial-velocity survey for northern extrasolar planets led by M. Mayor was conducted from 1994 to 2004 on ELODIE. After the success of the detection of 51 Peg b (Mayor and Queloz 1995), the sample of solar-type stars originally including 142 stars was progressively extended to 372 stars (Perrier et al. 2003). In total, 19 exoplanets were found as part of this early programme, including the planet found to be also transiting HD 209458 (Charbonneau et al. 2000; Mazeh et al. 2000; Henry et al. 2000). In 2004, a new exoplanet research programme was initiated by S. Udry, targeting metal-rich stars expected to have a higher probability of hosting planets. The aim of this project was to improve the statistical studies of planetary parameters, but also to efficiently detect short-period giant planets, ideal candidates for the detection of photometric transits. The strategy consisted in selecting a sample of 1060 dwarf stars of spectral type F8 to M0 with magnitude V < 8.5 and vsini 80,000) and a stable guiding (D centering of the stellar image on the fiber tip) to better than 0.05 arcsec. 3. High spectral resolution: The calibration of a spectrograph is performed using sources of spectral reference (emission and absorption) lines of high stability. Knowing the nominal wavelength of the reference, it can be associated with the pixel coordinate of the detector at which the photocenter of the line is recorded. The function describing the relation between the wavelength and the pixel coordinate is called wavelength solution. The quality of the wavelength solution depends also on the detection process and eventually on how precisely a reference line is sampled and reproduced. Supposing unresolved spectral lines, the photocenter of an individual line will be determined with a precision increasing with resolving power. The relation between precision and resolving power has been studied and described in several occasions (Butler et al. 1996; Bouchy et al. 2001; Pepe et al. 2014). This can be intuitively understood in the way that the error on the line position stays, for a fixed number of photons, in relative terms to the line width. Furthermore, given the fact that the “signal” is given by the line depths, it is also natural that the signal increases with resolving power. When increasing the resolving power and ensuring a sufficient pixel sampling of the line, each pixel will furthermore represent a smaller chunk of wavelength. As a consequence, any geometry, extra- and intra-pixel response non-uniformity, charge-transfer effects, etc. will eventually result in a smaller effect in terms of wavelength and thus radial velocity. For all these reasons, HARPS was specified to have a resolving power of at least R D 80,000 and a minimum sampling (pixels per FWHM of an unresolved spectral line) of

40 High-Precision Spectrographs for Exoplanet Research: CORAVEL,: : :

869

three pixels. This latter requirement resulted from a study (Wildi 1998, private communication) showing that three pixels were needed to preserve at least 90% of the information content.

Design Choices and Design The design resulting from the previous conceptual considerations is described in Pepe et al. (2002) and Pepe (2001) and the instrument’s characteristics summarised in Table 3. HARPS is a fiber-fed, cross-dispersed echelle spectrograph using the simultaneous reference technique. The fiber-feed is equipped with a doubled scrambler and a guiding-centering system to improve illumination stability. High instrumental stability is achieved by optimizing the thermomechanical design. In particular, the whole spectrograph (all the optics after the fiber feed) is placed in vacuum. The optical design is based on the simplest possible white-pupil mount (Fig. 5) in totally fixed configuration, i.e., there are no moving parts, degrees of freedom were reduced to the absolute necessary minimum, and no adjustment mechanisms where allowed (alignment by machining and shimming). The structure holding the optics is made of welded high-quality carbon steel, thermally cycled and remachined to remove stresses and obtain the desired mechanical accuracy, and is Nickel-plated for protection against corrosion and oxidation. This choice was preferred with regard to Invar and stainless steel because of shortand long-term structural stability, ratio between thermal conductivity and thermal expansion coefficient (more favorable than stainless steel and similar to Invar), simplicity of machining and costs. Other solutions like granite, aluminum, SiC, etc., were considered but discarded because of “risks” of various nature, for instance lack

Table 3 Main characteristics of HARPS/HARPS-N Parameters Telescope

Specification 4-m class

Feed

Fiber-fed spectrograph

Design

Cross-dispersed echelle spectrograph 4k4 CCD 2/3.3

Detector Sampling (pixels per FWHM) Spectral domain Resolving power R Total efficiency RV precision

383–690 nm 115,000 6%(4%) 0.5–1 m s1

Comments HARPS was installed on the 3–6-m telescope at the ESO-la Silla observatory (1998) Equipped with a double scrambler. HARPS was upgraded to octagonal fibers in may 2015 Under vacuum and thermally controlled Mosaic of two 4k2 CCDs Pixels per FWHM of unresolved line

(HARPS, before upgrade in 2015)

870

F. Pepe et al.

Fig. 5 Isometric view of the optical ray-tracing (left) and of the optomechanical layout of HARPS (right)

of experience and examples, availability and price, complexity of machining, and assembling. This latter point was even more emphasized by the choice of placing the spectrograph in a vertical plane instead of a “table-top” configuration. The whole assembly is mounted thus fully symmetrically with respect to a plane containing the gravity vector (the optical axis remains always in this plane). Furthermore, the main dispersion direction (the direction along which the radial velocity is measured) is perpendicular to this plane: Conceptually, the design being symmetric in this direction, any thermal or structural “instability” should have a zero net effect in terms of radial velocity. In the vertical direction, the cross-dispersion direction, the spectrograph would be more sensitive to such effects but they would not result in a radial-velocity effect, at least not at first order. We have to point out that the “vertical” mechanical layout allowed to mount the echelle grating on its edge (instead of face-up or face-down), which ensured better “stiffness” of the optical axis. This detail became even more important when it was decided to use the largest available “monolithic” R4 echelle gratings available at that time, i.e., the UVES@VLT-like mosaic of 840  208 mm. The choice of this grating became necessary to “minimize” the size and thus the costs of the spectrograph given the aimed spectral resolving power and the 1 arcsec field-of-view of the fiber (see Pepe et al. 2002 for the formula relating these quantities). Both parameters do contribute to increase the photonic precision of the radial-velocity measurement and had to be maximized in order to achieve the scientific goals of HARPS. It must be recalled that the use of an image slicer was methodically discarded because of the high risks of introducing Image Profile (IP) distortions at the only location that would not be tracked by the simultaneous reference fiber spectrum. Cross-dispersion was achieved by a grism, which combined relatively good efficiency and high dispersion. This solution in transmission is again less sensitive to thermomechanical instability and could be designed such that the beam is not deviated, making the spectrograph very compact. The “classical” camera lens produces finally a spectral format perfectly matching the 4k4 detector (mosaic of

40 High-Precision Spectrographs for Exoplanet Research: CORAVEL,: : :

871

Fig. 6 View of the vacuum vessel containing the integrated spectrograph just before final closing. HARPS provided continuous operation from September 2003 to May 2015 without ever being opened

two 2k4 E2V CCDs) for the full 380–690 nm wavelength range and provides an effective sampling of 3.3 pixels. Also, or in particular, because of the thermal sensitivity of the camera focus, a thermal compensation system had to be introduced in the camera design. The aimed thermal stability for the whole spectrograph further contributed to the overall stability of the spectrum on the chip. The detector is mounted inside an ESO standard continuous-flow cryostat that ensures continuous operations and avoids thermal shocks (e.g., in bath cryostat) or vibrations (e.g., in closed-cycle cryocooler). The spectrograph is installed in a vacuum chamber (Fig. 6) inside the CoudéWest room of the 3.6-m telescope. A triple system of temperature controls and the passive insulation by the vacuum vessel provide mK stability of the spectrograph on timescales of a night and of a few tens of mK over longer timescales. The optical fibers feeding the spectrograph are connected to the HARPS Cassegrain Fiber Adapter located in the Cassegrain cage of the telescope. The adapter provides all the functionalities for acquisition, guiding, and injection of the calibration light that eventually follows exactly the same path as the stellar light. Last but not least, we have to mention the HARPS data reduction (DRS or pipeline). HARPS had been conceived right from the beginning as an “experiment” or “facility,” and the DRS had to be part of this “machinery.” A big step had been made from CORAVEL to ELODIE by converting the physical cross-correlation into a numerical cross-correlation. A second big step had to be made: Extraction and reduction had to be improved in order not to lose signal and not to introduce biases. The DRS had to be able to preserve information and determine the average spectral line position with a precision of 1/1000th of the CCD pixel size! Although an excellent level was achieved right at start of operations, it should not be forgotten that a continuous effort of optimization has been conducted over the lifetime of the instrument to always improve the data quality (see, e.g., Lovis and Pepe 2007) and is still continuing (e.g., Coffinet et al. in prep.).

872

F. Pepe et al.

Results and Achievements of HARPS The first on-sky tests conducted during the instrument’s commissioning phase and early GTO period (Mayor et al. 2003) immediately showed that the instrument was delivering the expected efficiency and precision performances. At timescales of a night; on the bright and quiet star alpha Cen B sub-ms1 was demonstrated. The difference to preceding instruments became even more apparent by the “direct” detection of p-modes (RV-modulations seen directly on the RV-timeseries) on several standard G- and K-stars, for which the amplitude and periodicity of the modulation could actually be measured. Still during commissioning, known planetary stars were observed and the radial-velocity curve reproduced with a dispersion of 1.7 ms1 . Less than 1 year after the first light, a first discovery by HARPS of a planet around the star HD 330075 was published (Pepe et al. 2004). The reported dispersion is of 2.0 ms1 , which includes instrumental effects, photon noise, and, of course, possible stellar jitter and potential undetected additional planets, demonstrating a performance never achieved before. The tipping point was however achieved with the discovery of Ara c, a 10.5 M˚ planet on a 9.6 days orbit (Santos et al. 2004a). This was, together with 55 Cancri e (McArthur et al. 2004), the first “super-Earth” – a planet with mass lower than the mass of Neptune – to be discovered. The planet was actually found by “chance” since the primary purpose of the intensive measurement campaign was to study the stellar parameters through asteroseismologic measurements. Being the measurements reproducible at the ms1 level over long-time scales, the slow variation of the nightly averaged stellar velocities could be reconducted to the presence of an extrasolar planet orbiting the star. The fabulous result on Ara c gave further confidence in the precision of HARPS and paved the path to many other discoveries in the following years. Most of the “super-Earths” discovered by means of the Doppler technique have actually been discovered by HARPS. This led very early to the conclusion that low-mass planets are much more frequent than massive planets: Lovis et al. (2009) reported, after a first analysis of the full HARPS data, that at least 30% of the stars harbor at least one planet. Furthermore, they found more evidence for the fact that low-mass planets are very rarely “single” but rather part of multiplanetary systems. We may mention the examples of HD 69830 (Lovis et al. 2006, a trio of Neptunes), Gl 581 (Udry et al. 2007, two super-Earths and one Neptune), HD 10180 (Lovis et al. 2011, 7 planets in a single system), or HD 40307 (Mayor et al. 2009, 3 super-Earths), but the many results and achievements of HARPS are best summarized in the statistical study by Mayor et al. (2011) and Bonfils et al. (2013). We should not miss to mentioning a few very specific cases of very challenging single planets systems and/or very closeby objects that are particularly suitable for follow-up programs, for instance the controversial Earth-mass planet alpha Cen Bb (Dumusque et al. 2012), the 3.6 Earthmass planet in the habitable zone of HD 85512 (Pepe et al. 2011), or, more recently, Proxima Centauri b (Anglada-Escudé et al. 2016). In other cases, HARPS was successfully employed to characterize and measure the mass of transiting planets,

40 High-Precision Spectrographs for Exoplanet Research: CORAVEL,: : :

873

for instance CoRoT-7 (Queloz et al. 2009) or GJ 1214 (Charbonneau et al. 2009). Although some of these planets deserve further investigation, either to confirm their Nature or because they are particularly suitable for follow-up measurement with other techniques, they all have in common tiny, ms1 -level signal which could not be detected without the extreme precision of HARPS. The HARPS concept has been an inspiration for other instruments like SOPHIE (see next section) or its twin in the northern hemisphere, HARPS-N. These instruments together contributed significantly to the search and characterization of extrasolar planets. In addition, they have been the main contributors to the radialvelocity follow-up of planets discovered by the transit technique. HARPS-N, for instance, has significantly contributed in populating the Mass-Radius diagram of low-mass exoplanets with precise mass measurements of candidates previously detected by the Kepler satellite (see for instance Pepe et al. 2013b, Dumusque et al. 2014, Dressing et al. 2015). Another remarkable system is HD 219134, which hosts several super-Earths (Motalebi et al. 2015). After having been first discovered by HARPS-N, subsequent Spitzer observations revealed that the two innermost planets where actually transiting, making HD 219134 the closest system with transiting exoplanets to our Earth. Finally, it should not be forgotten that the precision of HARPS-N is based on an exquisite spectroscopic fidelity. This latter is actually fundamental for a new branch of the extrasolar planet science: spectroscopy of planetary atmospheres. It has been demonstrated by Wyttenbach et al. (2015, 2017) that ground-based transit spectroscopy of exoplanets with HARPS can attain or even surpass the quality achieved with Hubble at wavelength observable from the ground. Another example is 51 Peg b, from which Martins et al. (2015) have detected the reflected light using HARPS (Table 3).

SOPHIE: An Extension to the Northern Hemisphere Rationale Taking into account the limitations of the ELODIE spectrograph (Baranne et al. 1996) and taking benefits from experience gained on HARPS (Pepe et al. 2002; Mayor et al. 2003), the Haute-Provence Observatory (OHP) team undertook in early 2003 the development and building of a new spectrograph called SOPHIE to be operated at the 1.93-m telescope as a northern counterpart of HARPS. The top-level requirements for SOPHIE were to improve the overall optical throughput of ELODIE by a factor of 10, to increase the spectral resolution from 42,000 to 75,000, and to improve long-term spectral stability and radial-velocity precision by a factor of 2 to 3 compared to ELODIE. SOPHIE’s first light was achieved in summer 2006. During the first 10 years of operation, continuous improvements of subsystems were done with the goal to reduce any systematic instrumental effect below 2 ms1 .

874

F. Pepe et al.

Table 4 Main characteristics of SOPHIE Parameters Telescope

Specification 2-m class

Feed

Fiber-fed spectrograph

Design Detector Sampling

Cross-dispersed echelle spectrograph 4k2 CCD 2.7/6.7

Spectral domain Resolving power R Total efficiency RV precision

387–694 nm 750 000/390 000 4.3%/10.4% 1–2 m s1

Comments SOPHIE is installed on the 193 cm telescope at the Haute-Provence Observatory (2006) SOPHIEC was equipped with octagonal fibers in 2011 Dispersive components in a constant pressure tank Pixels per FWHM of unresolved line [HR/HE] HR/HE HR/HE (at 550 nm) HR

Spectrograph Design The entire optical design (Perruchot et al. 2008) was oriented to privilege longterm spectral stability and throughput. To minimize the number of separate optical surfaces, and hence increase the mechanical stability, SOPHIE was conceived as a double-pass Schmidt echelle spectrograph, with a single mirror as both collimator and focuser. Parameters are given in Table 4. Light exiting the fiber link (Fig. 7) is collimated by a spherical mirror and, after folding by a plane mirror, comes through the Schmidt chamber where main dispersion is achieved by the R2 echelle grating, while the cross-dispersion is done with a prism in double pass. The dispersed, returned collimated beams are reflected on the folding mirror and are then focused by the spherical mirror to form through a field lens an image of the spectrum on the CCD behind the pierced plane mirror. Apart from thermal precautions, the keypoint for stability is the encapsulation of the dispersive components in a constant pressure tank of dry Nitrogen sealed by the Schmidt plate. This solution stabilizes the air refractive index and thus removes the sensitivity to atmospheric pressure variations. To avoid any gravity effect, the echelle grating grooves and the cryostat are set vertical. Front End: Starlight is collected at the telescope’s focal plane and feed to the spectrograph through an optical fiber link. Two observation modes are available: the High-Efficiency mode (HE) and the High-Resolution mode (HR), to favor high throughput or better radial velocity precision, respectively. Each mode used two fibers: one for the star and the other (located 1.86 arcmin away) for the sky spectrum or simultaneous-reference lamp exposure. The ELODIE front-end was slightly modified during the first year of operation. The atmospheric dispersion correctors were realigned in October 2008. A new guiding system on the fiber entrance was installed in September 2009 achieving a typical accuracy better than

40 High-Precision Spectrographs for Exoplanet Research: CORAVEL,: : :

875

Fig. 7 Design of the SOPHIE spectrograph

0.3 arcsec. The calibration lamps, previously inside the front end, were transferred to a calibration unit in a stabilized environment in fall 2013. All the control hardware of the front end was replaced in spring 2015. Detector and spectral format: Spectra are recorded by an e2V CCD (4kx2k with 15 m pixel size) cooled at 100 ı C. The detector covers 39 spectral orders for both “star” and “sky” fibers in each mode – HR or HE – from order 50 (680.6 nm to 694.4 nm, dispersion 2.25 Å/mm D 0.034 Å/pixel) to 88 (387.2 nm to 395.5 nm, dispersion 1.27 Å/mm D 0.019 Å/pixel). During the first months of operation, a strong dependency of the radial velocity with the signal-to-noise ratio (SNR) at low flux level was observed, typically below SNR D 70, which corresponds to a flux level of about 600 e per pixel. All the spectral lines were blue-shifted at lower flux. Considering that wavelength decreases with increasing distance to the serial register, it was a clear indication for charge transfer inefficiency (CTI). This effect is independent of the used readout mode (slow or fast readout mode). A CTI software correction (Bouchy et al. 2009b) was thus included in the data reduction pipeline of SOPHIE in order to correct for the charge loss during the readout process as a function of pixel coordinate (distance to the register) and flux level. Back-end environment: The spectrograph is installed in the Coudé room of the 1.93-m telescope building, minimizing the fiber length to 17 m. The whole instrument is mounted on shock absorbers supported by the telescope pillar structure to avoid vibrations. The only moving element, the exposure shutter with an

876

F. Pepe et al.

electromagnetic mechanism, has been cautiously damped to avoid transmission of any vibration through the granite bench. The spectrograph is installed in an isolated and thermally controlled box (itself located inside a two-stage thermalized room at 21 ı C) that reduces daily thermal variations to better than 0.01 ı C. A continuous LN2 -filing system for CCD cooling at 100 ı C was installed in summer 2010 to avoid daily thermal shock and weight change otherwise introduced by the original bath cryostat. A complete upgrade of the thermal control system of the insulated box was performed in spring 2016 in order to reduce the thermal shortcut between optical bench and floor and consequently reduce the residual yearly temperature variations of the instrument. Optical fiber link: Considering the average seeing on the OHP site of 2.5 arcsec, the field-of-view was chosen to be of 3 arcsec on sky. It is defined by stepindex multimode cylindrical optical fibers of 100 m diameter. To limit focal-ratio degradation (FRD) effects, a compromise is done injecting the light into the fibers at focal ratio f/3.6. A double-scrambler (symmetrical doublets arranged to exchange object and pupil spaces) is included in the HR fiber link to homogenize and to stabilize the spectrograph entrance illumination. In high-resolution (HR) mode, the spectrograph is fed by a 40.5 m slit bonded to the output of the 100 m fiber, reaching a spectral resolution of 75,000. In high-efficiency mode (HE), the spectrograph is directly fed by the 100 m fiber with a resolution power of 39,000. Figure 8 schematically describes the fiber link. The first years of operation showed that the RV precision obtained on stable stars was limited to 5–6 ms1 in the best cases. This RV limitation was identified as being mainly caused by the insufficient scrambling of the fiber link and the high sensitivity

Fig. 8 The SOPHIE fiber-link: It is composed from left to right of: (1) SOPHIE’s front-end interface; (2) aperture conversion from f/15 (telescope) to f/3.6 (spectrograph); (3) light transport from the telescope’s focal plane to the spectrograph in the Coudé room; (4) light feed of the spectrograph; (5) additional elements for the HR mode (slit and scramblers); and (6) sections of octagonal fiber inserted in 2011 to improve optical scrambling

40 High-Precision Spectrographs for Exoplanet Research: CORAVEL,: : :

877

of the spectrograph to illumination variations. The seeing effect was identified on the HR mode when a correlation was found between the SNR and the radial velocities on several standard stars observed with the same exposure time. This effect, first described by Boisse et al. (2010), appeared to be the limiting factor in SOPHIE’s precision. In this regime, when the seeing is improving, mainly the center of the fiber entrance is illuminated. Taking into account the double scrambler, the far field at the fiber output is then mainly illuminated on its central part. Ray-tracing simulations show that the instrumental profile (spot diagrams) depends on the pupil illumination of the spectrograph, and demonstrate that varying illumination of the far field projected on the SOPHIE grating induces pseudo radial-velocity changes that are not symmetric along the spectral orders and therefore introduce a net change in the measured stellar radial velocity. In summary, seeing variations directly translate into radial-velocity changes. To significantly improve the Doppler precision, in summer 2011 octagonal fibers of 1.5 m length were introduced on the paths of fiber A and B of the HR mode upfront the double scrambler. In the HE mode, both fiber A and B were cut and reconnected with a section of octagonal fibers of identical average diameter. The technical details about octagonal fibers, their implementation in the SOPHIE fiber link, as well as first tests are presented by Perruchot et al. (2011) and Bouchy et al. (2013), respectively. The flux loss due to the insertion of octagonal fiber was estimated to be 8% and 10% on fiber A of HR and HE mode, respectively. The SOPHIE spectrograph, with these new octagonal fibers, was renamed to SOPHIEC. The implementation of an octagonal fiber in the fiber link of the SOPHIE spectrograph improved the radial-velocity precision by a factor about 6. In December 2012, 1.5 m-long octagonal fibers were inserted into fiber A and B of the HR mode also after of the double scrambler, further improving the illumination stability. The residual limitation of SOPHIE due to its sensitivity to pupil illumination changes is in the current configuration strongly reduced thanks to the octagonal fibers. The typical precision of SOPHIEC is now in the range 1–2 ms1 on standard solar-type stars observed with the HRC mode. Calibration and exposure meter: A tungsten lamp is used to locate the position of the diffraction orders and to measure the flat field. More than one thousand lines of a thorium-argon lamp are used to calibrate the spectra in wavelength with high precision (1–2 ms1 ). Each lamp can illuminate both fibers in each mode (HR or HE). Illumination of the sky fiber by the thorium-argon lamp during a stellar exposure (simultaneous-reference technique) allows correcting the residual instrumental drifts. A short-pass rejection filter has been added in April 2007 in front of the lamp to avoid any background caused by wavelengths above 694.6 nm. Furthermore, the simultaneous-reference beam is optically attenuated in the front depending on the exposure time in order to get a constant flux level. The sky fiber can also be used to analyze simultaneously the sky background or the moonlight contamination. An exposure meter records continuously the flux entering the spectrograph. This information is also used to set the optimum observing time depending on atmospheric conditions. Furthermore, the exposure meter allows us to compute the true mean time of the exposures, which is critical for an accurate correction for the Earth’s barycentric velocity. In fall 2013, a new calibration

878

F. Pepe et al.

unit was developed and installed in a stable and thermal-controlled environment. This calibration unit included a Laser-Driven Light Source (LDLS) replacing the tungsten lamp, two Thorium-Argon Hollow-Cathode lamps. Finally, a Fabry-Pérot etalon illuminated in white light was added in spring 2017. This Fabry-Pérot replaces the thorium lamp for the simultaneous-reference observations and allows improving the wavelength solution.

Main Science Results The SOPHIE performances allowed for a dramatic extension of the number of observed stars while pursuing the long-term observations of the ELODIE era (since 1994). SOPHIE plays a very efficient role in the search for northern extrasolar planets including studies on the transition massive planets – brown dwarf (Bouchy et al. 2009a; Diaz et al. 2012), Jupiter analogs (Boisse et al. 2012; Rey et al. 2017), multigiant-planetary systems (Hébrard et al. 2010; Moutou et al. 2014), Neptunelike planet (Courcol et al. 2015), and stellar activity of planetary host stars (Boisse et al. 2009; Aigrain et al. 2012). SOPHIE is also intensively used for the RV followup of transit surveys SWASP (Pollacco et al. 2006; Collier Cameron et al. 2007), HAT (Shporer et al. 2009), CoRoT (Barge et al. 2008; Alonso et al. 2008), and Kepler (Santerne et al. 2011; Bouchy et al. 2011a). It led to studies on bloated hotJupiter (Hebb et al. 2009; Almenara et al. 2015), Saturn-size planets (Bouchy et al. 2010; Hébrard et al. 2014), misaligned spin-orbit giant planets (Hébrard et al. 2008; Moutou et al. 2009), mass-radius characterization of massive companions in the planet-brown dwarf boundary (Bouchy et al. 2011b; Diaz et al. 2014), and rate of false positive among the Kepler giant candidates (Santerne et al. 2012, 2016).

Conclusions From CORAVEL to HARPS and SOPHIE, there has been an impressive evolution of technology based on an approach of small steps, every time improvements had become necessary to overcome precision limitations. This approach is continued with the ESPRESSO instrument (Pepe et al. 2013b), which represents the next step of evolution on an 8-m class telescope. In parallel, great effort is being done to adapt the “HARPS-concepts” developed for high-precision visible spectrographs to the near infra-red domain, and to bring a new generation of NIR instruments like CARMENES@Calar Alto, SPIROU@CFHT, and NIRPS@La Silla to the 1 m s1 level of precision. ESPRESSO and the infrared spectrographs are however an intermediate step to the next level, i.e., precision spectrographs for extremely large telescopes. A phase A study is currently being conducted to propose such an instrument for the European ELT (HIRES Phase A). Looking forward to these future instruments, we finally would like to remind the reader that instrumental and photon-noise precision have to evolve jointly. Instrumental precision should therefore not be sacrificed on the altar of photons.

40 High-Precision Spectrographs for Exoplanet Research: CORAVEL,: : :

879

Acknowledgments This work has been carried out within the frame of the National Centre for Competence in Research PlanetS supported by the SNSF. The authors acknowledge the financial support of the SNSF. The authors would like to thank in the most sincere way all the persons who have contributed through their passionate and competent, but often unapparent, work of administration, management, engineering, technology, programming, etc., to the great scientific success of the described instruments.

References Aigrain S, Pont F, Zucker S (2012) A simple method to estimate radial velocity variations due to stellar activity using photometry. Mon Not R Astron Soc 419:3147 Almenara JM, Daminai C, Bouchy F et al (2015) SOPHIE velocimetry of Kepler transit candidates. XV. KOI-614b, KOI-206b, and KOI-680b: a massive warm Jupiter orbiting a G0 metallic dwarf and two highly inflated planets with a distant companion around evolved F-type stars. Astron Astrophys 575:71 Alonso R, Auvergne M, Baglin A et al (2008) Transiting exoplanets from the CoRoT space mission. II. CoRoT-Exo-2b: a transiting planet around an active G star. Astron Astrophys 482:L21 Anglada-Escudé G, Amado PJ, Barnes J et al (2016) A terrestrial planet candidate in a temperate orbit around Proxima Centauri. Nature 536(7617):437–440 Avila G, Buzzoni B, Casse M (1998) Fiber characterization and compact scramblers at ESO. Opt Astron Instrum 3355:900–904 Baranne A, Mayor M, Poncet JL (1979) Coravel – a new tool for radial velocity measurements. Vistas Astron 23:279–316 Baranne A, Queloz D, Mayor M et al (1996) ELODIE: a spectrograph for accurate radial velocity measurements. Astron Astrophys Suppl 119:373–390 Barge P, Baglin A, Auvergne M et al (2008) Transiting exoplanets from the CoRoT space mission. I. CoRoT-Exo-1b: a low-density short-period planet around a G0V star. Astron Astrophys 482:L17 Benz W, Mayor M (1981) A new method for determining the rotation of late spectral type stars. Astron Astrophys 93:235 Benz W, Mayor M (1984) Photoelectric rotational velocities of late-type dwarfs. Astron Astrophys 138:183 Boisse I, Moutou C, Vidal-Madjar A et al (2009) Stellar activity of planetary host star HD189733. Astron Astrophys 495:959 Boisse I, Eggenberger A, Santos NC et al (2010) The SOPHIE search for northern extrasolar planets. III. A Jupiter-mass companion around HD 109246. Astron Astrophys 523:88 Boisse I, Pepe F, Perrier C et al (2012) The SOPHIE search for northern extrasolar planets. V. Follow-up of ELODIE candidates: Jupiter-analogs around sun-like stars. Astron Astrophys 545:55 Bonfils X, Delfosse X, Udry S et al (2013) The HARPS search for southern extra-solar planets. XXXI. The M-dwarf sample. Astron Astrophys 549:75 Bouchy F, Pepe F, Queloz D (2001) Fundamental photon noise limit to radial velocity measurements. A&A 374:733–739 Bouchy F, Udry S, Mayor M et al (2005) ELODIE metallicity-biased search for transiting hot Jupiters. II. A very hot Jupiter transiting the bright K star HD 189733. Astron Astrophys 444:L15 Bouchy F, Hébrard G, Udry S et al (2009a) The SOPHIE search for northern extrasolar planets. I. A companion around HD 16760 with mass close to the planet/brown-dwarf transition. Astron Astrophys 505:853 Bouchy F, Isambert J, Lovis C et al (2009b) Charge transfer inefficiency effect for high-precision radial velocity measurements. EAS Publ Ser 37:247–253 Bouchy F, Hebb L, Skillen I et al (2010) WASP-21b: a hot-Saturn exoplanet transiting a thick disc star. Astron Astrophys 519:9

880

F. Pepe et al.

Bouchy F, Bonomo AS, Santerne A et al (2011a) SOPHIE velocimetry of Kepler transit candidates. III. KOI-423b: an 18 MJup transiting companion around an F7IV star. Astron Astrophys 533:83 Bouchy F, Deleuil M, Guillot T et al (2011b) Transiting exoplanets from the CoRoT space mission. XV. CoRoT-15b: a brown-dwarf transiting companion. Astron Astrophys 525:68 Bouchy F, Diaz R, Hébrard G et al (2013) SOPHIEC: first results of an octagonal-section fiber for high-precision radial velocity measurements. Astron Astrophys 549:49 Brown TM (1990) High precision Doppler measurements via echelle spectroscopy. In: CCDs in astronomy; Proceedings of the conference, 6–8 Sept 1989 (A91-45976 19-33). Astronomical Society of the Pacific, San Francisco/Tucson, pp 335–344 Butler RP, Marcy GW Williams E et al (1996) Attaining Doppler precision of 3 ms1 . Publ Astron Soc Pac 108:500 Campbell B, Walker G (1979) Precision radial velocities with an absorption cell. PASP 91:540 Charbonneau D, Brown TM, Latham DW, Mayor M (2000) Detection of planetary transits across a sun-like star. Astrophys J 529:45 Charbonneau D, Berta ZK, Irwin J et al (2009) A super-earth transiting a nearby low-mass star. Nature 462(7275):891–894 Collier Cameron A, Bouchy F, Hébrard G et al (2007) WASP-1b and WASP-2b: two new transiting exoplanets detected with SuperWASP and SOPHIE. Mon Not R Astron Soc 375:951 Courcol B, Bouchy F, Pepe F et al (2015) The SOPHIE search for northern extrasolar planets. VII. A warm Neptune orbiting HD 164595. Astron Astrophys 581:38 Da Silva R, Udry S, Bouchy F et al (2006) Elodie metallicity-biased search for transiting hot Jupiters. I. Two hot Jupiters orbiting the slightly evolved stars HD 118203 and HD 149143. Astron Astrophys 446:717 Diaz R, Santerne A, Sahlmann J et al (2012) The SOPHIE search for northern extrasolar planets. IV. Massive companions in the planet-brown dwarf boundary. Astron Astrophys 538:113 Diaz RF, Montagnier G, Leconte J et al (2014) SOPHIE velocimetry of Kepler transit candidates. XIII. KOI-189 b and KOI-686 b: two very low-mass stars in long-period orbits. Astron Astrophys 572:109 Dressing C, Charbonneau D, Dumusque X et al (2015) The mass of Kepler-93b and the composition of terrestrial planets. Astrophys J 800(2):7 Dumusque X, Pepe F, Lovis C et al (2012) An earth-mass planet orbiting ’ Centauri B. Nature 491:207–211 Dumusque X, Bonomo AS, Haywood RD et al (2014) The Kepler-10 planetary system revisited by HARPS-N: a hot rocky world and a solid Neptune-mass planet. Astrophys J 789(2):14 Felgett P (1953) Opt Acta 2:9 Griffin RF (1967) A photoelectric radial-velocity spectrometer. Astrophys J 148:465 Hebb L, Collier-Cameron A, Loeillet B et al (2009) WASP-12b: the hottest transiting extrasolar planet yet discovered. Astrophys J 693:1920 Hébrard G, Bouchy F, Pont F et al (2008) Misaligned spin-orbit in the XO-3 planetary system? Astron Astrophys 488:763 Hébrard G, Bonfils X, Ségransan D et al (2010) The SOPHIE search for northern extrasolar planets. II. A multiple planet system around HD 9446. Astron Astrophys 513:69 Hébrard G, Santerne A, Montagnier G et al (2014) Characterization of the four new transiting planets KOI-188b, KOI-195b, KOI-192b, and KOI-830b. Astron Astrophys 572:93 Henry GW, Marcy G, Butler RP, Vogt S (2000) A transiting “51 Peg-like” planet. Astrophys J 529:41 Hunter TR, Ramsey LW (1992) Scrambling properties of optical fibers and the performance of a double scrambler. Astron Soc Pac 104:1244–1251 Latham DW, Stefanik RP, Mazeh T, Mayor M, Burki G (1989) The unseen companion of HD114762 – a probable brown dwarf. Nature 339:38 Lovis C, Pepe F (2007) A new list of thorium and argon spectral lines in the visible. A&A 468(3):1115–1121 Lovis C et al (2006) An extrasolar planetary system with three Neptune-mass planets. Nature 441:305–309

40 High-Precision Spectrographs for Exoplanet Research: CORAVEL,: : :

881

Lovis C, Mayor M, Bouchy F et al. (2009) Towards the characterization of the hot Neptune/superEarth population around nearby bright stars. Transiting Planets. Proceedings of the International Astronomical Union, IAU symposium, vol 253, pp 502–505 Lovis C, Ségransan D, Mayor M et al (2011) The HARPS search for southern extra-solar planets. XXVIII. Up to seven planets orbiting HD 10180: probing the architecture of low-mass planetary systems. Astron Astrophys 528:16 Marmier M, Ségransan D, Udry S et al (2013) The CORALIE survey for southern extrasolar planets. XVII. New and updated long period and massive planets. Astron Astrophys 551:A90 Martins JHC, Santos N, Figueira P et al (2015) Evidence for a spectroscopic direct detection of reflected light from 51 Pegasi b. Astron Astrophys 576:9 Mayor M (1980) Metal abundance of F and G dwarfs determined by the radial velocity scanner CORAVEL. Astron Astrophys 87:1 Mayor M, Queloz D (1995) A Jupiter-mass companion to a solar-type star. Nature 378(6555): 355–359 Mayor M, Pepe F, Queloz D et al (2003) Setting new standards with HARPS. The Messenger (ISSN0722–6691) 114:20–24 Mayor M, Udry S, Lovis C et al (2009) The HARPS search for southern extra-solar planets. XIII a planetary system with 3 super-earths (4.2, 6.9 and 9.3 Mearths). Astron Astrophys 493(2): 639–644 Mayor M, Marmier M, Lovis C et al (2011) The HARPS search for southern extra-solar planets XXXIV. Occurrence, mass distribution and orbital properties of super-earths and Neptune-mass planets. arXiv 1109:2497 Mazeh T, Naef D, Torres G, Latham DW et al (2000) The spectroscopic orbit of the planetary companion transiting HD209458. Astrophys J 532:55 McArthur BE, Endl M, Cochran WD et al (2004) Detection of a Neptune-Mass Planet in the ¡ Cancri System Using the Hobby-Eberly Telescope. The Astrophysical Journal 614:L81 Motalebi F, Udry S, Gillon M et al (2015) The HARPS-N rocky planet search. I. HD 219134b, a transiting rocky planet in a multi-planet system at 6.5 pc from the sun. Astron Astrophys 584:72 Moutou C, Hébrard G, Bouchy F et al (2009) Photometric and spectroscopic detection of the primary transit of the 111-day-period planet HD 80 606 b Astronomy and Astrophysics 498:L5 Moutou C, Hébrard G, Bouchy F et al (2014) The SOPHIE search for northern extrasolar planets. VI. Three new hot Jupiters in multi-planet extrasolar systems. Astron Astrophys 563:22 Paresce F, Renzini A (1997) The search for extrasolar planets at ESO. The Messenger 90:15–18 Pepe F (2001) HARPS Final design and performance report, ESO Archive 3M6-TRE-HAR-331000013, 28 Feb 2001 Pepe F, Mayor M, Rupprecht G et al (2002) HARPS: ESO’s coming planet searcher. Chasing exoplanets with the La Silla 3.6-m telescope. The Messenger (ISSN 0722-6691) 110:9–14 Pepe F, Mayor M, Queloz D et al (2004) The HARPS search for southern extra-solar planets. I. HD 330075 b: a new “hot Jupiter”. A&A 423:385–389 Pepe F, Cameron AC, Latham DW et al (2013b) An earth-sized planet with an earth-like density. Nature 503(7476):377–380 Pepe F, Lovis C, Ségransan D et al (2013a) The HARPS search for earth-like planets in the habitable zone. I. Very low-mass planets around HD 20794, HD 85512, and HD 192310. Astron Astrophys 534:16 Pepe F, Molar P, Cristiani S et al (2013b) ESPRESSO: the next European exoplanet hunter. Astron Nachr 335(1):8 Pepe F, Ehrenreich D, Meyer M (2014) Instrumentation for the detection and characterization of exoplanets. Nature 513(7518):358–366 Perrier C, Sivan JP, Naef D et al (2003) The ELODIE survey for northern extra-solar planets. I. Six new extra-solar planet candidates. Astron Astrophys 410:1039 Perruchot S, Kohler D, Bouchy F et al (2008) The SOPHIE spectrograph: design and technical key-points for high throughput and high stability. Proc SPIE 7014:12 Perruchot S, Bouchy F, Chazelas B et al (2011) Higher-precision radial velocity measurements with the SOPHIE spectrograph using octagonal-section fibers. Proc SPIE 8151:815115

882

F. Pepe et al.

Pollacco D, Skillen I, Collier Cameron A et al (2006) The WASP project and SuperWASP camera. Astrophys Space Sci 304:253–255 Queloz D, Bouchy F, Moutou C et al (2009) The CoRoT-7 planetary system: two orbiting superearths. Astron Astrophys 506(1):303–319 Queloz D, Mayor M, Udry S et al (2011) From CORALIE to HARPS. The way towards 1 m s1 precision Doppler measurements. The Messenger (ISSN 0722-6691) 90:15–18 Rey J, Hébrard G, Bouchy F et al (2017) The SOPHIE search for northern extrasolar planets. XII. Three giant planets suitable for astrometric mass determination with Gaia. Astron Astrophys 601:9 Ricker G, Winn J, Vanderspek R, Latham DW, Bakos G et al (2015) Transiting exoplanet survey satellite (TESS). J Astron Telesc Instrum Syst 1:014003 Santerne A, Diaz RF, Bouchy F et al (2011) SOPHIE velocimetry of Kepler transit candidates. II. KOI-428b: a hot Jupiter transiting a subgiant F-star. Astron Astrophys 528:63 Santerne A, Diaz RF, Moutou C et al (2012) SOPHIE velocimetry of Kepler transit candidates. VII. A false-positive rate of 35% for Kepler close-in giant candidates. Astron Astrophys 545:7 Santerne A, Moutou C, Tsantaki M et al (2016) SOPHIE velocimetry of Kepler transit candidates. XVII. Astron Astrophys 587, 64 Santos NC, Israelian G, Mayor M (2001) The metal-rich nature of stars with planets. Astron Astrophys 373:1019 Santos NC, Israelian G, Mayor M, Rebolo R, Udry S (2003) Statistical properties of exoplanets. II. Metallicity, orbital parameters, and space velocities. Astron Astrophys 398:363 Santos NC, Bouchy F, Mayor M et al (2004a) The HARPS survey for southern extra-solar planets II. A 14 earth-masses exoplanet around Arae. A&A 426:L19–L23 Santos NC, Israelian G, Mayor M (2004b) Spectroscopic (Fe/H) for 98 extrasolar planet-host stars.Exploring the probability of planetary formation. Astron Astrophys 415:1153 Ségransan D, Udry S, Mayor M et al (2010) The CORALIE survey for southern extrasolar planets. XVI. Discovery of a planetary system around HD147018 and of two long period and massive planets orbiting HD171238 and HD204313. Astron Astrophys 511:A45 Shporer A, Bakos G, Bouchy F et al (2009) HAT-P-9b: a low-density planet transiting a moderately faint F star. Astrophys J 690:1393 Tamuz O, Ségransan D, Udry S et al (2008) The CORALIE survey for southern extra-solar planets. XV. Discovery of two eccentric planets orbiting HD4113 and HD156846. Astron Astrophys 480:33 Udry S, Mayor M, Naef D et al (2000) The CORALIE survey for southern extra-solar planets. II. The short-period planetary companions to HD75289 and HD130322. Astron Astrophys 356:590 Udry S, Bonfils X, Delfosse X et al (2007) The HARPS search for southern extra-solar planets. XI. Super-Earths (5 and 8 M˚) in a 3-planet system. Astron Astrophys 469(3):L43–L47 Wheatley P, West R, Goad M et al (2017) The Next Generation Transit Survey (NGTS). Mon Not R Astron Soc 469:2361. (in press) Wyttenbach A, Ehrenreich D, Lovis C et al (2015) Spectrally resolved detection of sodium in the atmosphere of HD 189733b with the HARPS spectrograph. Astron Astrophys 577:13 Wyttenbach A, Lovis C, Ehrenreich D (2017) Hot exoplanet atmospheres resolved with transit spectroscopy (HEARTS). I. Detection of hot neutral sodium at high altitudes on WASP-49b. Astron Astrophys 602:14

ESPRESSO on VLT: An Instrument for Exoplanet Research

41

Jonay I. González Hernández, Francesco Pepe, Paolo Molaro, and Nuno C. Santos

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Exoplanet Science with ESPRESSO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . An Ultra-stable High-Resolution Spectrograph for the VLT . . . . . . . . . . . . . . . . . . . . . . . . . . Observing Modes and Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Coudé Train . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Front End . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Fiber Link . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Optical Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Opto-Mechanics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Large-Area CCDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Data Flow System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . End-to-End Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

884 886 889 889 891 892 893 893 895 896 897 898 899

J. I. González Hernández () Departamento de Astrofísica, Universidad de La Laguna (ULL), La Laguna, Tenerife, Spain Instituto de Astrofísica de Canarias, La Laguna, Tenerife, Spain e-mail: [email protected] F. Pepe Département d’Astronomie, Observatoire de l’Université de Genéve, Versoix, GE, Switzerland e-mail: [email protected] P. Molaro INAF Osservatorio Astronomico di Trieste, Trieste, Italy e-mail: [email protected] N. C. Santos Instituto de Astrofísica e Ciências do Espaço, Universidade do Porto, CAUP, Porto, Portugal Departamento de Física e Astronomia, Faculdade de Ciências, Universidade do Porto, Porto, Portugal e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_157

883

884

J. I. González Hernández et al.

Abstract

ESPRESSO (Echelle SPectrograph for Rocky Exoplanets and Stable Spectroscopic Observations) is a VLT ultra-stable high-resolution spectrograph that will be installed in Paranal Observatory in Chile at the end of 2017 and offered to the community by 2018. The spectrograph will be located at the Combined Coudé Laboratory of the VLT and will be able to operate with one or (simultaneously) several of the four 8.2 m Unit Telescopes (UT) through four optical Coudé trains. Combining efficiency and extreme spectroscopic precision, ESPRESSO is expected to gaining about two magnitudes with respect to its predecessor HARPS. We aim at improving the instrumental radial velocity precision to reach the 10 cm s1 level, thus opening the possibility to explore new frontiers in the search for Earth-mass exoplanets in the habitable zone of quiet, nearby G to M dwarfs. ESPRESSO will be certainly an important development step toward high-precision ultra-stable spectrographs on the next generation of giant telescopes such as the E-ELT. Keywords

Instrumentation: spectrographs · Planetary systems · Techniques: spectroscopic

Introduction During the last decades, the exoplanet science has become one of the most exciting research fields in modern astrophysics. The Doppler spectroscopy or radial velocity (RV) technique (based on the determination of the projected velocity of stars in the direction of the line of sight) was the earliest method delivering the first detection of a low-mass companions (e.g., HD114762 b – Latham et al. 1989). This technique provided in 1995 the first discovery of a Jupiter-mass exoplanet orbiting the solar-type star 51 Pegasi (Mayor and Queloz 1995). After the discovery of 51 Peg b, direct imaging, microlensing, and specially transit searches both from the ground and space (Lissauer et al. 2014), together with the RV technique (Mayor et al. 2014), have produced an increasing number of detections of planets and planetary systems. The number of confirmed exoplanets discovered as of May 2017 is about 3,600, among which 2700 exoplanets have been detected using the transit technique, mostly from the space missions CoROT (Barge et al. 2006) and Kepler (Borucki et al. 2009). About 700 exoplanets have been discovered using the Doppler technique on high-resolution spectrographs. These detections from both transit and RV techniques have yielded a significant statistical contribution to our understanding of exoplanet population, such as the frequency of planets (Mayor et al. 2011; Howard et al. 2012), for different host spectral types and at different orbital distances including the habitable zone (i.e., in orbits where water is retained in liquid form on the planet surface).

41 ESPRESSO on VLT: An Instrument for Exoplanet Research

885

Now we know there is a wide variety of planetary systems with different planetary masses, sizes, orbital periods, eccentricities, and configurations of mass/sizedistances to the host stars. Most of the known planets appear to be in planetary systems and a significant fraction in multiple planet systems. Examples of complex planetary systems as the Solar System are the seven-planet system orbiting the G dwarf HD10180, detected using the RV technique with six confirmed planets in the mass range 11–65 ME and one planet candidate with a mass as small as 1.5 ME (Lovis et al. 2011), or the recent discovery using transit technique from the ground of also seven planets with similar size as that of the Earth orbiting the late M dwarf TRAPPIST-1 (Gillon et al. 2017). Those discoveries have required intensive and long-term monitoring of stars and the continuous development of astronomical instrumentation (Pepe et al. 2014a). This together with significant improvement in the observational strategies has made exoplanet science quickly evolve from the discovery of giant planets even more massive than Jupiter to rocky planets with similar mass as that of the Earth. This is demonstrated by the recent discoveries of Kepler 78 b (Pepe et al. 2013), a planet with similar density as that of the Earth, and Proxima Centauri b (Anglada-Escudé et al. 2016), a planet with similar minimum mass as that of the Earth in the habitable zone of the closest star to the Sun. Coupling the transit with the RV technique allows us to derive the mass and radius, and therefore to compute bulk density of exoplanets, a first step toward characterization of exoplanets. This in combination with detailed theoretical modeling allows us to get some insights about the bulk composition of some exoplanets and how they compared with the Earth composition (see Mayor et al. 2014; Lissauer et al. 2014, and references therein). The next step is to characterize the atmospheres of exoplanets, using, e.g., high-resolution spectroscopy via transmission with a technique applied to some transiting giant planets to detect Na and CO in their atmospheres (Charbonneau et al. 2002; Snellen et al. 2008, 2010; Wyttenbach et al. 2015) or via spectroscopic detection of reflected light of exoplanets (Charbonneau et al. 1999; Martins et al. 2015). Most of the known transiting exoplanets (namely, those discovered by the Kepler space telescope) orbit faint targets, which make it difficult to study their atmospheres with the current facilities. The recently found transiting super-Earth-like planet in the habitable zone of the faint M dwarf LHS 1140 (Dittmann et al. 2017) is an example of the need for larger telescopes equipped with high-resolution spectrograph to investigate the atmospheres of such interesting exoplanets. In particular, the difficulty raises when trying to find transiting Earth-like planets orbiting at the habitable zones of bright host stars. It appears quite necessary to design new techniques able to study the atmospheres of exoplanets by using the reflected light (Snellen et al. 2014), combining a high-contrast adaptive optics (AO) system with a high-resolution spectrograph to overcome the tiny planetto-star flux ratio in two stages. First, the AO system should spatially resolve the planet from the host star enhancing the planet-to-star contrast at the planet location, and second, the light beam at the planet location passes through the highresolution spectrographs. This technique has been recently applied to model the possible planetary atmospheric signal of Proxima Cen b by coupling the SPHERE

886

J. I. González Hernández et al.

high-contrast imager (Beuzit et al. 2008) and the ESPRESSO spectrograph at VLT (Lovis et al. 2017). The new generation of 20–40 m giant telescopes such as E-ELT equipped with sophisticated AO systems and with stable high-precision highresolution spectrographs will possibly be able to study the atmospheres of rocky exoplanets in the habitable zone (Snellen et al. 2015). The need of instrumentation for exoplanet science on the 8.2-m Very Large Telescopes (VLT) was highlighted in the ESO (European Southern Observatory)ESA(European Space Agency) working report on extrasolar planets. In October 2007, the ESO Science Advisory Committee recommended the development of new second-generation VLT instrumentation, later endorsed by the ESO Council. Among those instruments, ESPRESSO, a high-resolution ultra-stable spectrograph for the VLT Combined Coudé focus, was proposed. The ESPRESSO (Echelle SPectrograph for Rocky Exoplanets and Stable Spectroscopic Observations) project started with the kick-off meeting in February 2011. The ESPRESSO consortium is composed of Observatoire Astronomique de l’Université de Genéve (project head, Switzerland), Instituto de Astrofísica e Ciências do Espaço/Universidade de Porto and Universidade de Lisboa (Portugal), INAF-Osservatorio Astronomico di Brera (Italy), INAF-Osservatorio Astronomico di Trieste (Italy); Instituto de Astrofísica de Canarias (Spain), and Physikalisches Institut der Universität Bern (Switzerland). ESO participates to the ESPRESSO project as Associated Partner. The ESPRESSO instrument is expected to be commissioned at the VLT in Paranal Observatory in the fall 2017 and possibly offered to the community by 2018.

Exoplanet Science with ESPRESSO The main scientific drivers of ESPRESSO are: (i) search for rocky planets; (ii) measure the variation of physical constants; and (iii) analyze the chemical composition of stars in nearby galaxies. We refer to Pepe et al. (2014b) for details on (ii) and (iii), as well as very challenging additional scientific topics that ESPRESSO will be able to address. One of the main scientific topics in the next decades is the search and characterization of terrestrial planets in the habitable zone of their host stars, and one of the main drivers of the new generation of extremely large telescopes is the detection of their atmospheres (see, e.g., Marconi et al. 2016). At the end of the 1990s and the beginning of the last decade, high-resolution spectroscopic monitoring of stars yielded many detections of Jupiter-like planets. Only after 2003, the HARPS (Mayor et al. 2003) spectrograph installed at the 3.6-m ESO telescope in La Silla Observatory (Chile) opened a new window on the domain of Neptuneand super-Earth-like planets (see the Review by Pepe et al. in this HandBook of Exoplanets). This instrument (with a resolving power of R  115;000) is contained inside a vacuum vessel and uses a simultaneous calibration reference that allows to correct for (small) instrumental drifts due to (also small) changes of temperature and pressure, providing an extreme long-term RV precision below the 1 m s1 (see Fig. 1). Already in 2004, the discovery of a super-Earth with a minimum mass of

41 ESPRESSO on VLT: An Instrument for Exoplanet Research

887

HD10700 4

RV [m/s]

2

0

−2

−4

3000

3500

4000

4500

5000

5500

6000

JD − 2450000.0 [days] Fig. 1 Ten years HARPS RV data of the late G-type dwarf star  Ceti. The overall dispersion is about 1 m s1 . Time-binning of the data would decrease the dispersion down to about 20 cm s1 . The absence of any long-term trend in the data is remarkable, thus demonstrating the outstanding precision of the HARPS spectrograph (This figure has been taken from Pepe et al. 2014b)

about only 10 ME orbiting the G-type star  Arae (Santos et al. 2004) demonstrated the impressive capabilities of the HARPS instrument. Since then, and thanks to its precision, HARPS has provided the discovery of most of the sub-Neptunian mass planets with masses down to a few Earth masses inducing RV semiamplitudes as low as 0.5 m s1 (see, e.g., Pepe et al. 2011). The ESPRESSO instrument will improve a step further, aiming at achieving an RV precision of 10 cm s1 which is crucial in the path of detecting terrestrial planets in the habitable zone of host stars of different spectral types, in particular for G and K dwarfs. The Earth induces radial velocity variations in the Sun of 9 cm s1 , in comparison with RV signal of 12 m s1 induced by Jupiter. However, an Earth-mass exoplanet orbiting a M5V star in the habitable zone would cause a gravitational pull equivalent to 1.3 m s1 which, with the current instrument capabilities limited to 1 m s1 , can be detected (Bonfils et al. 2013a; Anglada-Escudé et al. 2016). Thus, the quest for low-mass planets in the habitable zone has favored RV studies on M dwarf in the last decade with numerous low-mass planet detections (see, e.g., Bonfils et al. 2013a,b; Anglada-Escudé et al. 2013; Affer et al. 2016; AstudilloDefru et al. 2017; Suárez Mascareño et al. 2017a,b). Today, several tens of planets with minimum masses below 10 ME have been discovered. Most of them have been detected orbiting cool dwarfs less massive than the Sun (see Fig. 2), using HARPS and HARPS-N (at 3.6-m TNG telescope located in the Observatorio del Roque de los Muchachos, La Palma, Spain; Cosentino et al. 2012) spectrographs. New dedicated instruments aiming for planet search around M dwarfs operating in the

888

J. I. González Hernández et al.

near infrared are already running such as CARMENES (at the 3.5-m telescope in the Observatorio de Calar Alto, Spain; Quirrenbach et al. 2016) or under construction such as NIRPS that will operate together with HARPS at the 3.6-m ESO telescope in La Silla (Chile) and SPIRou at the CHFT telescope in Mauna Kea (Hawaii, USA). A larger telescope size provides a lower photon noise level on Doppler signal for the same exposure time. ESPRESSO at the 8.2-m VLT at the Observatorio de Paranal is expected to achieve the 10 cm s1 Doppler precision and long-term stability. This will open the possibility to search for Earth-mass planets at different orbital distances, including the habitable zones of solar-type stars. A carefully selected sample of non-active, nonrotating, quiet G to M dwarf will allow to explore this new domain. In Fig. 2, we display the known planets at different orbital periods around stars with different apparent brightness. So far, most of the low-mass planets have been discovered around M dwarfs where the RV planetary signals are stronger. The larger telescope mirror of VLT and the RV precision provided by ESPRESSO will also allow to access a larger sample of fainter stars of different spectral types. However, stellar noise or jitter, which causes different radial velocity variations at

Fig. 2 Minimum mass – orbital period diagram for known planets orbiting solar-type stars. The color of the symbols is related to the mass of the host star given in solar mass (Mˇ ) according to the color bar on top. Inclined dashed-dotted lines show the RV semiamplitude of planets orbiting a late M dwarf star with 0.25 Mˇ (green line) and a G dwarf star with 1 Mˇ star (blue line) assuming a RV semiamplitude of 1 m s1 and 10 cm s1 , respectively, and null eccentricity. Planets of the solar system (filled black circles) are labeled. The “habitable zones” of 0.8–1.2 Mˇ and 0.2–0.3 Mˇ stars are indicated with blue and pink rectangles, respectively. These are regions where rocky planets with a mass in the interval 0.1–10 ME would retain liquid water on their surface

41 ESPRESSO on VLT: An Instrument for Exoplanet Research

889

different timescales and of different magnitude (see, e.g., Saar et al. 1998; Santos et al. 2000; Queloz et al. 2001; Boisse et al. 2011; Dumusque et al. 2011; Robertson et al. 2014; Suárez Mascareño et al. 2017c), still remains the main source of error and probably the strongest limitation toward the sub-m s1 precision. Therefore, continuous investigation and modeling on these stellar effects are required in the way toward finding rocky planets in the habitable zones of solar-type stars. The discovery of a new and large population of Earth-mass exoplanets orbiting solar-type stars will expand our knowledge of planet formation and will also deliver new candidates for follow-up observations using other techniques such as transit, astrometry, and Rossiter-McLaughlin (RM) effect. The detection with HARPS of the RM effect in occasion of the 2012 transit of Venus over the disk of the SUN provided a sort of preview of the kind of the physical information which we can hope to obtain by observing transits of exoplanets with a large telescope (Molaro et al. 2013b). ESPRESSO could also perform follow-up observations of ongoing and forthcoming transit surveys such as NGTS and MEarth from the ground or Kepler-K2, TESS, and Plato from the space. ESPRESSO will be possibly one of the rare instruments able to confirm long-period Earth-size transiting planets discovered by space missions like Plato, which hopefully will provide Earth-size candidates transiting bright targets in their habitable zones. The high resolution and stability of ESPRESSO will allow atmospheric characterization of exoplanets of different mass and size using the transmission and possibly reflection techniques (Charbonneau et al. 2002; Snellen et al. 2014; Martins et al. 2015; Lovis et al. 2017).

An Ultra-stable High-Resolution Spectrograph for the VLT ESPRESSO is a fiber-fed, cross-dispersed, high-resolution échelle spectrograph located in the Combined Coudé Laboratory (CCL) at the incoherent focus, where a front-end unit can combine the light from up to four Unit Telescopes (UT) of the VLT. The so-called Coudé train optical system will fed the light of each UT to the spectrograph. The sky light and the target will enter the instrument simultaneously through two separate fibers, which form together the slit of the spectrograph. ESPRESSO, unlike any other ESO instrument, will be able to receive the light from any of the four 8.2-m UTs and will be able to operate simultaneously with the light of either one UT or several UTs (see Fig. 3).

Observing Modes and Performance The extreme precision of ESPRESSO will be achieved based on well-known concepts provided by the HARPS experience. The light of one or several UTs is fed through the front-end unit into optical fibers that scramble the light and provide excellent illumination stability to the spectrograph. In order to improve light scrambling, noncircular but octagonal or square fiber shapes are used (Chazelas et al. 2012). The target fiber can be fed either with the light from the astronomical

890

J. I. González Hernández et al.

Fig. 3 Schematic view of the four 8.2 m Unit Telescopes of the VLT (multi-UT mode) feeding, through the Coudé train, the front-end unit of the ESPRESSO spectrograph located in the Combined Coudé Laboratory

Table 1 Observing modes of ESPRESSO

Par./mode Wave. range Resol. power Aper. on sky Spec. samp. Spat. samp. Sim. ref. Sky sub. Tot. eff.

HR (1UT) 380–780 nm 134,000 1:000 4.5 pix 11  2 pix Yes (no sky) Yes (no ref.) 11%

UHR (1UT) 380–780 nm 200,000 0:500 2.5 pix 5  2 pix Yes (no sky) Yes (no ref.) 5%

MR(1–4UTs) 380–780 nm 59,000 400  100 11 pix 22  2 pix Yes (no sky) Yes (no ref.) 11%

object or one of the calibration sources. The reference fiber will receive either sky light (faint source mode) or calibration light (bright source mode). In the latter case, the simultaneous reference technique successfully applied in HARPS will enable to track instrumental drifts down to the cm s1 level. In this mode, the measurement is photon-noise limited and detector readout noise negligible. In the faint-source mode, instead, detector noise and sky background may become significant. In this case, the second fiber will allow to measure the sky background, whereas a slower readout and high binning factor will reduce the detector noise. ESPRESSO will have three instrumental modes: singleHR, singleUHR, and multiMR (see Table 1). Each mode will be available with two different detector readout modes optimized for lowand high-SNR measurements, respectively.

41 ESPRESSO on VLT: An Instrument for Exoplanet Research 5

5

10

10 10

Pixel = 0.0011 nm RON=5, Bin=1x1 Resolution 134.000

10

Approx. 10 cm/s limit

10

Pixel = 0.0022 nm RON=3, Bin=2x4 Resolution 59.000

4

3

S/N

3

10 S/N

multiMR, 4UT

singleHR, fast

4

2

10

Approx. 100 cm/s limit

2

10

1

1

10

0

10

10

0

10

−1

−1

10

891

0

5

10 15 Magnitude mv

20

25

10

0

5

15 10 Magnitude mv

20

25

Fig. 4 Expected signal-to-noise ratio versus the stellar apparent visible magnitude for the singleHR mode (left panel) and the multiMR mode (right panel). Red, blue, and magenta curves indicate exposure times of 3600, 1200, and 60 s, respectively (This figure has been taken from Pepe et al. 2014b)

The observational efficiency of ESPRESSO is shown in Fig. 4. In the singleHR mode (R  134; 000), we estimate SNR = 10 per extracted pixel in 20 min on a V D 16:3 star or a SNR = 540 on a V D 8:6 star. We have estimated that at this resolution and a SNR = 540 will lead to 10 cm s1 RV precision for a nonrotating K5 star. For an F8 star, the same precision would be achieved for V D 8. In the multiMR mode, at R  59; 000, a SNR of 10 is achieved on a V D 19:4 star with an exposure of 20 min, a binning 2  4, and a slow readout of the CCD. In the following sections, we briefly describe the several subsystems of the ESPRESSO project (see Pepe et al. 2014b).

The Coudé Train The four VLT telescope are connected to the CCL through four tunnels with a length that goes from 48 to 69 m (see Fig. 3). A trade-off analysis among the different solutions to bring the light from the telescope to the CCL favored a full optical solution that includes prisms, mirrors, and lenses. The selected design uses 11 optical elements (see Fig. 5). The Coudé train takes the light beam with a prism at the Nasmyth-B platform and conduct it through the UT mechanical structure down to the Coudé room below each UT using a set of six prims and mirrors. The light is then routed from the UT Coudé room to the CCL, using two large lenses along the existing incoherent light ducts. The four Coudé trains relay a field of 17 arcsec around the acquired astronomical target to the CCL. The four beams from four UTs are combined in the CCL, where mode selection and beam conditioning is performed by the fore-optics of the front-end subsystem. At the time of writing, all four Coud trains have been installed and tested in Paranal. They have shown to provide seeing limited images to the front end that is installed at the CCL.

892

J. I. González Hernández et al.

Fig. 5 Schematic view of the Coudé train of ESPRESSO with the optical path through the telescope to the Combined Coudé Laboratory. The optical elements are highlighted at different positions along the path from the Nasmyth-B platform to the CCL with P (prism), R (mirror), and L (lens)

The Front End The front end conducts the light beam received at the CCL after correcting it for atmospheric dispersion with the ADC to the common focal plane where the spectrograph fiber feeds are located. A toggling mechanism handles the selection between the possible observational modes in a fully passive way. The beam conditioning is performed applying pupil and field stabilization (see Fig. 6). These are achieved via two independent control loops each consisting of a technical camera and a tip-tilt stage. Another dedicated stage delivers a focusing function. The front end also handles the injection of the calibration light from the calibration unit into the fibers and then into the spectrograph. A laser frequency comb (LFC) system is foreseen as main calibration source. It produces a regular spectrum of lines equally spaced in frequency with an accuracy and stability linked to an atomic clock. The short-term Doppler shift repeatability of the LFC system has been tested in HARPS spectrograph and demonstrated to achieve the cm s1 level (Wilken et al. 2012; Probst et al. 2016). The required repeatability of the order of =  1010 cannot be attained with currently used spectral sources such as thorium argon spectral lamps, iodine cells, etc., but can be obtained with a LFC that would provide a spectrum sufficiently wide, rich, stable, and uniform for this purpose (Lo Curto et al. 2012; Molaro et al. 2013a). However, two ThAr lamps for both simultaneous reference and calibration will be also available as backup calibration sources, together with one simultaneous stabilized Fabry-Pérot unit, also as a backup solution.

41 ESPRESSO on VLT: An Instrument for Exoplanet Research

893

Fig. 6 Front-end unit and the arrival of one UT beam at the CCL. The same is replicated for the other UTs

The Fiber Link The fiber-link subsystem transfers the light from the front end to the spectrograph and creates the a pseudo-slit in the output end inside the vacuum vessel. The 1-UT mode uses a bundle of two octagonal fibers each, one for the astronomical object and one for the sky or simultaneous reference. In the high-resolution (singleHR) mode, the fiber has a core of 140 m, equivalent to 100 on the sky; in the ultrahigh resolution (singleUHR) mode, the fiber core is 70 m, covering a field of view of 0:500 . The fiber entrances are organized in pickup heads that are moved to the focal plane of the Front End when the specific bundle of the specific mode is selected. In the 4-UT mode (multiMR), four object fibers and four sky/reference fibers are fed simultaneously by the four telescopes. The four object fibers and the four sky/reference fibers will finally feed two separate single square fibers of 280 m, for the object and for the sky/reference, respectively. In the 4-UT mode, the spectrograph will also see a pseudo-slit of four fiber square images twice as wide as the 1-UT fibers. One essential task performed by the fiber-link subsystem is the light scrambling. The use of a double-scrambling optical system will ensure both scrambling of the near field and far field of the light beam. A high scrambling gain, which is crucial to obtain the required RV precision in the 1-UT mode, is achieved by the use of octagonal fibers (Chazelas et al. 2012).

Optical Design Designing a high-efficiency and high-resolution spectrograph is not an easy task due to the large mirrors of the VLT telescopes and the 1 arcsec aperture of the

894

J. I. González Hernández et al.

Fig. 7 Layout of the ESPRESSO spectrograph and its optical elements

instrument. In order to minimize the size of the optics and, in particular, that of the main collimator and échelle grating, ESPRESSO implements an anamorphic optics, the APSU, which compresses the size of the pupil in the direction of the cross-dispersion. The pupil is then sliced in two by a pupil slicer, and the slices are overlapped on the échelle grating, leading to a doubled spectrum on the detector. This design reduces significantly the sizes of the optics and the échelle grating. Without this trick, the collimator beam size would have been 40 cm in diameter, and the size of the échelle grating would have reached a size of 240  40 cm. The size of current échelle grating of ESPRESSO is only 120  20 cm, and this also allows the use of much smaller optics (collimators, cross dispersers, etc.). The échelle grating will be an R4 Echelle of 31.6 l mm1 and a blaze angle of 76ı . This solution significantly reduces the overall costs. The drawback is that each spectral element will be covered by more detector pixels given the doubled image of the object fiber and its elongated shape on the CCD. In order to avoid to increase the detector noise, heavy binning will be done in the case of faint-object observations, especially in the 4-UT mode. The main components of the optical design are (see Fig. 7): • The anamorphic pupil slicing unit (APSU). At the spectrograph entrance, the APSU shapes the beam in order to compress it in cross-dispersion and splits in two smaller beams while superimposing them on the échelle grating to minimize its size. The rectangular white pupil is then re-imaged and compressed. • Dichroic. Given the wide spectral range, a dichroic beam splitter separates the beam in a blue and a red arm, which in turn allows to optimize each arm for image quality and optical efficiency. • Volume-phase holographic gratings (VPHGs). The cross-disperser enables to separate the dispersed spectrum in all its spectral orders. In addition, an anamorphism is reintroduced to make the pupil square and to compress the order height such that the inter-order space and the SNR per pixel are both maximized. Both

41 ESPRESSO on VLT: An Instrument for Exoplanet Research Blue chip

Red chip 780 nm

530 nm

510 nm

380 nm

895 Slit

Fig. 8 Spectral format of the red (left panel) and blue (middle panel) spectra, and a zoom of the pseudo-slit (right panel), showing the image of the target (bottom) and sky (top) fibers. Each fiber is re-imaged into two slices. The three sets of fibers, corresponding (from left to right) to the standard resolution 1-UT mode, ultra-high resolution 1-UT mode, and the mid-resolution 4-UT mode (shown simultaneously for comparison) (This figure has been taken from Pepe et al. 2014b)

functions are accomplished using volume-phase holographic gratings (VPHGs) mounted on prisms. • Fast cameras. Two optimized camera lens systems image the full spectrum from 380 to 780 nm on two large 92  92 mm CCDs with 10- m pixels. A sketch of the optical layout is depicted in Fig. 7. The spectral format covered by the blue and the red chips as well as the shape of the pseudo slit are displayed in Fig. 8. In order to precisely compute the relative Earth motion to be able to properly correct the RV measurement, it is necessary to calculate the weighted mean time of exposure. Thus, the spectrograph is also equipped with an advanced exposure meter that measures the flux entering the spectrograph as a function of time. Its innovative design (based on a simple diffraction grating) allows a flux measurement and an RV correction at different spectral channels, in order to cope with possible chromatic effects that could occur during the scientific exposures. The use of various channels also provides a redundant and thus more reliable evaluation of the mean time of exposure.

The Opto-Mechanics ESPRESSO has been designed to be an ultra-stable spectrograph enabling RV precisions of the order of 10 cm s1 , i.e., one order of magnitude better than its predecessor HARPS. ESPRESSO is therefore built with a totally fixed configuration and with the highest thermomechanical stability. The spectrograph optics are mounted in an optical bench specifically designed to keep the optical system within the thermomechanical tolerances required for high-precision RV measurements. The bench is mounted in a vacuum vessel in which 105 mbar class vacuum is maintained during the entire duty cycle of the instrument. An overview of the opto-

896

J. I. González Hernández et al.

Fig. 9 Opto-mechanics of the ESPRESSO spectrograph

mechanics is shown in Fig. 9. The temperature at the level of the optical system is required to be stable at the mK level in order to avoid both short-term drift and long-term mechanical instabilities. Such an ambitious requirement is obtained by locating the spectrograph in a multi-shell active thermal enclosure system as shown in Fig. 10. Each shell will improve the temperature stability by a factor 10, thus getting from typically Kelvin-level variations in the CCL down to 0.001 K stability inside the vacuum vessel and on the optical bench.

Large-Area CCDs The CCDs are another innovative solution in the ESPRESSO project. Large monolithic state-of-the-art CCDs have been chosen to use the optical field of ESPRESSO and to further improve the stability compared to the mosaic solution employed in HARPS. The sensitive area of the e2v chip is 92  92 mm covering 8:46  107 pixels of 10 m size. Fast readout of such a large chip is achieved by using its 16 output ports at high speed. Other requirements on CCDs are very demanding, e.g., in terms of charge transfer efficiency (CTE) and all the other parameters affecting the definition of the pixel position, immediately reflected into the radial velocity precision and accuracy. The RV precision of 10 cm s1 rms requires measuring spectral line position changes of 2 nm (physical) in the CCD plane, equivalent to only four times the silicon lattice constant. For better stability

41 ESPRESSO on VLT: An Instrument for Exoplanet Research

897

Fig. 10 Schematic view of the Combined Coudé Laboratory where ESPRESSO spectrograph will be located inside the vacuum vessel and several thermal enclosures in a multi-shell control system

and thermal expansion matching, the CCD package is made of silicon carbide. The package of the CCDs, the surrounding mechanics and precision temperature control inside the cryostat head and its cooling system, as well as the thermal stability and the homogeneous dissipation of the heat locally produced in the CCDs during operation are of critical importance. ESO has thus built a new superstable cryostat that has already demonstrated excellent short-term stability.

Data Flow System The ESPRESSO project has the final goal to provide the user with scientific data as complete and precise as possible in a short time (within minutes) after the end of an observation, to increase the overall efficiency and the ESPRESSO scientific output. For this purpose, a software-cycle integrated view from the observation preparation through instrument operations and control to the data reduction and analysis has been adopted. Coupled with a careful design, this will ensure optimal compatibility and will facilitate the operations and maintenance within the existing ESO Paranal Data Flow environment both in service and visitor mode. ESPRESSO Data Flow System presents the following main subsystems: • The ESPRESSO Observation Preparation Software (EOPS): a dedicated visitor tool (able to communicate directly with the VOT – Visitor Observing Tool) to help the observer to prepare and schedule ESPRESSO observations at the telescope according to the needs of planet-search surveys or other scientific programs. The tool will allow users to choose the targets best suited for a given

898

J. I. González Hernández et al.

night and to adjust the observation parameters in order to obtain the best possible quality of data. • The Data Reduction Software (DRS): ESPRESSO will have a fully automatic data reduction pipeline with the specific aim of delivering to the user high-quality reduced data, science ready, already in a short time after an observation has been performed. The computation of the RV at a precision better than 10 cm/s will be an integral part of the DRS. Coupled with the need to optimally remove the instrument signature, to take account the complex spectral and multi-HDU FITS format, the handling of the simultaneous reference technique and the multi-UT mode will make the DRS a truly challenging component of the DFS chain. • The Data Analysis Software (DAS): dedicated data analysis software will allow to obtain the best scientific results from the observations directly at the telescope. A robust package of recipes tailored to ESPRESSO, taking full advantage of the existing ESO tools (based on CPL and fully compatible with Reflex), will address the most important science cases for ESPRESSO by analyzing (as automatically as possible) stars and quasar spectra (among others, tasks will be performed such as line Voigt-profile fitting, estimation of stellar atmospheric parameters, normalization of stellar spectra and comparison with synthetic spectra, quasar continuum fitting, identification of absorption systems). • Templates and control: compared to other stand-alone instruments, the main reason for the complexity of the ESPRESSO acquisition and observation templates will be the possible usage of any combination of UTs, besides the proper handling of the simultaneous reference technique. ESPRESSO will contribute to open the new path for the control systems of future ESO instrumentation.

End-to-End Operation ESPRESSO will combine an unprecedented RV and spectroscopic precision with the largest photon collecting area available today at the European Southern Observatory, with a unique resolving power up to R  200;000. In the singleHR mode, ESPRESSO can be fed with the light of an astronomical object coming from any of the four 8.2 m VLT telescopes, which significantly improves the scheduling flexibility for ESPRESSO programs and surely will optimize the use of VLT time. The singleHR mode will operate at a resolution of 134,000 with an RV precision of 10 cm s1 , opening the possibility to explore a new population of rocky planets orbiting the habitable zones of solar-type stars. The scheduling flexibility is a fundamental advantage for survey programs like RV searches for exoplanets or time-critical programs like studies of transiting planets. The 1-UT mode also offers the possibility to enhance the resolving power up to 200,000 with an extremely stable wavelength accuracy which certainly will motivate new scientific projects, e.g., high-accuracy stellar astrophysics projects. In addition, in the multiMR mode, ESPRESSO will be able to collect the light of up to four UTs to generate a highresolution spectrum. The effectively larger telescope aperture of about 16m will provide access to faint astronomical targets at a resolution of 59,000. ESPRESSO

41 ESPRESSO on VLT: An Instrument for Exoplanet Research

899

shall not be considered as a stand-alone instrument but as a science-generating machine, certainly delivering full-quality scientific data in less than a minute after the end of an observation. Acknowledgements The ESPRESSO project is supported by the Swiss National Science Foundation program FLARE, Italian Institute of Astrophysics (INAF), Instituto de Astrofísica de Canarias (IAC, Spain), Instituto de Astrofísica e Ciências do Espaço/Universidade de Porto and Universidade de Lisboa (Portugal), and European Southern Observatory (ESO). J.I.G.H. acknowledges financial support from the Spanish Ministry of Economy and Competitiveness (MINECO) under the 2013 Ramón y Cajal programme MINECO RYC-2013-14875, and the Spanish ministry project MINECO AYA2014-56359-P. N.C.S. acknowledges the support by Fundação para a Ciência e a Tecnologia (FCT, Portugal) through the research grant through national funds and by FEDER through COMPETE2020 by grants UID/FIS/04434/2013&POCI-01-0145-FEDER007672 and PTDC/FIS-AST/1526/2014&POCI-01-0145-FEDER-016886, as well as through Investigador FCT contract nr. IF/00169/2012/CP0150/CT0002. The authors wish to acknowledge the exceptional work and enthusiasm delivered by all the members of the ESPRESSO team and warmly thank them for significantly contributing to the successful completion of the project.

References Affer L, Micela G, Damasso M et al (2016) HADES RV program with HARPS-N at the TNG GJ 3998: an early M-dwarf hosting a system of super-Earths. A&A 593:A117 Anglada-Escudé G, Tuomi M, Gerlach E et al (2013) A dynamically-packed planetary system around GJ 667C with three super-Earths in its habitable zone. A&A 556:A126 Anglada-Escudé G, Amado PJ, Barnes J et al (2016) A terrestrial planet candidate in a temperate orbit around Proxima Centauri. Nature 536:437–440 Astudillo-Defru N, Forveille T, Bonfils X et al (2017) The HARPS search for southern extra-solar planets. XLI. A dozen planets around the M dwarfs GJ 3138, GJ 3323, GJ 273, GJ 628, and GJ 3293. A&A 602:A88 Barge P, Léger A, Ollivier M et al (2006) Photometric search for transiting planets. In: Fridlund M, Baglin A, Lochard J, Conroy L (eds) The CoRoT mission pre-launch status – stellar seismology and planet finding, vol 1306. ESA Special Publication, p 83 Beuzit JL, Feldt M, Dohlen K et al (2008) SPHERE: a ‘Planet Finder’ instrument for the VLT. In: Ground-based and airborne instrumentation for astronomy II. Proc SPIE 7014:701418. https:// doi.org/10.1117/12.790120 Boisse I, Bouchy F, Hébrard G et al (2011) Disentangling between stellar activity and planetary signals. A&A 528:A4 Bonfils X, Delfosse X, Udry S et al (2013a) The HARPS search for southern extra-solar planets. XXXI. The M-dwarf sample. A&A 549:A109 Bonfils X, Lo Curto G, Correia ACM et al (2013b) The HARPS search for Southern extrasolar planets. XXXIV. A planetary system around the nearby M dwarf GJ 163, with a super-Earth possibly in the habitable zone. A&A 556:A110 Borucki W, Koch D, Batalha N et al (2009) KEPLER: search for Earth-size planets in the habitable zone. In: Pont F, Sasselov D, Holman MJ (eds) Transiting planets, IAU symposium, vol 253, pp 289–299. https://doi.org/10.1017/S1743921308026513 Charbonneau D, Noyes RW, Korzennik SG et al (1999) An upper limit on the reflected light from the planet orbiting the star  bootis. ApJ 522:L145–L148 Charbonneau D, Brown TM, Noyes RW, Gilliland RL (2002) Detection of an extrasolar planet atmosphere. ApJ 568:377–384 Chazelas B, Pepe F, Wildi F (2012) Optical fibers for precise radial velocities: an update. In: Modern technologies in space- and ground-based telescopes and instrumentation II. Proceedings of SPIE, vol 8450, p 845013. https://doi.org/10.1117/12.926188

900

J. I. González Hernández et al.

Cosentino R, Lovis C, Pepe F et al (2012) Harps-N: the new planet hunter at TNG. In: Ground-based and airborne instrumentation for Astronomy IV. Proceedings of SPIE, vol 8446, p 84461V. https://doi.org/10.1117/12.925738 Dittmann JA, Irwin JM, Charbonneau D et al (2017) A temperate rocky super-Earth transiting a nearby cool star. Nature 544:333–336 Dumusque X, Udry S, Lovis C, Santos NC, Monteiro MJPFG (2011) Planetary detection limits taking into account stellar noise. I. Observational strategies to reduce stellar oscillation and granulation effects. A&A 525:A140 Gillon M, Triaud AHMJ, Demory BO et al (2017) Seven temperate terrestrial planets around the nearby ultracool dwarf star TRAPPIST-1. Nature 542:456–460 Howard AW, Marcy GW, Bryson ST et al (2012) Planet occurrence within 0.25 AU of solar-type stars from Kepler. ApJS 201:15 Latham DW, Stefanik RP, Mazeh T, Mayor M, Burki G (1989) The unseen companion of HD114762 – a probable brown dwarf. Nature 339:38–40 Lissauer JJ, Dawson RI, Tremaine S (2014) Advances in exoplanet science from Kepler. Nature 513:336–344 Lo Curto G, Pasquini L, Manescau A et al (2012) Astronomical spectrograph calibration at the exo-earth detection limit. The Messenger 149:2–6 Lovis C, Ségransan D, Mayor M et al (2011) The HARPS search for southern extra-solar planets. XXVIII. Up to seven planets orbiting HD 10180: probing the architecture of low-mass planetary systems. A&A 528:A112 Lovis C, Snellen I, Mouillet D et al (2017) Atmospheric characterization of Proxima b by coupling the SPHERE high-contrast imager to the ESPRESSO spectrograph. A&A 599:A16 Marconi A, Di Marcantonio P, D’Odorico V et al (2016) EELT-HIRES the high-resolution spectrograph for the E-ELT. In: Ground-based and airborne instrumentation for astronomy VI. Proceedings of SPIE, vol 9908, p 990823. https://doi.org/10.1117/12.2231653 Martins JHC, Santos NC, Figueira P et al (2015) Evidence for a spectroscopic direct detection of reflected light from 51 Pegasi b. A&A 576:A134 Mayor M, Queloz D (1995) A Jupiter-mass companion to a solar-type star. Nature 378:355–359 Mayor M, Pepe F, Queloz D et al (2003) Setting new standards with HARPS. The Messenger 114:20–24 Mayor M, Marmier M, Lovis C et al (2011) The HARPS search for Southern extra-solar planets XXXIV. Occurrence, mass distribution and orbital properties of super-Earths and Neptune-mass planets. ArXiv e-prints Mayor M, Lovis C, Santos NC (2014) Doppler spectroscopy as a path to the detection of Earth-like planets. Nature 513:328–335 Molaro P, Esposito M, Monai S et al (2013a) A frequency comb calibrated solar atlas. A&A 560:A61 Molaro P, Monaco L, Barbieri M, Zaggia S (2013b) Detection of the Rossiter-McLaughlin effect in the 2012 June 6 Venus transit. MNRAS 429:L79–L83 Pepe F, Lovis C, Ségransan D et al (2011) The HARPS search for Earth-like planets in the habitable zone. I. Very low-mass planets around HD 20794, HD 85512, and HD 192310. A&A 534:A58 Pepe F, Cameron AC, Latham DW et al (2013) An Earth-sized planet with an Earth-like density. Nature 503:377–380 Pepe F, Ehrenreich D, Meyer MR (2014a) Instrumentation for the detection and characterization of exoplanets. Nature 513:358–366 Pepe F, Molaro P, Cristiani S et al (2014b) ESPRESSO: the next European exoplanet hunter. Astron Nachr 335:8 Probst RA, Lo Curto G, Ávila G et al (2016) Relative stability of two laser frequency combs for routine operation on HARPS and FOCES. In: Ground-based and airborne instrumentation for Astronomy VI. Proceedings of SPIE, vol 9908, p 990864. https://doi.org/10.1117/12.2231434 Queloz D, Henry GW, Sivan JP et al (2001) No planet for HD 166435. A&A 379:279–287

41 ESPRESSO on VLT: An Instrument for Exoplanet Research

901

Quirrenbach A, Amado PJ, Caballero JA et al (2016) CARMENES: an overview six months after first light. In: Ground-based and airborne instrumentation for astronomy VI. Proceedings of SPIE, vol 9908, p 990812. https://doi.org/10.1117/12.2231880 Robertson P, Mahadevan S, Endl M, Roy A (2014) Stellar activity masquerading as planets in the habitable zone of the M dwarf Gliese 581. Science 345:440–444 Saar SH, Butler RP, Marcy GW (1998) Magnetic activity-related radial velocity variations in cool stars: first results from the lick extrasolar planet survey. ApJ 498:L153–L157 Santos NC, Mayor M, Naef D et al (2000) The CORALIE survey for Southern extra-solar planets. IV. Intrinsic stellar limitations to planet searches with radial-velocity techniques. A&A 361:265–272 Santos NC, Bouchy F, Mayor M et al (2004) The HARPS survey for Southern extra-solar planets. II. A 14 Earth-masses exoplanet around  Arae. A&A 426:L19–L23 Snellen IAG, Albrecht S, de Mooij EJW, Le Poole RS (2008) Ground-based detection of sodium in the transmission spectrum of exoplanet HD 209458b. A&A 487:357–362 Snellen IAG, de Kok RJ, de Mooij EJW, Albrecht S (2010) The orbital motion, absolute mass and high-altitude winds of exoplanet HD209458b. Nature 465:1049–1051 Snellen IAG, Brandl BR, de Kok RJ et al (2014) Fast spin of the young extrasolar planet ˇ Pictoris b. Nature 509:63–65 Snellen I, de Kok R, Birkby JL et al (2015) Combining high-dispersion spectroscopy with high contrast imaging: probing rocky planets around our nearest neighbors. A&A 576:A59 Suárez Mascareño A, González Hernández JI, Rebolo R et al (2017a) A super-Earth orbiting the nearby M dwarf GJ 536. A&A 597:A108 Suárez Mascareño A, González Hernández JI, Rebolo R et al (2017b) HADES RV programme with HARPS-N at TNG. V. A super-Earth on the inner edge of the habitable zone of the nearby M dwarf GJ 625. A&A 605:A92 Suárez Mascareño A, Rebolo R, González Hernández JI, Esposito M (2017c) Characterization of the radial velocity signal induced by rotation in late-type dwarfs. MNRAS 468:4772–4781 Wilken T, Curto GL, Probst RA et al (2012) A spectrograph for exoplanet observations calibrated at the centimetre-per-second level. Nature 485:611–614 Wyttenbach A, Ehrenreich D, Lovis C, Udry S, Pepe F (2015) Spectrally resolved detection of sodium in the atmosphere of HD 189733b with the HARPS spectrograph. A&A 577:A62

SPIRou: A NIR Spectropolarimeter/High-Precision Velocimeter for the CFHT

42

Jean-François Donati, D. Kouach, M. Lacombe, S. Baratchart, R. Doyon, X. Delfosse, Étienne Artigau, Claire Moutou, G. Hébrard, François Bouchy, J. Bouvier, S. Alencar, L. Saddlemyer, L. Parès, P. Rabou, Y. Micheau, F. Dolon, G. Barrick, O. Hernandez, S. Y. Wang, V. Reshetov, N. Striebig, Z. Challita, A. Carmona, S. Tibault, E. Martioli, P. Figueira, I. Boisse, Francesco Pepe, and the SPIRou Teams

J.-F. Donati () CNRS, Institut de Recherche en Astrophysique et Planétologie, Toulouse, France e-mail: [email protected] D. Kouach · M. Lacombe · S. Baratchart · L. Parès · Y. Micheau · N. Striebig · Z. Challita · A. Carmona IRAP/OMP, Toulouse, France R. Doyon · O. Hernandez · S. Tibault UdeM/UL, Montréal, QC, Canada X. Delfosse · G. Hébrard IAP/IdF, Paris, France E. Artigau Institut de Recherche sur les Exoplanètes, Département de Physique, Université de Montréal, Montréal, QC, Canada C. Moutou CNRS/CFHT, Kamuela, HI, USA CNRS, LAM, Laboratoire d’Astrophysique de Marseille, Aix Marseille University, Marseille, France UdeM/UL, Montréal, QC, Canada F. Bouchy Département d’Astronomie, Université de Genève, Versoix, GE, Switzerland Observatoire astronomique de l’Université de Genève, Versoix, Switzerland J. Bouvier · P. Rabou IPAG, Paris, France G. Barrick CFHT, Waimea, HI, USA © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_107

903

904

J.-F. Donati et al.

