Mercury ; The View after Messenger [1 ed.] 1107154456, 9781107154452

Observations from the first spacecraft to orbit the planet Mercury have transformed our understanding of the origin and

260 18 61MB

English Pages 596 [600] Year 2018

Report DMCA / Copyright


Polecaj historie

Mercury ; The View after Messenger [1 ed.]
 1107154456, 9781107154452

Table of contents :
List of Contributors
1 The MESSENGER Mission: Science and Implementation Overview
2 The Chemical Composition of Mercury
3 Mercury’s Crust and Lithosphere: Structure and Mechanics
4 Mercury’s Internal Structure
5 Mercury’s Internal Magnetic Field
6 The Geologic History of Mercury
7 The Geochemical and Mineralogical Diversity of Mercury
8 Spectral Reflectance Constraints on the Composition and Evolution of Mercury’s Surface
9 Impact Cratering of Mercury
10 The Tectonic Character of Mercury
11 The Volcanic Character of Mercury
12 Mercury’s Hollows
13 Mercury’s Polar Deposits
14 Observations of Mercury’s Exosphere: Composition and Structure
15 Understanding Mercury’s Exosphere: Models Derived from MESSENGER Observations
16 Structure and Configuration of Mercury’s Magnetosphere
17 Mercury’s Dynamic Magnetosphere
18 The Elusive Origin of Mercury
19 Mercury’s Global Evolution
20 Future Missions: Mercury after MESSENGER
Index of Place Names

Citation preview

Mercury The View after MESSENGER Observations from the first spacecraft to orbit the planet Mercury have transformed our understanding of the origin and evolution of rocky planets. This volume is the definitive resource about Mercury for planetary scientists, from students to senior researchers. Topics treated in depth include Mercury’s chemical composition; the structure of its crust, lithosphere, mantle, and core; Mercury’s modern and ancient magnetic field; Mercury’s geology, including the planet’s major geologic units and their surface chemistry and mineralogy, its spectral reflectance characteristics, its craters and cratering history, its tectonic features and deformational history, its volcanic features and magmatic history, its distinctive hollows, and the frozen ices in its polar deposits; Mercury’s exosphere and magnetosphere and the processes that govern their dynamics and their interaction with the solar wind and interplanetary magnetic field; the formation and large-scale evolution of the planet; and current plans and needed capabilities to explore Mercury further in the future. se an c. s o lomon is Director of the Lamont-Doherty Earth Observatory and William B. Ransford Professor of Earth and Planetary Science at Columbia University. He earlier served as Director of the Department of Terrestrial Magnetism at the Carnegie Institution of Washington and Professor of Geophysics at the Massachusetts Institute of Technology. He was the Principal Investigator for NASA’s MESSENGER mission to Mercury from the initial mission concept in 1996 to the end of the project in 2018. He also served on the science teams for the Magellan mission to Venus, the Mars Global Surveyor mission, and the Gravity Recovery and Interior Laboratory mission to the Moon. A member of the U.S. National Academy of Sciences and the American Academy of Arts and Sciences and former President of the American Geophysical Union, Solomon in 2014 was awarded the National Medal of Science by President Barack Obama. la r ry r. n i tt le r conducts laboratory research on extraterrestrial materials and remote sensing observations of planets at the Carnegie Institution of Washington. He served on NASA’s MESSENGER mission to Mercury as Participating Scientist from 2007 to 2012 and Deputy Principal Investigator from 2012 to 2018. He earlier participated in the Near Earth Asteroid Rendezvous, Stardust, and Genesis missions and is currently a science team member on the Japan Aerospace Exploration Agency’s Hayabusa2 asteroid sample return mission and the BepiColombo mission to Mercury. He received the 2001 Alfred O. C. Nier Prize of the Meteoritical Society and was named Fellow of that society in 2010. Asteroid 5992 Nittler is named in his honor. br i a n j. a n de r s o n is Principal Professional Staff Physicist at The Johns Hopkins University Applied Physics Laboratory, having served earlier as Magnetospheric Section supervisor and Space Physics Group supervisor. For MESSENGER he was Magnetometer Instrument Scientist from 1999 to 2009 and Deputy Project Scientist from 2007 to 2018 while also serving as Co-Investigator from 2009 to 2018. He was spacecraft magnetics lead and is on the science team of NASA’s Magnetospheric Multiscale mission. He is the Principal Investigator of the National Science Foundation’s Active Magnetosphere and Planetary Electrodynamics Response Experiment. His research includes the physics of magnetospheres, plasma wave–particle physics, and planetary magnetic fields.

Cambridge Planetary Science Series Editors: Fran Bagenal, David Jewitt, Carl Murray, Jim Bell, Ralph Lorenz, Francis Nimmo, Sara Russell Books in the Series: 1. Jupiter: The Planet, Satellites and Magnetosphere† Edited by Bagenal, Dowling and McKinnon 978-0-521-03545-3 2. Meteorites: A Petrologic, Chemical and Isotopic Synthesis† Hutchison 978-0-521-03539-2 3. The Origin of Chondrules and Chondrites† Sears 978-1-107-40285-0 4. Planetary Rings† Esposito 978-1-107-40247-8 5. The Geology of Mars: Evidence from Earth-Based Analogs† Edited by Chapman 978-0-521-20659-4 6. The Surface of Mars† Carr 978-0-521-87201-0 7. Volcanism on Io: A Comparison with Earth† Davies 978-0-521-85003-2 8. Mars: An Introduction to its Interior, Surface and Atmosphere† Barlow 978-0-521-85226-5 9. The Martian Surface: Composition, Mineralogy and Physical Properties Edited by Bell 978-0-521-86698-9 10. Planetary Crusts: Their Composition, Origin and Evolution† Taylor and McLennan 978-0-521-14201-4 11. Planetary Tectonics† Edited by Watters and Schultz 978-0-521-74992-3 12. Protoplanetary Dust: Astrophysical and Cosmochemical Perspectives† Edited by Apai and Lauretta 978-0-521-51772-0 13. Planetary Surface Processes Melosh 978-0-521-51418-7

14. Titan: Interior, Surface, Atmosphere and Space Environment Edited by Müller-Wodarg, Griffith, Lellouch and Cravens 978-0-521-19992-6 15. Planetary Rings: A Post-Equinox View (Second edition) Esposito 978-1-107-02882-1 16. Planetesimals: Early Differentiation and Consequences for Planets Edited by Elkins-Tanton and Weiss 978-1-107-11848-5 17. Asteroids: Astronomical and Geological Bodies Burbine 978-1-107-09684-4 18. The Atmosphere and Climate of Mars Edited by Haberle, Clancy, Forget, Smith and Zurek 978-1-107-01618-7 19. Planetary Ring Systems Edited by Tiscareno and Murray 978-1-107-11382-4 20. Saturn in the 21st Century Edited by Baines, Flasar, Krupp and Stallard 978-1-107-10677-2 21. Mercury: The View after MESSENGER Edited by Solomon, Nittler and Anderson 978-1-107-15445-2 †