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Science with SPIRou . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Planetary Systems of Nearby M Dwarfs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Magnetic Fields and Star/Planet Formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The SPIRou Legacy Survey (SLS) and the Synergy with Major Observatories . . . . . . . . . Additional Science Goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The SPIRou Science Consortium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The SPIRou Spectropolarimeter/Velocimeter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Cassegrain Unit and Calibration Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Fiber Link and Pupil Slicer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Cryogenic High-resolution Spectrograph . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Controlling the Instrument . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Data Simulator and Reduction Pipeline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Countdown to First Light and Science Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The SPIRou Project Team . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions and Prospects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross-References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

905 907 907 911 913 913 914 914 914 917 918 920 920 920 921 924 925 926

Abstract

SPIRou is a near-infrared (nIR) spectropolarimeter/velocimeter for the CanadaFrance-Hawaii Telescope (CFHT) that will focus on two forefront science topics, (i) the quest for habitable Earthlike planets around nearby M stars and (ii) the study of low-mass star/planet formation in the presence of magnetic fields. SPIRou will also efficiently tackle many key programs beyond these two main goals, from weather patterns on brown dwarfs to solar system planet and exoplanet atmospheres. SPIRou will cover a wide spectral domain in a single

I. Boisse · F. Dolon LAM/OHP, Marseille, France S. Alencar UFMG, Belo Horizonte, Brazil L. Saddlemyer · V. Reshetov NRC-H, Victoria, Canada S. Y. Wang ASIAA, Taipei, Taiwan E. Martioli LNA, Itajubá, Brazil P. Figueira CAUP, Porto, Portugal F. Pepe Département d’Astronomie, Observatoire de l’Université de Genéve, Versoix, GE, Switzerland the SPIRou Teams

42 SPIRou: A NIR Spectropolarimeter/High-Precision Velocimeter for the CFHT

905

exposure (0.98–2.44 m) at a resolving power of 70 K, yielding unpolarized and polarized spectra of low-mass stars with a 15% average throughput at a radial velocity (RV) precision of 1 m s1 . It consists of a Cassegrain unit mounted at the Cassegrain focus of CFHT and featuring an achromatic polarimeter, coupled to a cryogenic spectrograph cooled down at 80 K through a fluoride fiber link. SPIRou is currently integrated at IRAP/OMP and will be mounted at CFHT in 2018 Q1 for a first light scheduled in early 2018. Science operation is predicted to begin in 2018 S2, allowing many fruitful synergies with major ground and space instruments such as the JWST, TESS, ALMA, and later-on PLATO and the ELT.

Introduction Detecting and characterizing exoplanets, especially Earthlike ones located at the right distance from their host stars to lie in the habitable zone (HZ, where liquid water can pool at the planet surface), stands as one of the most exciting areas of modern astronomy and comes as an obvious milestone in our quest to understand the emergence of life (Gaidos and Selsis 2007). High-precision velocimetry, measuring RVs of stars and the periodic fluctuations that probe the presence of orbiting bodies, is currently the most reliable way to achieve this goal; in particular, velocimetry allows one to validate candidate planets detected with transit surveys (with, e.g., CoRoT, Kepler, TESS, and later-on PLATO) and to estimate the densities and study the bulk composition of the detected planets from their masses and radii (Lissauer et al. 2014). M dwarfs are key targets for this quest; beyond largely dominating the population of the solar neighborhood, they feature many low-mass planets (Dressing and Charbonneau 2015; Gaidos et al. 2016) and render HZ planets far easier to detect by shrinking the size of their HZs (thereby boosting RV wobbles and reducing orbital periods). Their monitoring with existing velocimeters like HARPS on the 3.6 m ESO telescope (Rupprecht et al. 2004) is however tricky, especially for the coolest ones, given their intrinsic faintness at visible wavelengths, preventing a deep-enough exploration to detect significant samples of HZ Earthlike planets (Bonfils et al. 2013). Moreover, M dwarfs are notorious for their magnetic activity, generating spurious RV signals (activity jitter) that can hamper planet detectability (Newton et al. 2016; Hébrard et al. 2016). Modeling the activity of M dwarfs and the underlying magnetic fields is thus crucial for filtering out the RV jitter and for maximizing the efficiency at detecting low-mass planets (Hébrard et al. 2016). Magnetic fields of low-mass stars are also expected to have a major impact on the evolution of close-in planets (Strugarek et al. 2015) as well as on their habitability (Güdel et al. 2014; Vidotto et al. 2013). Allowing one to detect and model large-scale fields of active stars, spectropolarimetry comes as the ideal complement to precision velocimetry, making it possible not only to maximize the efficiency of planet detection but also to characterize the impact of magnetic activity on the habitability of the detected close-in planets.

906

J.-F. Donati et al.

Investigating star/planet formation comes as the logical complement to studying exoplanetary systems of M dwarfs. Magnetic fields are known to have a major impact at the early stages of the life of low-mass stars and their planets, as they form from collapsing dense pre-stellar cores that progressively flatten into largescale magnetized accretion discs and eventually settle as young suns orbited by planetary systems (André et al. 2009). In this overall picture, the pre-main-sequence (PMS) phases, in which central protostars feed from surrounding planet-forming accretion discs, are crucial for our understanding of how worlds like our solar system form. Following a phase where they massively accrete from their discs (as class-I protostars, aged 0.1–0.5 Myr) while still embedded in dust cocoons, newly formed protostars progressively grow bright enough to clear out their dust envelopes (at ages 0.5–10 Myr), becoming classical T-Tauri stars (cTTSs) when still accreting from their planet-forming discs, then weak-line T-Tauri stars (wTTSs) once they have mostly exhausted their discs. These steps are key for benchmarking star/planet formation. Spectropolarimetry is the ideal tool for constraining the large-scale field topologies of PMS stars and their accretion discs and thereby quantitatively assessing the impact of magnetic fields on star/planet formation. ESPaDOnS and Narval, the twin high-resolution spectropolarimeters, respectively, mounted on CFHT (Donati 2003; Donati et al. 2006) and on the 2 m Télescope Bernard Lyot (TBL), already allowed to unveil for the first time magnetic topologies of PMS objects (Donati et al. 2005, 2010, 2013, 2014; Skelly et al. 2010) and to detect the youngest known hot Jupiters (hJs) to date (Donati et al. 2016, 2017; Yu et al. 2017, see Fig. 1), demonstrating that planet formation and planet-disc interaction are both quite efficient on timescales of less than 2 Myr. However, our knowledge of magnetic fields and planetary systems of PMS stars is still fragmentary, the intrinsic faintness of these objects in the visible drastically limiting their accessibility even to the most sensitive instruments. SPIRou was designed to address these two forefront issues with unprecedented efficiency (Delfosse et al. 2013; Artigau et al. 2014; Moutou et al. 2015). By operating in the nIR (including the K band), it will offer maximum sensitivity to both M dwarfs and PMS stars. Moreover, by coupling spectropolarimetry with highprecision velocimetry, SPIRou will allow us to model magnetic activity and to filter out RV curves more accurately than previously possible and thus to achieve major progress in our exploration of planetary systems of nearby M dwarfs and in our understanding of planetary formation at early stages of evolution. In particular, thanks to its widest nIR coverage among planet hunters coupled to its unique polarimetric and activity-filtering capabilities, SPIRou will be especially efficient at detecting and characterizing planets around late-M dwarfs whose high levels of magnetic activity are notorious. We describe below in more detail the science programs underlying these two prime goals, to which the SPIRou Legacy Survey (SLS) of about 500 CFHT nights will be dedicated, and mention the many other exciting programs that SPIRou will be able to efficiently tackle, thanks to its unique observational assets. We also provide a technical description of SPIRou, outline the expected performances on which our ambitious Legacy Survey relies, and summarize the overall project

42 SPIRou: A NIR Spectropolarimeter/High-Precision Velocimeter for the CFHT

907

Fig. 1 Artist view of the hot Jupiter recently detected around the wTTS V830 Tau, orbiting in the large-scale magnetic field of its host star. The field topology is reconstructed using tomographic imaging on a phase-resolved spectropolarimetric data set of V830 Tau (Donati et al. 2017)

characteristics in terms of schedule, budget, and manpower. We finally conclude with SPIRou-related prospects over the next decade.

Science with SPIRou We detail below the two main science goals to which our SPIRou Legacy Survey is dedicated; we also mention a few additional programs that SPIRou will be able to tackle and the worldwide science consortium, thanks to which SPIRou is coming to life.

Planetary Systems of Nearby M Dwarfs Much interest has recently been focused on planets of M dwarfs (Bonfils et al. 2013; Muirhead et al. 2015), with the conclusion that these stars host low-mass planets

908

J.-F. Donati et al.

more frequently than Sunlike stars do (Dressing and Charbonneau 2015; Gaidos et al. 2016). The recent discovery of a HZ planet around Proxima Cen (AngladaEscudé et al. 2016, see also dedicated section in this book) further triggered the motivation to detect and study low-mass planets and planetary systems around nearby red dwarfs. The main goals are to reveal the planet occurrence frequencies and system architectures, to investigate how they depend on the masses of the host stars (and thus on the masses and properties of the parent protoplanetary disc), and ultimately to better characterize the formation mechanism(s) that led to the observed distributions of planets and systems. Up to now, only a few such planets have been detected and characterized with RV observations, which required in particular focusing on the brightest M dwarfs as the only accessible targets for the few existing optical velocimeters capable of reaching a RV precision of 1 m s1 . This is clearly insufficient to achieve a proper statistical study of rocky exoplanets and more generally of exoplanetary systems around M dwarfs. This constraint also drastically limits our chances of detecting transiting rocky planets in the HZs of the nearest stars, i.e., the only ones for which atmospheric characterization with the JWST will be possible (Berta-Thompson et al. 2015). Carrying out an exploration of nearby M dwarfs extensive enough to detect and characterize hundreds of low-mass planets and planetary systems, whose existence is known, mandatorily requires RV observations in the nIR domain, where these stars are brightest. This is what SPIRou aims at with the SLS, concentrating the effort in two main directions, (i) a systematic RV monitoring of a large sample of nearby M dwarfs (called the SLS Planet Search) and (ii) a RV follow-up of the most interesting transiting planet candidates to be uncovered by future photometric surveys (called the SLS Transit Follow-up). In both cases, SPIRou will be observing in spectropolarimetric mode to simultaneously monitor stellar activity, unambiguously identify the rotation period (with which activity is modulated) and reconstruct the parent large-scale magnetic field triggering the activity. This will enable to implement novel and efficient ways of filtering out the polluting effect of activity from RV curves (Hébrard et al. 2016) and thus to boost the sensitivity of SPIRou to low-mass planets. This option will turn especially useful for late-M dwarfs, many of which are rather active as a result of their higher rotation rates (Newton et al. 2016) and show RV activity jitters of several m s1 (Gomes da Silva et al. 2012; Hébrard et al. 2016). The immediate objective of the SLS Planet Search is to: • identify at least 200 exoplanets with orbital periods ranging from 1 d to 1 year around stars with masses spanning 0.08–0.5 Mˇ to derive accurate planet statistics as a function of stellar mass; • identify a few tens of HZ terrestrial planets orbiting nearby M dwarfs, thanks to which we will infer a better description of the different types of planets located in HZs; • identifying several tens of multi-planet systems for studying the architecture of exoplanetary systems and their dynamical evolution;

42 SPIRou: A NIR Spectropolarimeter/High-Precision Velocimeter for the CFHT

909

Planet mass (Mearth)

10

1

0

200

400 Planet effective temperature (K)

600

800

Fig. 2 Simulation of an example SLS Planet Search assuming a sample size of 360 M dwarfs and 55 visits per star, for a total survey time of 300 CFHT nights. Filled circles indicate detected planets, open circles indicate undetected ones, and red circles (both filled and open) represent transiting planets. Blue lines show notional limits for the HZ, both in mass and temperature. Most planets more massive than 2 M˚ located in the HZ are detected, including two transiting. A sample of sub-M˚ planets hotter than 350 K is also detected. Using a smaller sample with more visits (as we now propose) improves the detectability of multi-planet systems

• identifying a large population of close-in planets to investigate how they form and interact with the magnetospheres of their host stars. Practically speaking, this implies carrying out a deep survey of at least 200 M dwarfs of different masses, with typically 100 visits per star (each yielding a spectrum with high-enough S/N to achieve 1 m s1 RV precision). Monte Carlo simulations (see, e.g., Fig. 2) suggest that the SLS planet search should detect at least 200 new exoplanets, including 150 with masses 70,000 H4RG-15 HgCdTe array, 40962 15 m pixels 306  154 mm 22 gr/mm R2 grating from Richardson-Lab 2 ZnSe prism and 1 silica prism (size 190  206 mm) 2.3 km s1 S/N = 110 per 2.3 km s1 bin at K'11 in 1 h for a M6 dwarf Circular & linear, sensitivity 10 ppm, crosstalk 20; 000, where  denotes the wavelength) operating in the visual (0.4–0.9 m), near-infrared (NIR, 1–2.5 m) and mid-infrared (3– 5 m), with a camera for direct imaging close to the diffraction limit of the telescope, and with an integral field unit operating in the visual and NIR (Table 1, adapted from Crossfield 2016).

53 Exoplanet Research in the Era of the Extremely Large Telescope (ELT)

1107

Lasers Altitude cradles for inclining the telescope

Instrument platforms sit either side of the rotatable telescope

The 2800-tonne telescope system can turn through 360 degrees

Seismic isolators

Five-mirror design 1. The 39.3-metre primary mirror collects light from the night sky and reflects it to a smaller mirror located above it. 2. The 4-metre secondary mirror reflects light back down to a smaller mirror nestled in the primary mirror. 3. The third mirror relays light to an adaptive flat mirror directly above. 4. The adaptive mirror adjusts its shape a thousand times a second to correct for distorsions caused by atmospheric turbulence. 5. A fifth mirror, mounted on a fast-moving stage, stabilises the image and sends the light to cameras and other instruments on the stationary paltform.

Fig. 1 Sketch explaining the ELT with its two instrument platforms (Credit: ESO; http://eso.org/ public/images/E-ELT_labels_5000/)

1108

F. Rodler

Table 1 Instruments for the three telescopes that will be relevant for the exoplanet research. Instruments marked with an asterisk ( ) are first-light instruments Instrument type/telescope High-resolution spectroscopy MIR High-resolution spectroscopy NIR High-resolution spectroscopy visual Direct imaging Integral field spectroscopy visual/NIR

ELT METIS HIRES HIRES EPICS HARMONI

TMT MICHI NIRES HROS PSI IRIS

GMT GMTNIRS GMTNIRS G-CLEF TIGER GMTIFS

Background Exoplanets are faint. Detecting a planet is an enormous challenge due to the brightness of the planetary system’s host star, which tends to overwhelm the relatively dim planets: the planet-to-star flux ratio for a hot Jupiter around a Sunlike star is on the order of 105 in the visual (Martins et al. 2015), while for an Earth analogue, it is on the order of 1010 . While the major constituent to the planetary flux in the visual is the reflected starlight, in the mid-infrared, it is the thermal emission coming from the planet itself. When observing in the infrared, the planet-to-star flux ratio can be drastically improved: for example, hot Jupiters glow already in the near-infrared due to their high temperatures of 1500–2000 K, which results in planet-to-star flux ratios on the order of 103 , being at least a factor of 100 better than in the visual. The thermal emission of Earth-like planets with an equilibrium temperature of 300 K peaks in the mid-infrared wavelength regime at 10 m and leads to an improved planet-to-star flux ratio of 107 with respect to the visual and near-infrared (Traub and Jucks 2002). To date, more than 3,700 planets beyond our solar system have been confirmed (Extrasolar Planets Encyclopaedia http://www.exoplanet.eu), plus more than 2,600 planet candidates waiting for confirmation. The vast majority of the confirmed exoplanets were detected by indirect detection methods: about 2,700 planets have been detected that periodically transit their planet host stars (Charbonneau et al. 2000; Henry et al. 2000), and around 700 exoplanets have been discovered by the radial velocity method (e.g., Mayor and Queloz 1995). Only about 90 exoplanet so far could be directly spotted, most of them being young giant planets which are still hot from their formation process and strongly radiate in the infrared. Direct imaging is at the verge of becoming a powerful tool for detecting and characterizing a large number of exoplanets. Instruments employing extreme adaptive optics (AO) systems and high-efficiency coronagraphs (e.g., SPHERE at ESO’s VLT; Beuzit et al. 2008) have opened a window to direct high-contrast imaging and atmospheric characterization of giant planets at wide orbital separations from their host stars (several AU). The extreme AO system corrects for the turbulences in our atmosphere and condenses the blurry seeing image of the star into a stable pointspread function with a full width half maximum (FWHM) close to the diffraction

53 Exoplanet Research in the Era of the Extremely Large Telescope (ELT)

1109

limit, thereby permitting to attain Strehl ratios >80%, and a coronagraphic mask blocks most of the starlight and facilitates the observation of companions at angular separations of a small fraction of an arc second (00 ). For the vast majority of exoplanets that appear too close to their host stars to be separated, two observing strategies have proven powerful for their atmospheric characterization: transmission spectroscopy of transiting exoplanets, phase curves and high-dispersion spectroscopy of the reflected starlight/thermal emission from the planets. Transmission spectroscopy during the passage of the planet in front of its host star is currently the most successful strategy to measure the composition of exoplanet atmospheres, as part of the starlight crosses the planetary atmosphere and the signature of chemicals therein gets imprinted in the light we measure from the star (Charbonneau et al. 2002; Snellen et al. 2010). The detectability of the spectral features in the planet atmosphere depends on the area ratio between those parts of the planet atmosphere that appear transparent (i.e., the transparent atmosphere ring surrounding the solid planetary disk) and the apparent stellar disk. For an Earthlike planet orbiting a Sun-like star, the flux ratio between the star and the light passing through the planet atmosphere is about 2  106 . However, for the same planet orbiting a red dwarf star of spectral type M, this flux ratio is drastically improved to values on the order of a few times 105 (Rodler and López-Morales 2014). Transmission spectroscopy is powerful but is limited to transiting planets and transparent atmospheres; high-altitude cloud decks and haze that block the passage of the transversing starlight may hide the spectral features in the planetary atmosphere. For short-period exoplanets (P < 20 days), high-dispersion spectroscopy with spectral resolutions larger than = D 20;000 is a robust technique to inspect the daysides of exoplanet atmospheres. The idea behind is to resolve the individual absorption lines in the planetary spectrum and monitor their radial velocity during the course of an orbit: as the planet orbits its host star, the planetary spectrum travels with respect to the stellar spectrum. The monitoring of this radial velocity change of the planetary spectrum allows us to disentangle the weak, but traveling planetary spectrum from both the dominating, but rather stationary stellar spectrum and the telluric absorption lines produced by our atmosphere (e.g., Brogi et al. 2012; Rodler et al. 2012), thereby unambiguously revealing the atomic and molecular composition of the planetary atmosphere. With high-resolution spectrographs mounted on 8 m class telescopes, this method permits to overcome a planet-to-star flux ratio of 105 and 103 , respectively, in the visual and infrared. High-dispersion spectroscopy of exoplanet atmospheres has also opened a window for the study of physical effects that impact the intrinsic shape of atmospheric spectral lines: atmospheric circulation, planetary rotation, pressure broadening, etc. In contrast to transmission spectroscopy, which requires a transparent atmosphere, high-dispersion spectroscopy is not limited by high-altitude cloud decks and haze. This technique has led to robust measurement of carbon monoxide (CO) and water vapor in the dayside spectra of hot Jupiters (e.g., Brogi et al. 2012; Rodler et al. 2012; Birkby et al. 2013).

1110

F. Rodler

The Bright Future: A Road Map for the Upcoming Years In 2018, the 2-year-lasting NASA Transiting Exoplanet Survey Satellite (TESS, Ricker et al. 2014) will commence a scan of almost the entire sky for transiting planets around relatively bright stars (magnitudes brighter than 12), permitting the determination of their masses by follow-up radial velocity observations. This satellite mission is expected to reveal thousands of planets, mostly with orbital periods less than 25 days, with hundreds of them being rocky planets residing in the habitable zone around red dwarf stars of spectral type M (more on that in the following). Being launched in 2020, the James Webb Space Telescope (JWST) with its 6.5 m primary mirror – the largest space telescope ever built – will inspect the atmospheres of transiting exoplanets and will reveal the composition of the atmospheres of superEarths and larger. The JWST will allow us to employ transmission spectroscopy of transiting exoplanets with medium spectral resolutions (=  1000) in the wavelength regime from 1 to 12 m and will largely extend our knowledge of the atmospheres of gaseous exoplanets and super-Earths. In addition, the JWST will permit high-contrast direct imaging of young Saturns and Jupiters. In 2018, the ultra-stable high-resolution spectrograph ESPRESSO (Pepe et al. 2010) which operates in the visual will become operational. This spectrograph mounted at all of the four ESO VLTs will permit the measurement of radial velocities of planet host stars with a precision of 10 cm s1 . This will not only allow us to detect and follow up Earth analogues around solar-type stars, but it will also play an important role in studying the reflectance of hot Jupiters (Martins et al. 2013). Konopacky et al. (2013) and Snellen et al. (2015) demonstrated the power of coupling direct imaging instruments with high-resolution spectrographs. The goal is to block most of the starlight with coronagraphs on the one side and to observe the planetary companion separated from its host star with high-dispersion spectroscopy on the other side. In the upcoming years, the high-contrast direct imager VLT/SPHERE might be coupled with high-resolutions spectrographs VLT/CRIRES and VLT/ESPRESSO, thereby permitting the observation of exoplanets with a planet-to-flux ratio of 108 . Lovis et al. (2017) presented feasibility studies and concluded that VLT/SPHERE combined with VLT/ESPRESSO could detect the planet Proxima b (Anglada-Escudé et al. 2016), if having a planet-to-star flux ratio of 107 in the visual, with 5 confidence. These authors found that Proxima b’s albedo could be measured in 20–40 nights of VLT telescope time, assuming an Earth-like atmospheric composition. In late 2018, ESA’s CHEOPS satellite (CHEOPS website: http://cheops. unibe.ch/) will be launched. This 30 cm telescope operating in the visual will follow up known transiting Earth-size planets, super-Earths, and hot Neptunes and refine their radius measurements. In 2019, the Breakthrough Initiatives Watch (Website: https:// breakthroughinitiatives.org/) will search the ˛ Centauri double star system, our neighbor stars at a distance of just 1.35 pc, for Earth-analogue planets. The

53 Exoplanet Research in the Era of the Extremely Large Telescope (ELT)

1111

observations will be carried out during 20 nights with the mid-infrared instrument VISIR (Lagage et al. 2004) which will be mounted at ESO’s VLT/UT4. VISIR will observe in the wavelength region between 10 and 12.5 m, where the thermal glow from an orbiting planet is expected to yield a planet-to-star flux ratio of 107 . This instrument will be enhanced by making use of AO to permit observations close to the diffraction limit of the 8.2 m telescope and will be equipped with an AGPM coronagraph (Mawet et al. 2005) to reduce starlight to reveal the atmospheric features of potential terrestrial planets (Kasper et al. 2017). The ongoing satellite mission GAIA (GAIA website: http://www.esa.int/Our_ Activities/Space_Science/Gaia), which is an all-sky survey dedicated to a precise measurement of the position, brightness, and motion of over one billion stars in our galaxy, is expected to reveal tens of thousands of exoplanets up to distances of 500 pc by tracing the astrometric reflex motion of the host star. In 2026, the PLATO mission (PLATO website: http://sci.esa.int/plato/) by the European Space Agency is foreseen to be launched. This satellite will vet the entire sky for planetary transits around solar-type stars and red dwarf stars of spectral type M. The emphasis of this satellite mission is to discover and characterize rocky exoplanets which are located in the habitable zone around their host stars, including solar-type stars. PLATO will target relatively bright stars, with magnitudes brighter than 12 (similar as TESS), thereby allowing the follow-up and atmospheric characterization of their surrounding planets. The planned lifetime of the mission is 5 years.

Selected Science Cases for the ELT Disclaimer: most of the instruments for the ELT are still in the design phase, so their technical specifications can change. In the mid-2020s, tens of thousands of worlds beyond our solar system will be known. Thus, the primary aim of the ELT will not be to detect more exoplanets but to inspect and characterize the atmospheres of the already known exoplanets.

Characterizing Potentially Habitable Planets The next decade will bring us closer to the answer of one of the most fundamental questions of mankind: does life exist beyond Earth? To date, only a few dozens of rocky exoplanets were found to reside in the habitable zone around their host star, i.e., the range of orbits where a planet residing inside it just receives the right amount of radiation from its host star to have the possibility to bear liquid water on its surface – the absolute requirement for life as we know it (for a detailed discussion on the concept of the habitable zone, see Kasting and Harman 2013; Seager 2013; Kopparapu et al. 2014). TESS, GAIA, and PLATO will reveal thousands of terrestrial planets of which hundreds will be hot candidates for life. The path to finding exoplanets either

1112

F. Rodler

inhabited or amenable for life as we know it on Earth includes the detection of greenhouse gases in their atmospheres and chemical compounds such as H2 O, CO2 , CH4 , and O3 , known as biomarkers (e.g., Schindler and Kasting 2000; Pavlov et al. 2000; Kaltenegger et al. 2007). Another important biomarker is gaseous oxygen, O2 , detected by Sagan et al. (1993) while analyzing the spectrum of Earth observed by the Galileo probe. In that spectrum, O2 appeared to be – together with CH4 in strong thermodynamical disequilibrium – a strong indicator for life on Earth. However, O2 can also be produced by abiotic processes in planetary atmospheres, e.g., by photolytic reactions with CO2 ; Meadows (2017) showed in a detailed study of O2 as a bio-signature that to search for life on a distant world, it is not enough to just detect O2 or O3 in the planetary atmosphere; other molecules, including O4 , CO, CO2 , CH4 , H2 O, and N4 must also be probed. All these gases have several spectral absorption bands at visible and infrared wavelengths, which could be detected using ground-based telescopes.

Searching for Oxygen: HIRES Most of the rocky planets found by TESS planets will have short orbital periods (a few days up to 25 days), and those orbiting M-dwarfs will likely reside in the habitable zone. As for transmission spectroscopy the most relevant aspect is the area ratio between the transparent atmosphere ring surrounding the planetary disk and the apparent stellar disk, the small M-dwarfs are excellent targets for atmospheric studies with transmission spectroscopy, with such area ratios on the order of a few times 105 . One drawback of M-dwarfs is that they are intrinsically faint; for this reason, only those potentially habitable terrestrial planets will be targets for searches for biomarkers in their atmospheres that orbit M-dwarfs in the close solar neighborhood (distances < 5 pc). Snellen et al. (2013) and Rodler and López-Morales (2014) presented feasibility studies to measure O2 absorption in the atmosphere of an Earth-like planet with high-resolution spectrographs mounted at extremely large telescopes, e.g., with HIRES, which is a proposed high-precision Echelle spectrograph for the ELT with a spectral resolving power of at least 100,000 in the wavelength range from 0.37 to 2.4 m (visual and near-infrared). To overcome a flux ratio of a few times 105 between the stellar spectrum and the spectrum of the planet atmosphere, highdispersion transmission spectroscopy is a powerful method to unveil the weak planetary signal and to resolve the atomic and molecular absorption bands in the planetary atmosphere into dense forests of individual absorption lines. Key to this method is a large number of well-resolved spectral features (e.g., absorption lines) which are traveling with respect to the stellar spectrum and the absorption spectrum from atoms and molecules present in our own atmosphere p (telluric lines). In the ideal case, the planetary signal can be boosted by a factor n, where n denotes the number of resolved spectral lines. As each atomic and molecular species produces its own, unique pattern of spectral features, this method allows an unambiguous and robust determination of the compositions of exoplanet atmospheres.

53 Exoplanet Research in the Era of the Extremely Large Telescope (ELT)

1113

transmission

1.0 0.8 0.6 0.4 0.2

flux

transmission

0.0 1.0 0.5 0.0 1.0 0.5 0.0 760

762

764 766 wavelength (nm)

768

770

Fig. 2 Top panel: transmission spectrum of the atmosphere of an Earth-like planet around 760 nm. Middle panel: telluric spectrum of our atmosphere for a zenith distance of 30ı (airmass X D 1:3). Bottom: model spectrum of an M4V star with a surface temperature of Teff D 3000 K, log g D 4:5 dex and solar abundance. All spectra are shown at a spectral resolution of 100,000. (From Rodler and López-Morales (2014))

Snellen et al. (2013) and Rodler and López-Morales (2014) concluded that indeed with a reasonable amount of observing time, it will be possible to detect the O2 A-band at 760 nm (Fig. 2) in the atmosphere of a transiting Earth-like planet orbiting an M-dwarf. However, they also pointed out that the measurement of the O2 signal would require the observations of a minimum of a few dozen transits with an instrument like HIRES. For example, it would require 21 transits and a total of 45 h of observing time to attain a 3 detection of O2 in the atmosphere of an Earthlike planet orbiting an M4V dwarf star. In addition, they remarked that in the case of stars smaller than M7V, the O2 might be easier to detect in the infrared around 1.27 m since those stars will be brighter at near-infrared wavelengths than in the visible. Rodler and López-Morales (2014) discussed the issue that from the ground the same absorption spectral features in the exoplanet atmosphere would be targeted which are also present in our atmosphere. Consequently, such observations could only be carried out when the telluric lines of our atmosphere were shifted by 20–30 km s1 due to the orbital motion of the Earth (barycentric motion) with respect to the spectral features of the planet atmosphere to be investigated. Due to this scheduling constraint, the observation of the aforementioned 21 transits would drag on over a total of 2 years in the ideal case.

1114

F. Rodler

Direct Imaging of Terrestrial Planets: METIS, EPICS The Mid-infrared ELT Imager and Spectrograph (METIS; Brandl et al. 2012) is foreseen to be the third instrument to be commissioned at the ELT. METIS will be an extreme AO-assisted imager that will permit direct imaging close to the diffraction limit in the wavelength range from 2.9 to 14 m. It will offer an 1800  1800 field of view, high-contrast coronagraphy, medium-resolution (=  5000) long-slit spectroscopy, and polarimetry. In addition, an integral field spectrograph (IFS) will provide a spectral resolution of 100;000 for a field of view of 0:400  1:500 in small wavelength chunks in the wavelength range from 2.9 to 5.3 m (atmospheric L- and M-band). Similar to the coupling SPHERE+CRIRES/ESPRESSO, METIS will combine ground-based high-contrast imaging (AO-assisted imaging with a coronagraph) with high-dispersion spectroscopy. METIS will operate in the mid-infrared and will permit the observation of the thermal emission of terrestrial planets at contrast levels star/planet on the order of 108 at angular distances of >60 milliarcseconds (mas) from the host star. Snellen et al. (2015) demonstrated that METIS would be capable to image and characterize an Earth analogue orbiting our neighbor star ˛ Centauri A in the habitable zone. These authors assumed planets with radii from 1 to 1.5 REarth , with equilibrium temperatures between 255 K (like our Earth) and 355 K, which correspond to semimajor axes between 0.63 and 1.22 AU (angular distances of 0:4800 0:9300 in the sky). These authors simulated observations of the O3 and CO2 absorption line forest in the wavelength range from 4.82 to 4.89 m and found that already with one night of observing time, it is possible to overcome the thermal background of the night sky and to detect the planetary signal of a planet with a radius of R D 1:5 REarth and a twin-Earth thermal spectrum of Teq D 300 K at a signal-to noise (S/N) of 5 (Fig. 3). Skidmore et al. (2015) studied the capability of the Thirty Meter Telescope to directly image potentially habitable exoplanets with an extreme AO-assisted coronagraphic instrument and to search for biomarkers in their atmospheres. These authors draw an optimistic image of the capabilities of direct imaging of superEarths located in the habitable zone around nearby stars: with 5 hours of observing time on a 30 m class telescope, it would be possible to attain a 5 -detection of the O2 absorption band of 1.27 m in the atmosphere of a super-Earth orbiting a nearby early M-dwarf (< 7 pc). These authors further concluded that nearby M-dwarfs are the ideal type of stars for these studies, as potentially habitable planets orbiting stars warmer than M-dwarfs would be located at larger orbital distances, which would result in a very low planet-to-star flux (Fig. 4). Likewise, possibly habitable planets orbiting M-dwarfs located further away from us than 7 pc would appear too close to their host stars to be properly separated. Given the size of the ELT, these values could be even further pushed: for the ELT, EPICS (Kasper et al. 2010) could give insights in the atmospheres of potentially habitable super-Earths via direct imaging. EPICS is a planned extreme AO-assisted high-contrast imager with high-efficiency coronagraphs similar to VLT/SPHERE. It is foreseen to become operational in the late 2020s and will permit direct imaging of companions with contrast ratios of 108 and 109 , respectively, at angular

53 Exoplanet Research in the Era of the Extremely Large Telescope (ELT)

1115

Fig. 3 The contrast curve for a 30 h observation of an Earth-like planet in a Venus orbit around ˛ Cen A with METIS (Snellen et al. 2015)

Fig. 4 Detectability of the 1.27 m oxygen absorption band in the atmosphere of an Earth-like planet with a 30 m class telescope and 5 h exposure time. The color of the filled circles corresponds to the spectral types of all possible planet host stars within 10 pc distance. The black solid line shows a theoretical best-case 5 detection limits when observing a star with an apparent J-band magnitude of 6 and assuming photon-limited performance; about 10 stars fall above this line (Skidmore et al. 2015)

distances of 30 and 100 mas, and will offer an integral field spectrograph (IFS) in the near-infrared for low-, medium-, and high-resolution spectroscopy with spectral resolutions of 100, 1500, and 20;000 in the wavelength range from 0.6 to 1.7 m.

1116

F. Rodler

The Atmospheres of Gas Planets and Super-Earths Without any doubt, the JWST will lead to a revolution in the topic of exoplanet atmospheres down to sub-Neptunes and super-Earths. JWST will offer both imaging and low-resolution spectroscopy with extreme sensitivity. However, as it will operate in the near-infrared and mid-infrared, it will have a limited lifetime of 5– 10 years due to limited fuel capacities for the cooling system. The ELT will have several major advantages over the JWST, including a 36 times larger collecting area and a six times better spatial resolution (diffraction limit of the ELT: 7 mas at 1 m, 33 mas at 5 m), thereby opening a window into highcontrast imaging of planets at orbital distances from their host stars of a fraction of 1 AU. In addition, the ELT will be equipped with high-dispersion spectrographs with spectral resolving powers of 100,000, ranging from the visible to the midinfrared wavelength regimes, thereby offering the opportunity to measure the composition, structure, dynamics, and cloud patterns of exoplanet atmospheres. In comparison to the four 8.2 m VLTs on Cerro Paranal, the ELT will permit to attain a 4–5 times higher signal-to-noise ratio in the data with the same observing time, thereby opening a window into smaller-sized planets. For a detailed discussion on possible science cases for the extremely large telescopes, see Crossfield (2016).

High-Dispersion Spectroscopy of Irradiated Hot Exoplanets: METIS, HARMONI, and HIRES Short-period giant exoplanets (P < 10 days) subjected to strong irradiation from their host stars have opened a window to exotic planet atmospheres (e.g., Sing et al. 2016). Our current knowledge of exoplanet atmospheres is almost entirely based on observations of hot gaseous planets: hot Jupiters and hot Neptunes. The ELT will offer the opportunity to probe the atmospheres of irradiated hot exoplanets with high-dispersion spectroscopy, particularly, for their: • Composition and structure • Dynamics and weather patterns HARMONI (Thatte et al. 2016) is a first-light instrument of the ELT and will offer near-infrared integral field spectroscopy with spectral resolutions of 3,500, 7,500, and 20,000 and an instantaneous wavelength coverage spanning from 0.5 to 2.4 m. Similar to the high-dispersion spectrographs METIS and HIRES, HARMONI in the 20,000-resolution mode will allow us to obtain transmission spectra of the atmospheres of transiting exoplanets down to the size of super-Earths. Beyond that, HIRES and METIS will permit high-dispersion spectroscopy to monitor Doppler shifts and line profile variations of the atmospheric features, indicating atmospheric rotation and wind patterns (cf. Snellen et al. 2010; de Kok et al. 2014; Brogi et al. 2016). Martins et al. (2013) presented feasibility studies to observe the stellar spectrum reflected from its orbiting planet with HIRES and to extract its albedo in the visual.

53 Exoplanet Research in the Era of the Extremely Large Telescope (ELT)

1117

Fig. 5 A global cloud map of the brown dwarf Luhman 16B. The rotation period of the brown dwarf is 4.9 h. (From Crossfield et al. 2014)

These authors concluded that with 10 h of observing time, it is possible to measure the reflected spectrum of a hot Neptune having a wavelength-independent gray albedo of 0.3 with 19 confidence. At infrared wavelengths, HIRES and METIS will extend the observations of atomic and molecular features in the thermal spectra from currently hot Jupiters (e.g., Brogi et al. 2012; Rodler et al. 2012) to hot Neptunes. For hot Jupiters, HIRES and METIS will even open a window to create two-dimensional weather maps of the atmospheres of hot Jupiters via Doppler imaging (cf. Crossfield et al. 2014), thereby permitting a direct measurement of the atmospheric dynamics, of cloud patches and hot spots in the atmospheres (Fig. 5).

Direct Imaging of Young Exoplanets and Mature Gas Giants: METIS, HARMONI, and EPICS High-precision astrometry conducted by GAIA will unveil tens of thousands of exoplanets, waiting to be further characterized. Owing to the size of the ELT and its diffraction limit (7–30 mas in the near- and mid-infrared), it will open a window into high-contrast direct imaging of young exoplanets and mature, cooler gas giants. HARMONI will offer high-contrast imaging down to flux ratios of 106 at angular distances >100 mas from the star; this contrast ratio might be even further boosted when employing the high-dispersion spectroscopy mode. EPICS will permit high-contrast imaging of planets with planet-to-star flux ratios of 108 at angular distances >30 mas, which translates to orbital distances of >0.3 AU at 10 pc. During their formation process, young gaseous exoplanets can reach high luminosities resulting in favorable planet-to-star flux ratios of 104 in the infrared (Mordasini et al. 2012). High-dispersion spectroscopy of these young giant protoplanets with METIS will not only allow us to learn about their atmospheric

1118

F. Rodler

composition but will also open a window into atmospheric circulation and planetary rotation (e.g., Konopacky et al. 2013; Snellen et al. 2014). In addition to that, METIS will be capable of conducting Doppler imaging of giant planetary atmospheres (cf. Crossfield et al. 2014), thereby producing two-dimensional maps of the atmospheric dynamics, including the formation and dissipation of clouds in these young atmospheres (Fig. 5). Toward cooler (500 K) and more mature gas giants, HARMONI and EPICS will permit high-contrast imaging on the order of 107 of cool gas giants at orbital distances less than 1 AU. Reflected light studies in the visual will reveal information about their albedos and atmospheric composition.

Conclusions Owing to its unprecedented size and the first-light instruments HARMONI and METIS, the ELT will be a high-contrast direct imaging machine, permitting the characterization of young exoplanets and mature and cooler giant planets in the solar neighborhood. The combination of high-contrast direct imaging with highdispersion spectroscopy, as realized in METIS, will open a window into measuring the thermal emission of terrestrial planets in the habitable zone around nearby M-dwarfs, thereby bringing us much closer to the answer of one of the most fundamental questions of mankind: does life exist beyond our planet? Being first-light instruments, METIS and HARMONI will play an immensely important role in investigating these exoplanet atmospheres in the next decade. EPICS and HIRES are instruments expected to become available in the late 2020s; once operational, EPICS will enhance the capabilities of high-contrast imaging in the visual and near-infrared wavelength regime. The high-dispersion spectrographs HIRES and METIS will permit transmission spectroscopy of the atmospheres of transiting giant planets down to rocky Earth-like planets. These instruments will furthermore allow us to inspect the day- and nightside spectra of (non-)transiting gaseous exoplanets and to produce weather maps of giant planets. Exciting times are ahead!

References Anglada-Escudé G, Amado PJ, Barnes J et al (2016) A terrestrial planet candidate in a temperate orbit around Proxima Centauri. Nature 536:437–440 Beuzit JL, Feldt M, Dohlen K et al (2008) SPHERE: a ’Planet Finder’ instrument for the VLT. In: Ground-based and airborne instrumentation for astronomy II. Proceedings of SPIE, vol 7014, p 701418. https://doi.org/10.1117/12.790120 Birkby JL, de Kok RJ, Brogi M et al (2013) Detection of water absorption in the day side atmosphere of HD 189733 b using ground-based high-resolution spectroscopy at 3.2 m. MNRAS 436:L35–L39 Brandl BR, Lenzen R, Pantin E et al (2012) METIS: the thermal infrared instrument for the EELT. In: Ground-based and Airborne Instrumentation for Astronomy IV. Proceeding of SPIE, vol 8446, p 84461M. https://doi.org/10.1117/12.926057

53 Exoplanet Research in the Era of the Extremely Large Telescope (ELT)

1119

Brogi M, Snellen IAG, de Kok RJ et al (2012) The signature of orbital motion from the dayside of the planet  Boötis b. Nature 486:502–504 Brogi M, de Kok RJ, Albrecht S et al (2016) Rotation and winds of exoplanet HD 189733 b measured with high-dispersion transmission spectroscopy. ApJ 817:106 Charbonneau D, Brown TM, Latham DW, Mayor M (2000) Detection of planetary transits across a sun-like star. ApJ 529:L45–L48 Charbonneau D, Brown TM, Noyes RW, Gilliland RL (2002) Detection of an extrasolar planet atmosphere. ApJ 568:377–384 Crossfield IJM (2016) Exoplanet atmospheres and giant ground-based telescopes. ArXiv e-prints Crossfield IJM, Biller B, Schlieder JE et al (2014) A global cloud map of the nearest known brown dwarf. Nature 505:654–656 de Kok RJ, Birkby J, Brogi M et al (2014) Identifying new opportunities for exoplanet characterisation at high spectral resolution. A&A 561:A150 Henry GW, Marcy GW, Butler RPVogt SS (2000) A transiting “51 Peg-like” planet. ApJ 529:L41– L44 Kaltenegger L, Traub WA, Jucks KW (2007) Spectral evolution of an Earth-like planet. ApJ 658:598–616 Kasper M, Beuzit JL, Verinaud C et al (2010) EPICS: direct imaging of exoplanets with the EELT. In: Ground-based and Airborne Instrumentation for Astronomy III. Proceedings of SPIE, vol 7735, pp 77352E–77352E–9. https://doi.org/10.1117/12.856850 Kasper M, Arsenault R, Käufl HU et al (2017) NEAR: low-mass planets in ˛ Cen with VISIR. Messenger 169:16–20 Kasting JF, Harman CE (2013) Extrasolar planets: inner edge of the habitable zone. Nature 504:221–223 Konopacky QM, Barman TS, Macintosh BA, Marois C (2013) Detection of carbon monoxide and water absorption lines in an exoplanet atmosphere. Science 339:1398–1401 Kopparapu RK, Ramirez RM, SchottelKotte J et al (2014) Habitable zones around main-sequence stars: dependence on planetary mass. ApJ 787:L29 Lagage PO, Pel JW, Authier M et al (2004) Successful commissioning of VISIR: the mid-infrared VLT instrument. Messenger 117:12–16 Lovis C, Snellen I, Mouillet D et al (2017) Atmospheric characterization of Proxima b by coupling the SPHERE high-contrast imager to the ESPRESSO spectrograph. A&A 599:A16 Martins JHC, Figueira P, Santos NC, Lovis C (2013) Spectroscopic direct detection of reflected light from extrasolar planets. MNRAS 436:1215–1224 Martins JHC, Santos NC, Figueira P, et al (2015) Evidence for a spectroscopic direct detection of reflected light from 51 Pegasi b. A&A 576:134 Mawet D, Riaud P, Absil O, Surdej J (2005) Annular groove phase mask coronagraph. ApJ 633:1191–1200 Mayor M, Queloz D (1995) A Jupiter-mass companion to a solar-type star. Nature 378:355–359 Meadows VS (2017) Reflections on O2 as a biosignature in exoplanetary atmospheres. Astrobiology 17:1022–1052 Mordasini C, Alibert Y, Georgy C et al (2012) Characterization of exoplanets from their formation. II. The planetary mass-radius relationship. A&A 547:A112 Pavlov AA, Kasting JF, Brown LL, Rages KA, Freedman R (2000) Greenhouse warming by CH4 in the atmosphere of early Earth. J Geophys Res 105:11,981–11,990 Pepe FA, Cristiani S, Rebolo Lopez R et al (2010) ESPRESSO: the Echelle spectrograph for rocky exoplanets and stable spectroscopic observations. In: Ground-based and airborne instrumentation for astronomy III. Proceedings of SPIE, vol 7735, p 77350F. https://doi.org/ 10.1117/12.857122 Ricker GR, Winn JN, Vanderspek R et al (2014) Transiting exoplanet survey satellite (TESS). In: Space telescopes and instrumentation 2014: optical, infrared, and millimeter wave. Proceeding of SPIE, vol 9143, p 914320. https://doi.org/10.1117/12.2063489 Rodler F, López-Morales M (2014) Feasibility studies for the detection of O2 in an Earth-like exoplanet. ApJ 781:54

1120

F. Rodler

Rodler F, Lopez-Morales M ,Ribas I (2012) Weighing the non-transiting hot Jupiter  boo b. ApJ 753:L25 Sagan C, Thompson WR, Carlson R, Gurnett D, Hord C (1993) A search for life on Earth from the Galileo spacecraft. Nature 365:715–721 Schindler TL Kasting JF (2000) Spectral Signatures of Biomarker Gases in Simulated Earth-Like Atmospheres. In: Schürmann B (ed) Darwin and astronomy : the infrared space interferometer. ESA special publication, vol 451. p 159 Seager S (2013) Exoplanet habitability. Science 340:577–581 Sing DK, Fortney JJ, Nikolov N et al (2016) A continuum from clear to cloudy hot-Jupiter exoplanets without primordial water depletion. Nature 529:59–62 Skidmore W, TMT International Science Development Teams Science Advisory Committee T (2015) Thirty meter telescope detailed science case: 2015. Res Astron Astrophys 15:1945 Snellen I, de Kok R, Birkby JL et al (2015) Combining high-dispersion spectroscopy with high contrast imaging: probing rocky planets around our nearest neighbors. A&A 576:A59 Snellen IAG, de Kok RJ, de Mooij EJW, Albrecht S (2010) The orbital motion, absolute mass and high-altitude winds of exoplanet HD209458b. Nature 465:1049–1051 Snellen IAG, de Kok RJ, le Poole R, Brogi M, Birkby J (2013) Finding extraterrestrial life using ground-based high-dispersion spectroscopy. ApJ 764:182 Snellen IAG, Brandl BR, de Kok RJ et al (2014) Fast spin of the young extrasolar planet ˇ Pictoris b. Nature 509:63–65 Thatte NA, Clarke F, Bryson I et al (2016) The E-ELT first light spectrograph HARMONI: capabilities and modes. In: Ground-based and airborne instrumentation for astronomy VI. Proceedings of SPIE, vol 9908, p 99081X. https://doi.org/10.1117/12.2230629 Traub WA, Jucks KW (2002) A possible aeronomy of extrasolar terrestrial planets. Geophysical monograph series, vol 130. American Geophysical Union, Washington, DC, p 369

Section VI Space Missions for Exoplanet Research Malcolm Fridlund

Malcolm Fridlund received his PhD in Astrophysics at the University of Stockholm in 1987. He was hired as a postdoc by ESA/ESTEC in 1988 and as a staff member in 1989. Until 2013 he worked as study scientist on many missions including Darwin and PLATO and as ESA’s project scientist on the CoRoT mission. He holds a professorship at Leiden Observatory, Netherlands, and an affiliated professorship at Chalmers University of Technology, Sweden. Presently, he is also a member of the Science Team of the CHEOPS mission. His research is focused on space observations of exoplanets.

Space Missions for Exoplanet Research: Overview and Introduction

54

Malcolm Fridlund

Contents Introduction: The Case for Space-Based Observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The First Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phase Two: Development and Deployment of Facilities in Space . . . . . . . . . . . . . . . . . . . . . CoRoT and Kepler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Further Developments in Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Where Do We Go from Here? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1123 1125 1126 1129 1130 1133 1133

Abstract

This chapter makes a brief summary of the space-based research in exoplanetology during the last 25 years that is described in detail in the other chapters of this section. In this period, research from space has made important if not the most important advances to this field. These activities are destined to continue during the next 10–20 years with promises of eventually leading up to research placing the Earth and the life on it into the greater context of the Universe we live in.

Introduction: The Case for Space-Based Observations Already in antiquity, approximately 2,500 years ago, philosophers were discussing the nature of the lights in the sky called stars and planets. Many thinkers considered it likely, not only that some of them were worlds like the solid Earth they were M. Fridlund () Leiden Observatory, Leiden, The Netherlands Department of Space, Earth and Environment, Chalmers University of Technology, Onsala, Sweden e-mail: [email protected]; [email protected]; [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_77

1123

1124

M. Fridlund

standing on but also that there would be some of them that would be inhabited by – as Epikuros (b. 341 BC) stated – “life forms, some of them like us and some unlike us.” The search for and study of such objects, planets orbiting stars other than our Sun or “exoplanets,” can trace its modern phase to the publication of the first detailed proposal (or plan) for a “proper” (i.e., feasible) search to a paper by O. Struve (1952). In this 1 ½ page paper, Struve suggested observations of exoplanetary transits, noting that the newly developed detectors based on photomultiplier tubes could easily detect the 1% diminishing in the light output of the Sun observed at interstellar distances during a transit of Jupiter. He also expected the radial velocity detection of large planets in short period orbits, based on the mass spectrum of contact binary stars. The radial velocity signature of such objects would be observable, in principle, with spectrographs available in 1952. Nevertheless, it was much subsequent to that paper that the first discoveries of bodies with masses comparable to those found within our Solar System followed during the late 1980s and in the beginning of the 1990s (Latham et al. 1989; Campbell et al. 1988). Most spectacular in this context was the detection of Earthmass planets orbiting pulsars by Wolszczan and Frail (1992). It was generally assumed that most if not all solar systems would look exactly as our own, i.e., small rocky planets distributed in the inner, hotter areas of the system, followed by gas giants in orbits further out. There were physical explanations for assuming this (such as gas giants having to form further out in the system in cooler regions), but the major reason was a paradigm going back to Copernican times that our Solar System and all it contained could not be special in any way but must conform to an average. The results so far, during the last 20C years of exoplanetary research, are instead indicative of an incredible diversity in structure and physical conditions found in exoplanetary systems. This diversity, quite contrary to the opinion quoted above about all exosystems being similar to our own, guides us toward the next steps needed to be carried out in order to make further progress. And there is great interest in such results. The extensive interest in exoplanets is of course related to ourselves. “Are we alone in the Universe” or is it filled by a multitude of life forms? Since we do believe that the most likely place for life is on or close to a planetary surface, the discovery and study of exoplanets became paramount, and it has regularly been listed as one of the two or three major issues for space science during the twentyfirst century (e.g., “Cosmic Vision: Space Science for Europe 2015–2025,” ESA BR-247 October 2005). It is obvious with such an important scientific issue that all possible tools and methods should be brought to bear on it, and indeed very large efforts have been made and are being carried out as this is written. What about resources deployed in space? Given the complexity, cost, and scale of efforts in order to launch the simplest of instruments into space, one really has to address the requirement to carry out such an effort. This section of the Exoplanet handbook describes many past, present, and future space projects, and it is important to understand this context. It may appear

54 Space Missions for Exoplanet Research: Overview and Introduction

1125

obvious that an instrument in space should be the preferred choice for any research into the physics of exoplanets, but this is not necessarily so. For instance, in the case of when we want to carry out the radial velocity measurements of exoplanetary host stars in order to determine the masses of the planetary bodies, the effort of going to space is not justified by the improvement in the accuracy of the results. On the contrary, the technical requirements for a high-precision spectrograph are such that they cannot easily be met on a space platform. In this case it makes more sense to construct the instrument on the ground where such requirements can be more easily met. Other techniques do require a space environment. As was pointed out by W. Borucki during the 1980s (see below and Chap. 56, “Space Missions for Exoplanet Science: Kepler/K2”), the photometric requirements (precision, stability, long durance, and high cadence) could not at all or only with the greatest difficulty be met from the ground. Also as noted in  Chap. 62, “Space Missions for Exoplanet Science: PLATO” the precision and cadence required photometrically when observing the so-called p-modes (acoustical variations in the surface layers of stars and penetrating to large depths resulting in micro variations in the light output) in solar-type stars do require deployment of the detector to space. And it is just these techniques that will be required in order to study more systems in detail so that we can find out how and when planetary systems form and how they evolve. In order to make direct comparisons with the data from our own Solar System, we also require data on each exosystems’ age at a precision so far unheard of. Also here asteroseismology and the study of the p-modes will come to our aid, and again access to space will be required.

The First Phase The époque which we can designate the first phase of exoplanetary discovery begins with the seminal paper by Mayor and Queloz (1995). Based on radial velocity (RV) observations of the line-of-sight movements of a solar-type star, they could report a bona fide Jupiter-size planet. This object is orbiting the solar-type star 51 Pegasi located about 50 light-years away from the Sun. This latter planet was found to have a minimum mass of about 0.5 Jupiter masses and an orbital period of 4.23 days equivalent to an orbital distance of 0.05 astronomical units. The discovery of such a planet with these characteristics was unexpected by the scientific community, with the exception of Struve’s prediction in 1952 who had first speculated on the existence of such “hot Jupiters,” based on the existence of contact binary stars. The discovery of 51 Pegasi b changed the paradigm that systems must look like ours and introduced a new phase, extending until the year 2000, that was going to see an increased number of planets discovered by the RV technique. A number of the first detected planets were similar to 51 Peg b, i.e., very massive planets in short period orbits, and were dubbed “hot Jupiters.” Of course, these objects were the ones that were most easily found with the RV technology, and as this

1126

M. Fridlund

observational method improved, the percentage of “hot Jupiters” began to fall. As the total number of planets increased, the effort to find smaller planets also took on a larger importance, with the design and deployment of the HARPS spectrograph at the European Southern Observatory’s 3.6 m telescope in Chile (Pepe et al. 2000). It should be remembered that all interpretations of the periodic RV changes of the host star, as caused by the orbital movement of an exoplanet, were challenged at this time. As a consequence of the details of the method, the interpretation of the RV signatures as planetary objects was questioned by a number of scientists (e.g., Gray 1997; Gray and Hatzes 1997; Hatzes et al. 1997), who justly pointed out that inherent activity and stellar oscillations could explain the radial velocity signatures as well. This was especially true when periodic changes in the shape of the spectral line profiles could also be detected. More data, however, kept pointing toward an exoplanetary signature (Gray 1998). The issue was finally settled by the detection of a planet transiting its host star (Charbonneau et al. 2000). There should be a 1–5% probability of detecting an exoplanet transiting its star depending on the distance. There were about 70 planets that had been detected with the RV method when the first transit (of HD 209458) was reported, consistent with such statistics. With such statistics it was very clear that only by observing large numbers of stars for long uninterrupted times one could detect planetary transits. From the light curve of the transit, it would then be possible to derive the physical parameters of both planet and host star (see, e.g., Seager and Mallén-Ornelas 2003). While a large number of ground-based facilities (networks) were immediately begun to be developed in order to find large numbers of exoplanetary transits, it was clear from the beginning that the detection of small planets was going to be very difficult from the ground. Here facilities deployed in space were already seen to have a large advantage, as had been long advocated by W. Borucki. He had realized in the 1980s (e.g., Borucki and Summers 1984) that the detection of an earth-size object transiting a solar-type star would require a telescope deployed in space. The “dip” in light being 1/10000th of a magnitude was 1–2 orders of magnitude more challenging than possible at the time, which induced Borucki to begin carrying out a number of experiments at NASA Ames which showed promise. This resulted in a sequence of unsuccessful proposals from Borucki during the beginning of the 1990s which were going to parallel the efforts – also unsuccessful – in Europe (e.g., STARS). But these efforts were going to eventually be crowned with success because they were going to lead to the CoRoT and Kepler missions. It was also realized that the verification of the planetary nature of such (small) objects by the detection of the RV signature was going to be very challenging.

Phase Two: Development and Deployment of Facilities in Space What is interesting is that space research played a prominent role in the discovery and study of exoplanets almost from the beginning of the observational phase of this topic. This was due to the fact that, based on the structure of our own Solar

54 Space Missions for Exoplanet Research: Overview and Introduction

1127

System which was assumed to be a “run-of-the-mill” system, it was considered to be necessary to deploy very large instruments in space in order to be able to discover “true” planets. So, interestingly enough, as ground-based scientists were obtaining the data (collecting RV observations of solar-type stars) that would result in the discovery of the first exoplanets, suggestions to search for exoplanets from space had already been made to some major space agencies, i.e., the US National Aeronautics and Space Administration (NASA) and the European Space Agency (ESA). These proposals could trace their origin to repeated workshops during the 1980s that dealt with the possibilities of carrying out optical interferometry from space platforms or even the moon. Here the detection of exoplanets was considered one of the prime objectives. At the same time, one was also considering what kind of telescope would be the follow-up to the Hubble Space Telescope (HST) that was then under development and launched in 1990. Here very large optical telescopes (maybe 20 m diameter collecting area) were considered, in part because of the requirements of being able to detect and study exoplanets (not discovered at the time). The scientific objectives of such technology was deemed important enough to include space interferometry in the technology elements of science plans such as the Horizon 2000 program of ESA and intended for the next-generation space missions to be carried out in the new millennium after the required technology had been developed and verified. As a result of these exercises, already in 1994, ESA received specific proposals to design a “nulling” interferometer, known as “Darwin,” deployed in an orbit close to that of the planet Jupiter in order to search for transiting exoplanets around nearby solar-type stars (Léger et al. 1995). In the meantime, ESA had also received a proposal updating the ongoing study of the so-called STARS mission, originally dedicated to carry out asteroseismological observations of stars, to simultaneously carry out a search for exoplanetary transits (Badiali et al. 1996). Asteroseismology was then a new topic where one can study deep layers of the Sun (called helioseismology) or stars by measuring the miniscule (a few parts per million) variations in light output of a star caused by acoustical waves traveling deep into the star, and after being reflected there emerges on the surface leading to wave patterns (with an amplitude of a few meters) there. The study of the proposed STARS mission took place at roughly the same time as scientists in the USA were beginning to study what would 1 day fly in space as the Kepler mission (e.g., Koch et al. 1996). The requirements on the STARS mission in order to add the topic of planetary transits were minimal, but the STARS study was terminated in the spring of 1996 when the project was replaced by what was to become ESA’s Planck mission studying the cosmic background radiation. Simultaneously, however, the CoRoT mission (see  Chap. 55, “CoRoT: The First Space-Based Transit Survey to Explore the Close-in Planet Population”) was being considered by the French space agency CNES. Also here a mission from the beginning dedicated to asteroseismology was seen as being able to simultaneously search for the signatures of exoplanetary transits in the light curves. As STARS was discontinued, ESA began an assessment study in the fall of 1996 of what was originally called the InfraRed Space Interferometer or IRSI. The

1128

M. Fridlund

proposed name “Darwin” was – temporarily as it turned out to be – put to the side, because it was felt that the study should begin with a critical review of both the technology that could best address the scientific issues and those issues themselves. And those issues were ambitious. The task at hand was defined as detecting a planet of the same size as the Earth itself, orbiting an early G-type main sequence star at distances up to 30 pc. The distance was specified by the requirement of having a large enough sample of truly solar-like stars to observe. And, after detecting such a planet, the mission was also required to obtain a spectrum with enough resolution and sensitivity to detect an Earth-like atmosphere and, most important, judge whether there were so-called biomarkers – indicators of biological activity – present. After the initial phase of the study, it was decided that the technology best responding to the requirements was indeed a version of the one proposed as “Darwin” but with significant differences. The original interferometer that used telescopes mounted on booms which also carried light pipes to transmit the light from the individual telescope units to a central beam combiner was replaced with telescopes free flying in space. And, the initially conceived orbit of the interferometer in the vicinity of Jupiter was found to be unnecessary. The origin of that idea had been that the zodiacal dust in the vicinity of the Earth (at a temperature of 300 K) would produce a background signal literally drenching the light expected from an exo-Earth. Careful calculations demonstrated that observations in the antiSun direction would negate this problem to a large extent. Since one of the major problems in obtaining a spectrum of a terrestrial exoplanet is the contrast between the host star which is more than 109–10 times brighter than the planet, the interferometer would be based on a so-called ‘nulling’ configuration. This means that the array is using delays in the various combining beams such that destructive interference takes place on the optical axis. However, rays originating from a point offset a small angle from the optical axis would experience constructive interference. The beams from the various telescopes would add together. The end result would be that the light from the host star would be suppressed, while the light from any planet orbiting at a certain distance (depending on the actual separation of individual telescopes in the interferometer – see Chap. 59, “Interferometric Space Missions for Exoplanet Science: Legacy of Darwin/TPF”) would be enhanced. Already in 1996/1997, contact between ESA and NASA was established in this context, leading to a successful collaboration between Darwin and the NASA parallel study designated as the Terrestrial Planet Finder (TPF – see Beichman et al. 2002). These joint activities lasted until 2007 when both projects where postponed indefinitely. Other architectures were and have been continued to be studied by NASA. Chief among these are large coronographic telescopes or telescopes using a distant (tens of thousands kilometers away from the primary mirror) occulting disks (see  Chap. 64, “Future Exoplanet Space Missions: Spectroscopy and Coronographic Imaging”). The coronographic option is currently included in the Design Reference Mission (DRM) for NASA’s WFIRST mission.

54 Space Missions for Exoplanet Research: Overview and Introduction

1129

CoRoT and Kepler At the same time, other space-based exoplanetary projects were pursued such as CoRoT in France by the French space agency CNES – the acronym stands for COnvection ROtation et Transits planetaires (see  Chap. 55, “CoRoT: The First Space-Based Transit Survey to Explore the Close-in Planet Population”) – and Eddington (a successor to STARS studied by ESA). Both were telescopes intended to survey areas of the sky for long periods of time hoping to catch transiting exoplanets of small sizes. Both were also in various aspects intended to study asteroseismological variations in stars (including the exoplanetary host stars) in order to make major breakthroughs in fundamental stellar physics. Eddington was discontinued in 2003, but the development of CoRoT continued, and it was successfully launched on 27 December 2006. From then on, it operated successfully for the next 6 years, discovering more than 40 confirmed exoplanets, as well as roughly equal numbers of candidates that have not been finally confirmed with ground-based RV detections (due to the relative faintness of the host stars). In 2009, the CoRoT team published the discovery of the first rocky exoplanet, CoRoT-7b, with an exactly measured mass and diameter (Léger et al. 2009; Queloz et al. 2009). This allowed the determination of a precise average density, opening up the detailed study of the physics of small (Earth-like) planets. That planet itself turned out to be, however, not very Earth-like in that although it orbits a solar-type star, it does so in an orbit taking only 20 h to make a full revolution and has a dayside temperature of more than 2000 K. Of extreme importance as well was CoRoT’s detection and characterization, from space, of asteroseismic (acoustic p-mode) variations in large numbers of red giant stars (Hekker et al. 2008, 2009; Miglio et al. 2009) providing a tool for the study of the evolution of our Galaxy, as well as gravitational g-mode variations in very massive objects (e.g., Degroote et al. 2010). Halfway through CoRoT’s mission, it was joined by its NASA colleague Kepler (Borucki et al. 2009). This mission made the next large breakthrough in exoplanetology and initiated the next phase by detecting several thousands of (mainly) small exoplanets (see Chap. 56, “Space Missions for Exoplanet Science: Kepler/K2”) within the same very large field of view that was observed for a total of 4 years (between May 2009 and May 2013). This has allowed the first proper statistical studies to be performed, and important aspects of exoplanets have been noted. As usual the mission poses more questions, and it is now becoming very clear in what direction one should proceed: Kepler detected many thousands of small planets, a number of which are located in the habitable zone of their stars. A difficulty lies, however, in the follow-up of these planets, particularly the determination of the stellar parameters with the same precision. Also for the smaller planets (one to several earth-masses), it is very difficult or impossible to detect any radial velocities from ground-based follow-up. What can and should be done is to obtain better stellar data, as well as observe brighter stars which will make it possible to detect radial velocities of their small planets. This will be done by the next

1130

M. Fridlund

generation of space missions, namely, TESS, CHEOPS ( Chap. 60, “CHEOPS: CHaracterizing ExOPlanets Satellite”), and PLATO ( Chap. 62, “Space Missions for Exoplanet Science: PLATO”). In May 2013, the second of the originally four stabilizing reaction wheels onboard Kepler failed, which effectively put a stop to the basic mission. As described in the separate chapter on the Kepler mission, the mission has been retargeted and is now designated as K2 and observes different fields of view every 80 days, where the mission has found further about 500 candidates of which so far 150 have been confirmed as bona fide exoplanets.

Further Developments in Space There have been, of course, contributions to exoplanet science from space missions that were not specifically designed for it. After the first transiting exoplanet had been recorded from the ground with a 10 cm aperture (Charbonneau et al. 2000), its light curve was also acquired using the Hubble Space Telescope (Brown et al. 2001). The Canadian satellite MOST detected the transit of the planet 55 Cancri e in 2011 at a period similar to that of CoRoT-7b and Kepler 10-b and determined its radius to be about two Earth radii (Winn et al. 2011). Spitzer, formerly the Space InfraRed Telescope Facility (SIRTF), was launched in 2003 and designed for a mission lifetime of 2.5 years extendable as long as the liquid He on board had not boiled off. This event took place after 6 years in the year of 2009 after which the instruments designed for longer wavelengths (>5 m) were no longer available. Since the telescope and the focal plane stabilized at a temperature of 29 K, the two shortest channels at 3.6 and 4.5 m remain operational, and thus the telescope was given an extended lifetime with a more focused set of objectives, including exoplanetology. Here this mission has contributed extensively (see  Chap. 61, “Observing Exoplanets with the Spitzer Space Telescope”). All these missions have demonstrated the value of observing from space, particularly as what concerns the detection of exoplanetary transits. The space environment allows for very detailed observations and all should be fine but there is a significant problem. The precision with which we can determine the planetary diameters using space-based telescopes is exquisite, but the planetary diameter will be expressed in terms of the stellar diameter. The same is true for the planetary masses that are mainly acquired from ground-based RV observations. Here the mass is expressed in terms of the assumed stellar mass. And we have again the same problem – the precision of the planetary parameters is dependent on the precision of the stellar ones. The mass-radius relation for stars is also problematic for just the kind of stars that we are especially interested in, namely, stars similar in mass or smaller than our Sun, given their special interest in the search for life in the Universe. But if we want to make systematic studies of the small planets (presumably “rocky”) that we are beginning to find in increasing numbers, we need to have

54 Space Missions for Exoplanet Research: Overview and Introduction

1131

accurate values of the planetary masses and radii. The current data for stellar masses and radii in the context of exoplanetology is unknown to at least factors >10–20%. Also of particular interest when studying exoplanets is to determine the stage of evolution of the system in question. For main sequence planet-hosting stars, it is not unusual to find uncertainties similar to the time during which all of life’s history on the Earth is contained ( 4Gyr). Therefore, data such as the average density of exoplanets contain large uncertainties, something which is especially important when analyzing data where one is trying to draw conclusions about the geophysics of the planet in question. As spectroscopy (ground based and space based) is progressing, it will also become more and more important to be able to determine accurate masses and radii in order to properly interpret such data. The most important next step in exoplanetology is therefore to obtain higher quality data. This is going to happen in a number of steps – of which the spacebased projects are described in chapters in this section. So, at the moment, the status of the acquired data is that the precision with which we learn about the planetary parameters is mainly limited by the knowledge of our data. Albeit we have acquired a large mass of interesting – indeed exquisite – data, there are limits to the progress with the current ones. We do need both higher quality data as well as different data. The obvious way to proceed is to obtain more and better data of the kind which we already have. Observing planetary transits around brighter stars than hitherto observed are going to be the first stage, and there are two space missions in immediate development. NASA’s TESS and the Swiss-ESA CHEOPS mission (see  Chap. 60, “CHEOPS: CHaracterizing ExOPlanets Satellite”). The former of these two missions (to be launched in 2019) is targeted toward the discovery of an order of magnitude more than the small planets of the same type as those discovered previously by CoRoT, Kepler, and K2 but around stars several magnitudes brighter than what those missions observed. It does so by observing nearly the whole sky for periods similar to what CoRoT or K2 did. CHEOPS on the other hand will observe already discovered transiting planets individually, and essentially only very bright stars. Within the immediate time frame, there are also the improving results from the Gaia (ESA) astrometric mission (see  Chap. 58, “Space Astrometry Missions for Exoplanet Science: Gaia and the Legacy of Hipparcos”). As well as discovering larger exoplanets from the astrometric “wobble” of host stars as they are moving across the sky, the photometry of the mission will also find a large number of transits. And the much improved precision in stellar distances (for literally billions of stars) will also improve the precision in planetary parameters for large numbers of exoplanets. In the same time frame, the James Webb Space Telescope (JWST) will at last be launched (see  Chap. 61, “Observing Exoplanets with the James Webb Space Telescope”). This massive (6.5 m main mirror) passively cooled telescope has as one of its key objectives “planetary systems and the origin of life” which can be taken as meaning the obtaining of the first high-sensitivity spectroscopic observations of exoplanetary atmospheres.

1132

M. Fridlund

Finally, ESA has begun the construction of PLATO (see  Chap. 62, “Space Missions for Exoplanet Science: PLATO”). This mission is filling ESA’s M3 (median-sized mission) slot within the Cosmic Vision program. It was first proposed to ESA in 2007. After its adoption into ESA’s launch program in the summer of 2017, it is now expected to be launched in 2026. The key element of the PLATO mission is the goal of achieving systematically asteroseismological data from the hosts star at the same time as detecting the light curves of transiting exoplanets, including a significant number of Earth-like ones. Although TESS will obtain some asteroseismic data for very bright stars, it will fall on PLATO to carry out a large systematic study of exoplanetary host stars. It is expected that PLATO, together with the Gaia parallaxes, will improve the values for fundamental stellar parameters such as mass, radius, and age by orders of magnitude. This will allow the analysis of the exoplanetary fundamental parameters to result in a similarly higher precision. In the decade of 2020–2030, we can thus expect a number of space missions to deliver new fundamental results. It is even possible that spectroscopic data will exist for (super-)Earth-like planets that will result in a first set of atmospheric parameters. This would require the detection of atmospheres of nearby terrestrial-type planets if such atmospheres exist. Possible targets have already been found, e.g., the Trappist1 system (Gillon et al. 2017), or tentatively identified, e.g., Proxima Centauri b (Anglada-Escudé et al. 2016). It is clear from the design of the presently planned space-based assets that success in this area will depend very critically on the properties of the targets (O’MalleyJames and Kaltenegger 2017). We do not yet know enough about the properties of exoplanets, especially when it concerns smaller (and thus presumably terrestrial) objects. The indications from the Kepler mission is that such planets are common, but we do not yet know how common they are in the HZ as a function of stellar masses and other parameters. And of course we still do not have a good enough sample of such objects orbiting stars similar to our Sun. Instead, the hope is now that we will be able to obtain spectra of the atmospheres of planets such as those in, e.g., the Trappist-1 system, which is host stars with masses significantly lower than that of our Sun. But the similarities of such objects to solar system terrestrial planets may be only superficial. Systematic studies of terrestrial planets in the HZ orbits around stars more similar to our own Sun will require more ambitious projects. As can be discerned from chapters in this section, there have been studies of very ambitious projects requiring interferometric systems, large telescopes (8 m class) with sophisticated coronographic masks, or even telescopes with occulting disks (blocking disturbing light from the host star) flying up to 50,000 km away from the telescope, while maintaining perfect alignment. While such a sophisticated coronographic option is included on WFIRST, there are currently no firm plans for developing and deploying an instrument with a free-flying occulter, and they are therefore far away on the planning horizon. It is clear that we will need the results from space missions like CHEOPS, TESS, JWST, and PLATO and possibly from relatively small dedicated spectroscopic missions (e.g., the ARIEL mission (https://ariel-spacemission.eu/) that is currently being evaluated by ESA for the M4 slot) before plans for systems like the ones mentioned above can be carried out

54 Space Missions for Exoplanet Research: Overview and Introduction

1133

without unacceptable risk. Also observations with very large (20–40 m collecting aperture diameters) ground-based telescopes will contribute in this field over the next 15 years. But the good news is that these preparing studies are being carried out, and the next generation of observing facilities is being constructed. We may therefore hope that the next 25 years will see results that may begin to answer the fundamental questions about the commonality of the planet Earth and all those life forms inhabiting it.

Where Do We Go from Here? The ultimate objective of studying exoplanets is closely connected to the issue of life itself – its presence here and in the rest of the Universe. Is the Earth unique and life itself rare or even possibly nonexistent anywhere else than here on our own planet? Or is it as some have speculated present everywhere if the conditions are right? These issues will probably not be answered within the near future nor by the exciting future space projects described in this section (see, however,  Chap. 150, “Future Exoplanet Research: Science Questions and How to Address Them” and Others in the Section on Future Exoplanet Research). Only by the greatest of luck will, for instance, the JWST detect an atmosphere on an exoplanet carrying the signs of biological activity. But what is important is that we have finally begun to address these issues in a scientific way. The technology available is improving continuously and rapidly approaching the precisions required. After millennia of speculation, such instruments that could detect traces of life at interstellar distances could be made available within the not too distant future. And this is what makes the current science so exciting.

References Anglada-Escudé G, Amado PJ, Barnes J, Berdiñas ZM et al (2016) A terrestrial planet candidate in a temperate orbit around Proxima Centauri. Nature 536:437 Badiali M, Catala C, Favata F, Fridlund M et al (1996) STARS – Seismic Telescope for Astrophysical Research from Space. Report on the phase A study. http://adsabs.harvard.edu/abs/1996star.book.....B Beichman CA, Coulter DR, Lindensmith C, Lawson PR (2002) Selected mission architectures for the terrestrial planet finder (TPF): large, medium, and small. SPIE 4835:115 Borucki WJ, Summers AL (1984) The photometric method of detecting other planetary systems. ICAR 58:121 Borucki WJ, Koch D, Batalha N, Caldwell D et al (2009) KEPLER: search for Earth-size planets in the habitable zone. IAUS 253:289 Brown TM, Charbonneau D, Gilliland RL, Noyes RW, Burrows A (2001) Hubble space telescope time-series photometry of the transiting planet of HD 209458. ApJ 552:699 Campbell B, Walker GAH, Yang S (1988) A search for substellar companions to solar-type stars. ApJ 331:902

1134

M. Fridlund

Charbonneau D, Brown TM, Latham DW, Mayor M (2000) Detection of planetary transits across a Sun-like Star. ApJ 529:L45 Degroote P, Aerts C, Baglin A, Miglio A et al (2010) Deviations from a uniform period spacing of gravity modes in a massive star. Nature 464:259 Gillon M, Triaud AHMJ, Demory BO, Jehin E, Agol E et al (2017) Seven temperate terrestrial planets around the nearby ultracool dwarf star TRAPPIST-1. Nature 542:456 Gray DF (1997) Absence of a planetary signature in the spectra of the star 51 Pegasi. Nature 385:795 Gray DF (1998) A planetary companion for 51 Pegasi implied by absence of pulsations in the stellar spectra. Nature 391:153 Gray DF, Hatzes AP (1997) Non-radial oscillation in the Solar-temperature Star 51 Pegasi. ApJ 490:412 Hatzes AP, Cohran WD, Johns-Krull CM (1997) Testing the planet hypothesis: a search for variability in the spectral-line shapes of 51 Pegasi. ApJ 478:374 Hekker S, Barban C, Kallinger T et al (2008) Solar-like oscillations in red giants in the CoRoT exofield. CoAst 157:319 Hekker S, Kallinger T, Baudin F et al (2009) Characteristics of solar-like oscillations in red giants observed in the CoRoT exoplanet field. A&A 506:465 Koch D, Borucki W, Cullers K et al (1996) System design of a mission to detect Earth-sized planets in the inner orbits of solar-like stars. JGR 101:9297 Latham DW, Stefanik RP, Robert P, Mazeh T, Mayor M, Burki G (1989) The unseen companion of HD114762 – a probable brown dwarf. Nature 339:38–40 Léger A, Puget JL, Mariotti JM, Rouan D, Schneider J (1995) DARWIN: an IR space observatory with interferometric rejection to search for primitive life on extra-solar planets. Ap&SS 223:172 Léger A, Rouan D, Schneider J et al (2009) Transiting exoplanets from the CoRoT space mission. VIII. CoRoT-7b: the first super-Earth with measured radius. A&A 506:287 Mayor M, Queloz D (1995) A Jupiter-mass companion to a solar-type star. Nature 378:355 Miglio A, Montalbán J, Baudin F et al (2009) Probing populations of red giants in the galactic disk with CoRoT. A&A 503:L21 O’Malley-James JT, Kaltenegger L (2017) UV surface habitability of the TRAPPIST-1 system. MNRAS 469:L26 Pepe F, Mayor M, Delabre B et al (2000) HARPS: a new high-resolution spectrograph for the search of extrasolar planets. SPIE 4008:582 Queloz D, Bouchy F, Moutou C et al (2009) The CoRoT-7 planetary system: two orbiting superEarths. A&A 506:303 Seager S, Mallén-Ornelas G (2003) A unique solution of Planet and Star parameters from an extrasolar planet transit light curve. ApJ 585:1038 Struve O (1952) Proposal for a project of high-precision stellar radial velocity work. Observatory 72:199–200 Winn JN, Matthews JM, Dawson R et al (2011) A super-Earth transiting a naked-eye Star. ApJ 37:L18 Wolszczan A, Frail DA (1992) A planetary system around the millisecond pulsar PSR1257 C 12. Nature 355:145

55

CoRoT: The First Space-Based Transit Survey to Explore the Close-in Planet Population Magali Deleuil and Malcolm Fridlund

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Spacecraft . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Scientific Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Exoplanetary Science with CoRoT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Detection and Vetting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Exoplanetary Yield: Giants and the First Rocky Planet . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1136 1137 1141 1142 1145 1147 1154 1154

Abstract

The CoRoT (COnvection, internal ROtation and Transiting planets) space mission was launched in the last days of 2006, becoming the first major space mission dedicated to the search for and study of exoplanets, as well as doing the same for asteroseismological studies of stars. Designed as a small mission, it became highly successful, with, among other things, discovering the first planet proved by the measurements of its radius and mass to be definitely “Rocky” or Earthlike in its composition and the first close-in brown dwarf with a measured radius. Designed for a lifetime of 3 years, it survived in a 900 km orbit around the

M. Deleuil () LAM (Laboratoire d’Astrophysique de Marseille), CNRS, CNES, UMR 7326, Aix Marseille Université, Marseille, France e-mail: [email protected] M. Fridlund Leiden Observatory, Leiden, RA, The Netherlands Department of Space, Earth and Environment, Chalmers University of Technology, Onsala, Sweden e-mail: [email protected]; [email protected]; [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_79

1135

1136

M. Deleuil and M. Fridlund

Earth for 6 years discovering in total 37 planetary systems or brown dwarfs, as well as about 100 planet candidates and 2269 eclipsing binaries, detached or in contact. In total CoRoT acquired 177,454 light curves, varying in duration from about 30–150 days. CoRoT was also a pioneer in the organization and archiving of such an exoplanetary survey. The development and utilization of this spacecraft has left a legacy of knowledge, both as what concerns the scientific objectives as well as the technical know-how, that is, currently being utilized in the construction of the European CHEOPS and PLATO missions.

Introduction The CoRoT mission is a space observatory launched by CNES, the French space agency in December 2006. It has its roots in the new science of asteroseismology, which in itself originated from helioseismology, the study of the microvariations of our Sun. Such observations could literally look inside the Sun and determine parameters such as density profiles, internal rotation, and age among other things. The mission was first proposed in 1993, when CNES issued a call for ideas for what they called small missions. This gave French scientists the opportunity to propose and develop a much more ambitious mission than the smaller asteroseismology instrument EVRIS that had already flown on a Russian spacecraft. This was the origin of CoRoT, devoted to the study of stellar COnvection and internal ROTation. The original objective of CoRoT was to carry out very high-precision observations of stellar oscillation mode frequencies, for a dozen of bright solar to F-type stars, in order to detect p-modes (microvariations where pressure, p, is the restoring force) in order to obtain constraints for models of internal structure and to begin to quantify the internal rotation of stars other than the Sun. CNES selected the project at the end of 1994 for a launch in 1998! The detection of the first exoplanet, 51 Peg, (Mayor and Queloz 1995) led to the realization that the CoRoT requirements should also allow the detection of transiting exoplanets. The detection of transiting planets was added in 1997 to the new scientific program of CoRoT whose name was changed to COnvection, internal ROtation and Transiting planets. Different financial and administrative problems led to enlarge the participation to other countries: Austria, Belgium, Brazil, Germany, and Spain, and the ESA Science Program decided to contribute to the project, giving to CoRoT an European and even wider impact. The final mission selection took place only in 2000, with a launch foreseen in 2006 after a development phase that started in 2003. The latter is described fully in Baglin et al. (2016) and Fridlund et al. (2006) and will not be detailed here. The development time was 4 years only, short compared to what is usually required, but the mission succeeded in being launched on December 27, 2006. CoRoT was placed very accurately (errors of order a few hundred meters) by the Soyuz/Fregat launch vehicle into the desired orbit. The launcher was the first version of the Soyuz-Fregat rocket later to be used by the ESA as its medium-size launch vehicle (Lam-Trong 2006b). The fact that the Fregat stage could steer CoRoT

55 CoRoT: The First Space-Based Transit Survey to Explore the Close-in: : :

1137

into the exact orbit saved enough propulsion capacity to last for the whole extended mission. The spacecraft was placed into orbit in perfect condition. The tests and evaluations of the spacecraft used up only half the allotted time, and on January 17, 2007, observations of the first field began.

The Spacecraft The CoRoT satellite was sent into a circular polar orbit with an altitude of 896 km and remained operative there until November 2, 2012, when a computer error terminated the mission. CoRoT was designed as a low-cost mission utilizing a proven spacecraft bus of the PROTEUS family (of which 5 were manufactured for a number of missions) allowing a faster and cheaper development. Together the plateform and the instrument measure 4.2 m along the longest dimension and with a launch wet mass of 626 kg, and the payload comprising 300 kg was thus relatively small. The payload consisted of a 27 cm off-axis telescope, the associated camera, and the mechanical structures and electronics (Fig. 1). The spacecraft bus consumed 300 W of power, while the payload required another 150 W (Lam-Trong 2006a). In order to comply with its scientific objectives, the instrument had to deliver a very stable signal. This stringent stability requirement implied: (i) a high level

Fig. 1 Satellite overview

1138

M. Deleuil and M. Fridlund

of straylight rejection, most of it due to the nearby Earth, (ii) a high pointing stability, and (iii) a high level of performance for the thermal control subsystem. The stability requirements associated with this required the use of hyper stable materials to control the various sources of noise and ensure to reach the photon noise limit (Boisnard and Auvergne 2006). The opto-mechanical design for the telescope, its control of jitter, and its high-performance compact baffling concept were implemented for the first time, and its performance has now been fully demonstrated in flight. To accommodate the two prime scientific objectives, instead of having two separate instruments, the adopted approach consisted in splitting the focal plane in two parts, each dedicated to one of the mission goal. At that time, the exoplanet and seismology observations aimed at targeting stars of different brightnesses. Indeed, while achieving the detection of solar oscillations with a precision of the order of 0.1 Hz required to match the photon noise on bright stars with a very high temporal cadence of one measurement every second, for the transit detection the low transit probability required to observe thousands of stars in order to increase the chance of detection. To fulfill this requirement with the limited field of view of the instrument, the exoplanet program targeted stars in the range 11 to nearly 16. A pair of CCDs was thus dedicated to each program, and they could not be interchanged. The asteroseismology program concentrated on very bright stars, typically in the magnitude range 6–9. The CCDs for the asteroseismology program were defocused, while the exoplanet CCDs were on focus but with a small biprism inserted above the devices (Fig. 2). The resulting point spread function (PSF) in the faint star channel is an on-axis spectrum at a very low spectral resolution. The goal was to provide a chromatic information in order to disentangle true planetary transits from stellar activity features, like spots, or background eclipsing binaries (Rouan et al. 1999).

Fig. 2 The sky on one of the faint star channel (left) and one of the bright star channel (right) (Chaintreuill, priv. com.). The prism in front of the CCDs confers this slightly blurred aspect to the stars of the faint channel, while those in the bright channel are simply defocused

55 CoRoT: The First Space-Based Transit Survey to Explore the Close-in: : :

1139

The two onboard data processing units (DPU) controlled each two CCDs at the time but one for asteroseismology and one for exoplanetology, thereby providing redundancy. The breakdown of the first data processing unit (DPU1), which occurred in March 2009, caused the loss of one CCD in each of the exoplanet and seismology channels and reduced the field-of-view by half but didn’t cause the stop of one of the programs. In the exoplanet channel, the onboard processing and telemetry capacity of the satellite enabled the observation of up to 6000 target stars per CCD. At the start of each observing run, an image of the complete field of view was obtained and downloaded to the ground. Based on this image, each target star was automatically assigned a photometric aperture selected from a library of 254 predefined masks, built so as to optimize the signal-to-noise ratio of the integrated flux (Llebaria and Guterman 2006). For these target stars, the photometry was carried out onboard, and only the light curves were downloaded to Earth. In addition, twenty 10 by 15 pixels windows were downloaded from each CCD in order to provide sky reference images and monitor changes in the background level. A further 80 such windows, designated “imagettes,” initially foreseen as calibration, were assigned to selected special (bright) targets of interest and were downloaded as pixel-level data enabling a dedicated photometric analysis on the ground. The nominal magnitude range of the mission, for the targets in the exoplanet channel, were of magnitude 11–16, but a number of brighter stars were also observed, despite being saturated, and the data from most of these were downloaded as imagettes allowing a more precise photometry to be optimized in later processing. For stars with magnitude 15 or brighter, the photometric aperture was divided along detector column boundaries into three regions of the PSF corresponding approximately to the red, green, and blue parts of the visible spectrum. This way three color light curves were acquired for each such object for up to 5000 stars per CCD. These color light curves were summed together on the ground to give a corresponding white light curve. For stars with magnitudes larger than 15, only white light curves were extracted, and no color information was available. The cadence of the observations could be set to either 32 s or 512 s in the faint channel mode, while the asteroseismology channel allowed settings down to 1 s. The basic faint channel integration time was 32 s, but the flux of 16 readouts was coadded on board over a 8.5 min time span before being downloaded to accommodate with the telemetry budget. The nominal sampling time of 32 s was however preserved for 1000 selected targets (500 per CCD), known as oversampled targets. These targets were selected at the beginning of each run, but the list was then updated every week, thanks to a quick look analysis of the crudely processed light curves and the predetection of transits. In 2016, the complete set of CoRoT light curves that is those from both the bright and faint channels were homogeneously processed with the latest version of the pipeline and released to the community (Chaintreuil et al. 2016). A complete description of the different steps of the final data reduction pipeline and of the associated algorithms is provided in Ollivier et al. (2016). In addition to the regular

1140

M. Deleuil and M. Fridlund

Fig. 3 Section of a light curve in the faint star channel before (top) and after (bottom) correction by the latest version of the data reduction pipeline, including systematics. The jumps which are clearly visible on the raw light curve have been automatically detected and corrected

corrections, such as crosstalk or background contribution corrections, which were already included in the pipeline (Auvergne et al. 2009), but updated in this last version, new corrections were implemented. With this latest release, the user can thus get ready-to-use light curves corrected from the jumps in the photometry such as those induced by a change in temperature or by impacts of protons onto the CCD but also from systematics (Guterman et al. 2016) (Fig. 3). As mentioned above, CoRoT was based on the multipurpose PROTEUS spacecraft bus. This configuration, while saving cost and time in development, restricted the mission to a low Earth orbit (LEO). Because of this limitation, and in order to be able to observe the same field of view for as long time as possible, at least up to 6 months, and at the same time without allowing either too much scattered solar light to enter the telescope or experience too many occultations by the Earth, the satellite had to be injected into a polar orbit and restricted to observe along line of sights roughly perpendicular to the orbital plane. Every 6 months (in April and October), to avoid blinding by the Sun, the satellite was rotated by 180ı with respect to the polar axis, and a new observation period started in the opposite direction. As a consequence, the continuous viewing zones of CoRoT were two almost circular

55 CoRoT: The First Space-Based Transit Survey to Explore the Close-in: : :

1141

regions of 10ı radius, called the CoRoT eyes. They are centered on the galactic plane at 6h 50d (near the galactic anticenter) and 18h 50d (near the galactic center), respectively.

The Scientific Organization In CoRoT, two scientific topics were coexisting in the same instrument. The asteroseismology segment had as its objective to divulge the internal physical parameters of stars for the first time. The objective of the exoplanet element was to discover new transiting exoplanets, to measure their diameter with unprecedented precision and further determine their properties. While these two objectives could at first appear very different, it is actually true that both the technology behind the actual measurements, namely, ultrahigh-precision photometry, as well as the science in them, have a deep connection. One of the superb achievements brought by the CoRoT results is that the understanding of exoplanets is based on an equally deep understanding of the host star. Planets are orbiting a host star, the physics of which governs their evolution and properties. But the opposite is also true. During the formation phase of a planetary system, both star and planets are tightly connected through the transfer of angular momentum and through chemical changes in the accretion disk that must take place as star and planets accrete from the original material. Nevertheless, we will here mainly discuss the exoplanetary part of the mission. The CoRoT project was committed to deliver to the scientific community light curves that were properly reduced and corrected for the main instrumental defects and ready for scientific analyses. The detection of planetary transits was thus left to the discretion of the science team with the exception of a real-time detection carried out in the Alarm Mode on a weekly basis (Quentin et al. 2006; Surace et al. 2008). The Alarm Mode was carried out to retune the cadence of the observations of that particular star to sample the light curve every 32 s instead of 512 s, allowing the possibility to observe forthcoming transit events with higher temporal resolution. In order to interpret the actual transit shape in terms of physical parameters of both the host star and the planet, it is imperative to have the highest temporal resolution. In addition, this real-time detection allows to save time in the follow-up process of the planet candidates. During the years that preceded the launch of the satellite, the scientists who were involved in the exoplanetary program of the mission made the decision to work as a single international team. The goal was to share the workload and results so that to increase the scientific return of the mission and to avoid time wasted in competition. This team thus came to consist of individual scientists from all the nations who were partners in the project and including members from ESAs science department. It took the name of CoRoT Exoplanet Science Team (CEST) and organized every aspects of the analysis, starting from the transit-like features detection to the detailed analysis of the planet’s properties. One important aspect of this collaborative work was the follow-up observations of planet candidates. The CoRoT exoplanet program

1142

M. Deleuil and M. Fridlund

has been indeed supported by a large accompanying ground-based observation program (Deleuil et al. 2006). Operating more than a dozen of telescopes in various places, Europe, Hawaii, Israel, and Chile, with size varying from 1 to 8 m, the team used various techniques: photometric observations (Deeg et al. 2009), highcontrast imaging (Guenther et al. 2013), and spectroscopy including radial velocity measurements (Bouchy et al. 2009). The goal was to identify false positives, to fully secure planets, and to determine their complete set of parameters in order to derive the planetary properties.

Exoplanetary Science with CoRoT In total, CoRoT made observations of 163,665 targets over 26 stellar fields in the faint star channel. The various fields are identified by five digits: the two first refer to the duration of the run but the first run named as Initial Run (IR) – LR for runs with a duration longer than 70 days (Long Runs) and SR for runs with a much shorter duration (Short Run). The third digit “a” or “c” means galactic anticenter or center direction, respectively. Finally, the last two digits are just the chronological order of the given type of field in a given direction. According to the revised exoplanet input catalog (Damiani et al. 2016), 61,174 of these targets have been classified as luminosity class V. In the CoRoT exoplanet context, we were interested in dwarf stars, i.e., luminosity classes IV and V, and then a total of 101,083 stars were available to the mission. The dwarf/giant identification was based on a simple color-magnitude diagram (Deleuil et al. 2009; Damiani et al. 2016), and while this is reliable from a statistical point of view, individual targets could be misclassified. The scope of the preparatory classification and of the targets selection was thus to ensure the observation of all possible dwarfs, at least those with a size suitable for planetary transit detection in a given pointing. Nevertheless, in a given field, the number of stars of classes V and IV could not populate all the available photometric windows, and thus the remaining ones were allocated to targets selected for complementary science programs. There are however noticeable differences in the ratio of dwarf stars to giants from one field to another due to variations in the stellar populations, given their position in the Galaxy and from different reddening between the fields (Table 1). CoRoT observations resulted in 177,454 light curves whose duration ranges from 5.0 days for the stars observed in the SRc03 up to 148.3 days for those that were observed in the LRa03. The SRc03 field was a special pointing dedicated to the sole observation of one further transit of CoRoT-9b (Deeg et al. 2010; Lecavelier des Etangs et al. 2017; Bonomo et al. 2017). Not only was its duration very short, but the number of targets was also very limited (652 only) all of which had been targeted by a previous observation. Table 1 provides a summary of the CoRoT runs and targets that were observed in each of them. Among the 163,665 targets, 12,005 were observed twice and 892 three times. More than 4000 transit-like features were detected in total by the detection teams. Their depth could be as low as 0.01% and periods range from 0.27 to 95 days.

55 CoRoT: The First Space-Based Transit Survey to Explore the Close-in: : :

1143

Table 1 Overview of the CoRoT observed fields and targets. Column 5 gives the number of stars monitored during the pointing, column 6 the number of those targets that were observed in this field only, column 7 the number of those targets that were observed in another field, column 8 the number of those targets that were observed in two other fields, and column 9 the number of targets classified as dwarfs Field IRa01 LRa01 SRa01 SRa02 LRa02 LRa03 SRa03 LRa04 LRa05 SRa04 SRa05 LRa06 LRa07 SRc01 LRc01 LRc02 SRc02 LRc03 LRc04 LRc05 LRc06 LRc07 SRc03 LRc08 LRc09 LRc10 Total

CCD Duration (days) 2 54:3 2 131:5 2 23:4 2 31:8 2 114:7 1 148:3 1 24:3 1 77:6 1 90:5 1 52:3 1 38:7 1 76:7 1 29:3 2 25:6 2 142:1 2 145 2 20:9 1 89:2 1 84:2 1 87:3 1 77:4 1 81:3 1 20:9 1 83:6 1 83:6 1 83:5

Overlap LRa01/LRa06 IRa01/LRa06 SRa05 LRa07

SRa01 LRa01/IRa01 SRa02

LRc06/LRc05

LRc10 LRc06 LRc02/LRc05 LRc08/LRc10 LRc02/LRc06 LRc07/LRc10 LRc04/LRc07

Observed Targ. ] 1 Targ. ] 2 targets 9921 8216 821 11;448 11;448 0 8190 5822 2368 10;305 10;305 0 11;448 11;448 0 5329 5329 0 4169 4169 0 4262 4262 0 4648 4648 0 5588 5588 0 4213 4213 0 5724 1356 3484 4844 4390 454 7015 7015 0 11; 448 11; 448 0 11;448 11;448 0 11;448 11;448 0 5724 5724 0 5724 5724 0 5724 5724 0 5724 3836 1880 5724 3953 1771 652 85 559 5724 5724 0 5724 5724 0 5286 4618 668 177;454 163;665 12;005

Targ. ] 3 Dwarfs (IV/V) 884 6550 0 8961 0 4218 0 7990 0 9410 0 3862 0 3038 0 2967 0 3332 0 3840 0 2452 884 947 0 3173 0 4484 0 4922 0 6239 0 3477 0 3639 0 4200 0 2456 8 2029 0 1784 8 0 0 2658 0 2630 0 1825 892 101;083

Among the detected transit-like features, 116 displayed only 1 transit in any given run. Their duration ranges from nearly 2 h up to 112.16 h for the longest event. From this list, subsequent analyses identified a total of 824 false alarms of various kinds. This category includes simple false alarms due to errors in the detection software or to a discontinuity in the light curve (see Fig. 3), but also at least 211 signals due to a bright eclipsing binary whose light leaked over one or more pixel columns and left its photometric imprint in the light curve(s) of other nearby target(s). Further 2269 eclipsing binaries among which 616 are contact and 1653 are detached binaires, as well as a total of about 600 transit events initially classified as planet candidates

1144 Fig. 4 Distribution of the detected transit-like events among planet candidates (Cand), detached eclipsing binaries (EB), false alarms (FA), and contact binaries (CB)

M. Deleuil and M. Fridlund

Candidats 16.0% False Alarms 20.8%

Contact Binaries 18.9% detached EB 44.3%

Fig. 5 Period–depth diagram of all candidates (violet circle) and EB (gray triangle) detected in the CoRoT fields. The plain lines show the transit signal to noise ratio (SNR)

were detected (Deleuil et al. 2018). Figure 4 shows the distribution of the detected transit-like features over the different classes of events. Out of the planet candidates, 37 exoplanets and brown dwarf systems have been confirmed, with 2 multiplanet systems only. These numbers appear paltry when comparing with the thousands of exoplanets that have been confirmed subsequently, particularly through the Kepler mission and the succeeding K2 mission. The reason for these differences is mostly due to the difference in sensitivity between the two instruments (see Fig. 5). With typical durations of the runs of 69.4 days for the fields in the anticenter and 83.6 days for those in the center, CoRoT has been well adapted for the exploration of the close-in giant population. Two thirds of both the candidates and the EBs have

55 CoRoT: The First Space-Based Transit Survey to Explore the Close-in: : :

1145

Fig. 6 Period distribution (stacked histograms) of EBs (pink) and candidates (gray). The dash lines give the median of each distribution

orbital periods shorter than 10 days with a peak value of 1.5 day and 90% shorter than 25 days (Fig. 6). Transit events at orbital periods in excess of 100 days were also reported. Some of those are single transit events, but some were detected as periodic events during long runs, as was the case for CoRoT-9b (Deeg et al. 2010). The CoRoT team followed the principle of obtaining both a firm mass and a precise radius, the latter owing to the exquisite photometry delivered by the instrument and the former to the great amount of follow-up work carried out. The ensemble of CoRoT planets is thus a very good sample of bodies on which to test theories of planetary structure as well as planetary formation theories.

Detection and Vetting While at the time of writing (December 2017) there are some robust planet candidates still in the final stage of the validation process, as illustrated by the recent publication of CoRoT-32b (Boufleur et al. 2018), CoRoT currently accounts for 38 transiting planets detected and securely confirmed. These are all fully characterized thanks to comprehensive and systematic ground-based follow-up program. Of these 38 objects labeled as planets, there are in fact 2 brown dwarfs, CoRoT-15b (Bouchy et al. 2011) and CoRoT-33b (Csizmadia et al. 2015) and one object, CoRoT-3b (Deleuil et al. 2008), whose exact nature, light brown dwarf or massive planet, depends on how these objects are defined, something that remains the subject of controversies (see Schneider et al. 2011; Hatzes and Rauer 2015) and  Chap. 29, “Definition of Exoplanets and Brown Dwarfs”.

1146

M. Deleuil and M. Fridlund

Follow-up observations significantly increased the scientific return of the mission but required a large effort on various facilities. It involved ground-based photometry taken during and just outside the transits with larger telescopes at higher spatial resolution and spectroscopic observations of the host star. The former was used to verify whether the transits occurred on the main target or a fainter nearby star. Highcontrast imaging helped also identify background binaries or physical triple systems further. Radial velocity measurements enabled identifying objects with multiple sets of spectral lines and to measure the masses of any actual planets, together with the eccentricity of their orbits. Spectra collected for RV measurements or additional spectroscopic data were also used to estimate the host star s parameters: effective temperature, gravity, metallicity, and further deduce their mass, radius, and age. Follow-up observations started as soon as possible after the end of each run and even while CoRoT observations were still ongoing thanks to detections done by the Alarm’s software. Among the 594 candidates which were at one point deemed worthy of follow-up, a bit more of 70% were observed by at least one groundbased facility. Because follow-up observations were triggered as soon as possible while the candidates vetting and light curves modeling continued for months with a process which evolved continuously during the mission, 88 of the candidates which were initially included in the follow-up program were later discarded on the basis of a refined light curve analysis. In some cases, for example, secondary eclipses that indicate an EB were later detected in the full CoRoT light curve of what first appeared as an interesting candidate. Meanwhile the candidate had however already been observed. Follow-up observations concentrated on the brightest stars, with a completeness close to 100% for candidates with r-mag < 14. Much fainter targets were also followed-up, but the fraction drops to 78% for 14  r-mag < 15 and 63% for the faintest targets with r -mag > 15. Figure 7 shows how the transit features initially classified as candidates finally distribute after follow-up observations or a deeper analysis of their light curve. The eclipsing binary class gathers the cases where the source of the transiting signal coincides with the target, which includes spectroscopic binaries and configurations in which the source of the transit has been identified as an eclipsing binary other than the target but whose light contributes to the target measured in CoRoT s photometric aperture. These we denoted also as CEB for contaminating eclipsing binaries. The unresolved category comprises candidates that were either not followed-up or those whose follow-up observations remained inconclusive. The later correspond to cases where the host star is either a hot star or a fast rotator, which prevents assessing the nature of the detected companion with the usual current techniques, but also some bright targets for which ground-based photometry didn’t point out any contaminating star as the source of the signal but for which repeated RV observations reveal no significant variation consistent with the ephemeris of the transits. Confirmed planets/brown dwarfs account for only 6% of the initial list of candidates. Considering only the candidates whose nature has been resolved, from both follow-up and light curves detailed analysis, on-target EB accounts for 53.7% of

55 CoRoT: The First Space-Based Transit Survey to Explore the Close-in: : :

Anticentre

Centre

Planets/BD 6.1%

EB 52.5%

1147

Planets/BD 6.0%

Unres. 41.4%

EB 53.4%

Unres. 40.6%

Fig. 7 Distribution of all features initially selected as candidates among: confirmed planets and brown dwarfs (P/BD), detached eclipsing binaries (EB) finally identified through follow-up observations or a deeper analysis of their light curve, and candidates whose nature still remains unknown (Unres.) in the two pointing directions of CoRoT

them, CEB for 36.1%, and planets/brown dwarfs are 10.2%. This gives a false positive rate close to 90% for CoRoT over the complete mission. Among the candidates that remain unresolved, a bit more of 100 still present some chances of being of planetary nature. Nearly half of them were subject to ground-based complementary observations which however remained inconclusive about their exact nature or would just require some deeper analysis as exemplified by CoRoT-32b. For the other half, the faintness of the targets (usually r-mag > 14) did not allow a proper characterization of the transiting body. Indeed, even for Jupiter-size planets, radial velocity measurements remain difficult for targets with magnitudes greater than 14.5. Taking into account the spectral classification of these targets, the results achieved for candidates, and assuming that the unresolved cases follow the same distribution as those whose nature has been secured, one can estimate that 8 ˙ 3 planets are still to be identified as such. This highlights the challenge of establishing the planetary nature of faint candidates which are still challenging the current spectrograph performances. This experience also motivates the search for transiting planets on brighter stars, as will be done by the TESS and PLATO missions.

The Exoplanetary Yield: Giants and the First Rocky Planet With its high-precision photometry and its typical pointing duration of a few tens of days, CoRoT was well adapted to explore the close-in planet population, and its contribution to the properties of this population has been pioneering. Table 2 provides a summary of parameters of the published CoRoT planetary systems.

10b10

9b9

8b8

7c7

7b7

6b6

5b5

4b4

3b3

2b2

1b1

Planet

Period Rp (days) (Rjup / 1:5089557 1:49 0:00000064 0:08 1:7429964 1:47 0:0000017 0:03 4:2567994 1:01 0:000004 0:07 9:20205 1:19 0:00037 0:06 4:037896 1:33 0:000002 0:05 8:886593 1:17 0:000004 0:04 0:85359163 0:141 5:8E  7 0:009 3:698 – 0:003  6:21229 0:57 0:00003 0:02 95:273804 0:94 0:0014 0:04 13:2406 0:97 0:0002 0:07

Mp (Mjup / 1:03 0:12 3:31 0:16 21:77 1:0 0:72 0:08 0:47 0:05 2:96 0:34 0:017 0:003 0:026 0:003 0:22 0:03 0:84 0:07 2:75 0:16 0.006 0.012 0.036 0.033 0.012 0.01 0.27 0.15 0.086 0.07 0.18 0.12 0.137 0.094 0 Fixed 0 Fixed 0.11 0.039 0.53 0.04

e

a (AU) 0:0254 0:0004 0:0281 0:0009 0:057 0:003 0:090 0:001 0:0495 0:0003 0:0855 0:0015 0:017 1:6E4 0:046 – 0:063 0:001 0:407 0:005 0:1055 0:0021

M? (Mˇ ) 0:95 0:15 0:97 0:06 1:37 0:09 1:16 0:03 1:00 0:02 1:05 0:05 0:91 0:02 0:91 0:03 0:88 0:04 0:99 0:04 0:89 0:05

R? (Rˇ ) 1:11 0:05 0:90 0:02 1:56 0:09 1:17 0:03 1:19 0:04 1:025 0:03 0:82 0:02 0:82 0:04 0:77 0:02 0:94 0:04 0:79 0:05

Teff (K) 5950 150 5625 120 6740 140 6190 60 6100 65 6090 50 5275 60 5275 60 5080 80 5625 80 5075 75

v sin i (km/s) 5:2 1:0 11:85 0:5 18 3:0 6:4 1:0 1:0 1:0 7:5 1:0 1:5 1:0 1:5 1:0 2:0 1:0

3.5 1:0 2:0 0:5 0.3 0.25 0.04 0.1 0.02 0.06 +0.05 0.07 0.25 0.06 0.20 0.1 +0.12 0.06 +0.12 0.06 0.3 0.1 0.01 0.06 +0.26 0.07

Fe/H

Prot (days) 10.7 2.2 4.52 0.14 4.6 0.4 8.9 1.1 50.0 10 6.4 0.5 23.6 0.1 23.6 0.1 20.0 5  14.0 5 2.0 0.5

0.5–3.0

0.5–8.0

1.32 0.75 1.32 0.75 0.5–3.0

2.0–4.0

5.5–8.3

0.7–2.0

1.6–2.8

0.03–0.3

Age (Gyr)

Table 2 The 35 confirmed and published CoRoT planets and brown dwarf stars. The second row in each planetary entry gives the accuracy of the result. Note that when asymmetrical error bars were provided, we give here the largest value

1148 M. Deleuil and M. Fridlund

20b20

19b19

18b18

17b17

16b16

15b15

14b14

13b13

12b12

11b11

2:99433 0:000011 2:828042 0:000013 4:03519 0:00003 1:51214 0:00013 3:06036 0:00003 5:35227 0:0000 3:7681 0:0000 1:900069 0:0000 3:89713 0:0000 9:24285 0:0000

1:43 0:03 1:44 0:13 0:89 0:01 1:09 0:07 1:12 0:30 1:17 0:15 1:02 0:07 1:31 0:18 1:29 0:03 0:84 0:04

2:33 0:34 0:92 0:07 1:31 0:07 7:6 0:6 63:3 4:1 0:54 0:09 2:43 0:30 3:47 0:38 1:11 0:06 4:24 0:23 0.35 0.03 0.07 0.06 0. Fixed 0. Fixed 0 Fixed 0.33 0.10 0. Fixed 0.04 0.04 0.047 0.045 0.562 0.013

0.0436 0.005 0.0402 0.0009 0.051 0.0031 0.027 0.002 0.045 0.014 0.0618 0.0015 0.0461 0.0008 0.0295 0.0016 00518 0.0008 0.0902 0.0021

1:27 0:05 1:08 0:08 1:09 0:02 1:13 0:09 1:32 0:12 1:1 0:08 1:04 0:1 0:95 0:15 1:21 0:05 1:14 0:08

1:37 0:03 1:1 0:1 1:01 0:03 1:21 0:08 1:46 0:31 1:19 0:14 1:59 0:07 1:00 0:13 1:65 0:04 1:02 0:05

6440 120 5675 80 5945 90 6035 100 6350 200 5650 10 5740 80 5440 100 6090 70 5880 90

40:0 5:0 1:0 1:0 4:0 1:0 9:0 0:5 19 1:0 0:5 1:0 4:5 0:5 8:0 1:0 6:0 1:0 4:5 1:0

0.03 0.08 +0.16 0.1 +0.01 0.07 +0.05 0. +0.1 0.2 +0.19 0.06 +0.00 0.1 0.10 0.1 0.02 0.1 +0.14 0.12 1.7 0.2 68.0 10 13.0 5 5.7 15 3.0 0.1 60.0 10 20.0 5 5.4 0.4 15.0 5 11.5 3 (continued)

0.06–0.9

4.0–6.0

0.05–1

9.7–11

3.7–9.7

1.1–3.4

0.4–8.0

0.1–3.2

3.2–9.4

1–3

55 CoRoT: The First Space-Based Transit Survey to Explore the Close-in: : : 1149

27b26

26b25

25b25

24c24

24b24

23b23

22b22

21b21

Planet

Period (days) 2:72474 0:0000 9:75598 0:00011 3:6313 0:0001 5:1134: 0:0006 11:759 0:0063 4:86069 0:00006 4:20474 0:00005 3:57532 0:00006

Table 2 (continued)

Rp (Rjup / 1:30 0:14 0:435 0:035 1:05 0:13 0:33 0:04 0:44 0:04 1:08 0:1 1:26 0:13 1:007 0:044

0:088 0:035 0:27 0:04 0:52 0:05 10:39 0:55

Mp (Mjup / 2:26 0:31 0:038 0:044 2:8 0:3 0:018

0.0 Fixed 0.0 Fixed 4.6

12.0 1.5 4.5 3.5 3.0 3.7 4.7 2.2 –

References to table: 1 Barge et al. (2008), 2 Alonso et al. (2008); Bouchy et al. (2008), 3 Deleuil et al. (2008), 4 Aigrain et al. (2008), 5 Rauer et al. (2009), 6 Fridlund et al. (2010), 7 Léger et al. (2009); Queloz et al. (2009); Barros et al. (2014); Haywood et al. (2014), 8 Bordé et al. (2010), 9 Deeg et al. (2010), 10 Bonomo et al. (2010), 11 Gandolfi et al. (2010), 12 Gillon et al. (2010), 13 Cabrera et al. (2010), 14 Tingley et al. (2011), 15 Bouchy et al. (2011), 16 Ollivier et al. (2012), 17 Csizmadia et al. (2011), 18 Hébrard et al. (2011), 19 Guenther et al. (2012), 20 Deleuil et al. (2012), 21 Pätzold et al. (2012), 22 Moutou et al. (2014), 23 Rouan et al. (2012), 24 Alonso et al. (2014), 25 Almenara et al. (2013), 26 Almenara et al. (2013), 26 Parviainen et al. (2014), 27 Cabrera et al. (2015), 28 Bordé et al. (2018),29 Boufleur et al. (2018), 30 Csizmadia et al. (2015)

33b30

32b29

31b28

30b28

29b27

28b27

55 CoRoT: The First Space-Based Transit Survey to Explore the Close-in: : : 1151

1152

M. Deleuil and M. Fridlund

While much more modest than Kepler, CoRoT accounts for number of first results on the properties of close-in transiting populations, a large fraction of them being highlighted in Moutou et al. (2013): – The first terrestrial planet, CoRoT-7b (Léger et al. 2009; Queloz et al. 2009). – The first transiting substellar companion/brown dwarfs, CoRoT-3b (Deleuil et al. 2008) and CoRoT-15b (Bouchy et al. 2011) well in the gap between planets and low mass stellar companions. – The first temperate Jupiter-like planet, CoRoT-9b (Deeg et al. 2010) – The first massive planet around a fast CoRoT-11 Gandolfi et al. (2010) – The first phase curve and planet occultation, CoRoT-2b (Snellen et al. 2009; Alonso et al. 2009) – The first mapping of active regions at the stellar surface, CoRoT-4b (Lanza et al. 2009) – the first detection of the ellipsoidal and the relativistic beaming effects for substellar companion, CoRoT-3b (Mazeh and Faigler 2010) In total, CoRoT accounts today for 39 planets, with the last 4 still to be published (Grziwas et al. in prep). Among them, 33 are Saturn- and Jupiter-like planets. For these massive planets, CoRoT light curves have enabled detailed analyses on their properties. It allowed to probe their interior composition and to quantify their various enrichment in heavy elements (e.g. Bordé et al. 2010; Deleuil et al. 2012). Close-in planets with non-zero eccentricity like CoRoT-20b (Deleuil et al. 2012) and CoRoT-23b (Rouan et al. 2012) bring observational constraints to ongoing tidal dissipation in planets and their circularization time (Ferraz-Mello 2016). Despite the complexity of the CoRoT detection and vetting processes which involved different pipelines and methods that evolved during the mission lifetime and the lack of a proper estimate of the mission detection sensitivity, these planets were used to derive first-order occurrence rates (Deleuil et al. 2018). For close-in massive objects, brown dwarfs and hot-Jupiters, limiting the sample to those with an orbital period less than 10 days and a magnitude less than mag-r = 15.1 ensured the completeness of the follow-up process. In this range of sizes, it seems likely that only a few of them could have been missed, and a detection completeness of 90% seems reasonable. This gives approximate occurrence rates of 0.98 ˙ 0.26% for close-in hot-Jupiters and of 0.07 ˙ 0.05% for brown dwarfs in the CoRoT fields. For hot-Jupiters, this result is in agreement with the occurrence rates estimated for this planet population from radial velocity surveys (Wright et al. 2012; Mayor et al. 2011) but not with those from Kepler data (Howard et al. 2012; Fressin et al. 2013; Santerne et al. 2016). In particular, this is about twice the estimate derived by Santerne et al. (2016) for the Kepler giants based on planet candidate follow-up observations. On the reverse, the brown dwarf frequency estimate, treating CoRoT-3b as such, is about four times smaller than the one for the Kepler field. Santerne et al. (2016) found indeed 0:29 ˙ 0:17% for Kepler, but the latter covers orbital periods up to 400 days, while the three CoRoT brown dwarfs all have P < 6 days.

55 CoRoT: The First Space-Based Transit Survey to Explore the Close-in: : :

1153

For giant planets at longer orbital period, between 10 and 100 days, the occurrence rate of 1.86 ˙ 0.68% is more in agreement with studies from other sources (Mayor et al. 2011; Fressin et al. 2013; Santerne et al. 2016). Because of the typical duration of CoRoT runs and a likely much higher incompleteness of the follow-up in this range of periods, this number was calculated not only based on the confirmed planets but by including some candidates whose planetary nature could not be fully confirmed. The detection of CoRoT-7b (Léger et al. 2009) has opened the domain of the super-Earth planets, a population that was not predicted by planet formation models but was later demonstrated by Kepler to be numerous. The super-Earth population is however at the limit of CoRoT photometric precision for solar-type stars, and CoRoT detections remained rare in this domain of small-size planets. While still, in principle, within CoRoT detection capability, a dearth of Neptunesized planets was reported by Bonomo et al. (2012) from their analysis of six CoRoT fields only. This was later confirmed by Deleuil et al. (2018) who did the final analysis of all CoRoT observations. They indeed estimated that, compared to the frequency of Kepler planets for the class of small Neptunes and super-Earths (Fressin et al. 2013), the occurrence of small-size planets with Rp < 5R˚ orbiting GKM dwarfs within 10 days in CoRoT is still too low by more than a factor of three. This number was derived using the same detection completeness of 36.6 ˙ 6.4% as in Bonomo et al. (2012). Among the various reasons, this discrepancy could have its origin in differences in the stellar population targeted by these two missions. In addition, a fraction of these small-size planets might be missed due to discontinuities in the CoRoT light curves caused by hot pixels, light curves which were still used when the final CoRoT catalog was established. With the final version of the CoRoT pipeline that corrects for a number of discontinuities being available, it will be possible to investigate this issue further. While multi-planet systems account for about 40% of the Kepler objects of interest (KOI), only one multi-planet transiting system, CoRoT-24 (Alonso et al. 2014), has been reported by CoRoT. This system hosts two Neptune-sized planets. The second multi-planet system is CoRoT-7 (Queloz et al. 2009). CoRoT-7c does not transit, but its existence has been definitely established from the intensive radial velocity campaigns carried out to measure CoRoT-7b’ mass (e.g. Haywood et al. 2014). A third planet, CoRoT-7d, with a mass representative of super-Earth or Neptune size is indicated but not as well secured in the RV data (Hatzes et al. 2010). This observed low number of multi-planet detections is consistent with Kepler’s results, which show that such systems are indeed numerous but are mainly found in the low mass regime of Neptune- and Earth-size planets and in the long orbital period range, a domain which is beyond the limit of CoRoT sensitivity. A search for circumbinary planets (CBPs, planets that orbit both components of a binary star) in CoRoT’s entire sample of over 2000 EBs found three candidates based on the observation of single transits (Klagyivik et al. 2017). For the time being, none of these can be verified as a planet, however. Abundance maxima for CBPs derived from that search agree with results from the Kepler mission, which had a sample of EBs of similar size. The CBPs detected by Kepler (see

1154

M. Deleuil and M. Fridlund

 Chap. 128, “Two Suns in the Sky: The Kepler Circumbinary Planets”) have

periods that are mostly too long to produce multiple transits within the durations of the CoRoT runs, while their small size would have made them only marginally detectable with CoRoT. CoRoT had however good detection capability for larger or shorter-periodic (p . 25 days) CBPs, but such planets either don’t exist or are very rare. This survey doubled the sample size on which similar conclusions had been drawn previously (Martin et al. 2015; Hamers et al. 2016), based on searches in Kepler data.

Conclusion The CoRoT mission has been the pioneer for space-based transiting planet detection. It was the first instrument that provided high-precision photometric observations, with almost uninterrupted coverage for weeks. Supported by an extensive program of ground-based follow-up observations, it also managed to accurately determine not only the planets diameter but also their mass and the complete set of their orbital parameters, providing observational constraints to models of formation and evolution of the close-in planet population. Surpassed by Kepler, CoRoT accounts for a number of first results on the properties of the close-in transiting populations (Moutou et al. 2013; Deleuil et al. 2018, e.g.). The approach the CoRoT team chooses has demonstrated the interest of combining space-based and ground-based observations in their respective domain of best performance. It also has become the role model for ESA’s small satellite program and led the European community to propose both the CHEOPS and PLATO missions (Broeg et al. 2013; Rauer et al. 2014), now under development. CoRoT was designed for 2 1/2 years of operation but surpassed all expectations and continued to operate for almost 6 years producing a number of light curves similar to much more ambitious (later) missions. The legacy of CoRoT continues as is demonstrated by the number of exoplanetary systems carrying the CoRoT designation that have been further studied and continue to be studied in order to clarify all the unsolved questions relating to this topic.

References Aigrain S, Collier Cameron A, Ollivier M et al (2008) Transiting exoplanets from the CoRoT space mission. IV. CoRoT-Exo-4b: a transiting planet in a 9.2 day synchronous orbit. A&A 488:L43– L46 Almenara JM, Bouchy F, Gaulme P et al (2013) Transiting exoplanets from the CoRoT space mission. XXIV. CoRoT-25b and CoRoT-26b: two low-density giant planets. A&A 555:A118 Alonso R, Auvergne M, Baglin A et al (2008) Transiting exoplanets from the CoRoT space mission. II. CoRoT-Exo-2b: a transiting planet around an active G star. A&A 482:L21–L24 Alonso R, Guillot T, Mazeh T et al (2009) The secondary eclipse of the transiting exoplanet CoRoT-2b. A&A 501:L23–L26 Alonso R, Moutou C, Endl M et al (2014) Transiting exoplanets from the CoRoT space mission. XXVI. CoRoT-24: a transiting multiplanet system. A&A 567:A112

55 CoRoT: The First Space-Based Transit Survey to Explore the Close-in: : :

1155

Auvergne M, Bodin P, Boisnard L et al (2009) The CoRoT satellite in flight: description and performance. A&A 506:411–424 Baglin A, Lam-Trong T, Vandermarcq O, Donny C Burgaud S (2016) I.3 The CoRoT story. In: CoRot Team (ed) The CoRoT legacy book: the adventure of the ultrahigh-precision photometry from space, p 11. https://doi.org/10.1051/978-2-7598-1876-1.c013 Barge P, Baglin A, Auvergne M et al (2008) Transiting exoplanets from the CoRoT space mission. I. CoRoT-Exo-1b: a low-density short-period planet around a G0V star. A&A 482:L17–L20 Barros SCC, Almenara JM, Deleuil M et al (2014) Revisiting the transits of CoRoT-7b at a lower activity level. A&A 569:A74 Boisnard L Auvergne M (2006) CoRoT in brief. In: Fridlund M, Baglin A, Lochard J Conroy L (eds) The CoRoT mission pre-launch status – stellar seismology and planet finding, vol 1306. ESA Special Publication, Noordwijk, p 19 Bonomo AS, Santerne A, Alonso R et al (2010) Transiting exoplanets from the CoRoT space mission. X. CoRoT-10b: a giant planet in a 13.24 day eccentric orbit. A&A 520:A65 Bonomo AS, Chabaud PY, Deleuil M et al (2012) Detection of neptune-size planetary candidates with CoRoT data. Comparison with the planet occurrence rate derived from Kepler. A&A 547:A110 Bonomo AS, Hébrard G, Raymond SN et al (2017) A deeper view of the CoRoT-9 planetary system. A small non-zero eccentricity for CoRoT-9b likely generated by planet-planet scattering. A&A 603:A43 Bordé P, Bouchy F, Deleuil M et al (2010) Transiting exoplanets from the CoRoT space mission. XI. CoRoT-8b: a hot and dense sub-Saturn around a K1 dwarf. A&A 520:A66 Bordé P, Díaz R, Creevey O et al (2018, submitted) Transiting exoplanets from the CoRoT space mission. XXIX. The hot Jupiters CoRoT-30 and CoRoT-31. A&A Bouchy F, Queloz D, Deleuil M et al (2008) Transiting exoplanets from the CoRoT space mission. III. The spectroscopic transit of CoRoT-Exo-2b with SOPHIE and HARPS. A&A 482:L25–L28 Bouchy F, Moutou C, Queloz D, CoRoT Exoplanet Science Team (2009) Radial velocity follow-up for confirmation and characterization of transiting exoplanets. In: Pont F, Sasselov D, Holman MJ (eds) Transiting planets. IAU symposium, vol 253, pp 129–139. https://doi.org/10.1017/ S174392130802632X Bouchy F, Deleuil M, Guillot T et al (2011) Transiting exoplanets from the CoRoT space mission. XV. CoRoT-15b: a brown-dwarf transiting companion. A&A 525:A68 Boufleur RC, Emilio M, Janot-Pacheco E et al (2018) A modified CoRoT detrend algorithm and the discovery of a new planetary companion. MNRAS 473:710–720 Broeg C, Fortier A, Ehrenreich D et al (2013) CHEOPS: a transit photometry mission for ESA’s small mission programme. Eur Phys J Web Conf, Eur Phys J Web Conf 47:03005. https://doi. org/10.1051/epjconf/20134703005 Cabrera J, Bruntt H, Ollivier M et al (2010) Transiting exoplanets from the CoRoT space mission . XIII. CoRoT-13b: a dense hot Jupiter in transit around a star with solar metallicity and supersolar lithium content. A&A 522:A110 Cabrera J, Csizmadia S, Montagnier G et al (2015) Transiting exoplanets from the CoRoT space mission. XXVII. CoRoT-28b, a planet orbiting an evolved star, and CoRoT-29b, a planet showing an asymmetric transit. A&A 579:A36 Chaintreuil S, Bellucci A, Baudin F et al (2016) II.5 where to find the CoRoT data?, p 109. https:// doi.org/10.1051/978-2-7598-1876-1.c025 Csizmadia S, Moutou C, Deleuil M et al (2011) Transiting exoplanets from the CoRoT space mission. XVII. The hot Jupiter CoRoT-17b: a very old planet. A&A 531:A41 Csizmadia S, Hatzes A, Gandolfi D et al (2015) Transiting exoplanets from the CoRoT space mission. XXVIII. CoRoT-33b, an object in the brown dwarf desert with 2:3 commensurability with its host star. A&A 584:A13 Damiani C, Meunier JC, Moutou C et al (2016) Stellar classification of CoRoT targets. A&A 595:A95 Deeg HJ, Gillon M, Shporer A et al (2009) Ground-based photometry of space-based transit detections: photometric follow-up of the CoRoT mission. A&A 506:343–352

1156

M. Deleuil and M. Fridlund

Deeg HJ, Moutou C, Erikson A et al (2010) A transiting giant planet with a temperature between 250K and 430K. Nature 464:384–387 Deleuil M, Moutou C, Deeg HJ et al (2006) Complemetary Observations for the CoRoT exoplanet program. In: Fridlund M, Baglin A, Lochard J Conroy L (eds) The CoRoT mission pre-launch status – stellar seismology and planet finding, vol 1306. ESA Special Publication, Noordwijk, p 341 Deleuil M, Deeg HJ, Alonso R et al (2008) Transiting exoplanets from the CoRoT space mission . VI. CoRoT-Exo-3b: the first secure inhabitant of the brown-dwarf desert. A&A 491: 889–897 Deleuil M, Meunier JC, Moutou C et al (2009) Exo-dat: an information system in support of the CoRoT/exoplanet science. AJ 138:649–663 Deleuil M, Bonomo AS, Ferraz-Mello S et al (2012) Transiting exoplanets from the CoRoT space mission. XX. CoRoT-20b: a very high density, high eccentricity transiting giant planet. A&A 538:A145 Deleuil M, Aigrain S, Moutou C et al (2018) Planets, candidates, and binaries from the CoRoT/exoplanet program, the CoRoT catalog. A&A. https://doi.org/10.1051/0004-6361/ 201731068 Ferraz-Mello S (2016) III.9-2 tidal evolution of CoRoT massive planets and brown dwarfs and of their host stars, p 169. https://doi.org/10.1051/978-2-7598-1876-1.c239 Fressin F, Torres G, Charbonneau D et al (2013) The false positive rate of Kepler and the occurrence of planets. ApJ 766:81 Fridlund M, Baglin A, Lochard J, Conroy L (2006) The CoRoT mission pre-launch status – stellar seismology and planet finding. In: Fridlund M, Baglin A, Lochard J, Conroy L (eds) The CoRoT mission pre-launch status – stellar seismology and planet finding, vol 1306. ESA Special Publication, Noordwijk Fridlund M, Hébrard G, Alonso R et al (2010) Transiting exoplanets from the CoRoT space mission. IX. CoRoT-6b: a transiting “hot Jupiter” planet in an 8.9d orbit around a lowmetallicity star. A&A 512:A14 Gandolfi D, Hébrard G, Alonso R et al (2010) Transiting exoplanets from the CoRoT space mission. XIV. CoRoT-11b: a transiting massive “hot-Jupiter” in a prograde orbit around a rapidly rotating F-type star. A&A 524:A55 Gillon M, Hatzes A, Csizmadia S et al (2010) Transiting exoplanets from the CoRoT space mission. XII. CoRoT-12b: a short-period low-density planet transiting a solar analog star. A&A 520:A97 Guenther EW, Díaz RF, Gazzano JC et al (2012) Transiting exoplanets from the CoRoT space mission. XXI. CoRoT-19b: a low density planet orbiting an old inactive F9V-star. A&A 537:A136 Guenther EW, Fridlund M, Alonso R et al (2013) High angular resolution imaging and infrared spectroscopy of CoRoT candidates. A&A 556:A75 Guterman P, Mazeh T, Faigler S (2016) II.3 exposure based algorithm for removing systematics out of the CoRoT light curves, p 55. https://doi.org/10.1051/978-2-7598-1876-1.c023 Hamers AS, Perets HB, Portegies Zwart SF (2016) A triple origin for the lack of tight coplanar circumbinary planets around short-period binaries. MNRAS 455:3180–3200 Hatzes AP, Rauer H (2015) A definition for giant planets based on the mass-density relationship. ApJ 810:L25 Hatzes AP, Dvorak R, Wuchterl G et al (2010) An investigation into the radial velocity variations of CoRoT-7. A&A 520:A93 Haywood RD, Collier Cameron A, Queloz D et al (2014) Planets and stellar activity: hide and seek in the CoRoT-7 system. MNRAS 443:2517–2531 Hébrard G, Evans TM, Alonso R et al (2011) Transiting exoplanets from the CoRoT space mission. XVIII. CoRoT-18b: a massive hot Jupiter on a prograde, nearly aligned orbit. A&A 533:A130 Howard AW, Marcy GW, Bryson ST et al (2012) Planet occurrence within 0.25 AU of solar-type stars from Kepler. ApJS 201:15 Klagyivik P, Deeg HJ, Cabrera J, Csizmadia S, Almenara JM (2017) Limits to the presence of transiting circumbinary planets in CoRoT data. A&A 602:A117

55 CoRoT: The First Space-Based Transit Survey to Explore the Close-in: : :

1157

Lam-Trong T (2006a) General presentation of CoRoT. In: Fridlund M, Baglin A, Lochard J, Conroy L (eds) The CoRoT mission pre-launch status – stellar seismology and planet finding, vol 1306. ESA Special Publication, Noordwijk, p 103 Lam-Trong T (2006b) The launcher. In: Fridlund M, Baglin A, Lochard J, Conroy L (eds) The CoRoT mission pre-launch status – stellar seismology and planet finding, vol 1306. ESA Special Publication, Noordwijk, p 255 Lanza AF, Aigrain S, Messina S et al (2009) Photospheric activity and rotation of the planet-hosting star CoRoT-4a. A&A 506:255–262 Lecavelier des Etangs A, Hébrard G, Blandin S et al (2017) Search for rings and satellites around the exoplanet CoRoT-9b using spitzer photometry. A&A 603:A115 Léger A, Rouan D, Schneider J et al (2009) Transiting exoplanets from the CoRoT space mission. VIII. CoRoT-7b: the first super-Earth with measured radius. A&A 506:287–302 Llebaria A, Guterman P (2006) Building up photometric apertures for the exoplanet channel. In: Fridlund M, Baglin A, Lochard J, Conroy L (eds) The CoRoT mission pre-launch status – stellar seismology and planet finding, vol 1306. ESA Special Publication, Noordwijk, p 293 Martin DV, Mazeh T, Fabrycky DC (2015) No circumbinary planets transiting the tightest Kepler binaries – a possible fingerprint of a third star. MNRAS 453:3554–3567 Mayor M, Queloz D (1995) A Jupiter-mass companion to a solar-type star. Nature 378:355–359 Mayor M, Marmier M, Lovis C et al (2011) The HARPS search for southern extra-solar planets XXXIV. Occurrence, mass distribution and orbital properties of super-Earths and Neptune-mass planets. ArXiv e-prints Mazeh T, Faigler S (2010) Detection of the ellipsoidal and the relativistic beaming effects in the CoRoT-3 lightcurve. A&A 521:L59 Moutou C, Deleuil M, Guillot T et al (2013) CoRoT: harvest of the exoplanet program. Icarus 226:1625–1634 Moutou C, Almenara JM, Díaz RF et al (2014) CoRoT-22 b: a validated 4.9 R‹ exoplanet in 10-d orbit. MNRAS 444:2783–2792 Ollivier M, Gillon M, Santerne A et al (2012) Transiting exoplanets from the CoRoT space mission. XXII. CoRoT-16b: a hot Jupiter with a hint of eccentricity around a faint solar-like star. A&A 541:A149 Ollivier M, Deru A, Chaintreuil S et al (2016) II.2 description of processes and corrections from observation to delivery, p 41. https://doi.org/10.1051/978-2-7598-1876-1.c022 Parviainen H, Gandolfi D, Deleuil M et al (2014) Transiting exoplanets from the CoRoT space mission. XXV. CoRoT-27b: a massive and dense planet on a short-period orbit. A&A 562:A140 Pätzold M, Endl M, Csizmadia S et al (2012) Transiting exoplanets from the CoRoT space mission. XXIII. CoRoT-21b: a doomed large Jupiter around a faint subgiant star. A&A 545:A6 Queloz D, Bouchy F, Moutou C et al (2009) The CoRoT-7 planetary system: two orbiting superEarths. A&A 506:303–319 Quentin CG, Barge P, Cautain R et al (2006) Software applications for oversampling of transit candidates. In: Fridlund M, Baglin A, Lochard J, Conroy L (eds) The CoRoT mission pre-launch status – stellar seismology and planet finding, vol 1306. ESA Special Publication, Noordwijk, p 409 Rauer H, Queloz D, Csizmadia S et al (2009) Transiting exoplanets from the CoRoT space mission. VII. The “hot-Jupiter”-type planet CoRoT-5b. A&A 506:281–286 Rauer H, Catala C, Aerts C et al (2014) The PLATO 2.0 mission. Exp Astron 38:249–330 Rouan D, Baglin A, Barge P et al (1999) Searching for exosolar planets with the COROT space mission. Phys Chem Earth C 24:567–571 Rouan D, Parviainen H, Moutou C et al (2012) Transiting exoplanets from the CoRoT space mission. XIX. CoRoT-23b: a dense hot Jupiter on an eccentric orbit. A&A 537:A54 Santerne A, Moutou C, Tsantaki M et al (2016) SOPHIE velocimetry of Kepler transit candidates. XVII. The physical properties of giant exoplanets within 400 days of period. A&A 587:A64 Schneider J, Dedieu C, Le Sidaner P, Savalle R, Zolotukhin I (2011) Defining and cataloging exoplanets: the exoplanet.eu database. A&A 532:A79

1158

M. Deleuil and M. Fridlund

Snellen IAG, de Mooij EJW, Albrecht S (2009) The changing phases of extrasolar planet CoRoT1b. Nature 459:543–545 Surace C, Alonso R, Barge P et al (2008) The oversampling mode for CoRoT exo-field observations. In: Advanced software and control for astronomy II. Proceedings of SPIE, vol 7019, p 70193B. https://doi.org/10.1117/12.789334 Tingley B, Endl M, Gazzano JC et al (2011) Transiting exoplanets from the CoRoT space mission. XVI. CoRoT-14b: an unusually dense very hot Jupiter. A&A 528:A97 Wright JT, Marcy GW, Howard AW et al (2012) The frequency of hot Jupiters orbiting nearby solar-type stars. ApJ 753:160

Space Missions for Exoplanet Science: Kepler/K2

56

William J. Borucki

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Proposals to the Discovery Program and Validation of Technical Readiness . . . . . . . . . . . . Mission Goals, Science Requirements, Key System Features, and Ancillary Programs . . . . Flight System Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Flight System Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Field-Flattening Lenses and Wavelength Response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Selection of FOV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Selection and Characterization of Target Stars: Kepler Input Catalog . . . . . . . . . . . . . . . . Orbit and Commissioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Instrument Operation and Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Detector Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Saturation and Dynamic Range . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Read Noise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Smear . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . On-Orbit Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Data Processing, Vetting, and Archiving . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mission Completion and Transition to K2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Transition to the K2 Mission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Representative Science Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1160 1161 1163 1164 1164 1167 1167 1167 1169 1169 1169 1170 1170 1170 1171 1171 1172 1173 1173 1174 1176

Abstract

The Kepler Mission is a space observatory launched in 2009 to monitor 170,000 stars over a period of 4 years to explore the structure and diversity of planetary systems, in particular to determine the frequency of Earth-size and larger planets

W. J. Borucki () NASA Ames Research Center, Moffett Field, CA, USA e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_80

1159

1160

W. J. Borucki

in and near the habitable zone (HZ) of Sun-like stars, the size and orbital distributions of these planets, and the types of stars they orbit. Kepler is the tenth in the series of NASA Discovery Program missions that were competitively selected, PI-directed, medium-cost missions. The Kepler instrument is based on a 0.95 m aperture Schmidt-design telescope with a 113 sq deg FOV with 84 channels of CCD detectors. The spacecraft orbits the Sun in 53-week orbit and points at a single field of stars to avoid missing transits. It achieves a photometric precision of 10 ppm for bright, quiet stars for planets with transit times of several hours. During the 4 years of its operation, an analysis of its data detected over 4,600 planetary candidates which include several hundred Earth-size planetary candidates, over 3,458 confirmed planets, and 21 Earth-size and super-Earthsize planets in the HZ. These discoveries provide the information required for estimates of the frequency of planets in our galaxy. The mission results show that most stars have planets, many of these planets are similar in size to the Earth, and systems with several planets are common. In combination with radial velocity measurements, the Kepler results were able to distinguish rocky planets from planets that might be composed mostly of water or gas. The Kepler Mission also made major contributions to astrophysical sciences including the detection of “heartbeat” stars, supernovae, and the interior structure and characteristics of red giant stars. At the end of its 4-year mission, the Kepler Mission was re-proposed as a community facility (“K2”) that provided high-precision photometric timeseries data for tens of thousands of stars in each of 13 FOVs along the ecliptic.

Introduction A first step in finding the extent of life in our galaxy is to discover whether Earth-size planets in the HZ of other stars are common or rare. The HZ is the region around a star in which a rocky planet would receive the appropriate amount of radiative flux to allow water to be liquid on its surface (Kasting 1993; Kopparapu et al. 2014). The Kepler MissionKepler MissionKepler Mission was designed specifically to determine whether Earth-size planets in the HZ are common or rare. The mission was named after Johannes Kepler who developed the laws of planetary motion in the seventeenth century and was an early advocate of the Copernican theory that planets orbit the Sun. The first quantitative discussion of the use of transit photometry for detecting exoplanets by searching for patterns of transits to get size and orbital period was a paper by Rosenblatt (1971). The size of the planet can be determined from the depth of the transits, the orbital period from the repetition of transits, and the semimajor axis of the orbit and the incident flux at the planet from measurements of the host stars and Kepler’s third law. A paper by Borucki and Summers (1984) corrected the detection probability in the Rosenblatt paper and pointed out that ground-based observations of at least 13,000 stars simultaneously should be sufficient to detect

56 Space Missions for Exoplanet Science: Kepler/K2

1161

Jupiter-size planets in long-period orbits, but that the detection of Earth-size planets would require space-based observations. Limitations to the detectability of small planets caused by stellar variations were recognized (Borucki et al. 1985). They concluded that F-, G-, and K-type stars would be the best targets and that the observations could also detect brightness variations caused by acoustic (p-mode) oscillations of the brightest stars. Starting in 1983, NASA funded a small program to construct and test detectors and photometers and to develop a prototype multichannel photometer capable of detecting Earth-size planets (Borucki et al. 1988a, b, c). Laboratory tests of a CCD detector were conducted at the University of California Lick Observatory that showed that CCDs could produce 10-ppm photometric precision for an ensemble of simulated stars when systematic errors were measured and corrected (Robinson et al. 1995). An automated observatory was developed at Lick Observatory that demonstrated that the simultaneous photometry of thousands of stars with the precision required to detect extrasolar planets was possible (Borucki et al. 2001). Following its development, a laboratory facility was constructed at NASA Ames to test a prototype photometer that demonstrated 10-ppm photometric precision in the presence of the noise sources expected from on-orbit operation. This paper outlines the development of the Kepler Mission, presents its technical capabilities, and presents two examples of the results of the science program. Section “Proposals to the Discovery Program and Validation of Technical Readiness” discusses the series of proposals and the technical demonstrations that proved the technical readiness of the approach and that led to the acceptance of the mission. Section “Mission Goals, Science Requirements, Key System Features, and Ancillary Programs” presents the mission goals, science requirements, key system features, and descriptions of the programs for education and public outreach, participating scientists, and guest observers. The flight system design is covered in sections “Flight System Design” and “Instrument Operation and Performance” provides detailed information on the instrument operation and performance. Data processing and archiving are outlined in sections “Data Processing, Vetting, and Archiving” and “Mission Completion and Transition to K2” describes the mission completion and the transition to the K2 mission. Section “Representative Science Results” presents some representative science results. Section “Summary” provides the summary and conclusion. To a large extent, the material in these sections summarizes that from Borucki (2016).

Proposals to the Discovery Program and Validation of Technical Readiness The Kepler Mission was a NASA “Discovery-class” mission. This type of mission (i.e., a “PI-led mission”) is distinct from larger missions that are organized by NASA and assigned to a NASA Center for execution. Discovery-class missions are led by a principal investigator (PI) that organizes the team, chooses the scientific

1162

W. J. Borucki

objectives and requirements (with the help of the science team), and selects the NASA and industrial partners who provide mission management and develop the flight instrument and spacecraft. The Discovery-class missions were designed to have shorter development times, to cost substantially less than flagship missions, and to quickly address topical questions. The Kepler Mission was designed and proposed to be consistent with this philosophy. Proposals for the Kepler Mission were presented five times (1992, 1994, 1996, 1998, and 2000) before being accepted for development (Borucki 2016). In the first two proposals, the mission concept was titled the “FRESIP (FRequency of Earth-size Inner Planets)” Mission (Borucki et al. 1996). At the urging of several members of the science team, the title of the later proposals was changed to “Kepler Mission” to honor the sixteenth century astronomer Johannes Kepler who developed the laws of planetary motion. The mission was proposed at each Discovery Program opportunity starting in 1992 and culminated in the acceptance of the proposal submitted in 2000. During this period, the mission design evolved in response to review panel comments and suggestions. However, the basic goals and features of the instrument did not change appreciably. The core of all proposals was the determination of the frequency of Earth-size planets, the size and semimajor distributions of planets, and the existence of planets in multiplestar systems orbiting Sun-like stars. In each proposal, the design of the mission was based on an orbiting telescope/photometer that would conduct a census of extrasolar terrestrial planets by observing the flux of many thousands of stars to detect the pattern of dimming caused by planetary transits. The point design for the Kepler Mission was based on a total noise value of 20 ppm for a 6.5-h transit of a 12th-magnitude solar-like star by an Earth-size planet using a 1 m aperture telescope. The stellar noise was assumed to be similar to that of the Sun, i.e., 10 ppm for periods similar to those of transit durations (Willson et al. 1981; Jenkins 2002). With shot noise of 15 ppm and a stellar variability of 10 ppm, an instrument with a precision of 7 ppm for a 6.5-h transit duration was required to obtain a root-sum-square (rss) noise of 20 ppm. This performance provides a signal-to-noise ratio (SNR) of about 4 per transit. Because of the large number of stars surveyed and because the search covered periods between 1 day and 1.33 years, about 1011 statistical tests were required to search for patterns of transits for 170,000 stars. Consequently, to reduce the expectation of a false-alarm (FA) rate due to statistical fluctuations to less than 1 FA (assuming Gaussian noise) over the entire mission duration, the stellar flux time-series data were examined only when the detected transit pattern had a total SNR that exceeded 7.1¢ (Jenkins et al. 2002). Statistically, the transit patterns that just meet this threshold can be recognized only 50% of the time, while those with a transit pattern with an 8¢ detection can be recognized 84% of the time. To obtain at least 50 planets in the HZ of a Sun-like star, the mission was designed to monitor at least 170,000 stars for a period of 4 years to observe a minimum of four transits. Because astrophysical phenomena also produce events that mimic planetary transits, methods to recognize and remove both FA and astrophysical false-positive events were part of the mission development (Couglin et al. 2014).

56 Space Missions for Exoplanet Science: Kepler/K2

1163

Mission Goals, Science Requirements, Key System Features, and Ancillary Programs In 2000, NASA received 26 proposals to their AO-00-02-OSS-020 Discovery-class flight opportunity. The Kepler Mission proposal was one of the three selected for further competition. After submission of the required Concept Study Report (CSR), the Kepler Mission was chosen as Discovery Mission #10. As stated in the 2000 Discovery proposal and in the CSR, the general goal of the mission was the exploration of the structure and diversity of planetary systems, with a special emphasis on the detection of Earth-size planets in the HZs surrounding other stars. The specific goals of the mission were: 1. Determine the frequency of 0.8 Earth-radii and larger planets in or near the HZ of a wide variety of stars. 2. Determine the distributions of sizes and orbital semimajor axes of these planets. 3. Estimate the frequency and orbital distributions of planets orbiting multiple-star systems. 4. Determine the distributions of semimajor axis, albedo, size, mass, and density of short-period giant planets. 5. Identify additional members of each photometrically discovered planetary system using complementary techniques. 6. Determine the properties of those stars that harbor planetary systems. The specific mission requirements needed to accomplish the science objectives are listed in Tables 1 and 2. The values listed in these tables are representative of the “as-built” instrument and differ somewhat from those in the CSR. See Borucki (2016) for more detailed descriptions of the CSR, mission characteristics, and for the team membership. Three ancillary programs were conducted as part of the mission: 1. An education and public outreach program (EPO) program to capitalize on the public excitement associated with the discovery of Earth-size planets to

Table 1 Science Requirements Requirement Target stars

System photometric precision Continuous observing Mission lifetime

Required value Monitor 170,000 stars at a 30-min cadence in a single FOV. Monitor 512 stars at 1-min cadence for p-mode observations. The subset can be changed every 3 months. 1.9  105 (19 ppm) total for 6.5-h integration for 12th-mag G2 dwarf Single, inertial FOV for targets. No obscuration by Sun, Earth, Moon, and planets 4 years to observe four transits of planets in 1-year orbits

1164

W. J. Borucki

Table 2 Key features and enabling design Key features Wide field of view Large focal plane with replaceable detector modules Tolerant focus error budget Detectors

Passive focus retention Stable spacecraft Orbit type Data collection No mechanisms

Enabling design Schmidt-design telescope: 0.95 m aperture Schmidt corrector; f#1.0,1.4 m mirror; 113 sq deg active detector area Module approach for CCD packaging (two channels per CCD, two CCDs per module). One design repeated 21 times Images flattened by sapphire field-flattening lenses to compensate for strongly curved focal surface 42 four-phase, 1,024 rows  2,048 column (plus masked and overclocked rows and columns) CCDs with charge injection, 27 pixels, two readouts per CCD. Operation at 85 ı C Low thermal expansion materials for metering structure Heliocentric orbit to minimize thermal variations Earth-trailing heliocentric, 372.5-day period Parallel electronics with serial readout strategy No shutter, quarterly rolls using a fixed solar array, fixed antennas Reaction wheels are only exception

stimulate student learning, to motivate teachers, and to advance public interest in astronomy and physics 2. A Participating Scientist Program that invited community members to participate in the Kepler Science Team activities such as data analysis, archiving, and publication 3. A Guest Observer (GO) Program that provided mission resources to the astrophysical science community to observe astrophysical objects not otherwise being observed by the exoplanet program

Flight System Design The Kepler Mission relied on precise differential photometry to detect the slight signal variations due to the transits of small planets, e.g., a transit of Earth across the Sun causes only an 84-ppm decrease in intensity. Such precision also required superb instrument stability on timescales up to several times the duration of the transits, systematic error removal to much better than 20 ppm, and pointing precision of 0.01 arc seconds based on half-hour samples. The following material discusses the methods used to obtain the needed precision.

Flight System Structure The flight segment (FS) consisted of the photometer and supporting spacecraft, both built by the Ball Aerospace and Technologies Corporation (BATC). Figure 1a is a cutaway diagram of the Kepler photometer. The basic design is a Schmidt optical system with the array of 42 CCD detectors at the prime focus. Focus mechanisms are

56 Space Missions for Exoplanet Science: Kepler/K2

1165

Fig. 1 (a) Cutaway view showing the optics, detector array, detector electronics, thermal radiator, and sunshade (Courtesy of BATC). (b) Detector array with sapphire field-flattening lenses and the four small fine guidance sensors (FGS) in the corners of the array (Courtesy of BATC). (c) Cartoon showing the mirror composed of slabs of fused silica frit-bonded to a hollow hexagonal structure to reduce its mass by 87%. Also shown are the focus mechanisms attached to the spacecraft backplate (Courtesy of BATC). (d) Photograph of the mirror and its focus and mounting mechanisms prior to coating the mirror (Courtesy of BATC)

mounted at the rear of the primary mirror. The CCDs have been custom-designed for this application. They are passively cooled to –85ı C to make the effects of dark current and radiation damage negligible. Four fine guidance sensors (FGS) are mounted near the corners of photometer focal surface to ensure stable pointing and for accurate repositioning after the monthly rotations to download the data. The FS is rotated 90ı every quarter spacecraft-year to keep the solar arrays pointed at the Sun and the thermal radiator pointed to deep space. A large sunshade allows the telescope to point to within 55ı of the Sun without any light coming into the sunshade or the optics (Fig. 1a). Below the sunshade is a 0.95-m clear-aperture Schmidt corrector that corrects for spherical aberration associated with the large FOV and the spherical mirror. A 1.4 m aperture mirror is at the lower end of the telescope structure that focuses the light onto the detector array that is at the prime focus. The detectors are passively cooled by two heat pipes that conduct the heat from the detector array to the thermal radiator. The electronics

1166

W. J. Borucki

Fig. 2 (a) Sketch of the instrument mated to the spacecraft. The two star trackers and two of the four reaction wheels are visible on the lower left (Courtesy of BATC). (b) Assembled flight system in the BATC clean room with the telescope dust cover in place. Dust cover was ejected after launch. Gold-colored material is the thermal insulation. Note the person at the lower left for scale (Courtesy of BATC)

are directly behind (i.e., just above in the figure) the detector array and are cooled by radiating to space through the corrector. Each pair of detectors is combined to form a square module, and each module is fitted with a 5-cm-square-coated sapphire lens shown in Fig. 1b. These lenses are used to correct the focus at the detectors because the focal surface is highly curved while the detectors are flat. The mirror has been reduced in mass by 87% from that of a solid mirror and is composed of ultralow-expansion glass with a multicoated antireflection coating over the silver coating. Figure 1c, d show the hollow mirror structure, the mounting and focusing mechanisms, the backplane structure that supports the instrument, and the ring that attaches the spacecraft to the rocket booster. A sketch of the instrument mated to the spacecraft is shown in Fig. 2a, and a photo of the assembled flight system is shown in Fig. 2b. The solar panels are mounted to the hexagonal structure that forms the spacecraft and to which are mounted the various spacecraft components such as the omni and high-gain antennas, thrusters, star trackers, transmitter and receiver, battery, reaction wheels, and computers.

56 Space Missions for Exoplanet Science: Kepler/K2

1167

Field-Flattening Lenses and Wavelength Response Much of the spectral power of the stellar UV variability is caused by the Ca II H and K lines that are features of stellar spectral types from F4 to M. To reduce the variability introduced by these lines, an interference filter was coated onto the backsides of the field-flattening lenses. The bandpass filter limits the transmission response at 423 nm and 897 nm to 5% to remove both the stellar UV variability and the variability from “fringing” of reflected light in the near-IR due to the transparency of silicon detectors in this region. The field-flattening lenses and the pairs of detectors in each module are visible in Fig. 1b.

Selection of FOV To monitor the maximum number of appropriate stars, the FOV was pointed at an area of the celestial sphere rich in stars, i.e., near the galactic equator. Other constraints included: 1. FOV must be above 55ı ecliptic latitude to keep the Sun from shining into the sunshade and scattering light into the telescope. 2. Very bright stars must be completely off the focal plane to avoid scattered light that would produce a high background level and could saturate large portions of the detector array. 3. Other bright stars in the FOV must be placed on the blackened areas that lie between the detectors. To reduce the fraction of giant stars in the FOV, the position of the FOV was chosen to be slightly above the galactic equator, i.e., l, b D 76.53ı , C13.29ı (Fig. 3a, b).

Selection and Characterization of Target Stars: Kepler Input Catalog To select appropriate targets for the mission, spectrophotometric observations of the 4.4  106 stars in the Kepler FOV were conducted by a team led by David Latham at the Smithsonian Astrophysical Observatory (SAO). The results of the analysis (Brown et al. 2011) were used to generate the Kepler Input Catalog (KIC) from which the target stars were chosen. The selected stars were primarily latetype dwarfs selected to maximize numbers that are both bright and small enough to show detectable transit signals for small planets in and near the HZ (Batalha et al. 2010). Although the KIC enabled the selection of the most promising stars, it did not provide accurate estimates of stellar sizes that were needed to get accurate planet sizes and incident fluxes. Huber et al. (2014) generated improved values for the stellar properties of the target stars based on recently available stellar spectra and model results.

1168

W. J. Borucki

Fig. 3 (a) Position of the Kepler FOV relative to the galactic plane (Photo of galaxy by Carter Roberts, astrophotographer, with permission). (b) Superposition of detectors on the Kepler FOV (Star chart from the Sky Software with permission (Bisque.com))

Fig. 4 Kepler orbit. Average distance from the Earth increases by approximately 0.12 AU/year (NASA image)

56 Space Missions for Exoplanet Science: Kepler/K2

1169

Table 3 Kepler orbital parameters, epoch JD2456974.5 (11/13/2014) (reference frame, Earth Mean Orbit of J2000) Period (days) Semimajor axis (km) Eccentricity Inclination (deg.) Longitude of node(deg.) Argument of periapsis (deg.) Mean anomaly (deg.)

372.5038 ˙ 0.0001 151,570,124 ˙ 40 0.0360438 ˙ 0.0000005 0.44739 ˙ 0.00002 w.r.t. ecliptic 154.082 ˙ 0.0020 w.r.t. ecliptic 65.893 ˙ 0.003 74.7977 ˙ 0.0008

Orbit and Commissioning On 9 March 2009, the Kepler spacecraft was launched into a heliocentric orbit (Fig. 4) with a period of 372.5 days. The orbital parameters are given in Table 3. The 67 days of commissioning (6 March through 11 May 2009) included several tests to characterize the photometer performance prior to the ejection of the aperture cover, e.g., bias level, dark current, energetic particle flux distribution, and noise for all 84 channels. The first-light image taken on 8 April showed nominal operation for all detectors. This step was followed by the calibration of the fine guidance sensors, a checkout of focal plane geometry, and measurements of the focus and scattered light. Science operations began on 12 May (13 May UTC) 2009.

Instrument Operation and Performance Detector Properties The detectors were manufactured by the E2V Corporation, and all detectors have 1,024 rows by 2,200 columns with 27  27 pixels. Each detector has two separate readouts: each reads out a section of 1,024 rows by 1,100 columns plus overclocked rows and columns (Caldwell et al. 2010). The focal plane consists of 84 separate science readout channels and four fine guidance sensor channels. The fine guidance sensors (FGS) at the four corners of the focal surface are read out continuously at 10 Hz. There are exactly five FGS readouts per science CCD readout, and all are read out synchronously (albeit staggered) so that all the clock pulse phases for all readouts repeat exactly each time a pixel is read out. Each detector channel has several regions available to collect calibration data as discussed below. The detectors use four-phase architecture to increase the well depth to reduce saturation of bright stars. The science CCDs are operated with an electrical charge injection feature that injects charge at the top of each CCD for four consecutive rows in the center of

1170

W. J. Borucki

the virtual smear region at a signal level approximately 40% of full well. The signal appears entirely in the virtual smear and does not result in the overwriting of any stars in the science FOV. Charge injection serves the dual purpose of filling radiation-induced traps in the CCDs and providing a stable signal for monitoring the charge transfer efficiency of the readout electronics.

Saturation and Dynamic Range The high quantum efficiency of the back-illuminated CCDs combined with the broad bandpass causes the central pixel of the star images for most stars brighter than magnitude 11.3 to saturate. Measurements show that the photometric precision of saturated images can match those for unsaturated stars if the pixel aperture is chosen large enough to capture all the electrons that bleed down the columns. By using apertures of several hundred pixels, stars as bright as magnitude 4.5 have been monitored, i.e., ™-Cygni.

Read Noise Because of the large number of photons collected from typical targets, a low value for the read noise is not critical, and the focal plane median value of 95 electrons per readout, or approximately 1 digital number (DN) per readout, is satisfactory (Caldwell et al. 2010). The CCD dark currents are negligible because the focal plane is maintained at 85 ı C. The Kepler CCDs are operated such that the full well of 1.1 million electrons is not clipped by the 14-bit analog-to-digital converter.

Smear To avoid the possibility that a mechanical shutter could malfunction while in the closed position thereby preventing further images and terminating the science operations, the instrument does not include a shutter. Consequently, the stars in the FOV briefly overwrite the images as they are clocked down the columns, i.e., the images are “smeared.” Each channel has several regions available to collect calibration (“collateral”) data that provides information needed to correct for smear, bias, and background. There are 12 leading and 20 trailing virtual column readouts in the serial register and 26 virtual row readouts after the actual columns and rows. Twenty rows at the bottom of each CCD are covered by an aluminum mask to measure smear, bias, and artifacts from instrument noise (Caldwell et al. 2010). The “smear” caused by the additional flux added as rows are clocked across star images during readout is measured for each column of each LC image and subtracted. Smear values are typically small, since each pixel only “sees” a star for the short readout time (0.52 s) divided by the total number of rows in a column.

56 Space Missions for Exoplanet Science: Kepler/K2

1171

Table 4 Global roll-up of noise terms (Gilliland et al. 2011) Component Intrinsic stellar Poisson C readout Intrinsic detector Quarter dependent Total

Variance (ppm2 ) 380.5 283.0 116.2 60.1 839.8

Noise (ppm) 19.5 16.8 10.8 7.8 29.0

Baseline noise (ppm) 10.0 14.1 10.0 20.0

On-Orbit Performance Table 4 summarizes the measured on-orbit photometric performance. Although the photometric precision is beyond anything previously attained, the precision was reduced from the expected values mostly because the stellar variability was about double the predicted values. In particular, the high level of stellar variability of Sunlike stars was not expected because two-thirds of G dwarfs were expected to be older and less variable than the Sun. The results of Gilliland et al. (2011) show that most G dwarfs are more variable than the Sun at the timescales of transit durations, 3–12 h. However, Basri et al. (2013) find that for periods longer than 6 h, the variability of the Sun appears to be quite similar to other G dwarfs. See Borucki (2016) and references therein for a more comprehensive discussion of the instrument design and performance.

Data Processing, Vetting, and Archiving Data for all stars were recorded at a cadence of one per 29.4244 min (hereafter, long cadence or LC). Data for a subset of up to 512 stars were also recorded at a cadence of one per 58.85 s (hereafter, short cadence or SC). The SC observations were sufficiently short to provide high resolution of the ingress and egress times of planetary transits and to conduct asteroseismic observations needed for measurement of the sizes, masses, and ages of dwarf stars. The LC data have 270 readouts of slightly more than 6.5-s duration (e.g., 6.01982-s integration and 0.51895-s readout time) and are co-added to 29.4-min intervals. Each SC is the sum of nine exposures. For a full discussion of the LC data and their reduction, see Jenkins et al. (2010b, c). The SC data are discussed in Gilliland et al. (2010). Each month about 1.1 GBytes of engineering data and 9 GBytes of science data were downlinked by way of the Deep Space Network using Ka band for the science and stored engineering data and using X-band for real-time engineering data. The onboard storage was designed to save 2 months of data before overwriting the data. This procedure allowed a second opportunity to return the data if errors were found in the previous transmission or if a ground station pass was missed. Besides reducing the number of cadences for some targets due to the Module 3 failure (9 Jan 2010), each quarterly time series contains gaps, some much larger

1172

W. J. Borucki

than others. These gaps are due to a variety of occurrences including monthly breaks for data downlink, occasional safe-mode events, manually excluded cadences, loss of tracking at the precision required for science data acquisition, attitude tweaks, and solar storms. All missing cadences are tabulated in the Anomaly Summary Table in section “Data Processing, Vetting, and Archiving” of the Data Release Notes. These Notes are archived at MAST (http://archive.stsci.edu/Kepler/Kepler_ fov/search.php). Results from the data analysis are archived at NASA’s Exoplanet Archive (http://exoplanetarchive.ipac.caltech.edu). Pixel data downloaded from the spacecraft were converted to instrumental fluxes via Kepler pipeline software modules that calibrated pixel data (Quintana et al. 2010), performed aperture photometry (Twicken et al. 2010), and corrected for systematic errors (Stumpe et al. 2012, 2014; Smith et al. 2012). The pipeline software is documented in the Kepler Data Processing Handbook (KSCI-19081) located at MAST. Transit signals were identified using an adaptive, wavelet-based-matched filter that explicitly took the power spectral density of the observation noise (stellar variability C shot noise C residual instrument noise) into account in formulating the detection statistics for each light curve (Jenkins et al. 2010c). Because of the presence of false alarms caused by statistical fluctuations in the measurements as well as astrophysical sources that mimic planetary transits, it was necessary to thoroughly vet the data. See Bryson et al. (2013), Couglin et al. (2014), Lissauer et al. (2014), Rowe et al. 2014, and Batalha (2014) for discussions of the procedures used See Morton et al. 2016 for an estimate of the false positive probabilities. Data at each stage of the pipeline analysis (Jenkins et al. 2010a), from pixellevel to the error-corrected light curves, are stored at the MAST and are available to the public. The Kepler Instrument Handbook (KSCI-19033) and the Kepler Data Characteristics Handbook (KSCI-19040) (2016) describe the acquisition and analysis of the data and are also available at the MAST.

Mission Completion and Transition to K2 The spacecraft needed a minimum of three reaction wheels to keep the photometer pointed precisely at the chosen star field and to periodically rotate the spacecraft to point the high-gain antenna toward the Earth to download the engineering data and science data. The pointing was determined from four fine guidance sensors (FGS) on the focal plane that monitored a minimum of ten stars each. Because the Kepler FOV was so large, the stars at the edges of the FOV moved as much as ˙0.6 pixel over a season (quarter of a spacecraft year). The pointing performance was excellent, 0.0100 3¢ for durations of 30 min. Shortly after 4 years of operation, a second reaction wheel failed (14 May 2013) causing the cessation of data acquisition. In the following months, the scientific effort was focused on determining the frequency and orbital distribution of the small exoplanet candidates (Rp < 2.5R˚ ) and on providing well-documented legacy

56 Space Missions for Exoplanet Science: Kepler/K2

1173

products to enable the science community to examine the analysis and results and to revise them if appropriate.

Transition to the K2 Mission After the loss of two of the reaction wheels, a search for an alternative use for the Kepler spacecraft was conducted. Engineers at BATC found that the effects of the torque caused by solar photon flux could be minimized by pointing the spacecraft in the direction of the orbit and adjusting the roll angle to balance the solar torque on the solar panels. By frequently correcting the roll angle with reaction jets and by moving the FOV to new locations along the ecliptic four times per year, high-precision photometry (80 ppm for 6 hr integrations of a 12th magnitude quiet star) could be conducted for more than 10,000 targets in each FOV (Howell et al. 2014). The new mission (“K2”) is currently operating as a community facility to observe targets proposed by the individual researchers. K2 results are providing new discoveries of short-period planets (Montet et al. 2015). However, three (i.e., modules #3, #4, and #7) of the 21 detector modules are no longer functioning, thereby reducing the FOV to 95 sq deg.

Representative Science Results During the 4 years of operation, Kepler made an unprecedented set of timeseries observations of 190,000 stars and discovered over 4,600 planetary candidates that range in size from slightly larger than the Moon to planets over three times larger than Jupiter and have orbital periods from a few hours to several years. By combining these observations with those of RV and transit-timing variations (TTV) to get masses and densities (Nesvorny and Morbidelli 2008; Masuda 2014), rocky planets were distinguished from low-density (water, gas, and ice) planets (Marcy et al. 2014). Ten planets were found orbiting binary stars (Orosz et al. 2012; Welsh et al. 2015), and several planets were found in orbits that are at large angles relative to the stellar equator. Figure 5a shows the distribution of 4,600 planets and planetary candidates found by Kepler based on the data from the first 47 months of science operations (Mullally et al. 2015). Most of these planets are between the size of Earth and Neptune, i.e., sizes not found in our Solar System. Because of the large bias introduced by the difficulty of detecting small planets versus large planets orbiting noisy and/or large stars, the correction for the size bias is expected to greatly increase their fraction. Similarly, the bias associated with the relatively low number of transits that occur for long orbital periods could be the explanation for the paucity of small planets with longer orbital periods seen in the lower right-hand portion of Fig. 5a. Based on models of the Kepler search that correct for such biases, the average number of planets per star with sizes between 1R˚ < Rp < 5R˚ is estimated to

1174

W. J. Borucki

be approximately two for orbital periods between 3 and 200 days (Boviard and Lineweaver 2017). Currently, 24 Earth-size and super-Earth-size planets have been found in the HZ. Diagrams of two examples of verified planetary systems with terrestrial planets in the HZ are shown in Fig. 5b, c. The Kepler-62 planetary system (Borucki et al. 2013) of five planets orbits a star somewhat cooler (4,925ı K vs. 5,780ı K) and smaller than the Sun (0.64 vs. 1.00 Rˇ ). Both planets are somewhat larger than the Earth (1.4 R˚ and 1.7 R˚ , respectively) and could be “water planets” (Leger et al. 2004; Kaltenegger et al. 2013). A planet much closer in size to the Earth (1.1 ˙ 0.2 R˚ ) was found (Quintana et al. 2014) in the HZ of a five-planet system (Kepler-186) orbiting an M-dwarf star (3,761ı K and R* D 0.46 Rˇ ). A planetary system where three of the seven Earth-size planets are in HZ has also been discovered (Gillon et al. 2017) orbiting an even cooler and smaller star (2,550 K, i.e., 2,000 times less luminous than the Sun and with R* D 0.11Rˇ ) The Kepler Mission also generated time series of high-precision stellar brightness measurements that span months or years that are valuable for asteroseismic studies of over 14,000 stars, including dwarfs, subgiants, and red giants. The data show (Fig. 5d) a rich spectrum of overtones of oscillations and pulsations from pressure-mode (p-mode) and gravity-mode (g-mode) waves. These measurements allow model-based estimation of fundamental stellar properties, including size, density, and age (Chaplin and Miglio 2013). Gilliland (2011) describes the international organization that was formed to conduct asteroseismology investigations using the Kepler Mission results.

Summary During the 4 years of operation, Kepler made an unprecedented set of time-series observations of 190,000 stars (170,000 program stars plus 20,000 Guest Observer stars) and discovered over 4,600 planetary candidates (confirming 3,548 as planets) that range in size from slightly larger than the Moon to planets over three times larger than Jupiter. By combining these observations with those of RV and TTV to get masses and densities, rocky planets were distinguished from low-density (water, gas, and ice) planets. Ten planets were found orbiting binary stars, and several planets were found in orbits that are at large angles to the stellar equator. Because of the large number of exoplanets discovered and characterized, useful estimates of the parent distributions have been derived. The results show that most stars have planets, many of these planets are Earth-size, and a significant fraction is in the HZ of the host stars. The Kepler Mission also made substantial contributions to astrophysical sciences, including heartbeat stars, white dwarfs, and supernovae. Observations of the photometric variations of stellar brightness show rich spectra of overtones of oscillations and pulsations from pressure-mode (p-mode) and gravity-mode (gmode) waves. These measurements provide accurate values for stellar and planet sizes as well as provide information on the interior structure of stars.

Fig. 5 (a) Kepler observations of exoplanet candidates and confirmed planets (NASA image). (b) Comparisons of the Kepler-62 and planetary systems relative to the solar system (Borucki 2016). (c) Comparisons of the Kepler-186 planetary systems relative to the solar system (Borucki 2016). (d) Power spectra of five Kepler stars including dwarfs, subgiants, and giants (Chaplin and Miglio 2013, reprinted with permission, copyright 2013 Annual Reviews)

56 Space Missions for Exoplanet Science: Kepler/K2 1175

1176

W. J. Borucki

The stated goal of the mission “ : : : the exploration of the structure and diversity of planetary systems, with a special emphasis on the detection of Earth-size planets in the HZs surrounding other stars” has largely been achieved. Acknowledgments Kepler was competitively selected as the tenth Discovery mission with funding provided by NASA’s Science Mission Directorate. The mission success was accomplished through the strenuous efforts of BATC personnel to develop the instrument and spacecraft; the space controllers at Laboratory for Atmospheric and Space Physics; the managers at NASA HQ, Ames, Marshall, and Jet Propulsion Laboratory; the Kepler team that analyzed the data and provided the science results; the thousands of members of the worldwide science community who were responsible for many of the exciting discoveries; the SETI Institute and Lawrence Hall of Science personnel who provided education and public outreach; and the teams at Space Telescope Science Institute and California Institute of Technology who archived and provided the data to the community. Other organizations that made important contributions to the mission includes Carnegie Institute of Washington, Harvard-Smithsonian Center for Astrophysics, W. M. Keck Observatory, University of California Lick Observatory, Lowell Observatory, NASA Goddard Space Flight Center, NASA Kennedy Spaceflight Center, University of Aarhus, University of California Berkeley, University of Texas Austin, and University of Washington Seattle.

References Basri G et al (2013) Comparison of Kepler photometric variability with the sun on different time scales. ApJ 769:37 Batalha NM (2014) Exploring exoplanet populations with NASA’s Kepler mission. PNAS 1111264B Batalha NM et al (2010) Selection, prioritization, and characteristics of Kepler target stars. ApJL 713:L109 Borucki WJ (2016) Kepler mission: development and overview. Repts Prog Phys 79:036901 Borucki WJ, Summers AL (1984) The photometric method of detecting other planetary systems. Icarus 58:121 Borucki WJ et al (1985) Detectability of extrasolar planetary transits. ApJ 291:852 Borucki WJ et al (1988a) High precision photometry with fiber optics. ASPC 3:247B Borucki WJ et al (1988b) A photometric approach to detecting Earth-size planets. ASSL 144:107B Borucki WJ et al (1988c) Tests of a multichannel photometer based on silicon diode detectors. NASA Conf. Publ., NASA CP-10015, 47–55 Borucki WJ et al (1996) FRESIP: a mission to determine the character and frequency of extra-solar planets around sun-like stars. A&SS 241:111B Borucki WJ et al (2001) The vulcan photometer: a dedicated photometer for extrasolar planet searches. PASP 113:439–451 Borucki WJ et al (2013) Kepler-62: a five-planet system with planets of 1.4 and 1.6 Earth radii in the habitable zone. Science 340:587 Boviard T, Lineweaver C H (2017) A flat inner disk model as an alternative to the Kepler dichotomy in the Q1 to Q16 planet population. MNRAS accepted for pub Brown TM et al (2011) Kepler input catalog: photometric calibration and stellar classification. AJ 142:112 Bryson ST et al (2013) Identification of background false positives from Kepler data. PASP 125:889 Caldwell DA et al (2010) Instrument performance in Kepler’s first months. ApJL 713:L92 Chaplin WJ, Miglio A (2013) Asteroseismology of solar-type and red-giant stars. Annu Rev Astron Astrophys 51:353–392

56 Space Missions for Exoplanet Science: Kepler/K2

1177

Couglin J et al (2014) Contamination in the Kepler field. Identification of 685 KOIs as false positives via ephemeris matching based on Q1–Q12 data. AJ 147:119 Gilliland RL (2011) Asteroseismology of cool dwarfs and giants with Kepler. ASPC 448:167G Gilliland RL et al (2010) Initial characteristics of Kepler short cadence data. ApJL 713:L160 Gilliland RL et al (2011) Kepler mission stellar and instrument noise properties. ApJS 197:6 Gillon M et al (2017) Temperate Earth-sized planets transiting a nearby ultracool dwarf star. Nature 533:221–224 Howell SB et al (2014) The K2 mission: characterization and early results. PASP 126:398–408 Huber D et al (2014) Revised stellar properties of Kepler targets for the quarter 1–16 transit detection run. ApJS 211:2 Jenkins JM et al (2002) Some tests to establish confidence in planets discovered by transit photometry. ApJ 564:495–507 Jenkins JM (2002) The impact of solar-like variability on the detectability of transiting terrestrial planets. ApJ 575:493–505 Jenkins JM et al (2010a) Overview of the Kepler science processing pipeline. ApJL 713:L87 Jenkins JM et al (2010b) Initial characteristics of Kepler long cadence data for detecting transiting planets. ApJL 713:L12 Jenkins JM et al (2010c) Transiting planet search in the Kepler pipeline. SPIE 7740:77400D Kaltenegger L et al (2013) Water-planets in the habitable zone: atmospheric chemistry, observable features, and the case of Kepler-62e and -62f. ApJL 775:L47 Kasting JE (1993) Habitable zones around main sequence stars. Icarus 101:108 Kepler Data Characteristic Handbook (KSCI-19040) (2016), Section 5.2, Milkuski archive at STScI. http://archive.stsci.edu/kepler Kepler Data Processing Handbook (KSCI-19081) (2017) Milkuski archive at STScI. http://archive.stsci.edu/kepler Kepler Instrument Handbook; (KSCI-19033) (2016) Milkuski archive at STScI. http://archive.stsci.edu/kepler Koch DG et al (2010) Kepler mission design, realized photometric performance, and early science. ApJL 713:L79 Kopparapu RK et al (2014) Habitable zones around main-sequence stars: dependence on planetary mass. ApJL787: L29. Leger A et al (2004) A new family of planets? “ocean planets”. Icarus 169:499 Lissauer JJ et al (2014) Validation of Kepler’s multiple planet candidates. II. Refined statistical framework and descriptions of systems of special interest. ApJ 784:44L Marcy GW et al (2014) Masses, radii, and orbits of small Kepler planets: the transition from gaseous to rocky planets. ApJS 210:20 Masuda K (2014) Very low density planets around Kepler-51 revealed with transit timing variations and an anomaly similar to a planet-planet event. ApJ 783:53 Montet BJ et al (2015) Stellar and planetary properties of K2 campaign 1 candidates and validation of 18 systems, including a planet receiving Earth-like insolation. ApJ 809:25 Morton T et al (2016) False positive probabilities for all Kepler objects of interest: 1284 newly validate planets and 428 likely false positives. ApJ 822:86 Mullally F et al (2015) Planetary candidates observed by Kepler VI. Planet sample from Q1–Q16 (47 months). ApJS 217:31 Nesvorny D, Morbidelli A (2008) Mass and orbit determination from transit timing variations of exoplanets. ApJ 688:636 Orosz JA et al (2012) Kepler-47: a transiting circumbinary multiplanet system. Science 337:1511O Quintana EV et al (2014) An Earth-sized planet in the habitable zone of a cool star. Science 344:277 Quintana EV et al (2010) Pixel-level calibration in the Kepler science operations center pipeline. SPIE 7740: 77401X Robinson LB et al (1995) Test of CCD limits for differential photometry. PASP 107:1094–1098 Rosenblatt F (1971) A two-color photometric method for detection of extra solar planetary systems. Icarus 14:71R

1178

W. J. Borucki

Rowe JF et al (2014) Validation of Kepler’s multiple planet candidates. III. Light curve analysis and announcement of hundreds of new multi-planet systems. ApJ 784:45 Smith JC et al (2012) Kepler presearch data conditioning II –a Bayesian approach to systematic error correction. PASP 124:919 Stumpe MC et al (2012) Kepler presearch data conditioning I – architecture and correction in Kepler light curves. PASP 124:985S Stumpe MC et al (2014) Multiscale systematic error correction via wavelet-based band splitting in Kepler data. PASP 126:100S Twicken JD et al (2010) Photometric analysis in the Kepler science operations center pipeline. SPIE 7740E:23T Welsh W et al (2015) Kepler 453b – the 10th Kepler transiting circumbinary planet. ApJ 809:26W Willson RC et al (1981) Observations of solar irradiance variability. Science 211:700

Observing Exoplanets with the Spitzer Space Telescope

57

Charles A. Beichman and Drake Deming

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . History of Spitzer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Instrument Suite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Exoplanet Science Themes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Searches for Planetary Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Debris Disks as Signposts of Exoplanets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Incidence of Debris Disks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Disks and Planets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Brown Dwarfs as Exoplanet Analogs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Spitzer and Microlensing Planets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Characterization of Known Planets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . “First Light” from Exoplanets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Atmospheric Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Transit Photometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Validation and Ephemerides Refinement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Challenge of Data Acquisition and Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Optimization of Spitzer’s Data Acquisition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Post-processing Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Limits to Photometric Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1180 1181 1182 1182 1183 1183 1183 1186 1187 1188 1189 1189 1190 1194 1195 1197 1198 1199 1200 1200 1201

C. A. Beichman () NASA Exoplanet Science Institute, California Institute of Technology and Jet Propulsion Laboratory, Pasadena, CA, USA e-mail: [email protected] D. Deming Department of Astronomy, University of Maryland, College Park, MD, USA e-mail: [email protected] © This is a U.S. Government work and not under copyright protection in the US; foreign copyright protection may apply 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_78

1179

1180

C. A. Beichman and D. Deming

Abstract

At its launch in 2003, Spitzer did not have exoplanet science among its primary goals. Yet in the second half of its lifetime, Spitzer’s exoplanet observations came to be among its most important scientific contributions, including the detection of seven planets – three of them Earth analogs in the habitable zone – transiting the late M-dwarf star TRAPPIST-1. We discuss how Spitzer became the first telescope to detect light from a mature exoplanet, to probe the vertical and horizontal structure of exoplanet atmospheres, to validate and improve our knowledge of transiting exoplanet candidates, and to characterize planets detected via microlensing. In related research topics, Spitzer observed the debris left over from the formation of planetary systems and studied Y dwarfs, the cold, free-floating analogs of Jovian mass objects. We also discuss how Spitzer observations and post-processing techniques were optimized to make these challenging exoplanet observations possible.