Reissued as a paperback

MERCURY The View after MESSENGER Edited by SEAN C. SOLOMON Lamont-Doherty Earth Observatory, Columbia University, New York, USA

LARRY R. NITTLER Carnegie Institution of Washington, Washington, DC, USA

BRIAN J . ANDERSON The Johns Hopkins University Applied Physics Laboratory, Laurel, Maryland, USA

University Printing House, Cambridge CB2 8BS, United Kingdom One Liberty Plaza, 20th Floor, New York, NY 10006, USA 477 Williamstown Road, Port Melbourne, VIC 3207, Australia 314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre, New Delhi – 110025, India 79 Anson Road, #06–04/06, Singapore 079906 Cambridge University Press is part of the University of Cambridge. It furthers the University’s mission by disseminating knowledge in the pursuit of education, learning, and research at the highest international levels of excellence. Information on this title: DOI: 10.1017/9781316650684 © Cambridge University Press 2018 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 2018 The MESSENGER mission was sponsored by the U.S. Government (contract #s NAS5-97271/24 and NASW-00002). Printed and bound in Great Britain by Clays Ltd, Elcograf S.p.A. A catalogue record for this publication is available from the British Library. Library of Congress Cataloging-in-Publication Data Names: Solomon, Sean C., editor. | Nittler, Larry R., editor. | Anderson, Brian J., editor. Title: Mercury : The view after MESSENGER / edited by Sean C. Solomon (Lamont-Doherty Earth Observatory, Columbia University, New York), Larry R. Nittler (Carnegie Institution of Washington, Washington, DC), Brian J. Anderson (The Johns Hopkins University, Applied Physics Laboratory, Laurel, Maryland). Other titles: View after MESSENGER Description: Cambridge : Cambridge University Press, [2018] | Includes bibliographical references. Identifiers: LCCN 2018022383 | ISBN 9781107154452 Subjects: LCSH: Mercury (Planet) – Observations. | MESSENGER (Spacecraft) Classification: LCC TL796.6.M47 M47 2018 | DDC 559.9/21–dc23 LC record available at ISBN 978-1-107-15445-2 Hardback Additional resources for this publication available at Cambridge University Press has no responsibility for the persistence or accuracy of URLs for external or third-party internet websites referred to in this publication and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.


List of Contributors Preface

page xi xv


The MESSENGER Mission: Science and Implementation Overview sean c. solomon and br ian j . anderson



The Chemical Composition of Mercury larr y r. nittler, nancy l. chabot, t imothy l. grove, and p atrick n. peplowski



Mercury’s Crust and Lithosphere: Structure and Mechanics r o ge r j. p h i l li p s , p a u l k . byr n e , p e t er b . ja m e s , e r w a n mazarico, gregory a. neumann, and mark e. p erry



Mercury’s Internal Structure jean-luc margot, steven a. hauck, ii, erwan mazarico, sebastiano padovan, and stanton j . peale



Mercury’s Internal Magnetic Field catherine l. johnson, brian j. anderson, haje korth, roge r j. p hillips, and l ydia c. p hilpott



The Geologic History of Mercury brett w. denevi, c arolyn m. ern s t , l o u i s e m . p r o c k t er , and mark s . r obinson



The Geochemical and Mineralogical Diversity of Mercury timothy j . mccoy, patrick n. peplowski, francis m. mccubb in, and shoshana z. weider



Spectral Reflectance Constraints on the Composition and Evolution of Mercury’s Surface s c ot t l . mu r c h i e, rac he l l . kl i ma, noam r . ize nb e r g, de bor ah l . d om i n gue , davi d t . b l e we t t , and jo¨ r n h e lb e r t




Impact Cratering of Mercury clark r. c hapman, david m. h. b aker, olivier s . b arnoui n, c al eb i. f as s et t , s im one marc hi , wi l li am j . merl ine, lillian r. ostr ach, louise m. prockter, and r o be r t g . s t r o m


The Tectonic Character of Mercury paul k. byrne, christian klimczak, and a. m. celaˆ l s¸ e ngo¨ r






The Volcanic Character of Mercury paul k. byrne, jennifer l. whitten, christian klimczak, francis m. mccubbin, and lillian r . ostrach



Mercury’s Hollows d av i d t . b l e w e t t , ca r o l y n m. e r n s t , sc o t t l . m u r c h i e, and f aith vilas



Mercury’s Polar Deposits nancy l. chabot, david j. lawrence, gregory a. neumann, william c. feldman, and david a. p aige



Observations of Mercury’s Exosphere: Composition and Structure william e. mcclintock, timothy a. cassidy, aimee w. merkel, rosemary m. killen, matthew h. burger, and ron a l d j . v e r v a c k, j r .



Understanding Mercury’s Exosphere: Models Derived from MESSENGER Observations rosemary m. killen, matthew h. burger, ronald j. vervack, jr., and timothy a. cassidy



Structure and Configuration of Mercury’s Magnetosphere 430 haje korth, brian j. anderson, catherine l. johnson, james a. s lavin, jim m. raines, and thomas h. zurbuchen


Mercury’s Dynamic Magnetosphere james a. s lavin, daniel n. b aker, daniel j . gershman, george c. ho, s uzanne m. imber, stamatios m. krimigis, and t orbjo¨ r n s u n d b e rg



The Elusive Origin of Mercury denton s. ebel and s arah t. stewart



Mercury’s Global Evolution steven a. hauck, ii, matthias grott, paul k. byrne, br e t t w . de n e vi , sab i ne s t anle y, and ti mot hy j . mc c oy



Future Missions: Mercury after MESSENGER ra l p h l . mc n u t t , j r. , j o h a n n es be n k h o f f , m a s a k i fujimoto, and brian j . anderson


Index Index of Place Names

570 582


den ton s. e bel Department of Earth and Planetary Sciences, American Museum of Natural History, New York, NY 10024, USA

b rian j. a nderson The Johns Hopkins University Applied Physics Laboratory, Laurel, MD 20723, USA

car ol yn m . ern st The Johns Hopkins University Applied Physics Laboratory, Laurel, MD 20723, USA

d a v i d m . h. ba k e r Solar System Exploration Division, NASA Goddard Space Flight Center, Greenbelt, MD 20771, USA and Department of Earth, Environmental and Planetary Sciences, Brown University Providence, RI 02912, USA

cal eb i . fassett NASA Marshall Space Flight Center, Huntsville, AL 35805, USA and Department of Astronomy, Mount Holyoke College, South Hadley, MA 01075, USA

d a n i e l n. ba k e r Laboratory for Atmospheric and Space Physics, University of Colorado, Boulder, CO 80303, USA

william c. f eldman Planetary Science Institute, Tucson, AZ 85719, USA

o l i v i e r s . ba r n o u i n The Johns Hopkins University Applied Physics Laboratory, Laurel, MD 20723, USA

ma s a k i f u j i m ot o Institute of Space and Astronautical Science, Japan Aerospace Exploration Agency, Sagamihara, Kanagawa, 252–5210, Japan

j o hannes benkhoff European Space Research and Technology Centre, European Space Agency, 2201 AZ Noordwijk, The Netherlands

daniel j. gershman Geospace Physics Laboratory, NASA Goddard Space Flight Center, Greenbelt, MD 20771, USA

david t. blewet t The Johns Hopkins University Applied Physics Laboratory, Laurel, MD 20723, USA

ma t t h i a s g r ot t Institute of Planetary Research, German Aerospace Center (DLR), 12489 Berlin, Germany

matth ew h . burger Space Telescope Science Institute, Baltimore, MD 21218, USA

timothy l. g rove Department of Earth, Atmospheric and Planetary Sciences, Massachusetts Institute of Technology, Cambridge, MA 02139, USA

p a ul k. byrn e Department of Marine, Earth, and Atmospheric Sciences, North Carolina State University, Raleigh, NC 27695, USA and Department of Terrestrial Magnetism, Carnegie Institution of Washington, Washington, DC 20015, USA

steven a. h au ck, ii Department of Earth, Environmental, and Planetary Sciences, Case Western Reserve University, Cleveland, OH 44106, USA

t i m o t h y a. ca s s i d y Laboratory for Atmospheric and Space Physics, University of Colorado, Boulder, CO 80303, USA

jo¨ rn helbert Institute of Planetary Research, German Aerospace Center (DLR), 12489 Berlin, Germany

n an cy l. ch ab ot The Johns Hopkins University Applied Physics Laboratory, Laurel, MD 20723, USA

george c. ho The Johns Hopkins University Applied Physics Laboratory, Laurel, MD 20723, USA

c lar k r. cha p m an Southwest Research Institute, Boulder, CO 80302, USA

suzan ne m. i mber Department of Physics and Astronomy, University of Leicester, Leicester, LE1 7RH, United Kingdom

b rett w. de nevi The Johns Hopkins University Applied Physics Laboratory, Laurel, MD 20723, USA

noam r. ize nberg The Johns Hopkins University Applied Physics Laboratory, Laurel, MD 20723, USA

d e b o r a h l. d o m i n g u e Planetary Science Institute, Tucson, AZ 85719, USA



List of Contributors

p eter b. james Department of Geosciences, Baylor University, Waco, TX 76706, USA and Lunar and Planetary Institute, Universities Space Research Association, Houston, TX 77058, USA cath erin e l. j o hnson Department of Earth, Ocean and Atmospheric Sciences, University of British Columbia, Vancouver, BC V6T 1Z4, Canada and Planetary Science Institute, Tucson, AZ 85719, USA rosemary m. killen Solar System Exploration Division, NASA Goddard Space Flight Center, Greenbelt, MD 20771, USA rach el l. klima The Johns Hopkins University Applied Physics Laboratory, Laurel, MD 20723, USA ch ristian k limczak Department of Geology, University of Georgia, Athens, GA 30602, USA and Department of Terrestrial Magnetism, Carnegie Institution of Washington, Washington, DC 20015, USA haj e korth The Johns Hopkins University Applied Physics Laboratory, Laurel, MD 20723, USA s t a m a t i os m . k r i m i g i s The Johns Hopkins University Applied Physics Laboratory, Laurel, MD 20723, USA and Office of Space Research and Technology, Academy of Athens, Athens 10679, Greece d a v i d j. la w r e n c e The Johns Hopkins University Applied Physics Laboratory, Laurel, MD 20723, USA simone marc hi Southwest Research Institute, Boulder, CO 80302, USA j e an - l u c m a r g o t Department of Earth, Planetary, and Space Sciences, University of California, Los Angeles, Los Angeles, CA 90095, USA e r w an m a z a r i c o Solar System Exploration Division, NASA Goddard Space Flight Center, Greenbelt, MD 20771, USA

a i m e e w. m e r k e l Laboratory for Atmospheric and Space Physics, University of Colorado, Boulder, CO 80303, USA william j . me rlin e Southwest Research Institute, Boulder, CO 80302, USA s c ott l. m ur c hi e The Johns Hopkins University Applied Physics Laboratory, Laurel, MD 20723, USA gregory a. n eumann Solar System Exploration Division, NASA Goddard Space Flight Center, Greenbelt, MD 20771, USA l arr y r. nittler Department of Terrestrial Magnetism, Carnegie Institution of Washington, Washington, DC 20015, USA l illian r. ostrach U.S. Geological Survey, Flagstaff, AZ 86001, USA and Solar System Exploration Division, NASA Goddard Space Flight Center, Greenbelt, MD 20771, USA s e b as t i a n o p a d ov a n Institute of Planetary Research, German Aerospace Center (DLR), 12489 Berlin, Germany david a . paige Department of Earth, Planetary, and Space Sciences, University of California, Los Angeles, Los Angeles, CA 90095, USA st an ton j . peale (deceased) Department of Physics, University of California, Santa Barbara, Santa Barbara, CA 93106, USA p a tric k n. p eplo w s k i The Johns Hopkins University Applied Physics Laboratory, Laurel, MD 20723, USA mark e. perr y The Johns Hopkins University Applied Physics Laboratory, Laurel, MD 20723, USA r oger j. p hillip s Department of Earth and Planetary Sciences and McDonnell Center for the Space Sciences, Washington University, St. Louis, MO 63130, USA

william e. mcclintock Laboratory for Atmospheric and Space Physics, University of Colorado, Boulder, CO 80303, USA

l yd i a c. p h i l p o t t Department of Earth, Ocean and Atmospheric Sciences, University of British Columbia, Vancouver, BC V6T 1Z4, Canada

timothy j. mccoy Department of Mineral Sciences, National Museum of Natural History, Smithsonian Institution, Washington, DC 20560, USA

l ou i s e m . p r o c k t e r Lunar and Planetary Institute, Universities Space Research Association, Houston, TX 77058, USA

f rancis m . m ccub bin NASA Johnson Space Center, Houston, TX 77058, USA

j im m . ra in es Climate and Space Sciences and Engineering Department, University of Michigan, Ann Arbor, MI 48109, USA

ralp h l. m cnutt, j r. The Johns Hopkins University Applied Physics Laboratory, Laurel, MD 20723, USA

mark s. robin s on School of Earth and Space Exploration, Arizona State University, Tempe, AZ 85287, USA