Introduction The Spitzer Space Telescope has played a dramatic yet unplanned role in the ongoing revolution in exoplanet science. Although the development of Spitzer occurred in parallel with the discovery of the first radial velocity planets, the study of exoplanets was never one of its original scientific goals. During its primary “cold” mission phase Spitzer made only a few exoplanet observations, although these were dramatic. In 2004 Spitzer became the first telescope to detect light from mature exoplanets with its measurements of the eclipses of HD 209458b and TrES-1b as these planets disappeared behind their host stars. During the last years of the “cold mission” when all three of the Spitzer instruments were operational, Spitzer made a two-dimensional map of the surface of a hot Jupiter and probed the composition and vertical structure in exoplanet atmospheres. But it was in later years, after Spitzer’s cryogen was exhausted and only two short-wave IRAC channels remained operational, that Spitzer became a critical platform for exoplanet characterization and discovery. During the “warm mission” exoplanet observations grew to account for 25% of Spitzer’s overall program. Perhaps Spitzer’s greatest discovery came with the observations of the transiting system TRAPPIST-1 which Spitzer showed to host seven roughly Earth-sized planets, three in the habitable zone, orbiting a nearby late M dwarf. The story of Spitzer’s unexpected role in exoplanet research is also a tale of how a powerful observatory is capable of many more discoveries than originally thought by its original proponents and designers. The Spitzer team was able to increase Spitzer’s ability to make exoplanet measurements by an order of magnitude or more through diligent experiment design, control of the systematic errors in the telescope and instruments, and the development of powerful post-processing of the data.

57 Observing Exoplanets with the Spitzer Space Telescope

1181

History of Spitzer The Spitzer Space Telescope was recommended by the 1980 National Research Council (NRC) Decadal Review (Field et al. 1992) as an infrared telescope to be launched on a regular basis as an attached payload on NASA’s space shuttle (Werner 2006; Rieke 2006). Originally called the Space Telescope Infrared Facility, SIRTF was gradually transformed into a free-flying observatory to become the last of NASA’s four great observatories after the Hubble Space Telescope (HST), the Chandra X-ray Observatory, and the Compton Gamma Ray Observatory. After being recommended by the 1990 NRC Decadal Review (the “Bahcall Report,” Bahcall et al. 1991), SIRTF continued to undergo design refinements and await a funding line in the NASA budget. The burst of excitement in the mid-1990s which accompanied the controversial discovery of “microbial fossils” in a Martian meteorite (ALH 84001, aka the “Mars Rock”) and of the first exoplanet, 51 Pegasi b, led to an increase in the NASA astrophysics budget sufficient to start SIRTF, begin the design of the Space Interferometer Mission (SIM, later canceled in 2001), and to initiate a study for what would become the James Webb Space Telescope (JWST). SIRTF was launched on a Delta rocket on August 25, 2003, and was renamed the Spitzer Space Telescope after noted Princeton astrophysicist Lyman Spitzer (Fig. 1). Solar Panel Dust Cover

Secondary Mirror Outer Shell Primary Mirror Instrument Package • Infrared Array Camera (IRAC) • Infrared Spectrograph (IRS) • Multiband Imaging Photometer (IRS) Helium Tank

Star Trackers Spacecraft Bus

Fig. 1 The Spitzer spacecraft, telescope, and instrument complement are shown in this cutaway drawing (Courtesy Ball Aerospace)

1182

C. A. Beichman and D. Deming

Instrument Suite Spitzer hosted three instruments spanning wavelengths from 3 m to 160 m: 1. The Infrared Array camera (IRAC, Fazio et al. 2004) provided simultaneous broadband imaging at wavelengths of 3.6, 4.5, 5.8, and 8.0 m using four 256  256 detector arrays. The two short wavelength arrays used In:Sb detector material, while the two longer wavelength arrays used Si:As technology. Only the two In:Sb detector channels remained operational in the “warm mission.” 2. The Infrared Spectrometer (IRS, Houck et al. 2004) provided low (R 60–130) and moderate (R600) resolution spectroscopic capabilities from 5.2 to 38 m. 3. The Multiband Imaging Photometer for Spitzer (MIPS, Rieke et al. 2004) provided imaging at 24, 70, and 160 m and low resolution spectroscopy between 55 and 95 m. A vital part of Spitzer’s capability was not a specific instrument but its heliocentric orbit in which Spitzer slowly drifted away from the Earth, at roughly 0.1 AU per year. This orbit allowed the Spitzer telescope to achieve a low, passively cooled temperature of 30 K through careful shielding from the Sun and the absence of radiation from the Earth. This low temperature allowed a modest amount of liquid helium (360 L) to cool the telescope to as low as 5 K and portions of the instrument chamber to 10 the level of our own Kuiper Belt) is 20C8 7 % for FGK stars based on the latest Herschel

1184

C. A. Beichman and D. Deming

Fig. 2 A composite image showing the ˇ Pic’s edge-on disk as well as the ˇ Pic planet (Lagrange et al. 2010) (Image courtesy of ESO)

analysis (Montesinos et al. 2016) and consistent with earlier Spitzer estimates (Beichman et al. 2005a; Bryden et al. 2009; Sierchio et al. 2014; Montesinos et al. 2016). There is evidence for a decline in the incidence of debris disks with later spectral types and with increasing age. The sporadic nature of debris disks across a wide range of ages, from one to many Gyr, is evidence for major dynamical events stirring up the material and generating the small grains seen in these disks (Wyatt 2008). MIPS and IRAC observations along with ground-based data found debris disks around nearby white dwarfs (WDs) (Jura, Farihi and Zuckerman 2007; Farihi et al. 2009). IRS spectroscopy revealed silicate-rich, carbon-poor material similar to that found in the inner reaches of our own solar system (Jura, Farihi and Zuckerman 2009). The explanation for this material is the presence of small grains generated by collisions between minor planets or asteroids orbiting these stars in sufficient quantity to be detected in 1–3% of WDs with cooling ages 2 m, the extension of the IO approach to the mid-infrared in the 3–30 m range requires an adequate material and technological platform to manufacture high-quality optical chips. Starlight Suppression A considerable expertise has been developed on starlight suppression over the past 20 years, both in academic and industrial centers across the globe. Approximately 35 PhD theses were dedicated to this topic and more than 40 refereed papers. These efforts culminated with laboratory demonstrations at room temperature mainly at the Jet Propulsion Laboratory (JPL) in the United States. For instance, work with the Adaptive Nuller has indicated that mid-infrared nulls of 105 are achievable with a bandwidth of 34% and a mean wavelength of 10 m (Peters et al. 2010). Another testbed, the planet detection testbed, was developed in parallel and demonstrated

59 Interferometric Space Missions for Exoplanet Science: Legacy of Darwin/TPF

1.5

1245

× 10–4

1

Signal (V)

0.5 0 –0.5 –1 –1.5 0

200

400

600

800

1000 1200 Time (s)

1400

1600

1800

2000

Fig. 7 Planet signal detected with the Planet Detection Testbed (Martin et al. 2012). Markers indicate planet signal from nulling detector with star shuttered off. On this scale, the star would be 250 V. Each point is an item of data from the 2-s chop cycle, and the whole trace shows the null signal obtained over a 360ı effective rotation of the interferometer array. The line is a fit to the signal from a planet at a nominal angular radius of 6:35  107 rad (or 132 mas) from the star. By comparison, the equivalent angular fringe distance from the short baseline is 4:7  107 rad. Near the center and at the ends of the plot, the planet crosses the null fringe. Right: equivalent sensitivity map of the interferometer array. Array rotation causes the planet location to orbit (solid line) around the central null fringe (gray), and thus its signal is modulated both by the higher-frequency fringes on the long baseline and by the chopping

the main components of a high performance four-beam nulling interferometer at a level matching that needed for the space mission (see Fig. 7). At 10 m with 10% bandwidth, it has achieved nulling of 8  106 (the flight requirement is 105 ), starlight suppression of 108 after post-processing, and actual planet detection at a planet-to-star contrast of 3  107 , i.e., the Earth-Sun contrast, but with fluxes much higher than those expected from stars and planets allowing working at room temperature without being disturbed by the thermal emission of the environment. The phase chopping technique (Mennesson et al. 2005) has also been implemented and validated on sky with the Keck Nuller Interferometer (Colavita et al. 2009). A null stability of a few 103 was achieved, mainly limited by the large thermal background and variable water vapor content, both effects specific to ground-based mid-infrared observations. The next step would be to reproduce this experiment in cryogenic conditions and would require the successful validation of cryogenic spatial filters and the implementation of a cryogenic deformable mirror, which is now within reach (Enya et al. 2009). In parallel, the operation of high-precision ground-based interferometers has matured in both Europe and the United States. In particular, Europe has gained a strong expertise in the field of fringe sensing, tracking, and stabilization with the operation of the Very Large Telescope Interferometer (VLTI). In the United States, a lot of technical expertise was gained by operating several nulling interferometers such as the Keck Interferometer Nuller (KIN, Colavita et al. 2009), the Palomar Fiber Nuller (PFN, Mennesson et al. 2011b), and the Large Binocular Telescope

1246

D. Defrère et al.

Interferometer (Hinz et al. 2014). All have produced excellent scientific results (e.g., Mennesson et al. 2014; Defrère et al. 2015) and pushed high-resolution midinfrared imaging to new limits (Defrère et al. 2016). New innovative data reduction techniques have also been developed to improve the accuracy of nulling instruments (Hanot et al. 2011; Mennesson et al. 2011a), but more work is required to adapt this technique to four-telescope configurations.

Prospects Current Context Interferometric telescope arrays operating at infrared wavelengths provide already exciting science at ground-based facilities, such as ESO’s VLTI and the US CHARA, LBT, and NPOI facilities. This leads to opportunities for testing technologies on the ground and to push new developments ahead in collaboration with the existing interferometry community. Within this community, there is currently a science-driven, international initiative to develop the roadmap for a future groundbased facility that will be optimized to image planet-forming disks on the spatial scale where the protoplanets are assembled, which is the Hill sphere of the forming planets. This Planet Formation Imager (PFI, Monnier et al. 2016) shall detect and characterize protoplanets during their first 100 million years and trace how the planet population changes due to migration processes, unveiling the processes that determine the final architecture of exoplanetary systems. With 20 telescope elements and baselines of 3 km, the PFI concept is optimized for imaging complex scenes at mid-infrared wavelengths (3–12 m) and at 0.1 mas resolution, complementing the capabilities of a space interferometer that would be optimized to achieve the sensitivity and contrast required to characterize the atmospheres of mature exoplanets. A space-based interferometer and PFI will share many common technology challenges, for instance, on mid-infrared beam combination and nulling schemes, and we are well positioned to exploit synergies resulting between these projects. Regarding space agencies, NASA is currently exclusively focused on visible (and possibly near infrared) wavelengths for future exoplanet missions (internal and external coronagraphs) such as the Habitable Exoplanet Imaging mission (HabEx, Mennesson et al. 2016) and the Large Ultraviolet Optical Infrared mission (LUVOIR, Crooke et al. 2016) currently studied in preparation for the 2020 US Decadal Survey in Astronomy. Although NASA’s consideration of flagship missions capable of detecting and characterizing Earth-like planets is focused solely on direct imaging with large single apertures, there are opportunities for smaller-scale missions design to use other techniques (interferometry, astrometry, transit spectroscopy). In particular, interferometry has been identified as a key technique for future astrophysical observatories in the NASA Astrophysics Roadmap.

59 Interferometric Space Missions for Exoplanet Science: Legacy of Darwin/TPF

1247

Precursor Concepts Several precursor instruments have been proposed and realized to prepare the way for a DARWIN/TPF-I mission. In particular, NASA funded two groundbased nulling interferometers (i.e., the KIN and the LBTI, see section “Starlight Suppression”) for precursor science and technology demonstration. In Europe, a similar project was also seriously considered (GENIE, Fridlund and Gondoin 2003; Gondoin et al. 2004). Regarding space-based precursor missions, there are the free-flying demonstrators, PRISMA and PROBA-3, already discussed in section “Formation Flying.” In addition, concepts of small-scaled space-based infrared nulling interferometers were seriously considered both in Europe and in the United States. • FKSI (Fourier-Kelvin Stellar Interferometer, NASA) is a project consisting of a structurally connected infrared space interferometer with 0.5-m diameter telescopes on a 12.5-m baseline, passively cooled down to 60 K (Danchi et al. 2008). The project was studied to the phase A level by the Goddard Space Flight Center in preparation for submission as a Discovery-class mission. With an angular resolution of 40 mas at a 5 m center wavelength, FKSI would exceed the angular resolution of JWST by a factor of 5. The main scientific goals would be the detection and characterization of extrasolar planets (including super-Earth around M stars), debris disks, and exozodiacal dust (e.g., Defrère et al. 2008a). • Pegase is a project consisting of a free-flying infrared space interferometer dedicated to hot Jupiter and exozodiacal disks (Ollivier et al. 2009). It was initially proposed in the framework of the 2004 call for ideas by the French space agency (CNES) for its formation flying demonstrator mission. CNES performed a Phase 0 study in 2005 and concluded that the mission is feasible within an 8–9 years development plan, but the mission was not selected for budgetary reasons.

Required Technological Developments The main development required to bring the technology of the proposed concept to TRL 5 is the implementation of a cryogenic interferometer system that achieves the necessary starlight suppression and actual planet detection from 6 to 20 m with optical fluxes similar to the astronomical ones. To achieve this goal, preliminary system studies are required to (i) define the cryogenic design for passive cooling of the optics and active cooling of the detectors, (ii) characterize and optimize the vibrations of the interferometer in cryogenic conditions, and (iii) develop spatial filters (if needed) and beam combiners that can provide the necessary performance from 6 to 20 m under cryogenic conditions. Specific developments in terms of fringe tracking (taking into account residual vibrations) and data reduction will undoubtedly be needed to reach the required level of performance in terms of starlight rejection. We also expect that dedicated developments will be required in

1248

D. Defrère et al.

the field of mid-infrared detectors, although the JWST legacy will be particularly useful in this context. Acknowledgements The authors thank A. Léger, M. Fridlund, and B. Mennesson for reading and commenting on the manuscript. The authors would also like to thank F. Selsis, H. Rauer, M. Godolt, A. Garcia Munoz, J.L. Grenfell, and F. Tian for providing figures and/or running simulations for Proxima Cen b. This work was partly funded by the European Research Council under the European Union’s Seventh Framework Program (ERC Grant Agreement n. 337569) and by the French Community of Belgium through an ARC grant for Concerted Research Action. Some of research described in this publication was carried out in part at the Jet Propulsion Laboratory, California Institute of Technology, under a contract with the National Aeronautics and Space Administration.

Cross-References  Atmospheric Biosignatures  Atmospheric Retrieval of Exoplanets  Characterizing Exoplanet Habitability  Direct Imaging as a Detection Technique for Exoplanets  Earth: Atmospheric Evolution of a Habitable Planet  Exoplanet Atmosphere Measurements from Direct Imaging  Exoplanet Phase Curves: Observations and Theory  Future Astrometric Space Missions for Exoplanet Science  Future Exoplanet Research: High-Contrast Imaging Techniques  Space Missions for Exoplanet Research: Overview and Introduction  Spectroscopic Direct Detection of Exoplanets  The Habitable Zone: The Climatic Limits of Habitability  The Spectral Zoo of Exoplanet Atmospheres

References Angel JRP, Cheng AYS, Woolf NJ (1986) A space telescope for infrared spectroscopy of Earth-like planets. Nature 322:341–343. https://doi.org/10.1038/322341a0 Angel JR, Burge JH, Woolf NJ (1997) Detection and spectroscopy of exo-planets like Earth. In: Ardeberg AL (ed) Optical Telescopes of Today and Tomorrow. Proceedings SPIE, vol 2871, pp 516–519. https://doi.org/10.1117/12.269076 Anglada-Escudé G, Amado PJ, Barnes J, Berdiñas ZM, Butler RP, Coleman GAL, de la Cueva I, Dreizler S, Endl M, Giesers B, Jeffers SV, Jenkins JS, Jones HRA, Kiraga M, Kürster M, López-González MJ, Marvin CJ, Morales N, Morin J, Nelson RP, Ortiz J, Ofir A, Paardekooper SJ, Reiners A, Rodríguez E, Rodriguez-López C, Sarmiento LF, Strachan JP, Tsapras Y, Tuomi M, Zechmeister M (2016) A terrestrial planet candidate in a temperate orbit around proxima centauri. Nature 536(7617):437–440 Beichman C (2000) NASA’s terrestrial planet finder (TPF). In: Schürmann B (ed) Darwin and astronomy: the infrared space interferometer, vol 451. ESA Special Publication, Noordwijk, p 239

59 Interferometric Space Missions for Exoplanet Science: Legacy of Darwin/TPF

1249

Beichman CA, Woolf NJ, Lindensmith CA (1999) The terrestrial planet finder (TPF): a NASA origins program to search for habitable planets. JPL publication Beichman CA, Fridlund M, Traub WA, Stapelfeldt KR, Quirrenbach A, Seager S (2007) Comparative planetology and the search for life beyond the solar system. Protostars and planets V, pp 915–928. astro-ph/0601469 Beichman C, Benneke B, Knutson H, Smith R, Lagage PO, Dressing C, Latham D, Lunine J, Birkmann S, Ferruit P, Giardino G, Kempton E, Carey S, Krick J, Deroo PD, Mandell A, Ressler ME, Shporer A, Swain M, Vasisht G, Ricker G, Bouwman J, Crossfield I, Greene T, Howell S, Christiansen J, Ciardi D, Clampin M, Greenhouse M, Sozzetti A, Goudfrooij P, Hines D, Keyes T, Lee J, McCullough P, Robberto M, Stansberry J, Valenti J, Rieke M, Rieke G, Fortney J, Bean J, Kreidberg L, Ehrenreich D, Deming D, Albert L, Doyon R, Sing D (2014) Observations of transiting exoplanets with the James Webb Space Telescope (JWST). PASP 126:1134. https:// doi.org/10.1086/679566 Blackwood GH, Serabyn E, Dubovitsky S, Aung M, Gunter SM, Henry C (2003) System design and technology development for the Terrestrial Planet Finder infrared interferometer. In: Coulter DR (ed) Techniques and Instrumentation for Detection of Exoplanets, Proceedings of SPIE, vol 5170, pp 129–143. https://doi.org/10.1117/12.521311 Bolmont E, Libert AS, Leconte J, Selsis F (2016) Habitability of planets on eccentric orbits: limits of the mean flux approximation. A&A 591:A106. https://doi.org/10.1051/0004-6361/ 201628073, 1604.06091 Bracewell RN (1978) Detecting nonsolar planets by spinning infrared interferometer. Nature 274:780. https://doi.org/10.1038/274780a0 Brandl BR, Feldt M, Glasse A, Guedel M, Heikamp S, Kenworthy M, Lenzen R, Meyer MR, Molster F, Paalvast S, Pantin EJ, Quanz SP, Schmalzl E, Stuik R, Venema L, Waelkens C (2014) METIS: the mid-infrared E-ELT imager and spectrograph. In: Ground-Based and Airborne Instrumentation for Astronomy V, Proceedings of SPIE, vol 9147, p 914721. https://doi.org/ 10.1117/12.2056468, 1409.3087 Burrows AS (2014) Highlights in the study of exoplanet atmospheres. Nature 513:345–352. https:// doi.org/10.1038/nature13782, 1409.7320 Cockell CS, Léger A, Fridlund M, Herbst TM, Kaltenegger L, Absil O, Beichman C, Benz W, Blanc M, Brack A, Chelli A, Colangeli L, Cottin H, Coudé du Foresto V, Danchi WC, Defrère D, den Herder JW, Eiroa C, Greaves J, Henning T, Johnston KJ, Jones H, Labadie L, Lammer H, Launhardt R, Lawson P, Lay OP, LeDuigou JM, Liseau R, Malbet F, Martin SR, Mawet D, Mourard D, Moutou C, Mugnier LM, Ollivier M, Paresce F, Quirrenbach A, Rabbia YD, Raven JA, Rottgering HJA, Rouan D, Santos NC, Selsis F, Serabyn E, Shibai H, Tamura M, Thiébaut E, Westall F, White GJ (2009) Darwin-A mission to detect and search for life on extrasolar planets. Astrobiology 9:1–22. https://doi.org/10.1089/ast.2007.0227, 0805.1873 Colavita MM, Serabyn E, Millan-Gabet R, Koresko CD, Akeson RL, Booth AJ, Mennesson BP, Ragland SD, Appleby EC, Berkey BC, Cooper A, Crawford SL, Creech-Eakman MJ, Dahl W, Felizardo C, Garcia-Gathright JI, Gathright JT, Herstein JS, Hovland EE, Hrynevych MA, Ligon ER, Medeiros DW, Moore JD, Morrison D, Paine CG, Palmer DL, Panteleeva T, Smith B, Swain MR, Smythe RF, Summers KR, Tsubota K, Tyau C, Vasisht G, Wetherell E, Wizinowich PL, Woillez JM (2009) Keck interferometer nuller data reduction and on-sky performance. PASP 121:1120–1138. https://doi.org/10.1086/606063 Cowan NB, Voigt A, Abbot DS (2012) Thermal phases of Earth-like planets: estimating thermal inertia from eccentricity, obliquity, and diurnal forcing. ApJ 757:80. https://doi.org/10.1088/ 0004-637X/757/1/80, 1205.5034 Crooke JA, Roberge A, Domagal-Goldman SD, Mandell AM, Bolcar MR, Rioux NM, Perez MR, Smith EC (2016) Status and path forward for the large ultraviolet/optical/infrared surveyor (LUVOIR) mission concept study. In: Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, Proceedings SPIE, vol 9904, p 99044R. https://doi.org/10.1117/12. 2233084 Crossfield IJM, Hansen BMS, Harrington J, Cho JYK, Deming D, Menou K, Seager S (2010) A new 24 m phase curve for + andromedae b. ApJ 723:1436–1446. https://doi.org/10.1088/ 0004-637X/723/2/1436, 1008.0393

1250

D. Defrère et al.

Danchi WC, Barry RK, Lawson PR, Traub WA, Unwin S (2008) The Fourier-Kelvin stellar interferometer (FKSI): a review, progress report, and update. In: Optical and Infrared Interferometry, Proceedings of SPIE, vol 7013, p 70132Q. https://doi.org/10.1117/12.790649 Defrère D (2009) Detection of exozodiacal dust; a step toward Earth-like planet characterization with infrared interferometry. PhD thesis, Liège University, Liège Defrère D, Absil O, Coudé du Foresto V, Danchi WC, den Hartog R (2008a) Nulling interferometry: performance comparison between space and ground-based sites for exozodiacal disc detection. A&A 490:435–445. https://doi.org/10.1051/0004-6361:200810248, 0808.3713 Defrère D, Lay O, den Hartog R, Absil O (2008b) Earth-like planets: science performance predictions for future nulling interferometry missions. In: Optical and Infrared Interferometry, Proceedings of SPIE Defrère D, Absil O, den Hartog R, Hanot C, Stark C (2010) Nulling interferometry: impact of exozodiacal clouds on the performance of future life-finding space missions. A&A 509:A9. https://doi.org/10.1051/0004-6361/200912973, 0910.3486 Defrère D, Hinz PM, Skemer AJ, Kennedy GM, Bailey VP, Hoffmann WF, Mennesson B, MillanGabet R, Danchi WC, Absil O, Arbo P, Beichman C, Brusa G, Bryden G, Downey EC, Durney O, Esposito S, Gaspar A, Grenz P, Haniff C, Hill JM, Lebreton J, Leisenring JM, Males JR, Marion L, McMahon TJ, Montoya M, Morzinski KM, Pinna E, Puglisi A, Rieke G, Roberge A, Serabyn E, Sosa R, Stapeldfeldt K, Su K, Vaitheeswaran V, Vaz A, Weinberger AJ, Wyatt MC (2015) First-light LBT nulling interferometric observations: warm exozodiacal dust resolved within a few AU of  Crv. ApJ 799:42. https://doi.org/10.1088/0004-637X/799/ 1/42, 1501.04144 Defrère D, Hinz PM, Mennesson B, Hoffmann WF, Millan-Gabet R, Skemer AJ, Bailey V, Danchi WC, Downey EC, Durney O, Grenz P, Hill JM, McMahon TJ, Montoya M, Spalding E, Vaz A, Absil O, Arbo P, Bailey H, Brusa G, Bryden G, Esposito S, Gaspar A, Haniff CA, Kennedy GM, Leisenring JM, Marion L, Nowak M, Pinna E, Powell K, Puglisi A, Rieke G, Roberge A, Serabyn E, Sosa R, Stapeldfeldt K, Su K, Weinberger AJ, Wyatt MC (2016) Nulling data reduction and on-sky performance of the large binocular telescope interferometer. ApJ 824:66. https://doi.org/10.3847/0004-637X/824/2/66, 1601.06866 Des Marais DJ, Harwit MO, Jucks KW, Kasting JF, Lin DNC, Lunine JI, Schneider J, Seager S, Traub WA, Woolf NJ (2002) Remote sensing of planetary properties and biosignatures on extrasolar terrestrial planets. Astrobiology 2:153–181. https://doi.org/10.1089/15311070260192246 Eisenhauer F, Perrin G, Brandner W, Straubmeier C, Perraut K, Amorim A, Schöller M, Gillessen S, Kervella P, Benisty M, Araujo-Hauck C, Jocou L, Lima J, Jakob G, Haug M, Clénet Y, Henning T, Eckart A, Berger JP, Garcia P, Abuter R, Kellner S, Paumard T, Hippler S, Fischer S, Moulin T, Villate J, Avila G, Gräter A, Lacour S, Huber A, Wiest M, Nolot A, Carvas P, Dorn R, Pfuhl O, Gendron E, Kendrew S, Yazici S, Anton S, Jung Y, Thiel M, Choquet É, Klein R, Teixeira P, Gitton P, Moch D, Vincent F, Kudryavtseva N, Ströbele S, Sturm S, Fédou P, Lenzen R, Jolley P, Kister C, Lapeyrère V, Naranjo V, Lucuix C, Hofmann R, Chapron F, Neumann U, Mehrgan L, Hans O, Rousset G, Ramos J, Suarez M, Lederer R, Reess JM, Rohloff RR, Haguenauer P, Bartko H, Sevin A, Wagner K, Lizon JL, Rabien S, Collin C, Finger G, Davies R, (2011) GRAVITY: observing the Universe in motion. The Messenger 143: 16–24 Enya K, Kataza H, Bierden P (2009) A micro electrical mechanical systems (MEMS)-based cryogenic deformable mirror. PASP 121:260–265. https://doi.org/10.1086/598171 Errmann R, Minardi S, Labadie L, Muthusubramanian B, Dreisow F, Nolte S, Pertsch T (2015) Interferometric nulling of four channels with integrated optics. Appl Opt 54:7449. https://doi. org/10.1364/AO.54.007449 Fischer PD, Knutson HA, Sing DK, Henry GW, Williamson MW, Fortney JJ, Burrows AS, Kataria T, Nikolov N, Showman AP, Ballester GE, Désert JM, Aigrain S, Deming D, Lecavelier des Etangs A, Vidal-Madjar A (2016) HST Hot-Jupiter transmission spectral survey: clear skies for cool saturn WASP-39b. ApJ 827:19. https://doi.org/10.3847/0004-637X/827/1/19, 1601.04761 Fridlund CVM (2000) Darwin – the infrared space interferometer. In: Schürmann B (ed) Darwin and astronomy: the infrared space interferometer, vol 451. ESA Special Publication, Noordwijk, p 11

59 Interferometric Space Missions for Exoplanet Science: Legacy of Darwin/TPF

1251

Fridlund CVM (2004) The DARWIN project – an ESA cornerstone candidate mission. In: Penny A (ed) Planetary Systems in the Universe, IAU Symposium, vol 202, p 451 Fridlund M, Gondoin P (2003) GENIE – the Darwin demonstrator. Ap&SS 286:93–98. https://doi. org/10.1023/A:1026166313620 Fridlund CVM, d’Arcio L, den Hartog R, Karlsson A (2006) Status and recent progress of the Darwin mission in the cosmic vision program. In: Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, Proceedings SPIE, vol 6268, p 62680R. https://doi.org/ 10.1117/12.671040 Fridlund M, Eiroa C, Henning T, Herbst T, Lammer H, Léger A, Liseau R, Paresce F, Penny A, Quirrenbach A, Röttgering H, Selsis F, White GJ, Absil O, Defrère D, Schneider J, Tinetti G, Karlsson A, Gondoin P, den Hartog R, D’Arcio L, Stankov AM, Kilter M, Erd C, Beichman C, Coulter D, Danchi W, Devirian M, Johnston KJ, Lawson P, Lay OP, Lunine J, Kaltenegger L (2010) The search for worlds like our own. Astrobiology 10:5–17. https://doi.org/10.1089/ast. 2009.0380 Gargaud M, Amils R, Quintanilla JC, Cleaves HJ, Irvine WM, Pinti DL, Viso M (2011) Encyclopedia of astrobiology. https://doi.org/10.1007/978-3-642-11274-4 Gómez-Leal I (2013) Spectrophotometry of the infrared emission of Earth-like planets. Ph.D thesis, University of Bordeaux, France Gondoin PA, Absil O, den Hartog RH, Wilhelm RC, Gitton PB, D’Arcio LL, Fabry P, Puech F, Fridlund MC, Schoeller M, Glindemann A, Bakker EJ, Karlsson AL, Peacock AJ, Volonte S, Paresce F, Richichi A (2004) Darwin-GENIE: a nulling instrument at the VLTI. In: Traub WA (ed) New Frontiers in Stellar Interferometry, Proceedings of SPIE, vol 5491, p 775. https://doi. org/10.1117/12.549411 Grenfell JL, Grießmeier JM, Patzer B, Rauer H, Segura A, Stadelmann A, Stracke B, Titz R, Von Paris P (2007) Biomarker response to galactic cosmic ray-induced NOx and the methane greenhouse effect in the atmosphere of an Earth-like planet orbiting an M dwarf star. Astrobiology 7:208–221. https://doi.org/10.1089/ast.2006.0129, astro-ph/ 0702622 Hanot C, Mennesson B, Martin S, Liewer K, Loya F, Mawet D, Riaud P, Absil O, Serabyn E (2011) Improving interferometric null depth measurements using statistical distributions: theory and first results with the palomar fiber nuller. ApJ 729:110. https://doi.org/10.1088/0004-637X/ 729/2/110, 1103.4719 Hinz P, Bailey VP, Defrère D, Downey E, Esposito S, Hill J, Hoffmann WF, Leisenring J, Montoya M, McMahon T, Puglisi A, Skemer A, Skrutskie M, Vaitheeswaran V, Vaz A (2014) Commissioning the LBTI for use as a nulling interferometer and coherent imager. In: Optical and Infrared Interferometry IV, Proceedings of SPIE, vol 9146, p 91460T. https://doi.org/10. 1117/12.2057340 Hystad G, Downs RT, Grew ES, Hazen RM (2015) Statistical analysis of mineral diversity and distribution: Earth’s mineralogy is unique. Earth Planet Sci Lett 426:154–157. https://doi.org/ 10.1016/j.epsl.2015.06.028 Jovanovic N, Martinache F, Guyon O, Clergeon C, Singh G, Kudo T, Garrel V, Newman K, Doughty D, Lozi J, Males J, Minowa Y, Hayano Y, Takato N, Morino J, Kuhn J, Serabyn E, Norris B, Tuthill P, Schworer G, Stewart P, Close L, Huby E, Perrin G, Lacour S, Gauchet L, Vievard S, Murakami N, Oshiyama F, Baba N, Matsuo T, Nishikawa J, Tamura M, Lai O, Marchis F, Duchene G, Kotani T, Woillez J (2015) The Subaru coronagraphic extreme adaptive optics system: enabling high-contrast imaging on solar-system scales. PASP 127:890. https:// doi.org/10.1086/682989, 1507.00017 Kaltenegger L, Traub WA, Jucks KW (2007) Spectral evolution of an Earth-like planet. ApJ 658:598–616. https://doi.org/10.1086/510996, astro-ph/0609398 Kaltenegger L, Eiroa C, Fridlund CVM (2010) Target star catalogue for Darwin nearby stellar sample for a search for terrestrial planets. Ap&SS 326:233–247. https://doi.org/10.1007/ s10509-009-0223-3, 0810.5138 Karlsson AL, Wallner O, Perdigues Armengol JM, Absil O (2004) Three telescope nuller based on multibeam injection into single-mode waveguide. In: Traub WA (ed) Proceedings of SPIE, vol 5491, pp 831–842

1252

D. Defrère et al.

Kitzmann D, Patzer ABC, von Paris P, Godolt M, Rauer H (2011) Clouds in the atmospheres of extrasolar planets. III. Impact of low and high-level clouds on the reflection spectra of Earth-like planets. A&A 534:A63. https://doi.org/10.1051/0004-6361/201117375, 1108.3274 Kleidon A (2010) Life, hierarchy, and the thermodynamic machinery of planet Earth. Phys Life Rev 7:424–460. https://doi.org/10.1016/j.plrev.2010.10.002 Knutson HA, Charbonneau D, Allen LE, Fortney JJ, Agol E, Cowan NB, Showman AP, Cooper CS, Megeath ST (2007) A map of the day-night contrast of the extrasolar planet HD 189733b. Nature 447:183–186. https://doi.org/10.1038/nature05782, 0705.0993 Krissansen-Totton J, Bergsman DS, Catling DC (2016) On detecting biospheres from chemical thermodynamic disequilibrium in planetary atmospheres. Astrobiology 16:39–67. https://doi. org/10.1089/ast.2015.1327, 1503.08249 Ksendzov A, Lay O, Martin S, Sanghera JS, Busse LE, Kim WH, Pureza PC, Nguyen VQ, Aggarwal ID (2007) Characterization of mid-infrared single mode fibers as modal filters. Appl Opt 46:7957–7962. https://doi.org/10.1364/AO.46.007957 Ksendzov A, Lewi T, Lay OP, Martin SR, Gappinger RO, Lawson PR, Peters RD, Shalem S, Tsun A, Katzir A (2008) Modal filtering for midinfrared nulling interferometry using single mode silver halide fibers. Appl Opt 47:5728. https://doi.org/10.1364/AO.47.005728 Lawson P, Traub W (2006) Earth-Like exoplanets: the science of NASA’s navigator program. JPL Publication, Pasadena Lawson PR, Lay OP, Johnston KJ, Beichman CA (2007) Terrestrial Planet Finder Interferometer Science Working Group Report. NASA STI/Recon Technical Report N 8 Lay OP (2004) Systematic errors in nulling interferometers. Appl Opt 43:6100–6123. https://doi. org/10.1364/AO.43.006100 Lay OP, Martin SR, Hunyadi SL (2007) Planet-finding performance of the TPF-I Emma architecture. In: Techniques and Instrumentation for Detection of Exoplanets III, Proceedings of SPIE, vol 6693, p 66930A. https://doi.org/10.1117/12.732230 Le Bouquin JB, Berger JP, Lazareff B, Zins G, Haguenauer P, Jocou L, Kern P, Millan-Gabet R, Traub W, Absil O, Augereau JC, Benisty M, Blind N, Bonfils X, Bourget P, Delboulbe A, Feautrier P, Germain M, Gitton P, Gillier D, Kiekebusch M, Kluska J, Knudstrup J, Labeye P, Lizon JL, Monin JL, Magnard Y, Malbet F, Maurel D, Ménard F, Micallef M, Michaud L, Montagnier G, Morel S, Moulin T, Perraut K, Popovic D, Rabou P, Rochat S, Rojas C, Roussel F, Roux A, Stadler E, Stefl S, Tatulli E, Ventura N (2011) PIONIER: a 4-telescope visitor instrument at VLTI. A&A 535:A67. https://doi.org/10.1051/0004-6361/201117586, 1109.1918 Lederberg J (1965) Signs of life: criterion-system of exobiology. Nature 207:9–13. https://doi.org/ 10.1038/207009a0 Leger A, Pirre M, Marceau FJ (1993) Search for primitive life on a distant planet: relevance of 02 and 03 detections. A&A 277:309 Léger A, Mariotti JM, Mennesson B, Ollivier M, Puget JL, Rouan D, Schneider J (1996a) Could we search for primitive life on extrasolar planets in the near future? Icarus 123:249–255. https:// doi.org/10.1006/icar.1996.0155 Léger A, Mariotti JM, Mennesson B, Ollivier M, Puget JL, Rouan D, Schneider J (1996b) The DARWIN project. Ap&SS 241:135–146. https://doi.org/10.1007/BF00644221 Léger A, Selsis F, Sotin C, Guillot T, Despois D, Mawet D, Ollivier M, Labèque A, Valette C, Brachet F, Chazelas B, Lammer H (2004) A new family of planets? “Ocean-Planets”. Icarus 169:499–504. https://doi.org/10.1016/j.icarus.2004.01.001, astro-ph/0308324 Léger A, Fontecave M, Labeyrie A, Samuel B, Demangeon O, Valencia D (2011) Is the presence of oxygen on an exoplanet a reliable biosignature? Astrobiology 11:335–341. https://doi.org/ 10.1089/ast.2010.0516 Lovelock JE (1965) A physical basis for life detection experiments. Nature 207:568–570. https:// doi.org/10.1038/207568a0 Martin SR, Booth AJ (2010) Demonstration of exoplanet detection using an infrared telescope array. A&A 520:A96. https://doi.org/10.1051/0004-6361/201014942

59 Interferometric Space Missions for Exoplanet Science: Legacy of Darwin/TPF

1253

Martin S, Booth A, Liewer K, Raouf N, Loya F, Tang H (2012) High performance testbed for fourbeam infrared interferometric nulling and exoplanet detection. Appl Opt 51:3907–3921. https:// doi.org/10.1364/AO.51.003907 Martin S, Serabyn G, Liewer K, Mennesson B (2017) Achromatic broadband nulling using a phase grating. Optica 4(1):110–113. https://doi.org/10.1364/OPTICA.4.000110, http://www. osapublishing.org/optica/abstract.cfm?URI=optica-4-1-110 Maurin AS, Selsis F, Hersant F, Belu A (2012) Thermal phase curves of nontransiting terrestrial exoplanets. II. Characterizing airless planets. A&A 538:A95. https://doi.org/10.1051/00046361/201117054, 1110.3087 Mayor M, Queloz D (1995) A Jupiter-mass companion to a solar-type star. Nature 378:355–359. https://doi.org/10.1038/378355a0 Mennesson B, Mariotti JM (1997) Array configurations for a space infrared nulling interferometer dedicated to the search for Earth-like extrasolar planets. Icarus 128:202–212. https://doi.org/10. 1006/icar.1997.5731 Mennesson B, Ollivier M, Ruilier C (2002) Use of single-mode waveguides to correct the optical defects of a nulling interferometer. J Opt Soc Am A 19:596–602. https://doi.org/10.1364/ JOSAA.19.000596 Mennesson B, Léger A, Ollivier M (2005) Direct detection and characterization of extrasolar planets: the Mariotti space interferometer. Icarus 178:570–588. https://doi.org/10.1016/ j.icarus.2005.05.012 Mennesson B, Hanot C, Serabyn E, Liewer K, Martin SR, Mawet D (2011a) High-contrast stellar observations within the diffraction limit at the palomar hale telescope. ApJ 743:178. https://doi. org/10.1088/0004-637X/743/2/178 Mennesson B, Serabyn E, Hanot C, Martin SR, Liewer K, Mawet D (2011b) New constraints on companions and dust within a few AU of vega. ApJ 736:14. https://doi.org/10.1088/0004-637X/ 736/1/14 Mennesson B, Millan-Gabet R, Serabyn E, Colavita MM, Absil O, Bryden G, Wyatt M, Danchi W, Defrère D, Doré O, Hinz P, Kuchner M, Ragland S, Scott N, Stapelfeldt K, Traub W, Woillez J (2014) Constraining the exozodiacal luminosity function of main-sequence stars: complete results from the Keck nuller mid-infrared surveys. ApJ 797:119. https://doi.org/10. 1088/0004-637X/797/2/119 Mennesson B, Gaudi S, Seager S, Cahoy K, Domagal-Goldman S, Feinberg L, Guyon O, Kasdin J, Marois C, Mawet D, Tamura M, Mouillet D, Prusti T, Quirrenbach A, Robinson T, Rogers L, Scowen P, Somerville R, Stapelfeldt K, Stern D, Still M, Turnbull M, Booth J, Kiessling A, Kuan G, Warfield K (2016) The Habitable Exoplanet (HabEx) Imaging Mission: preliminary science drivers and technical requirements. In: Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, Proceedings SPIE, vol 9904, p 99040L. https://doi.org/10.1117/12. 2240457 Monnier JD, Ireland MJ, Kraus S, Baron F, Creech-Eakman M, Dong R, Isella A, Merand A, Michael E, Minardi S, Mozurkewich D, Petrov R, Rinehart S, ten Brummelaar T, Vasisht G, Wishnow E, Young J, Zhu Z (2016) Architecture design study and technology road map for the Planet Formation Imager (PFI). In: Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, Proceedings SPIE, vol 9907, p 99071O. https://doi.org/10.1117/12.2233311, 1608.00580 Moskovitz NA, Gaidos E, Williams DM (2009) The effect of lunarlike satellites on the orbital infrared light curves of Earth-analog planets. Astrobiology 9:269–277. https://doi.org/10.1089/ ast.2007.0209, 0810.2069 Ollivier M, Absil O, Allard F, Berger JP, Bordé P, Cassaing F, Chazelas B, Chelli A, Chesneau O, Coudé du Foresto V, Defrère D, Duchon P, Gabor P, Gay J, Herwats E, Jacquinod S, Kern P, Kervella P, Le Duigou JM, Léger A, Lopez B, Malbet F, Mourard D, Pelat D, Perrin G, Rabbia Y, Rouan D, Reiss JM, Rousset G, Selsis F, Stee P, Surdej J (2009) PEGASE, an infrared interferometer to study stellar environments and low mass companions around nearby stars. Exp Astron 23:403–434. https://doi.org/10.1007/s10686-008-9133-6

1254

D. Defrère et al.

Owen T (1980) The search for early forms of life in other planetary systems – future possibilities afforded by spectroscopic techniques. In: Papagiannis MD (ed) Strategies for the search for life in the Universe. Astrophysics and space science library, vol 83, p 177. https://doi.org/10.1007/ 978-94-009-9115-6-17 Peters RD, Lay OP, Lawson PR (2010) Mid-infrared adaptive nulling for the detection of Earth-like exoplanets. PASP 122:85–92. https://doi.org/10.1086/649850 Quanz SP, Crossfield I, Meyer MR, Schmalzl E, Held J (2015) Direct detection of exoplanets in the 3–10 m range with E-ELT/METIS. Int J Astrobiol 14:279–289. https://doi.org/10.1017/ S1473550414000135, 1404.0831 Rauer H, Gebauer S, Paris PV, Cabrera J, Godolt M, Grenfell JL, Belu A, Selsis F, Hedelt P, Schreier F, (2011) Potential biosignatures in super-Earth atmospheres. I. Spectral appearance of super-Earths around M dwarfs. A&A 529:A8. https://doi.org/10.1051/0004-6361/201014368 Rugheimer S, Kaltenegger L, Zsom A, Segura A, Sasselov D (2013) Spectral fingerprints of Earth-like planets around FGK stars. Astrobiology 13:251–269. https://doi.org/10.1089/ast. 2012.0888, 1212.2638 Sagan C, Thompson WR, Carlson R, Gurnett D, Hord C (1993) A search for life on Earth from the Galileo spacecraft. Nature 365:715–721. https://doi.org/10.1038/365715a0 Seager S, Bains W (2015) The search for signs of life on exoplanets at the interface of chemistry and planetary science. Sci Adv 1:e1500,047. https://doi.org/10.1126/sciadv.1500047 Seager S, Turner EL, Schafer J, Ford EB (2005) Vegetation’s red edge: a possible spectroscopic biosignature of extraterrestrial plants. Astrobiology 5:372–390. https://doi.org/10.1089/ast. 2005.5.372, astro-ph/0503302 Segura A, Krelove K, Kasting JF, Sommerlatt D, Meadows V, Crisp D, Cohen M, Mlawer E (2003) Ozone concentrations and ultraviolet fluxes on Earth-Like planets around other stars. Astrobiology 3:689–708. https://doi.org/10.1089/153110703322736024 Segura A, Kasting JF, Meadows V, Cohen M, Scalo J, Crisp D, Butler RAH, Tinetti G (2005) Biosignatures from Earth-like planets around M dwarfs. Astrobiology 5:706–725. https://doi. org/10.1089/ast.2005.5.706, astro-ph/0510224 Selsis F (2004) The atmosphere of terrestrial exoplanets: detection and characterization. In: Beaulieu J, Lecavelier Des Etangs A, Terquem C (eds) Extrasolar planets: today and tomorrow. Astronomical Society of the Pacific conference series, vol 321, p 170. http://adsabs.harvard.edu/ abs/2004ASPC..321..170S Selsis F, Despois D, Parisot JP (2002) Signature of life on exoplanets: can Darwin produce false positive detections? A&A 388:985–1003. https://doi.org/10.1051/0004-6361:20020527 Selsis F, Wordsworth RD, Forget F (2011) Thermal phase curves of nontransiting terrestrial exoplanets. I. Characterizing atmospheres. A&A 532:A1. https://doi.org/10.1051/0004-6361/ 201116654, 1104.4763 Selsis F, Maurin AS, Hersant F, Leconte J, Bolmont E, Raymond SN, Delbo’ M (2013) The effect of rotation and tidal heating on the thermal lightcurves of super Mercuries. A&A 555:A51. https://doi.org/10.1051/0004-6361/201321661, 1305.3858 Shao M, Unwin SC, Beichman C, Catanzarite J, Edberg SJ, Marr JC IV, Marcy G (2007) Finding Earth clones with SIM: the most promising near-term technique to detect, find masses for, and determine three-dimensional orbits of nearby habitable planets. In: Techniques and Instrumentation for Detection of Exoplanets III, Proceedings SPIE, vol 6693, p 66930C. https:// doi.org/10.1117/12.734671, 0704.0952 Simoncini E, Virgo N, Kleidon A (2013) Quantifying drivers of chemical disequilibrium: theory and application to methane in the Earth’s atmosphere. Earth Syst Dynam 4:317–331. https://doi. org/10.5194/esd-4-317-2013 Snellen IAG, Brandl BR, de Kok RJ, Brogi M, Birkby J, Schwarz H (2014) Fast spin of the young extrasolar planet ˇ Pictoris b. Nature 509:63–65. https://doi.org/10.1038/nature13253 Tian F, France K, Linsky JL, Mauas PJD, Vieytes MC (2014) High stellar FUV/NUV ratio and oxygen contents in the atmospheres of potentially habitable planets. Earth Planet Sci Lett 385:22–27. https://doi.org/10.1016/j.epsl.2013.10.024

59 Interferometric Space Missions for Exoplanet Science: Legacy of Darwin/TPF

1255

Traub WA, Kaltenegger L, Jucks KW, Turnbull MC (2006) Direct imaging of Earth-like planets from space (TPF-C). In: Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, Proceedings SPIE, vol 6265, p 626502. https://doi.org/10.1117/12.676257 Velusamy T, Angel RP, Eatchel A, Tenerelli D, Woolf NJ (2003) Single and double Bracewell nulling interferometer in space. In: Fridlund M, Henning T, Lacoste H (eds) Earths: DARWIN/TPF and the search for extrasolar terrestrial planets, vol 539. ESA Special Publication, Noordwijk, pp 631–636 von Paris P, Cabrera J, Godolt M, Grenfell JL, Hedelt P, Rauer H, Schreier F, Stracke B (2011) Spectroscopic characterization of the atmospheres of potentially habitable planets: GL 581 d as a model case study. A&A 534:A26. https://doi.org/10.1051/0004-6361/201117091, 1108.3670 von Paris P, Hedelt P, Selsis F, Schreier F, Trautmann T (2013) Characterization of potentially habitable planets: retrieval of atmospheric and planetary properties from emission spectra. A&A 551:A120. https://doi.org/10.1051/0004-6361/201220009, 1301.0217 Wallner O, Leeb WR, Winzer PJ (2002) Minimum length of a single-mode fiber spatial filter. J Opt Soc Am A 19:2445–2448. https://doi.org/10.1364/JOSAA.19.002445 Weber V, Barillot M, Haguenauer P, Kern PY, Schanen-Duport I, Labeye PR, Pujol L, Sodnik Z (2004) Nulling interferometer based on an integrated optics combiner. In: Traub WA (ed) New Frontiers in Stellar Interferometry, Proceedings of SPIE, vol 5491, p 842 Woolf N, Angel JR (1998) Astronomical searches for Earth-like planets and signs of life. ARA&A 36:507–538. https://doi.org/10.1146/annurev.astro.36.1.507

60

CHEOPS: CHaracterizing ExOPlanets Satellite Willy Benz, David Ehrenreich, and Kate Isaak

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Science . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Science Objectives of CHEOPS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Science Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The CHEOPS Mission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Payload . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Platform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The CHEOPS Ground Segment, Launch, Orbit, and Operations . . . . . . . . . . . . . . . . . . . . Observing with CHEOPS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Current Status and Next Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1258 1259 1259 1261 1266 1269 1269 1269 1272 1274 1275 1278 1279

Abstract

The Characterising Exoplanet Satellite (CHEOPS) was selected on October 19, 2012, as the first small mission (S-mission) in the ESA Science Programme. CHEOPS will be the first mission dedicated to the search for transits of exoplanets by means of ultrahigh precision photometry on bright stars already

W. Benz () Physikalisches Institut, Universität Bern, Bern, Switzerland e-mail: [email protected] D. Ehrenreich Astronomical Observatory of the University of Geneva, Sauverny, Switzerland e-mail: [email protected] K. Isaak Science Support Office, European Space Agency - ESTEC, AZ, Noordwijk, The Netherlands e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_84

1257

1258

W. Benz et al.

known to host planets. It will provide the unique capability of determining accurate radii for a subset of those planets for which the mass has already been estimated from ground-based spectroscopic surveys. It will also provide precise radii for new planets discovered by the next generation of ground- or space-based transit surveys (Neptune-size and smaller). By unveiling transiting exoplanets with high potential for in-depth characterization, CHEOPS will also provide prime targets for future instruments suited to the spectroscopic characterization of exoplanetary atmospheres. To reach these goals, CHEOPS will measure photometric signals with a precision of 20 ppm over 6 h for a magnitude 9 star in the V band. This precision will be achieved by using a single-frame transfer, back-side illuminated CCD detector located in the focal plane assembly of a 33-cm diameter onaxis telescope. The optical design is based on a Ritchey-Chrétien telescope that produces a defocused image of the target star while minimizing the stray light contamination with a dedicated field stop and a baffling system. The spacecraft (280 kg) will be launched as a secondary passenger on a Soyuz rocket from Kourou into a 700 km altitude Sun-synchronous orbit and will have a nominal operational lifetime of 3.5 years.

Introduction The Characterising Exoplanet Satellite (CHEOPS) mission is the first small, socalled S-class, mission in the Science Programme of the European Space Agency (ESA). Implemented in partnership with Switzerland, and with important contributions from a Mission Consortium comprising Austria, Belgium, France, Germany, Hungary, Italy, Portugal, Spain, Sweden, and the United Kingdom, the mission is dedicated to the search for transits of bright stars by known exoplanets with masses between Earth and Neptune. CHEOPS will utilize the technique of ultrahigh precision photometry to provide the unique capability of determining accurate radii for a subset of planets for which the mass has already been estimated from ground-based spectroscopic surveys. By combining measurements of radius with existing mass determinations, CHEOPS will provide a first-step characterization of exoplanets. It will also provide precise radii for new planets discovered in the next generation of ground- or space-based transit surveys (Neptune-size). By identifying exoplanets with a high potential for in-depth characterization, CHEOPS will also provide prime targets for future instruments suited to the spectroscopic characterization of exoplanetary atmospheres. CHEOPS was selected for study by the Science Programme Committee (SPC) in October 2012 and subsequently adopted in February 2014. S-class missions are much smaller in scope and scale than ESA medium (M) – (e.g., Euclid) or large (L) – class (e.g., Athena) missions and are subject to the following programmatic constraints: a development time not exceeding 4 years, a high level of technology readiness at proposal selection (ESA Technology Readiness Level > 5 (ESA TRL), a total cost to the ESA Science Programme limited to 50 MA C (at 2012 economic

60 CHEOPS: CHaracterizing ExOPlanets Satellite

1259

Table 1 Division of responsibilities in the CHEOPS mission Role Mission architect Platform procurement S/C assembly, integration, and test Launch procurement LEOP and IOC Instrument

Responsible party ESA ESA ESA ESA ESA CMC

Science team

CMC

Mission operations

CMC

Science operations

CMC

Remarks Overall mission definition Procured via the space craft (S/C) contractor Via S/C contractor, supported by AIT service provider Shared launch opportunity Executed by the S/C contractor Delivered as CFI to S/C contractor (CCD procured by ESA) Chaired by consortium member, ESA attending by invitation Implemented via Mission Operations Centre (CDTI, ES) Implemented via Science Operations Centre (UniGe, CH)

conditions), and a total cost of less than 150 MA C (approximately, ESA C member states). To meet the programmatic constraints, the division of responsibilities between ESA and the CHEOPS Mission Consortium (CMC) is also very different from a typical ESA mission and is summarized in Table 1. In the following sections, we present the key elements of the CHEOPS mission. Starting with the overview of the science goals and the science requirements of the mission, we continue with the presentation of the spacecraft itself discussing separately the payload and the platform. Finally, we present the ground segment and the operation concept that, owing to the special nature of the CHEOPS project, is unique among ESA science missions.

Science Background Most known exoplanets are found transiting their host star. The special geometry of these systems makes them particularly interesting, since for these planets the orbital inclination as well as the radius can be derived. In practice, a thorough analysis of the light curve and follow-up observations are needed to understand the nature of the transiting object. Usually, these follow-up observations consist of spectroscopic observations and precise radial velocity measurements (Doppler follow-up). The improvements and intensive efforts made during the last decade by teams carrying out Doppler surveys have led to the identification of numerous planetary systems. Among those, sub-Neptune-mass planets and super-Earths are receiving particular

1260

W. Benz et al.

attention as they have no counterpart in our solar system, and thus their exact nature is still unknown. These discoveries clearly indicate that low-mass planets orbiting solar-type stars must be very common (Mayor et al. 2011; Butler et al. 2017; Udry et al. 2017). Analyses of the Kepler mission results have similarly revealed that sub-Neptunesize planets are overwhelmingly more abundant than giants (Howard et al. 2012; Fulton et al. 2017). Furthermore, Kepler uncovered a large population of “packed” planetary systems (Borucki et al. 2011; Lissauer et al. 2011) with about 17% of the stars hosting multi-planet systems. While their existence has been demonstrated, the exact nature of super-Earth planets remains a matter of fierce debate. From the handful of small planets with both mass and size estimates, characterizing their structure remains challenging. On the ground, transit surveys have been successful in detecting planets among which WASP (Collier Cameron et al. 2007; Triaud et al. 2017) and HAT (Bakos et al. 2007; Hartman et al. 2015) stand out as the most prolific, with on average at least one planet discovered every 3 weeks. Targeted surveys, pointing at one star at a time, such as MEarth and TRAPPIST, have lower yields than the field surveys quoted above but critically detect benchmark systems such as GJ 1214 (Charbonneau et al. 2009) or TRAPPIST-1 (Gillon et al. 2017). In addition, new surveys, such as NGTS (Wheatley et al. 2013) or MASCARA (Talens et al. 2017), have started recently. The main advantage of ground-based surveys is their ability to search the whole sky and hence to find planets orbiting bright stars. Current surveys have, within their design limitations, mapped almost the whole sky and found hundreds of short-period giant planets. The first space transit mission, CoRoT, was successfully launched in December 2006 (see  Chap. 55, “CoRoT: The First Space-Based Transit Survey to Explore the Close-in Planet Population”). It was a pioneer in its use of ultra-precision photometry with high sampling rate and long duration observations. The satellite primarily observed two fields of view, each of 4 square degrees – once per year. Each field comprised 5–6 thousand dwarf stars with magnitudes ranging from roughly 11 to 16. After 5 years of operations, CoRoT has discovered dozens of exoplanets (CoRoT Team 2016), among them the first super-Earth planet at very short orbital period, identified as having a rocky core: CoRoT-7-b (Léger et al. 2009). Launched 2 years after CoRoT, the Kepler mission (see  Chap. 56, “Space Missions for Exoplanet Science: Kepler/K2”) has turned out to be a landmark in transiting planet searches. Kepler measured continuously, and for 4 years, brightness variations of about 100,000 solar-like stars to a precision of order 10 ppm in a single field of view of approximately 100-square degrees. Kepler found many thousands of transiting planetary candidates, some of them with radii as small as the Earth radius (R˚ ) and many multiple transiting systems. With these discoveries, Kepler has provided the community not only with several benchmark systems – circumbinary planets, compact systems of sub-Earth-size planets, or systems exhibiting transit timing variations (e.g., Lissauer et al. 2014) – but also with a large uniform database of potential planetary systems enabling the derivation of distribution functions for planetary orbits, radii, and hierarchical structure of systems (Borucki et al. 2011).

60 CHEOPS: CHaracterizing ExOPlanets Satellite

1261

CoRoT and Kepler have successfully pioneered the field of space-borne transit photometry. Together, they have demonstrated the diversity of the exoplanet population and provided important statistics helping constraining planet formation models. Yet, in terms of exoplanet characterization, much remains to be done. As both missions aimed at discovering new planets, they stared at a single or a very small number of restricted fields in the sky. While this approach allowed them to monitor a large number (of order 105 ) of stars and thus to discover many planets, the drawback is that, on average, the host stars are typically between V  11 and 16 magnitude. Measuring sufficiently precise stellar properties or radial velocities for stars this faint, to obtain a reliable detection and mass determination for small planets, is extremely difficult. The example of CoRoT-7 shows that with the HARPS spectrograph, it is possible to measure the mass of small planets (5 M˚ , 2 R˚ ) in the super-Earth domain (Queloz et al. 2009; Hatzes et al. 2011), located on short-period orbit (< 1 day) for stars brighter than V  11 magnitude. Similar measurements for planets with longer orbital periods on even fainter stars typical of Kepler candidates require a prohibitive amount of telescope time. In total Kepler has found 19 small (2–4 R˚ which have no analog in our own solar system (Dong and Zhu 2013), and the ubiquity of 1–4 R˚ planets orbiting cool M stars (Dressing and Charbonneau 2015). As will be discussed below the recent discovery of 7 planets transiting the very cool M8 star, Trappist-1 (Gillon et al. 2017) highlights the opportunities for studying even Earth-sized planets in the coming decade. Each of these classes will provide valuable targets for detailed characterization with JWST. The exoplanet community has made good first steps in characterizing the physical and chemical properties of a small number of planets with three different techniques: combining transit measurements (which yield the planet’s radius) with precision radial velocity (PRV) measurements (which determine its mass) yields the planet’s bulk density; measuring the brightness and/or spectrum of a planet during transit, secondary eclipse or through a full orbit yields detailed atmospheric information; making filter photometric or spectroscopic observations of directly imaged planets leads to detailed atmospheric information. Among the key results from these efforts are: (a) the distribution of planet densities, from gas-dominated systems with radii (masses) >1.5 R˚ (>5 M˚ ) and rocky planets with smaller radii and masses (Fulton et al. 2017); (b) the presence of a variety of atomic and molecular species in the atmospheres of transiting and directly imaged planets; (c) the vertical and horizontal thermal structure of exoplanet atmospheres including the presence of supersonic winds and planetary rotation. But many important questions remain including: • How do the bulk and atmospheric composition vary as a function of key exoplanet characteristics, such as mass, radius, level of insolation and location within the planetary system? • What can we learn about vertical structure of exoplanet atmospheres and global atmospheric motions? • What can we learn about the formation of exoplanets from, for example, differences in their atmospheric C/O ratio or overall metallicity compared to that of their host stars?

1286

C. A. Beichman and T. P. Greene

• Do massive planets found on distant orbits (>50 AU) have a different formation mechanism compared to those on closer orbits, i.e., disk fragmentation vs. core accretion? • Can we distinguish between low vs. high initial entropy states for forming massive planets? • What are the atmospheric constituents of mini-Neptunes, super-Earths, and even terrestrial planets? • How does the presence of planets affect the structure of debris disks? • Is the presence of a hot or luminous debris disk a signpost of dynamical activity such as a period of “Late Heavy Bombardment”? • What can the composition of debris disk material tell us about the formation of planets? • Do young planetary systems with known massive planets also contain lower mass planets? JWST is poised to make dramatic advances in all of these areas – and more. In the sections that follow we discuss the various methods of planets searches, and more importantly, planet characterization which JWST will bring to the exoplanet community. The ideas described here are drawn primarily from projects being proposed by the instrument teams for the first JWST cycles. The broader community will come up with many new projects, some of these before launch, but the most important and most innovative uses of JWST will doubtless come after launch as we learn about the telescope and the instruments and more importantly as we learn more about exoplanets themselves.

The JWST Instrument Suite JWST has four science instruments, and all have modes that will be useful for observing exoplanets: • NIRCam (Rieke et al. 2005) provides imaging in broad, medium, and narrowband filters from 0.6–5.0 mm as well as grism spectroscopy from 2.4–5. mm. In addition, NIRCam has coronagraphic capabilities with a series of Lyot masks, including 3 circular and 2 wedge-shaped occulters (Fig. 2; Krist et al. 2010). • NIRspec (Ferruit et al. 2012) provides single,multi-object, and Integral Field (IFS) spectroscopy from 0.7 to 5 m at spectral resolutions ranging from 100 to 3,000. • MIRI (Rieke et al. 2015) provides direct imaging with a variety of filters, slit and IFS spectroscopy at low and medium resolution, and coronagraphy in 4 fixed filters at wavelengths from 5.6 to 28.8 m (Fig. 2 Boccaletti et al. 2015). • NIRISS (Doyon et al. 2012) provides imaging in a variety of filters out to 5 m, wide-field and single-object grism spectroscopy at wavelengths from 0.7 to 2.5 m, and aperture masking interferometric imaging at 3.8 and 4.8 m.

HWHM = 0.40” (6λ/D @ 2.1 μm)

60 mm

HWHM = 0.64” HWHM = 0.82” (6λ/D @ 3.35 μm) (6λ/D @ 4.3 μm)

2.5’’ x 2.5’’ ND Square

5” x 5” ND Square (OD = 3)

6λ/D

HWHMc = 0.58” HWHMc = 0.27” (4λ/D @ 4.6 μm) (4λ/D @ 2.1 μm)

2λ/D

20 arcsec 24 x 24 arcsec 4QPMs

4.7 x 0.51arcsec LRS slit

10.65μm

11.4μm

15.5μm

23μm

14μm 74 arcsec

LRS spectrum

5μm

Image FOV

113 arcsec

Fig. 2 (Left) NIRCam has coronagraphic capabilities with a series of Lyot masks, including 3 circular and 2 wedge-shaped occulters optimized for different wavelengths; (right) the MIRI focal plane has three Four Quadrant Phase Masks at 10.6, 11.4, 15.5 m and a Lyot coronagraph at 23 m. (Image courtesy of NASA)

12 mm

30 x 30 arcsec Lyot Mask

61 Observing Exoplanets with the James Webb Space Telescope 1287

1288

C. A. Beichman and T. P. Greene

Coronagraphy and Direct Spectroscopy of Young Planets Planet Characterization JWST offers unprecedented sensitivity for the characterization of exoplanets via direct imaging and spectroscopy. While JWST does not offer the smallest Wave Front Error (WFE) or the most aggressive Inner Working Angle (IWA) compared with Extreme Adaptive Optics coronagraphs on large ground-based telescopes, its stability, low temperature and low background in space make JWST an extremely sensitive observatory for exoplanet observations longward of 3 m where young planets are bright. JWST is not, however, optimized for carrying out large scale coronagraphic surveys for exoplanets, given the high overheads associated with observations of single targets. Thus, much of JWST’s coronagraphic imaging will be focused on characterizing planets within 200 of their host stars in previously known systems. NIRSpec’s IFS and MIRI’s spectrometer will be used to obtain complete spectra of more widely separated planets. JWST’s great strength for direct imaging of planets comes from the stability of its Point Spread Function over a time scale of many hours and its operation in the low background space environment. The Inner Working Angles (IWA) of the NIRCam coronagraphs are relatively coarse, 4/D or 6/D for the wedge and circular masks, respectively, where D is JWST’s 6.5 m aperture. Thus NIRCam’s best IWAs are 0:400 –0:600 at 3.0 and 4:4 m. The effective IWA or the MIRI Four Quadrant Phase Masks (FQPM) is roughly 2/D, or 0:700 at 11:4 m. Contrast limits at 100 are predicted to approach 105 for NIRCam and 104 for MIRI (Fig. 3;

JWST Contrast Ratios (5 sigma)

Log(Contrast Ratio)

−3

−4 NRM

MIRI 11.4

−5 NRM

NIRCam 4.6

−6

−7

0

1

2

3

4

Theta (arcsec) Fig. 3 Contrast ratio cures for JWST’s three coronagraphs: MIRI at 11.4 m (green), NIRCam at 4.6 m (black), and the NIRISS Non-redundant Aperture Mask at 4.8 m (the blue dotted lines reflect two assumptions about its sensitivity) (Beichman et al. 2010)

61 Observing Exoplanets with the James Webb Space Telescope

1289

Beichman et al. 2010). The on-orbit performance of the coronagraphs will depend sensitively on the drift in telescope wavefront error (WFE) over the duration of a combined target and reference star sequence as well as on the ability of the telescope to hold the star carefully centered on the coronagraphic masks. The NIRISS Non-Redundant Mask (NRM) or Aperture Masking Interferometer (AMI) operates as an interferometer with an IWA set by the longest baseline across JWST’s aperture, roughly 0:5/D or 0.0700 at 4.8 m, out to an outer working angle defined by the size of an individual aperture in the mask, or about 0:400 (Artigau et al. 2014). The AMI does not block the light of the central star so the sensitivity is affected by the full photon noise of the stellar signal. The AMI mode can operate at 4 wavelengths: 2.8, 3.8, 4.3 and 4.8 m. The coronagraphs will be used to characterize previously imaged planets across a suite of NIRCam and MIRI filters. From this information it will be possible to infer such basic properties as total luminosity, effective temperature, and thus effective radius. From this information it will be possible to estimate the initial entropy of the planet which is a direct indicator of its formation mechanism via a “hot” or “cold” start (Fig. 5; Marley et al. 2007; Spiegel and Burrows 2012). JWST’s sensitivity is such that it can measure the known young planets at separations of  100 in a suite of filters in less than an hour of integration time. Figure 4 shows the signal-to-

HR8799

500 e

d

c

51 Eri 2MASS1207 HD 95086

HD8799b FW Tau GJ504

100 50

spot_430_f444w

SNR

spot_335_f356w spot_430_f430m spot_335_f360m

10 5

1

0

1

2 3 Radius (arcsec)

4

5

Fig. 4 The signal-to-noise ratio of a 5 MJup planet orbiting HR8799 in 600 s is shown as a function of separation from the star at two medium band and two wide band filters using the circular occulters. Planets ‘b’ and ‘c’ are easily detected while planets ‘d’ and ‘e’ present more of a challenge. Across the top of the figure are markers showing the angular separations of various well known planets. Close to the star the SNR is dominated by residual stellar photons and speckle subtraction while at larger separations (>200 ) the noise for a relatively bright, massive planet is dominated by shot noise from the planet itself

1290

C. A. Beichman and T. P. Greene

Table 1 Illustrative JWST targets for direct imaging and spectroscopy Object LkCa15 51 Eri b HR8799 bcde HD95086 b 2MASS 1207-3932 b & Eri ˇ Leo Vega 2MASS J2236C4751 HD106906 b Fomalhaut b GU Psc b WD 0806-66 b

Mass MJup 0.5? 2 5–10 5 4 >0.1 >0.3 >1 xx 11 2? 11 8

Age (Myr) 2 20 50 15 8 400 45 450 xx 13 15 100 2000

Separation (00 ) 0.1 0.5 0.6–1.7 0.6 0.9 1 >2 >2 3.7 7.1 15 42 130

Instrument NIRISS NIRCam, MIRI NIRCam, MIRI, NIRSpec NIRCam, MIRI NIRCam, NIRSpec,MIRI NIRCam NIRCam NIRCam NIRSpec, MIRI NIRCam, MIRI, NIRSpec NIRCam, MIRI NIRSpec, MIRI NIRCam, MIRI, NIRSpec

Commenta Char,Search Char, Search Char, Search, Spectra Char, Search Char, Search, Spectra Search Search Search Spectra Char, Search Char, Search Spectra Char, Search, Spectra

a “Char” denotes “Characterization” with medium and/or broad band filters; “Search” implies a deep search for planets with masses as low as 1 Saturn-mass;“Spectra” denotes high resolution spectroscopy obtained with either NIRSpec and/or MIRI

noise on a 50 MYr old 5 MJup planet orbiting HR 8799 in a variety of medium and broad band filters. NIRCam’s medium passband filters, F300M, F335M, F360M, F410M, F430M, and F460M will be used to characterize exoplanet atmospheres, measuring both the continuum and searching for signatures of CH4 , CO and CO2 . MIRI’s three FQPM filters will isolate a band of NH3 expected in cool (1 Gyr) all but the most massive planets (or brown dwarfs) will be either too cold or too close to their host star to be detected with NIRCam. JWST will explore the morphology and composition of disks via filter photometry with the NIRCam coronagraph looking for evidence of ices and tholins which characterize Kuiper Belt objects in our own solar system. The MIRI and NIRSpec spectrometers will explore spatially unresolved disks including many of the disk systems with 22 m emission discovered by the WISE satellite (Patel et al. 2014). The compositional information may yield surprises such as the silica grains found toward  Corvi which were interpreted to be due the remnants of a very energetic collisional event, perhaps the impact of a large Kuiper Belt object onto a terrestrialsized planet (Lisse et al. 2012; Su 2015).

Brown Dwarfs Brown Dwarfs (BDs) are the more massive cousins of Jovian-mass planets, starting at the Deuterium burning limit of 13 MJup and extending upward to the Hydrogen burning limit of 70 MJup . The first examples, the L dwarfs with effective temperatures 2000 K were discovered in Sloan (SDSS) and 2MASS sky surveys of the 1990s (Kirkpatrick et al. 1999) and as bound companions to nearby stars, e.g GJ1229B (Nakajima et al. 1995). Subsequently, the WISE survey pushed the detection limits to cooler temperatures, the T and Y classes with temperatures below 1000 K (Kirkpatrick 2013). The space motions of BDs are similar to those of late type M dwarfs (Faherty et al. 2012), so it is likely that these objects represent the lowest mass end of the initial mass function, slowing falling in number as the efficiency of star formation decreases to lower mass (Kirkpatrick et al. 2011). Because the temperature and luminosity of BDs slowly decline with age (Burrows et al. 1989) determining the mass of a BDs of unknown age is a challenge. Some BDs, found in young clusters, are quite hot, i.e., appearing as L dwarf or even a late M dwarf, but not very massive, such as an 8 MJup object identified in Chamaeleon (Luhman et al. 2005). These young objects can be identified through their association with clusters and through spectroscopic features such as low surface gravity (Gagné et al. 2014). JWST imaging surveys with NIRCam of young clusters could identify many more such objects. Among the most dramatic discoveries of the WISE mission was the detection of the coldest ( 25 days i? for these stars. They find that this method fails for rotation periods of P  because the rotational line broadening is too small to be measured reliably. Among

1716

P. F. L. Maxted

the shorter period systems, they find most stars are consistent with i?  90ı if some allowance is made for systematic errors in the estimates of Vrot sin i? , as is expected if most transiting planets have orbits with low obliquity. One exception is Kepler-9, a star that shows transits from at least three planetary companions for which i?  45ı . Hirano et al. (2014) found similar results from their sample of 32 KOIs for which they measured Vrot sin i? and identified three other multi-planet systems that may have significant nonzero obliquity. These systems are useful for studying the dynamical effects that can occur during planet formation (Lai 2014; Hansen 2017). The large number of stars that have planets that are not detected makes it difficult to interpret any observed difference in the rotation periods between KOIs and nonKOIs in the Kepler sample, particularly given that there is a detection bias against planets that transit rapidly rotating solar-type stars because of additional noise from stellar activity.

Wide-Area Surveys and Hot Jupiters KOIs are typically quite faint, with most having apparent magnitudes in the Kepler bandpass Kp D 12–16, so most planet host stars bright enough for detailed characterization come from wide-area surveys such as WASP (Pollacco et al. 2006) and HAT (Bakos et al. 2004). The quality of the photometry achievable by these > 0:5 R ground-based surveys restricts them to discovering planets with radii R Jup < with orbital periods less P 10 days, i.e., hot Jupiter systems. We also discuss in this subsection results for other bright hot Jupiter systems such as those discovered by the CoRoT mission. Pont (2009) found tentative evidence that the host stars of some hot Jupiters rotate faster than typical stars of the same spectral type, particularly for systems where the tidal interaction between the star and the planet is large. This study was based on a sample of about 25 transiting systems with published Vrot sin i? or Prot estimates available at that time. Maxted et al. (2015) used Bayesian techniques to compare the age estimates from stellar model isochrones to gyrochronological age estimates for 28 transiting exoplanet host stars with accurate mass and radius estimates and directly measured rotation periods. The gyrochronological age estimate was significantly lower than the isochronal age estimate for about half of the stars in that sample. They found no clear correlation between the gyrochronological age estimate and the strength of the tidal force on the star due to the innermost planet, leading them to conclude that tidal spin-up is a reasonable explanation for this discrepancy in some cases, but not all. For example, they could find no satisfactory explanation for the discrepancy between the young age for COROT-2 estimated from gyrochronology and supported by its high lithium abundance and the extremely old age for its K-type stellar companion inferred from its lack of magnetic activity. Maxted et al. suggest that this may point to problems with the stellar models for some planet host stars.

83 Rotation of Planet-Hosting Stars

1717

Cross-References  Ages for Exoplanet Host Stars  Characterizing Host Stars using Asteroseismology  Exoplanet Atmosphere Measurements from Transmission Spectroscopy and

Other Planet Star Combined Light Observations  Radial Velocities as an Exoplanet Discovery Method  Space Missions for Exoplanet Science: Kepler/K2  The Impact of Stellar Activity on the Detection and Characterization of Exoplan-

ets  The Rossiter-McLaughlin Effect in Exoplanet Research  Tidal Star-Planet Interactions: A Stellar and Planetary Perspective  Transit Photometry as an Exoplanet Discovery Method

References Albrecht S, Winn JN, Johnson JA et al (2012) Obliquities of hot Jupiter host stars: evidence for tidal interactions and primordial misalignments. ApJ 757:18. https://doi.org/10.1088/0004637X/757/1/18, 1206.6105 Alves S, Do Nascimento JD Jr, de Medeiros JR (2010) On the rotational behaviour of parent stars of extrasolar planets. MNRAS 408:1770–1777. https://doi.org/10.1111/j.1365-2966.2010.17243. x, 1007.0145 Angus R, Aigrain S, Foreman-Mackey D, McQuillan A (2015) Calibrating gyrochronology using Kepler asteroseismic targets. MNRAS 450:1787–1798. https://doi.org/10.1093/mnras/stv423, 1502.06965 Auvergne M, Bodin P, Boisnard L et al (2009) The CoRoT satellite in flight: description and performance. A&A 506:411–424. https://doi.org/10.1051/0004-6361/200810860, 0901.2206 Bakos G, Noyes RW, Kovács G et al (2004) Wide-Field millimagnitude photometry with the HAT: a tool for extrasolar planet detection. PASP 116:266–277. https://doi.org/10.1086/382735, astroph/0401219 Barnes SA (2010) A simple nonlinear model for the rotation of main-sequence cool stars. I. Introduction, implications for gyrochronology, and color-period diagrams. ApJ 722:222–234. https://doi.org/10.1088/0004-637X/722/1/222 Barnes SA, Kim YC (2010) Angular momentum loss from cool stars: an empirical expression and connection to stellar activity. ApJ 721:675. https://doi.org/10.1088/0004-637X/721/1/675, 1104.2350 Barnes SA, Spada F, Weingrill J (2016a) Some aspects of cool main sequence star ages derived from stellar rotation (gyrochronology). Astronomische Nachrichten 337:810. https://doi.org/10. 1002/asna.201612377 Barnes SA, Weingrill J, Fritzewski D, Strassmeier KG, Platais I (2016b) Rotation periods for cool stars in the 4 Gyr old open cluster M67, the solar-stellar connection, and the applicability of gyrochronology to at least solar age. ApJ 823:16. https://doi.org/10.3847/0004-637X/823/1/16, 1603.09179 Beatty TG, Gaudi BS (2015) Astrophysical sources of statistical uncertainty in precision radial velocities and their approximations. PASP 127:1240. https://doi.org/10.1086/684264, 1507.00780 Béky B, Holman MJ, Kipping DM, Noyes RW (2014) Stellar rotation-planetary orbit period commensurability in the HAT-P-11 system. ApJ 788:1. https://doi.org/10.1088/0004-637X/788/ 1/1, 1403.7526

1718

P. F. L. Maxted

Bolmont E, Raymond SN, Leconte J, Matt SP (2012) Effect of the stellar spin history on the tidal evolution of close-in planets. A&A 544:A124. https://doi.org/10.1051/0004-6361/201219645, 1207.2127 Borucki W, Koch D, Basri G et al (2008) Finding earth-size planets in the habitable zone: the Kepler mission. In: Sun YS, Ferraz-Mello S, Zhou JL (eds) Exoplanets: detection, formation and dynamics, IAU symposium. vol 249, pp 17–24. https://doi.org/10.1017/S174392130801630X Borucki WJ, Koch D, Basri G et al (2010) Kepler planet-detection mission: introduction and first results. Science 327:977. https://doi.org/10.1126/science.1185402 Boué G, Montalto M, Boisse I, Oshagh M, Santos NC (2013) New analytical expressions of the Rossiter-McLaughlin effect adapted to different observation techniques. A&A 550:A53. https:// doi.org/10.1051/0004-6361/201220146, 1211.3310 Bouvier J (2013) Observational studies of stellar rotation. In: Hennebelle P, Charbonnel C (eds) EAS publications series, EAS publications series, vol 62, pp 143–168. https://doi.org/10.1051/ eas/1362005, 1307.2891 Cegla HM, Lovis C, Bourrier V et al (2016) The Rossiter-McLaughlin effect reloaded: probing the 3D spin-orbit geometry, differential stellar rotation, and the spatially-resolved stellar spectrum of star-planet systems. A&A 588:A127. https://doi.org/10.1051/0004-6361/201527794, 1602.00322 Cohen O, Drake JJ, Kashyap VL, Sokolov IV, Gombosi TI (2010) The impact of hot Jupiters on the spin-down of their host stars. ApJ 723:L64–L67. https://doi.org/10.1088/2041-8205/723/1/ L64, 1009.5955 Collier Cameron A, Guenther E, Smalley B et al (2010) Line-profile tomography of exoplanet transits – II. A gas-giant planet transiting a rapidly rotating A5 star. MNRAS 407:507–514. https://doi.org/10.1111/j.1365-2966.2010.16922.x, 1004.4551 Collins GW II, Truax RJ (1995) Classical rotational broadening of spectral lines. ApJ 439:860– 874. https://doi.org/10.1086/175225 Damiani C, Lanza AF (2015) Evolution of angular-momentum-losing exoplanetary systems. Revisiting Darwin stability. A&A 574:A39. https://doi.org/10.1051/0004-6361/201424318, 1411.3802 do Nascimento JD Jr, García RA, Mathur S et al (2014) Rotation periods and ages of solar analogs and solar twins revealed by the Kepler mission. ApJ 790:L23. https://doi.org/10.1088/20418205/790/2/L23, 1407.2289 Dobbs-Dixon I, Lin DNC, Mardling RA (2004) Spin-orbit evolution of short-period planets. ApJ 610:464–476. https://doi.org/10.1086/421510, astro-ph/0408191 Douglas ST, Agüeros MA, Covey KR et al (2016) K2 rotation periods for low-mass hyades and the implications for gyrochronology. ApJ 822:47. https://doi.org/10.3847/0004-637X/822/1/47, 1603.00419 Ferraz-Mello S, Tadeu dos Santos M, Folonier H et al (2015) Interplay of tidal evolution and stellar wind braking in the rotation of stars hosting massive close-in planets. ApJ 807:78. https://doi. org/10.1088/0004-637X/807/1/78, 1503.04369 Fressin F, Torres G, Charbonneau D et al (2013) The false positive rate of Kepler and the occurrence of planets. ApJ 766:81. https://doi.org/10.1088/0004-637X/766/2/81, 1301.0842 Frölich C et al (1995) VIRGO: experiment for helioseismology and solar irradiance monitoring. Sol Phys 162:101–128. https://doi.org/10.1023/A:1004929108864 Gallet F, Bouvier J (2013) Improved angular momentum evolution model for solar-like stars. A&A 556:A36. https://doi.org/10.1051/0004-6361/201321302, 1306.2130 García RA, Ceillier T, Salabert D et al (2014) Rotation and magnetism of Kepler pulsating solarlike stars. Towards asteroseismically calibrated age-rotation relations. A&A 572:A34. https:// doi.org/10.1051/0004-6361/201423888, 1403.7155 Gizon L, Ballot J, Michel E et al (2013) Seismic constraints on rotation of sun-like star and mass of exoplanet. Proc Natl Acad Sci 110:13,267–13,271. https://doi.org/10.1073/pnas.1303291110, 1308.4352 Gonzalez G (2011) Parent stars of extrasolar planets – XII. Additional evidence for trends with v sin i, condensation temperature and chromospheric activity. MNRAS 416:L80–L83. https://doi. org/10.1111/j.1745-3933.2011.01102.x, 1106.5002

83 Rotation of Planet-Hosting Stars

1719

Gray DF (1982) The rotation of cool main-sequence stars. ApJ 261:259–264. https://doi.org/10. 1086/160336 Güdel M (2004) X-ray astronomy of stellar coronae. A&A Rev 12:71–237. https://doi.org/10.1007/ s00159-004-0023-2, astro-ph/0406661 Hansen BMS (2017) Perturbation of compact planetary systems by distant giant planets. MNRAS https://doi.org/10.1093/mnras/stx182, 1608.06300 Hirano T, Suto Y, Winn JN et al (2011) Improved modeling of the Rossiter-McLaughlin effect for transiting exoplanets. ApJ 742:69. https://doi.org/10.1088/0004-637X/742/2/69, 1108.4430 Hirano T, Sanchis-Ojeda R, Takeda Y et al (2014) Measurements of stellar inclinations for Kepler planet candidates. II. Candidate spin-orbit misalignments in single- and multiple-transiting systems. ApJ 783:9. https://doi.org/10.1088/0004-637X/783/1/9, 1401.1229 Howard AW, Johnson JA, Marcy GW et al (2010) The California planet survey. I. Four new giant exoplanets. ApJ 721:1467–1481. https://doi.org/10.1088/0004-637X/721/2/1467, 1003.3488 Janes KA (2017) Rotation periods of wide binaries in the Kepler field. ApJ 835:75. https://doi.org/ 10.3847/1538-4357/835/1/75, 1612.00070 Kovács G (2015) Are the gyro-ages of field stars underestimated? A&A 581:A2. https://doi.org/ 10.1051/0004-6361/201525920, 1506.06267 Lai D (2014) Star-disc-binary interactions in protoplanetary disc systems and primordial spin-orbit misalignments. MNRAS 440:3532–3544. https://doi.org/10.1093/mnras/stu485, 1402.1907 Levrard B, Winisdoerffer C, Chabrier G (2009) Falling transiting extrasolar giant planets. ApJ 692:L9–L13. https://doi.org/10.1088/0004-637X/692/1/L9, 0901.2048 Lo Curto G, Mayor M, Benz W et al (2010) The HARPS search for southern extra-solar planets. XXII. Multiple planet systems from the HARPS volume limited sample. A&A 512:A48. https:// doi.org/10.1051/0004-6361/200913523 Matsumura S, Peale SJ, Rasio FA (2010) Tidal evolution of close-in planets. ApJ 725:1995–2016. https://doi.org/10.1088/0004-637X/725/2/1995, 1007.4785 Maxted PFL, Anderson DR, Collier Cameron A et al (2011) WASP-41b: a transiting hot Jupiter planet orbiting a magnetically active G8V star. PASP 123:547–554. https://doi.org/10.1086/ 660007, 1012.2977 Maxted PFL, Serenelli AM, Southworth J (2015) Comparison of gyrochronological and isochronal age estimates for transiting exoplanet host stars. A&A 577:A90. https://doi.org/10.1051/00046361/201525774, 1503.09111 McQuillan A, Aigrain S, Mazeh T (2013a) Measuring the rotation period distribution of field M dwarfs with Kepler. MNRAS 432:1203–1216. https://doi.org/10.1093/mnras/stt536, 1303.6787 McQuillan A, Mazeh T, Aigrain S (2013b) Stellar rotation periods of the Kepler objects of interest: a dearth of close-in planets around fast rotators. ApJ 775:L11. https://doi.org/10.1088/20418205/775/1/L11, 1308.1845 McQuillan A, Mazeh T, Aigrain S (2014) Rotation periods of 34,030 Kepler main-sequence stars: the full autocorrelation sample. ApJS 211:24. https://doi.org/10.1088/0067-0049/211/2/ 24, 1402.5694 Meibom S, Barnes SA, Latham DW et al (2011) The Kepler cluster study: stellar rotation in NGC 6811. ApJ 733:L9. https://doi.org/10.1088/2041-8205/733/1/L9, 1104.2912 Meibom S, Barnes SA, Platais I et al (2015) A spin-down clock for cool stars from observations of a 2.5-billion-year-old cluster. Nature 517:589–591. https://doi.org/10.1038/nature14118, 1501.05651 Metzger BD, Giannios D, Spiegel DS (2012) Optical and X-ray transients from planet-star mergers. MNRAS 425:2778–2798. https://doi.org/10.1111/j.1365-2966.2012.21444.x, 1204.0796 Newton ER, Irwin J, Charbonneau D et al (2016) The rotation and galactic kinematics of Mid M Dwarfs in the solar neighborhood. ApJ 821:93. https://doi.org/10.3847/0004-637X/821/2/93, 1511.00957 Ogilvie GI (2014) Tidal dissipation in stars and giant planets. ARA&A 52:171–210. https://doi. org/10.1146/annurev-astro-081913-035941, 1406.2207 Ohta Y, Taruya A, Suto Y (2005) The Rossiter-McLaughlin effect and analytic radial velocity curves for transiting extrasolar planetary systems. ApJ 622:1118–1135. https://doi.org/10.1086/ 428344, astro-ph/0410499

1720

P. F. L. Maxted

Pollacco DL, Skillen I, Collier Cameron A et al (2006) The WASP project and the SuperWASP cameras. PASP 118:1407–1418. https://doi.org/10.1086/508556, astro-ph/0608454 Pont F (2009) Empirical evidence for tidal evolution in transiting planetary systems. MNRAS 396:1789–1796. https://doi.org/10.1111/j.1365-2966.2009.14868.x, 0812.1463 Privitera G, Meynet G, Eggenberger P et al (2016) Star-planet interactions. I. Stellar rotation and planetary orbits. A&A 591:A45. https://doi.org/10.1051/0004-6361/201528044, 1604.06005 Radick RR, Thompson DT, Lockwood GW, Duncan DK, Baggett WE (1987) The activity, variability, and rotation of lower main-sequence Hyades stars. ApJ 321:459–472. https://doi. org/10.1086/165645 Radick RR, Lockwood GW, Skiff BA, Thompson DT (1995) A 12 year photometric study of lower main-sequence Hyades stars. ApJ 452:332. https://doi.org/10.1086/176305 Reiners A, Schmitt JHMM (2002) On the feasibility of the detection of differential rotation in stellar absorption profiles. A&A 384:155–162. https://doi.org/10.1051/0004-6361:20011801 Siess L, Livio M (1999) The accretion of brown dwarfs and planets by giant stars – II. Solar-mass stars on the red giant branch. MNRAS 308:1133–1149. https://doi.org/10.1046/j.1365-8711. 1999.02784.x, astro-ph/9905235 Silva-Valio A (2008) Estimating stellar rotation from starspot detection during planetary transits. ApJ 683:L179. https://doi.org/10.1086/591846, 0808.2156 Skumanich A (1972) Time scales for CA II emission decay, rotational braking, and lithium depletion. ApJ 171:565. https://doi.org/10.1086/151310 Struve O (1930) On the axial rotation of stars. ApJ 72. https://doi.org/10.1086/143256 Teitler S, Königl A (2014) Why is there a dearth of close-in planets around fast-rotating stars? ApJ 786:139. https://doi.org/10.1088/0004-637X/786/2/139, 1403.5860 Tinney CG, Butler RP, Marcy GW et al (2001) First results from the Anglo-Australian planet search: a brown dwarf candidate and a 51 peg-like planet. ApJ 551:507–511. https://doi.org/10. 1086/320097, astro-ph/0012204 Torres G, Andersen J, Giménez A (2010) Accurate masses and radii of normal stars: modern results and applications. A&A Rev 18:67–126. https://doi.org/10.1007/s00159-009-0025-1, 0908.2624 Tregloan-Reed J, Southworth J, Tappert C (2013) Transits and starspots in the WASP-19 planetary system. MNRAS 428:3671–3679. https://doi.org/10.1093/mnras/sts306, 1211.0864 Triaud AHMJ, Collier Cameron A, Queloz D et al (2010) Spin-orbit angle measurements for six southern transiting planets. New insights into the dynamical origins of hot Jupiters. A&A 524:A25. https://doi.org/10.1051/0004-6361/201014525, 1008.2353 Twicken JD, Jenkins JM, Seader SE et al (2016) Detection of potential transit signals in 17 quarters of Kepler data: results of the final Kepler mission transiting planet search (DR25). AJ 152:158. https://doi.org/10.3847/0004-6256/152/6/158, 1604.06140 Unsold A (1955) Physik der Sternatmospharen, MIT besonderer Berucksichtigung der Sonne. Springer Berlin Valenti JA, Piskunov N (1996) Spectroscopy made easy: a new tool for fitting observations with synthetic spectra. A&AS 118:595–603 van Saders JL, Ceillier T, Metcalfe TS et al (2016) Weakened magnetic braking as the origin of anomalously rapid rotation in old field stars. Nature 529:181–184. https://doi.org/10.1038/ nature16168, 1601.02631 Walkowicz LM, Basri GS (2013) Rotation periods, variability properties and ages for Kepler exoplanet candidate host stars. MNRAS 436:1883–1895. https://doi.org/10.1093/mnras/stt1700, 1309.2159

83 Rotation of Planet-Hosting Stars

1721

Weber EJ, Davis L Jr (1967) The angular momentum of the solar wind. ApJ 148:217–227. https:// doi.org/10.1086/149138 Weise P, Launhardt R, Setiawan J, Henning T (2010) Rotational velocities of nearby young stars. A&A 517:A88. https://doi.org/10.1051/0004-6361/201014453 Wright JT (2005) Radial velocity jitter in stars from the California and Carnegie planet search at Keck observatory. PASP 117:657–664. https://doi.org/10.1086/430369, astro-ph/0505214

84

Stellar Coronal Activity and Its Impact on Planets Giuseppina Micela

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stellar High-Energy Radiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Heating and Evaporation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Planets Orbiting dM Stars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1724 1726 1727 1729 1733 1734

Abstract

In this chapter the relevance of stellar coronal activity in determining the physical conditions of planetary atmospheres is discussed. We still lack a comprehensive, self-consistent picture of the role of stellar activity, during the star lifespan, in determining and shaping the evolution of planetary atmospheres, after the circumstellar disk has been cleared out, but many efforts in this direction are today ongoing. Here we focus on high-energy radiation since it penetrates deeply the atmosphere ionizing and heating the gas, thus affecting its chemistry with consequences very different from those of optical or UV radiation. Stellar activity is inhomogeneous and variable; flares in particular can be very frequent and intense in young and dM stars. Depending on their duty cycle, they may drive the atmospheric gas toward different chemical regimes.