List of Contributors a. m. celaˆ l s¸ eng o¨ r Department of Geology, Faculty of Mines and Eurasia Institute of Earth Sciences, Istanbul Technical University, Istanbul 34810, Turkey j a me s a . s l a v i n Climate and Space Sciences and Engineering Department, University of Michigan, Ann Arbor, MI 48109, USA s e an c . s o l o m o n Lamont-Doherty Earth Observatory, Columbia University, Palisades, NY 10964, USA sabine stanley Department of Earth and Planetary Sciences, Johns Hopkins University, Baltimore, MD 21218, USA


tor bjo¨ rn sundberg Queen Mary University of London, London, E1 4NS, United Kingdom ron a ld j. vervack, jr. The Johns Hopkins University Applied Physics Laboratory, Laurel, MD 20723, USA fait h v ilas Planetary Science Institute, Tucson, AZ 85719, USA shoshana z. weider Department of Terrestrial Magnetism, Carnegie Institution of Washington, Washington, DC 20015, USA

sarah t. s tewart Department Earth and Planetary Sciences, University of California, Davis, Davis, CA 95616, USA

jennife r l. wh itt en Center for Earth and Planetary Studies, National Air and Space Museum, Smithsonian Institution, Washington, DC 20013, USA

r obe rt g. strom Lunar and Planetary Laboratory, University of Arizona, Tucson, AZ 85721, USA

thomas h. zurb uchen Climate and Space Sciences and Engineering Department, University of Michigan, Ann Arbor, MI 48109, USA


Mercury, like Earth, is one of only four rocky planets in our solar system, yet the exploration of the innermost planet by spacecraft has lagged substantially behind that of Earth’s two nearest neighbors, Venus and Mars. The first spacecraft to visit Venus was Mariner 2, which was developed by the National Aeronautics and Space Administration (NASA) and flew by Venus in December 1962. The first successful spacecraft flyby of Mars was by Mariner 4 in July 1965. The first spacecraft to orbit Mars was Mariner 9, which arrived at the red planet in November 1971. The first probe to orbit Venus was the Soviet Union’s Venera 9 orbiter, which arrived at Venus in October 1975. Both planets have been visited dozens of times since those early missions by other spacecraft sent by multiple nations and space agencies. In contrast, the first spacecraft encounter of Mercury was not until March 1974, when Mariner 10 completed the first of its three Mercury flybys. The third and final Mariner 10 flyby of Mercury was one year later in March 1975, and no spacecraft visited Mercury again for more than three decades. The spacecraft exploration of Mercury resumed when NASA’s MErcury Surface, Space ENvironment, GEochemistry, and Ranging (MESSENGER) probe flew by Mercury three times in 2008–2009 and became the first spacecraft to orbit Mercury on 18 March 2011. MESSENGER operated in orbit about Mercury for more than four years, until 30 April 2015, and acquired the first global observations of Mercury’s surface, interior, exosphere, magnetosphere, and heliospheric environment. The MESSENGER project and its science team continued to validate, archive, and analyze data acquired during the mission for more than two additional years, until the project formally ended on 30 September 2018. The first spacecraft orbital mission to any planet, like Mariner 9 and the Venera 9 orbiter, enables many important discoveries and produces large new data sets. Collectively those first orbital data sets from a planet drive major increases in scientific understanding and raise multiple new scientific questions, not only about the target body but also about planetary and solar system processes more generally. So it has been with observations made by the MESSENGER spacecraft. The wealth of new data returned by MESSENGER and archived with NASA’s Planetary Data System by the MESSENGER team continues to foster new investigations of this nearby yet remarkably distinctive sibling of Earth, and at the same time prompts new questions that expand the rationale for continued Mercury exploration. This book is intended to synthesize the findings from the MESSENGER mission into a description of our current scientific understanding of Mercury. The book is timely, for two reasons. First, it was written after the end of data collection by MESSENGER, so that all of the measurements acquired over the course of the mission could be integrated and our markedly improved knowledge of Mercury could serve to update our understanding of the formation and evolution of the inner solar system’s rocky planets. Second, the book was completed approximately eight years before the scheduled arrival of the next spacecraft at Mercury, the dual probes of the BepiColombo mission of the European Space Agency and the Japan Aerospace Exploration Agency.



Preface The editors of this volume owe considerable thanks to many colleagues whose efforts contributed to the technical and scientific success of the MESSENGER mission. Among these individuals are members of the science team who, because of other responsibilities and interests, were not able to share in the writing for this book but in myriad other ways participated in the analysis and interpretation of observations from the mission. Hundreds of engineers, technicians, software developers, managers, and support personnel contributed to the successful design, construction, testing, launch, and operation of the MESSENGER spacecraft. Among those, eight warrant special thanks: the four individuals at the Johns Hopkins University Applied Physics Laboratory who served successively as MESSENGER Project Manager – Max R. Peterson, David G. Grant, Peter D. Bedini, and Helene L. Winters – and the four who served successively as MESSENGER’s Mission Systems Engineer – Andrew G. Santo, James C. Leary, Eric J. Finnegan, and Daniel J. O’Shaughnessy. Each played a vital leadership role at a critical stage in the MESSENGER mission. The editors are also indebted to the authors of the 20 chapters in this volume, most drawn from the MESSENGER science team but a few from outside the project who bring special expertise on a topic of importance to their chapter. Each of the chapters was reviewed not only by other members of the MESSENGER science team but also by an expert scientist from outside the project. The editors appreciate the thoughtful reviews of individual chapters by Erik Asphaug, Wolfgang Baumjohann, Doris Breuer, Masaki Fujimoto, Walter S. Kiefer, François Leblanc, H. Jay Melosh, Edwin J. Mierkiewicz, Stephen W. Parman, David A. Rothery, Christopher T. Russell, Richard A. Schultz, Matthew A. Siegler, Krista M. Soderlund, S. Alan Stern, David J. Stevenson, Jessica M. Sunshine, G. Jeffrey Taylor, Rebecca J. Thomas, and David A. Williams. The editors are deeply grateful to Kimberly Schermerhorn for her substantial assistance with the preparation of all of the material for this book. We thank Nancy Chabot and Brett Denevi for their design of the maps located inside the front and back covers of the book; these maps provide a compact illustration of much of the data returned by MESSENGER as well as a ready means to locate many of the major features on Mercury’s surface mentioned in the book’s chapters. Brett also designed the image on the front cover and selected the MESSENGER images shown on the back cover. We thank Magda Saina for lending her graphical expertise to convert a number of the figures in this volume to publication quality. Finally, we thank the editors at Cambridge University Press who worked with us from early discussions, to the writing of a formal book proposal, through the preparation of all of the chapters and supporting material, to copy editing and final production. We are particularly grateful for the sustained guidance of Lucy Edwards, Vince Higgs, and Esther Migueliz. It is our hope that this volume will provide a standard reference on the planet Mercury for a number of years, at least until the next spacecraft after MESSENGER arrive to renew humankind’s exploration of our solar system’s innermost world. Sean C. Solomon, Larry R. Nittler, and Brian J. Anderson