G. Micela () Istituto Nazionale di Astrofisica – Osservatorio Astronomico di Palermo Giuseppe S. Vaiana, Palermo, Italy e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_19

1723

1724

G. Micela

Introduction Stars and planetary systems are born and evolve in a “cold” environment. Molecular clouds that give origin to stellar systems typically have temperatures of tens of kelvin. When a cloud collapses in a solar-type star, its stellar surface may reach temperatures of 5000–6000 K during the hydrogen-burning stage, while more massive stars may reach temperatures of several tens of thousands degrees. At the same time, planetary systems form circumstellar disks having much colder temperatures. Therefore, low energies are most suitable to study stars and planetary systems: the radio band for their initial phases and the infrared band for protostars, which are still embedded in their original nebula and disks, while the optical band is the best suited to observe more evolved stars. In the last decades, the “cold” scenario has undergone a substantial revision, thanks to the sensitivity and image capability of X-ray observatories. Since the early 1980s, it has become clear that young stars are strong X-ray emitters and that the “cold” scenario does not fully describe the real environment and processes of stellar birth and evolution. High-energy radiation, even being a relatively small fraction of the total energy available, can influence significantly stellar system formation and evolution. The study of planets cannot neglect the presence of their host star: formation and evolution of planets are driven by stellar properties, as they form from the same material, and the energy source of the planets comes from their host stars. Furthermore, in close-in planets, star and planets can interact directly, with the star affecting the planet but also vice versa. Tidal interactions may occur in systems with massive planets moving around stars with convection cells of comparable mass, as in dF stars; at the same time, magnetic interactions can take place between the stellar magnetosphere and, if present, the planetary magnetosphere. Several authors have tried to find observational evidence of this interaction with mixed results. In some cases a hot spot rotating in phase with the planet or modulation of X-ray emission has been observed (Shkolnik et al. 2005, 2008; Gurdemir et al. 2012; Saar et al. 2008), while in other cases, observations failed to find any evidence of such an interaction (e.g., Poppenhaeger et al. 2011; Scandariato et al. 2013). However, theoretical models indicate that enhancement of stellar activity due to interaction between the stellar and planetary magnetospheres may occur (e.g., McIvor et al. 2006; Lanza 2008, 2013; Cohen et al. 2011). Some results suggest that the magnetic interaction produces non-stationary phenomena. Beside the direct interaction of magnetospheres, the planetary atmosphere may be powered by the surplus of energy produced by stellar activity, in particular highenergy radiation, EUV, and X-rays. The temperature of the atmosphere rises, and, if heated sufficiently, it may evaporate, as observed in a couple of hot Jupiters from the Ly’ line (Vidal-Madjar et al. 2003; Lecavelier Des Etangs et al. 2010, 2012) and from other chemical species (Sing et al. 2008; Linsky et al. 2010). Evaporation appears to be variable, as expected, if the source of heating is itself variable. The

84 Stellar Coronal Activity and Its Impact on Planets

1725

presence of a planet magnetic field may limit the evaporation rate (Trammell et al. 2011; Adams 2011). X-ray radiation is one of the many manifestations of stellar activity that shows up in several bands since it involves many stellar regions with different temperatures. Granulation, chromospheric and coronal emission, magnetic cycles, photometric variability, spots, and rotation are examples of different aspects of stellar activity (e.g., Eberhard and Schwarzschild 1913; Wilson 1963 1978; Baliunas et al. 1995; Lockwood et al. 2013; Vogt and Penrod 1983; Strassmeier 2009). All these phenomena decay with stellar age, as recognized by Skumanich (1972), who found that rotation and Ca II H and K lines decay with time following a square root law. The Skumanich work was confirmed and extended by several authors who used the relationship in the reverse direction to estimate stellar age from activity indicators – the so-called gyrochronology (Mamajek and Hillenbrand 2008; Epstein and Pinsonneault 2014; Barnes 2003, 2007; Meibom et al. 2011, 2015). Using rotational periods derived from the K2 mission of the 4.2-Gyr-old open cluster M69, Barnes et al. (2016) have shown that at approximately the solar age, it is possible to derive individual stellar age with an error of 0.7 Gyr (see Fig. 1). These authors suggest that such an age error can be achieved for host planet field stars, which

Fig. 1 Rotation period as a function of B-V color and age. Red points mark M69 members (From Barnes et al. 2016)

1726

G. Micela

would allow the recovery of the past activity history. Today gyrochronology is the most accurate way we have to date field single stars, except for the limited number of cases for which asteroseismology can be applied. The purpose of this work is not the study of the stellar activity per se, but the effect that it may have on the circumstellar environment. For this reason this work will focus on the most energetic phenomena, i.e., coronal emission and flares, that is expected to be those affecting most the environment and hence the planets.

Stellar High-Energy Radiation Since the first X-ray stellar observations, it was evident that (1) X-ray emission is common to all late-type stars and (2) stars lying on a given position of the HR diagram may have a different level of X-ray luminosity (Vaiana et al. 1981). The main parameter responsible for X-ray level is the stellar rotational velocity (Pallavicini et al. 1981), pointing to a magnetic dynamo mechanism as the basis of the corona origin. As a consequence, since rotation evolves with age, coronal emission evolves with stellar age, too. Indeed young solar-type stars emit X-rays at a level 3–4 orders of magnitude higher than the present-day Sun, both during the pre-main sequence phase, when the emission is dominated by intense flares (Feigelson et al. 2003; Favata et al. 2005), and during the first phases of the main sequence, when a solar-like star has an Xray luminosity in the range 1029 1030 erg/s (e.g., Micela 2002). Around 108 year, the X-ray luminosity of a solar-type star starts to decrease, and, at about 5 Gyr, it reaches the (average) value of LX  1027 erg/sec observed for the Sun. The most commonly adopted approach to study the evolution of coronal emission has been through the systematic observation of stellar associations and open clusters for which age is fairly well known. This approach is very reliable for an age younger than approximately one billion of year, range well populated by clusters in the solar neighborhood. On the contrary, nearby older clusters are rare, since the dissipation time is shorter than their age and the survivors are not very prominent in the sky as they lack massive stars; furthermore, since X-ray luminosity decreases with age, they should be within a distance less than few hundreds of parsecs. For this reason, while the decay of LX at a young age relies on open clusters, for older stars it relies on individual stars for which age could not be very accurately determined. All the observations show a very clear decrease in X-ray luminosity with increasing stellar age. Stars maintain more or less the same activity level up to the ’-Persei age (about 50 Myr); then luminosity starts to slowly decrease up to the Hyades age (about 600 Myr), declining much more steeply for an older age. The sketched scenario is valid for single stars: indeed X-ray luminosity decreases more slowly in close binaries, because the tidal forces acting between the components slow down the loss of angular momentum, with the stars being fast rotators even at advanced ages. Since their activity level is driven by stellar dynamo, which is directly related to the rotational velocity, binaries maintain high X-ray luminosity even at old ages. It is not entirely clear whether stars with hot Jupiters evolve

84 Stellar Coronal Activity and Its Impact on Planets

1727

similarly to single stars or binary systems. In the latter case, the determination of their age with gyrochronology would be underestimated. The evolution of X-ray luminosity goes with spectral evolution. In particular the spectrum of X-ray emission of the youngest stars is much harder than that of older stars, including the Sun: the mean coronal temperatures of pre-main sequence stars range from tens of millions kelvin up to one hundred million in the extreme cases, while solar-like coronae have temperatures in the 1–3 (or even less) million kelvin range. From the study of the Sun as a star (Orlando et al. 2004), this fact has been interpreted as an evolution of the relative weights of different coronal components, with the brightest and hottest (flares, cores of active regions) being dominant in the first stellar phases. The X-ray spectral evolution is important in the present context because, while soft X-ray photons are absorbed by a thin material column (parameterized by a hydrogen column – NH – assuming a solar chemical composition), hard photons are much more penetrating, thus affecting the circumstellar environment at larger distances. Therefore, X-ray emission is an important spectral component of a young solartype star (Ribas et al. 2005): for a 100-Myr-old star, the X-ray (0.1–10 nm) flux is larger than the extreme ultraviolet (EUV, 10–90 nm) one, and it remains within a factor of two for stars as old as 1 Gyr; the present Sun ratio is about 0.25 (Cecchi Pestellini et al. 2009). Such a copious hard emission must have affected the processing of circumstellar material. Furthermore the high-energy emission of young stars is highly variable, with huge flares whose luminosity may overwhelm the quiescent stellar X-ray luminosity, and thus much harder X-ray emission than the present Sun. Plasma temperatures are around 107 K, and during the most powerful events, it may increase up to an order of magnitude. Their spectra are very hard, often with significant emission above 10 keV, beyond the passband of Chandra or XMM/Newton, the most sensitive X-ray observatories currently in orbit. The frequency and conditions in which such huge flares may occur are not yet well known. The amount of energy emitted in the X-ray band is only a modest fraction of the entire stellar energy budget, with a maximum fractional value of 103 (the so-called saturation level) except during very bright flares, in which it may further increase. Nevertheless, because of their high energy, X-ray photons produce phenomena that cannot be caused by radiation in any other of the lower energy bands, regardless of their larger flux. As matter of fact, X-ray photons ionize the circumstellar material, changing the physical properties of the environment (interstellar material, circumstellar disks, or already formed planetary systems) and also the star, through complex feedback mechanisms, mainly driven by the magnetic fields.

Heating and Evaporation A unique feature of high-energy irradiation is that all the relevant processes are dominated by secondary electron cascades generated by primary photoelectrons. The cascade of secondary electrons ionizes more atoms, penetrating very deeply

1728

G. Micela

in the medium. This makes a qualitative difference between the effects of X-ray and EUV photons. At the end of the cascade, when the residual energy is below the threshold of the least energetic inelastic process, it heats the material increasing the temperature. Therefore the impact involves not only the outer layers but also the internal regions, depending on the energy of the incident photons and the local physical properties of the medium. The global effect is a substantial rise in the ionization level that modifies the viscosity of the circumstellar medium and its coupling to magnetic fields, also inducing significant chemical effects. Depending on the physical conditions of the environment, the heating may contribute to evaporating the material in the disk and in planetary atmospheres as well as changing the habitability conditions around the star. These effects occur at several stages of the stellar life: during the first phases, when the star is still surrounded by an envelope, the conditions of the circumstellar material may be less conducive to stellar formation; then the physical conditions of the circumstellar disk may be substantially changed, with EUV photons mainly acting at the disk surface and X-rays penetrating more deeply. Planetary formation and migration may be favored or inhibited depending on the induced turbulence and viscosity and on the effects of X-rays on dusts. During its very early stages (classes I and II), a protostar, still embedded in a high-density environment, copiously emits hard photons that become the main ionization source of the nearby interstellar gas up to several astronomical units (AUs), depending on X-ray intensity, spectrum, and density of the circumstellar material. Within this ionized “bubble,” the conditions may be less conducive for further star formation, with a net result of a defect of very young stellar objects around very bright X-ray sources (Sternberg and Dalgarno 1995; Maloney et al. 1996). During the following evolutionary stages, X-ray observations show that very bright and long-duration flares may originate in magnetic loops of length comparable with star-disk separation (Favata et al. 2005) that may constitute the magnetic tubes through which accretion occurs, regulating stellar mass accretion, angular momentum transport, and possibly the outflows often observed in very young stars. There exists direct evidence of the interaction between high-energy photons and the cold neutral or weakly ionized gas in the disk through the observation of fluorescent emission lines. Indeed, Fe I and Fe II lines at 6.4 keV have been observed in a number of cases (e.g., Tsujimoto et al. 2005) as a response to photons with energy larger than 7.1 keV, while Ne II and Ne III IR lines, likely induced by photons of about 1 keV, have been observed with Spitzer and VISIR/VLT (Flaccomio et al. 2009). Other lines may be excited by the interaction between X-rays and the gas in the disk (Hollenbach and Gorti 2009), but little work has been done so far. X-ray photons may also interact with dust in the disk affecting the evolution of grain growth and crystallization. The analysis of the properties of the 10 m silicate emission feature in the disks shows a wide range in the degree of crystallinity and grain sizes (e.g., Sargent et al. 2009). At the same age, disks may be dominated by very small grains or by very large silicate grain aggregates. Analogously the degree of crystallinity may cover a large range. Such differences may be very

84 Stellar Coronal Activity and Its Impact on Planets

1729

relevant for following the growth of planetesimals. Many stellar parameters have been investigated to understand the origin of such diversity (mass, accretion, binarity, etc.), but the only stellar property that shows some correlation with the dust processing mechanism is the stellar X-ray luminosity (e.g., Riaz et al. 2012, for a sample of young brown dwarfs). One possible explanation for such relation is that X-ray-induced ionization increases the active turbulent mixing in the disk, inhibiting sedimentation. Therefore a significant fraction of large grains may reside in the upper disk layers responsible for the silicate emission feature (e.g., Dullemond and Dominik 2008). Systematic spectroscopic NIR observations together with X-ray detailed studies of selected targets would clarify this confused scenario. During more evolved stages, when the disk has been dissipated and planets are already formed, stellar high-energy radiation contributes to heating the planet atmospheres, to changing the habitability conditions and in extreme cases to partially or even totally evaporating the planet. The most important effects occur during the first billion years on the main sequence and are more relevant for lowdensity, low-mass planets (Penz and Micela 2008). Interestingly, close giant planets may drive active regions on stellar surface, enhancing the stellar X-ray emission (e.g., Kashyap et al. 2008). Penz et al. (2008) and Penz and Micela (2008), using a simple modified energylimited escape approach, have evaluated the atmospheric mass loss for a broad range of planetary parameters, both for solar-like and dM stars. They found that G stars located in the high-energy tail of the X-ray luminosity distribution can evaporate most of their planets within 0,05 AU, while a significant fraction of planets can survive if exposed to a moderate X-ray luminosity. Low-density Neptunian planets are more affected than high-density Jupiters (see Fig. 2). As a consequence, the planetary mass distribution of inner planets changes with time. In the case of dM stars, the X-ray flux is significantly lower than the flux from dG stars for a given orbital distance; therefore, the effects on planets are less relevant: mass loss is negligible for hydrogen-rich Jupiter-mass planets at orbits >0.02 AU, while Neptune-mass planets are influenced up to 0.05 AU. Sanz Forcada et al. (2010, 2011) analyzed a sample of known planetary systems for which X-ray observations were available and found that the distribution of planetary mass with X-ray flux is consistent with a scenario in which planet atmospheres have undergone erosion by coronal X-ray and EUV emission, even if a larger sample and a more detailed model are needed to explain in detail the observations.

Planets Orbiting dM Stars Planets orbiting around dM stars are of special interest because low-mass stars are the most common stars in the Galaxy. Furthermore their low intrinsic luminosity, low mass, and small size produce a favorable planet-star contrast making them the best targets for detection and characterization of small planets. Finally, they are the ideal targets to search habitable planets since their habitable zone is pretty close to the star at a distance corresponding to orbital periods of the order of tens of days.

1730

G. Micela 1 0.2

0.9 0.4 4.0

Planetary Mass Distribution

0.8

1.0 10.0

0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

0

0.1

0.2

0.4

0.3

0.5 0.6 M [Mjup]

0.7

0.8

0.9

1

1 0.2

0.9

Planetary Mass Distribution

0.8

0.4 4.0

0.7

1.0

10.0 0.6 0.5 0.4 0.3 0.2 0.1 0

0

0.1

0.2

0.3

0.4

0.5 0.6 M [Mnep]

0.7

0.8

0.9

1

Fig. 2 Mass distribution at 0.02 AU for a Jupiter-mass planet (upper panel, density is 0.4 g/cm3 ) and a Neptune-mass planet (lower panel, density is 2 g/cm3 ) around a dG star at different ages of the system (From Penz et al. 2008)

84 Stellar Coronal Activity and Its Impact on Planets

1731

As said above, the X-ray flux from a dM star incident at a given distance is lower than that from a solar-like star. However, low-mass stars are strongly variable with a flare frequency much higher than the active Sun. The EUV and X-rays’ high-energy radiation modifies the ionization of the atmosphere producing chemical effects; if the flare frequency is high enough, the atmosphere has not sufficient time to come back to the equilibrium. In principle such phenomenon can be observable with future, more advanced spectroscopic instruments. A few works have touched the problem, but we still lack a complete self-consistent picture. Segura et al. (2010) explored the effects on habitability of an earth-like planet, while Rugheimer et al. (2015) derived the spectra of an earth-like planet around active and inactive dM stars. Venot et al. (2016) have studied through a one-dimensional thermophotochemical model the spectra produced during a primary transit of a hot and a warm super-Earth around a dM star as active as AD Leo, subject to a large flare. Results indicate that abundance of several molecules changes during the flares – some increasing, some decreasing – and that the post-flare chemical composition is different from the pre-flare one (see Fig. 3). Since flares are not isolated but follow a distribution, with the weaker flares being more frequent and the stronger flares rarer, the “steady-state” composition can be very different from that of an inactive star, and abundance can change during the largest flares. Most of these variations will be detected if planets will be observed with a high enough signal to noise ratio, with the new forthcoming instruments. Furthermore, the solar largest flares are often accompanied by coronal mass ejection (CME) (Compagnino et al. 2017 and reference therein), i.e., blobs of coronal plasma expelled at high velocity from the star. Both flares and CMEs are related to magnetic reconnection that releases the energy stored in stressed magnetic fields. Considering that solar CMEs may have destructive effects on the terrestrial magnetosphere, we can speculate that larger effects can be present around dM stars where flares are stronger and the habitability zone closer to the star. After the discovery of a super-Earth in the habitable zone (HZ) of Proxima Centauri (Anglada-Escudé et al. 2016), our nearest star, several authors have studied its characteristics and the properties that may hamper or allow habitability in Proxima b. The stellar radiation environment on the planet is very different from that on the Earth; thus many papers focus on characterizing and speculating on its effects. Prox Cen is a very low-mass and cold star whose evolution followed a path along the vertical Hayashi track with a substantial change in its intrinsic luminosity and small temperature variations. With time such an evolution shifts the location of the habitable zone, defined as the region where liquid water may survive on the planetary surface, bringing it closer to the star. As a consequence, Proxima b entered the HZ only after a long phase in which it was much hotter. Prox Cen is a very active star with frequent and energetic flares (e.g., Reale et al. 1988; Haisch et al. 1983; Fuhrmeister et al. 2011) affecting the environment of the planet. Ribas et al. (2016) have evaluated the possibility that the planet has maintained liquid water on its surface, taking into account the history of the high-energy radiation flux received by the planet. Today it receives 60 and 10 times more flux than the Earth in the XUV and far-UV, respectively (see Fig. 4). Ribas et al. evaluated also that the particle flux

1732

G. Micela –4

10–4

10

10–3 –6 10

CO2

Teq = 412 K Pressure (mbar)

Pressure (mbar)

CO2

Initial steady state 100 s 200 s 300 s 400 s 500 s 600 s 700 s 800 s 912 s 1038 s 10–5

10–4 Molar fraction

10–3 –6 10

10–3

10–4

Teq = 412 K Pressure (mbar)

Pressure (mbar)

10–3

10–2

NO

0-1038 s

10–1 100 10

–10

10

10

–9

–8

–7

10 10 Molar fraction

–6

10

–5

10–2

–4

10 Molar fraction

–3

10

Teq = 412 K No 1130-2586 s

–1

10

10–11

10

10–4

–10

10

10–9

10–8 10–7 Molar fraction

10–6

10–5

10–4

Teq = 412 K Pressure (mbar)

Teq = 412 K Pressure (mbar)

10–5

100 –11

10

Initial steady state 1130 s 1222 s 1313 s 1405 s 1497 s 1589 s 1680 s 1864 s 2219 s 2586 s

10–4

10–3

10–3

Teq = 412 K

OH

–2

10–1

10–3

OH

10–2 10–1

0-1038 s

1130-2586 s 100

100 10–1510–1410–1310–1210–1110–10 10–9 10–8 10–7 10–6 10–5 10–4 Molar fraction

10–1510–1410–1310–1210–1110–10 10–9 10–8 10–7 10–6 10–5 10–4 Molar fraction

Fig. 3 Evolution of H, NH2, and NH3 mixing ratios during an extreme flare event with the thermal profile corresponding to Teq D 412 K (From Venot et al. 2016)

received by the planet (due to the stellar wind) is a factor 4–80 higher than that received today by the Earth. The resulting present water abundance on Proxima b is very uncertain because it depends on its initial conditions, the disk dissipation history, the abundance of water present in the system at start, and various other details. The authors have shown that several scenarios are possible: the planet may have lost a substantial mass of volatiles, water, and other elements, thus becoming unsuitable to sustain life, or, on the contrary, it may have lost only a modest fraction of water, being left with a dense humid atmosphere. In either case the history of the high-energy radiation has an important role in determining the present water abundance. Garraffo et al. (2016) model more accurately the particle flux received by Proxima b from the stellar wind. They explore 3D MHD models including wind

84 Stellar Coronal Activity and Its Impact on Planets

1733

Fig. 4 Comparison between the high-energy spectral irradiance received by Proxima b and the Earth (From Ribas et al. 2016)

and magnetic field around the star for all the combinations of orbital parameters consistent with the observations. They found that the pressure experienced by the planet is at least 2000 times higher than that of the Earth, with significant changes, from one to three orders of magnitude, on short time scale, during the orbital motion. Periodically this modifies very rapidly the planetary magnetosphere size, a phenomenon that is not present on Earth and that should have a significant effect on the atmosphere.

Summary This paper discusses the relevance of stellar activity, and X-ray coronal emission in particular, in shaping circumstellar disks and planetary atmospheres, especially in the youngest systems. High-energy radiation penetrates deeply in the atmosphere, heating it significantly, thanks to secondary electrons generated by primary photoelectrons. In the most extreme cases (bright X-ray stars and close planets), part of the planet may evaporate losing a substantial fraction of its mass. The effect is more relevant for solar-type stars and for low-mass, low-density planets. Moreover the EUV and X-ray photons may change the conditions in the canonical habitability region around an active star. Around dM stars the effects are enhanced by the occurrence of high-frequency flares accompanied by CMEs. Besides temperature, also chemistry is modified. The discovery of the Proxima b “habitable” planet offers a unique opportunity to explore the properties of a rocky planet in the habitable zone of a star different from the Sun. Many efforts are being devoted to building a comprehensive, self-consistent picture of the evolution of physical conditions in such a kind of systems.

1734

G. Micela

References Adams FC (2011) Magnetically controlled outflows from hot Jupiters. ApJ 730:27–44 Anglada-Escudé G, Amado PJ, Barnes J et al (2016) A terrestrial planet candidate in a temperate orbit around Proxima Centauri. Nature 536:437–440 Baliunas SL, Donahue RA, Soon WH et al (1995) Chromospheric variations in main-sequence stars. ApJ 438:269–287 Barnes SA (2003) On the rotational evolution of solar- and late-type stars, its magnetic origins, and the possibility of stellar gyrochronology. ApJ 586:464–479 Barnes SA (2007) Ages for illustrative field stars using gyrochronology: viability, limitations, and errors. ApJ 669:1167–1189 Barnes SA, Weingrill J, Fritzewski D, Strassmeier KG, Platais (2016) Rotation periods for cool stars in the 4 Gyr old open cluster M67, the solar-stellar connection, and the applicability of gyrochronology to at least solar age. ApJ 823:16 Cecchi Pestellini C, Ciaravella A, Micela G, Penz T (2009) The relative role of EUV radiation and x-rays in the heating of hydrogen-rich exoplanet atmospheres. A&A 496:863–868 Cohen O, Kashyap V, Drake JJ et al (2011) The dynamics of stellar coronae harboring hot Jupiters. I. A time-dependent magnetohydrodynamic simulation of the interplanetary environment in the HD 189733 planetary system. ApJ 738:67–79 Compagnino A, Romano P, Zuccarello F (2017) A statistical study of CME properties and of the correlation between flares and CMEs over solar cycles 23 and 24. SoPh 292:5–23 Dullemond CP, Dominik C (2008) Size-sorting dust grains in the surface layers of protoplanetary disks. A&A 487:205–209 Eberhard G, Schwarzschild K (1913) On the reversal of the calcium lines H and K in stellar spectra. ApJ 38:292–295 Epstein C, Pinsonneault M (2014) How good a clock is rotation? The stellar rotation-mass-age relationship for old field stars. ApJ 780:159–183 Favata F, Flaccomio E, Reale F et al (2005) Bright x-ray flares in Orion young stars from COUP: evidence for star-disk magnetic fields? ApJS 160:469–502 Feigelson E, Gaffney JAIII, Garmire G, Hillebrand LA, Townsley (2003) X-rays in the Orion nebula cluster: constraints on the origins of magnetic activity in pre-main-sequence star. ApJ 584:911–930 Flaccomio E, Stelzer B, Sciortino S et al (2009) Results from DROXO. II. [Ne II] and x-ray emission from ¡ Ophiuchi young stellar objects. A&A 505:695–706 Fuhrmeister B, Lalitha S, Poppenhaeger K et al (2011) Multi-wavelength observations of Proxima Centauri. A&A 544:133–149 Garraffo C, Drake JJ, Cohen O (2016) The space weather of Proxima Centauri b. ApJ 833:L4–L9 Gurdemir L, Redfield S, Cuntz M (2012) Planet-induced emission enhancements in HD 179949: results from McDonald observations. PASA 29:141–149 Haisch BM, Linsky JL, Bornmann PL et al (1983) Coordinated Einstein and IUE observations of a disparitions brusques type flare event and quiescent emission from Proxima Centauri. ApJ 267:280–290 Hollenbach D, Gorti U (2009) Diagnostic line emission from extreme ultraviolet and x-rayilluminated disks and shocks around low-mass stars. ApJ 703:1203–1223 Kashyap VL, Drake JJ, Saar SH (2008) Extrasolar giant planets and x-ray activity. ApJ 687:1339– 1354 Lanza AF (2008) Hot Jupiters and stellar magnetic activity. A&A 487:1163–1170 Lanza AF (2013) Star-planet magnetic interaction and evaporation of planetary atmospheres. A&A 557:31–43 Lecavelier Des Etangs A, Ehrenreich D, Vidal-Madjar A et al (2010) Evaporation of the planet HD 189733b observed in H I Lyman-’. A&A 514:72–81 Lecavelier Des Etangs A, Bourrier V, Wheatley PJ et al (2012) Temporal variations in the evaporating atmosphere of the exoplanet HD 189733b. A&A 514:4–8

84 Stellar Coronal Activity and Its Impact on Planets

1735

Linsky JL, Yang H, France K et al (2010) Observations of mass loss from the transiting exoplanet HD 209458b. ApJ 717:1291–1299 Lockwood GW, Henry GW, Hall JC, Radick RR (2013) Decadal variations of sun-like stars. In: Chavez M et al (ed) ASP conference series 472, New Quests in Stellar Astrophysics III. ASP, San Francisco, pp 203–212 Maloney PR, Hollenbach DJ, Tielens AGGM (1996) X-ray – irradiated molecular gas. I. Physical processes and general results. ApJ 466:561–584 Mamajek E, Hillenbrand LA (2008) Improved age estimation for solar-type dwarfs using activityrotation diagnostic. ApJ 687:1264–1293 McIvor T, Jardine M, Holzwarth V (2006) Extrasolar planets, stellar winds and chromospheric hotspots. MNRAS 367:L1–L5 Meibom S, Barnes SA, Latham DW et al (2011) The Kepler cluster study: stellar rotation in NGC 6811. ApJ 733:9–13 Meibom S, Barnes SA, Platais I et al (2015) A spin-down clock for cool stars from observations of a 2.5-billion-year-old cluster. Nature 517:589–591 Micela G (2002) Evolution of stellar coronal activity on the main sequence. ASP Conf Ser 269:107 Orlando S, Peres G, Reale F (2004) The sun as an x-ray star: active region evolution, rotational modulation, and implications for stellar x-ray variabilità. A&A 424:677–689 Pallavicini R, Golub L, Rosner R et al (1981) Relations among stellar x-ray emission observed from Einstein, stellar rotation and bolometric luminosità. ApJ 248:279–290 Penz T, Micela G (2008) X-ray induced mass loss effects on exoplanets orbiting dM stars. A&A 479:579–584 Penz T, Micela G, Lammer H (2008) Influence of the evolving stellar x-ray luminosity distribution on exoplanetary mass loss. A&A 477:309–314 Poppenhaeger K, Lenz LF, Reiners A, Schmitt JHMM, Shkolnik E (2011) A search for starplanet interactions in the ¤ Andromedae system at x-ray and optical wavelengths. A&A 528: 58–64 Reale F, Peres G, Serio S, Rosner R, Schmitt JHMM (1988) Hydrodynamic modeling of an x-ray flare on Proxima Centauri observed by the Einstein telescope. ApJ 328:256–264 Riaz B, Honda M, Campins H et al (2012) The radial distribution of dust species in young brown dwarf discs. MNRAS 420:2603–2624 Ribas I, Guinan EF, Guedel M, Audard M (2005) Evolution of the solar activity over time and effects on planetary atmospheres. I. High-energy irradiances (1–1700 Å). ApJ 622:680–694 Ribas I, Bolmont E, Selsis F et al (2016) The habitability of Proxima Centauri b. I. Irradiation, rotation and volatile inventory from formation to the present. ApJ 596:111–128 Rugheimer S, Kaltenegger L, Segura A, Linsky J, Mohanty S (2015) Effect of UV radiation on the spectral fingerprints of earth-like planets orbiting M stars. ApJ 809:57–72 Saar SH, Cuntz M, Kashyap VL, Hall JC (2008) First observation of planet-induced X-ray emission: the system HD 179949. IAU Symp 249:79–81 Sanz Forcada J, Ribas I, Micela G et al (2010) A scenario of planet erosion by coronal radiation. A&A 511:L8–L13 Sanz-Forcada J, Micela G, Ribas I et al (2011) Estimation of the XUV radiation onto close planets and their evaporation. A&A 532:6–23 Sargent BA, Forrest WJ, Tayrien C et al (2009) Dust processing and grain growth in protoplanetary disks in the Taurus-Auriga star-forming region. ApjS 182:477–508 Scandariato G, Maggio A, Lanza AF et al (2013) A coordinated optical and X-ray spectroscopic campaign on HD 179949: searching for planet-induced chromospheric and coronal activity. A&A 552:7–19 Segura A, Walkowicz LM, Meadows V, Kasting J, Hawley S (2010) The effect of a strong stellar flare on the atmospheric chemistry of an earth-like planet orbiting an M dwarf. Biol 10:751–771 Shkolnik E, Walker GAH, Bohlender DA, Gu P-G, Kürster M (2005) Hot Jupiters and hot spots: the short- and long-term chromospheric activity on stars with giant planets. ApJ 622:1075–1090 Shkolnik E, Bohlender DA, Walker GAH, Collier Cameron A (2008) The on/off nature of starplanet interactions. ApJ 676:628–638

1736

G. Micela

Sing DK, Vidal-Madjar A, Lecavelier des Etangs A et al (2008) Determining atmospheric conditions at the terminator of the hot Jupiter HD 209458b. ApJ 686:667–673 Skumanich A (1972) Time scales for CA II emission decay, rotational braking, and lithium depletion. ApJ 171:565–567 Sternberg A, Dalgarno A (1995) Chemistry in dense photon-dominated regions. ApJ 99:565–607 Strassmeier KG (2009) Starspots. A&A R17:251–308 Trammell GB, Arras P, Li Z-Y (2011) Hot Jupiter magnetospheres. ApJ 728:152–175 Tsujimoto M, Feigelson ED, Grosso N et al (2005) Iron fluorescent line emission from young stellar objects in the Orion nebula. ApJS 160:503–510 Vaiana GS, Cassinelli JP, Fabbiano G et al (1981) Results from an extensive Einstein stellar survey. ApJ 245:163–182 Venot O, Rocchetto M, Carl S, Roshni HA, Decin L (2016) Influence of stellar flares on the chemical composition of exoplanets and spectra. Ap J 830:77–91 Vidal-Madjar A, Lecavelier des Etangs A, Désert J-M et al (2003) An extended upper atmosphere around the extrasolar planet HD209458b. Nature 422:143–146 Vogt SS, Penrod GD (1983) Doppler imaging of spotted stars – application to the RS Canum Venaticorum star HR 1099. PASP 95:565–576 Wilson OC (1963) A probable correlation between chromospheric activity and age in mainsequence stars. pJ 138:832–851 Wilson OC (1978) Chromospheric variations in main-sequence stars. ApJ 226:379–396

Signatures of Star-Planet Interactions

85

Evgenya L. Shkolnik and Joe Llama

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Planet-Induced and Orbit-Phased Stellar Emission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Scaling Law to Measure Planetary Magnetic Field Strengths . . . . . . . . . . . . . . . . . . . . . . . . Models and Observations of Planet-Induced Variability at Many Wavelengths . . . . . . . . . . Statistical Studies of Magnetic SPI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Planetary Effects on Stellar Angular Momentum Evolution . . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross-References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1738 1740 1741 1743 1745 1748 1749 1750 1750

Abstract

Planets interact with their host stars through gravity, radiation, and magnetic fields, and for those giant planets that orbit their stars within 10 stellar radii (0.1 AU for a sun-like star), star-planet interactions (SPI) are observable with a wide variety of photometric, spectroscopic, and spectropolarimetric studies. At such close distances, the planet orbits within the sub-Alfvénic radius of the star in which the transfer of energy and angular momentum between the two bodies is particularly efficient. The magnetic interactions appear as enhanced stellar activity modulated by the planet as it orbits the star rather than only by stellar rotation. These SPI effects are informative for the study of the internal dynamics and atmospheric evolution of exoplanets. The nature of magnetic SPI

E. L. Shkolnik () ASU School of Earth and Space Exploration, Tempe, AZ, USA e-mail: [email protected] J. Llama Lowell Observatory, Flagstaff, AZ, USA e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_20

1737

1738

E. L. Shkolnik and J. Llama

is modeled to be strongly affected by both the stellar and planetary magnetic fields, possibly influencing the magnetic activity of both, as well as affecting the irradiation and even the migration of the planet and rotational evolution of the star. As phase-resolved observational techniques are applied to a large statistical sample of hot Jupiter systems, extensions to other tightly orbiting stellar systems, such as smaller planets close to M dwarfs become possible. In these systems, star-planet separations of tens of stellar radii begin to coincide with the radiative habitable zone where planetary magnetic fields are likely a necessary condition for surface habitability.

Introduction Giant planets located 0.1 MJ , a < 0.1 AU orbiting stars with 4200 K < Teff < 6200 K, with a weaker correlation with planet mass. In another study of 210 systems, Krejˇcová and Budaj (2012) found statistically significant evidence that the equivalent width of the Ca II K line emission and log R0HK of the host star correlate with smaller semi-major axis and larger mass of the exoplanet, as would be expected for magnetic and tidal SPI. The efficiency of extracting data from large photometric catalogs has made studying stellar activity of many more planet hosts possible in both the ultraviolet (UV) and X-ray. A study of 72 exoplanet systems by Poppenhaeger et al. (2010) showed no significant correlation between the fractional luminosity .LX =Lbol / with planet properties. They did, however, report a correlation of stellar X-ray luminosity with the ratio of planet mass to semi-major axis .Mp sin i =a/, suggesting that massive, close-in planets tend to orbit more X-ray luminous stars. They attributed this correlation to biases of the radial velocity (RV) planet detection method, which favors smaller and further-out planets to be detected around less active, and thus X-ray faint, stars. A study of both RV- and transit-detected planets by Shkolnik (2013) of the farUV (FUV) emission as observed by the Galaxy Evolution Explorer (GALEX) also searched for evidence of increased stellar activity due to SPI in 300 FGK planet hosts. This investigation found no clear correlations with a or Mp , yet reported tentative evidence for close-in massive planets (i.e., higher Mp /a) orbiting more FUV-active stars than those with far-out and/or smaller planets, in agreement with

85 Signatures of Star-Planet Interactions

1747

Fig. 6 Fractional N V (at 1240Å) luminosity from a sample of 11 K and M dwarf planet hosts is weakly correlated with a measure of the star-planet interaction strength Mp =a, where Mp is the mass of the most massive planet in the system (in Earth masses) and a is the semi-major axis (in AU). The Pearson coefficient and statistical likelihood of a null correlation is shown at the top. This provides tentative evidence that the presence of short-period planets enhances the transition region activity on low-mass stars, possibly through the interaction of their magnetospheres (France et al. 2016)

past X-ray and Ca II results (Fig. 5). There may be less potential for detection bias in this case as transit-detected planets orbit stars with a more normal distribution of stellar activity than those with planets discovered with the RV method. To confirm this, a sample of transiting small and distant planets still needs to be identified. The first statistcal SPI test for lower mass (K and M) systems was reported by France et al. (2016) in which they measured a weak positive correlation between the fractional N V luminosity, a transition region FUV emission line, with Mp =a for the most massive planet in the system. They found tentative evidence that the presence of short-period planets (ranging in Mp sini from 3.5 to 615 MEarth ) enhances the transition region activity on low-mass stars, possibly through the interaction of their magnetospheres (Fig. 6). Cohen et al. (2015) modeled the interaction between an M-dwarf and a nonmagnetized planet like Venus. Their work shows very different results for the localized space-weather environments for the planet for sub- and super-Alfvénic stellar wind conditions. The authors postulate that these dynamic differences would lead to additional heating and additional energy being deposited into the atmosphere of the planet. In all their simulations, they find that the stellar wind penetrates much deeper into the atmosphere than for the magnetized planets simulated in Cohen

1748

E. L. Shkolnik and J. Llama

et al. (2014), suggesting that for planets orbiting M dwarfs a magnetosphere may be necessary to shield the planet’s atmosphere. Vidotto et al. (2014) modeled the stellar wind of six M stars ranging from spectral type M0 to M2.5 to study the angular momentum of the host star and the rotational evolution of the star. They found the stellar wind to be highly structured at the orbital separation of the planet and found that the planetary magnetospheric radii could vary by up to 20% in a single orbit. This will result in high variability in the strength of SPI signatures as the planet orbits through regions of closed and open magnetic field, implying that a larger, statistical study may be the most efficient path forward, especially for M dwarfs.

Planetary Effects on Stellar Angular Momentum Evolution As the evidence continues to mount that star-planet interactions measurably increase stellar activity, and now for a wider range of planetary systems, there remains an ambiguity in the larger statistical, single-epoch studies as to whether or not this effect is caused by magnetic SPI, tidal SPI, or planet search selection biases. Although no tidal SPI has been observed as stellar activity modulated by half the planet’s orbital period (Cuntz et al. 2000a), there may be other effects due to the presence of the planets or planet formation process on the angular momentum evolution of the stars, which might increase the stellar rotation through tidal spin-up or decrease the efficiency of stellar magnetic breaking (Lanza 2010; Cohen et al. 2011). In both cases, the star would be more active than expected for its mass and age. For main sequence FGK stars, the magnetized stellar wind acts as a brake on the stellar rotation, decreasing the global stellar activity rate as the star ages. This well observed process has given rise to the so-called “age-rotation-activity” relationship. However, the presence of a short-period giant planet may affect the star’s angular momentum. Under this scenario, the age-activity relation will systematically underestimate the star’s age, potentially making “gyrochronology” inapplicable to these systems. This poses an issue for evolutionary studies of exoplanets and their host stars, including planet migration models and planet atmospheric evolution. Several studies have found that stars hosting giant planets rotate faster than the evolutionary models predict. This increase in rotation rate is thought to be the direct consequence of tidal spin-up of the star by the planet. Additional evidence for the tidal spin-up of stars by giant planets has been found using two hot Jupiter systems by Schröter et al. (2011) and Pillitteri et al. (2011). These studies searched for X-ray emission from M dwarf companions to the active planet hosts CoRoT-2 and HD 189733. Both systems showed no X-ray emission indicating the age of the systems to be > 2Gyr; however, the rotation-age relation places these systems between 100300 Myr for CoRoT-2 and 600 Myr for HD 189733. A study by Lanza (2010) showed that tides alone cannot spin-up the star to the levels seen in CoRoT-2 and HD 189733. Rather, his study postulated that the

85 Signatures of Star-Planet Interactions

1749

excess rotation is a consequence of interactions between the planetary magnetic field and the stellar coronal field. He proposed that these interactions would result in a magnetic field topology where the majority of the field lines are closed. This configuration would therefore limit the efficiency of the stellar wind to spin-down the star through angular momentum loss. By computing a simple linear force-free model, Lanza (2010) was able to compute the radial extension of the stellar corona and its angular momentum loss. He found that stars that host hot Jupiters show a much slower angular momentum loss rate than similar stars without a short-period giant planet, similar to Cohen et al. (2011). In order to disentangle the possible causes of the observed increased stellar activity of HJ hosts observed from single-epoch observations, it is necessary to monitor the activity throughout the planet’s orbit and over the stellar rotation period. Such studies can better characterize the star’s variability, generate firmer statistical results of any planet induced activity, and assess the underlining physical processes involved. The first and only attempt to date of this was reported by Shkolnik et al. (2008) in which they monitored 13 HJ systems (all FGK stars) in search of orbitphased variability and then found a correlation between the median activity levels modulated by the planet and the Mp sin i =Porb (Figs. 2 and 3). In the case of multiplanet systems, the planet with the largest Mp sin i =Porb should have the strongest SPI effects.

Summary Detecting exoplanetary magnetic fields enables us to probe the internal structures of the planets and to place better constraints on their atmospheric mass loss through erosion from the stellar wind. Searching for the observational signatures of magnetic SPI in the form of planet induced stellar activity has proved to be the most successful method to date for detecting magnetic fields of hot Jupiters. Single-epoch statistical studies in search SPI signatures show that indeed there are significant differences in the activity levels between stars with close-in giant planets compared to those without. However, the cause of this remains ambiguous with four possible explanations. • Induced stellar activity in the form of interactions between the stellar and planetary magnetic fields. • The inhibition of magnetic breaking and thus faster than expected stellar rotation and increased stellar activity. • Tidal spin-up of the star due the presence of the close-in planet. • Lastly, the selection biases of planet hunting techniques. These potential underlying causes of such a result highlight the need for further monitoring campaigns across planetary orbit and stellar rotation periods to clearly identify planet-induced excess stellar activity.

1750

E. L. Shkolnik and J. Llama

The vast majority of SPI studies, both individual monitoring as well as larger single-epoch statistical studies, have concentrated on main sequence FGK stars as they are the dominant hosts of hot Jupiters. These stars have the advantage of being relatively quiescent compared to M dwarfs, and thus teasing out signals produced by magnetic SPI from intrinsic stellar activity is simpler. But they also have the disadvantage of lower stellar magnetic field strengths compared to M dwarfs, lowering the power produced by the interaction. The modeling of magnetic SPI, especially with realistic stellar magnetic maps from ZDI surveys, continues to advance and aid in the interpretation of observed planet phased enhanced activity across the main sequence. Additional models enable quantitative predictions of the radio flux density for stars displaying signatures of SPI. Radio detections of at least a few of these systems will help calibrate the relative field strengths, and provide for the first time, true magnetic field strengths for hot Jupiters. Ongoing and future studies of magnetic SPI in a large sample of systems are necessary for improved statistics and distributions of magnetic fields of exoplanets. Extensions of these techniques to other tightly orbiting stellar systems, such as smaller planets close to M dwarfs, are challenging but possible. In these systems, star-planet separations of tens of stellar radii begin to coincide with the radiative habitable zone where planetary magnetic fields are likely a necessary condition for surface habitability. As more close-in planets around relatively bright M dwarfs are discovered by missions such as TESS, the search for magnetic star-planet interactions will be extended to these low-mass stars.

Cross-References  Accretion of Planetary Material onto Host Stars  Gravitational Interactions and Habitability  Magnetic Fields in Planet-Hosting Stars  Models of Star-Planet Magnetic Interaction  Planetary Interiors, Magnetic Fields, and Habitability  Star-Planet Interactions and Habitability: Radiative Effects  Star-Planet Interactions in the Radio Domain: Prospect for Their Detection  Stellar Coronal Activity and Its Impact on Planets  Stellar Coronal and Wind Models: Impact on Exoplanets

References Bastian TS, Dulk GA, Leblanc Y (2000) A search for radio emission from extrasolar planets. ApJ 545:1058–1063 Bouchy F, Udry S, Mayor M et al (2005) ELODIE metallicity-biased search for transiting Hot Jupiters. II. A very hot Jupiter transiting the bright K star HD 189733. A&A 444:L15–L19 Butler RP, Wright JT, Marcy GW et al (2006) Catalog of nearby exoplanets. ApJ 646:505–522

85 Signatures of Star-Planet Interactions

1751

Christensen UR (2010) Dynamo scaling laws and applications to the planets. Space Sci Rev 152:565–590 Christensen UR, Holzwarth V, Reiners A (2009) Energy flux determines magnetic field strength of planets and stars. Nature 457:167–169 Cohen O, Drake JJ, Kashyap VL et al (2009) Interactions of the magnetospheres of stars and closein giant planets. ApJ 704:L85–L88 Cohen O, Kashyap VL, Drake JJ et al (2011) The dynamics of stellar coronae harboring hot Jupiters. I. A time-dependent magnetohydrodynamic simulation of the interplanetary environment in the HD 189733 planetary system. ApJ 733:67 Cohen O, Drake JJ, Glocer A et al (2014) Magnetospheric structure and atmospheric joule heating of habitable planets orbiting M-dwarf stars. ApJ 790:57 Cohen O, Ma Y, Drake JJ et al (2015) The interaction of Venus-like, M-dwarf planets with the stellar wind of their host star. ApJ 806:41 Cranmer SR, Saar SH (2007) Exoplanet-induced chromospheric activity: realistic light curves from solar-type magnetic fields. ArXiv Astrophysics e-prints Cuntz M, Saar SH, Musielak ZE (2000a) On stellar activity enhancement due to interactions with extrasolar giant planets. ApJ 533:L151–L154 Cuntz M, Saar SH, Musielak ZE (2000b) On stellar activity enhancement due to interactions with extrasolar giant planets. ApJ 533:L151–L154 do Nascimento JD Jr, Vidotto AA, Petit P et al (2016) Magnetic field and wind of kappa ceti: toward the planetary habitability of the young sun when life arose on earth. ApJ 820:L15 Donati JF, Moutou C, Farès R et al (2008) Magnetic cycles of the planet-hosting star  Bootis. MNRAS 385:1179–1185 Dressing CD, Charbonneau D (2015) The occurrence of potentially habitable planets orbiting M dwarfs estimated from the full kepler dataset and an empirical measurement of the detection sensitivity. ApJ 807:45 Fares R, Donati JF, Moutou C et al (2009) Magnetic cycles of the planet-hosting star  Bootis – II. A second magnetic polarity reversal. MNRAS 398:1383–1391 Fares R, Donati JF, Moutou C et al (2010) Searching for star-planet interactions within the magnetosphere of HD189733. MNRAS 406:409–419 Fares R, Donati JF, Moutou C et al (2012) Magnetic field, differential rotation and activity of the hot-Jupiter-hosting star HD 179949. MNRAS 423:1006–1017 Fares R, Moutou C, Donati JF et al (2013) A small survey of the magnetic fields of planet-host stars. MNRAS 435:1451–1462 Farrell WM, Desch MD, Zarka P (1999) On the possibility of coherent cyclotron emission from extrasolar planets. J Geophys Res 104:14,025–14,032 France K, Parke Loyd RO, Youngblood A et al (2016) The MUSCLES treasury survey. I. Motivation and overview. ApJ 820:89 Gurdemir L, Redfield S, Cuntz M (2012) Planet-induced emission enhancements in HD 179949: results from McDonald observations. PASA 29:141–149 Hartman JD (2010) A correlation between stellar activity and the surface gravity of hot Jupiters. ApJ 717:L138–L142 Hussain GAJ, Alvarado-Gómez JD, Grunhut J et al (2016) A spectro-polarimetric study of the planet-hosting G dwarf, HD 147513. A&A 585:A77 Jardine M, Collier Cameron A (2008) Radio emission from exoplanets: the role of the stellar coronal density and magnetic field strength. A&A 490:843–851 Jeffers SV, Petit P, Marsden SC et al (2014) & Eridani: an active K dwarf and a planet hosting star? The variability of its large-scale magnetic field topology. A&A 569:A79 Krejˇcová T, Budaj J (2012) Evidence for enhanced chromospheric Ca II H and K emission in stars with close-in extrasolar planets. A&A 540:A82 Lanza AF (2008) Hot Jupiters and stellar magnetic activity. A&A 487:1163–1170 Lanza AF (2009) Stellar coronal magnetic fields and star-planet interaction. A&A 505:339–350 Lanza AF (2010) Hot Jupiters and the evolution of stellar angular momentum. A&A 512:A77

1752

E. L. Shkolnik and J. Llama

Lanza AF (2012) Star-planet magnetic interaction and activity in late-type stars with close-in planets. A&A 544:A23 Lanza AF (2013) Star-planet magnetic interaction and evaporation of planetary atmospheres. A&A 557:A31 Lazio J, Bastian T, Bryden G et al (2009) Magnetospheric emissions from extrasolar planets. In: astro2010: the astronomy and astrophysics decadal survey, astronomy, vol 2010 Lazio TJW, Shkolnik E, Hallinan G, Planetary habitability study Team (2016) Planetary magnetic fields: planetary interiors and habitability. Technical Report Llama J, Wood K, Jardine M et al (2011) The shocking transit of WASP-12b: modelling the observed early ingress in the near-ultraviolet. MNRAS 416:L41–L44 Maggio A, Pillitteri I, Scandariato G et al (2015) Coordinated X-ray and optical observations of star-planet interaction in HD 17156. ApJ 811:L2 Matsakos T, Uribe A, Königl A (2015) Classification of magnetized star-planet interactions: bow shocks, tails, and inspiraling flows. A&A 578:A6 McIvor T, Jardine M, Holzwarth V (2006) Extrasolar planets, stellar winds and chromospheric hotspots. MNRAS 367:L1–L5 Mengel MW, Marsden SC, Carter BD et al (2017) A BCool survey of the magnetic fields of planethosting solar-type stars. MNRAS 465:2734–2747 Mignone A, Bodo G, Massaglia S et al (2007) PLUTO: a numerical code for computational astrophysics. ApJS 170:228–242 Mignone A, Zanni C, Tzeferacos P et al (2012) The PLUTO code for adaptive mesh computations in astrophysical fluid dynamics. ApJS 198:7 Miller BP, Gallo E, Wright JT, Pearson EG (2015) A comprehensive statistical assessment of starplanet interaction. ApJ 799:163 Moutou C, Donati JF, Savalle R et al (2007) Spectropolarimetric observations of the transiting planetary system of the K dwarf HD 189733. A&A 473:651–660 Pagano I, Lanza AF, Leto G et al (2009) CoRoT-2a magnetic activity: hints for possible star-planet interaction. Earth Moon Planets 105:373–378 Pillitteri I, Wolk SJ, Cohen O et al (2010) XMM-Newton observations of HD 189733 during planetary transits. ApJ 722:1216–1225 Pillitteri I, Günther HM, Wolk SJ, Kashyap VL, Cohen O (2011) X-ray activity phased with planet motion in HD 189733? ApJ 741:L18 Pillitteri I, Wolk SJ, Lopez-Santiago J et al (2014a) The corona of HD 189733 and its X-ray activity. ApJ 785:145 Pillitteri I, Wolk SJ, Sciortino S, Antoci V (2014b) No X-rays from WASP-18. Implications for its age, activity, and the influence of its massive hot Jupiter. A&A 567:A128 Pillitteri I, Maggio A, Micela G et al (2015) FUV variability of HD 189733. Is the star accreting material from its hot Jupiter? ApJ 805:52 Piskunov N (1996) Doppler imaging of eclipsing binaries. In: Strassmeier KG, Linsky JL (eds) Stellar surface structure, IAU symposium, vol 176, p 45 Poppenhaeger K, Robrade J, Schmitt JHMM (2010) Coronal properties of planet-bearing stars. A&A 515:A98 Powell KG, Roe PL, Linde TJ, Gombosi TI, De Zeeuw DL (1999) A solution-adaptive upwind scheme for ideal magnetohydrodynamics. J Comput Phys 154:284–309 Preusse S, Kopp A, Büchner J, Motschmann U (2006) A magnetic communication scenario for hot Jupiters. A&A 460:317–322 Saar SH, Cuntz M, Shkolnik E (2004) Stellar activity enhancement by planets: theory and observations. In: Dupree AK, Benz AO (eds) Stars as Suns: activity, evolution and planets, IAU symposium, vol 219, p 355 Scandariato G, Maggio A, Lanza AF et al (2013) A coordinated optical and X-ray spectroscopic campaign on HD 179949: searching for planet-induced chromospheric and coronal activity. A&A 552:A7 Schröter S, Czesla S, Wolter U et al (2011) The corona and companion of CoRoT-2a. Insights from X-rays and optical spectroscopy. A&A 532:A3

85 Signatures of Star-Planet Interactions

1753

Shkolnik EL (2013) An ultraviolet investigation of activity on exoplanet host stars. ApJ 766:9 Shkolnik E, Walker GAH, Bohlender DA (2003) Evidence for planet-induced chromospheric activity on HD 179949. ApJ 597:1092–1096 Shkolnik E, Walker GAH, Bohlender DA, Gu PG, Kürster M (2005a) Hot Jupiters and hot spots: the short- and long-term chromospheric activity on stars with giant planets. ApJ 622:1075–1090 Shkolnik E, Walker GAH, Rucinski SM, Bohlender DA, Davidge TJ (2005b) Investigating Ca II emission in the RS canum venaticorum binary ER vulpeculae using the broadening function formalism. AJ 130:799–808 Shkolnik E, Bohlender DA, Walker GAH, Collier Cameron A (2008) The on/off nature of starplanet interactions. ApJ 676:628–638 Tóth G, Sokolov IV, Gombosi TI et al (2005) Space weather modeling framework: a new tool for the space science community. J Geophys Res Space Phys 110(A12):n/a–n/a. https://doi.org/10. 1029/2005JA011126, a12226 Tóth G, van der Holst B, Sokolov IV et al (2012) Adaptive numerical algorithms in space weather modeling. J Comput Phys 231:870–903 Treumann RA (2006) The electron-cyclotron maser for astrophysical application. A&A Rev 13:229–315 Vidotto AA, Opher M, Jatenco-Pereira V, Gombosi TI (2009) Three-dimensional numerical simulations of magnetized winds of solar-like stars. ApJ 699:441–452 Vidotto AA, Jardine M, Helling C (2010) Early UV ingress in WASP-12b: measuring planetary magnetic fields. ApJ 722:L168–L172 Vidotto AA, Fares R, Jardine M et al (2012) The stellar wind cycles and planetary radio emission of the  boo system. MNRAS 423:3285–3298 Vidotto AA, Jardine M, Morin J et al (2013) Effects of M dwarf magnetic fields on potentially habitable planets. A&A 557:A67 Vidotto AA, Jardine M, Morin J et al (2014) M-dwarf stellar winds: the effects of realistic magnetic geometry on rotational evolution and planets. MNRAS 438:1162–1175 Walker GAH, Croll B, Matthews JM et al (2008) MOST detects variability on  Bootis a possibly induced by its planetary companion. A&A 482:691–697 Wood BE, Müller HR, Zank GP, Linsky JL, Redfield S (2005) New mass-loss measurements from astrospheric Ly˛ absorption. ApJ 628:L143–L146

Magnetic Fields in Planet-Hosting Stars

86

Claire Moutou, Rim Fares, and Jean-François Donati

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Observation Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Magnetic Fields in the Sun, Sun-Like, and Cool Stars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Observations of Planet-Host Magnetic Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Exploration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Full Characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . General Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions and Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1756 1757 1761 1762 1762 1763 1767 1769 1770

Abstract

The stellar magnetic field is a prime ingredient in the interactions between a parent star and its planets. Impacts on the stellar surface or dynamo as well as on the planetary atmosphere and internal structure are expected from these interactions. The magnetic field also plays a huge role in the formation and

C. Moutou () CNRS/CFHT, Kamuela, HI, USA CNRS, LAM, Laboratoire d’Astrophysique de Marseille, Aix Marseille University, Marseille, France UdeM/UL, Montréal, QC, Canada e-mail: [email protected] R. Fares INAF – Osservatorio Astrofisico di Catania, Catania, Italy e-mail: [email protected] J.-F. Donati CNRS, Institut de Recherche en Astrophysique et Planétologie, Toulouse, France e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_21

1755

1756

C. Moutou et al.

evolution of the system. The magnetic properties of planet-host stars, however, are barely known. Although it is impossible to spatially resolve the stellar surface of any star other than the Sun, spectropolarimetry allows probing the global large-scale magnetic strength and orientation at a given time. Then, monitoring polarized signatures over time as the star rotates gives us the possibility to reconstruct the topology of the magnetic field at the stellar surface. The method, the planet-host star sample observed so far, and the conclusions obtained from such observations are presented. Fifteen stars with planets have a detected and characterized magnetic field, including the Sun. Although global properties of stars with planets apparently resemble those of stars without known planets, detailed characterization of specific systems has opened a way to probe the energetic environment of exoplanets, with applications on radio emission, habitability, stellar wind/planetary atmosphere interactions, orbital decay, and Ohmic dissipation.

Introduction The magnetic field of the host star is one of the main ingredients in the making of a planetary system, together with the mass of the disc and its composition. When inward migration takes place and some of the formed planets get very close to the star, again the stellar magnetic field may have a huge impact on the evolution of the planets, mainly on the atmospheric escape (Lammer et al. 2012; Tanaka et al. 2014) and orbital decay (Lin et al. 1996). More distant planets, for instance, those in the habitable zone of the system, are still under the influence of the stellar wind and particle flux, both regulated by the stellar magnetic field. More basically, this magnetic field shapes the activity behavior at the surface of the star, which hampers the mere detection of the planetary companions, when using the indirect methods as radial velocity, astrometry, and (in a lesser extent) transit techniques. Precise interaction mechanisms between the star and planets are complex, and they are not constant with time. As the cool star rotates and progresses through its dynamo cycle, different effects might balance each other. For planets, the expected impacts are the worse: evaporation and erosion, dissipation of energy into internal structure, and orbital evolution. The star may also see some impacts from interacting with its planets, especially if nearby and massive: induced activity or dynamo, mass loss, instability of the rotation axis, and rotation rate. Both the star and the planet are affected when the inward migration results in the planet crushing into the star. It is thus primordial to characterize the magnetic field of planet-host stars in the best possible way. In the following, we will review the state of the art in characterizing the magnetic field of stars with planets. Investigating the surface magnetic topology of planet-host stars may have different goals and implications: (i) exploring in quantified details a few specific systems, (ii) collecting global properties to compare with non-planet-forming stars or multiple stellar systems, and (iii) detecting signature of star-planet interactions such as induced stellar activity or anomalous stellar magnetic cycles.

86 Magnetic Fields in Planet-Hosting Stars

1757

Observation Techniques Indirect manifestations of the magnetic field are observed through emission in chromospheric lines (like CaII H & K or H˛), bisector analyses of mean stellar lines, X-ray, or UV flux. As shown, e.g., by Meunier and Delfosse (2011) or Hébrard et al. (2014), the relation between these indirect tracers and the magnetic features themselves is far from simple. In the following, we focus the attention on the measurement of the magnetic field itself. Magnetic fields impact the spectral lines of stellar atmospheres through the Zeeman effect. It results in a splitting and a polarization in lines. Studying the Zeeman effect in spectral lines thus gives information about the intensity and orientation of the magnetic field prevailing in the region where such lines are formed. With a high-resolution spectrograph, it is possible to resolve the splitting of the different Zeeman components of a line if the splitting is larger than the rotational broadening (very strong fields). The magnetic field also induces a polarization of these lines which depends on the relative orientation of the magnetic field with respect to the line of sight. By using high-resolution spectropolarimetry, it is thus possible to recover the parameters of the magnetic field produced by stars on large scales, by measuring the polarization of spectral lines through the Stokes parameters. Circular polarization gives a direct measurement of the longitudinal component of the field Bl , e.g., the magnetic field projected over the line of sight and averaged over the visible hemisphere (Donati and Brown 1997): 11

Bl D 2:14  10

R 0 geff

vV .v/d v R .Ic  I .v//d v

(1)

where I and V are the intensity and circular polarization Stokes parameters, Ic is the continuum intensity level, geff is the effective Landé factor (sensitivity to the magnetic field), 0 is the line central wavelength, and v is the radial velocity. This method can measure weak magnetic fields. Bl is also called the net longitudinal magnetic field strength on the visible stellar hemisphere; other types of magnetic field measurements are summarized in, e.g., Mathys (2012) and Linsky and Schöller (2015). In quiet main-sequence stars such as those usually targeted for exoplanet searches, the circularly polarized signal precisions are very small, of the order of 103 of the nonpolarized continuum or less. However, by combining the information from thousand spectral lines, it is sometimes possible to detect and measure the circular polarization in the lines of those quiet stars. Then, by monitoring the stellar polarized spectrum over its rotational cycle, one gets a temporal evolution of the signature profiles and an estimate of the longitudinal field with rotational phase. The linear polarization signal is even fainter and usually undetectable. Figure 1 shows some examples of observed V Stokes profiles as a function of radial velocity. This illustrates the variety of signatures, their complexity, and their

V/Ic(%)

HD 46375

HD102195

HD73256

V830 Tau

Fig. 1 Circular polarization profiles of a few example data sets of planet-host stars: Kepler-78 (Moutou et al. 2016), HD 46375, HD 102195, HD 73256 (Fares et al. 2013), and the T Tauri star V830 Tau (Donati et al. 2015) (from left to right). The observed and synthetic profiles are shown in black and red, respectively. On the left of each profile, we show a ˙1 error bar, while on the right, the rotational cycles are indicated

Kepler-78

1758 C. Moutou et al.

86 Magnetic Fields in Planet-Hosting Stars

1759

sometimes fast variation with time as the stellar surface rotates. The constancy of the profile observed for HD 46375 (a slowly rotating star with a dominant dipolar field) contrasts with the rapidly evolving signatures of HD 102195 and HD 73256 or the multicomponent features in the wide profiles of V830 Tau. High-resolution, wide spectral range, polarimetric capabilities and high throughput are necessary to detect these weak magnetic field signals. Relevant instruments are ESPaDOnS at the 3.6-m Canada-France-Hawaii Telescope (Mauna Kea) (Donati et al. 2006a), its twin NARVAL at the 2-m Télescope Bernard Lyot (France), and HARPSpol at the 3.6-m telescope at La Silla (Chile) (Snik et al. 2011). They are operating in the optical domain (370–1050 nm for ESPaDOnS and NARVAL and 380–690 nm for HARPSpol). The spectral resolutions and throughputs are, respectively, 65,000 (10–15%) and 110,000 (2–3%) for ESPaDOnS/NARVAL and HARPSpol. Monitoring requirements are intrinsic to the observation strategy, since signatures must be collected over the full rotation cycle of the star and their repeatability checked over a couple of rotation cycles. Iterative inversion methods are then used to reconstruct the global, large-scale topology of the magnetic field at the stellar surface, in a way similar to medical imaging and described in Donati and Brown (1997) and Donati et al. (2006b). This tomographic technique is called Zeeman-Doppler imaging (ZDI). The magnetic field is decomposed in its poloidal and toroidal components expressed as spherical harmonics expansions. To find the best-fit solution, maximum entropy criteria are used, so that the reconstructed field is the simplest possible. The models of each observed profile are shown in Fig. 1 in red. The output of this analysis technique is the map of active regions at the stellar surface, as well as the orientation of the magnetic field in those regions. For fast-rotating stars, a slightly different reconstruction method is also used to model the Stokes V profiles that give results in agreement with ZDI (Kochukhov 2015). Two examples of reconstructed maps are shown in Fig. 2 for the planet hosts Kepler-78 (Moutou et al. 2016) and V830 Tau (Donati et al. 2015). The sampling of Kepler-78’s map, shown with the ticks, is particularly regular although the time series is short (of the order of a single rotation period); in cases where the rotation period is not well known, or if differential rotation may be significant, it is preferable to observe several consecutive rotation periods, as was obtained in this map of V830 Tau. The colored areas on the reconstructed map give the position, amplitude, and orientation of the magnetic field at the surface of the stars. Two parameters are important in this reconstruction process: the inclination and rotation of the star. Typical precisions of ˙10ı for the inclination and 1–10% of the rotation period can be determined from a well-sampled data set. The techniques may be sensitive to differential rotation, at a level lower than observed on the Sun (Morin et al. 2008a). The resolution obtained in the map then depends both on the rotation rate and the complexity and strength of the field topology. It is also possible to get the map of the surface brightness, for fast rotators as done by Donati et al. (2015) or for slow rotators as done by Hébrard et al. (2016). When the polarized signal is poor and/or the sampling is scarce, the reconstructed field has lower

1760

C. Moutou et al. Meridional magnetic field 0.50

Azimuthal magnetic field 0.50

0.75

0.25

0.75 0.25

0.75

–20

0.25

0

20

Radial magnetic field 0.50

0.00

0.00

0.00

Meridional magnetic field 0.50

Azimuthal magnetic field 0.50

0.75

0.75

0.75 0.25

0.25

–500

0.25

0

500

Radial magnetic field 0.50

0.00

0.00

0.00

Fig. 2 Magnetic map of Kepler-78 (top, from Moutou et al. 2016) and V830 Tau (bottom, from Donati et al. 2015). The three components of the field in a spherical coordinate system in a flattened polar view of the star are presented, down to latitude 30ı . The bold circle represents the equator. The small radial ticks around the star represent the rotational phases of our observations. The radial, azimuthal, and meridional field maps are labeled in G and have the same color scale. Note the different scales for both stars

strength than what would be reconstructed with higher-quality data, as quantified by Fares et al. (2012). The surface maps can then be used to extrapolate the structure of the magnetosphere in the extended atmosphere of stars or their close environment (Jardine et al. 2002; Vidotto et al. 2015), where exoplanets may form and evolve. It is then possible to estimate the magnetic energy available for interactions with a potential planetary magnetosphere and the expansion of the planet magnetosphere (assuming a planetary magnetic field strength). This extrapolation technique assumes a potential field. A source surface represents the Alfven radius, beyond which the field becomes purely radial due to the stellar wind and which encompasses closed field lines. Closed loops connect regions of different magnetic polarities and reach different heights in the corona. When the planet is closer to the star than the Alfven radius, it crosses regions of null magnetic energy (neutral lines), followed by closed loops and open lines. When it orbits outside the Alfven radius, its magnetic field can reconnect with open field lines, where the stellar wind is transported. In all cases, the values of the magnetic energy received by the planet are not constant along its orbit. An example of extrapolated field in a region of several stellar radii around the star is shown in Fig. 3 and features a complex coronal field.

86 Magnetic Fields in Planet-Hosting Stars

1761

Fig. 3 The extrapolated magnetic field of Kepler-78 (left) and V830 Tau (right) as seen from a random angle. White lines correspond to the closed magnetic lines and blue ones to the open field lines. The magnetic strength at the surface of V830 Tau is 20 times greater than for Kepler-78 (Figures from Moutou et al. 2016 and Donati et al. 2015)

Magnetic Fields in the Sun, Sun-Like, and Cool Stars The question of the origin and evolution of magnetic fields of cool stars (with outer convective zones) still challenges theories. If the dynamo process carries general consensus, its detailed mechanisms are not yet understood. The Sun, however, exhibits a vast variety of magnetic phenomena: sunspots, prominences, massive ejections, hot corona, and a 22-year cycle of polarity reversals. Observations of the magnetic Sun have been exploring these phenomena since a century with increasing accuracy and have shown the complexity of their inter-coupling. As the Sun is the only star for which spatial and temporal resolution are accessible, it has been the benchmark for magneto-hydro-dynamical models and numerical simulations, later applied and modified for other stars (e.g., Brun et al. 2015a, b). It is known from observations that magnetic fields are present in stars of all evolutionary stages and masses. Their strength and topology evolve with time and vary with mass and rotation period (Donati and Landstreet 2009). For stars cooler than 6800 K, the magnetic field is mostly driven by convection (while hotter star’s field would be fossil remnants from their forming cloud). The observation of magnetic fields in cool stars of varying mass, rotation rate, and age thus offers additional constraints supporting the dynamo models, beyond the precise observation of the Sun. For instance, it is observed that very young stars have the strongest magnetic energy, as shown in the age-magnetic field trend study of Vidotto et al. (2014), with the following power law between the unsigned magnetic strength B and age t : B / t 0:655˙0:045 . It is also observed that solar-type stars more active than the Sun have strong azimuthal components in their magnetic morphology,

1762

C. Moutou et al.

especially when they rotate fast (Donati et al. 1992, 2003; Petit et al. 2008; Linsky and Schöller 2015). Main-sequence stars having a mass lower than 0:5Mˇ (also called M dwarfs) may have different behaviors than solar analogs, since they are close or below the threshold where stars are fully convective (0:35Mˇ ). Two distinct categories of magnetic topologies are found for those stars: either strong axisymmetric dipolar fields or weak fields with non-axisymmetry or toroidal components (Morin et al. 2008b, 2010, 2011). Although not yet observed in M stars hosting planets, this bistable behavior could be related to the dynamo processes, or to other mechanisms impacting stellar evolution, as the existence of as-yet-unknown planets.

Observations of Planet-Host Magnetic Fields Exploration Since the stellar surface is not resolved and cancelations of the polarized signal take place, the final polarization signal of planet-host stars tends to be very weak and may remain undetected for moderately active stars. An exploratory phase is thus necessary, where a few polarized sequences are collected on a wide sample of stars. Such surveys have been performed, e.g., by the Bcool collaboration, and are greatly informative on the type of main-sequence stars where magnetic fields can be characterized (Marsden et al. 2014; Mengel et al. 2017). The detection is easier on late-type stars for which the number of spectral lines in the optical range is the greatest and the magnetic strength is the largest. Fast rotators and active stars are more favorable for magnetic field observations but tend to be rare among planet-host stars, due to detection biases in exoplanet searches. In addition to providing a useful comparison sample (stars without planets), it is expected that the Bcool sample actually contains present and potential planet hosts, for which stellar characterization will already be known if exoplanets ever get discovered. Easier targets for magnetic field investigations are bright and/or active stars. In the exoplanet-host population, this corresponds to either young stars with an intense field, a short rotation period and strong activity signatures, or bright main-sequence, late-type stars, with moderate rotation and low activity levels. Most of planet-host stars are not in these categories: either too faint, as most transit survey planet systems, or too quiet. Their polarized spectra require more photons than can be collected in a reasonable time with a 4 m class telescope, even equipped with a highthroughput, wide spectral-range optical spectropolarimeter. Also, finding planets of very young and active stars is challenging. That explains why only a small part of the full exoplanet system sample has so far been observed in spectropolarimetry for a magnetic field investigation. Fares et al. (2013) gathered a first sample of ten planet-bearing stars where the magnetic field was searched for. Focusing on the brightest hot Jupiter hosts, this survey has a good detection rate of 70%. Most of the stars were actually monitored and they will be discussed in the next section. CoRoT-7, XO-3, and HAT-P-2 are the

86 Magnetic Fields in Planet-Hosting Stars

1763

three stars without detection: all three originating from transit surveys are relatively fainter than the radial velocity survey stars, and detection limits of several tens of Gauss were found for them. Mengel et al. (2017) has collected snapshot observations of 33 planet-host stars with NARVAL, and only 24% of them show a magnetic signature, mostly the ones having an activity S -index larger than 0.18 (the mean S-index value of the Sun is found by Egeland et al. 2016 to be 0.17). A more extensive exploratory program is being executed with the most efficient ESPaDOnS spectropolarimeter, which will add about 70 new investigated planet-host systems (Moutou in prep). The detection rate, however, will remain low for this challenging population of stars.

Full Characterization A full characterization of the magnetic field of a planet-host star is obtained from a series of polarized spectra acquired typically over two to three consecutive rotational cycles. Most stars will have noticeable variations of their surface magnetic regions beyond three rotation cycles, which make the modelization more complex and less reliable. On the other hand, monitoring the amplitude of these modulations over longer timescales is also important for their impact on close-in planets. In the observational techniques being very time-consuming, it is nevertheless very difficult to get enough data for an extensive description of the magnetic evolution, even for a limited sample of stars (in average 10 h of telescope time is required for each map of 2 rotation periods, for a dozen visits of a quiet 8-mag star). There are 14 systems for which the magnetic field topology has been reconstructed from observations. Their characteristics are summarized in Table 1, where the Sun was added as a reference (since it does bear planets). Some of these systems ( Boo, HD 189733, " Eri, HD 179949, V830 Tau, and Tap 26) have been observed repeatedly over several months to years. Here is a summary of their most important characteristics:  Boo (F6V) magnetic field has been first detected by Catala et al. (2007) and then followed up over 10 years (Donati et al. 2008; Fares et al. 2009, 2013; Mengel et al. 2016). This F6V star has a shallow convective envelope and a massive close-in planet (6 MJup ) that appears to be synchronized with the rotation of the stellar surface at a latitude of 45ı , with a period of 3.31 days. Multiple polarization reversals have been witnessed since 2008 (see references above) that tend to show that the magnetic cycle of the star is extremely short, with periods of 720 or 240 days (Fares et al. 2013). This cycle is also related to a possible chromospheric activity cycle of 120 days (Mengel et al. 2016). The role of the close-in planet in accelerating this magnetic cycle is suggested (Donati et al. 2008), with the massive planet potentially synchronizing the convective envelope only, thus generating an enhanced shear at the base of the shallow convective zone and to tidal interactions of this zone with the massive planet. Elliptical instability processes due to tidal interactions can also contribute to the generation of a dynamo (Cébron and Hollerbach 2014). The magnetic field of  Boo is mostly poloidal with a contribution of the toroidal

Mass Mˇ 1.38 0.82 1.28 1.07 0.9 0.87 0.97 1.05 0.79 0.856 1.0 1.0 0.81 0.88 1.0

Prot Days 3.0 12.5 7.6 10.0 7.0 12.3 42 14 26 11.68 2.7 0.7 12.56 43.4 24.47

Age Gy 0.9 0.6 2.05 0.45 0.88 2.4 4.96 0.83 0.35 0.44 0.002 0.017 0.625 5.13 4.6

Mp sin(i) MJup 5.84 1.14 0.92 1.21 3.37 0.453 0.2272 1.869 1.043 1.55 0.77 1.9 0.0059 0.23 1.0 4.94 9–11 0.355 62.2 12

Porb Days 3.3125 2.2186 3.0925 528.4 133.7 4.1138 3.0236 2.5486 10.708 ESP/NAR ESP/NAR ESP HARPS-Pol HARPS-Pol ESP ESP ESP ESP NAR ESP ESP ESP NAR SDO

Instr

B G 1.7–3.9 22–36 2.6–3.7 1–3.2 10 12.4 2–3.2 2.7 2.5 10–20 340 120 16 3.93 5–8 F13, Me16 F10 F12 H16 AG15 F13 G10 F13 F13 J14 D16 Y17 M16 S15 Ha16

Ref

References (for magnetic field measurements): Gaulme et al. (2010) (G10), Fares et al. (2010) (F10), Fares et al. (2012) (F12), Fares et al. (2013) (F13), Jeffers et al. (2014) (J14), Alvarado-Gómez et al. (2015) (AG15), See et al. (2015) (S15), Moutou et al. (2016) (M16), Hussain et al. (2016) (H16), Mengel et al. (2016) (Me16), Donati et al. (2016) (D16), Haywood et al. (2016) (Ha16), Yu et al. (2017, in press) (Y17)

 Boo HD 189733 HD 179949 HD 147513 HD 1237 HD 102195 HD 46375 HD 73256 HD 130322 " Eri V830 Tau Tap 26 Kepler-78 HD 3651 Sun

Star

Table 1 Summary of observed data and properties of the planetary systems

1764 C. Moutou et al.

86 Magnetic Fields in Planet-Hosting Stars

1765

component fluctuating with time; all components sum up to a total mean energy of 1.7–3.9 G depending on the epoch of observation (Fares et al. 2013; Mengel et al. 2016). From the extrapolated coronal field of  Boo, it was possible to estimate the mass loss and the angular momentum loss along the magnetic cycle of the star, as well as the frequency and flux of a radio emission that would be due to star-planet interactions (Vidotto et al. 2012). HD 179949 (F8V) has been the focus of early attention when chromospheric activity observed on the star with the orbital period of the planet was interpreted as activity triggered by star-planet magnetic connection (Shkolnik et al. 2003). This activity signal, however, seems intermittent and had disappeared in a follow-up campaign (Shkolnik et al. 2008). HD 179949 features a mostly poloidal magnetic topology with 40% of the energy being axisymmetric and a strength of about 2.6– 3.7 G during 2007 and 2009 observing campaigns (Fares et al. 2012). A marginal detection of augmented chromospheric activity at the system’s beat period is reported, i.e., when the average magnetic field is the strongest of both epochs. That would support the theory that star-planet interactions need a given energy threshold to induce magnetospheric connections (Shkolnik et al. 2008). HD 189733 (K2V) has a much stronger magnetic field than the first two ones, with an average amplitude of 22–36 G, as measured in observing campaigns spanning 2006 through 2008 (Moutou et al. 2007; Fares et al. 2010). As expected for a K dwarf, the magnetic field structure is predominantly toroidal and presents more structures than earlier-type stars as  Boo and HD 179949. Although more chromospherically active than previous examples, there is no evidence of an additional contribution of activity at the beat period of the system – where the planet-star-observer configuration is the same. HD 189733’s magnetic field is worth continuous monitoring for quantifying the processes of atmospheric escape that depend on the (varying) stellar wind (Lecavelier des Etangs et al. 2012; Bourrier and Lecavelier des Etangs 2013). " Eri (K2V) is a younger and more active star than HD 189733. Its magnetic topology has been observed at six different epochs from 2007 to 2013, and relations with the chromospheric activity level were searched by Jeffers et al. 2014. The mean field energy varies from 10 to 20 G over this period, and the topology evolves from mostly poloidal to mostly toroidal and back to poloidal. It must be noted that the existence of an outer giant planet in this system is still controversial; in this respect, the investigation of stellar magnetic activity over timescale similar to the orbital period is relevant for the mere exoplanet interpretation of the radial velocity signal (Anglada-Escudé and Butler 2012). Kepler-78 (G7V) is a 625 My star and harbors an Earth-like planet in an extreme 8.5-hr orbit. Its magnetic properties were investigated in 2015; it features a dipole, quadrupole, and octupole components of similar energy and a mean energy of 16 G. The toroidal field has a 40% contribution and the poloidal field is mainly nonaxisymmetric, as is expected for such a star. The reconstructed map is shown in Fig. 2 (top). At the very short distance where the planet orbits its parent star, it crosses regions of open field lines alternating with closed field lines. In such a configuration, unipolar induction and Ohmic dissipation can occur and may induce

1766

C. Moutou et al.

heating of the planet’s interior as well as orbital decay (Laine and Lin 2012). The coronal field shows when the planet crosses regions where the field lines are open or closed along its orbit, which influences the nature of the interactions with the planetary magnetosphere. Figure 3 shows the coronal configuration inferred from spectropolarimetric observations of Kepler-78 (Moutou et al. 2016). V830 Tau and Tap 26 are two young T Tauri stars, for which the observation of the magnetic field and the detection of the hot Jupiter planetary companion have been concomitant (Donati et al. 2016; Yu et al. 2017, in press). Both young stars were observed in 2014–2016 with ESPaDOnS. At the same time, their magnetic field topology was obtained from Stokes V profiles; the Stokes I profile distortions from surface activity were modeled at the stellar rotation periodicity and removed. The exoplanet signals were then found in the residuals of these profile variations. Both objects are important discoveries, for the inward migration of giant planets can be constrained within timescales of a few million years. V830 Tau features a 340 G dipole field with smaller contributions of smaller-scale regions, higher-level poloidal components, and a weak structured toroidal field (Donati et al. 2017). Its reconstructed map is shown in Fig. 2 (bottom). The faster rotator Tap 26 has a fainter dipole of 120 G and a stronger contribution of the toroidal field than V830 Tau, in agreement with the fact that the star is older (17 Myr), more evolved, and largely radiative (Yu et al. 2017, in press). HD 1237 (G8V) hosts a giant planet in a 134-day orbit and was observed with HARPSpol in 2012. Its magnetic field displays a strong ringlike toroidal structure of 90 G energy , while more complex structures arise at a lower energy level in the radial field (Alvarado-Gómez et al. 2015). Such topologies are also observed in nonplanet-bearing fastly rotating solar analogs (Folsom et al. 2014). Even at a relatively large orbital distance, the mass loss and stellar wind expected from such a magnetic topology may have a significant impact on the orbital and atmospheric evolution of the exoplanet, as shown in Alvarado-Gómez et al. (2016). Finally, it is useful to add the Sun to this sample of planet-host stars with magnetic field characterization, as it is the only star whose surface can be resolved. It is thus possible to relate the behavior of individual magnetic regions to the global properties of the Sun seen as a star. It has been done with HARPS, either by observing the asteroid Vesta with simultaneous observations of the Solar Dynamics Observatory (Haywood et al. 2016) or the Sun directly (Dumusque et al. 2015). These studies have focused primarily on the impact of magnetic regions of the Sun on the radial velocity jitter produced by its inhomogeneous surface, showing that the inhibition of convection is the dominant source of activity-induced velocity variations on solar-type surfaces. Haywood et al. (2016) also shows that these velocity variations best correlate with the Sun full-disc magnetic flux density. Vidotto (2016) models the solar magnetic field in the same formalism and number of spherical harmonics than used in the stellar studies when spatial resolution is unavailable and shows that the Sun field energy is 90% in the poloidal component, with a typical strength of 1 G (for the considered epoch).

86 Magnetic Fields in Planet-Hosting Stars

1767

General Properties Within the 14 exoplanet-host stars whose magnetic topology has been reconstructed at least for one epoch, a wide variety of topology and field strength has been found. All these planet-host stars, in addition to a comparison sample of main-sequence solar-type stars without known exoplanets, are represented in Fig. 4. The magnetic strength and configuration are shown as a function of the position of these stars in the mass-rotation period plane. The shape of the symbol shows the complexity of the field (increasing for non-axisymmetric components and non-poloidal components). It shows that stellar surface fields, for the current sample, are usually more complex than a simple dipole or quadrupole, with the current sample. It is thus necessary to observe and reconstruct the actual magnetic topology, rather than assuming it is a

1.5

1.4 HD 75332

HD 171488

1.1

HD 147513 HD 73256

)

HD 35296 V830 Tau 1.0

AB Dor

V819 Tau HD 73350

Ek Dra Tap 26 LQ Hya

0.9

0.8

2.4e+05

0.6

8.1e+04

0.4

2.7e+04

0.2

9.0e+03

0.0

3.0e+03

HD 76151

HD 141943 HD 179949

1.2

Mass (M

HD 78366

7.3e+05

Sun

HD 190771 HD 29615 Xi Boo A HD 102195

HD 46375

HD 1237 HD 3651 Eps Eri

2 2 < B > (G ) [size]

Tau Boo

Axisymmetry [shape] / Poloidal [color]

1.3

1.0

Kepler-78 0.8 HD 189733 HD 130322 61 Cyg A 0.7

10

1 Prot (d)

Fig. 4 Mass-rotation diagram of 30 reconstructed stellar magnetic fields. Planet-host stars have their names indicated in green (14), while other stars without detected exoplanet have their names indicated in black (16) and are shown for reference. The dashed line represents Rossby number = 1.0 (the Rossby number is defined as the ratio between the rotational period and the convective turnover time). The size of the symbol represents the field strength, its color the contribution of the poloidal component to the field (blue and red for purely toroidal and purely poloidal fields, respectively), and its shape how axisymmetric the poloidal component is (decagons and stars for purely axisymmetric and non-axisymmetric poloidal fields, respectively). When several epochs are available, we show the field for one epoch of observation

1768

C. Moutou et al.

dipole, in models where magnetic fields are used as inputs (e.g., Cohen et al. 2009; Vidotto et al. 2015; Strugarek et al. 2015). Comparison star data come from the literature (e.g., Donati and Landstreet 2009; Waite et al. 2015). Trends in the magnetic properties of planet-host stars are similar to the ones observed for stars without a known planet, although the magnetic strength of the first sample tends to be fainter than the second. It is likely an observational bias due to the fact that indirect planet searches are sensitive to stellar activity and active stars are thus often rejected from radial velocity surveys. Trends with rotation periods are in agreement with the planet-host star sample. For instance, toroidal fields dominate in solar-type stars rotating faster than 12 days as shown by Petit et al. (2008) or in stars with Rossby numbers smaller than 1 (Donati and Landstreet 2009). Multiple observations of the same star, however, may show important variations on the year timescales (e.g., for  Boo, Fares et al. 2009; Mengel et al. 2016). In a global study, See et al. (2015) has studied the origin of the toroidal field in 55 stars having a magnetic map (nine of them being planet-host stars) and found a relationship between the stellar rotation period and a power law between the energy in toroidal and poloidal fields. Even though planet-bearing stars do not represent a sufficient sample to investigate any effect of the presence of the planet on the magnetic morphology, studying these stars brings some new lights on dynamo study in general. Different global properties may yet arise once the samples are large enough. A better knowledge of the magnetic processes in stars is necessary to further explore potential roles of planets. Exploring the magnetic field of planet-host stars has an impact on the knowledge of stars and planets in several ways: either by enabling the planet detection (or hampering it) by estimating the stellar jitter of magnetic origin at different timescales, by quantifying the stellar wind and magnetic flux injected into starplanet interactions, by measuring the rotational period, or by revealing shorter magnetic cycles that could be due to tidal interactions with a massive planet at short distance and a low-mass convective zone. Each characterization of the magnetic topology requires an intensive observing campaign and is sometimes coupled to another observing facility, as X-ray or radio monitoring. Scandariato et al. (2013), for instance, shows the chromospheric and coronal activity monitoring of HD 179949 over time, at a time where spectropolarimetric observations were taking place and further analyzed in Fares et al. (2012). In the course of these investigations, clear and systematic detections of induced activity due to the planet did not occur. If there are hints of those at the orbit period, they are intermittent (Shkolnik et al. 2008); in addition, they should rather be expected at the beat period of the system if due to magnetospheric interactions, i.e., when the configuration between the star, planet, and observer is repeated, rather than the orbital period (Fares et al. 2010). Such signatures of induced activity have been searched in chromospheric emission lines (and thus do not require the more complex use of spectropolarimetry). In order to detect an anomaly in the magnetic topology of the star due to the planet, it would be necessary to disentangle this effect from the intrinsically moving structures of the magnetic field at the surface

86 Magnetic Fields in Planet-Hosting Stars

1769

of the star. We know from solar studies that the surface magnetic field has a strong variability and is structured at small spatial resolution, so the task of separating a planet-induced event on the unresolved surface of a distant star having its own magnetic activity seems challenging. Some extreme cases, however, are promising and should be continuously monitored. Stellar wind models use the observed magnetic topologies to derive mass loss rates and wind density at planet orbit. It allows to set constraints on the environment of planets and ultimately derive how much radio emission is expected from the interactions between the stellar wind and the magnetosphere of the planet (Zarka, this book, and Vidotto et al. 2015). Among the systems with a hot Jupiter planet in the sample of Table 1, the system of V830 Tau is the most promising in terms of radio emission due to star-planet interactions (Vidotto and Donati 2017, in prep). An order of magnitude below comes from the systems of  Boo and HD 189733. While active research focuses on those radio detection, no definite detection has been claimed so far (Lecavelier des Etangs et al. 2013). However, magnetic topology reconstructions have also shown the inhomogeneity of the stellar magnetic field up to the planetary orbit, which makes the interaction-driven radio flux dependent on the stellar longitude and position of the planet, and thus variable in time. Only multiple observations can handle this difficult search, and radio emission should not be counted as a new planet detection technique. Finally, such a radio detection coupled with a thorough characterization of the stellar magnetic field would represent the only way to estimate the magnetic field strength of the planet and as such offer extremely valuable perspectives for the field of starplanet interactions (Hess and Zarka 2011) and exoplanet characterization in general. Planetary magnetic fields are indeed primordial ingredients of the atmospheric escape, surface habitability, and radiative shielding (e.g., See et al. 2014; Vidotto et al. 2013). Although not concerned by habitability questions, the extreme system of Kepler78 with an ultra-short orbital period of 8.5h is a peculiar laboratory for star-planet interactions. Laine and Lin (2012) predicted the creation of hot spots at the footprint of the planet due to Ohmic dissipation. Such hot spots were not observed at a level larger than 10% of the stellar intrinsic variability (Moutou et al. 2016). The properties of the magnetic field of Kepler-78, however, were inferred, and they imply that the theoretical inference of the planet’s conductivity is then comparable to that of partially molten rock. Impacts on the internal structure of the planet are thus expected.

Conclusions and Perspectives Magnetic topologies are key to understand the survival of the planet’s atmosphere, the global evolution of planets, and their habitability, even at orbital distance larger than 0.1 au. It is thus primordial to further explore the magnetic properties of host stars. For instance, low-mass stars have shown a bistability in their magnetic topology: what impact does it have on exoplanets in the habitable zone of these

1770

C. Moutou et al.

stars (closer in than for solar analogs), as a function of their actual topology? Such questions will probably find answers with the survey of low-mass stars planned with the new instrument SPIRou/CFHT (Donati et al., this book). A large population of low-mass main-sequence stars will be observed continuously in search for planets using the radial velocity method. SPIRou being a spectropolarimeter, the magnetic field signature of the parent star will be obtained simultaneously to each radial velocity visit, providing us with the most complete exploration of exoplanet parent hosts, covering several timescales. Another science objective of the SPIRou survey is to probe the early stages of stellar and planetary formation; here again, the observation of young planetary systems will allow combining magnetic property and exoplanet formation studies, on a large scale. Not only will this combination allow a more robust correction for stellar activity jitter but the physical conditions for exoplanet evolution and interactions with the parent star will gain a lot of new observational constraints.

References Alvarado-Gómez JD, Hussain GAJ, Grunhut J et al (2015) Activity and magnetic field structure of the Sun-like planet-hosting star HD 1237. A&A 582:A38. https://doi.org/10.1051/0004-6361/ 201525771, 1507.04117 Alvarado-Gómez JD, Hussain GAJ, Cohen O et al (2016) Simulating the environment around planet-hosting stars. II. Stellar winds and inner astrospheres. A&A 594:A95. https://doi.org/10. 1051/0004-6361/201628988, 1607.08405 Anglada-Escudé G, Butler RP (2012) The HARPS-TERRA Project. I. Description of the algorithms, performance, and new measurements on a few remarkable stars observed by HARPS. ApJS 200:15. https://doi.org/10.1088/0067-0049/200/2/15, 1202.2570 Bourrier V, Lecavelier des Etangs A (2013) 3D model of hydrogen atmospheric escape from HD 209458b and HD 189733b: radiative blow-out and stellar wind interactions. A&A 557:A124. https://doi.org/10.1051/0004-6361/201321551, 1308.0561 Brun AS, Browning MK, Dikpati M, Hotta H, Strugarek A (2015a) Recent advances on solar global magnetism and variability. Space Sci Rev 196:101–136. https://doi.org/10.1007/s11214-0130028-0 Brun AS, García RA, Houdek G, Nandy D, Pinsonneault M (2015b) The solar-stellar connection. Space Sci Rev 196:303–356. https://doi.org/10.1007/s11214-014-0117-8, 1503.06742 Catala C, Donati JF, Shkolnik E, Bohlender D, Alecian E (2007) The magnetic field of the planethosting star  Bootis. MNRAS 374:L42–L46. https://doi.org/10.1111/j.1745-3933.2006.00261. x, astro-ph/0610758 Cébron D, Hollerbach R (2014) Tidally driven dynamos in a rotating sphere. ApJ 789:L25. https:// doi.org/10.1088/2041-8205/789/1/L25, 1406.3431 Cohen O, Drake JJ, Kashyap VL et al (2009) Interactions of the magnetospheres of stars and closein giant planets. ApJ 704:L85–L88. https://doi.org/10.1088/0004-637X/704/2/L85, 0909.3093 Donati JF, Brown SF (1997) Zeeman-Doppler imaging of active stars. V. Sensitivity of maximum entropy magnetic maps to field orientation. A&A 326:1135–1142 Donati JF, Landstreet JD (2009) Magnetic fields of nondegenerate stars. ARA&A 47:333–370. https://doi.org/10.1146/annurev-astro-082708-101833, 0904.1938 Donati JF, Semel M, Rees DE (1992) Circularly polarized spectroscopic observations of RS CVn systems. A&A 265:669–681

86 Magnetic Fields in Planet-Hosting Stars

1771

Donati JF, Collier Cameron A, Semel M et al (2003) Dynamo processes and activity cycles of the active stars AB Doradus, LQ Hydrae and HR 1099. MNRAS 345:1145–1186. https://doi.org/ 10.1046/j.1365-2966.2003.07031.x Donati JF, Catala C, Landstreet JD, Petit P (2006a) ESPaDOnS: the new generation stellar spectropolarimeter. Performances and first results. In: Casini R, Lites BW (eds) Solar polarization 4: Proceedings of the conference held, Boulder, 19–23 Sept 2005. Astronomical Society Pacific conference series, vol 358, p 362 Donati JF, Howarth ID, Jardine MM et al (2006b) The surprising magnetic topology of  Sco: fossil remnant or dynamo output? MNRAS 370:629–644. https://doi.org/10.1111/j.1365-2966. 2006.10558.x, astro-ph/0606156 Donati JF, Moutou C, Farès R et al (2008) Magnetic cycles of the planet-hosting star  Bootis. MNRAS 385:1179–1185. https://doi.org/10.1111/j.1365-2966.2008.12946.x, 0802.1584 Donati JF, Hébrard E, Hussain GAJ et al (2015) Magnetic activity and hot Jupiters of young Suns: the weak-line T Tauri stars V819 Tau and V830 Tau. MNRAS 453:3706–3719. https://doi.org/ 10.1093/mnras/stv1837, 1509.02110 Donati JF, Moutou C, Malo L et al (2016) A hot Jupiter orbiting a 2-million-year-old solar-mass T Tauri star. Nature 534:662–666. https://doi.org/10.1038/nature18305, 1606.06236 Donati JF, Yu L, Moutou C et al (2017) The hot Jupiter of the magnetically active weak-line T Tauri star V830 Tau. MNRAS 465:3343–3360. https://doi.org/10.1093/mnras/stw2904, 1611.02055 Dumusque X, Glenday A, Phillips DF et al (2015) HARPS-N observes the sun as a star. ApJ 814:L21. https://doi.org/10.1088/2041-8205/814/2/L21, 1511.02267 Egeland R, Soon W, Baliunas S et al (2016) The Mount Wilson observatory S-index of the sun. ArXiv e-prints 1611.04540 Fares R, Donati JF, Moutou C et al (2009) Magnetic cycles of the planet-hosting star  Bootis – II. A second magnetic polarity reversal. MNRAS 398:1383–1391. https://doi.org/10.1111/j.13652966.2009.15303.x, 0906.4515 Fares R, Donati JF, Moutou C et al (2010) Searching for star-planet interactions within the magnetosphere of HD189733. MNRAS 406:409–419. https://doi.org/10.1111/j.1365-2966. 2010.16715.x, 1003.6027 Fares R, Donati JF, Moutou C et al (2012) Magnetic field, differential rotation and activity of the hot-Jupiter-hosting star HD 179949. MNRAS 423:1006–1017. https://doi.org/10.1111/j.13652966.2012.20780.x, 1202.4472 Fares R, Moutou C, Donati JF et al (2013) A small survey of the magnetic fields of planet-host stars. MNRAS 435:1451–1462. https://doi.org/10.1093/mnras/stt1386, 1307.6091 Folsom CP, Petit P, Bouvier J, Donati JF, Morin J (2014) The evolution of surface magnetic fields in young solar-type stars. In: Petit P, Jardine M, Spruit HC (eds) Magnetic fields throughout stellar evolution, IAU symposium, vol 302, pp 110–111. https://doi.org/10.1017/S1743921314001835, 1310.2073 Gaulme P, Deheuvels S, Weiss WW et al (2010) HD 46375: seismic and spectropolarimetric analysis of a young Sun hosting a Saturn-like planet. A&A 524:A47. https://doi.org/10.1051/ 0004-6361/201014142, 1011.2671 Haywood RD, Collier Cameron A, Unruh YC et al (2016) The Sun as a planet-host star: proxies from SDO images for HARPS radial-velocity variations. MNRAS 457:3637–3651. https://doi. org/10.1093/mnras/stw187, 1601.05651 Hébrard ÉM, Donati JF, Delfosse X et al (2014) Detecting planets around active stars: impact of magnetic fields on radial velocities and line bisectors. MNRAS 443:2599–2611. https://doi.org/ 10.1093/mnras/stu1285 Hébrard ÉM, Donati JF, Delfosse X et al (2016) Modelling the RV jitter of early-M dwarfs using tomographic imaging. MNRAS 461:1465–1497. https://doi.org/10.1093/mnras/stw1346, 1606.01775 Hess SLG, Zarka P (2011) Modeling the radio signature of the orbital parameters, rotation, and magnetic field of exoplanets. A&A 531:A29. https://doi.org/10.1051/0004-6361/201116510

1772

C. Moutou et al.

Hussain GAJ, Alvarado-Gómez JD, Grunhut J et al (2016) A spectro-polarimetric study of the planet-hosting G dwarf, HD 147513. A&A 585:A77. https://doi.org/10.1051/0004-6361/ 201526595, 1510.02127 Jardine M, Wood K, Collier Cameron A, Donati JF, Mackay DH (2002) Inferring X-ray coronal structures from Zeeman-Doppler images. MNRAS 336:1364–1370. https://doi.org/10.1046/j. 1365-8711.2002.05877.x, astro-ph/0207522 Jeffers SV, Petit P, Marsden SC et al (2014) & Eridani: an active K dwarf and a planet hosting star? The variability of its large-scale magnetic field topology. A&A 569:A79. https://doi.org/ 10.1051/0004-6361/201423725 Kochukhov O (2015) Diagnostic of stellar magnetic fields with cumulative circular polarisation profiles. A&A 580:A39. https://doi.org/10.1051/0004-6361/201526318, 1505.07266 Laine RO, Lin DNC (2012) Interaction of close-in planets with the magnetosphere of their host stars. II. Super-Earths as unipolar inductors and their orbital evolution. ApJ 745:2. https://doi. org/10.1088/0004-637X/745/1/2, 1201.1584 Lammer H, Güdel M, Kulikov Y et al (2012) Variability of solar/stellar activity and magnetic field and its influence on planetary atmosphere evolution. Earth Planets Space 64:179–199. https:// doi.org/10.5047/eps.2011.04.002 Lecavelier des Etangs A, Bourrier V, Wheatley PJ et al (2012) Temporal variations in the evaporating atmosphere of the exoplanet HD 189733b. A&A 543:L4. https://doi.org/10.1051/ 0004-6361/201219363, 1206.6274 Lecavelier des Etangs A, Sirothia SK, Gopal-Krishna Zarka P (2013) Hint of 150 MHz radio emission from the Neptune-mass extrasolar transiting planet HAT-P-11b. A&A 552:A65. https://doi.org/10.1051/0004-6361/201219789, 1302.4612 Lin DNC, Bodenheimer P, Richardson DC (1996) Orbital migration of the planetary companion of 51 Pegasi to its present location. Nature 380:606–607. https://doi.org/10.1038/380606a0 Linsky JL, Schöller M (2015) Observations of strong magnetic fields in nondegenerate stars. Space Sci Rev 191:27–76. https://doi.org/10.1007/s11214-015-0143-1 Marsden SC, Petit P, Jeffers SV et al (2014) A BCool magnetic snapshot survey of solar-type stars. MNRAS 444:3517–3536. https://doi.org/10.1093/mnras/stu1663, 1311.3374 Mathys G (2012) Magnetic fields across the Hertzsprung-Russell diagram. In: Shibahashi H, Takata M Lynas-Gray AE (eds) Progress in solar/stellar physics with helio- and asteroseismology: Proceedings of a Fujihara Seminar held at Hakone, Japan, 13–17 March 2011. ASP conference proceeding, vol 462. Astronomical Society of the Pacific, San Francisco, p 295 Mengel MW, Fares R, Marsden SC et al (2016) The evolving magnetic topology of  Boötis. MNRAS 459:4325–4342. https://doi.org/10.1093/mnras/stw828, 1604.02501 Mengel MW, Marsden SC, Carter BD et al (2017) A BCool survey of the magnetic fields of planet-hosting solar-type stars. MNRAS 465:2734–2747. https://doi.org/10.1093/mnras/ stw2949, 1611.07604 Meunier N, Delfosse X (2011) Filaments and the magnetic configuration. I. Observation of the solar case. A&A 532:A18. https://doi.org/10.1051/0004-6361/201116834 Morin J, Donati JF, Forveille T et al (2008a) The stable magnetic field of the fully convective star V374 Peg. MNRAS 384:77–86. https://doi.org/10.1111/j.1365-2966.2007.12709.x, 0711.1418 Morin J, Donati JF, Petit P et al (2008b) Large-scale magnetic topologies of mid M dwarfs. MNRAS 390:567–581. https://doi.org/10.1111/j.1365-2966.2008.13809.x, 0808.1423 Morin J, Donati JF, Petit P et al (2010) Large-scale magnetic topologies of late M dwarfs. MNRAS 407:2269–2286. https://doi.org/10.1111/j.1365-2966.2010.17101.x, 1005.5552 Morin J, Dormy E, Schrinner M, Donati JF (2011) Weak- and strong-field dynamos: from the Earth to the stars. MNRAS 418:L133–L137. https://doi.org/10.1111/j.1745-3933.2011.01159. x, 1106.4263 Moutou CEA (in prep) A spectro-polarimetric exploration of planet host stars Moutou C, Donati JF, Savalle R et al (2007) Spectropolarimetric observations of the transiting planetary system of the K dwarf HD 189733. A&A 473:651–660. https://doi.org/10.1051/00046361:20077795, 0707.2016

86 Magnetic Fields in Planet-Hosting Stars

1773

Moutou C, Donati JF, Lin D, Laine RO, Hatzes A (2016) The magnetic properties of the star Kepler-78. MNRAS 459:1993–2007. https://doi.org/10.1093/mnras/stw809 Petit P, Dintrans B, Solanki SK et al (2008) Toroidal versus poloidal magnetic fields in Sunlike stars: a rotation threshold. MNRAS 388:80–88. https://doi.org/10.1111/j.1365-2966.2008. 13411.x, 0804.1290 Scandariato G, Maggio A, Lanza AF et al (2013) A coordinated optical and X-ray spectroscopic campaign on HD 179949: searching for planet-induced chromospheric and coronal activity. A&A 552:A7. https://doi.org/10.1051/0004-6361/201219875, 1301.7748 See V, Jardine M, Vidotto AA et al (2014) The effects of stellar winds on the magnetospheres and potential habitability of exoplanets. A&A 570:A99. https://doi.org/10.1051/0004-6361/ 201424323, 1409.1237 See V, Jardine M, Fares R, Donati JF, Moutou C (2015) Time-scales of close-in exoplanet radio emission variability. MNRAS 450:4323–4332. https://doi.org/10.1093/mnras/stv896 Shkolnik E, Walker GAH, Bohlender DA (2003) Evidence for planet-induced chromospheric activity on HD 179949. ApJ 597:1092–1096. https://doi.org/10.1086/378583 Shkolnik E, Bohlender DA, Walker GAH, Collier Cameron A (2008) The on/off nature of starplanet interactions. ApJ 676:628–638. https://doi.org/10.1086/527351, 0712.0004 Snik F, Kochukhov O, Piskunov N et al (2011) The HARPS polarimeter. In: Kuhn JR, Harrington DM, Lin H et al (eds) Solar polarization 6. Astronomical society of the pacific conference series, vol 437, p 237. 1010.0397 Strugarek A, Brun AS, Matt SP, Réville V (2015) Magnetic games between a planet and its host star: the key role of topology. ApJ 815:111. https://doi.org/10.1088/0004-637X/815/2/111, 1511.02837 Tanaka YA, Suzuki TK, Inutsuka SI (2014) Atmospheric escape by magnetically driven wind from gaseous planets. ApJ 792:18. https://doi.org/10.1088/0004-637X/792/1/18, 1311.0972 Vidotto AA (2016) The magnetic field vector of the Sun-as-a-star. MNRAS 459:1533–1542. https://doi.org/10.1093/mnras/stw758, 1603.09226 Vidotto A, Donati JF (2017) Predicting radio emission from the newborn hot Jupiter V830 Tau b. A&A 602:39. doi:10.1051/0004-6361/201629700 Vidotto AA, Fares R, Jardine M et al (2012) The stellar wind cycles and planetary radio emission of the  Boo system. MNRAS 423:3285–3298. https://doi.org/10.1111/j.1365-2966.2012.21122. x, 1204.3843 Vidotto AA, Jardine M, Morin J et al (2013) Effects of M dwarf magnetic fields on potentially habitable planets. A&A 557:A67. https://doi.org/10.1051/0004-6361/201321504, 1306.4789 Vidotto AA, Gregory SG, Jardine M et al (2014) Stellar magnetism: empirical trends with age and rotation. MNRAS 441:2361–2374. https://doi.org/10.1093/mnras/stu728, 1404.2733 Vidotto AA, Fares R, Jardine M, Moutou C, Donati JF (2015) On the environment surrounding close-in exoplanets. MNRAS 449:4117–4130. https://doi.org/10.1093/mnras/stv618, 1503.05711 Waite IA, Marsden SC, Carter BD et al (2015) Magnetic fields on young, moderately rotating Sunlike stars – I. HD 35296 and HD 29615. MNRAS 449:8–24. https://doi.org/10.1093/mnras/ stv006, 1502.05788 Yu L, Donati JF, Hébrard E, Moutou C Malo (2017) A hot Jupiter around the very active weak-line T Tauri star TAP 26. MNRAS 467:1342. doi:10.1093/mnras/stx009

Star-Planet Interactions in the Radio Domain: Prospect for Their Detection

87

Philippe Zarka

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Plasma Flow-Obstacle Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Magnetized Obstacle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Unmagnetized Obstacle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dissipated Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Radio Signatures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Energetics of Radio Signatures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interpreting Radio Signatures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross-References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1776 1776 1776 1778 1782 1783 1783 1785 1787 1788 1789

Abstract

All possible types of interaction of a magnetized plasma flow with an obstacle (magnetized or not) are considered, and those susceptible to produce a radio signature are identified. The role of the sub-Alfvénic or super-Alfvénic character of the flow is discussed. Known examples in the solar system are given, as well as extrapolations to star-planet plasma interactions. The dissipated power and the fraction that goes into radio waves are evaluated in the frame of the radio-magnetic scaling law, the theoretical bases and validity of which are discussed in the light of recent works. Then it is shown how radio signatures can be interpreted, in the frame of the cyclotron-maser theory (developed for

P. Zarka () LESIA, Observatoire de Paris, CNRS, PSL, UPMC/SU, UPD, Meudon, France Station de Radioastronomie de Nançay, Observatoire de Paris, CNRS, PSL, Univ. Orléans, Nançay, France e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_22

1775

1776

P. Zarka

explaining the generation of solar system planetary auroral and satellite-induced radio emissions), for deducing many physical parameters of the system studied, including the planetary or stellar magnetic field. Prospects for the detection of such radio signatures with new generation low-frequency radiotelescopes are then outlined.

Introduction Stars interact in many ways with their planets in orbit: through gravitation that constrains the orbit and, at short range – i.e., for close in exoplanets – causes tidal effects of each body upon the other (Cuntz et al. 2000); through stellar light – especially at short wavelengths – that ionizes the planet’s upper atmosphere (e.g., Encrenaz et al. 2004); and through plasma and magnetic fields. Here we are mainly interested with the latter that may lead to the generation of intense radio emissions. We use the generic name “plasma interaction” for interactions involving plasma flows and magnetic fields.

Plasma Flow-Obstacle Interactions Various types of plasma interactions are observed in our solar system, involving the solar wind and magnetized or unmagnetized planets, as well as magnetized planets and their satellites. A coherent frame for sorting these various interactions is the general frame of flow-obstacle interactions (Zarka 2007). The former is of course a flow of plasma that can be strongly or weakly magnetized (the hypothetical case of a completely unmagnetized flow is not interesting for us because no radio signature is expected). The latter is a conductive or insulating body, with or without an atmosphere, and possessing or not an intrinsic magnetic field (Lepping 1986).

Magnetized Obstacle When the obstacle is magnetized, its interaction with the magnetized flow is likely to occur through magnetic reconnection at their interface (where both magnetic field amplitudes are comparable but their orientations different), leading to energy release in the form of plasma waves and particle acceleration susceptible to produce radio waves. The location in the system where radio emission may occur depends on the region of space accessible to the accelerated electrons along magnetic field lines. This in turn depends on the Mach number (M D Vflow /Vsound ) and especially the Alfvénic Mach number (MA D Vflow /VAlfvén ) of the flow in the obstacle’s frame that will define the topology of the interaction (Alfvén waves carry magnetic perturbations and associated electric currents along magnetic field lines).

87 Star-Planet Interactions in the Radio Domain: Prospect for Their Detection

1777

When the flow is super-Alfvénic (MA > 1), a bow shock forms upstream of the obstacle that slows down and heats the flow, making it become sub-Alfvénic (MA < 1) before it hits the obstacle and flows around it. This shock somewhat “disconnects” the obstacle from the source of the flow and thus prevents most of the waves, particles, and energy to travel back to this source. Accelerated electrons will rather focus toward the magnetic poles of the obstacle, where they collectively acquire non-Maxwellian distributions susceptible to produce intense radio waves. This is what happens in the auroral regions of the magnetospheres of magnetized planets in the solar wind (the large-scale cavities carved in the solar wind by planetary magnetic fields (Bagenal 2001; Encrenaz et al. 2004)), and it should also result from the interaction of stellar winds with magnetized exoplanets, if they are in the super-Alfvénic region of the wind. When the flow is sub-Alfvénic the perturbations of magnetic field lines (Alfvén waves) excited by reconnection at the flow-obstacle interface can propagate along magnetic field lines toward the source of the flow as well as toward the magnetic poles of the obstacle. These Alfvén waves propagate in two “wings”, symmetrically (or oppositely) oriented relative to the direction of the flow in the plane that contains the flow velocity Vflow and magnetic field Bflow if Vflow and Bflow are perpendicular. Currents are carried by these wings that may also lead to electron acceleration to keV energy or more. The angle between these Alfvén wings and the flow is defined by the ratio between VA along magnetic field lines and Vflow across them. In the limit where VA > > Vflow , Alfvén wings are simply the magnetic flux tube connecting the obstacle to the source of the flow. If the flow is itself strongly magnetized, the conditions (accelerated electron distributions, plasma density, and magnetic field amplitude) at the footprints of the perturbed field lines can be favorable for the production of radio emission. In the solar system, this does not happen in the solar wind, which is weakly magnetized and super-Alfvénic at all planetary orbits, but it does happen in the interaction of the magnetized satellite Ganymede with the rotating magnetic field of its parent planet Jupiter (Kivelson et al. 2004), which leads to magnetic reconnection, electron acceleration, and radio emission generation near the footprints in Jupiter’s ionosphere of the Ganymede flux tube (and also marginally near Ganymede). A similar interaction should exist between magnetized hot jupiters, orbiting close enough to their parent star to be in the region where the wind is still sub-Alfvénic. If the magnetic field amplitude at the stellar surface is 30–100 times that of the Sun, radio emission should be produced not only in the exoplanet’s auroral regions but also near the footprints on the stellar surface of the magnetic field lines connecting the hot jupiter and the star (Zarka 2006). Note that while both ends of Ganymede’s flux tube are connected to Jupiter’s ionosphere, at most one Alfvén wing of a hot jupiter may travel upstream to the stellar surface except if the planet lies in the closed magnetic field lines region (Preusse 2006). A particular case of the interaction between an exoplanet and a strongly magnetized wind concerns planets orbiting pulsars. Pulsar winds (beyond the light

1778

P. Zarka

cylinder) are expected to be relativistic, so that both Alfvén wings of a planetary obstacle are carried away by the flow, independent of the magnetization of the obstacle (Mottez and Heyvaerts 2011). It has been proposed that the plasma conditions in these wings are favorable to radio emission generation, and that relativistic focussing of the produced emission can explain the elusive fast radio bursts detected at cosmological distances (Mottez and Zarka 2014). This also applies when the obstacle in the pulsar wind is not magnetized, but only conductive (see below).

Unmagnetized Obstacle When the obstacle is not magnetized, its interaction with the plasma flow will again depend on the flow magnetization and Alfvén Mach number, and on the obstacle’s conductivity. In the weakly magnetized super-Alfvénic solar wind, conductive obstacles will develop an induced magnetosphere via the (temporary) pile-up of the wind’s magnetic field on the obstacle’s nose, caused by the slow large-scale perpendicular diffusion of plasma across magnetic fields. Conductivity is high in the ionospheres of Venus, Mars, Titan or comets close to the Sun, and possibly in the interior for metallic asteroids. Similar induced magnetospheres are expected to exist for unmagnetized exoplanets orbiting within a weakly magnetized super-Alfvénic stellar wind. The absence of a large-scale magnetic field of the obstacle prevents the focussing of accelerated electrons toward auroral regions and thus causes the absence of any intense radio emission for these systems. When the flow is sub-Alfvénic, Alfvén wings similar to those described above develop in the flow past the obstacle. The difference with the previous case is that the Alfvénic perturbations are not caused by magnetic reconnection but by the deviation of the flow (and its embedded magnetic field) by the conductive obstacle. Because they are only driven by the relative motion between the obstacle and the flow, these Alfvén wings are said to result from a “Unipolar Inductor” (UI) interaction (the theory was initially developed for artificial satellites in the ionosphere (Drell et al. 1965)). The first and most famous example of a natural UI interaction is that of Io with the strong rotating Jovian magnetic field (Neubauer 1980), which is now known to apply also to the Europa-Jupiter interaction. The conductivity of the obstacle is ensured by the existence of a volcanic atmosphere (and thus an ionosphere) around Io, and a subsurface ocean at Europa. Electrons accelerated in the Alfvén wings follow Jovian magnetic field lines down to the Jovian ionosphere. Above the northern and southern magnetic footprints of these two Galilean moons, the plasma conditions are favorable for the production of intense radio emission (Bigg 1964; Louis et al. 2017). Similar interactions may exist between some satellites of Saturn and the planet’s magnetic field, but because Saturn’s magnetic field is much weaker

87 Star-Planet Interactions in the Radio Domain: Prospect for Their Detection

1779

than Jupiter’s, the UI energy budget (see below) is not expected to give rise to strong satellite-induced radio emissions at Saturn, and indeed none has been observed until now. Unipolar induction is expected to take place in star-planet systems when the stellar wind is sub-Alfvénic (which should be the case for many hot jupiters) and the planet unmagnetized. If the star is weakly (Sun-like) magnetized, then no radio emission is expected similar to the Saturn-satellite case. If the star is strongly magnetized ( 30–100 times the Sun), then the star-planet system is a giant analogue of the Io-Jupiter (or Europa-Jupiter) system and strong radio emission is expected to be generated near one or both footprints of the magnetic flux tube connecting the exoplanet to the stellar surface (Zarka 2006). Willes and Wu (2004, 2005) studied the particular case of terrestrial planets in close orbits around magnetic white dwarf stars, acting as a unipolar inductor, and generating radio emission with large flux densities. These systems could be remnants of main sequence stars with a planet that has survived the stellar expansion phase and has settled again on a stable orbit. As mentioned above, the case of a conductive obstacle (planet, asteroid, dwarf companion) orbiting in a pulsar wind is the relativistic analogue of the Io-Jupiter interaction, for which Mottez and Heyvaerts (2011) have shown that although close to c, Vflow is still likely to be lower than VA , thus the interaction is expected to be a UI one. It is tempting to extend the above conclusions to the case of a strongly magnetized super-Alfvénic stellar wind flowing past an unmagnetized planet: in that case, the UI interaction produces an induced magnetosphere, the envelope of which (the magnetopause) can be considered as the limiting case of the planet’s Alfvén wings in a super-Alfvénic flow. It is remarkable that the transition between a super-Alfvénic and a sub-Alfvénic flow has been actually observed and simulated at Earth (Chané et al. 2012, 2015). In the super-Alfvénic case, the accelerated electrons cannot focus toward the planet’s auroral regions and they are not expected either to propagate upstream to the distant stellar surface, but it is not excluded that a mechanism similar to the one proposed by Mottez and Zarka (2014) takes place, i.e., an instability might develop along the magnetopause and produce radio emissions. This is speculative though, as no example of such an interaction exists in our solar system. Finally, let us remind that when the obstacle is not only unmagnetized but also insulating, it only absorbs the flow that impacts it, creating a plasma cavity in its wake, which is progressively refilled due to charged particle motion along the magnetic field lines permeating the flow. This is the case for the Earth’s Moon (and probably also some rocky asteroids) in the solar wind. No bow shock nor any radio emission is produced in that configuration. Table 1 expands and generalizes the Table 1 from Zarka (2007), attempting to list all flow – obstacle interactions described above in a synthetic way, extrapolate them to star-planet interactions and predict their radio signatures.

Flow MA b

"

#

"

#

"

#

Flow jBja

jBj #

jBj #

jBj #

jBj #

jBj "

jBj "

Pulsar – planet (reconnection)

Obstacle Known examples in the solar system (mechanism) Expected star-planet interaction (mechanism) B Solar wind – planet: Mercury, Earth, Jupiter, Saturn, Uranus, Neptune ➔ magnetospheres (reconnection) Stellar wind – magnetized exoplanet (reconnection) B – Stellar wind – magnetized hot jupiter (reconnection) No B Solar wind – conductive obstacle: Venus, Mars, Titan, Asteroids [metallic], comets [close to the Sun] (induced magnetosphere) Solar wind – insulating obstacle: Moon, Asteroids [rocky] (flow absorption, no shock, wake) Stellar wind – unmagnetized exoplanet (induced magnetosphere) No B Saturn – satellites (AW/UI)d Stellar wind – unmagnetized hot jupiter (AW/UI) B – Magnetized star – magnetized exoplanet (reconnection) B Jupiter – Ganymede (reconnection) Magnetized star – magnetized hot jupiter (reconnection)

– – – – From exoplanet’s aurora At Ganymede flux tube footprints From exoplanet’s aurora and/or at star-planet flux tube footprints In Alfvén wing ➔ fast radio bursts?

–

Observed CMIc radio signature Expected CMI radio signature From planet’s aurora (Earth, Jupiter, Saturn, Uranus, Neptune) From exoplanet’s aurora – From exoplanet’s aurora –

Table 1 Possible types of flow-obstacle interactions, examples, radio signatures, and extrapolation to star-planet plasma interactions

1780 P. Zarka

B No B

No B

No B

– Magnetized star – unmagnetized exoplanet (AW/UI ! induced magnetosphere) Jupiter – Io, Europa (AW/UI) Magnetized star – unmagnetized hot jupiter (AW/UI) Pulsar – planet (AW/UI) Particle precipitations into cusp (low energy) Plasma wake

b

Flow magnetic field jBj: " D strong, # D weak (the solar wind is weakly magnetized). MA : " D super-Alfvénic, # D sub-Alfvénic c Cyclotron maser instability d AW Alfvén wings, UI unipolar inductor

a

#

jBj "

No B No B

"

jBj "

At Io and Europa – flux tube footprints At star-planet flux tube footprints In Alfvén wing ➔ fast radio bursts? – –

– Along magnetopause?

87 Star-Planet Interactions in the Radio Domain: Prospect for Their Detection 1781

1782

P. Zarka

Dissipated Power Zarka (2007) proposed that the electromagnetic power Pd dissipated in all known flow-obstacle interactions can be expressed as:   Pd D © Vflow B? 2 = o  Robs 2

(1)

with B? the flow’s magnetic field component perpendicular to the flow direction in the obstacle’s frame. Equation (1) is thus simply the fraction © of the Poynting flux (or magnetic energy flux) Vflow B? 2 / o convected on the obstacle of cross section  Robs 2 . This expression, and the variation as a function of the distance to the Sun of B? in the frame of a planet in Keplerian orbit, allowed Zarka et al. (2001) and Zarka (2007) to identify a corridor at 0.17 AU where the flow and the magnetic field are nearly aligned with each other and thus no or very little Poynting flux is generated in the interaction. This corridor should exist in all stellar winds near 0.1–0.3 AU (Saur et al. 2013). Theoretical considerations suggest that the efficiency of the dissipation © is comprised between MA (in the sub-Alfvénic case) and 1 (Zarka et al. 2001; Zarka 2007). Measured efficiencies for the Earth’s magnetosphere and satelliteJupiter interactions are © D 0:2 ˙ 0:1. Electromagnetic signatures of flow-obstacle interactions cannot exceed a small fraction of Pd , estimated to 5–25% in the UV and 1–5% in the radio domain (see below and Zarka 2007). As a consequence, observed signatures can help to put constraints on the power dissipated in an interaction. Zarka (2006, 2007) noted for example that in the case of HD 179949, the system for which the first optical evidence of a star-planet interaction was detected by Shkolnik et al. (2003, 2004), the demands put on the dissipated power by the energetics of the optical signature suggest a stellar magnetic field 30–100 times stronger than the solar one, a wind strength much larger than the Sun’s, an obstacle size much larger than  RJ 2 (extended magnetosphere or exoionosphere), or a combination of these factors. This conclusion is supported by Saur et al. (2013), who computed analytically the magnetic energy fluxes dissipated in sub-Alfvénic interactions. These authors reach consistent – albeit more detailed – conclusions as those drawn from Eq. (1), with dissipated energy fluxes 1019 W in star-planet interactions (to be compared with 1012 W per hemisphere in the Io-Jupiter case). These conclusions must be compared to the measurement of the stellar magnetic field by Fares et al. (2012) giving a value of a few Gauss. The topology of sub-Alfvénic star-planet magnetic interactions has been studied via MHD simulation, focussing on close-in exoplanets (Ip et al. 2004; Strugarek et al. 2014, 2015). For the super-Alfvénic case, dissipated magnetic energy fluxes have been computed for the solar wind-Mercury interaction by Varela et al. (2016) in order to make predictions on radio emissions from Mercury.

87 Star-Planet Interactions in the Radio Domain: Prospect for Their Detection

1783

Radio Signatures From Table 1 it appears clearly that the existence of a strong magnetic field in the flow, the obstacle, or both, together with that of accelerated electrons, leads to the generation of intense radio emission from the system. This is systematically verified in the solar system. This can be understood considering that the most efficient and ubiquitous mechanism for generating magnetospheric (auroral and satelliteplanet) radio emissions is the Cyclotron Maser Instability (CMI – Zarka 1998). The CMI is a wave-particle instability that directly converts the perpendicular energy of electrons gyrating in a large-scale magnetic field into electromagnetic (radio) energy (Wu and Lee 1979). It has been demonstrated that up to a few percent of the electrons’ energy can be converted into radio waves, if three conditions are met: the presence of a strong magnetic field, of accelerated electrons with keV (or tens of keV) energies, and a relatively depleted and strongly magnetized plasma at the source (quantified by fpe /fce 1 in the solar corona except at specific locations, e.g., at footprints of strong magnetic loops – Morosan et al. 2016).

Energetics of Radio Signatures In order to predict emitted radio powers or flux densities, it is necessary to estimate the fraction of the dissipated power that goes into electron acceleration. This was first done via comparison, in the solar system, of emitted radio powers to electromagnetic power involved in the corresponding flow-obstacle interactions (Zarka et al. 2001; Zarka 2007). It was found that in all cases the emitted radio power Pr follows the relation:   Pr D “ Vflow B? 2 = o  Robs 2 D “ Pd =©

(2)

with an efficiency factor “ D 2–10  103 . Taking the conservative values “ D 2  103 and © D 0:2, this implies that a large fraction of the dissipated energy goes into electron acceleration, and 1% of the electrons’ energy then goes into radio waves (this number 1% is the canonical value used in all theoretical predictions).

1784

P. Zarka

Fig. 1 Radio-magnetic scaling law relating magnetospheric (E, J, S, U, N D Earth, Jupiter, Saturn, Uranus, and Neptune) and satellite-induced (I, G, C D Io, Ganymede, Callisto) radio power to incident Poynting flux of the plasma flow on the obstacle. Dashed line has slope 1, emphasizing the proportionality between ordinates and abscissae, with a coefficient 2  103 . The thick bar extrapolates to hot jupiters the magnetospheric interaction (solid) and satelliteplanet electrodynamic interactions (dashed). The orange dot illustrates the case of the RS CVn magnetic binary V711 Tauri discussed in the text. Inset sketch the types of interaction (solar windmagnetosphere, Jupiter-Io, Jupiter-Ganymede)

This radio-magnetic scaling law was found to apply to all solar system auroral or satellite-Jupiter radio emissions. It is illustrated in Fig. 1. Several theoretical works attempted to explore the causes and efficiency of the conversion of the incident magnetic energy flux into the energy of accelerated electrons. Jardine and Cameron (2008) proposed that runaway electrons are accelerated by the electric field resulting from magnetic reconnection at the flow-obstacle interaction, but this does not seem to be a significant process for generating solar system radio emissions. Nichols (2011) and Nichols and Milan (2016) studied the magnetosphere-ionosphere coupling in Jupiter-like and Earth-like magnetospheres. Magnetosphere-ionosphere coupling can be qualitatively understood as a limiting case of flow-obstacle interaction where the obstacle is the inertia of the plasma that resists displacement along with planetary magnetic field lines. As a consequence, magnetic field-aligned-currents arise to communicate the torque between the magnetosphere and the ionosphere. These currents are generally larger than that which can be carried by unaccelerated magnetospheric electrons, so that fieldaligned potential drops appear, that cause electron acceleration to keV energies, and ultimately radio emissions. These processes are suspected to play a role in the

87 Star-Planet Interactions in the Radio Domain: Prospect for Their Detection

1785

generation of radio emission from cool dwarf stars, via the coupling of the stellar magnetic field with its plasma environment (Hallinan et al. 2015). For Jupiter-like, rotation driven dynamics, Nichols (2011) found that the radio output can reach 1016 W for fast rotating magnetospheres of strongly magnetized planets orbiting at a few AU from their parent star, with a strong stellar X-UV luminosity being a favorable factor maximizing the magnetosphere-ionosphere coupling. For Earth-like, convection driven dynamics (magnetospheric convection results from dayside and nightside reconnections between the planetary magnetic field and the solar wind one), Nichols and Milan (2016) found that the radio output can reach 1015 W for close in exoplanets, somewhat smaller than that predicted by the radio-magnetic scaling law due to possible saturation of the convection that prevents to dissipate the full available incident Poynting flux. To test the extrapolation of the radio-magnetic scaling law toward high Poynting fluxes and high radio powers, Zarka (2010) analyzed the literature on radio emission from magnetized binary stars. They found that for the RS CVn stellar system V711 Tauri, measured radio flux densities (0.1–1 Jy – Budding et al. 1998; Richards et al. 2003) and distance (29 pc), and estimated magnetic fields (10–30 G at the interaction region) and bandwidth ( 8 GHz) allow to estimate that 104  “  102 (see the detailed calculation in (Mottez and Zarka 2014)). This good agreement with the solar system value of “ ( 2  103 ) suggests that the radio-magnetic scaling law holds – at least approximately – for 10 orders of magnitude above the range of solar system planets.

Interpreting Radio Signatures CMI radio emissions are expected to occur in the spectral range below a few tens of MHz, unless exoplanets much more strongly magnetized than Jupiter exist. For close-in exoplanets, that are probably spin-orbit locked, the relatively long rotation period (equal to the orbital period) may lead to a decay of the planetary dynamo and thus of the magnetic field (Sanchez-Lavega 2004; Reiners and Christensen 2010). But some of the star-planet interactions described in Table 1 may lead to radio emissions at star-planet flux tube footprints, i.e., at cyclotron frequencies governed by the stellar magnetic field. These emissions might reach hundreds of MHz or more. At these frequencies, the angular resolution of a few milli-arcseconds that would be necessary to separate the radio emission of an exoplanet from that of its parent star will not be available in the foreseeable future. Emission will thus be detected in dynamic spectra, i.e., measurements of the intensity and polarization as a function of time and frequency. With these data, discrimination between the stellar (coronal) and exoplanetary or exoplanet-induced CMI radio emissions can be done via the presence of circular polarization and the modulation at the exoplanet’s orbital period. As the CMI theory provides today a well-understood, quantitative framework for understanding radio emissions properties, the measurement of the radio intensity

1786

P. Zarka

a

b

Frequency

Frequency

Mag. moment

c

Time Revolution period Type of interaction Mag. moment

e

Polarization

f

Frequency

Intensity

Frequency

d

Time Revolution period Orbit inclination Type of interaction Mag. moment

h

Polarization

Frequency

Intensity

Frequency

g

i

Time

Revolution period

Rotation period

Dipole tilt

Orbit inclination Type of interaction

Fig. 2 Simulated dynamic spectra in intensity (a), (b), (d), (g) and circular polarization (e), (h) and associated parameters of the system that can be determined: (a), (b), (c) exoplanetary magnetic field aligned with the rotation axis and 0ı orbit inclination, for emission (a) of a full exoplanet’s auroral oval and (b) an auroral active sector fixed in local time. (d), (e), (f) aligned exoplanetary magnetic field and 15ı orbit inclination, for emission of a full auroral oval. (g), (h), (i) exoplanetary magnetic field tilted by 15ı and 0ı orbit inclination, for emission of a full auroral oval (From Hess and Zarka (2011))

87 Star-Planet Interactions in the Radio Domain: Prospect for Their Detection

1787

and polarization as a function of time and frequency will provide powerful diagnostics of the plasma environments and processes at work in star-planet interactions. Simulations by Hess and Zarka (2011), well tested on Jupiter’s radio emissions (Hess et al. 2008), have shown that such measurements will give access to the type of interaction at work (exoplanet’s auroral emission or exoplanet-induced emission in the stellar field), its energetics, the exoplanet’s magnetic field intensity and tilt (if emission comes from the exoplanet’s magnetosphere), the planetary rotation and revolution periods (effectively testing tidal spin-orbit synchronization), and the orbit inclination (resolving the ambiguity on the planet’s mass). Figure 2 illustrates how these parameters can be deduced from the observed dynamic spectra. Temporal modulations different from the planetary rotation and orbital periods could additionally reveal the presence of moons or the signature of the stellar wind activity. Most of these parameters, especially those concerning the exoplanet’s magnetic field and thus its interior structure, are very difficult or impossible to determine by other means than radio measurements. It can be noted that the existence of an exoplanet’s magnetosphere is an essential ingredient in favor of the possibility to develop life: it ensures shielding of the planet’s atmosphere and surface, preventing O3 destruction by cosmic rays bombardment as well as atmospheric erosion by the stellar wind or CMEs; it also limits atmospheric escape, as ionized material following dipolar field lines returns to the atmosphere (Grie“meier et al. 2004, 2005). Radio detection may eventually evolve as an independent discovery tool, complementary to radial velocities or transit measurements because it is more adapted to finding planets (giant or terrestrial) around and interacting with active, magnetic or variable stars.

Conclusion Star-planet plasma interactions are expected to be common and diverse examples of flow-obstacle interactions. The proposed radio-magnetic scaling law summarized by Eqs. (1) and (2) is likely to be an approximation, and various works suggest that deviations of one order of magnitude (up to two in extreme cases) are expected (Nichols 2011; Nichols and Milan 2016; Saur et al. 2013). Conversely, the recent detection and study of the energetics of the Ganymede-Jupiter radio emission is fully compatible with this scaling law (Zarka et al. 2017). And at least one case of magnetized binary stars interaction suggests that the scaling law still applies over 10 orders of magnitude above the range of solar system planets, and can thus be used for predicting radio flux densities from exoplanets and star-planet interacting systems. As a result, radio emissions up to 105–6 times more intense than Jupiter’s may exist, especially in hot jupiter systems. Even if hot jupiters are weakly or not magnetized due to slow sidereal rotation, the possibility of an interaction with a strongly magnetized parent star via Alfvén waves (Alfvén Wings or UI interaction) offers serious perspectives for radio detection, at frequencies up to hundreds of MHz, larger than those expected for magnetospheric emissions (tens of MHz). Clearly the measurement of stellar magnetic fields via Zeeman Doppler

1788

P. Zarka

spectropolarimetry (e.g., Donati et al. 2006) is a crucial information for selecting candidates for radio emission. Radio emissions 103–4 times stronger than Jupiter’s should be detectable at stellar distances (pc to tens of pc) with existing large low-frequency radiotelescopes (UTR-2 in Ukraine, LOFAR in Europe, NenuFAR in France), and SKA will make detection possible for emissions 101–2 times stronger than Jupiter’s (Zarka et al. 2015). SKA will perform large surveys, and the observing parameters adapted to the radio detection of exoplanets and star-planet interactions are also adapted to the study of stellar coronal bursts or CMI emission from cool stars. Based on the spatial in situ exploration of planetary magnetospheres in our solar system, a reliable interpretation frame is ready (Hess and Zarka 2011), waiting for the first unambiguous detection, that has not occurred yet although tentative detections have been made (see Zarka et al. 2015 and references therein). Unless most interacting star-exoplanet systems are on the (very) low side of the radiomagnetic scaling law, radio detections should occur in the coming years. Those will open the new field of comparative exo-magnetospheric physics and allow to probe flow-obstacle interactions in various regimes and over a large range of parameters (star-planet distance, stellar magnetic field and wind strength, stellar X-UV flux, planetary magnetic field, rotation, orbit inclination, etc.). Measured radio fluxes should allow us to infer the energy dissipation involved in the corresponding star-planet interactions, and thus to confirm or refine the radio-magnetic scaling law.

Cross-References  Electromagnetic Coupling in Star-Planet Systems  Future Exoplanet Research: Radio Detection and Characterization  Models of Star-Planet Magnetic Interaction  Magnetic Fields in Planet-Hosting Stars  Models of Star-Planet Magnetic Interaction  Planet and Star Interactions: Introduction  Planetary Interiors, Magnetic Fields, and Habitability  Radio Emission from Ultracool Dwarfs  Radio Observations as an Exoplanet Discovery Method  Signatures of Star-Planet Interactions  Stellar Coronal and Wind Models: Impact on Exoplanets  The Impact of Stellar Activity on the Detection and Characterization of Exoplan-

ets  Tidal Star-Planet Interactions: A Stellar and Planetary Perspective Acknowledgments PZ acknowledges funding from the programs PNP, PNST, PNPS, and AS SKA-LOFAR of CNRS/INSU.

87 Star-Planet Interactions in the Radio Domain: Prospect for Their Detection

1789

References Bagenal F (2001) Planetary magnetospheres. In: Murdin P (ed) Encyclopedia of astronomy and astrophysics. IOP Publishing, Bristol. article 2329 Bigg EK (1964) Influence of the satellite Io on Jupiter’s decametric emission. Nature 203:1008– 1010 Budding E, Slee OB, Jones K (1998) Further discussion of binary star radio survey data. PASA 15:183–188 Chané E, Saur J, Neubauer FM et al (2012) Observational evidence of Alfvén wings at the Earth. J Geophys Res 117:A09217 Chané E, Raeder J, Saur J et al (2015) Simulations of the Earth’s magnetosphere embedded in sub-Alfvénic solar wind on 24 and 25 May 2002. J Geophys Res 120:8517–8528 Cuntz M, Saar SH, Muzeliak ZE (2000) On stellar activity enhancement due to interactions with extrasolar giant planets. Astrophys J 533:L151–L154 Donati JF, Howarth ID, Bouret JC (2006) Discovery of a strong magnetic field on the O star HD 191612: new clues to the future of ™1 Orionis C*. MNRAS Lett 365(1):L6–L10 Drell SD, Foley HM, Ruderman MA (1965) Drag and propulsion of large satellites in the ionosphere: an Alfvén propulsion engine in space. J Geophys Res 70(13):3131–3145 Encrenaz T, Bibring JP, Blanc M et al (2004) The solar system, 3rd edn. A&A Library, Springer, Berlin. http://www.springer.com/in/book/9783540002413 Fares R, Donati JF, Moutou C et al (2012) Magnetic field, differential rotation and activity of the hot-jupiter-hosting star HD 179949. MNRAS 423:1006–1017 Grie“meier JM, Stadelmann A, Penz T et al (2004) The effect of tidal locking on the magnetospheric and atmospheric evolution of “hot jupiters”. Astron Astrophys 425:753–762 Grie“meier JM, Motschmann U, Mann G, Rucker HO (2005) The influence of stellar wind conditions on the detectability of planetary radio emissions. Astron Astrophys 437:717–726 Hallinan G, Littlefair SP, Cotter G et al (2015) Magnetospherically driven optical and radio aurorae at the end of the stellar main sequence. Nature 523(7562):568–571 Hess SLG, Zarka P (2011) Modeling the radio signature of the orbital parameters, rotation, and magnetic field of exoplanets. Astron Astrophys 531:A29 Hess S, Cecconi B, Zarka P (2008) Modeling of Io-Jupiter decameter arcs, emission beaming and energy source. Geophys Res Lett 35:L13107 Ip WH, Kopp A, Hu JH (2004) On the star–magnetosphere interaction of close-in exoplanets. Astrophys J 602:L53–L56 Jardine M, Cameron AC (2008) Radio emission from exoplanets: the role of the stellar coronal density and magnetic field strength. Astron Astrophys 490:843–851 Kivelson MG, Bagenal F, Kurth WS et al (2004) Magnetospheric interactions with satellites. In: Bagenal F, McKinnon W, Dowling T (eds) Jupiter: the planet, satellites, and magnetosphere. Cambridge University Press, Cambridge, pp 513–536 Lepping RP (1986) Magnetic configuration of planetary obstacles. In: Comparative study of magnetospheric systems. Cepadues/CNES ed, Toulouse, pp 45–75 Louis CK, Lamy L, Zarka P, Cecconi B, Hess SLG (2017) Detection of Jupiter decametric emissions controlled by Europa and Ganymede with Voyager/PRA and Cassini/RPWS. J Geophys Res (in press) Morosan DE, Gallagher PT, Zucca P et al (2016) LOFAR tied-array imaging and spectroscopy of solar S bursts. Astron Astrophys 580:A65 Mottez F, Heyvaerts J (2011) Magnetic coupling of planets and small bodies with a pulsar wind. Astron Astrophys 532:A21 Mottez F, Zarka P (2014) Radio emissions from pulsar companions: a refutable explanation for galactic transients and fast radio bursts. Astron Astrophys 569:A86 Neubauer FM (1980) Nonlinear standing Alfvén wave current system at Io: theory. J Geophys Res 85(A3):1171–1178

1790

P. Zarka

Nichols JD (2011) Magnetosphere–ionosphere coupling at Jupiter-like exoplanets with internal plasma sources: implications for detectability of auroral radio emissions. MNRAS 414:2125– 2138 Nichols JD, Milan SE (2016) Stellar wind–magnetosphere interaction at exoplanets: computations of auroral radio powers. MNRAS 461:2353–2366 Preusse S, Kopp A, Büchner J, Motschmann U (2006) A magnetic communication scenario for hot jupiters. Astron Astrophys 460:317–322 Reiners A, Christensen UR (2010) A magnetic field evolution scenario for brown dwarfs and giant planets. Astron Astrophys 522:A13 Richards MT, Waltman EB, Ghigo FD, Richards DSP (2003) Statistical analysis of 5 year continuous radio flare data from “ Persei, V711 Tauri, • Librae, and Ux Arietis. Astrophys J Suppl Ser 147:337–361 Sanchez-Lavega A (2004) The magnetic field in giant extrasolar planets. Astrophys J 609:L87–L90 Saur J, Grambusch T, Duling S, Neubauer FM, Simon S (2013) Magnetic energy fluxes in subAlfvénic planet star and moon planet interactions. Astron Astrophys 552:A119 Shkolnik E, Walker GAH, Bohlender DA (2003) Evidence for planet-induced chromospheric activity on HD 179949. Astrophys J 597:1092–1096 Shkolnik E, Walker GAH, Bohlender DA (2004) Erratum: evidence for planet-induced chromospheric activity on HD 179949. Astrophys J 609:1197 Strugarek A, Brun AS, Matt SP, Reville V (2014) On the diversity of magnetic interactions in close-in star-planet systems. Astrophys J 795(1):86 Strugarek A, Brun AS, Matt SP, Reville V (2015) Magnetic games between a planet and its host star: the key role of topology. Astrophys J 815(2):111 Varela J, Reville V, Brun AS, Pantellini F, Zarka P (2016) Radio emission in Mercury magnetosphere. Astron Astrophys 595:A69 Willes AJ, Wu K (2004) Electron-cyclotron maser emission from white dwarf pairs and white dwarf planetary systems. MNRAS 348:285–296 Willes AJ, Wu K (2005) Radio emissions from terrestrial planets around white dwarfs. Astron Astrophys 432:1091–1100 Wu CS, Lee LC (1979) A theory of the terrestrial kilometric radiation. Astrophys J 230:621–626 Zarka P (1998) Auroral radio emissions at the outer planets: observations and theories. J Geophys Res 103:20159–20194 Zarka P (2006) Hot jupiters and magnetized stars: giant analogs of the satellite-jupiter system? In: Rucker HO, Kurth WS, Mann G (eds) Planetary radio emissions VI. Austrian Academy of Science Press, Vienna, pp 543–569 Zarka P (2007) Plasma interactions of exoplanets with their parent star and associated radio emissions. Planet Space Sci 55:598–617 Zarka P (2010) Radioastronomy and the study of exoplanets. In: Coudé du Foresto V, Gelino DM, Ribas I (eds) Pathways towards habitable planets, ASP conference series, vol 430. Astronomical Society of the Pacific, San Francisco, pp 175–180 Zarka P, Treumann RA, Ryabov BP, Ryabov VB (2001) Magnetically-driven planetary radio emissions and applications to extrasolar planets. Astrophys Space Sci 277:293–300 Zarka P, Lazio TJW, Hallinan G (2015) Magnetospheric radio emissions from exoplanets with the SKA. In: Advancing astrophysics with the square kilometre array, Giardini Naxos. SKA Organisation (Dolman Scott Ltd), Jodrell Bank Observatory, Macclesfield Zarka P, Marques M, Louis C et al (2017) Radio emission from the Ganymede-Jupiter interaction and consequence for radio emission from exoplanets. Submitted

The Impact of Stellar Activity on the Detection and Characterization of Exoplanets

88

Andrew Collier Cameron

Contents Introduction: Stellar Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Photometric Effects of Stellar Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Spectroscopic Effects of Stellar Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Proxy Indicators for Activity-Driven RV Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sampling and Decorrelation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary and Future Prospects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1792 1792 1794 1795 1797 1797 1798

Abstract

At the level of a stellar photosphere, stellar magnetic fields manifest themselves as the stellar equivalents of sunspots and faculae. The dark, localized spots give rise to a rotationally modulated background signal that both increases the fractional depth of exoplanet transits and increases the variability of the background against which they are detected. The convective motions of the electrically conducting photospheric gas are inhibited in spots and in facular regions, again producing rotationally modulated variability in spectral line shapes as the star rotates. Here I discuss the physical phenomena that give rise to these forms of variability, their impact on the detection and characterization of extrasolar planets, and proxy indicators and observing strategies that can mitigate this impact.

A. C. Cameron () Centre for Exoplanet Science, SUPA School of Physics and Astronomy, University of St Andrews, St Andrews, UK e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_23

1791

1792

A. C. Cameron

Introduction: Stellar Activity Stellar analogues of the Sun’s dark spots and bright faculae are ubiquitous among stars possessing outer convective zones. The magnetic fields that drive this activity arise through the complex interaction between differential rotation and cyclonic convection in the rotating, convecting, electrically conducting fluid of a star’s outer convective envelope. The dependence of chromospheric activity on rotation and convection was established observationally during a long-term survey, begun at Mt Wilson in 1966, of chromospheric Ca II H&K emission in a large number of solar-type stars. Vaughan et al. (1981) established that the Ca II emission of solartype stars exhibited short-term rotational modulation as well as decade-scale activity cycles. They found further that stars of a given color follow an inverse relationship between the logarithmic activity index and the stellar rotation period inferred from the emission modulation. Noyes et al. (1984) showed that if the chromospheric 0 emission flux was expressed as a fraction RHK of bolometric stellar luminosity, its dependence on convection and rotation was a simple function of the ratio Ro of the stellar rotation period to the theoretically derived convective turnover time at the base of the convective zone. The same magnetic fields that heat the chromosphere also confine and heat the stellar corona to temperatures sufficiently high that the coronal plasma is gravitationally unbound (Parker 1958). The resulting hot stellar wind is channeled along open magnetic field lines which rotate with the star, exerting a magnetic torque that transfers angular momentum from the star to the escaping wind material (Weber and Davis 1967; Mestel and Spruit 1987). The resulting decline in stellar spin rates with age is clearly seen among main sequence stars later than mid-F type in open clusters (Skumanich 1972) and underpins the use of gyrochronology as a means of estimating the ages of individual stars (Barnes 2003, 2007).

Photometric Effects of Stellar Activity Sunspots, and by analogy starspots, appear where intense concentrations of magnetic flux erupt through the stellar photosphere. The photospheric gas is sufficiently highly ionized that it is tied to the field lines. This inhibits the upward convective transport of heat through the visible photospheric layers. As a star rotates, starspot groups form, are carried across the face of the star, and eventually decay as they are eaten away by turbulent diffusion. A single spot or spot group depresses the observable photospheric flux by an amount that depends on the spot’s area and surface brightness, as well as foreshortening, limb darkening, and the stellar equivalent of the solar Wilson depression, which reflects the vertical opacity structure of the spot. On the Sun, the largest spots can occupy up to about 0.1% of the visible hemisphere. With umbral surface brightnesses typically 0.2–0.3 times that of the quiet photosphere, and taking limb dark-

88 The Impact of Stellar Activity on the Detection and Characterization. . .

1793

ening into account, the amplitude of the solar optical flux modulation is also about 0.1%. The form of the modulation seen by an Earth-based observer depends on the latitude of the spot, the inclination of the stellar rotation axis to the line of sight, and the ratio of the spot lifetime to the rotation period. On the Sun, spots seldom last more than a single rotation, so it is relatively uncommon for the same spot group to traverse the visible hemisphere twice. Faster-rotating stars tend to be more active and to have larger spots. Larger spots take longer than small ones to decay under the effects of turbulent diffusion. The combination of longer spot lifetimes and shorter stellar rotation periods results in multiple successive passages across the visible hemisphere. The resulting light curve shows a more easily discernible periodicity than is seen in a star like the Sun. A planet in an orbit of period P around a star of mass M has an orbital semimajor axis a D .P =2/2=3 .GM /1=3 . If the planet’s orbital axis is inclined at 90ı to the line of sight, the full duration at half depth W of the transit occupies a fraction of the orbital period W 2=a .P =2/2=3 .GM /1=3 D D : P 2R R

(1)

For a planet with an orbital period of a few days, W is typically 2–4 h. This is significantly shorter than the spin period of any but the very youngest solar-type stars, so the timescales of transit signals are significantly shorter than that of the starspot modulation. Even if the transit depth is significantly less than the amplitude of the starspot modulation, Fourier or other filtering techniques are generally effective in allowing a smooth approximation to the starspot signal to be fitted as part of the transitsearch algorithm. Aigrain et al. (2004) devised Bayesian filtering schemes that, when applied to simulated space photometry, recovered injected transit signals with an efficiency approaching the theoretical statistical limit (Aigrain and Irwin 2004). For planet characterization, starspots present a more challenging problem. Transmission spectroscopy of planetary atmospheres relies on accurate measurement of the radius of the planet’s silhouette as a function of wavelength. Spots that are not occulted by the planet reduce the flux received from the visible stellar hemisphere. The flux blocked by a planet traversing bright, quiet photosphere is unchanged. The ratio of flux blocked in transit to total flux out of transit is therefore increased if unocculted spots are present. Pont et al. (2008) pointed out that this can lead to a wavelength-dependent overestimation of the planet radius. The spot contrast is greater at blue wavelengths, so the inferred planet radius will increase at shorter wavelengths unless the spot area, temperature, and limb darkening are modeled carefully as an integral part of the radius estimation calculation (Pont et al. 2013). In a recent study of the transmission spectrum of the mini-Neptune orbiting GJ1214, Rackham et al. (2016) have shown that bright emission from faculae also needs to be modeled correctly.

1794

A. C. Cameron

Spectroscopic Effects of Stellar Activity High-precision radial velocity measurements constitute an essential part of the exoplanet discovery chain, in both radial velocity surveys and transit surveys. Relatively early in the history of radial velocity surveys, Saar and Donahue (1997) studied the impact of both starspots and convective inhomogeneities on stellar radial velocity measurements. In the data that were sampled relatively sparsely on the timescale of the stellar rotation, an additional “noise” component was identified whose amplitude increased with both rotation rate and the level of surface activity. The effects of dark spots were discussed further by Hatzes (2002), Desort et al. (2007), and Lagrange et al. (2010). On a rotating star, the “missing” light blocked by dark surface features contains both continuum and line absorption features, Doppler shifted by an amount that depends on the stellar equatorial rotation speed, the inclination of the rotation axis to the line of sight, and the projected distance of the spot from the stellar rotation axis. The resulting distortion of the rotation profile consists of a bright “bump,” which migrates across the line absorption profile from blue to red as the spot traverses the visible hemisphere. Queloz et al. (2001) noted that the 3.7987-day radial velocity variation in the star HD 166435, which at first appeared to be caused by a planetary companion, was accompanied by variations in the line asymmetry, with a quarter-cycle phase shift as expected from starspot modulation. A number of other false detections (e.g., Henry et al. 2002; Desidera et al. 2004; Figueira et al. 2010; Kane et al. 2016; Anglada-Escudé et al. 2016) have been uncovered since, underlining the importance of independent proxy indicators for activity. In slowly rotating stars such as the Sun, however, the spots are relatively small. Meunier et al. (2010b) found that in such stars, the line profile distortions induced by starspots are not the major contributor to the astrophysical radial velocity signal. By partitioning solar images from the MDI instrument aboard the SOHO spacecraft into starspots, faculae, and quiet-sun photosphere, Meunier et al. (2010b) established that the dominant radial velocity signal came from magnetic suppression of granular convection in facular regions. This convective radial velocity signal has an amplitude of several meters per second on the timescale of the solar rotation, superimposed on a longer-term trend of order 6 m s1 in the mean velocity that varies with the solar activity cycle. The visible solar photosphere lies at the top of the solar convective zone. The fundamental convective elements are granules, where hot material from the deep photosphere flows upward, splays out, and cools before sinking into the dark intergranular lanes. High-dispersion long-slit spectroscopy with sufficient spatial resolution to resolve the granulation pattern shows characteristic “wiggly lines” reflecting the difference in Doppler shift and surface brightness between the upflows and downflows. When averaged over the million or so granules on the visible solar hemisphere, the resulting lines display profile asymmetries and wavelength shifts (Dravins et al. 1981). Gray and Toner (1985) and Dravins (1987a, b) studied the photospheric line asymmetries of F, G, and K stars, finding that the velocity

88 The Impact of Stellar Activity on the Detection and Characterization. . .

1795

difference between the granular upwellings and the intergranular downflows appears to increase with effective temperature. Small magnetic structures in the solar photosphere take the form of single flux tubes. These can appear as dark pores when viewed near the center of the solar disc, but toward the solar limb, the bright inner walls of the flux tubes become visible. These flux tubes congregate in large diffuse structures known as faculae, which are spatially associated with but cover a much larger area than sunspot groups. Elsewhere on the Sun, small magnetic flux tubes tend to be swept to the edges of the large convective upwellings known as supergranules, forming the bright photospheric and chromospheric network (Title et al. 1989). Viewed at high-spatial resolution, granules in magnetic regions appear distorted, and the velocity contrast between the granules and the lanes is reduced. Toner and Gray (1988) noted that the active star  Boo A displayed a periodic modulation in both line shift and asymmetry, suggesting that magnetic activity was influencing the globally averaged convective line asymmetry to an observable extent. More recently, Cegla et al. (2013) have performed radiative transfer calculations on 3D magnetohydrodynamic simulations of the solar convection pattern, to parameterize the absorption line shifts and asymmetries from granules and intergranular lanes in both magnetic and nonmagnetic regions, over the full range of foreshortening angles. These early studies showed clearly that stellar magnetic activity affects the global convection signature in spectral lines in a way that will give rise to radial velocity variations even in the absence of planetary reflex motion. The advent of space-based photometry from missions such as MOST, CoRoT, and Kepler/K2 has made it possible to detect the transits of planets with radii as small as that of Earth. The radial velocity follow-up of such small planets is essential for the determination of their bulk densities, in order to distinguish rocky superEarths from volatile-rich mini-Neptunes. With amplitudes of order 1 m s1 or less, the orbital reflex motions of such planets may be overwhelmed by the convective variability of even a moderately inactive host star such as the Sun.

Proxy Indicators for Activity-Driven RV Signals Understanding the origin of the activity-driven radial velocity signals is the first step toward disentangling them from planetary orbital motion. To do this successfully, it is necessary to find some observable photometric or spectroscopic characteristic of the star, whose time variability correlates with the radial velocity signal. Aigrain et al. (2012) developed a photometric proxy for the effects of dark starspots in cases where simultaneous high-precision photometry was available. They showed that, to first order, the radial velocity signatures of dark starspots can be approximated by the product of the flux variation F and its first derivative F 0 with respect to time. They also postulated that the suppression of convective blueshift in facular regions could be modeled by assuming the faculae to be spatially correlated with the spot groups. Haywood et al. (2014) applied this method to radial velocities

1796

A. C. Cameron

of the CoRoT-7 system obtained with the HARPS instrument on the ESO 3.6-m telescope on 26 consecutive nights with simultaneous photometry from the CoRoT spacecraft. They found that the photometric FF 0 and F 2 proxies accounted for some, but not all, of the stellar activity signal. The residual activity signal exhibited covariance properties that were well represented by a Gaussian process regression with a quasiperiodic covariance model. Such covariance models have many features in common with the autocorrelation function of a spotted star’s light curve, with a recurrence period equal to the stellar rotation period and a decay envelope with a timescale comparable with the lifetimes of the largest active regions. Faria et al. (2016) showed that even in the absence of simultaneous photometry, a Gaussian process regression model of the stellar activity recovered the same planetary signals in the CoRoT-7 system successfully. Rajpaul et al. (2015) developed a latent variable version of the FF 0 approach, for cases where simultaneous photometry was not available. Using a Gaussian process regression with the latent variable as input, they were able to model simultaneously the Ca II H&K emission from plages and the proxy indicators for the line profile width and asymmetry. Haywood et al. (2016) observed the radial velocity variations of sunlight reflected from asteroid 4/Vesta over a period of 2 months. They applied the methodology of Meunier et al. (2010b) to magnetograms, Dopplergrams, and continuum images obtained simultaneously with the HMI instrument on the Solar Dynamics Observatory. The resulting synthetic solar radial velocities were found to give a close match to the 4/Vesta observations. The line-of-sight magnetic flux density averaged over the visible hemisphere was correlated strongly with the radial velocity signal. This confirmed that the magnetic suppression of granular convection in faculae dominated the signal. Other efforts to model the radial velocity variability have used a more physical modeling approach, using a low-order representation of the stellar surface brightness distribution to model the radial velocity variations. Lanza et al. (2010) used this approach on nonsimultaneous data from CoRoT and HARPS for the CoRoT-7 system. A similar forward modeling tool, the Spot Oscillation And Planet code (SOAP Boisse et al. 2012; Oshagh et al. 2013; Dumusque et al. 2014), combines starspot modulation with a model for suppression of convective blueshift in plage regions. This code was used to synthesize artificial radial velocity data for a recent blind benchmarking exercise (Dumusque 2016), in which several teams applied a wide range of methods to the problem. The most efficient methods combined modeling of the different activity proxies with rigorous Bayesian model selection to quantify the evidence for the number of coherent planetary signals present. In a more extreme context, Donati et al. (2016) used a Doppler imaging-based method to separate the planetary and activity signatures in the very young, active, and rapidly rotating weak-line T Tauri star V830 Tau, revealing the presence of a hot Jupiter in a 5-day orbit. Petit et al. (2015) incorporated a model for Keplerian reflex motion in a maximum entropy Doppler imaging procedure, allowing full optimization of the orbit and the stellar surface brightness distribution for the spotdominated case. These approaches have the advantage over purely velocity-based methods, in that they make full use of the information encoded in the shape of

88 The Impact of Stellar Activity on the Detection and Characterization. . .

1797

the line profile, allowing a cleaner and physically self-consistent separation of line profile variability and the dynamical reflex motion due to an orbiting body.

Sampling and Decorrelation The Gaussian process regression and forward modeling approaches both require that the activity-driven modulation signal and the planetary orbit be adequately sampled. There is an extensive literature on Doppler imaging, which is beyond the scope of the present chapter. In essence, however, the effects of foreshortening and limb darkening dictate that the radial velocity curve of an active star is unlikely to have more than three local maxima per rotation cycle. In order to characterize such a waveform, it is necessary to obtain at least 6–8 samples, approximately uniformly spaced in rotation phase. These need not be obtained within a single stellar rotation, provided the sampling is done within the lifetime of a typical active region. Inadequate sampling leaves the form of the activity signal unconstrained, as López-Morales et al. (2016) demonstrated in their analysis of the Kepler-21 system.

Summary and Future Prospects Stellar activity impacts the detection thresholds for extrasolar planets via the transit and radial velocity methods, by increasing the level of correlated background noise. The impact on transit detection thresholds is not severe, provided the rotationally modulated activity signal is modeled simultaneously with the transits. The rotation periods of exoplanet host stars suitable for radial velocity follow-up are generally of order days to weeks, while transit durations are measured in hours. This difference in timescale makes it relatively straightforward to separate the starspot modulation from transits. The presence of unocculted spots and faculae on the visible stellar hemisphere during a transit does, however, bias the estimated transit depth. This bias is wavelength dependent, being more severe at blue wavelengths where the brightness contrast between active and quiet photosphere is greatest. Magnetic suppression of granular convection in faculae is the dominant contributor to the spectral line profile distortions that mimic radial velocity signals. To a lesser extent, spectral lines are also distorted by the Doppler-shifted flux deficits and enhancements in spots and faculae, respectively. If the rotation period or any of its submultiples falls close to the orbital period of a planet, great care is needed in the design of the observing program. It must sample the stellar rotation cycle adequately on timescales comparable to the lifetime of an active region. Moreover, observations must continue for long enough that the phase and amplitude of the activity signal evolve significantly. Given such a dataset, time-series analysis methods such as Gaussian process regression are effective at separating the coherent Keplerian signals of planetary orbits from the quasiperiodic variations caused by stellar activity. This approach works particularly well if proxy indicators for the instantaneous line-of-light magnetic flux on the visible stellar hemisphere are

1798

A. C. Cameron

present in the same spectra used to determine the radial velocities. New infrared échelle spectrometer instruments such as CARMENES (Quirrenbach et al. 2012), SPIRou (Artigau et al. 2014) and NIRPS (Conod et al. 2016) will have the capability to achieve this through direct measurement of Zeeman splitting of magnetically sensitive lines at near-infrared wavelengths.

References Aigrain S, Irwin M (2004) Practical planet prospecting. MNRAS 350:331–345 Aigrain S, Favata F, Gilmore G (2004) Characterising stellar micro-variability for planetary transit searches. A&A 414:1139–1152 Aigrain S, Pont F, Zucker S (2012) A simple method to estimate radial velocity variations due to stellar activity using photometry. MNRAS 419:3147–3158 Anglada-Escudé G, Tuomi M, Arriagada P et al (2016) No evidence for activity correlations in the radial velocities of Kapteyn’s star. ApJ 830:74 Artigau É, Kouach D, Donati JF et al (2014) SPIRou: the near-infrared spectropolarimeter/highprecision velocimeter for the Canada-France-Hawaii telescope. In: Ground-based and Airborne Instrumentation for Astronomy V. Proc SPIE 9147:914715. https://doi.org/10.1117/12.2055663 Barnes SA (2003) On the rotational evolution of solar- and late-type stars, its magnetic origins, and the possibility of stellar gyrochronology. ApJ 586:464–479 Barnes SA (2007) Ages for illustrative field stars using gyrochronology: viability, limitations, and errors. ApJ 669:1167–1189 Boisse I, Bonfils X, Santos NC (2012) SOAP. A tool for the fast computation of photometry and radial velocity induced by stellar spots. A&A 545:A109 Cegla HM, Shelyag S, Watson CA, Mathioudakis M (2013) Stellar surface magneto-convection as a source of astrophysical noise. I. Multi-component parameterization of absorption line profiles. ApJ 763:95 Conod U, Blind N, Wildi F, Pepe F (2016) Adaptive optics for high resolution spectroscopy: a direct application with the future NIRPS spectrograph. In: Adaptive optics systems V. Proceedings of the SPIE, vol 9909, p 990941. https://doi.org/10.1117/12.2233651 Desidera S, Gratton RG, Endl M et al (2004) No planet around HD 219542 B. A&A 420:L27–L30 Desort M, Lagrange AM, Galland F, Udry S, Mayor M (2007) Search for exoplanets with the radial-velocity technique: quantitative diagnostics of stellar activity. A&A 473:983–993 Donati JF, Moutou C, Malo L et al (2016) A hot Jupiter orbiting a 2-million-year-old solar-mass T Tauri star. Nature 534:662–666 Dravins D (1987a) Stellar granulation: part two – stellar photospheric line asymmetries. A&A 172:211 Dravins D (1987b) Stellar granulation. I – the observability of stellar photospheric convection. A&A 172:200–224 Dravins D, Lindegren L, Nordlund A (1981) Solar granulation – influence of convection on spectral line asymmetries and wavelength shifts. A&A 96:345–364 Dumusque X (2016) Radial velocity fitting challenge. I. Simulating the data set including realistic stellar radial-velocity signals. A&A 593:A5 Dumusque X, Boisse I, Santos NC (2014) SOAP 2.0: a tool to estimate the photometric and radial velocity variations induced by stellar spots and plages. ApJ 796:132 Faria JP, Haywood RD, Brewer BJ et al (2016) Uncovering the planets and stellar activity of CoRoT-7 using only radial velocities. A&A 588:A31 Figueira P, Marmier M, Bonfils X et al (2010) Evidence against the young hot-Jupiter around BD +20 1790. A&A 513:L8 Gray DF, Toner CG (1985) Inferred properties of stellar granulation. PASP 97:543–550 Hatzes AP (2002) Starspots and exoplanets. Astronomische Nachrichten 323:392–394

88 The Impact of Stellar Activity on the Detection and Characterization. . .

1799

Haywood RD, Collier Cameron A, Queloz D et al (2014) Planets and stellar activity: hide and seek in the CoRoT-7 system. MNRAS 443:2517–2531 Haywood RD, Collier Cameron A, Unruh YC et al (2016) The Sun as a planet-host star: proxies from SDO images for HARPS radial-velocity variations. MNRAS 457:3637–3651 Henry GW, Donahue RA, Baliunas SL (2002) A false planet around HD 192263. ApJ 577: L111–L114 Kane SR, Thirumalachari B, Henry GW et al (2016) Stellar activity and exclusion of the outer planet in the HD 99492 System. ApJ 820:L5 Lagrange AM, Desort M, Meunier N (2010) Using the Sun to estimate Earth-like planets detection capabilities . I. Impact of cold spots. A&A 512:A38 Lanza AF, Bonomo AS, Moutou C et al (2010) Photospheric activity, rotation, and radial velocity variations of the planet-hosting star CoRoT-7. A&A 520:A53 López-Morales M, Haywood RD, Coughlin JL et al (2016) Kepler-21b: a rocky planet around a V = 8.25 magnitude star. AJ 152:204 Mestel L, Spruit HC (1987) On magnetic braking of late-type stars. MNRAS 226:57–66 Meunier N, Desort M, Lagrange AM (2010) Using the Sun to estimate Earth-like planets detection capabilities . II. Impact of plages. A&A 512:A39 Noyes RW, Hartmann LW, Baliunas SL, Duncan DK, Vaughan AH (1984) Rotation, convection, and magnetic activity in lower main-sequence stars. ApJ 279:763–777 Oshagh M, Boisse I, Boué G et al (2013) SOAP-T: a tool to study the light curve and radial velocity of a system with a transiting planet and a rotating spotted star. A&A 549:A35 Parker EN (1958) Dynamics of the interplanetary gas and magnetic fields. ApJ 128:664 Petit P, Donati JF, Hébrard E et al (2015) A maximum entropy approach to detect close-in giant planets around active stars. A&A 584:A84 Pont F, Knutson H, Gilliland RL, Moutou C, Charbonneau D (2008) Detection of atmospheric haze on an extrasolar planet: the 0.55–1.05 m transmission spectrum of HD 189733b with the HubbleSpaceTelescope. MNRAS 385:109–118 Pont F, Sing DK, Gibson NP et al (2013) The prevalence of dust on the exoplanet HD 189733b from Hubble and Spitzer observations. MNRAS 432:2917–2944 Queloz D, Henry GW, Sivan JP et al (2001) No planet for HD 166435. A&A 379:279–287 Quirrenbach A, Amado PJ, Seifert W et al (2012) CARMENES. I: instrument and survey overview. In: Ground-based and airborne instrumentation for astronomy IV. Proceedings of the SPIE, vol 8446, p 84460R. https://doi.org/10.1117/12.925164 Rackham B, Espinoza N, Apai D et al (2017) ACCESS I: an optical transmission spectrum of GJ 1214b reveals a heterogeneous stellar photosphere. ApJ 834:151 Rajpaul V, Aigrain S, Osborne MA, Reece S, Roberts S (2015) A Gaussian process framework for modelling stellar activity signals in radial velocity data. MNRAS 452:2269–2291 Saar SH, Donahue RA (1997) Activity-related radial velocity variation in cool stars. ApJ 485: 319–327 Skumanich A (1972) Time scales for CA II emission decay, rotational braking, and lithium depletion. ApJ 171:565 Title AM, Tarbell TD, Topka KP et al (1989) Statistical properties of solar granulation derived from the SOUP instrument on Spacelab 2. ApJ 336:475–494 Toner CG, Gray DF (1988) The starpatch on the G8 dwarf XI Bootis A. ApJ 334:1008–1020 Vaughan AH, Preston GW, Baliunas SL et al (1981) Stellar rotation in lower main-sequence stars measured from time variations in H and K emission-line fluxes. I – initial results. ApJ 250: 276–283 Weber EJ, Davis L Jr (1967) The angular momentum of the solar wind. ApJ 148:217–227

89

Tidal Star-Planet Interactions: A Stellar and Planetary Perspective Stéphane Mathis

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tides: General Principles, Waves, and Dissipation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tides in Host Stars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Equilibrium Tide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Dynamical Tide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tides in Giant Planets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Internal Structure and Dynamics of Giant Planets . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tidal Dissipation Within Giant Planets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tides in Telluric Planets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Multilayered Telluric Planets and Tidal Forcing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tidal Dissipation Within Telluric Planets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1802 1803 1805 1807 1807 1812 1813 1815 1820 1820 1822 1824 1824

Abstract

Since 22 years, more than 3600 exoplanets have now been discovered in 2700 planetary systems, among which 610 host several planets. These discoveries reveal us a large diversity of planets, from super-Earths to hot Jupiters, orbiting host stars of different masses, ages, and metallicities. Moreover, the detected exoplanetary systems present a broad variety of orbital architectures that shows that the solar system is not typical. In addition, some of the detected telluric planets orbit within the habitable zone of their host stars. This challenges our

S. Mathis () Laboratoire AIM Paris-Saclay, IRFU/DAp Centre de Saclay, CEA/DRF – CNRS – Université Paris Diderot, Gif-sur-Yvette Cedex, France LESIA, Observatoire de Paris, PSL Research University, CNRS, Sorbonne Universités, UPMC Univ. Paris 06, Univ. Paris Diderot, Meudon, France e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_24

1801

1802

S. Mathis

understanding of the formation, the evolution, and the stability of planetary systems. In this context, tidal interactions should be coherently modeled taking advantage of the different observational signatures they have in our Earth-Moon system, in the solar system, and in exoplanetary systems. They reveal how tidal dissipation in celestial bodies varies strongly as a function of their internal structure and dynamics. In this chapter, we make a complete review of the physical mechanisms driving tidal friction in stars, giant planets, and telluric planets. For each type of celestial bodies, we discussed in details the state of the art of their modeling and how the resulting predictions compare to observations. We show the importance to adopt a stellar and planetary physicist perspective when studying tidal dissipation in planetary systems.