1 The MESSENGER Mission: Science and Implementation Overview S E AN C. S OLOM ON AND BRIA N J. AND ERS ON

1 .1 I N TR O D UCT I O N

After the Mariner 10 mission, the next logical step in the exploration of Mercury was widely viewed to be an orbiter mission (COMPLEX, 1978), and several notable discoveries by groundbased astronomers in the years since the Mariner 10 encounters (e.g., Potter and Morgan, 1985, 1986; Slade et al., 1992; Harmon and Slade, 1992) provided a wealth of new information about Mercury that whetted the appetite of the planetary science community for orbital observations. Nevertheless, substantial advances were needed in mission design, thermal engineering, and miniaturization of instruments and spacecraft subsystems before such a mission could be considered technically ready. The MErcury Surface, Space ENvironment, GEochemistry, and Ranging (MESSENGER) mission to orbit Mercury was proposed under NASA’s Discovery Program in 1996 and again in 1998 (Solomon et al., 2001; Gold et al., 2001; Santo et al., 2001) and was selected for flight in 1999. Development, construction, integration, and testing of the spacecraft and its instruments began in January 2000 and spanned the four and a half years leading to launch on 3 August 2004 (McNutt et al., 2006). MESSENGER completed gravity-assist flybys of Earth once, Venus twice, and Mercury three times (Figure 1.1) during a mission cruise phase that lasted 6.6 years. MESSENGER was inserted into orbit about Mercury on 18 March 2011 and conducted orbital observations of the innermost planet for more than four years, until 30 April 2015. In this chapter we provide an overview of the MESSENGER mission from a historical perspective, including the mission’s scientific objectives; the payload characteristics, data acquisition planning, and operational procedures adopted to achieve those objectives; and the scientific findings from flyby and orbital operations. We begin with summaries of the mission objectives, spacecraft, payload instruments, and orbit design. We then describe the procedures adopted to optimize the scientific return from the complex series of orbital data acquisition operations. We follow with an account of the primary mission, including the Mercury flybys and the first year of orbital observations. We then outline the rationale for and accomplishments of MESSENGER’s first extended mission, conducted over the second year of orbital operations, and the second extended mission, conducted over the final two years of orbital operations. The second extended mission included a distinctive lowaltitude campaign completed at the culmination of the mission. A concluding section briefly introduces the other chapters of this book.

Although a sibling of Earth, Venus, and Mars, the planet Mercury is an unusual member of the family (Solomon, 2003). Among the planets of our solar system, it is the smallest, at little more than 5% of an Earth mass, but its bulk density corrected for the effect of internal compression is the highest. Mercury’s orbit is the most eccentric of the planets, and it is the only known solar system object in a 3:2 spin–orbit resonance, in which three sidereal days equal two periods of Mercury’s revolution about the Sun. Mercury is the only inner planet other than Earth to host an internal magnetic field and an Earth-like magnetosphere capable of standing off the solar wind. The closest planet to the Sun, Mercury experiences a variation in surface temperature at the equator of 600°C over the course of a solar day, which because of Mercury’s slow spin rate equals two Mercury years. The permanently shadowed floors of Mercury’s highlatitude craters nonetheless are sufficiently cold to have trapped water ice and other frozen volatiles. Thought to have been created by the same processes as the other inner planets and at the same early stage in the history of the solar system, Mercury with its unusual attributes has long held out the promise of deepening our understanding of how Earth and other Earth-like planets formed and evolved. Yet Mercury is not an easy object to study. Never separated from the Sun by more than 28° of arc when viewed from Earth, Mercury is forbidden as a target for the Hubble Space Telescope and other astronomical facilities because their optical systems would be severely damaged by exposure to direct sunlight. Located deep within the gravitational potential well of the Sun, Mercury has also long presented a challenge to spacecraft mission design. The first spacecraft to view Mercury at close range was Mariner 10, which after flying once by Venus encountered the innermost planet three times in 1974–1975. The encounters occurred nearly at Mercury’s greatest distance from the Sun and were spaced approximately one Mercury solar day apart, so the same hemisphere of the planet was in sunlight at each flyby. Mariner 10 obtained images of 45% of the surface, discovered the planet’s global magnetic field, assayed three neutral species (H, He, and O) in Mercury’s tenuous atmosphere, and sampled the magnetic field and energetic charged particles in Mercury’s dynamic magnetosphere (Dunne and Burgess, 1978).



The MESSENGER Mission Figure 1.1. Image mosaic of Mercury acquired on departure from MESSENGER’s first Mercury flyby on 14 January 2008. Mercury Dual Imaging System wide-angle camera images acquired through the narrow-band filters centered at 1000, 700, and 430 nm are projected in red, green, and blue in this color representation. Much of the area shown had not been imaged by Mariner 10.

1 .2 M IS SION OBJECTIV ES , SPA CECR AFT, PA YLOA D, AND OR BIT D ES IGN 1.2.1 Key Scientific Questions The MESSENGER mission was designed to address six key scientific questions. The questions were motivated by the knowledge of Mercury available at the time the mission was proposed, were capable of being substantially addressed by measurements that could be made from orbit, and would yield answers that would bear not only on the nature of Mercury but more generally on the origin and comparative evolution of the inner planets as a group. Those questions and a brief summary of the rationale for each were as follows.

impact (e.g., Cameron, 1985; Wetherill, 1988; Benz et al., 1988). Those theories differed in their predictions for the bulk composition of the silicate fraction of the planet (e.g., Lewis, 1988), including the upper crust, which would be visible to geochemical remote sensing instruments on an orbiting spacecraft. Moreover, ground-based telescopic measurements of Mercury’s surface reflectance showed few if any absorption features commonly seen in reflectance spectra of the Moon, Mars, and asteroids and attributable to the presence of ferrous iron in silicate minerals (e.g., Vilas, 1988), indicating both a low abundance of ferrous iron on Mercury’s surface and the need to rely heavily on elemental remote sensing instruments to gain compositional information. What Is the Geological History of Mercury? What Planetary Formational Processes Led to the High Ratio of Metal to Silicate in Mercury? The Mariner 10 spacecraft carried no elemental remote sensing instruments, so at the time the MESSENGER mission was proposed the single most important piece of information about the planet’s bulk composition was its high uncompressed density, which implied that Mercury has an iron-rich core that occupies much higher fractions of the planet’s mass and volume than do the cores of the other inner planets (e.g., Siegfried and Solomon, 1974). A variety of theories for the origin and early evolution of Mercury had been advanced to account for its high metal fraction, including formation from metal-enriched precursors resulting from either high-temperature fractionation or aerodynamical sorting in the solar nebula (e.g., Weidenschilling, 1978; Lewis, 1988) or removal of an initially larger silicate crust and mantle by evaporation or giant