Introduction Twenty-two years ago, a revolution occurred in astronomy and astrophysics with the discovery of the first exoplanet, 51 Pegasi b, by Mayor and Queloz (1995). After two decades, more than 3600 exoplanets have now been detected using high-precision ground-based instruments and space missions, particularly with radial velocities and transit methods (see, e.g., Perryman 2011, and references therein). In this rich and exciting context, new types of planets such as the hot Jupiters, super-Earths, and mini-Neptunes, for instance, have been identified orbiting around a broad diversity of host stars with different masses, ages, metallicities, and rotation. Moreover, the orbital architectures of exoplanetary systems, with potentially ultracompact systems like the one discovered recently around Trappist-1 (Gillon et al. 2017), show that the solar system is not a typical system. These discoveries challenge our understanding of the formation, the evolution, and the stability of planetary systems. In this context, a global and detailed understanding of star-planet and planet-planet interactions within the large diversity of observed configurations is mandatory. Among these interactions, tides are a fundamental but complex physical mechanism. They deeply impact the orbital and rotational dynamics of planets, moons, and stars constituting planetary systems as observed both in the solar and extrasolar systems. Tidal orbital migration is also essential to predict the position of a planet, particularly relatively to the habitable zone of its host star. In addition, the impact on celestial body rotation modifies their magnetism. Finally, tides generate internal heating in planetary interiors with important consequences as, for instance, the volcanism induced on Io because of its strong tidal interactions with Jupiter. Therefore, tides should also be understood when studying planetary systems habitability, a fundamental question for modern astronomy. In this chapter, we propose to give a complete review of the state of the art of the physics of tides in the different observed types of host stars and exoplanets. In the first part, we provide a summary of the general principles driving the physics of tides. Then, we successively make a detailed review of the large diversity of tidal mechanisms occurring in stars and in gaseous and telluric planets, which

89 Tidal Star-Planet Interactions: A Stellar and Planetary Perspective

1803

are generally treated separately in the literature. For each process, we point out the importance of stellar and planetary internal structure and dynamics. In the conclusion, we finally summarize the important points that one should remember on star-planet tidal interactions when adopting the point of view of stellar and planetary physics.

Tides: General Principles, Waves, and Dissipation In planetary systems, planets and their host star(s) are submitted to mutual gravitational interactions. When one goes beyond the punctual mass approximation for a given body (called the primary), the gravitational force exerted by another body is different at its surface and at its center of mass. A differential gravitational force is thus applied, which is by definition the tidal force (fT ). It can be derived from the tidal potential (UT ) as fT D r UT

where

UT D

GMc P2 Œcos .d  r/ ; d3

(1)

where G is the universal gravity constant, Mc the mass of the considered tidal companion treated here as a point-like body with a negligible angular momentum, d (d) the distance (vector) between the centers of mass of the primary and the perturber, r the current position vector in the primary, and P2 the quadrupolar Legendre polynomial. In this chapter, our choice is to contribute to the literature by making a complete review of tidal processes in host stars and in the different types of planets, from giant planets to super-Earths, which are generally studied separately. We thus refer the reader to reference books (Murray and Dermott 1999) and other reviews (Mathis et al. 2013; Ogilvie 2014) for a complete description of the different complex mathematical formalisms that are used in each case while focusing here on fundamental processes and on the impact of stellar and planetary structure and dynamics. Both in fluid and solid celestial bodies, the tidal force perturbs the hydrostatic balance from which mass redistribution and perturbations of the gravitational potential and pressure result. The mass redistribution in the direction of the companion is called the tidal bulge. It induces large-scale velocity (elastic displacement) in fluid (solid) layers, the so-called fluid (elastic) equilibrium tide (e.g., Zahn 1966a; Tobie et al. 2005). However, this velocity (displacement), depending on the value of the tidal frequency   lnorb  m˝, where norb and ˝ are the angular velocities of the orbit and of the rotation of the primary, respectively, and l and m are integers, must be completed by fluid (elastic) eigenmodes of oscillation of the primary, the socalled dynamical tide. In fluid bodies, the dynamical tide is constituted by inertial, gravity, and acoustic waves excited by the tidal force (when ignoring any magnetic field). Their restoring forces are the Coriolis acceleration, buoyancy (in convectively stable layers), and the compressibility of the fluid, respectively (e.g., Rieutord 2015). In solid bodies, the dynamical tide is constituted by seismic waves (e.g., Tobie

1804

S. Mathis

Fig. 1 General principle of tides: a companion (B), which orbits around the primary (A) with an orbital angular velocity norb at a distance d , generates a tidal bulge because of tidal forces. In an idealized case, where any internal friction in A is ignored, a tidal bulge is created in the direction of the line of centers; this corresponds to an adiabatic tidal adjustment. When tidal friction is taken into account, the response of A presents a delay, and the tidal bulge becomes misaligned with the tidal angle (ı); this is the tidal dissipative adjustment. Tidal friction, in the case of stable systems, then tends to synchronize the rotation of A (˝) with the orbital motion (norb ) of B, thanks to the tidal torque (Eq. 2)

et al. 2005). In the case where the primary will be frictionless, the equilibrium and dynamical tides will be in phase with the tidal potential and the resulting tidal bulges aligned along the line of centers. Moreover, the tidal torque applied on the rotation of the primary and on the orbit vanishes when averaged on a tidal period in this adiabatic situation. However, this is an ideal situation that never exists and the equilibrium and dynamical tides are submitted to friction mechanisms such as viscosity and heat diffusion. Their kinetic and potential energies are converted into heat; this is the tidal heating that impacts the structure and the evolution of the body. Moreover, because of the dissipation, the direction of tidal bulges has an angle, the tidal angle ı, with the line of centers. This general principle of tidal interactions is summarized in Fig. 1. Therefore, a net tidal torque is applied both on rotation and the orbit. Following Zahn (2013), the torque applied to the spin of the primary body can be approximated in the simplest case of a circular coplanar system as .˝  norb / ( D tfriction



Mc Mp

2

Mp Rp2



Rp d

6 ;

(2)

where Mp and Rp are the mass and the radius of the primary and tfriction is the time characterizing tidal dissipation. This expression shows the importance of understanding the physics of tidal dissipation from first-principles equations to obtain a robust evaluation of tfriction and predict the evolution of planetary systems. Moreover, it allows one to identify the crucial importance of the socalled corotation radius at which norb D ˝. On the one hand, if norb > ˝, then the companion migrates inward and the primary rotation accelerates. On the other

89 Tidal Star-Planet Interactions: A Stellar and Planetary Perspective

1805

hand, if norb < ˝, the companion migrates outward and the primary rotation slows down. This is the configuration of the Earth-Moon system where the Moon migrates outward by 3:8 centimeters per year. In the first case (norb > ˝), if Lorb  3Lp , where Lorb and Lp are the angular momentum of the orbit and of the primary, respectively, the companion spirals toward the primary until it is tidally disrupted at the Roche limit (Hut 1980). If Lorb 3Lp , the binary system constituted by the primary and the companion tends when isolated toward an equilibrium state where the orbit is circular and the rotational spins and the orbital one are aligned and synchronized. In this framework, the time characterizing dissipation (tfriction ) allows one to predict circularization, alignment, and synchronization time scales. Finally, it is usual to express tfriction as a function of a corresponding tidal quality factor Q (MacDonald 1964) inspired by the theory of forced oscillators (e.g., Greenberg 2009). A strong dissipation corresponds to a low-quality factor and friction time (and vice versa). This quality factor is always combined with the quadrupolar Love number k2 , defined as the ratio of the self-gravitation potential perturbation and of the tidal potential at the surface of the studied body; it measures its mass concentration. Ogilvie and Lin (2007) also introduced the modified tidal quality factor Q0 D 3Q= .2k2 /, which reduces to Q for a homogeneous body. Finally, it is important to point out that each dissipation mechanism has a specific frequency dependence (i.e., smooth or resonant; we refer to Mathis et al. 2013; Ogilvie 2014, for mathematical formalisms), particularly for the resonant fluid dynamical tide, with important consequences for systems dynamical evolution (Zahn and Bouchet 1989; Witte and Savonije 2002; Efroimsky and Lainey 2007; Auclair-Desrotour et al. 2014). Each tidal dissipation mechanism, in each type of celestial bodies, should thus be carefully examined.

Tides in Host Stars From 51 Pegasi b, the first discovered exoplanet (Mayor and Queloz 1995), to the recently detected Trappist-1 system (Gillon et al. 2017), a large number of the known exoplanets orbit very close to their host stars. In such compact configurations, star-planet tidal interactions, and more specifically the dissipation of tides in stars, play a key role to shape systems orbital architecture. A first signature of its action is the small orbital eccentricity of systems with small orbital periods (e.g., 1.5 M§

Verbunt and Phinney 1995; Meibom and Mathieu 2005), that tidal dissipation in stars varies strongly with stellar mass, age, and rotation. This may have major consequences for exoplanetary systems, which have been observed around a large diversity of host stars. A detailed study of the three components of tidal dissipation in stellar interiors: the dissipation of the equilibrium tide, the dissipation of tidal inertial waves in their convective zones, and the dissipation of tidal gravito-inertial waves in stably stratified radiative zones, is thus needed. In Fig. 3, we recall the distribution of convective and radiative regions in main-sequence stars as a function of their mass.

89 Tidal Star-Planet Interactions: A Stellar and Planetary Perspective

1807

The Equilibrium Tide The founder of the theory of the equilibrium tide in stellar interiors was JeanPaul Zahn during his PhD thesis. He computed its velocity field (Zahn 1966a, see also Remus, Mathis and Zahn 2012 (Fig. 4 – left panel)) and its dissipation by the friction applied by turbulence and the diffusion of heat in convective and radiative zones, respectively (Zahn 1966a, b). He demonstrated that the dissipation of the equilibrium tide is efficient in the convective envelope of low-mass stars (from M- to F-type stars), but it is negligible in the convective core of intermediatemass and massive stars (from A- to O-type stars) and in stellar radiation zones (Zahn 1977). For the case of the convective envelope of low-mass stars, it was thus important to propose a robust modeling of the turbulent friction applied by convection on the velocity field of the equilibrium tide. Assuming a space-scale separation between convective eddies and the equilibrium tide and the isotropy of turbulence, which corresponds to neglect the impact of rotation and magnetic fields on convection, Zahn (1966a) proposed a mixing-length model based on characteristic velocity (Vc ) and length scales (lc ) for convection. Moreover, Zahn (1966a) identified that in the case of tidal periods (Pt D 2= where  is the tidal frequency defined above) short in comparison to the convective turnover time (the so-called rapid tides regime), the turbulent friction becomes less efficient because convection has not enough time to apply an efficient damping to tidal flows. Zahn (1966a) proposed that this loss of efficiency scales linearly with the ratio between the tidal period (Pt ) and the characteristic convective time (Pc D lc =vc ); Goldreich and Keeley (1977) proposed a scaling proportional to .Pt =Pc /2 based on the Kolmogorov theory for turbulence. These two propositions have been tested using direct numerical simulations of convective turbulence submitted to periodic shears and are still debated (e.g., Penev et al. 2007; Ogilvie and Lesur 2012, (Fig. 4 – right panel)). More recently, Mathis et al. (2016) examined the impact of rotation. Using scaling laws derived theoretically by Stevenson (1979) and verified by direct numerical simulations of turbulent convective rotating flows (Barker et al. 2014), they demonstrated how rotation weakens convection efficiency and the related turbulent friction. The dissipation of the equilibrium tide in rapidly rotating stars should thus be less efficient than in slow rotators. In many cases (e.g., Meibom and Mathieu 2005), the equilibrium tide is not strong enough to explain the observations; the dynamical tide should thus be introduced.

The Dynamical Tide When the orbital period (Porb ) is such that Porb > 1=2 Ps in a coplanar circular system, where Ps is the rotational period of the star, and assuming as a first step a uniform rotation, tidal inertial waves are excited in stellar convective regions (Bolmont and Mathis 2016).

1808

S. Mathis

-5

-4

-3

-2

-1

0

log(D)

a

b

1.0

Diss 0.8

z

0.6

0.4

0.2

0.0 0.0

0.2

0.4

0.6

0.8

1.0

s

c

Fig. 4 Left: (a) velocity field of the equilibrium tide at the surface of a convective star or planet represented by black arrows. The red and orange arrows represent the rotation axis and the direction of the companion, respectively. The colors from blue to red give the intensity of the tidal potential (Taken from Remus et al. 2012a, courtesy of Astronomy & Astrophysics). Middle: (b) viscous dissipation of the kinetic energy of a tidal inertial wave attractor in the convective envelope of a solar-type star with the solar latitudinal differential rotation (see also Guenel et al. 2016a). Right: (c) direct numerical simulation of the nonlinear interactions between the convective turbulence and an oscillatory tidal flow (Taken from Ogilvie and Lesur 2012, courtesy of Monthly Noticies of the Royal Astronomical Society)

In the case of convective cores (for stars whose masses are above 1:1Mˇ , where Mˇ is the solar mass) and fully convective M-type stars (with masses 1

1019 Stotal : total Poynting flux [W ]

Fig. 9 Left Poynting flux as a function of topology () and Alfvén Mach number Ma . Right Estimate of the Alfvén wing Poynting flux for detected exoplanets (as of 2013). Both estimates are based on the analytical model of Saur et al. (2013), from where the figures were adapted

1018 1017 1016 1015 1014 1013 1012 –2 10

10 –1

10 0 10 1 10 2 r exo : star distance [AU]

10 3

10 4

1848

A. Strugarek

today, ˛N has thus been neglected in applications to close-in exoplanets. Finally, using their analytical Alfvén wing model and simple Parker-like wind solution, Saur et al. estimated the Poynting flux through the wings for the close-in exoplanets known in 2013. They found that the Poynting could vary from 1014 to a few 1019 W depending on the orbital distance. Nevertheless, no observational constraints were available to characterize properly the wind of the planet-hosting stars; hence, these estimates have to be taken with caution. In real exoplanetary systems, the interaction state is likely to vary on short (inhomogeneities along the planetary orbit) and large (reversals of the stellar and/or planetary magnetic fields) time scales. Large uncertainties on the stellar wind characteristics of the central star and large uncertainties on the ionospheric properties of the exoplanet (value of the efficiency parameter ˛) N also make the quantitative estimates of SPMIs perilous. Numerical simulations can be used to tackle some of these aspects, which we now turn to.

Numerical Models Numerical models of star-planet interactions take their roots in the pioneering work of Ip et al. (2004) who first simulated the local interaction of a close-in giant exoplanet with the ambient stellar wind plasma. Nevertheless, SPMIs depend on the plasma characteristics from the base of the stellar corona to the vicinity of the planet and on the planet magnetic configuration. Global simulations were developed for the first time by Cohen et al. (2009, 2010, 2011), in which advanced, solar-calibrated stellar wind simulations were adapted to include a close-in orbiting planet (treated as a boundary condition). In their later work, they introduced a boundary condition inside the simulation domain moving with time to follow the orbital path of the planet. These early simulations strikingly showed the natural time variability one expects from SPMIs, as well as possibly very large lags between the orbital phase of the planet and the stellar subpoint where the magnetic interaction connects. Hybrid approaches for which the stellar wind and the planet vicinity are modeled separately for the same system have also been carried out by Kopp et al. (2011), Cohen et al. (2014, 2015), Alvarado-Gómez et al. (2016a, b). Building on the 2.5D approximation used in Strugarek et al. (2014c, 2015) also developed a global 3D model incorporating both the orbiting planet and the star, using as a first approach simple axisymmetric magnetic configurations (see top panels in Fig. 10). A special effort was carried out to develop adequate boundary condition for both the star and the planet, which are critical to correctly quantify SPMIs. The stellar wind boundary condition consists of a three-layer boundary condition, ensuring accurate conservation properties throughout the simulated stellar wind (Zanni and Ferreira 2009; Strugarek et al. 2014a, b). The boundary condition at the planet is defined by a buffer layer in which only the magnetic field of the planet is allowed to change, mimicking crudely a thick ionospheric layer (similarly to the approach developed in Jia et al. 2009, in the context of Ganymede in the Jovian system). More advanced planetary boundary conditions have been developed in the recent years. Duling et al. (2014) developed a generic

90 Models of Star-Planet Magnetic Interaction

Rorb Rorb Rorb Rorb

= 3R = 5R = 6R = 7R

Ma−0.56

Aeff /ΛαP [πRP2 ]

¯ χ [πR2 ] P/Λ P P

101

1849

Ma1.09

100

101

Rorb Rorb Rorb Rorb

= 3R = 5R = 6R = 7R

Ma0.02 Ma1.40

10−1

Anti-aligned Aligned

Anti-aligned Aligned

Ma

100

Ma

100

Fig. 10 Top panels 3D numerical model of Alfvén wings in the aligned configuration. The colored volumes in blue/red trace the currents underlying the wings (the volume is cropped to make the planet apparent). The stellar magnetic field lines are color-coded by the magnetic field strength, and the planetary field is shown by the gray tubes (Adapted from Strugarek et al. 2015). Bottom panels Scaling-laws of the Poynting flux in one Alfvén wing (left) and of the torque applied to the planet (right) deduced from the numerical model of SPMIs of Strugarek et al. 2015. Both are shown as a function of the Alfvénic Mach number Ma (Adapted from Strugarek 2016. Reproduced with the permission of AAS)

boundary condition for nonconductive planetary bodies and showed that the choice of boundary condition for the planet has a drastic impact on the development of the magnetic interaction (see, e.g., the change in the current system on their Fig. 3). Ultimately, the interaction of a planetary magnetosphere with the wind of its star could be modeled with a higher precision using a magnetosphereionosphere coupling boundary condition, as already developed for simulations of the magnetosphere of the Earth (e.g., Goodman 1995; Merkin and Lyon 2010). The simple geometry used in Strugarek et al. (2015) allowed a quantitative comparison between the numerically modeled Alfvén wings and the analytical work of Saur et al. (2013). A good agreement was found between the two models, with Poynting fluxes of similar amplitude. In the 3D numerical model, the nonlinear interaction between the orbiting planet and the star and its wind leads to a slightly

1850

A. Strugarek

more elongated Alfvén wing cross section along v0 . The two opposite topologies of the dipolar interaction were also shown to lead to radically different properties of the magnetic interaction with this model, leading to at least an order of magnitude changes in magnetic torque and Poynting flux. For both aspects, these simulations helped realized how much the effective area of the interaction (or, if one prefers, the obstacle) significantly changes with the topology. In the closed magnetosphere case (anti-aligned configuration), the magnetospheric size can be well approximated using the pressure ratio between the magnetospheric magnetic pressure on the planetary side and the total (thermal plus magnetic plus ram) pressure on the stellar wind side of the interaction. Assuming a spherical magnetosphere composed of a dipole field, this leads to the well-known expression 1=6

Robst D RP )P D RP



BP2 8 Pt

1=6 ;

(10)

where the subscript P denotes values at the planetary surface and Pt is the total pressure of the stellar wind at the planetary orbit. This simple estimate nevertheless fails to describe the area of interaction in the aligned case (see Fig. 10), as in this case the magnetic field lines connecting the planet and the wind that are part of the Alfvén wings act as well as an obstacle. As a result, the area of interaction is far larger in the aligned case. In the extreme case where the travel time of Alfvénic perturbations is short compared to the orbital period (see Eq. 2), the waves can propagate back and forth between the planet location and the stellar surface, and the effective area of interaction is the full Alfvén wing from the planet location to the stellar surface (see also Fleck 2008). In general, though, the effective obstacle will be only composed of a subpart of the Alfvén wings (see Figure 5 in Strugarek 2016). Relying on the good agreement between the numerical and analytical models, a parameter study using a self-consistent stellar wind model was undertaken in Strugarek (2016). By changing the orbital radius and magnetic field strength of the planet, empirical scaling laws have been derived from a large set of nonlinear numerical simulations for the Poynting flux and magnetic torque associated with SPMI. They can be summarized as follows (see bottom panels in Fig. 10 and Strugarek 2016):       T / cd Pt Maˇ  RP2 Rorb  )˛P ;      * P / cd Sw Ma  RP2  )P :

(11) (12)

We recall here that the coupling coefficient is defined as cd D .4=c 2 /˙A vo and the wind Poynting flux is Sw D vo Bw2 =.4/. The interested reader may find more detailed scaling laws in Strugarek (2016), including their dependency on the resistive properties of the plasma and of the modeled ionospheric layer. The exponents (˛; ˇ; ; *) vary with the topology of the interaction and are given in Strugarek (2016). The various terms in Eqs. 11 and 12 have been rearranged in three blocks from left to right: terms that depend only on the star and its wind,

90 Models of Star-Planet Magnetic Interaction

1851

only on the planet properties, and on a combination of both ()P ). As a result, these scaling laws may help relate observed anomalous activity on distant stars (e.g., Shkolnik et al. 2008) with the magnetic properties (strength, topology) of the close-in planet, provided the stellar wind and planetary radius can be inferred or constrained observationally.

Conclusions The study of star-planet magnetic interactions is a young and promising field of research. In this review, we have focused our discussion on the modeling efforts that have been undertaken by the community in the past decades, motivated by the many intriguing phenomena observed in exoplanetary systems. Rather than listing again the effects initiated by magnetic interactions, we list here several routes of improvement of the models that need to be followed in support of the future observation missions dedicated to the characterization of exoplanets and their magnetic properties: • In the context of both the unipolar and dipolar interaction cases, more realistic models of the interior of planets and their magnetosphere are needed. For example, an accurate response of the magnetospheric system to the impacting stellar wind is needed to assess aurorae and possible planetary emissions, to assess the steady state of interaction and the energy conversion the interaction is able to operate, and to constrain the properties of the hypothetical bow shock at the nose of the interaction. These improvements can be carried out as a first step, e.g., by considering a magnetosphere-ionosphere coupling model in the case of the dipolar interaction and more realistic planetary interior models in the context of the unipolar interaction. Ultimately, models including kinetic effects (beyond the standard magneto-hydrodynamic framework) will be needed for some of these aspects. • For the sake of simplicity, models have so far generally considered circular planetary orbits. This is often justified as a reasonable approximation, as closein planets are likely in a tidally locked state. Tidal theory nevertheless predicts departures from this simple picture see  Chap. 89, “Tidal Star-Planet Interactions: A Stellar and Planetary Perspective”. As a result, eccentric orbits should be more systematically included in star-planet magnetic interaction models. • With a growing sample of stars hosting close-in exoplanets for which spectropolarimetric observations are available, it is today possible to simulate realistic stellar winds of particular star-planet systems. Because of the variable nature of the magnetic interaction along the planetary orbit, it is essential to use these observational constraints to model close-in systems in order to assess the robustness of the simplified models and quantitatively test our understanding of magnetic star-planet interactions. • Last but not least, a significant effort has to be made to self-consistently compare magnetic effects with other physical mechanisms at stake in star-planet

1852

A. Strugarek

interactions. These include, but are not limited to, tides (torque, heating), radiation (ionization, atmospheric escape), and particle acceleration in the planetary magnetosphere (magnetic reconnection, instabilities). We hope this review will arouse multiple interests on this promising and multidisciplinary subject of research and will encourage further efforts in developing models of SPMIs in all their complexities to provide critical insights for the future observations of (close-in) exoplanetary systems.

Cross-References  Dynamical Evolution of Planetary Systems  Electromagnetic Coupling in Star-Planet Systems  Magnetic Fields in Planet-Hosting Stars  Planetary Evaporation Through Evolution  Planetary Interiors, Magnetic Fields, and Habitability  Rotation of Planet-Hosting Stars  Signatures of Star-Planet Interactions  Star-Planet Interactions in the Radio Domain: Prospect for Their Detection  Stellar Coronal and Wind Models: Impact on Exoplanets  Tidal Star-Planet Interactions: A Stellar and Planetary Perspective Acknowledgements A. Strugarek acknowledges enlightening discussions about star-planet interactions with A.S. Brun, J. Bouvier, D. Cébron, A. Cumming, S. Matt, V. Réville, and P. Zarka. This review was written while A. Strugarek was partially supported by the Canada’s Natural Sciences and Engineering Research Council, the ANR Blanc 2011 Toupies, and the Centre National d’Etudes Spatiales.

References Adams FC (2011) Magnetically controlled outflows from hot Jupiters. ApJ 730(1):27. https://doi. org/10.1088/0004-637X/730/1/27 Alvarado-Gómez JD, Hussain GAJ, Cohen O et al (2016a) Simulating the environment around planet-hosting stars. A&A 588:A28. https://doi.org/10.1051/0004-6361/201527832 Alvarado-Gómez JD, Hussain GAJ, Cohen O et al (2016b) Simulating the environment around planet-hosting stars. II. Stellar winds and inner astrospheres. A&A 594:A95. https://doi.org/10. 1051/0004-6361/201628988 Bouvier J, Cébron D (2015) Protostellar spin-down: a planetary lift? MNRAS 453(4):3720–3728. https://doi.org/10.1093/mnras/stv1824 Brun AS, Browning MK, Dikpati M, Hotta H, Strugarek A (2015a) Recent advances on solar global magnetism and variability. Space Sci Rev 196(1):101–136. https://doi.org/10.1007/s11214-0130028-0

90 Models of Star-Planet Magnetic Interaction

1853

Brun AS, Garcia RA, Houdek G, Nandy D, Pinsonneault M (2015b) The solar-stellar connection. Space Sci Rev 196(1):303–356. https://doi.org/10.1007/s11214-014-0117-8 Cauley PW, Redfield S, Jensen AG et al (2015) Optical Hydrogen Absorption Consistent with a Thin Bow Shock Leading the Hot Jupiter Hd 189733b. ApJ 810(1):13. https://doi.org/10.1088/ 0004-637X/810/1/13 Clarke JT, Ajello J, Ballester G et al (2002) Ultraviolet emissions from the magnetic footprints of Io, Ganymede and Europa on Jupiter. Nature 415:997–1000 Cohen O, Drake JJ, Kashyap VL et al (2009) Interactions of the magnetospheres of stars and closein giant planets. ApJ 704:L85. https://doi.org/10.1088/0004-637X/704/2/L85 Cohen O, Drake JJ, Kashyap VL, Sokolov IV, Gombosi TI (2010) The impact of hot Jupiters on the spin-down of their host stars. ApJ 723(1):L64–L67. https://doi.org/10.1088/2041-8205/723/ 1/L64 Cohen O, Kashyap VL, Drake JJ et al (2011) The dynamics of stellar coronae harboring hot Jupiters. I. A time-dependent magnetohydrodynamic simulation of the interplanetary environment in the Hd 189733 planetary system. ApJ 733(1):67. https://doi.org/10.1088/0004637X/733/1/67 Cohen O, Drake JJ, Glocer A et al (2014) Magnetospheric structure and atmospheric joule heating of habitable planets orbiting M-dwarf stars. ApJ 790(1):57. https://doi.org/10.1088/0004-637X/ 790/1/57 Cohen O, Ma Y, Drake JJ et al (2015) The interaction of Venus-like, M-dwarf planets with the stellar wind of their host star. ApJ 806(1):41. https://doi.org/10.1088/0004-637X/806/1/41 Cuntz M, Saar SH, Musielak ZE (2000) On stellar activity enhancement due to interactions with extrasolar giant planets. ApJ 533(2):L151–L154. https://doi.org/10.1086/312609 Damiani C, Lanza AF (2015) Evolution of angular-momentum-losing exoplanetary systems. Revisiting Darwin stability. A&A 574:A39. https://doi.org/10.1051/0004-6361/201424318 Donati JF, Landstreet JD (2009) Magnetic fields of nondegenerate stars. Annu Rev A&A 47:333. https://doi.org/10.1146/annurev-astro-082708-101833 Duling S, Saur J, Wicht J (2014) Consistent boundary conditions at nonconducting surfaces of planetary bodies: applications in a new Ganymede MHD model. J Geophys Res 119(6):4412– 4440. https://doi.org/10.1002/2013JA019554 Figueira P, Santerne A, Suárez Mascareño A et al (2016) Is the activity level of HD 80606 influenced by its eccentric planet? A&A 592:A143. https://doi.org/10.1051/0004-6361/201628981 Fleck RC (2008) A magnetic mechanism for halting inward protoplanet migration: I. Necessary conditions and angular momentum transfer timescales. Ap&SS 313(4):351–356. https://doi.org/ 10.1007/s10509-007-9703-5 Fossati L, Ayres TR, Haswell CA et al (2013) Absorbing Gas around the WASP-12 Planetary System. ApJ 766(2):L20. https://doi.org/10.1088/2041-8205/766/2/L20 Goodman ML (1995) A three-dimensional, iterative mapping procedure for the implementation of an ionosphere-magnetosphere anisotropic Ohm’s law boundary condition in global magnetohydrodynamic simulations. Ann Geophys 13(8):843–853. https://doi.org/10.1007/s00585-9950843-z Grießmeier JM, Zarka P, Spreeuw H (2007) Predicting low-frequency radio fluxes of known extrasolar planets. A&A 475(1):359–368. https://doi.org/10.1051/0004-6361:20077397 Ip WH, Kopp A, Hu JH (2004) On the star-magnetosphere interaction of close-in exoplanets. ApJ 602(1):L53–L56. https://doi.org/10.1086/382274 Jia X, Walker RJ, Kivelson MG, Khurana KK Linker JA (2009) Properties of Ganymede’s magnetosphere inferred from improved three-dimensional MHD simulations. J Geophys Res 114(A9):n/a–n/a. https://doi.org/10.1029/2009JA014375 Jones CA (2011) Planetary magnetic fields and fluid dynamos. Annu Rev Fluid Mech 43(1):583– 614. https://doi.org/10.1146/annurev-fluid-122109-160727 Khodachenko ML, Shaikhislamov IF, Lammer H, Prokopov PA (2015) Atmosphere expansion and mass loss of close-orbit giant exoplanets heated by stellar Xuv. Ii. Effects of planetary magnetic

1854

A. Strugarek

field; structuring of inner magnetosphere. ApJ 813(1):50. https://doi.org/10.1088/0004-637X/ 813/1/50 Kopp A, Schilp S, Preusse S (2011) Magnetohydrodynamic simulations of the magnetic interaction of hot Jupiters with their host stars: a numerical experiment. ApJ 729(2):116. https://doi.org/10. 1088/0004-637X/729/2/116 Laine RO, Lin DNC (2012) Interaction of close-in planets with the magnetosphere of their host stars. II. Super-Earths as unipolar inductors and their orbital evolution. ApJ 745(1):2. https:// doi.org/10.1088/0004-637X/745/1/2 Laine RO, Lin DNC, Dong S (2008) Interaction of close-in planets with the magnetosphere of their host stars. I. Diffusion, ohmic dissipation of time-dependent field, planetary inflation, and mass loss. ApJ 685(1):521–542. https://doi.org/10.1086/589177 Lanza AF (2008) Hot Jupiters and stellar magnetic activity. A&A 487(3):1163–1170. https://doi. org/10.1051/0004-6361:200809753 Lanza AF (2009) Stellar coronal magnetic fields and star-planet interaction. A&A 505(1):339–350. https://doi.org/10.1051/0004-6361/200912367 Lanza AF (2012) Star-planet magnetic interaction and activity in late-type stars with close-in planets. A&A 544:23. https://doi.org/10.1051/0004-6361/201219002 Lanza AF (2013) Star-planet magnetic interaction and evaporation of planetary atmospheres. A&A 557:31. https://doi.org/10.1051/0004-6361/201321790 Lanza AF (2015) Star-planet interactions. 18th Cambridge workshop on cool stars, vol 18, pp 811–830 Lanza AF, Shkolnik EL (2014) Secular orbital evolution of planetary systems and the dearth of close-in planets around fast rotators. MNRAS 443(2):1451–1462. https://doi.org/10.1093/ mnras/stu1206 Llama J, Vidotto AA, Jardine M et al (2013) Exoplanet transit variability: bow shocks and winds around HD 189733b. MNRAS 436(3):2179–2187. https://doi.org/10.1093/mnras/stt1725 Lovelace RVE, Romanova MM, Barnard AW (2008) Planet migration and disc destruction due to magneto-centrifugal stellar winds. MNRAS 389(3):1233–1239. https://doi.org/10.1111/j.13652966.2008.13617.x Matsakos T, Uribe A, Königl A (2015) Classification of magnetized star-planet interactions: bow shocks, tails, and inspiraling flows. A&A 578:A6. https://doi.org/10.1051/0004-6361/ 201425593 McQuillan A, Mazeh T, Aigrain S (2013) Stellar rotation periods of the Kepler objects of interest: a dearth of close-in planets around fast rotators. ApJ 775(1):L11. https://doi.org/10.1088/20418205/775/1/L11 Mengel MW, Marsden SC, Carter BD et al (2016) A BCool survey of the magnetic fields of planethosting solar-type stars. MNRAS 465(3):2734–2747. https://doi.org/10.1093/mnras/stw2949 Merkin VG, Lyon JG (2010) Effects of the low-latitude ionospheric boundary condition on the global magnetosphere. J Geophys Res 115(A):A10,202. https://doi.org/10.1029/2010JA015461 Miller BP, Gallo E, Wright JT, Pearson EG (2015) A comprehensive statistical assessment of starplanet interaction. ApJ 799(2):163. https://doi.org/10.1088/0004-637X/799/2/163 Neubauer FM (1998) The sub-Alfvénic interaction of the Galilean satellites with the Jovian magnetosphere. J Geophys Res 103(E):19,843–19,866. https://doi.org/10.1029/97JE03370 Pillitteri I, Wolk SJ, Sciortino S, Antoci V (2014) No X-rays from WASP-18. Implications for its age, activity, and the influence of its massive hot Jupiter. A&A 567:A128. https://doi.org/10. 1051/0004-6361/201423579 Pont F (2009) Empirical evidence for tidal evolution in transiting planetary systems. MNRAS 396(3):1789–1796. https://doi.org/10.1111/j.1365-2966.2009.14868.x Poppenhaeger K, Wolk SJ (2014) Indications for an influence of hot Jupiters on the rotation and activity of their host stars. A&A 565:L1. https://doi.org/10.1051/0004-6361/201423454 Preusse S, Kopp A, Büchner J, Motschmann U (2005) Stellar wind regimes of close-in extrasolar planets. A&A 434(3):1191–1200. https://doi.org/10.1051/0004-6361:20041680 Preusse S, Kopp A, Büchner J, Motschmann U (2006) A magnetic communication scenario for hot Jupiters. A&A 460(1):317–322. https://doi.org/10.1051/0004-6361:20065353

90 Models of Star-Planet Magnetic Interaction

1855

Rubenstein EP, Schaefer BE (2000) Are superflares on solar analogues caused by extrasolar planets? ApJ 529(2):1031–1033. https://doi.org/10.1086/308326 Saur J, Neubauer FM, Strobel DF, Summers ME (1999) Three-dimensional plasma simulation of Io’s interaction with the Io plasma torus: asymmetric plasma flow. J Geophys Res 104(A):25,105–25,126. https://doi.org/10.1029/1999JA900304 Saur J, Grambusch T, Duling S, Neubauer FM, Simon S (2013) Magnetic energy fluxes in subAlfvénic planet star and moon planet interactions. A&A 552:119. https://doi.org/10.1051/00046361/201118179 Shkolnik EL, Bohlender DA, Walker GAH, Cameron AC (2008) The On/Off nature of star-planet interactions. ApJ 676(1):628–638. https://doi.org/10.1086/527351 Staab D, Haswell CA, Smith GD et al (2017) SALT observations of the chromospheric activity of transiting planet hosts: mass-loss and star–planet interactions. MNRAS 466(1):738–748. https://doi.org/10.1093/mnras/stw3172 Stanley S, Glatzmaier GA (2010) Dynamo models for planets other than earth. Space Sci Rev 152:617. https://doi.org/10.1007/s11214-009-9573-y Strugarek A (2016) Assessing magnetic torques and energy fluxes in close-in star–planet systems. ApJ 833(2):140. https://doi.org/10.3847/1538-4357/833/2/140 Strugarek A, Brun AS, Matt S (2012) On close-in magnetized star-planet interactions. In: Boissier S (eds) SF2A-2012: Proceedings of the annual meeting of the French society of astronomy and astrophysics, Montpelier, pp 419–423. https://ui.adsabs.harvard.edu/#abs/2012sf2a.conf..419S/ abstract Strugarek A, Brun AS, Matt SP, Réville V (2014a) Modeling magnetized star-planet interactions: boundary conditions effects. Nat Promin Role Space Weather 300:330–334. https://doi.org/10. 1017/S1743921313011162 Strugarek A, Brun AS, Matt SP, Réville V (2014b) Numerical aspects of 3D stellar winds. In: 18th Cambridge workshop on cool stars, stellar systems, and the sun, proccedings of Lowell Observatory, Flagstaff, vol 1410, p 3537 Strugarek A, Brun AS, Matt SP, Réville V (2014c) On the diversity of magnetic interactions in close-in star-planet systems. ApJ 795(1):86. https://doi.org/10.1088/0004-637X/795/1/86 Strugarek A, Brun AS, Matt SP et al (2014d) Modelling the corona of HD 189733 in 3D. Proceeding of the SFA conference, Cancun, vol 1411, p 2494 Strugarek A, Brun AS, Matt SP, Réville V (2015) Magnetic games between a planet and its host star: the key role of topology. ApJ 815(2):111. https://doi.org/10.1088/0004-637X/815/2/111 Tremblin P, Chiang E (2013) Colliding planetary and stellar winds: charge exchange and transit spectroscopy in neutral hydrogen. MNRAS 428(3):2565–2576. https://doi.org/10.1093/mnras/ sts212 Turner JD, Christie D, Arras P, Johnson RE, Schmidt C (2016a) Investigation of the environment around close-in transiting exoplanets using CLOUDY. MNRAS 458(4):3880–3891. https://doi. org/10.1093/mnras/stw556 Turner JD, Pearson KA, Biddle LI et al (2016b) Ground-based near-UV observations of 15 transiting exoplanets: constraints on their atmospheres and no evidence for asymmetrical transits. MNRAS 459(1):789–819. https://doi.org/10.1093/mnras/stw574 Vidotto AA, Jardine M, Morin J et al (2014) M-dwarf stellar winds: the effects of realistic magnetic geometry on rotational evolution and planets. MNRAS 438(2):1162–1175. https://doi.org/10. 1093/mnras/stt2265 Weber EJ, Davis LJ (1967) The angular momentum of the solar wind. ApJS 148:217. https://doi. org/10.1086/149138 Zanni C, Ferreira J (2009) MHD simulations of accretion onto a dipolar magnetosphere. I. Accretion curtains and the disk-locking paradigm. A&A 508:1117. https://doi.org/10.1051/ 0004-6361/200912879 Zarka P (2007) Plasma interactions of exoplanets with their parent star and associated radio emissions. Planet Space Sci 55(5):598–617. https://doi.org/10.1016/j.pss.2006.05.045

Stellar Coronal and Wind Models: Impact on Exoplanets

91

Aline A. Vidotto

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Observationally Derived Properties of Stellar Coronae . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Observationally Derived Properties of Stellar Winds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Models of Stellar Coronal Winds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stellar Wind Effects on Exoplanets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Effects on Potentially Habitable Planets Around M Dwarf Stars . . . . . . . . . . . . . . . . . . . . Effects on Potentially Habitable Planets Orbiting Young Stars . . . . . . . . . . . . . . . . . . . . . Final Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1857 1858 1861 1864 1867 1868 1869 1870 1870

Abstract

Surface magnetism is believed to be the main driver of coronal heating and stellar wind acceleration. Coronae are believed to be formed by plasma confined in closed magnetic coronal loops of the stars, with winds mainly originating in open magnetic field line regions. In this chapter, we review some basic properties of stellar coronae and winds and present some existing models. In the last part of this chapter, we discuss the effects of coronal winds on exoplanets.

Introduction About 90% of the currently known exoplanets orbit around low-mass stars. These stars (0:1 . M? =Mˇ . 1:3), while in the main-sequence phase, have convective interiors that vary in extension as a function of the stellar mass. Below 0:4Mˇ ,

A. A. Vidotto () School of Physics, Trinity College Dublin, The University of Dublin, Dublin-2, Ireland e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_26

1857

1858

A. A. Vidotto

these stars are fully convective. Above this mass threshold, there is an appearance of a radiative core, whose size is larger for more massive stars. In turn, the convective part of the star is limited to the outer layers and becomes progressively smaller as one goes toward more massive stars. At 1:3Mˇ , the outer convective envelope is already very small. As convection is one of the key ingredients in the generation of magnetic fields, main-sequence low-mass stars have surface magnetic fields. This magnetism gives rise to a multitude of phenomena, from small and localized features (spots, active regions, prominences) to large-scale ones (global magnetism, coronal holes, helmet streamers). Surface magnetism is also believed to be the main driver of coronal heating and stellar wind acceleration. However, at present, there is no consensus of the basic physical mechanisms involved in these processes. Even for the Sun, heating of the solar corona and acceleration of the solar wind are still currently being debated, with possible scenarios relating to propagation and dissipation of waves and turbulence in open magnetic flux tubes and/or reconnection between open and closed magnetic flux tubes (Cranmer 2009). In this chapter, we start by reviewing basic properties of stellar coronae and winds. We then present a review of some existing models. The last part of this chapter is dedicated to the impact of coronal winds on exoplanets.

Observationally Derived Properties of Stellar Coronae Low-mass stars harbor hot coronae with average temperatures on the order of 106 –107 K (Guedel 2004; Telleschi et al. 2005; Johnstone and Guedel 2015). The hot stellar coronae are detected in X-ray wavelengths (e.g., Pizzolato et al. 2003; Guedel 2004; Telleschi et al. 2005; Maggio et al. 2011; Wright et al. 2011; Scandariato et al. 2013; Pillitteri et al. 2014; Johnstone and Guedel 2015), during both quiescent and flaring states. Coronae are believed to be formed by plasma confined in closed magnetic coronal loops of the stars. An indication that coronae have indeed their origins in stellar magnetism comes from the observed correlation between X-ray emission and stellar magnetic fields (Pevtsov et al. 2003; Vidotto et al. 2014a). In this section, we highlight a few observed properties of stellar coronae. An interested reader will find comprehensive reviews of X-ray stellar coronae in, e.g., Guedel (2004), Guedel and Nazé (2009), and Testa et al. (2015). X-ray coronae and stellar rotation: Earlier studies have shown the connection between stellar rotation and chromospheric activity (Kraft 1967). Similarly, X-ray emission has also been recognized to correlate with stellar rotation, with the exception of fast-rotating stars (e.g., Pallavicini et al. 1981; Pizzolato et al. 2003; Jeffries et al. 2011; Wright et al. 2011; Reiners et al. 2014). For this reason, the rotation–activity relation is usually divided into two parts. Fast-rotating stars have X-ray emission that is roughly independent of rotation. They are in the so-called saturated regime. These stars have X-ray luminosities that account for about 0:1%

91 Stellar Coronal and Wind Models: Impact on Exoplanets

1859

of their bolometric luminosities. For slower rotators, in the unsaturated regime, X-ray luminosities increase with rotation rate ˝? as (Reiners et al. 2014) Lx / ˝?2:01˙0:05 :

(1)

The rotation rate at which stars transition from unsaturated to saturated regimes corresponds to (Johnstone and Guedel 2015)   ˝?;sat M? 1:08 ' 13:53 ; ˝ˇ Mˇ

(2)

where ˝ˇ D 2:67  106 rad s1 . The saturation threshold is mass dependent, with lower-mass stars transitioning from the saturated to the unsaturated regime at lower rotation rates. As rotation decreases with the square root of the age of stars (Skumanich 1972), stars in the lowest-mass range (e.g., M dwarfs) remain saturated even at relatively old ages (note also that these stars have longer lifetimes). Another observed link between rotation and X-ray emission is seen in X-ray light curves. Because X-ray emission arises in closed magnetic coronal loops and since the distribution of closed/open magnetic field line regions at the surface of the star is inhomogeneous, stars can also show rotational modulations in X-ray (Hussain et al. 2005, 2007). X-ray coronae and temperatures It has also been shown that stars with hot coronae have high X-ray emission (e.g., Telleschi et al. 2005). For low-mass mainsequence stars, there is a tight relation between X-ray flux Fx and average coronal temperature TQc (Johnstone and Guedel 2015) TQc Fx D 0:9 106 K

!3:8 erg cm2 s1 :

(3)

This empirical relation is useful for estimating the average coronal temperature of stars, once Fx is known. Fx can either be determined observationally or by using the rotation–activity relation (e.g., Eq. 1). As we will see in this chapter, the stellar wind temperature is an unknown in the models. Models that relate the temperature of the wind to the temperature of the corona can benefit from the empirical relation (3). X-ray coronae and magnetism The link between coronae and magnetism has long been identified. For this reason, X-ray emission is often used as a proxy for stellar magnetism. One way to validate this is by confronting observed values of X-ray luminosities/fluxes with observations of stellar magnetism. Two methods are mostly used to measure stellar magnetism. The Zeemaninduced line broadening of unpolarized light (Stokes I), or Zeeman broadening (ZB) technique (e.g., Solanki 1994; Saar 1996, 2001; Johns-Krull et al. 1999; Johns-Krull 2007; Reiners et al. 2009), yields estimates of the average of the total unsigned

1860

A. A. Vidotto

surface field strength hjBI ji (small- and large-scale structures). This technique does not provide information of the topology of the field. The Zeeman Doppler imaging (ZDI) technique (Stokes V), on the other hand, is able to reconstruct the intensity and topology of the stellar magnetic field (e.g., Donati and Brown 1997; Donati and Landstreet 2009; Morin et al. 2013), but cannot reconstruct the small-scale field component, which is missed within the resolution element of the reconstructed ZDI maps (Johnstone et al. 2010; Arzoumanian et al. 2011; Lang et al. 2014). As a consequence, the ZDI magnetic maps are limited to measuring large-scale magnetic field. Pevtsov et al. (2003) found that the X-ray luminosities are related to the unsigned magnetic fluxes ˚I measured by the ZB technique Lx / ˚I1:13˙0:05 ;

(4)

where ˚I D hjBI ji4R?2 : To derive this relation, Pevtsov et al. (2003) considered magnetic field observations of the Sun (X-ray bright points, active regions, quiet Sun, and integrated solar disk) and pre- and main-sequence stars. This empirical relation can be seen in Fig. 1a, spanning about 12 orders of magnitude in magnetic flux. Similarly, Vidotto et al. (2014a) found that Lx / ˚V0:913˙0:054 ;

(5)

where ˚V D hjBV ji4R?2 is the unsigned magnetic flux as derived from the ZDI technique (i.e., only contains the large-scale component of the stellar magnetic field). To be consistent with the method from Pevtsov et al. (2003), the relation above considers both main-sequence stars and pre-main-sequence (accreting) stars. The slope found by Vidotto et al. (2014a) is consistent to the nearly linear trend found by Pevtsov et al. (2003). Figure 1a (The data provided in Fig. 1a are from: Donati et al. (1999, 2003, 2008a, b, c, 2010a, b, 2011a, b, c, 2012, 2013), Marsden et al. (2006, 2011), Catala et al. (2007), Morin et al. (2008a, b, 2010), Petit et al. (2008, 2009), Hussain et al. (2009), Fares et al. (2009, 2010, 2012, 2013), Morgenthaler et al. (2011, 2012), Waite et al. (2011, 2015, 2017), do Nascimento et al. (2016), and Folsom et al. (2016) and from Petit et al. in prep.) shows that the pre-main-sequence stars (open circles) are under-luminous as compared to the empirical fit (solid line). When considering only the sample of 16 G, K and M dwarf stars (i.e., no solar data .dwarfs/ nor accreting PMS stars), Pevtsov et al. (2003) found that Lx / ˚I0:98˙0:19 .

91 Stellar Coronal and Wind Models: Impact on Exoplanets

1861

Fig. 1 (a) Relation between the X-ray luminosities and the unsigned magnetic fluxes ˚I measured by the Zeeman broadening technique (Pevtsov et al. 2003) (Reproduced by permission of the AAS). (b) The same as in (a) but for magnetic field measurements done with the Zeeman Doppler imaging technique. The different symbols mainly indicate different types of surveys. The solid line shows the empirical fit through the data, while the dashed line (at an arbitrary vertical offset) is indicative of the slope found by the Zeeman broadening technique, when considering the 16 G, K, M stars. From Vidotto et al. (2014 MNRAS, 441, 2361)

Considering the same types of objects, the relation derived from ZDI data yields .dwarfs/ / ˚V1:80˙0:20 (based on a larger sample of 61 dwarf stars). This is shown LX in Fig. 1b. Given the larger errors in the exponents of the fits, both relations are consistent to each other within 3 . Still, this is a topic worth of future investigation. For example, finding a different power law for ˚V and ˚I might clarify on how the small- and large-scale field structures contribute to X-ray emission.

Observationally Derived Properties of Stellar Winds Low-mass stars undergo mass loss through winds during their entire lives. Contrary to the Sun, in which the solar wind can be probed in situ, the existence of winds around low-mass stars is known indirectly, e.g., from the observed rotational evolution of stars (e.g., Bouvier et al. 2014). Measuring the wind mass-loss rates MP of cool, low-mass stars is challenging, as these winds are rarefied and difficult to be directly detected.

1862 Table 1 Characteristics of Proxima Centauri and its wind

A. A. Vidotto Physical property Mass (Mˇ ) Radius (Rˇ ) Rotation period (days) Fx (106 erg cm2 s1 ) TQc .106 K/ Spectral type Total magnetic flux (G) MP .MP ˇ D 2  1014 Mˇ yr1 / ... ...

Value 0:123 0:141 83 1:2 2:7 M5.5 600 1. If MA < 1, then pronounced Alfvén wings can form and no bow shock exists. Note if MA < 1, then MF < 1 as well. On average only once every two years the solar wind has such exotic low plasma densities that Earth looses its bow shock and forms Alfvén wings as well (e.g., Chané et al. 2012, 2017). The situation is significantly different at exoplanets, where a large number of them orbit their host star at radial distances less than 0.1 AU where the Alfvén Mach number is expected to be MA < 1, and SPI should thus be active because of the slow stellar winds.

Alfvén Wing Model The Alfvén wing model is a model which describes the farfield. It was originally developed to characterize Io’s interaction with the plasma of Jupiter’s magnetosphere. An extended set of analytical, numerical, and observational studies on this topic can be found in the literature (e.g., Goldreich and Lynden-Bell 1969; Neubauer 1980; Goertz 1980; Acuña and Ness 1980; Saur et al. 1999; Kivelson et al. 2004; Jacobsen et al. 2007). The Alfvén wing model has been applied to exoplanets in a number of studies (e.g., Ip et al. 2004; Preusse et al. 2005, 2006, 2007; Kopp et al. 2011). The main concept of the Alfvén wing model is that the interaction of the flowing plasma with the planetary obstacles slows the plasma in the planet’s vicinity to values v D .1  ˛/v N 0 (e.g., Goertz 1980; Neubauer 1980). Here we used the interaction strength parameter 0  ˛N  1 (Saur et al. 2013). The slowing occurs, e.g., through collisions of the plasma with the neutrals in the exoplanet’s atmosphere. For a sufficiently dense atmosphere, the flow is expected to be stagnant, and the interaction strength is maximum with ˛N  1. As the magnetic field is partly frozen into the flow, i.e., convected with the flow, the slowed flow bends the magnetic field lines and generates magnetic tensions. These tensions propagate along the magnetic field lines as Alfvén waves. The Alfvén waves propagate away from the exoplanets with group velocities strictly parallel and antiparallel to the background magnetic field. This property is the primary reason for the importance of the Alfvén mode in space and astrophysical plasmas. Due to this property, the Alfvén waves can carry energy confined within a magnetic flux tube and without dispersion over very large distances. In the farfield, the fast mode starts to have insignificant wave amplitude because it propagates away fairly isotropic, but the Alfvén wave amplitude stays finite. In this case, quantities called Elsässer variables or Alfvén characteristics c˙ A D v ˙ vA

(1)

with vA D vA .B=jBj/ are conserved quantities if the plasma is sufficiently smooth and the plasma beta, i.e., the ratio of thermal to magnetic pressure, is sufficiently

1886

J. Saur

small (Elsässer 1950; Neubauer 1980). The Elsässer variables relate the flow v and the magnetic field through vA . The plus and minus signs in expression (1) refer to the propagation parallel and antiparallel to the background field. The directions of the Alfvén wings are given by the directions of the Elsässer variables as indicated in Fig. 1. Expressions in (1) are exact solutions of the incompressible fully nonlinear MHD equations (Elsässer 1950; Neubauer 1980). When the strength of the slowed velocity near the exoplanet is known, this allows for an analytic description of the magnetic field in the Alfvén wings. The slowed velocity can be expressed through the electric field E D v  B under the assumption of the frozen-in-field theorem. With (1) and under the assumption of steady state and homogeneous background fields, the magnetic field B in the rest frame of the exoplanet is given by B? D .Oz  E/0 ˙A˙

(2)

q 2 Bz D ˙ B02  B?

(3)

and

with zO the unit vector along the c˙ directions and ? the direction perpendicular to A it (Neubauer 1980). The Alfvén conductance ˙A˙ is given by ˙A˙ D

1 N 1=2 0 vA .1 C MA2 ˙ 2MA cos /

(4)

where N is the angle between the flow v0 and the magnetic field B0 . The electric field E D r˚ can in this case be written by an electric potential ˚ which characterizes the slowdown of the plasma around the planet through (e.g., Neubauer 1998; Saur et al. 1999) N ˚ D E0 y .1  ˛/   ˚ D E0 y 1  ˛N .R=r/2

for r  R

(5)

for r R:

(6)

Here we used Alfvén wing coordinates where the z axis is along the wing, the y axis is in v0  B0 direction, and the x axis completes a right-handed coordinate system. The radius R represents the effective radius of the obstacle and r D .x 2 C y 2 /1=2 . Without intrinsic magnetic field, the effective radius is given by the radius of the planet including its atmosphere. If the exoplanet possesses an intrinsic magnetic field, the effective radius can be significantly larger or even smaller than the radius of the body depending on the orientation of the external magnetic field B0 and the internal magnetic moment Mexo . If both are parallel, the effective radius is maximum and the magnetosphere is open. In the antiparallel case, the magnetosphere is closed, and the effective radius is approximately zero. Quantitative expressions for the effective radius can be found in Saur et al. (2013) and Strugarek (2016).

92 Electromagnetic Coupling in Star-Planet Systems

1887

The Alfvén wings are bent back compared to the background magnetic field B0 by the angles A˙ with sin A˙ D

MA sin N : N 1=2 .1 C MA2 ˙ 2MA cos /

(7)

The Alfvén wings are invariant along the z direction. They carry electromagnetic energy given by the Poynting flux, which in ideal MHD is equal to S D .E  B/=0 D

B2 v? : 0

(8)

The Poynting flux thus describes transport of magnetic enthalpy carried by the flow perpendicular to the magnetic field v? . The local Poynting flux can be calculated with the expressions for E and B provided in this subsection. Note, the amplitude and the direction of the Poynting flux depend in an nonintuitive way on the frame of reference. The flux of kinetic and electric energy is negligible (Saur et al. 2013). The total Poynting flux in the wing is obtained by integration over a reference plane perpendicular to the background magnetic field in a rest frame moving with the star because we are interested in the energy flux arriving at the star (Fig. 3). In general the total Poynting flux cannot be written in closed analytic form. However for small Alfvén Mach number MA , the total Poynting flux can be approximated by Stotal D 2R2

N 2 .˛M N A B0 sin / vA : 0

(9)

Properties of the Alfvén Wings in SPI Here we discuss several dependencies of the Alfvén wing couplings. In Fig. 5 left, we display the estimated Alfvén Mach number near 850 exoplanets known until November 2012 (Saur et al. 2013). The stellar wind velocity was calculated with the Parker solar wind model (Parker 1958). The stellar properties were either taken from observed values provided on http://exoplanet.eu/, derived values based on the spectral classes, or were taken in analogy to the sun. The solid line in Fig. 5 left displays the average Alfvén Mach number in the solar wind as a function of radial distance from the sun. The figure shows that many exoplanets follow the same trend because many of the observed exoplanets orbit solar-type stars. Within 0.1 AU 77% of the exoplanets are subject to sub-Alfvénic interaction. Considering all distances, i.e., all of the observed 850 exoplanets, we find 35% of them in the subAlfvénic range. These numbers might however not fully reflect the true distribution due to the selection effects of the detection methods. In Fig. 5 right, we show the Poynting fluxes generated within the Alfvén wings at each exoplanet with MA < 1. The values of the Poynting fluxes vary by many

1888

MA: Alfv´en Mach number

103

n/a B A F G K M−

102

cA,r>0 solar system

101

100

10−1

10−2 10−2

10−1

100

101

102

103

104

rexo : star distance [AU] 1020 Stotal : total Poynting flux [W ]

Fig. 5 (Left) Estimated Alfvén Mach number MA as a function of radial distance for the sample of 850 planets known until November 2012. (Right) Total Poynting fluxes carried in Alfvén wings as a function of distance from the star for all 258 planets with sub-Alfvénic plasma interaction, which connect to the central star. The spectral classes of the exoplanet hosting stars are color-coded. The thin solid line represents MA in the solar wind of our solar system. Planets with MA < 1 that are marked with a cross generate two Alfvén wings, which both point away from the star and thus do not connect to the central star (Reproduced with permission ©ESO. Saur et al. 2013)

J. Saur

10

n/a B A F G K M 2 2 (1 + α ¯ ) sin ΘA>1

19

1018 1017 1016 1015 1014 1013 1012 10−2

10−1

100 101 102 103 rexo : star distance [AU]

104

orders of magnitude and reach values as high as a few times 1019 W. The largest Poynting fluxes are reached for exoplanets very close to their host star with rapidly decreasing values further out. The lowest Poynting fluxes are reached in a “corridor” between 0.1 and 0.3 AU. The reason is that in this region, the orbital velocities of the exoplanets are such that the relative velocity between the stellar wind and the exoplanets is aligned with the magnetic field embedded in the stellar wind. As a result, the motional electric field vanishes, and very little magnetic stress in the stellar wind is generated which propagates away as Alfvén waves (Zarka et al. 2001; Zarka 2007; Saur et al. 2013). Enhanced Poynting fluxes are again seen around 1.0 AU. These cases are related to particular large stars, large exoplanets,

92 Electromagnetic Coupling in Star-Planet Systems

1889

and large assumed planetary magnetic fields derived from a scaling law after Olson and Christensen (2006). Not all exoplanets which are embedded in sub-Alfvénic flows with MA < 1 can couple to their host star. A necessary conditions is that the radial stellar wind velocity is smaller than the radial component of the Alfvén velocity. In this case, one Alfvén wing is directed toward the host star as can be seen in Fig. 4 for a hypothetical planet located at a radial distance such that one wing slowly spirals in (shown in green). The exoplanet HD 11977 b possesses two Alfvén wings, but the radial velocities of the flow and the wave velocities are such that both wings point away from the star (shown in red). Figure 4 indicates another important aspect of the Alfvén wing model. The location where the wings intersect with the stellar atmosphere is not located on the same longitude as the exoplanet but shifted in phase similar to observational findings by Shkolnik et al. (2008). There are two reasons for this: (1) the stellar wind magnetic field is structured in an Archimedean spiral due to the rotation of the star (Parker 1958). A typical magnetic field line is shown in black in Fig. 4. (2) The Alfvén wings do not exactly follow the magnetic field lines but are bent back by an angle A˙ (see Equation (7)). The reason is that the waves travel along the magnetic field, which is at the same time carried away from the star by the stellar wind. The estimated Poynting flux for HD 179949 calculated with the Alfvén wing model renders 3  1017 W, which is about three orders of magnitude smaller compared to what has been derived from the observations (Shkolnik et al. 2005). Only with exotic properties of the star or of the exoplanet, e.g., if it had a surface magnetic field strength 4000 times larger than that of Jupiter, Poynting fluxes on the order of 1020 W can be reached. Thus in order to reach such large values, other mechanisms which release free energy and which might be triggered by the Alfvénic interaction are needed. Such mechanisms are discussed in the following subsection.

Models of Stress Release in Magnetic Loops The electromagnetic SPI has been investigated analytically with alternative models by Lanza (2008, 2009, 2012, 2013, 2015). In these models, it was assumed that the plasma and magnetic field environment of both the star and the planet are static and that the magnetic field is force-free, i.e., j  B D 0, which is possible either if the magnetic field is a potential field, i.e., j D 0, or if the electric current density j D r  B=0 is parallel to the magnetic field. The latter implies r  B D ˛B, with ˛ being constant along an individual field line. Under the assumptions of a current system connecting the star with the exoplanet, i.e., ˛ ¤ 0, and no current closed near the surface of the magnetized exoplanet, the electric current system can still close near the magnetopause of the exoplanet’s magnetosphere through reconnection. Lanza (2009) estimates that the energy dissipated in the reconnection near the magnetopause is given by

1890

J. Saur

Pd '

 2 .B? .a//2 Rm vorb 

(10)

with B? .a/ the stellar magnetic field at the location a of the exoplanet, Rm the radius of the magnetosphere, and 0 < < 1 is an efficiency factor that depends, e.g., on the orientation of both magnetic fields. Since the stellar properties are assumed static, the the relative velocity v0 in this model is given by the orbital velocity vorb of the exoplanet with respect to the stellar plasma and magnetic field environment. Applying typical properties for HD 179949, Lanza (2009) finds values for Pd on the order of 1017 W, which is also not sufficient to explain the energies derived from observations by Shkolnik et al. (2005). To overcome the missing source of energy flux, Lanza (2009, 2013) investigate the role of free magnetic energy in the form of magnetic helicity (i.e., from twisted field lines) or magnetic stresses in general. One suggestion by Lanza (2009) is that the motion in the stellar photosphere and the emergence of magnetic flux from the stellar convection zone power the buildup of stresses and helicity in the coronal magnetic field. This buildup of the free energy is independent of the presence of exoplanets, but the release of energy stored in the coronal loops is triggered by close-in exoplanets. Lanza (2009) estimates that this energy release for the case of HD 179949 can be on the order of 1020 to 1021 W. The buildup time for the stored energy in the loop is estimated to be on the order of a few times 104 s, which is shorter than the synodic rotation period of the host star as seen from the exoplanet and the orbital period of the exoplanet. An alternative model to explain the large SPI energy fluxes has been proposed by Lanza (2013). He suggests that constant stress accumulation and release in the flux tubes connecting the planet and the star (e.g., as displayed in Fig. 3 right) are being generated by the motion of the magnetized planets within the stellar field. The related energy fluxes in this model are equal to the average Poynting fluxes across the base of the flux tubes P D

2 2 fAP RP2 jE  Bj ' fAP RP2 BP2 vorb : 0 0

(11)

The factor fAP characterizes the fraction of the 2RP2 area which the flux loop penetrates on the surface of the exoplanet with radius RP . The model assumes that the flow is not modified, e.g., through an ionosphere, and thus E D vorb  B was applied. With typical values for close-in giant exoplanets, Lanza (2013) estimates P  1020  1021 W. A reduced flow velocity would reduce these numbers. This energy flux ultimately comes from the mechanical energy of the orbital motion of the exoplanets and not from the processes within the star as in the model of Lanza (2009). The dissipation of the energy is expected to occur over most of the volume of the loop that connects the planet and the star when a turbulent photospheric flow is assumed (Lanza 2013). This model can also account for a phase shift between orbital longitude of the exoplanet and those of the stellar hot spot in a twisted forcefree coronal field.

92 Electromagnetic Coupling in Star-Planet Systems

1891

Note, the Alfvén wing model turns into a force-free model, i.e., the classical unipolar inductor model of Goldreich and Lynden-Bell (1969), if the Alfvén travel time between the exoplanet and the star is much shorter than the travel time of the plasma past the exoplanet. In this case, the Alfvén waves can travel multiple times between the exoplanet and the star (e.g., Goldreich and Lynden-Bell 1969; Neubauer 1998).

Numerical Models Electromagnetic SPI has also been investigated with a series of numerical models, which we describe here only very briefly and mostly refer to the original work. The modeling work can be subdivided in two classes which focus on the local or the farfield interaction. The local interaction has been investigated with MHD models, for example, by Ip et al. (2004), Preusse et al. (2007), Strugarek et al. (2015), and Strugarek (2016). Based on these models, the authors investigate quantitatively the magnetospheric structures around the exoplanets and on how open or closed the magnetospheres are. These calculations provide subsequent constraints on the farfield and the impacts on the host star, such as total energy fluxes. Numerical simulations are often applied to describe the local interaction due to the large spatial gradients and the nonlinear coupling of all MHD modes near the exoplanets. The farfield interaction was studied, for example, by Preusse et al. (2005, 2006) and with the first full 3D simulations of the coupling being performed by Cohen et al. (2009, 2011). These models investigate how the Alfvén wings evolve in the inhomogeneous stellar winds and how coronal plasmas and the magnetic field environments of the host stars are modified by the electromagnetic effects of the exoplanets. Another series of models focus on the evolution of the stellar winds and the resulting space plasma environment convected to the exoplanets (e.g., Vidotto et al. 2015; Alvarado-Gómez et al. 2016; Cohen 2017).

Cross-References  Models of Star-Planet Magnetic Interaction  Planet and Star Interactions: Introduction  Signatures of Star-Planet Interactions  Star-Planet Interactions in the Radio Domain: Prospect for Their Detection  Stellar Coronal and Wind Models: Impact on Exoplanets

References Acuña MH, Ness NF (1980) The magnetic field of Saturn – Pioneer 11 observations. Science 207:444–446 Alvarado-Gómez JD, Hussain GAJ, Cohen O et al (2016) Simulating the environment around planet-hosting stars. II. Stellar winds and inner astrospheres. A&A 594:A95

1892

J. Saur

Baumjohann W, Treumann RA (1996) Basic space plasma physics. Imperial College Press, London Bonfond B, Grodent D, Gérard J et al (2008) UV Io footprint leading spot: a key feature for understanding the UV Io footprint multiplicity? Geophys Res Lett 35:L05,107 Chané E, Saur J, Neubauer FM, Raeder J, Poedts S (2012) Observational evidence of Alfvén wings at the Earth. J Geophys Res (Space Phys) 117(A16):A09217 Chané E, Saur J, Poedts S, Keppens R (2017) How is the Jovian Main auroral emission affected by the solar wind? J Geophys Res (Space Phys) 122:2016JA023,318 Clarke JT, Ajello J, Ballester GE et al (2002) Ultraviolet emissions from the magnetic footprints of Io, Ganymede and Europa on Jupiter. Nature 415:997–1000 Cohen O (2017) A comparison between physics-based and polytropic MHD models for stellar coronae and stellar winds of solar analogs. ApJ 835:220 Cohen O, Drake JJ, Kashyap VL et al (2009) Interactions of the magnetospheres of stars and closein giant planets. ApJ 704:L85–L88 Cohen O, Kashyap VL, Drake JJ et al (2011) The dynamics of stellar coronae harboring hot Jupiters. I. A time-dependent magnetohydrodynamic simulation of the interplanetary environment in the HD 189733 planetary system. ApJ 733:67 Connerney JEP, Baron R, Satoh T, Owen T (1993) Images of excited HC 3 at the foot of the Io flux tube in Jupiter’s atmosphere. Science 262(5316):1035–1038 Cuntz M, Saar SH, Musielak ZE (2000) On stellar activity enhancement due to interactions with extrasolar giant planets. ApJ 533:L151–L154 Elsässer W (1950) The hydromagnetic equations. Phys Rev 79:183 Goertz CK (1980) Io’s interaction with the plasma torus. J Geophys Res 85(A6):2949–2956 Goldreich P, Lynden-Bell D (1969) Io, a Jovian unipolar inductor. Astrophys J 156:59–78 Gurdemir L, Redfield S, Cuntz M (2012) Planet-induced emission enhancements in HD 179949: results from McDonald observations. PASA 29:141–149 Ip WH, Kopp A, Hu J (2004) On the star-magnetosphere interaction of close-in exoplanets. Astrophys J 602:L53–L56 Jacobsen S, Neubauer FM, Saur J, Schilling N (2007) Io’s nonlinear MHD-wave field in the heterogeneous Jovian magnetosphere. Geophys Res Lett 34:L10,202. https://doi.org/10.1029/ 2006GL029187 Kivelson MG, Bagenal F, Neubauer FM et al (2004) Magnetospheric interactions with satellites, Chap. 21. In: Bagenal F (ed) Jupiter. Cambridge University Press/University of Colorado, Cambridge, pp 513–536 Kopp A, Schilp S, Preusse S (2011) Magnetohydrodynamic Simulations of the magnetic interaction of hot Jupiters with their host stars: a numerical experiment. Astrophys J 729:116 Lanza AF (2008) Hot Jupiters and stellar magnetic activity. Astron Astrophys 487:1163–1170 Lanza AF (2009) Stellar coronal magnetic fields and star-planet interaction. A&A 505:339–350 Lanza AF (2012) Star-planet magnetic interaction and activity in late-type stars with close-in planets. A&A 544:A23 Lanza AF (2013) Star-planet magnetic interaction and evaporation of planetary atmospheres. A&A 557:A31 Lanza AF (2015) Star-planet interactions. In: van Belle GT, Harris HC (eds) 18th Cambridge workshop on cool stars, stellar systems, and the sun, Cambridge Workshop on Cool Stars, Stellar Systems, and the Sun, vol 18, pp 811–830 Neubauer FM (1980) Nonlinear standing Alfvén wave current system at Io: theory. J Geophys Res 85(A3):1171–1178 Neubauer FM (1998) The sub-Alfvénic interaction of the Galilean satellites with the Jovian magnetosphere. J Geophys Res 103(E9):19,843–19,866 Olson P, Christensen UR (2006) Dipole moment scaling for convection-driven planetary dynamos. Earth Planet Sci Lett 250:561–571 Parker EN (1958) Dynamics of the interplanetary gas and magnetic fields. ApJ 128:664 Pillitteri I, Maggio A, Micela G et al (2015) FUV variability of HD 189733. Is the star accreting material from its hot Jupiter? ApJ 805:52

92 Electromagnetic Coupling in Star-Planet Systems

1893

Poppenhaeger K, Schmitt JHMM (2011) A correlation between host star activity and planet mass for close-in extrasolar planets? ApJ 735:59 Poppenhaeger K, Robrade J, Schmitt J (2010) Coronal properties of planet-bearing stars. Astron Astrophys 515:A98 Poppenhaeger K, Lenz LF, Reiners A, Schmitt JHMM, Shkolnik E (2011) A search for star-planet interactions in the + Andromedae system at X-ray and optical wavelengths. Astron Astrophys 528:A58+ Preusse S, Kopp A, Büchner J, Motschmann U (2005) Stellar wind regimes of close-in extrasolar planets. Astron Astrophys 434:1191–1200 Preusse S, Kopp A, Büchner J, Motschmann U (2006) A magnetic communication scenario for hot jupiters. Astron Astrophys 460:317–322 Preusse S, Kopp A, Büchner J, Motschmann U (2007) MHD simulation scenarios of the stellar wind interaction with Hot Jupiter magnetospheres. Plant Space Sci 55:589–597 Saur J, Neubauer FM, Strobel DF, Summers ME (1999) Three-dimensional plasma simulation of Io’s interaction with the Io plasma torus: asymmetric plasma flow. J Geophys Res 104(A11):25,105–25,126 Saur J, Grambusch T, Duling S, Neubauer FM, Simon S (2013) Magnetic energy fluxes in sub-Alfvénic planet star and moon planet interactions. Astron Astrophys 552:A119. doi:10.1051/0004-6361/201118179 Scharf CA (2010) Possible constraints on exoplanet magnetic field strengths from planet-star interaction. Astrophys J 722:1547–1555 Shkolnik E, Walker GAH, Bohlender DA (2003) Evidence for planet-induced chromospheric activity on HD 179949. ApJ 597:1092–1096 Shkolnik E, Walker GAH, Bohlender DA, Gu P, Kürster M (2005) Hot Jupiters and hot spots: the short- and long-term chromospheric activity on stars with giant planets. ApJ 622:1075–1090 Shkolnik E, Bohlender DA, Walker GAH, Collier Cameron A (2008) The on/off nature of starplanet interactions. ApJ 676:628–638 Strugarek A (2016) Assessing magnetic torques and energy fluxes in close-in star-planet systems. ApJ 833:140 Strugarek A, Brun AS, Matt SP, Réville V (2015) Magnetic games between a planet and its host star: the key role of topology. ApJ 815:111 Vidotto AA, Fares R, Jardine M, Moutou C, Donati JF (2015) On the environment surrounding close-in exoplanets. MNRAS 449:4117–4130 Zarka P (2007) Plasma interactions of exoplanets with their parent star and associated radio emissions. Plant Space Sci 55:598–617 Zarka P, Treumann RA, Ryabov BP, Ryabov VB (2001) Magnetically-driven planetary radio emissions and applications to extrasolar planets. Astrophys Space Sci 277:293–300

Accretion of Planetary Material onto Host Stars

93

Brian Jackson and Joleen Carlberg

Contents Where Are Planetary Bodies Disrupted? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . How Are Planetary Bodies Disrupted? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Roche-Lobe Overflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Disruption Below the Stellar Photosphere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . When Are Planetary Bodies Disrupted? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . During the Pre- and Main Sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . During the Post-main Sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . In the Stellar Graveyard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1896 1899 1899 1901 1902 1902 1905 1907 1908

Abstract

Accretion of planetary material onto host stars may occur throughout a star’s life. Especially prone to accretion, extrasolar planets in short-period orbits, while relatively rare, constitute a significant fraction of the known population, and these planets are subject to dynamical and atmospheric influences that can drive significant mass loss. Theoretical models frame expectations regarding the rates and extent of this planetary accretion. For instance, tidal interactions between planets and stars may drive complete orbital decay during the main sequence. Many planets that survive their stars’ main sequence lifetime will still be engulfed when the host stars become red giant stars. There is some observational evidence supporting these predictions, such as a dearth of close-

B. Jackson () Department of Physics, Boise State University, Boise, ID, USA e-mail: [email protected] J. Carlberg Instruments Division, Space Telescope Science Institute, Baltimore, MD, USA e-mail: [email protected]; [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_28

1895

1896

B. Jackson and J. Carlberg

in planets around fast stellar rotators, which is consistent with tidal spin-up and planet accretion. There remains no clear chemical evidence for pollution of the atmospheres of main sequence or red giant stars by planetary materials, but a wealth of evidence points to active accretion by white dwarfs. In this article, we review the current understanding of accretion of planetary material, from the preto the post-main sequence and beyond. The review begins with the astrophysical framework for that process and then considers accretion during various phases of a host star’s life, during which the details of accretion vary, and the observational evidence for accretion during these phases.

Where Are Planetary Bodies Disrupted? Owing to detection biases, exoplanet discoveries skew heavily toward short-period or close-in planets, as shown in Fig. 1, and the nearest a planet can orbit its star and still remain intact depends largely on its self-gravity and, for smaller bodies, its tensile strength, friction, and/or internal viscosity. The Roche potential plays a key role either way, and a considerable literature has developed around the Roche potential (Kopal 1959; Paczy´nski 1971; Lai et al. 1993; Murray and Dermott 1999). The Roche potential is the potential field around a gravitating binary system in a bound orbit (Fig. 2) and is cast in a frame that rotates with the binary and so includes a centrifugal term. The potential is dominated by the planet’s gravity near the planet, but moving radially outward, the planet’s gravity drops off until the potential reaches a local maximum, which corresponds to zero net acceleration. The corresponding surface around the planet is defined as the Roche lobe (Murray and Dermott 1999). Such a surface also surrounds the star.

Fig. 1 Masses and separation of known exoplanets (Han et al. 2014). Blue symbols denote solar system planets. Vertical lines show the size of the stellar radius at various stages of the post-main sequence for a Sun-like star

93 Accretion of Planetary Material onto Host Stars

1897

Fig. 2 The Roche potential. The colored grid shows a three-dimensional representation of the potential surface in the x-y plane. Contours at the bottom show a projection of isopotentials onto the x-y plane. The Lagrange point L1 is labeled. For this figure, q D 0:2 and distances are measured in solar radii RSun , with the two masses separated by 0:718 RSun (Courtesy of René Heller)

The coordinate reference frame revolves with the system and is centered on the planet with the x-axis pointing from the planet to the star (which are separated by a fixed orbital distance a) and the z-axis pointing parallel to the orbital angular momentum. Approximating the two bodies as point masses, the effective potential (gravitational + centrifugal) field ˚ is ˚ D

h i GMp GMs 1   ˝ 2 .x  xcm /2 C y 2 ; jrj jaxO  rj 2

(1)

where G is the gravitational constant, Ms =Mp the stellar/planetary mass, r the location at which to evaluate ˚, and xO a unit vector pointing along x. The system’s  center of mass is at xcm D a Mp = Ms C Mp . ˝ is the orbital mean motion, and ˝ 2 D G Mp C Ms =a3 : The Roche lobe is not spherical, and there is no general closed-form expression for that surface. However, Eggleton (1983) provided a formula for the volumeequivalent radius of the Roche lobe RRL , i.e., the radius of a sphere with a volume equivalent to the given Roche lobe’s, accurate to 1% for all mass ratios q D Mp =Ms :

1898

B. Jackson and J. Carlberg

RRL D

0:49 q 2=3  a;  0:6 q 2=3 C ln 1 C q 1=3

(2)

RRL  0:49 q 1=3 a:

(3)

and for q 1

For planets (and stars) composed of slowly moving fluid, contours of ˚ coincide closely with density contours, and so the surface of a planet just filling its Roche lobe will correspond very nearly to the Roche lobe. To decide whether a planet is losing mass, we can compare the planet’s mean radius Rp to RRL . Incorporating Kepler’s third law, Rappaport et al. (2013) recast Eq. 3 to estimate the orbital period for the Roche limit, PRL , from the corresponding semimajor axis, aRL , for a planet with a bulk density p : s PRL 

3 .0:49/3 G p

  9:6 h

p 1; g cm3

1=2 :

(4)

Figure 3 compares the current periods P to PRL for several short-period planets. Importantly, the classical Roche limit here involves a number of assumptions, including treating the planet and star as point masses, assuming a circular orbit (Jackson et al. 2008) and synchronized planetary rotation (Showman and Guillot 2002), neglecting the effects of additional bodies (Van Laerhoven and Greenberg 2012; Hansen and Zink 2015), and neglecting tensile strength and friction (Richardson et al. 1998; Davidsson 1999).

Fig. 3 Roche limit periods PRL (Eq. 4) vs. orbital periods P for many short-period planets. Red triangles indicate planets for which tidal interactions with the host star may bring a planet with zero orbital eccentricity into its Roche limit in less than 1 Gyr (assuming a modified tidal dissipation parameter for the star Qs0 D 107 – Penev et al. 2012) – see “Tides in Star-Planet Systems” chapter

93 Accretion of Planetary Material onto Host Stars

1899

How Are Planetary Bodies Disrupted? Once some portion of the planet’s surface or atmosphere contacts the Roche lobe, it will begin losing mass (Roche-lobe overflow or RLO). Whether RLO proceeds stably or not depends on how the planet and its orbit respond to the mass loss. For dense planets with aRL below the stellar photosphere, other processes determine how the planet is disrupted.

Roche-Lobe Overflow For short-period exoplanets, tidal interactions with the host star usually transfer orbital angular momentum to the star’s rotation and reduce the semimajor axis. Even for zero eccentricity, tides can bring hot Jupiters (P  days) into Rochelobe contact within Gyrs (Jackson et al. 2008). Smaller planets raise smaller tides, driving slower orbital decay, but Earth-sized, rocky planets have been found with P  h and may also be accreted in Gyrs (Sanchis-Ojeda et al. 2013; Jackson et al. 2013; Sanchis-Ojeda et al. 2014; Adams et al. 2016). Material leaks out primarily through the L1 Lagrange point between the planet and the star. As the mass escapes, conservation of angular momentum can form a thin accretion disk around the star. Viscous stresses within the disk and tidal interactions with the disrupting planet can transfer some of the disk’s angular momentum back to the planet, driving the material inward (Ritter 1988). Torques between the planet and disk act in the opposite direction as the torque from the stellar tide (Lin and Papaloizou 1979), but mass loss would stop if the planet stopped filling its Roche lobe. Consequently, the disk torque cannot drive the planet beyond the Roche limit. However, if the tide raised on the host star pulled the planet inward through the Roche limit, the mass loss would increase (Ritter 1988), increasing the mass in the accretion disk and the strength of the disk’s torque. The resulting balance, stable RLO (Priedhorsky and Verbunt 1988), keeps the planet’s radius Rp  RRL . If the planet’s density drops as it loses mass, PRL increases (Eq. 4), and the gas disk would drive the planet out with the Roche limit (Valsecchi et al. 2015; Jackson et al. 2016). This balance relates the planet’s orbital evolution to its mass and radius evolution: MP p Mp

!

"    #  @ ln Rp 1 1 1 @a C D  : 2 a @t star @t Mp

"     #    @ ln Rp 1 @a aP 1 1 D  C .1  ı / ; a 3 a @t star @t Mp where Mp Ms .

(5)

(6)

1900

B. Jackson and J. Carlberg

pHere,  D @ ln Rp =@ ln Mp . The escaping gas has a specific angular momentum GMs a, but the angular momentum is not necessarily returned to the planet. Gas may escape from the system, driven by stellar wind (Bloecker 1995), or angular momentum is transferred to the star during accretion. The parameter ı expresses the fraction of gas that does not return its angular momentum: ı D 0 means all is returned, while ı D 1 means none is. These parameters combine to give  D =2  ı C 5=6 (5=6 comes from the Roche lobe’s mass dependence). A vast literature exists that provides many formulations for tidal interactions in binary systems (see Ferraz-Mello et al. 2008 for a thorough review). One well-worn model (Goldreich and Soter 1966) gives 

@a @t

 D star

9 2



G Ms

1=2

Rs5 Mp 11=2 a ; Qs0

(7)

where Qs0 is the modified tidal dissipation parameter (Ferraz-Mello et al. 2008). Unfortunately, the details of tidal dissipation for stars and gaseous planets are not well understood, and predictions for Qs0 span a wide range (cf. Ogilvie 2014). Consequently, the timescales for tidal decay and RLO are poorly known, but the outcomes for stable RLO should be insensitive to the rates. With a mass-radius relation, Eqs. 5 and 6 can be integrated to model the evolution of a disrupting gaseous planet. Valsecchi et al. (2015) and Jackson et al. (2016) employed the Modules in Stellar Astrophysics (MESA) suite (Paxton et al. 2011, 2013, 2015) to model disrupting hot Jupiters with a variety of initial conditions, and Fig. 4 illustrates the results. For a hot Jupiter   0 (Fortney et al. 2007), so mass loss initially reduces p , increases PRL , and drives planets out. Eventually, mass loss removes most of the gaseous envelope, p and PRL drop, and the tide can draw the planet back in. The outcome depends on the mass of the planet’s solid core, not on initial conditions (assuming they allow for RLO): the smaller the core, the more orbital expansion. Jackson et al. (2016) found that the gaseous planets that reach their Roche limits can completely shed their atmospheres, resulting in significant orbital expansion (which slows disruption and may save the remnant from accretion) or rapid accretion of the planet by the star. Stable RLO requires a specific relationship between evolution of Rp and of RRL : ı < =2 C 5=6:

(8)

If this inequality is not satisfied, unstable RLO may result. For example, if  1 N of the mass escaping from the and   0, stability requires no more than 5/6 (= 0:83) planet be lost without returning its angular momentum. The loss of orbital angular momentum is inevitable – Metzger et al. (2012) pointed out that material accreted at the stellar surface carries non-negligiblepangular momentum which will not be returned to the orbit, translating into ı D pR? =aRL . For Jupiter orbiting the Sun, a D aRoche  0:01 AU  2 RSun , and ı D 1=2.

93 Accretion of Planetary Material onto Host Stars

1901

a

b

Fig. 4 Mass (red lines) and orbital (blue lines) evolution for hot Jupiter systems with initial periods P0 D 3 days and Qs0 D 105 and a variety of initial envelope masses Menv, 0 and core masses Mcore . The different linestyles indicate different planetary parameters. (a) Hot Jupiters with Menv, 0 D 0.3, 1, and 3 MJup and Mcore fixed at 10 MEarth . (b) Hot Jupiters with Menv, 0 D 1 MJup and Mcore D 1, 5, 10, and 30 MEarth . These calculations assume ı D 0, i.e., the orbital angular momentum is completely conserved (From Jackson et al. 2016 and used with permission)

Valsecchi et al. (2015) explored loss of escaped material and found instability was likely, while Jia and Spruit (2017) found rocky planets between 1 and 10 Earth masses MEarth are likely to undergo unstable mass transfer. During unstable RLO, the mass loss timescale is comparable to the orbital period, but what happens next is unclear. Valsecchi et al. (2015) speculated that the escaped material may form a torus orbiting the host star, and interactions with the denuded planet may eject the torus, driving the planet into its Roche limit or the star.

Disruption Below the Stellar Photosphere The photosphere for a Sun-like, main sequence star corresponds to the Roche limit for p D 8 g cm3 , denser than nearly all super-Earths found by the Kepler mission (Weiss and Marcy 2014). Planets orbiting lower density, red giant stars, though, may enter the stellar photosphere relatively unscathed – Fig. 1.

1902

B. Jackson and J. Carlberg

Several outcomes are possible for such planets. The most massive companions may unbind the stellar envelopes, but only for asymptotic giant branch stars, when stars reach their largest radii and thus lowest binding energy (Nordhaus and Blackman 2006). In smaller red giants, companions can (1) enter into a common envelope stage if Mp 5–20 MJup , with the final mass of the accreting companion dependent on the star’s envelope mass; (2) fully evaporate in the stellar envelope; or (3) impact the stellar core (Livio and Soker 1984). For the latter two cases, evaporation within the star will occur near the depth where the temperature of the stellar interior exceeds the virial temperature of the planet Tv (Siess and Livio 1999a): Tv D

GMp p mH ; kRp

(9)

where p is the planet’s mean molecular weight, mH hydrogen’s mass, and k Boltzmann’s constant. Aguilera-Gómez et al. (2016) modeled dissolution for a range of companion planets and brown dwarfs and stellar masses and metallicities and found that companions with Mp  15 MJup will evaporate in the stellar envelope but more massive companions would probably impact the stellar core.

When Are Planetary Bodies Disrupted? Planetary accretion may occur from the pre-main sequence through the white dwarf stage, and different physical processes drive the accretion in different stages. The composition and rotation of a host star and orbital architecture of a planetary system may be long-lived signatures of planetary accretion, although unequivocal evidence is lacking for many stages.

During the Pre- and Main Sequence In principle, accretion of planetary material changes a star’s composition, an idea originally invoked to explain the planet-metallicity correlation (Fischer and Valenti 2005). The stellar mass fraction Xi for an element i is given as (Siess and Livio 1999b) Xi D

XiCZ MCZ C Xiacc Macc ; Macc C MCZ

(10)

where “CZ” refers to the stellar convection zone and “acc” to the accreted material, and M is the associated total mass. The changes to the photospheric abundances depend on the amount of accreted material and how much its composition differs from the star’s. For hydrogendominated, giant planets, the least stellar-like compositions are found in their cores.

93 Accretion of Planetary Material onto Host Stars

1903

The convection zone in Sun-like stars makes up a small fraction of the stars’ masses, and so a little material can cause large local abundance enhancements. The fact that the planet-metallicity correlation does not depend on convective envelope mass strongly suggests the correlation does not arise from planetary accretion (Fischer and Valenti 2005). However, accretion during the pre- and main sequence may impart other chemical signatures. Recent work suggests the last gas accreted onto stars is depleted in planet-forming materials, which are probably locked-up in terrestrial planets or gas giant’s cores. Meléndez et al. (2009) and Ramírez et al. (2009) found that the absence of 4 M˚ of rocky material could explain a depletion in refractory elements in the Sun’s photosphere compared to solar twins, fueling research into abundance comparisons between binary stars for which only one hosts a planet. For example, the 16 Cyg binary system is composed of two solar analogs, and one hosts a planet (Cochran et al. 1997). The coevality and nearly identical masses of the stars make 16 Cyg ideal for measuring abundance differences, and Tucci Maia et al. (2014) found a telling deficiency in the planet-hosting star corresponding to 1.5–6 M˚ of missing metals. Planetary accretion may induce downward mixing (Théado and Vauclair 2012), effectively erasing expected abundance enhancements (Eq. 10), but propitiously, this mixing can also create a new signature. The convection zone is the only region in Sun-like stars where fragile elements, like lithium, can survive, and downward mixing carries those elements to hotter regions, where they are destroyed, permanently depleting their surface abundances. Deal et al. (2015) argued that this scenario could explain the discrepant lithium abundances in the 16 Cyg system if the planet-hosting star, which has less lithium, accreted less than an Earth’s mass of material. Some migration likely brings short-period planets in from where they form, and two types have been considered: gas disk migration and dynamical excitation, followed by tidal circularization. In later stages, planets may enjoy long-lived stability, but, for short-period planets, tidal interactions can drive significant orbital decay – Fig. 3. Additional evidence for planetary accretion may come from considering these dynamical processes.

Due to Gas Disk Migration Gas disk migration involves gravitational interactions between a planet and its disk and can drive significant orbital evolution during the disk’s lifetime, 10 Myrs (Bell et al. 2013; Trilling et al. 1998; Chambers 2009). Lin et al. (1996) suggested a young star’s magnetosphere might clear the protoplanetary disk several stellar radii from the star, allowing gas disk migration to deposit planets in orbits near the disk’s inner edge (Lin and Papaloizou 1979; Goldreich and Tremaine 1979). Coupling between the magnetosphere and disk may synchronize the star’s rotation to the innermost disk (Koenigl 1991); thus, a young star’s rotation period may indicate the orbital period of the disk’s inner edge and the stopping point for gas disk migration. Irwin and Bouvier (2009) found young stars’ rotation periods range from 0:16 MJup ; data from Santerne et al. (2016) are for planets with a radius in the approximate range 0.5–2 RJup . Rates are reported in planets per star per  log P D 0:23 (the range indicated by the horizontal error bars). Downward-pointing arrows indicate upper limits

(2008) studied planets with a minimum mass m in the range from 0.3–10 MJup and P from 2–2000 days. They fitted a power law of the form given by Eq. 5, finding ˛ D 0:31 ˙ 0:20 and ˇ D 0:26 ˙ 0:10, normalized such that 10.5% of Sun-like stars have such a planet. (Technically, Cumming et al. (2008) calculated the fraction of stars with planets, rather than the average number of planets per star. The value of C is 1:04  103 when mass is measured in MJup and period is measured in days.) They found the data to be equally well described by a distribution uniform in log P from 2–300 days (i.e., ˇ D 0), followed by a sharp increase by a factor of 4–5 for longer periods. The Kepler data are also consistent with the latter description (Santerne et al. 2016). This uptick in planet occurrence at long periods might be related to the location of the “snow line” in protoplanetary disks, which plays a role in the theory of giant planet formation via core accretion; beyond this line, there is enough snow (frozen volatiles) to pack onto a growing protoplanet and help it to achieve the critical mass for runaway accretion of hydrogen and helium gas (Pollack et al. 1996; Lecar et al. 2006). Metallicity The earliest Doppler surveys revealed the occurrence of giant planets with periods shorter than a few years to be a steeply rising function of the host star’s metallicity (Santos et al. 2003; Fischer and Valenti 2005). This too is widely interpreted as support for core accretion theory. The logic is that the rapid assembly of a massive solid core – an essential step in the theory – is easier to arrange in a metal-rich protoplanetary disk. Fischer and Valenti (2005) found their sample of

96 Planet Occurrence: Doppler and Transit Surveys

1955

Doppler-detected giant planets to be compatible with n / z2 , where z is the iron-tohydrogen abundance relative to the solar value. Most recently, Petigura et al. (2018) used Kepler data to determine the best-fitting parameters of

(P;z D

@2 N D C P ˛ zˇ : @ log P @ log z

(6)

For hot Jupiters, they found ˇ D 3:4 ˙ 0:9, a remarkably strong dependence. However, for companions more massive than 4 MJup , Santos et al. (2017) found the association with high metallicity to be much weaker or absent, suggesting that such objects do not form through core accretion. Schlaufman (2018) reached the same conclusion with a sample spanning a larger range of companion masses and went so far as to say that companions more massive than 10 MJup should not be considered planets. The metallicity effect is also weaker for planets smaller than Neptune (Buchhave et al. 2012), although for orbital periods shorter than about 10 days, even small planets are associated with elevated metallicity (Mulders et al. 2016; Wilson et al. 2018; Petigura et al. 2018). Hot Jupiters Easy to detect, but intrinsically uncommon, hot Jupiters have an occurrence rate of 0.5–1% for periods between 1 and 10 days. They are even rarer for periods shorter than 1 day (Howard et al. 2012; Sanchis-Ojeda et al. 2014). There is a 2 discrepancy between the rate of 0.8–1.2% measured in Doppler surveys (Wright et al. 2012; Mayor et al. 2011) and 0.6% measured using Kepler data (Howard et al. 2012; Petigura et al. 2018). This is despite the similar metallicity distributions of the stars that were searched (Guo et al. 2017). While we should never lose too much sleep over 2 discrepancies, it might be caused by misclassified stars and unresolved binaries in the Kepler sample (Wang et al. 2015). Jupiter analogs A perennial question is whether the solar system is typical or unusual in some sense. It is difficult to answer because the current Doppler and transit surveys are only barely sensitive to the types of planets found in the solar system: the inner planets are too small and the outer planets have periods that are too long. The most easily detected planet in an extraterrestrial Doppler survey would probably be Jupiter; hence, a few groups have tried to quantify the occurrence of solar-like systems by searching for Jupiter-like exoplanets. Wittenmyer et al. (2016) presented the latest effort, finding the occurrence rate to be 6:2C2:8 1:6 % for planets of mass 0.3–13 MJup with orbital distances from 3–7 AU and eccentricities smaller than 0.3. Of course the rate depends on the definition of “Jupiter analog,” a term without a precise meaning. The same problem arises when trying to measure the occurrence of “Earth-like” planets. Long-period giants Regarding the more general topic of wide-orbiting giant planets, Foreman-Mackey et al. (2016) measured the occurrence of “cold Jupiters” with periods ranging from 2–25 years, by searching the Kepler data for stars

1956

J. N. Winn

showing only one or two transits over 4 years. For planets with R D 0:4–1.0 RJup , they found

(R;P D

@2 n D 0:18 ˙ 0:07: @ log R @ log P

(7)

Integrating over the specified ranges of radius and period gives a total occurrence rate of 0:42 ˙ 0:16 planets per star. Bryan et al. (2016) studied the occurrence of long-period giants conditioned on the detection of a shorter-period giant. Using high-resolution imaging and longterm Doppler monitoring, they searched for wide-orbiting companions to 123 giant planets with orbital distances ranging from 0.01 to 5 AU. They found the occurrence of outer companions to be higher than would be predicted by extrapolating the power law of Cumming et al. (2008) to longer periods. They also found d n=d log P to decline with period, unlike the more uniform distribution observed for closerorbiting giant planets. The occurrence rate was .53 ˙ 5/% for outer companions of mass 1–20 MJup and orbital distance 5–20 AU. Other properties The giant planet population is distinguished by other features. Their orbits show a broad range of eccentricities (see, e.g., Udry and Santos 2007). Their occurrence seems to fall precipitously for masses above 10 MJup , at least for orbital distances shorter than a few AU. Because of this low occurrence, the mass range from 10–80 MJup is often called the “brown dwarf desert” (Grether and Lineweaver 2006; Sahlmann et al. 2011; Triaud et al. 2017). As mentioned earlier, the inhabitants of this desert are not strongly associated with high-metallicity stars, unlike Jovian-mass planets (Santos et al. 2017; Schlaufman 2018). Occasionally we find two giant planets in a mean-motion resonance (Wright et al. 2011). The rotation of the star can be grossly misaligned with the orbit of the planet, especially if the star is more massive than about 1.2 Mˇ (Triaud, this volume). These and other topics were reviewed recently by Winn and Fabrycky (2015) and Santerne (this volume).

Smaller Planets Overall occurrence About half of Sun-like stars have at least one planet with an orbital period shorter than 100 days and a size in between those of Earth and Neptune. Planet formation theories generally did not predict this profusion of closeorbiting planets. Indeed some of the most detailed theories predicted that close-in “super-Earth” or “sub-Neptune” planets would be especially rare (Ida and Lin 2008). Their surprisingly high abundance led to new theories in which small planets can form in short-period orbits, rather than forming farther away from the star and then migrating inward (see, e.g., Hansen and Murray 2012; Chiang and Laughlin 2013).

96 Planet Occurrence: Doppler and Transit Surveys

1957

Doppler surveys provided our first glimpse at this population of planets and then Kepler revealed it in vivid detail. For planets with periods shorter than 50 days and minimum masses between 3 and 30 M˚ , two independent Doppler surveys found the occurrence rate to be .15 ˙ 5/% (Howard et al. 2010) and .27 ˙ 5/% (Mayor et al. 2011). For this same period range and planets with a radius between 2 and 4 R˚ , analysis of Kepler data gave an occurrence rate of .13:0 ˙ 0:8/% (Howard et al. 2012). The results of these surveys are compatible, given reasonable guesses for the relation between planetary mass and radius (Howard et al. 2012; Figueira et al. 2012; Wolfgang and Laughlin 2012). Size, mass, and period The surveys also agree that within this range of periods and planet sizes, the occurrence rate is higher for the smallest planets, roughly according to power laws (Howard et al. 2010, 2012): dn dn / m0:5 ; / R2 : d log m d log R

(8)

For even smaller or longer period planets, Kepler provides almost all the available information. Figure 4 shows some of the latest results (see also Fressin et al. 2013; Burke et al. 2015). The period distribution d n=d log P rises as P 2 between 1 and 10 days, before leveling off to a nearly constant value between 10 and 300 days. Multiple-planet systems Small planets occur frequently in closely spaced systems (Mayor et al. 2011; Lissauer et al. 2011), with as many as eight planets with periods shorter than a year (Shallue and Vanderburg 2018). The period ratios tend to be in the neighborhood of 1.5–5 (Fabrycky et al. 2014). In units of the mutual Hill radius,  aH 

Min C Mout 3M?

1=3 

ain C aout 2

 ;

(9)

more relevant to dynamical stability, the typical spacing is 10–30 (Fang and Margot 2013). At the lower end of this distribution, the systems flirt with instability (Deck et al. 2012; Pu and Wu 2015). A few percent of the Kepler systems are in (or near) mean-motion resonances, suggesting that the orbits have been sculpted by planet-disk gravitational interactions. These systems offer the gift of transit-timing variations (Agol & Fabrycky, this volume), the observable manifestations of planetplanet gravitational interactions that sometimes allow for measurements of planetary masses as well as orbital eccentricities and inclinations. Such studies and some other lines of evidence show that the compact multiple-planet systems tend to have orbits that are nearly circular (Hadden and Lithwick 2014; Xie et al. 2016; Van Eylen and Albrecht 2015) and coplanar (Fabrycky et al. 2014). Radius gap The radius distribution of planets with periods shorter than 100 days shows a dip in occurrence between 1.5 and 2 R˚ (Fulton et al. 2017; Van Eylen et al. 2017, see Fig. 5). Such a feature had been anticipated based on theoretical

1958

J. N. Winn

16

FGK dwarfs (Petigura et al. 2018)

radius [R⊕]

8

hot

2

un ept

rt

ese

ed

4 N

1 low detection probability

16

M dwarfs (Dressing et al. 2015)

radius [R⊕]

8 4 2 1 low detection probability

1

10 orbital period [days]

100

Fig. 4 Planet occurrence around FGK dwarfs (top) and M dwarfs (bottom) based on Kepler data. The blue dots represent a random sample of planets around 103 stars, drawn from the occurrence rate densities of Petigura et al. (2018) and Dressing and Charbonneau (2015). Compared to FGK stars, the M stars have a higher occurrence of small planets and a lower occurrence of giant planets. For the M dwarfs, occurrence rates for planets larger than 4 R˚ were not reported because only four planet candidates in that range were detected

calculations of the photo-evaporation of the atmospheres of low-mass planets by the intense radiation from the host star (Owen and Wu 2013; Lopez and Fortney 2013). Thus, the radius gap or “evaporation valley” seems to be a precious example in exoplanetary science of a prediction fulfilled, with many implications for the structures and atmospheres of close-orbiting planets (Owen and Wu 2017). Hot Neptunes As mentioned above, d n=d log P changes from a rising function for P . 10 days to a nearly constant value for P D 10–100 days. The critical period separating these regimes is longer for larger planets. The effect is to create a diagonal boundary in the space of log R and log P , above which the occurrence is very low (see Fig. 4). The same phenomenon is seen in Doppler data (Mazeh et al. 2016). This “hot Neptune desert” may be another consequence of atmospheric erosion. Interestingly those few hot Neptunes that do exist are strongly associated with metal-rich stars (Dong et al. 2017; Petigura et al. 2018), making them similar to giant planets and unlike smaller planets. The hot Neptunes are also similar to hot Jupiters in that they tend not to have planetary companions in closely spaced

96 Planet Occurrence: Doppler and Transit Surveys

1959

Fig. 5 From Fulton et al. (2017). Occurrence as a function of radius based on Kepler data, for orbital periods shorter than 100 days. The dip in the occurrence rate density between 1.5 and 2 R˚ has been attributed to the erosion of planetary atmospheres by high-energy radiation from the star

coplanar orbits (Dong et al. 2017). All this suggests that the hot Neptunes and close-in giant planets originate in similar circumstances, possibly from some type of dynamical instability. Earth-like planets A goal with broad appeal is measuring the occurrence rate of Earth-sized planets orbiting Sun-like stars within the “habitable zone,” the range of distances within which a rocky planet could plausibly have oceans of liquid water. The Kepler mission provided the best-ever data for this purpose. However, even Kepler was barely sensitive to such planets. The number of detections was of order 10, depending on the definitions of “Earth-sized,” “Sun-like,” and “habitable zone.” The desired quantity can be understood as an integral Z

Rmax

Z

Smax

˚  Rmin

Smin

@2 n d log S d log R; @ log S @ log R

(10)

where S is the bolometric flux the planet receives from the star. The integration limits are chosen to select planets likely to have a solid surface with a temperature and pressure allowing for liquid water. These limits depend on assumptions about the structure and atmosphere of the planet and the spectrum of the star (see, e.g., Kasting et al. 2014). Even if we set aside the problem of setting the integration limits, the measurement of (˚ 

ˇ ˇ @2 n ˇ @ log P @ log R ˇP D1 yr; RDR˚

(11)

1960

J. N. Winn

0.01

0.03

3.16

Γ⊕

10.00

Youdin

Direct Baseline Trimmed Direct Baseline Extrapolated Baseline

Petigura

Foreman-Mackey

2

1

1.00

Dong

2.5

1.5

0.32

Optimistic Efficiency Pessimistic Efficiency Original KIC DV Rp Low Reliability High Reliability

3 Probability Density

0.10

0.5 0 -2

-1.5

-1

-0.5 0 Log10[Γ⊕]

0.5

1

Fig. 6 From Burke et al. (2015). Estimates for (˚ based on Kepler data. The orange histogram is the posterior probability distribution considering only the uncertainties from counting statistics and extrapolation. The other curves illustrate the effects of some systematic errors: uncertainty in the detection efficiency, orbital eccentricities, stellar parameters, and reliability of weak planet candidates. These systematic effects led to a range in (˚ spanning an order of magnitude. The vertical lines show other estimates of (˚ by Foreman-Mackey et al. (2014), Petigura et al. (2013), Dong and Zhu (2013), and Youdin (2011)

has proven difficult and may require extrapolation from measurements of larger planets at shorter periods. The Kepler team has published a series of papers reporting steady advancement in the efficiency of detection, elimination of false positives, and understanding of instrumental artifacts. The most recent effort to determine (˚ found the data to be compatible with values ranging from 0.04 to 11.5 (see Fig. 6). Since then the Kepler team and other groups have clarified the properties of the stars that were searched (Petigura et al. 2017; De Cat et al. 2015), and the most recent installments by Twicken et al. (2016) and Thompson et al. (2017) quantified the sensitivity of the algorithms for planet detection and validation. These developments have brought us right to the threshold of an accurate occurrence rate for Earth-like planets.

Other Types of Stars Almost all the preceding results pertain to main-sequence stars with masses between 0.5 and 1.2 Mˇ , i.e., spectral types from K to late F. Stars with masses between

96 Planet Occurrence: Doppler and Transit Surveys

1961

0.1 and 0.5 Mˇ , the M dwarfs, are not as thoroughly explored, especially near the low end of the mass range. However these stars are very attractive for planet surveys because their small masses and sizes lead to larger Doppler and transit signals and because planets in the habitable zone have conveniently short orbital periods. Giant planets are relatively rare around M dwarfs, at least for periods shorter than a few years. Cumming et al. (2008) showed that if planet occurrence is modeled by the functional form of Eq. 5, then planets with masses exceeding 0.4 MJup and periods shorter than 5.5 years are 3–10 times less common around M dwarfs than around FGK dwarfs. Similar results were obtained by Bonfils et al. (2013). On the other hand, for smaller planets over the same range of periods, M dwarf occurrence rates exceed those of FGK dwarfs by a factor of 2–3 (Howard et al. 2012; Mulders et al. 2015). This result is based on Kepler data, which remains our best source of information on this topic despite the fact that only a few percent of the Kepler target stars were M dwarfs. Comprehensive analyses have been performed by Dressing and Charbonneau (2015) and Gaidos et al. (2016). Their results differ in detail but agree that, on average, M dwarfs have about two planets per star with a radius in between those of Earth and Neptune and an orbital period shorter than 100 days (see Fig. 4). One implication of this high occurrence rate is that the nearest habitable-zone planets are almost surely around M dwarfs. Indeed, Doppler surveys have turned up two candidates for “temperate” Earth-mass planets around M dwarfs within just a few parsecs: Proxima Cen (1.3 pc, AngladaEscudé et al. 2016) and Ross 128 (also known as Proxima Vir; 3.4 pc, Bonfils et al. 2017). Some other comparisons with FGK dwarfs have been made. Among the similarities are that M dwarfs often have compact systems of multiple planets (Muirhead et al. 2015; Gaidos et al. 2016; Ballard and Johnson 2016) and that high metallicity is associated with giant planet occurrence (Johnson et al. 2010; Neves et al. 2013). It remains unclear whether or not the occurrence of smaller planets is associated with high metallicity for M dwarfs (see, e.g., Gaidos and Mann 2014). There is also evidence that the planet population around M dwarfs exhibits both the “evaporation valley” between 1.5 and 2 R˚ and the “hot Neptune desert” (Hirano et al. 2017). Beyond the scope of this review, but nevertheless fascinating, are the occurrence rates that have been measured in Doppler and transit surveys of other types of stars: evolved stars (Johnson et al. 2010; Reffert et al. 2015), stars in open clusters (Mann et al. 2017) and globular clusters (Gilliland et al. 2000; Masuda and Winn 2017), binary stars (Armstrong et al. 2014), brown dwarfs (He et al. 2017), and white dwarfs (Fulton et al. 2014; van Sluijs and Van Eylen 2018).

Future Prospects The data from recent surveys offer opportunities for progress for at least another few years. The ongoing struggle to measure the occurrence rate of Earth-like planets has already been described. Another undeveloped area is the determination of joint and

1962

J. N. Winn

conditional probabilities; for example, given a planet with radius R1 and period P1 , what is the chance of finding another planet around the same star with radius R2 and period P2 ? Conditional rates, or the relative occurrence of different types of systems, may be more useful than overall occurrence rates for testing planet formation theories. Only a few cases have been studied, such as the mutual radius distribution of neighboring planets (Ciardi et al. 2013; Weiss et al. 2018) and the probability for giant planets to have wider-orbiting companions (Huang et al. 2016; Bryan et al. 2016; Schlaufman and Winn 2016). New data are also forthcoming. The stellar parallaxes soon to be available from the ESA Gaia mission (Gaia Collaboration et al. 2016) will clarify the properties of all the Kepler stars as well as the targets of future surveys. Among these future surveys is the Transiting Exoplanet Survey Satellite (TESS), which was launched in April 2018 (Ricker et al. 2015). This mission was not designed to measure planet occurrence rates but rather to pluck low-hanging fruit: short-period transiting planets around bright stars. Less well appreciated is that TESS may be superior to Kepler for measuring the occurrence of planets larger than Neptune with periods shorter than 10 days. When TESS was conceived, it was expected that limitations in data storage and transmission would restrict the search to 105 preselected stars, as was the case with Kepler. Later it became clear that entire TESS images could be stored and transmitted with 30-min time sampling. As a result, although 105 Sun-like stars will still be selected for finer time sampling, it should be possible to search millions of stars for large and short-period planets. This includes hot and massive stars, for which comparatively little is known. TESS should excel at finding rare, large-amplitude, short-period photometric phenomena of all kinds. For smaller planets around Sun-like stars, it will be difficult to achieve an orderof-magnitude improvement over the existing data. There is more room for advances in the study of low-mass stars, using new Doppler spectrographs operating at farred and infrared wavelengths and ground-based transit surveys focusing exclusively on low-mass stars. Particularly encouraging was the discovery of TRAPPIST-1, a system of seven Earth-sized planets orbiting an “ultracool dwarf” that barely qualifies as a star (Gillon et al. 2017). This system was found after searching 50 similar objects with a detection efficiency of around 60% (Burdanov et al., this volume; M. Gillon, private communication), and the transit probability for the innermost planet is 5%. This suggests that the occurrence of such systems is approximately .50  0:6  0:05/1 D 0:7. Thus, while TRAPPIST-1 seems extraordinary, it may represent a typical outcome of planet formation around ultracool dwarfs. In the decades to come, the domains of all the planet detection techniques – including direct imaging, gravitational microlensing, and astrometry – will begin overlapping. Some efforts have already been made to determine occurrence rate densities based on data from very different techniques (see, e.g., Montet et al. 2014; Clanton and Gaudi 2016). We can look forward to a more holistic view of the occurrence of planets around other stars, barring any civilization-ending cataclysm.

96 Planet Occurrence: Doppler and Transit Surveys

1963

References Anglada-Escudé G, Amado PJ, Barnes J et al (2016) A terrestrial planet candidate in a temperate orbit around Proxima Centauri. Nature 536:437–440 Armstrong DJ, Osborn HP, Brown DJA et al (2014) On the abundance of circumbinary planets. MNRAS 444:1873–1883 Ballard S, Johnson JA (2016) The Kepler Dichotomy among the M dwarfs: half of systems contain five or more coplanar planets. ApJ 816:66 Bonfils X, Delfosse X, Udry S et al (2013) The HARPS search for southern extra-solar planets. XXXI. The M-dwarf sample. A&A 549:A109 Bonfils X, Astudillo-Defru N, Díaz R et al (2017) A temperate exo-Earth around a quiet M dwarf at 3.4 parsecs. ArXiv e-prints Borucki WJ (2016) KEPLER mission: development and overview. Rep Prog Phys 79(3): 036901 Bryan ML, Knutson HA, Howard AW et al (2016) Statistics of long period gas giant planets in known planetary systems. ApJ 821:89 Buchhave LA, Latham DW, Johansen A et al (2012) An abundance of small exoplanets around stars with a wide range of metallicities. Nature 486:375–377 Burke CJ, Christiansen JL, Mullally F et al (2015) Terrestrial planet occurrence rates for the Kepler GK dwarf sample. ApJ 809:8 Chiang E, Laughlin G (2013) The minimum-mass extrasolar nebula: in situ formation of close-in super-Earths. MNRAS 431:3444–3455 Ciardi DR, Fabrycky DC, Ford EB et al (2013) On the relative sizes of planets within Kepler multiple-candidate systems. ApJ 763:41 Clanton C, Gaudi BS (2016) Synthesizing exoplanet demographics: a single population of longperiod planetary companions to M dwarfs consistent with microlensing, radial velocity, and direct imaging surveys. ApJ 819:125 Cumming A (2004) Detectability of extrasolar planets in radial velocity surveys. MNRAS 354:1165–1176 Cumming A, Butler RP, Marcy GW et al (2008) The Keck planet search: detectability and the minimum mass and orbital period distribution of extrasolar planets. PASP 120:531 De Cat P, Fu JN, Ren AB et al (2015) Lamost observations in the Kepler field. I. Database of low-resolution spectra. ApJS 220:19 Deck KM, Holman MJ, Agol E et al (2012) Rapid dynamical chaos in an exoplanetary system. ApJ 755:L21 Dong S, Zhu Z (2013) Fast rise of “Neptune-size” planets (4-8 R ˚ ) from P = 10 to 250 days— statistics of Kepler planet candidates up to 0.75 AU. ApJ 778:53 Dong S, Xie JW, Zhou JL, Zheng Z, Luo A (2017) LAMOST telescope reveals that Neptunian cousins of hot Jupiters are mostly single offspring of stars that are rich in heavy elements. ArXiv e-prints Dressing CD, Charbonneau D (2015) The occurrence of potentially habitable planets orbiting M dwarfs estimated from the full Kepler dataset and an empirical measurement of the detection sensitivity. ApJ 807:45 Fabrycky DC, Lissauer JJ, Ragozzine D et al (2014) Architecture of Kepler’s multi-transiting systems. II. New investigations with twice as many candidates. ApJ 790:146 Fang J, Margot JL (2013) Are planetary systems filled to capacity? A study based on Kepler results. ApJ 767:115 Feynman RP (1963) Feynman lectures on physics, vol 1, Addison-Wesley, Boston Figueira P, Marmier M, Boué G et al (2012) Comparing HARPS and Kepler surveys. The alignment of multiple-planet systems. A&A 541:A139 Fischer DA, Valenti J (2005) The planet-metallicity correlation. ApJ 622:1102–1117 Foreman-Mackey D, Hogg DW, Morton TD (2014) Exoplanet population inference and the abundance of Earth analogs from noisy, incomplete catalogs. ApJ 795:64

1964

J. N. Winn

Foreman-Mackey D, Morton TD, Hogg DW, Agol E, Schölkopf B (2016) The population of longperiod transiting exoplanets. AJ 152:206 Fressin F, Torres G, Charbonneau D et al (2013) The false positive rate of Kepler and the occurrence of planets. ApJ 766:81 Fulton BJ, Tonry JL, Flewelling H et al (2014) A search for planetary eclipses of white dwarfs in the pan-STARRS1 medium-deep fields. ApJ 796:114 Fulton BJ, Petigura EA, Howard AW et al (2017) The California-Kepler survey. III. A gap in the radius distribution of small planets. AJ 154:109 Gaia Collaboration, Prusti T, de Bruijne JHJ et al (2016) The Gaia mission. A&A 595:A1 Gaidos E, Mann AW (2014) M dwarf metallicities and giant planet occurrence: ironing out uncertainties and systematics. ApJ 791:54 Gaidos E, Mann AW, Kraus AL, Ireland M (2016) They are small worlds after all: revised properties of Kepler M dwarf stars and their planets. MNRAS 457:2877–2899 Gilliland RL, Brown TM, Guhathakurta P et al (2000) A lack of planets in 47 Tucanae from a hubble space telescope search. ApJ 545:L47–L51 Gillon M, Triaud AHMJ, Demory BO et al (2017) Seven temperate terrestrial planets around the nearby ultracool dwarf star TRAPPIST-1. Nature 542:456–460 Grether D, Lineweaver CH (2006) How dry is the brown dwarf desert? quantifying the relative number of planets, brown dwarfs, and stellar companions around nearby sun-like stars. ApJ 640:1051–1062 Guo X, Johnson JA, Mann AW et al (2017) The metallicity distribution and hot Jupiter rate of the Kepler field: hectochelle high-resolution spectroscopy for 776 Kepler target stars. ApJ 838:25 Hadden S, Lithwick Y (2014) Densities and eccentricities of 139 Kepler planets from transit time variations. ApJ 787:80 Hansen BMS, Murray N (2012) Migration then assembly: formation of Neptune-mass planets inside 1 AU. ApJ 751:158 He MY, Triaud AHMJ, Gillon M (2017) First limits on the occurrence rate of short-period planets orbiting brown dwarfs. MNRAS 464:2687–2697 Hirano T, Dai F, Gandolfi D et al (2017) Planetary systems around low-mass stars unveiled by K2. ArXiv e-prints Howard AW, Marcy GW, Johnson JA et al (2010) The occurrence and mass distribution of close-in super-Earths, Neptunes, and Jupiters. Science 330:653 Howard AW, Marcy GW, Bryson ST et al (2012) Planet occurrence within 0.25 AU of solar-type stars from Kepler. ApJS 201:15 Huang C, Wu Y, Triaud AHMJ (2016) Warm Jupiters are less lonely than hot Jupiters: close neighbors. ApJ 825:98 Ida S, Lin DNC (2008) Toward a deterministic model of planetary formation. V. Accumulation near the ice line and super-Earths. ApJ 685:584–595 Johnson JA, Howard AW, Bowler BP et al (2010) Retired a stars and their companions. IV. Seven Jovian exoplanets from Keck observatory. PASP 122:701 Kasting JF, Kopparapu R, Ramirez RM, Harman CE (2014) Remote life-detection criteria, habitable zone boundaries, and the frequency of Earth-like planets around M and late K stars. Proc Natl Acad Sci 111:12641–12646 Lecar M, Podolak M, Sasselov D, Chiang E (2006) On the location of the snow line in a protoplanetary disk. ApJ 640:1115–1118 Lissauer JJ, Ragozzine D, Fabrycky DC et al (2011) Architecture and dynamics of Kepler’s candidate multiple transiting planet systems. ApJS 197:8 Lopez ED, Fortney JJ (2013) The role of core mass in controlling evaporation: the Kepler radius distribution and the Kepler-36 density dichotomy. ApJ 776:2 Lovis C, Fischer D (2010) Radial velocity techniques for exoplanets. University of Arizona Press, Tucson, pp 27–53 Mann AW, Gaidos E, Vanderburg A et al (2017) Zodiacal exoplanets in time (ZEIT). IV. Seven transiting planets in the Praesepe cluster. AJ 153:64

96 Planet Occurrence: Doppler and Transit Surveys

1965

Masuda K, Winn JN (2017) Reassessment of the null result of the HST search for planets in 47 Tucanae. AJ 153:187 Mayor M, Marmier M, Lovis C et al (2011) The HARPS search for Southern extra-solar planets XXXIV. Occurrence, mass distribution and orbital properties of super-Earths and Neptune-mass planets. arxiv:11092497 Mazeh T, Holczer T, Faigler S (2016) Dearth of short-period Neptunian exoplanets: a desert in period-mass and period-radius planes. A&A 589:A75 Montet BT, Crepp JR, Johnson JA, Howard AW, Marcy GW (2014) The TRENDS high-contrast imaging survey. IV. The occurrence rate of giant planets around M dwarfs. ApJ 781:28 Muirhead PS, Mann AW, Vanderburg A et al (2015) Kepler-445, Kepler-446 and the occurrence of compact multiples orbiting mid-M dwarf stars. ApJ 801:18 Mulders GD, Pascucci I, Apai D (2015) An increase in the mass of planetary systems around lower-mass stars. ApJ 814:130 Mulders GD, Pascucci I, Apai D, Frasca A, Molenda-Zakowicz J (2016) A super-solar metallicity for stars with hot rocky exoplanets. AJ 152:187 Neves V, Bonfils X, Santos NC et al (2013) Metallicity of M dwarfs. III. Planet-metallicity and planet-stellar mass correlations of the HARPS GTO M dwarf sample. A&A 551:A36 Owen JE, Wu Y (2013) Kepler planets: a tale of evaporation. ApJ 775:105 Owen JE, Wu Y (2017) The evaporation valley in the Kepler planets. ApJ 847:29 Pepper J, Gould A, Depoy DL (2003) Using all-sky surveys to find planetary transits. Acta Astron 53:213–228 Petigura EA, Howard AW, Marcy GW (2013) Prevalence of Earth-size planets orbiting sun-like stars. Proc Natl Acad Sci 110:19,273–19,278 Petigura EA, Howard AW, Marcy GW et al (2017) The California-Kepler survey. I. High-resolution spectroscopy of 1305 stars hosting Kepler transiting planets. AJ 154:107 Petigura EA, Marcy GW, Winn JN et al (2018) The California-Kepler survey. IV. Metal-rich stars host a greater diversity of planets. AJ 155:89 Pollack JB, Hubickyj O, Bodenheimer P et al (1996) Formation of the giant planets by concurrent accretion of solids and gas. Icarus 124:62–85 Pu B, Wu Y (2015) Spacing of Kepler planets: sculpting by dynamical instability. ApJ 807:44 Reffert S, Bergmann C, Quirrenbach A, Trifonov T, Künstler A (2015) Precise radial velocities of giant stars. VII. Occurrence rate of giant extrasolar planets as a function of mass and metallicity. A&A 574:A116 Ricker GR, Winn JN, Vanderspek R et al (2015) Transiting exoplanet survey satellite (TESS). J Astron Telescopes Instrum Syst 1(1):014003 Sahlmann J, Ségransan D, Queloz D et al (2011) Search for brown-dwarf companions of stars. A&A 525:A95 Sanchis-Ojeda R, Rappaport S, Winn JN et al (2014) A study of the shortest-period planets found with Kepler. ApJ 787:47 Santerne A, Moutou C, Tsantaki M et al (2016) SOPHIE velocimetry of Kepler transit candidates. XVII. The physical properties of giant exoplanets within 400 days of period. A&A 587:A64 Santos NC, Israelian G, Mayor M, Rebolo R, Udry S (2003) Statistical properties of exoplanets. II. Metallicity, orbital parameters, and space velocities. A&A 398:363–376 Santos NC, Adibekyan V, Figueira P et al (2017) Observational evidence for two distinct giant planet populations. A&A 603:A30 Schlaufman KC (2018) Evidence of an upper bound on the masses of planets and its implications for giant planet formation. ApJ 853:37 Schlaufman KC, Winn JN (2016) The occurrence of additional giant planets inside the waterice line in systems with hot Jupiters: evidence against high-eccentricity migration. ApJ 825:62 Shallue CJ, Vanderburg A (2018) Identifying exoplanets with deep learning: a five-planet resonant chain around Kepler-80 and an eighth planet around Kepler-90. AJ 155:94 Tabachnik S, Tremaine S (2002) Maximum-likelihood method for estimating the mass and period distributions of extrasolar planets. MNRAS 335:151–158

1966

J. N. Winn

Thompson SE, Coughlin JL, Hoffman K et al (2017) Planetary candidates observed by Kepler. VIII. A fully automated catalog with measured completeness and reliability based on data release 25. ArXiv e-prints Tremaine S, Dong S (2012) The statistics of multi-planet systems. AJ 143:94 Triaud AHMJ, Martin DV, Ségransan D et al (2017) The EBLM project. IV. Spectroscopic orbits of over 100 eclipsing M dwarfs masquerading as transiting hot Jupiters. A&A 608:A129 Twicken JD, Jenkins JM, Seader SE et al (2016) Detection of potential transit signals in 17 quarters of Kepler data: results of the final Kepler mission transiting planet search (DR25). AJ 152:158 Udry S, Santos NC (2007) Statistical properties of exoplanets. ARA&A 45:397–439 Van Eylen V, Albrecht S (2015) Eccentricity from transit photometry: small planets in Kepler multi-planet systems have low eccentricities. ApJ 808:126 Van Eylen V, Agentoft C, Lundkvist MS et al (2017) An asteroseismic view of the radius valley: stripped cores, not born rocky. ArXiv e-prints van Sluijs L, Van Eylen V (2018) The occurrence of planets and other substellar bodies around white dwarfs using K2. MNRAS 474:4603–4611 Wang J, Fischer DA, Horch EP, Huang X (2015) On the occurrence rate of hot Jupiters in different stellar environments. ApJ 799:229 Weiss LM, Marcy GW, Petigura EA et al (2018) The California-Kepler survey. V. Peas in a pod: planets in a Kepler multi-planet system are similar in size and regularly spaced. AJ 155:48 Wilson RF, Teske J, Majewski SR et al (2018) Elemental abundances of Kepler objects of interest in APOGEE. I. Two distinct orbital period regimes inferred from host star iron abundances. AJ 155:68 Winn JN (2010) Exoplanet transits and occultations. University of Arizona Press, Tucson, pp 55–77 Winn JN, Fabrycky DC (2015) The occurrence and architecture of exoplanetary systems. ARA&A 53:409–447 Wittenmyer RA, Butler RP, Tinney CG et al (2016) The Anglo-Australian planet search XXIV: the frequency of Jupiter analogs. ApJ 819:28 Wolfgang A, Laughlin G (2012) The effect of population-wide mass-to-radius relationships on the interpretation of Kepler and HARPS super-Earth occurrence rates. ApJ 750:148 Wright JT, Veras D, Ford EB et al (2011) The California planet survey. III. A possible 2:1 resonance in the exoplanetary triple system HD 37124. ApJ 730:93 Wright JT, Marcy GW, Howard AW et al (2012) The frequency of hot Jupiters orbiting nearby solar-type stars. ApJ 753:160 Xie JW, Dong S, Zhu Z et al (2016) Exoplanet orbital eccentricities derived from LAMOST-Kepler analysis. Proc Natl Acad Sci 113:11,431–11,435 Youdin AN (2011) The exoplanet census: a general method applied to Kepler. ApJ 742:38

Occurrence Rates from Direct Imaging Surveys

97

Brendan P. Bowler and Eric L. Nielsen

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Calculating Occurrence Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Occurrence Rate of Giant Planets on Wide Orbits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Occurrence Rate of Brown Dwarf Companions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1968 1969 1973 1976 1978 1979

Abstract

The occurrence rate of young giant planets from direct imaging surveys is a fundamental tracer of the efficiency with which planets form and migrate at wide orbital distances. These measurements have progressively converged to a value of about 1% for the most massive planets (5–13 MJup ) averaged over all stellar masses at separations spanning a few tens to a few hundreds of AU. The subtler statistical properties of this population are beginning to emerge with everincreasing sample sizes: there is tentative evidence that planets on wide orbits are more frequent around stars that possess debris disks; brown dwarf companions exist at comparable (or perhaps slightly higher) rates as their counterparts in the planetary-mass regime; and the substellar companion mass function appears to be smooth and may extend down to the opacity limit for fragmentation. Within a few years, the conclusion of second-generation direct imaging surveys will

B. P. Bowler () Department of Astronomy, The University of Texas at Austin, Austin, TX, USA e-mail: [email protected]; [email protected] E. L. Nielsen () Kavli Institute for Particle Astrophysics and Cosmology, Stanford University, Stanford, CA, USA e-mail: [email protected]; [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_155

1967

1968

B. P. Bowler and E. L. Nielsen

enable more definitive interpretations with the ultimate goal of identifying the dominant origin of this population and uncovering its relationship to planets at smaller separations.

Introduction Direct imaging is the foremost method to study giant planets at wide orbital distances beyond about 10 AU and complements radial velocity, transit, microlensing, and astrometric discovery techniques which probe smaller separations closer to their host stars. High-contrast imaging from the ground makes use of adaptive optics systems that largely operate at near-infrared wavelengths, making this method most sensitive to thermal emission from massive, warm giant planets. As a result, direct imaging surveys predominantly focus on the closest and youngest stars before planets have cooled to faint luminosities and low temperatures. This makes target samples for imaging surveys unusual compared to other planet detection methods, which predominantly focus on old (several Gyr) field stars with lower activity and jitter levels. In addition to individual discoveries, high-contrast imaging surveys deliver statistical constraints on the occurrence rates and demographics of giant planets at large orbital distances. The frequency of giant planets and their mass-period distributions provide valuable information about the efficiency of planet formation and migration to large separations and have been used to both guide and test giant planet formation routes. Nearly two dozen young planets have now been imaged with inferred masses as low as 2 MJup and separations spanning a large dynamic range of 10–104 AU (Fig. 1). Several formation routes can explain the origin of gas giants at these unexpectedly wide separations: core and pebble accretion (Lambrechts and Johansen 2012), disk instability (e.g., Durisen et al. 2007; Kratter and Lodato 2016), turbulent fragmentation (Bate et al. 2003), and planet-planet scattering (Veras et al. 2009). In principle these pathways should imprint unique signatures on the resulting occurrence rates, mass-period distributions, and threedimensional orbital architecture of exoplanets, although these are challenging to discern in practice due to the low incidence of widely separated planets, a diversity of theoretical predictions, the potential for subsequent migration to occur, and the difficulty of constraining orbital elements for ultra-long-period planets (e.g., Blunt et al. 2017). One of the emerging goals for direct imaging is to untangle the dominant formation route(s) of this population, which is best accomplished in a statistical fashion with expansive high-contrast imaging surveys. The largest-scale surveys exploiting extreme adaptive optics systems like the Gemini Planet Imager (GPI) and the Spectropolarimetric High-contrast Exoplanet Research (SPHERE) are currently underway. These second-generation instruments utilize integral field spectrographs for speckle suppression through spectral differential imaging and achieve unprecedented on-sky contrasts at small (10 MJ

>4 MJ 20−100 AU

>2 MJ 50−295 AU

>4 MJ 59−460 AU 6−12 MJ 50−1000 AU

>4 MJ 20−500 AU

>5 MJ >80 AU >3 MJ >40 AU

>5 MJ 100−500 AU

3−14 MJ 5−320 AU

5−20 MJ 10−1000 AU 1−20 MJ 10−150 AU

5−70 MJ 10−100 AU

5−13 MJ 2−14 MJ 10−100 AU 0.5−14 MJ 8−400 AU 20−300 AU 5−13 MJ 30−300 AU

5−14 MJ 5−500 AU

La

fre

ni

er N ee ie ta ls en l. ( et 200 H ei 7) nz al. (2 N ee ie t a 008 ls ) en l. ( 20 C ha et al 10) uv .( in 2 Vi et a 010 ga ) l. ( n R am et 201 0 a ea ) l. (2 u Ja 0 e ns t al 12) .( o 20 N ne 13 ie t ) Ya lse al. (2 n m 01 am et 3 a ) l. ot ( o et 201 Bi 3) lle al. ( 20 W re 13 t ah ) ha al. ( 20 Br j et al 13) an . C dt e (20 ha 13 ta uv ) l .( in Bo et 201 4) al w . l G er e (20 al 15 t ic ) he al. r e (20 15 ta ) l. ( La Bow 201 nn 6) le r i M er e (20 es 16 hk t al. ) at (2 Vi et a 016 ga ) l. ( n et 201 7) al .( 20 17 )

0.00

Fig. 2 Occurrence rate measurements of giant planets from direct imaging surveys over the past decade. Most surveys have reported upper limits, but larger meta-analyses over the past few years are converging on a frequency of about 1% averaged over all host star spectral types. Arrows indicate upper limits with horizontal bars representing the upper limit value. For measurements (as opposed to upper limits), solid and dotted lines denote 68% and 95% credible intervals, respectively. Note that these surveys are not all independent; targets may overlap and many surveys incorporate previously published results into their statistical analysis

Nielsen et al. (2013), Wahhaj et al. (2013b), Brandt et al. (2014a), Bowler et al. (2015), Chauvin et al. (2015), Galicher et al. (2016), Lannier et al. (2016), Uyama et al. (2017), and Naud et al. (2017). Merging discoveries and detection limits from published surveys into everlarger samples has become standard practice in order to improve the precision of occurrence rate measurements. These large meta-analyses carry the most statistical weight and are necessary to test for potential correlations with stellar properties and environmental context, which are ultimately expected to yield clues about planet formation and migration pathways. The following results are based on compilations of several surveys and assume hot-start evolutionary models to infer planet masses and sensitivities. The first large-scale compilation of direct imaging surveys was carried out by Nielsen et al. (2008) and expanded to over 118 stars in Nielsen and Close (2010), who measured an upper limit of 4 MJup planets between 30–500 AU around Sunlike stars (at the 95% confidence level). The next major milestone for population statistics was carried out by Brandt et al. (2014b). They merged 5 surveys totaling a sample of 248 unique targets with spectral types from B to mid-M and found that 1.0–3.1% of stars host 5–70 MJup companions between 10 and 100 AU (at 68% confidence; the 95% credible interval spans 0.52–4.9%). In addition, they conclude that the substellar companions from these surveys are

97 Occurrence Rates from Direct Imaging Surveys

1975

consistent with a smooth companion mass and separation distribution (p.M; a/ / M 0:65˙0:60 a0:85˙0:39 ), implying that the planetary-mass companions identified in these surveys may represent the low-mass tail of brown dwarfs rather than a separate population of high-mass planets. Recently, Vigan et al. (2017) compiled the largest collection of imaging results focusing on FGK stars. Based on a sample of 199 targets from 12 previous imaging surveys, they derive a frequency of 0.90% for 5–14 MJup companions spanning 5–500 AU (the 68% credible interval is 0.70–2.95; the 95% range is 0.25–5.00%). For the entire substellar mass range from 5 to 75 MJup , they find a somewhat higher occurrence rate of 2.40% (with a 95% credible interval of 0.90–6.80%). By quantitatively comparing their results with population synthesis models of planets formed via disk instability, they lay the groundwork for direct population-level tests of planet formation theory and conclude that this mode of planet formation is probably inefficient at forming giant planets. At present there are no signs that host star multiplicity dramatically impacts the occurrence rates of circumbinary planets on wide orbits compared to single stars. Bonavita et al. (2016) measure a frequency of 1.3% for companions spanning 2–15 MJup between 10 and 500 AU based on 117 close binary systems (their 95% confidence range spans 0.35–6.85%), which is similar to the rate for their control sample of 205 single stars compiled from the literature. The largest statistical analyses spanning all host star masses were assembled by Bowler (2016) and Galicher et al. (2016). Based on results from nine published surveys totaling 384 unique and single young stars, Bowler (2016) found that the overall occurrence rate of 5–13 MJup giant planets spanning 30–300 AU is 0.6C0:7 0:5 % (68% credible interval). For an even wider range of separations from 10 to 1000 AU, the occurrence rate is 0.8C1:0 0:6 %. This is in good agreement with results from Galicher et al. (2016), who find a frequency of 1.05C2:8 0:7 % (95% credible interval) for 0.5–14 MJup planets between 20 and 300 AU based on their sample of 292 stars spanning B stars to M dwarfs. Altogether, the most massive giant planets reside around roughly 1% of stars in the interval between a few tens to a few hundred AU. Planetary-mass companions exist at even wider separations with comparably small frequencies. Durkan et al. (2016) carried out an analysis of Spi t zer/IRAC observations of young stars and found an upper limit of 9% for 0.5–13 MJup companions between 100 and 1000 AU. Naud et al. (2017) measure a frequency of 0.84C6:73 0:66 % (at 95% confidence intervals) for 5–13 MJup companions between 500 and 5000 AU, similar to the substellar companion frequency at ultrawide separations estimated by Aller et al. (2013). With these large samples in hand, the next natural step is to begin searching for correlations with stellar parameters and environment to better understand the context in which this population of planetary-mass companions forms. Several surveys have specifically focused on high-mass stars (Ehrenreich et al. 2010; Janson et al. 2011; Vigan et al. 2012; Nielsen et al. 2013) and low-mass stars (Delorme et al. 2012; Bowler et al. 2015; Lannier et al. 2016) to examine the occurrence rate of giant planets as a function of stellar host mass. Lannier et al. (2016) find intriguing hints of a trend with stellar host mass, but this has not been recovered with larger samples by Bowler (2016) and Galicher et al. (2016). Breaking their sample into spectral-type

1976

B. P. Bowler and E. L. Nielsen

bins, Bowler (2016) measures an occurrence rate of 2.8C3:7 2:3 % from 110 young BA stars, 500 stars with next-generation imagers capable of detecting brown dwarfs even closer to their parent stars. Already these programs have detected new brown dwarfs around HR 2562 (Konopacky et al. 2016b) and HD 206893 (as part of the SHARDDS campaign, Milli et al. 2017); the final surveys will be well-placed to make a more robust measurement of substellar occurrence rate and determine the nature of the stellar mass and companion mass distributions.

Conclusions Substantial progress has been made over the past decade constraining the statistical properties of giant planets and brown dwarf companions via direct imaging. Altogether, averaged across all host star spectral types (B stars to M dwarfs), the most precise measurements of the giant planet occurrence rate spanning masses of 5–13 MJup for the entire range of separations accessible to direct imaging (5–500 AU) are converging to a value near 1%. More nuanced correlations are just beginning to be explored, and there are initial indications that giant planets positively scale with the presence of debris disks and (to a lesser extent) host star mass. The frequency of brown dwarf companions (13–75 MJup ) is consistently found to be between 1 and 4% – albeit with considerable uncertainty among individual measurements – with no obvious signs that this value evolves with age or as a function of stellar host mass. This indicates that the frequency of brown dwarf companions is comparable to those of giant planets, with hints that brown dwarfs may exceed planets by a factor of a few. There remain many open questions that will be especially suitable to examine when the large (>500 star) surveys with second-generation AO instruments conclude and in the longer term with the 30-m class of ground-based extremely large telescopes. These include refining the functional form of the companion mass function and companion mass ratio distribution; more robust tests for correlations with stellar host mass, the presence of debris disks, and multiplicity; searching for signs of evolution in the overall occurrence rate or outer separation cutoff over time; a better understanding of planet multiplicity at wide separations; and eventually completing the bridge between the radial velocity and direct imaging planet distribution functions. The steady improvement of occurrence rate measurements and population statistics is a testament to the communal effort required to build ever-larger samples and address increasingly rich questions about planet formation, architectures, and evolution.

Cross-References  Direct Imaging as a Detection Technique for Exoplanets  Exoplanet Atmosphere Measurements from Direct Imaging  Planet Occurrence: Doppler and Transit Surveys

97 Occurrence Rates from Direct Imaging Surveys

1979

 Future Exoplanet Research: High-Contrast Imaging Techniques  Future Exoplanet Space Missions: Spectroscopy and Coronographic Imaging  HR8799: Imaging a System of Exoplanets  Imaging with Adaptive Optics and Coronographs for Exoplanet Research  Multi-Pixel Imaging of Exoplanets with a Hypertelescope in Space  Planet Populations as a Function of Stellar Properties

References Aller KM, Kraus AL, Liu MC et al (2013) A Pan-STARRS + UKIDSS search for young, wide planetary-mass companions in upper scorpius. Astrophys J 773(1):63 Baraffe I, Chabrier G, Barman TS, Allard F, Hauschildt PH (2003) Evolutionary models for cool brown dwarfs and extrasolar giant planets. The case of HD 209458. A&A 402:701 Baraffe I, Homeier D, Allard F, Chabrier G (2015) New evolutionary models for pre-main sequence and main sequence low-mass stars down to the hydrogen-burning limit. A&A 577:A42 Bate MR, Bonnell IA, Bromm V (2003) The formation of a star cluster: predicting the properties of stars and brown dwarfs. MNRAS 339:577 Best WMJ, Liu MC, Dupuy TJ, Magnier EA (2017) The young L dwarf 2MASS J111932541137466 is a planetary-mass binary . Astrophy J Lett 843(1):L4 Beust H, Augereau JC, Bonsor A et al (2014) An independent determination of Fomalhaut b’s orbit and the dynamical effects on the outer dust belt. A&A 561:A43 Biller BA, Close LM, Masciadri E et al (2007) An imaging survey for extrasolar planets around 45 close, young stars with the simultaneous differential imager at the very large telescope and MMT. Astrophys J Suppl Ser 173:143 Biller BA, Liu MC, Wahhaj Z et al (2010) The gemini NICI planet-finding campaign: discovery of a close substellar companion to the young debris disk star PZ tel. Astrophys J 720(1):L82–L87 Biller BA, Liu MC, Wahhaj Z et al (2013) The GEMINI/NICI planet-finding campaign: the frequency of planets around young moving group stars. ApJ 777(2):160 Blunt S, Nielsen EL, De Rosa, RJ et al (2017) Orbits for the impatient: a Bayesian rejectionsampling method for quickly fitting the orbits of long-period exoplanets. Astron J 153:229 Bonavita M, Chauvin G, Desidera S et al (2012) MESS (multi-purpose exoplanet simulation system). A&A 537:A67 Bonavita M, de Mooij EJW, Jayawardhana R (2013) Quick-MESS: a fast statistical tool for exoplanet imaging surveys. PASP 849(125) Bonavita M, Desidera S, Thalmann C et al (2016) SPOTS: the search for planets orbiting two stars. A&A 593:A38 Bowler BP (2016) Imaging extrasolar giant planets. Publ Astron Soc Pac 128(968):102,001 Bowler BP, Liu MC, Shkolnik EL et al (2012) Planets around low-mass stars (PALMS). I. A substellar companion to the young M DWARF 1RXS J235133.3+312720. Astrophys J 753(2):142 Bowler BP, Liu MC, Shkolnik EL, Tamura M (2015) Planets around low-mass stars (PALMS). IV. The outer architecture of M dwarf planetary systems. Astrophys J Suppl Ser 216(1):7 Bowler BP, Liu MC, Mawet D et al (2017) Planets around low-mass stars (PALMS). VI. Discovery of a remarkably red planetary- mass companion to the AB DOR moving group candidate 2MASS J22362452+4751425. Astron J 153(1):1–15 Brandt TD, McElwain MW, Turner EL et al (2013) New techniques for high-contrast imaging with ADI: the ACORNS-ADI seeds data reduction pipeline. Astrophy J 764(2):183 Brandt TD, Kuzuhara M, McElwain MW et al (2014a) the moving group targets of the seeds highcontrast imaging survey of exoplanets and disks: results and observations from the first three years. ApJ 786(1):1

1980

B. P. Bowler and E. L. Nielsen

Brandt TD, McElwain MW, Turner EL et al (2014b) A statistical analysis of seeds and other highcontrast exoplanet surveys: massive planets or low-mass brown dwarfs? ApJ 794(2):159 Cantalloube F, Mouillet D, Mugnier LM et al (2015) Direct exoplanet detection and characterization using the ANDROMEDA method: performance on VLT/NaCo data. A&A 582:A89 Chauvin G, Lagrange AM, Dumas C et al (2005) Giant planet companion to 2MASSWJ1207334393254. A&A 438(2):L25–L28 Chauvin G, Lagrange AM, Bonavita M et al (2010) Deep imaging survey of young, nearby austral stars. A&A 509:A52 Chauvin G, Vigan A, Bonnefoy M et al. (2015) The VLT/NaCo large program to probe the occurrence of exoplanets and brown dwarfs at wide orbits. A&A 573:A127 Chauvin G, Desidera S, Lagrange AM et al (2017) Discovery of a warm, dusty giant planet around HIP65426. A&A 605:L9 Cheetham AC, Kraus AL, Ireland MJ et al (2015) Mapping the shores of the brown dwarf desert. IV. Ophiuchus. ApJ 813(2):1–12 Clanton C, Gaudi BS (2016) Synthesizing exoplanet demographics: a single population of longperiod planetary companions to M dwarfs consistent with microlensing, radial velocity, and direct imaging surveys. ApJ 819(2):1–42 Currie T, Muto T, Kudo T et al (2014) Recovery of the candidate protoplanet HD 100546 b with GEMINI/NICI and detection of additional (planet-induced?) disk structure at small separations. Astrophys J Lett 796(2):L30 De Rosa RJ, Nielsen EL, Blunt SC et al (2015) Astrometric confirmation and preliminary orbital parameters of the young exoplanet 51 Eridani b with the gemini planet imager. Astrophys J Lett 814(1):1–7 Delorme P, Lagrange AM, Chauvin G et al (2012) High-resolution imaging of young M-type stars of the solar neighbourhood: probing for companions down to the mass of Jupiter. A&A 539:A72 Delorme P, Gagné J, Girard JH et al (2013) Direct-imaging discovery of a 12–14 Jupiter-mass object orbiting a young binary system of very low-mass stars. A&A 553:L5 Durisen RH, Boss AP, Mayer L et al (2007) Gravitational instabilities in gaseous protoplanetary disks and implications for giant planet formation. In: Reipurth B, Jewitt D, Keil K (eds) Protostars and planets V. University Arizona Press, Tucson, p 607 Durkan S, Janson M Carson JC (2016) High contrast imaging with spitzer: constraining the frequency of giant planets out to 1000 AU separations. Astrophys J 824(1):58 Ehrenreich D, Lagrange AM, Montagnier G et al (2010) Deep infrared imaging of close companions to austral A- and F-type stars. A&A 523:A73 Feldt M, Olofsson J, Boccaletti A et al (2017) SPHERE/SHINE reveals concentric rings in the debris disk of HIP 73145. A&A 601:A7 Fortney JJ, Marley MS, Saumon D, Lodders K (2008) Synthetic spectra and colors of young giant planet atmospheres: effects of initial conditions and atmospheric metallicity. Astrophy J 683:1104 Galicher R, Marois C, Macintosh B et al (2016) The international deep planet survey. A&A 594:A63 Heinze AN, Hinz PM, Kenworthy M et al (2010) Constraints on long-period planets from ANLANDM-band survey of nearby sun-like stars: modeling results. Astrophys J 714(2):1570–1581 Hinz JL, McCarthy DW, Simons DA et al (2002) A near-infrared, wide-field, proper-motion search for brown dwarfs. Astron J 123:2027 Ireland MJ, Kraus A, Martinache F, Law N, Hillenbrand LA (2011) Two wide planetary-mass companions to solar-type stars in upper Scorpius. Astrophys J 726:113 Janson M, Bonavita M, Klahr H et al (2011) High-contrast imaging search for planets and brown dwarfs around the most massive stars in the solar neighborhood. Astrophys J 736(2):89 Janson M, Brandt TD, Moro-Martín A et al (2013) The seeds direct imaging survey for planets and scattered dust emission in debris disk systems. ApJ 773(1):73 Jensen-Clem R, Mawet D, Gonzalez CAG et al (2018) A new standard for assessing the performance of high contrast imaging systems. Astron J 155(1):19

97 Occurrence Rates from Direct Imaging Surveys

1981

Kalas P, Graham JR, Fitzgerald MP, Clampin M (2013) STIS coronagraphic imaging of Fomalhaut: main belt structure and the orbit of Fomalhaut b. ApJ 775(1):56 Konopacky QM, Marois C, Macintosh BA et al (2016a) Astrometric monitoring of the HR 8799 planets: orbit constraints from self-consistent measurements . Astron J 152(2):1–18 Konopacky QM, Rameau J, Duchene G et al (2016b) Discovery of a substellar companion to the nearby debris disk host HR 2562. Astrophys J Lett 829(1):1–7 Kratter K, Lodato G (2016) Gravitational instabilities in circumstellar disks. Ann Rev Astro Astrophys 54(1):271–311 Lafrenière D, Doyon R, Marois C et al (2007a) The gemini deep planet survey. Astrophy J 670:1367 Lafrenière D, Marois C, Doyon R, Nadeau D, Artigau É (2007b) A new algorithm for pointspread function subtraction in high-contrast imaging: a demonstration with angular differential imaging. Astrophys J 660:770 Lafrenière D, Jayawardhana R, van Kerkwijk MH, Brandeker A, Janson M (2014) An adaptive optics multiplicity census of young stars in upper Scorpius. ApJ 785(1):47 Lagrange AM, Bonnefoy M, Chauvin G et al (2010) A giant planet imaged in the disk of the young star beta pictoris: supporting online material. Science 329(5987):57–59 Lambrechts M, Johansen A (2012) Rapid growth of gas-giant cores by pebble accretion. A&A 544:A32 Lannier J, Delorme P, Lagrange AM et al (2016) MASSIVE: a Bayesian analysis of giant planet populations around low-mass stars. A&A 596:A83 Liu MC, Magnier EA, Deacon NR et al (2013) The extremely red, young L dwarf PSO J318.5338– 22.8603: a free-floating planetary-mass analog to directly imaged young gas-giant planets. Astrophys J 777(2):L20 Lowrance PJ, Schneider G, Kirkpatrick JD et al (2000) A candidate substellar companion to HR 7329. Astrophys J 541:390 Macintosh B, Graham JR, Ingraham P et al (2014) First light of the Gemini Planet Imager. Proc Natl Acad Sci 111(35):12,661–12,666 Macintosh B, Graham JR, Barman T et al (2015) Discovery and spectroscopy of the young Jovian planet 51 Eri b with the gemini planet imager. Science 350:64 Marleau G-D, Cumming A (2014) Constraining the initial entropy of directly detected planets. MNRAS 437:1378 Marois C, Macintosh B, Barman T et al (2008) Direct imaging of multiple planets orbiting the star HR 8799. Science 322(5906):1348–1352 Marois C, Macintosh B, Véran JP (2010) Exoplanet imaging with LOCI processing: photometry and astrometry with the new SOSIE pipeline. Proc SPIE 7736:77,361J Marois C, Correia C, Galicher R et al (2014) GPI PSF subtraction with TLOCI: the next evolution in exoplanet/disk high-contrast imaging. In: Marchetti E, Close LM, Veran JP (eds) SPIE astronomical telescopes + instrumentation. SPIE, p 91480U Mawet D, Milli J, Wahhaj Z et al (2014) Fundamental limitations of high contrast imaging set by small sample statistics. ApJ 792(2):97 McCarthy C, Zuckerman B (2004) The brown dwarf desert at 75–1200 AU. Astron J 127:2871 Meshkat T, Bailey VP, Su KYL et al (2015a) Searching for planets in holey debris disks with the apodizing phase plate. ApJ 800(1):5 Meshkat T, Bonnefoy M, Mamajek EE et al (2015b) Discovery of a low-mass companion to the F7V star HD 984. Mon Not RAS 453(3):2379–2387 Meshkat T, Mawet D, Bryan ML et al (2017) A direct imaging survey of spitzer-detected debris disks: occurrence of giant planets in dusty systems . Astron J 154(6):245 Metchev SA, Hillenbrand LA (2009) The Palomar/Keck adaptive optics survey of young solar analogs: evidence for a universal companion mass function. Astrophys J Suppl Ser 181(1): 62–109 Meyer MR, Amara A, Reggiani M, Quanz SP (2017) M dwarf exoplanet surface density distribution: a log-normal fit from 0.07–400 AU. arXiv

1982

B. P. Bowler and E. L. Nielsen

Milli J, Hibon P, Christiaens V et al (2017) Discovery of a low-mass companion inside the debris ring surrounding the F5V star HD206893. A&A 597:L2 Mordasini C, Marleau G-D, Molliere P (2017) Characterization of exoplanets from their formation. III. The statistics of planetary luminosities. A&A 608:72 Mugrauer M, Neuh User R, Guenther EW et al (2004) HD?77407 and GJ?577: two new young stellar binaries. A&A 417(3):1031–1038 Mugrauer M, Vogt N, Neuhäuser R, Schmidt TOB (2010) Direct detection of a substellar companion to the young nearby star PZ Telescopii. A&A 523:L1 Nakajima T, Oppenheimer BR, Kulkarni SR et al (1995) Discovery of a cool brown dwarf. Nature 378(6556):463–465 Naud ME, Artigau E, Doyon R et al (2017) PSYM-WIDE: a survey for large-separation planetarymass companions to late spectral type members of young moving groups. Astron J 154:154–129 Nielsen EL, Close LM (2010) A uniform analysis of 118 stars with high-contrast imaging: longperiod extrasolar giant planets are rare around sun-like stars. Astrophy J 717(2):878–896 Nielsen EL, Close LM, Biller BA, Masciadri E, Lenzen R (2008) Constraints on extrasolar planet populations from VLT NACO/SDI and MMT SDI and direct adaptive optics imaging surveys: giant planets are rare at large separations. Astrophys J 674:466 Nielsen EL, Liu MC, Wahhaj Z et al (2013) The gemini NICI planet-finding campaign: the frequency of giant planets around young B and A stars. ApJ 776(1):4 Nielsen EL, Liu MC, Wahhaj Z et al (2014) The gemini NICI planet-finding campaign: the orbit of the young exoplanet ˇ pictoris b. ApJ 794(2):158 Nielsen Eric L, De Rosa RJ, Rameau J et al. (2017) Evidence that the directly imaged planet HD 131399 Ab is a background star. AJ 154:218 Oppenheimer BR, Golimowski DA, Kulkarni SR et al (2001) A coronagraphic survey for companions of stars within 8 Parsecs. Astron J 121:2189 Pueyo L (2016) Detection and characterization of exoplanets using projections on Karhunen– Loeve eigenimages: forward modeling . ApJ 824(2):1–29 Pueyo L, Crepp JR, Vasisht G et al (2012) Application of a damped locally optimized combination of images method to the spectral characterization of faint companions using an integral field spectrograph. Astrophys J Suppl Ser 199(1):6 Rameau J, Chauvin G, Lagrange AM et al (2013a) A survey of young, nearby, and dusty stars conducted to understand the formation of wide-orbit giant planets. A&A 553:A60 Rameau J, Chauvin G, Lagrange AM et al (2013b) Confirmation of the planet around HD 95086 by direct imaging. Astrophys J 779(2):L26 Rameau J, Nielsen EL, De Rosa RJ et al (2016) Constraints on the architecture of the HD 95086 planetary system with the gemini planet imager . Astrophys J Lett 822(2):1–7 Reggiani M, Meyer MR, Chauvin G et al (2016) The VLT/NaCo large program to probe the occurrence of exoplanets and brown dwarfs at wide orbits. A&A 586:A147 Ruffio JB, Macintosh B, Wang JJ et al (2017) Improving and assessing planet sensitivity of the GPI exoplanet survey with a forward model matched filter. arXiv (1):14 Samland M, Mollière P, Bonnefoy M et al (2017) Spectral and atmospheric characterization of 51 Eridani b using VLT/SPHERE. A&A 603:A57 Soummer R, Pueyo L, Larkin J (2012) Detection and characterization of exoplanets and disks using projections on Karhunen-Loève eigenimages. Astrophys J 755(2):L28 Todorov K, Luhman KL, Mcleod KK (2010) Discovery of a planetary-mass companion to a brown dwarf in Taurus. Astrophys J 714(1):L84–L88 Uyama T, Hashimoto J, Kuzuhara M et al (2017) The SEEDS high-contrast imaging survey of exoplanets around young stellar objects. Astron J 153(3):1–27 Veras D, Crepp JR, Ford EB (2009) Formation, survival, and detectability of planets beyond 100 AU. Astrophys J 696:1600 Vigan A, Patience J, Marois C et al (2012) The international deep planet survey. I. The frequency of wide-orbit massive planets around A-stars. A&A 544:9 Vigan A, Bonavita M, Biller B et al (2017) The VLT/NaCo large program to probe the occurrence of exoplanets and brown dwarfs at wide orbits. A&A 603:A3

97 Occurrence Rates from Direct Imaging Surveys

1983

Wagner K, Apai D, Kasper M et al (2016) Direct imaging discovery of a Jovian exoplanet within a triple-star system. Science Wahhaj Z, Liu MC, Biller BA et al (2013a) The gemini NICI planet-finding campaign: the companion detection pipeline. ApJ 779(1):80 Wahhaj Z, Liu MC, Nielsen EL et al (2013b) The gemini planet-finding campaign: the frequency of giant planets around debris disk stars. Astrophys J 773(2):179 Wahhaj Z, Cieza LA, Mawet D et al (2015) Improving signal-to-noise in the direct imaging of exoplanets and circumstellar disks with MLOCI. A&A 581:A24 Wang JJ, Graham JR, Pueyo L et al (2016) The orbit and transit prospects for ˇ Pictoris b constrained with one milliarcsecond astrometry . Astron J 152(4):1–16 Wertz O, Absil O, Gomez Gonzalez CA et al (2017) VLT/SPHERE robust astrometry of the HR8799 planets at milliarcsecond-level accuracy. A&A 598:A83 Yamamoto K, Matsuo T, Shibai H et al (2013) Direct imaging search for extrasolar planets in the Pleiades. PASJ 65:90

Populations of Extrasolar Giant Planets from Transit and Radial Velocity Surveys

98

Alexandre Santerne

Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Different Populations of EGP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hot Jupiter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Temperate/Cold Giants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Period-Valley Giants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lack of Very Short Period EGP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Multiplicity of EGP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Radius of EGP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Occurrence Rates of EGPs in Different Stellar Populations . . . . . . . . . . . . . . . . . . . . . . . . . . The Solar Neighborhood . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Kepler Field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The CoRoT eyes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ground-Based Transit Surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . In Open Clusters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Brown-Dwarf Desert . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Comparison of the Occurrence Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Relation with Host Star Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Stellar Metallicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Stellar Mass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross-References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1986 1986 1987 1988 1988 1989 1989 1990 1991 1991 1994 1996 1996 1997 1998 1998 1999 1999 2000 2001 2003 2003

Abstract

Transit and radial velocity surveys have deeply explored the population of extrasolar giant planets, with hundreds of objects detected to date. All these detections allow to understand their physical properties and to constrain their

A. Santerne () Aix Marseille Univ, CNRS, CNES, LAM, Marseille, France e-mail: [email protected] © Springer International Publishing AG, part of Springer Nature 2018 H. J. Deeg, J. A. Belmonte (eds.), Handbook of Exoplanets, https://doi.org/10.1007/978-3-319-55333-7_154

1985

1986

A. Santerne

formation, migration, and evolution mechanism. In this chapter, the observed properties of these planets are presented along with the various populations identified in the data. The occurrence rates of giant exoplanets, as observed in different stellar environment by various surveys, are also reviewed and compared. Finally, the presence and properties of the giant exoplanets are discussed in the regard of the properties of the host star. Over this chapter, the observational constraints are discussed in the context of the dominant planet formation, migration, and evolution scenarios.

Introduction Two decades of exploration of extrasolar giant planets (hereafter EGPs) with radial velocity and transit surveys, both from the ground and from space, have revolutionized our view on giant planets, in comparison with the solar system. At a time when Earth-sized exoplanets are discovered in the habitable zone of their star, many questions regarding the formation, migration, and evolution of EGPs are not yet fully understood. For instance, their dominant formation mechanisms are still debated: either by core-accretion (e.g., Mordasini et al. 2012) or disk instability (see Nayakshin 2017, for a review). The physical process causing the inflation of giant planets is also unclear (Baraffe et al. 2014). Even the definition of what is an EGP, with respect to brown dwarfs is actively discussed (Schneider et al. 2011; Chabrier et al. 2014; Hatzes and Rauer 2015). Nevertheless, hundreds of EGPs have been discovered and well characterized thanks to photometric surveys like SuperWASP (Pollacco et al. 2006) and HATNet (Bakos et al. 2004) from the ground and CoRoT (Baglin et al. 2006) and Kepler (Borucki et al. 2009) from space, as well as spectroscopic surveys like with the CORALIE and HARPS (Mayor et al. 2011), the SOPHIE (e.g., Hébrard et al. 2016), and the Lick and Keck (e.g., Marcy et al. 2005) instruments and observatories (for a more complete list of instruments and surveys, see the corresponding sections of this book). All these discoveries bring important insights into giant planet formation, migration, and evolution. This chapter highlights the main interpretation of all the EGP discoveries, starting with a description of the different populations of EGP, then the occurrence rates of EGP as determined in different stellar environments, the relation between the presence and properties of EGPs with respect to their host star, and finally the conclusions.

Different Populations of EGP To date, nearly 1000 EGPs have been detected and characterized by photometric and spectroscopic surveys. They are displayed in Fig. 1 together with their distribution. From all these detections, there is a clear limit between giant planets and lowermass, Neptune-like planets at about 0.1–0.2 M (about 30–60 M˚ ). In this regime, very few objects have been detected either by transit or RV. This cannot be explained

98 Populations of Extrasolar Giant Planets from Transit and Radial Velocity Surveys

1987

Fig. 1 Extrasolar planets discovered to date (Source: NASA Exoplanet Archive) by the transit (orange marks) and radial velocity (violet marks) surveys. The planetary mass (sky-projected minimum mass in case of RV planets) is shown as function of the semimajor axis of the orbit. The top and right histograms represent the raw distribution of extrasolar planets in their semimajor axis and (minimum) mass, respectively. This reveals the two main populations of EGP and a transitional population where few objects have been found. Solar system planets are also present for comparison

by an observational bias. This lower limit in mass for the giant planet is supported by the threshold at which the planetesimal starts the runaway accretion and also opens a gap in the disk changing their migration from type I and type II (e.g., Crida and Bitsch 2017). The upper limit in mass, arbitrarily set at 30M in Fig. 1 corresponds to the one used by the NASA exoplanet archive (Akeson et al. 2013).

Hot Jupiter Planets more massive than 0.1–0.2 M are clearly distributed in two main clusters. The first cluster is very close to the star, with semimajor axis of less than 0.1 AU (equivalent to about 10 days for sun-like stars). These are the so-called hot Jupiters as these planets are highly irradiated by their host star. The most typical member of this population is 51 Peg b (Mayor and Queloz 1995). This population of hot Jupiters has been deeply explored by ground-based photometric surveys with hundreds of detections. They are the easiest planets to detect, even within reach of amateur facilities (e.g., Santerne 2014).

1988

A. Santerne

Temperate/Cold Giants The second cluster is located at much larger separation, starting at about 0.4 AU (about 100 days for sun-like stars) up to several AU. This population is composed by temperate to cold giants and includes planets like Jupiter. Transit surveys from the ground are not very sensitive to planets with periods greater than about 10 days and are not able to probe this population. Only space-based photometry lasting several years like Kepler was able to detect giant planets at a few hundred days, but given their low transit probability (less than 0.5% at 1AU for a sun-like star), the number of temperate/cold giants detected in transit is small (Santerne et al. 2016). Note that the lack of planets detected beyond the orbit of Saturn is only due to observational bias. Two decades of spectroscopic observations allows for the detection of planets with periods up to typically 20 years. As a consequence, planets with orbital period longer than Saturn (29 years) present only partial orbits in the data which is therefore poorly determined for now. Radial velocity observations in the next decades will likely extend this population of planets toward larger separation, if these planets are common. Planets less massive than Saturn at a few AU are also not yet detected because of instrumental limitations.

Period-Valley Giants Between these two clusters resides a transition where relatively few EGP have been found. This transition is known as the period valley, first identified by Udry et al. (2003) based on radial velocity detections. If this transition population is clear in spectroscopic surveys (see also Udry and Santos 2007), it was first unconfirmed by the transit detections of the Kepler mission (Howard et al. 2012; Fressin et al. 2013), indicating only one, continuous population of EGPs. This period valley was however confirmed in the Kepler detections by Santerne et al. (2016) after a systematic removal of false-positive contaminations in the sample using the SOPHIE spectrograph. This valley in the period distribution of EGPs is a strong indication that temperate/cold giants and hot Jupiters have different formation and/or migration mechanisms. Nevertheless, this period-valley population appears narrower in the Kepler data (only restricted to within 10–20 days Santerne et al. 2016) compared to spectroscopic data (extended between 10 and about 100 days). One reason for this behavior is that Kepler detected several EGPs with orbital periods between 20 and 100 days that belong to multiple planetary systems, like Kepler-9 (Holman et al. 2010), Kepler-51 (Masuda 2014), Kepler-89 (Weiss et al. 2013), and Kepler-117 (Bruno et al. 2015). These EGPs might have stopped their migration in the period valley and did not end as hot Jupiters because of their companion, as it is proposed for the pair Jupiter – Saturn in the solar system (Morbidelli and Crida 2007). Note that the period valley is poorly explored by ground-based transit surveys as their sensitivity drops drastically above 10 days, i.e., at the upper limit of the hot Jupiter population. Therefore, the relative weight between the hot Jupiter, the

98 Populations of Extrasolar Giant Planets from Transit and Radial Velocity Surveys

1989

temperate/cold giants, and the period-valley populations seen in Fig. 1 is strongly dominated by observing bias.

Lack of Very Short Period EGP Another clear behavior observed with all the EGP detections is the lack of lowmass, hot, giant planets (less massive than Jupiter, down to super-Earth) orbiting very close to their star (up to 0.04 AU, see Fig. 1). This desert is fully described in Mazeh et al. (2016). Two main reasons have been proposed to explain this desert. The first reasons are that low-mass EGPs in the desert are too much irradiated by their host star, and consequently, they lose their atmosphere (as it is observed for some hot Jupiters, e.g., Vidal-Madjar et al. 2003). Thus, they migrate down in the mass-period diagram. The remnants of photo-evaporated hot Jupiters could be shortperiod super-Earths (Valsecchi et al. 2014). This was however recently ruled out by Winn et al. (2017) based on the different metallicity distribution of the host stars. The second hypothesis for this dearth of short-period EGP is that the distance at which planets stop their migration depends on the planet mass. This would be the case if the magnetospheric cavity in the inner-edge of protoplanetary disks, which is supposed to stop hot Jupiters migration (Chang et al. 2010), depends on the mass of the disk and indirectly to the mass of the planet (Mazeh et al. 2016). This would also be the case if hot Jupiters migrated through a high-eccentricity and tidal circularization, as the minimum distance between planets and their star depends on their Roche limit, which also depends on their mass (Matsakos and Königl 2016). However, there is evidence against this high-eccentricity migration model for hot Jupiters, as presented in Schlaufman and Winn (2016). Therefore, this dearth of planets is currently not clearly understood.

Multiplicity of EGP Spectroscopic surveys of hot Jupiters revealed that about half of them have longperiod (at several AU) massive companions, within the planetary or stellar regime (e.g., Knutson et al. 2014; Neveu-VanMalle et al. 2016). Except in the unique case of the WASP-47 system (Becker et al. 2015) where a hot Jupiter is sandwiched by two low-mass planets, this EGP population does not have nearby low-mass planets. They might have low-mass companions at wide separation, but current instrumentation cannot detect them. This supports the idea that hot Jupiters could have migrated with two main mechanisms: (1) through interaction with the disk (type I or II migration) or (2) by planet-planet interactions which would make a high eccentricity for the hot Jupiter progenitor caused either by the Lidov-Kozai effect or planet-planet scattering and before a tidal circularization (see Ford 2014, for a review). On the other hand, some temperate and cold giants as well as EGPs in the period valley have inner, low-mass planetary companions (see Fabrycky et al. 2014). This is evidence that these planets had a smooth disk migration that preserved the inner

1990

A. Santerne

planets, unlike the hot Jupiters. As aforementioned, in some circumstances, the presence of the companion could even be the reason for these EGPs to stay cool and prevent them from migrating inward and becoming hot Jupiters. The upcoming nextgeneration instruments, like ESPRESSO (Pepe et al. 2014), will be able to further probe the architecture of EGP systems and reveal their migration mechanism.

Radius of EGP The radius of EGPs that transit in front of their host star can be measured with relatively high precision. Figure 2 shows all EGP with measured radius as a function of their incident flux. Because the transit method is more sensitive to planet close to their star, most transiting EGPs known to date are highly irradiated. The EGPs receiving more than 109 erg.cm1 .s1 exhibit a radius up to 2.2 R (hence, 25 R˚ ). With current models of giant planet atmospheres and internal structure (that are calibrated on Jupiter and Saturn), it is not possible to explain such large radius for a gaseous planet, unless they are extremely young (Almenara et al. 2015). Several hypotheses are proposed (Baraffe et al. 2014) but they are still debated. Planets receiving an insolation flux of less than 108 erg.cm2 .s1 are yet poorly explored. They have orbital period typically longer than a month and require longduration space-based photometric surveys, like CoRoT and Kepler, to be detected. Nevertheless, the relatively few objects that were detected to date in this lowincident flux regime do not show signs of inflation. Their radii are in the range 0.5–1.2 R (equivalent to about 6–13 R˚ ).

2.5

Planet radius [RJup ]

2.0

1.5

1.0

0.5

0.0

1010

109

108

107

106

105

104

Stellar insolation flux [erg.cm−2 .s−1 ] Fig. 2 Radius of EGP as function of their stellar insolation flux. Jupiter and Saturn are displayed for comparison

98 Populations of Extrasolar Giant Planets from Transit and Radial Velocity Surveys

1991

As a summary, extrasolar giant planets are objects more massive than about 0.1–0.2 M . The upper limit in mass is a few tens of M but is unclear as the distinction between giant planets and brown dwarfs is still debated. In this regime of planetary mass resides two main populations of planets: the hot Jupiters, with orbital separation of less than 0.1 AU (about 10 days of orbital period) and the temperate/cold giants with semimajor axis greater than about 0.4 AU (about 100 days of period). In the transition, the so-called period valley, relatively few planets have been found. The RV and transit detections have also unveiled a clear desert of short-period low-mass giants, where no planet has been found. This desert might be the result of a dramatic evolution or a migration of giant exoplanets. Finally, most of the hot Jupiters only have wide separation more massive companions, while warm to cold giants might have inner, low-mass planetary companions, as detected with current instrumentation.

Occurrence Rates of EGPs in Different Stellar Populations Transit and RV surveys with well-defined and characterized stellar samples can be used to derive the occurrence rates of EGPs. One of the main challenges to derive occurrence rate is to correctly estimate the bias inherent of each technique or survey. In the case of the transit surveys, which have targeted relatively faint stars, another challenge is to characterize precisely the observed stellar sample (see Huber et al. 2014; Damiani et al. 2016, in the case of Kepler and CoRoT, respectively). Occurrence rates of EGPs have been derived based on different surveys that observed different stellar populations across the Galaxy. The values are reported in Table 1 and discussed below. The most up-to-date value of each survey is also displayed in Fig. 3 for comparison. For clarity of the plot, only values with a relative precision better than 50% are shown.

The Solar Neighborhood For now, only RV experiments substantially surveyed the solar neighborhood and were used to determine the occurrence rates of EGPs. The two main surveys of exoplanets in RV are the Californian Planet Search (CPS) and the Geneva-lead survey, initiated by Marcy et al. and Mayor et al., respectively. The CPS survey mostly used the Keck and Lick telescopes with their respective high-resolution iodine-cell-based spectrographs. Some observations were also performed with the Anglo-Australian Telescope. The occurrence rate estimates were first published in Marcy et al. (2005) and subsequently in Cumming et al. (2008), Howard et al. (2010), and Wright et al. (2012). The various studies used

1992

A. Santerne

Table 1 Occurrence rates of EGP for different ranges of orbital periods from different studies. All values are in percent Hot Jupiters P < 10 d [%] 1.2 ˙ 0.2 0.7 ˙ 0.5a 1.5 ˙ 0.6 0.83 ˙ 0.34b 1.20 ˙ 0.38

Period-valley giants 10