Because of Mercury’s size, intermediate between the Moon and Mars, as well as its high metal/silicate ratio, documenting the geological history of Mercury was viewed as crucial to understanding how terrestrial planet evolution depends on planet size and initial conditions. A broad geological history of Mercury had been developed from Mariner 10 images (e.g., Strom, 1979; Spudis and Guest, 1988), but the limited coverage and resolution of those images left many aspects of that history uncertain. Extensive plains units were documented by Mariner 10, and the youngest of those plains deposits were seen to be in stratigraphic positions similar to the volcanic lunar maria. Unlike the maria, however, the plains deposits on Mercury are not markedly lower in reflectance than the surrounding older terrain, and no volcanic landforms were visible at the resolution of Mariner 10 images, so both volcanic and impact ejecta processes for plains emplacement had been suggested (e.g., Strom et al., 1975; Wilhelms,

1.2 Objectives, Spacecraft, Payload, and Orbit Design 1976) and the importance of volcanism in Mercury’s history was thus uncertain. Deformational features on Mercury were seen to be dominantly contractional, leading to the proposal that such features were the expression of global contraction resulting from interior cooling (e.g., Strom et al., 1975), although the restricted imaging coverage meant that the global contraction hypothesis remained untested over slightly more than a full hemisphere of the planet. What Are the Nature and Origin of Mercury’s Magnetic Field? Measurements by Mariner 10 demonstrated that Mercury has an internal magnetic field (Ness et al., 1976) with a dipole component nearly orthogonal to the planet’s orbital plane and an estimated moment near 300 nT RM3, where RM is Mercury’s mean radius (Connerney and Ness, 1988). Because external sources can dominate the total field measured at Mercury, and because of the limited sampling of the field during the two Mariner 10 flybys that penetrated Mercury’s magnetosphere, the uncertainty in Mercury’s dipole moment derived from Mariner 10 data was a factor of 2, and higher-order terms were linearly dependent and thus not resolvable (Connerney and Ness, 1988). A variety of mechanisms for producing Mercury’s observed magnetic field had been proposed, including remanent or fossil fields in Mercury’s crust and lithosphere (Stephenson, 1976; Srnka, 1976; Aharonson et al., 2004), hydromagnetic dynamos in a fluid outer core (e.g., Schubert et al., 1988; Stanley et al., 2005; Christensen, 2006), and a thermoelectric dynamo driven by temperature differences along the top of the core (Stevenson, 1987; Giampieri and Balogh, 2002). The different field generation models made different predictions regarding the geometry of the field, particularly for terms of higher order than the dipole term, and so measurements made from orbit about the planet were seen to be needed to distinguish among hypotheses. What Are the Structure and State of Mercury’s Core? The size and physical state of Mercury’s core are key to understanding the planet’s bulk composition, thermal history, and magnetic field generation processes (Zuber et al., 2007). Peale (1976) realized that the existence and radius of a liquid outer core on Mercury can be determined by the measurement of Mercury’s obliquity, the amplitude of its physical libration forced by variations in the torque exerted by the gravitational pull of the Sun over the planet’s 88-day orbit period, and two quantities that define the shape of the planet’s gravity field at spherical harmonic degree and order 2. The required coefficients in the spherical harmonic expansion of Mercury’s gravity field had been estimated from radio tracking of the Mariner 10 flybys (Anderson et al., 1987) but not with high precision. All four quantities can be determined from measurements made by an orbiting spacecraft with sufficient precision to determine Mercury’s polar moment of inertia and the moment of inertia of the planet’s solid outer shell that participates in the 88-day libration (Peale, 1976: Peale et al., 2002), and from those quantities important aspects of Mercury’s internal structure can be


resolved. Mercury’s obliquity and forced libration amplitude can also be measured from Earth-based radar observations, and such measurements were reported by Margot et al. (2007) before MESSENGER was inserted into orbit around Mercury, and then refined several years later (Margot et al., 2012). Although the measurements of libration amplitude and obliquity indicated that Mercury does indeed possess a fluid outer core (Margot et al., 2007), the uncertainties in Mercury’s gravitational field coefficients at harmonic degree 2 dominated the uncertainty in the planet’s moments of inertia. Radio tracking of an orbiting spacecraft was required to improve the determination of these key quantities. What Are the Radar-Reflective Materials at Mercury’s Poles? The discovery in 1991 of radar-bright regions near Mercury’s poles and the similarity of the radar reflectivity and polarization characteristics of such regions to those of icy satellites and the south residual polar cap of Mars led to the proposal that these areas host deposits of surface or near-surface water ice (Slade et al., 1992; Harmon and Slade, 1992). Subsequent radar imaging at improved resolution confirmed that the radar-bright deposits are confined to the floors of near-polar impact craters (e.g., Harmon et al., 2011). Because of Mercury’s small obliquity, sufficiently deep craters are permanently shadowed and are predicted to be at temperatures at which water ice is stable for billions of years (Paige et al., 1992). Although a contribution from interior outgassing could not be excluded, impact volatilization of cometary and meteoritic material followed by transport of water molecules to polar cold traps was shown to provide sufficient polar ice to match the characteristics of the deposits (Moses et al., 1999). Two alternative explanations for the radar-bright polar deposits of Mercury were nonetheless suggested. One was that the polar deposits are composed of elemental sulfur, on the grounds that sulfur would be stable in polar cold traps and the presence of sulfides in the regolith can account for a high disk-averaged index of refraction and low microwave opacity of surface materials (Sprague et al., 1995). The second was that the permanently shadowed portions of polar craters are radar-bright not because of trapped volatiles but because of either unusual surface roughness (Weidenschilling, 1998) or low dielectric loss (Starukhina, 2001) of near-surface silicates at extremely cold temperatures. Geochemical remote sensing measurements made from orbit around Mercury were recognized as able to distinguish among the competing proposals. What Are the Important Volatile Species and Their Sources and Sinks on and near Mercury? Mercury’s atmosphere is a surface-bounded exosphere for which the composition and behavior are controlled by interactions with the magnetosphere and the surface. At the time the MESSENGER mission was under development, the atmosphere was known to contain at least six elements (H, He, O, Na, K, Ca). The Mariner 10 airglow spectrometer detected H and He and set an upper bound on O (Broadfoot et al., 1976), and ground-based spectroscopic observations led to the discovery



of exospheric Na (Potter and Morgan, 1985), K (Potter and Morgan, 1986), and Ca (Bida et al., 2000). Exospheric H and He were thought to be dominated by solar wind ions neutralized by recombination at the surface, whereas proposed source processes for other exospheric species included diffusion from the planet’s interior, evaporation, sputtering by photons and energetic ions, chemical sputtering by protons, and meteoroid impact and vaporization (e.g., Killen and Ip, 1999). That several of these processes play some role was suggested by the strong variations in exospheric characteristics observed as functions of local time, solar distance, and level of solar activity (e.g., Sprague et al., 1998; Hunten and Sprague, 2002; Leblanc and Johnson, 2003). It was long recognized that a spacecraft in orbit about Mercury can provide a range of opportunities for elucidating further the nature of the exosphere, through profiles of major exospheric neutral species versus time of day and solar distance and searches for new species (e.g., Domingue et al., 2007). In situ measurement of energetic and thermal plasma ions from orbit can also detect solar wind pickup ions that originated as exospheric neutral atoms (e.g., Koehn et al., 2002). 1.2.2 Scientific Objectives The six key questions above led to a set of scientific objectives for the MESSENGER mission and in turn to a set of project requirements (Solomon et al., 2001), a suite of payload instruments (Gold et al., 2001), and a measurement strategy (Section 1.3). The scientific objectives for MESSENGER’s primary mission are given in Table 1.1, and the project requirements for the primary mission are given in Table 1.2. The objective to characterize the chemical composition of Mercury’s surface led to a project requirement for global maps of major element composition at a resolution sufficient to discern the principal geological units and to distinguish material excavated and ejected by young impact craters from a possible veneer of cometary and meteoritic material. Information on surface mineralogy was also deemed important for this objective. The objective to determine the planet’s geological history led to a project requirement for global monochrome imaging at a resolution of hundreds of meters or better, for topographic profiles across key geological features from altimetry or stereo, and for spectral measurements of major geologic units at spatial resolutions of several kilometers or better. The objective to characterize Mercury’s magnetic field led to a project requirement for magnetometry,

Table 1.1. Scientific objectives for MESSENGER’s primary mission. 1. Determine the chemical composition of Mercury’s surface. 2. Determine Mercury’s geological history. 3. Determine the nature of Mercury’s magnetic field. 4. Determine the size and state of Mercury’s core. 5. Determine the volatile inventory at Mercury’s poles. 6. Determine the nature of Mercury’s exosphere and magnetosphere.

both near the planet and throughout the magnetosphere, as well as for energetic particle and plasma measurements so as to assist in the separation of external and internal fields. The objective to estimate the size and state of Mercury’s core led to the project requirement for altimetric measurement of the amplitude of Mercury’s physical libration as well as determination of the planet’s obliquity and low-degree gravitational field. The objective to assay the volatile inventory at Mercury’s poles led to the project requirement for ultraviolet spectrometry of the polar atmosphere and for gamma-ray and neutron spectrometry, imaging, and altimetry of polar-region craters. The objective to characterize the nature of Mercury’s exosphere and magnetosphere led to the project requirement to identify all major neutral species in the exosphere and charged species in the magnetosphere. 1.2.3 Spacecraft The design of the MESSENGER spacecraft (Figure 1.2) was driven largely by two requirements: to minimize mass and to survive the harsh thermal environment at Mercury (Santo et al., 2001; Leary et al., 2007). The largest launch vehicle available to the Discovery Program was the Delta II 7925-H, which could inject ~1100 kg into the required interplanetary trajectory. Because more than half of that total launch mass was needed for the propellant required to achieve the mission design, only 500 kg remained for the total spacecraft dry mass. To meet this constraint, the spacecraft structure was fabricated primarily with lightweight composite material and was fully integrated with a dual-mode propulsion system that

Table 1.2. Project requirements for MESSENGER’s primary mission. 1.

Provide major-element maps of Mercury to 10% relative uncertainty on the 1000-km scale and determine local composition and mineralogy at the ~20-km scale. 2a. Provide a global map with >90% coverage (monochrome) at 250-m average resolution and >80% of the planet imaged stereoscopically. 2b. Provide a global multispectral map at 2-km/pixel average resolution. 2c. Sample half of the northern hemisphere for topography at 1.5-m average height resolution. 3. Provide a multipole magnetic field model resolved through quadrupole terms with an uncertainty of less than ~20% in the dipole magnitude and direction. 4. Provide a global gravity field to degree and order 16 and determine the ratio of the solid-planet moment of inertia to the total moment of inertia to ~20% or better. 5. Identify the principal component of the radar-reflective material at Mercury’s north pole. 6. Provide altitude profiles at 25-km resolution of the major neutral exospheric species and characterize the major ionspecies energy distributions as functions of local time, Mercury heliocentric distance, and solar activity.

1.2 Objectives, Spacecraft, Payload, and Orbit Design Low-gain antennas +x Pitch Yaw Roll +y +z (anti-Sun)



Large velocity adjust (LVA) thruster


Figure 1.2. Engineering view of the MESSENGER spacecraft from behind the sunshade.

Helium tank

Front phased-array/ fanbeam antennas

Star trackers

Back phased-array/ fanbeam/low-gain antennas Solar array (back)

Launch vehicle adapter

Propellant tank (1 of 3)

featured lightweight tanks for propellant (hydrazine), oxidizer (nitrous tetroxide), and pressurant (gaseous helium). The propulsion system included a total of 17 thrusters: a single large velocity adjustment bipropellant thruster; four 22-N monopropellant thrusters for thrust-vector steering during large spacecraft maneuvers and for trajectory-correction maneuvers; and 12 4.4-N monopropellant thrusters for attitude control, angular momentum management, and small trajectory-correction maneuvers. A large number of mass-reduction measures were used in the development of the spacecraft. To avoid a cumbersome gimbaled antenna and the challenges associated with testing and operating it at high temperatures, an electronically steerable phased-array system was developed for the high-gain antenna. Used one at a time, each of two antennas – one on the spacecraft’s Sun-facing side and one aft – could be steered about one axis while the spacecraft body rolled about a second axis to point the antenna toward Earth at any point in the mission. The phased-array antennas were complemented with two mediumgain fanbeam antennas and four low-gain antennas. Radio signals were transmitted to and received from the MESSENGER spacecraft at X-band frequencies (7.2-GHz uplink, 8.4-GHz downlink) by the 34-m and 70-m antennas at NASA’s Deep Space Network stations in Goldstone, California; Madrid, Spain; and Canberra, Australia. Mass was also conserved by limiting the number of spacecraft components that moved. With the lone exception of the imaging system (see next section), all science instruments were hard-mounted to the spacecraft. As a consequence, spacecraft attitude often had to be changed continuously in orbit about Mercury to permit the instruments to make their observations. Spacecraft power was provided by two solar arrays (Figures 1.2 and 1.3), which could be articulated to manage array temperature, and by a battery during those orbits when


the spacecraft was on Mercury’s nightside and the Sun was eclipsed. In a fully redundant electronics system, a main processor performed all nominal spacecraft functions, while two other processors monitored spacecraft health and safety. The spacecraft attitude control system was three-axis stable and momentum biased and made use of four reaction wheels. Attitude knowledge was acquired through an inertial measurement unit, two star trackers, and a suite of Sun sensors as a backup to the primary attitude sensors. Primarily passive thermal management techniques were used to minimize heating of spacecraft subsystems by the Sun and the dayside surface of Mercury. To protect the spacecraft from solar heating, all systems except the solar arrays were kept behind a ceramic-cloth sunshade that pointed toward the Sun. This approach simplified the design of the subsystems, which could be built with conventional electronics, but added a substantial constraint to the operation of the spacecraft. Throughout its time within the inner solar system and in orbit about Mercury, MESSENGER was constrained to maintain the orientation of the normal to the central sunshade panel in the sunward direction to within ±10° in Sun-relative elevation angle (pitch) and ±12° in Sun-relative azimuth (yaw) at all times. 1.2.4 Instrument Payload The project requirements for MESSENGER’s primary mission were met by a suite of seven scientific instruments plus the spacecraft communication system (Gold et al., 2001). There was a dual imaging system for wide and narrow fields of view, monochrome and color imaging, and stereo; gammaray, neutron, and X-ray spectrometers for surface chemical mapping; a magnetometer; a laser altimeter; a combined ultraviolet–visible and visible–near-infrared spectrometer to survey both exospheric species and surface mineralogy; and a combined energetic particle and plasma spectrometer to



Figure 1.3. View of the MESSENGER spacecraft during vibration testing at the Johns Hopkins University Applied Physics Laboratory. The solar arrays (mirrored surfaces) are stowed in their positions at the time of launch. Also visible are the Magnetometer boom (center), similarly in its stowed position, and thermal blankets (gold).

sample charged species in the magnetosphere (Figure 1.4). Brief descriptions of the payload instruments are as follows. Mercury Dual Imaging System The Mercury Dual Imaging System (MDIS) on the MESSENGER spacecraft (Hawkins et al., 2007), shown in Figure 1.4, consisted of a monochrome narrow-angle camera (NAC) and a multispectral wide-angle camera (WAC). The NAC was an off-axis reflector with a 1.5° field of view (FOV) and was co-aligned with the WAC, a four-element refractor with a 10.5° FOV and a 12-color filter wheel. The focal-plane electronics of each camera were identical and used a 1024 × 1024 charge-coupled-device detector. Only one camera operated at a time, a design that allowed them to share a common set of control electronics. The NAC and the WAC were mounted on a pivoting platform that provided a 90° field of regard, from 40° sunward to 50° anti-sunward from the spacecraft z-axis (Figure 1.2) – the boresight direction of most of MESSENGER’s instruments. Onboard data compression provided capabilities for pixel binning, remapping of

12-bit data to 8 bits, and lossless or lossy compression. During MESSENGER’s primary mission, four main MDIS data sets were planned: a monochrome global image mosaic at near-zero emission angles and moderate incidence angles, a stereo complement map at off-nadir geometry and near-identical lighting, multicolor images at low incidence angles, and targeted high-resolution images of key surface features. It was further planned that those data would be used to construct a global image base map, a digital terrain model, global maps of color properties, and mosaics of high-resolution image strips. Gamma-Ray and Neutron Spectrometer The Gamma-Ray and Neutron Spectrometer (GRNS) instrument (Figure 1.4) included separate Gamma-Ray Spectrometer (GRS) and Neutron Spectrometer (NS) sensors (Goldsten et al., 2007). The GRS detector was a mechanically cooled crystal of germanium, and the sensor detected gamma-ray emissions in the energy range 0.1–10 MeV and achieved an energy resolution of 3.5 keV full width at half maximum for 60Co (1332 keV). Special construction techniques provided the necessary thermal isolation to

1.2 Objectives, Spacecraft, Payload, and Orbit Design

Gamma-Ray Spectrometer (GRNS/GRS)

Mercury Atmospheric and Surface Composition Spectrometer (MASCS)


Mercury Laser Altimeter (MLA)

X-Ray Spectrometer Solar Assembly (XRS/SAX) Fast Imaging Plasma Spectrometer (EPPS/FIPS)

Mercury Dual Imaging System (MDIS)

X-Ray Spectrometer Mercury Unit (XRS/MXU)

Energetic Particle Spectrometer (EPPS/EPS)

Data Processing Unit (DPU)

Magnetometer M t t (MAG) (at end of boom [not shown])

Neutron Spectrometer (GRNS/NS)

Figure 1.4. MESSENGER instruments and their locations on the spacecraft.

maintain the encapsulated detector at cryogenic temperatures (90 K) despite the high temperatures in Mercury’s environment. The outer housing of the GRS sensor was equipped with an anticoincidence shield (ACS) to reduce the background from charged particles. The NS sensor consisted of a sandwich of three scintillation detectors working in concert to measure the flux of neutrons in three energy ranges from thermal to ∼7 MeV. X-Ray Spectrometer The X-Ray Spectrometer (XRS) (Figure 1.4) measured the characteristic X-ray emissions induced on the surface of Mercury by the incident solar X-ray flux (Schlemm et al., 2007). The instrument detected the Kα lines for the elements Mg, Al, Si, S, Ca, Ti, Cr, Mn, and Fe. The planet-viewing sensor (Mercury X-ray Unit, MXU) consisted of three gas-filled proportional counters, one with a thin Mg foil over the entrance window, one with a thin Al foil over the entrance window, and one with no foil to separate the lower-energy lines from Mg, Al, and Si. The 12° field of view of the planetviewing sensor allowed a spatial resolution that ranged from 42 km at 200-km altitude to 3200 km at 15,000-km altitude. A small SiPIN detector (Solar Assembly for X-rays, SAX) mounted on the spacecraft sunshade (Leary et al., 2007) and directed sunward provided simultaneous measurement of the solar X-ray flux. The solar detector included a thermoelectric cooler that could also operate in a heater mode to anneal the sensor after radiation damage. Magnetometer MESSENGER’s Magnetometer (MAG) was a low-noise, triaxial fluxgate instrument (Anderson et al., 2007). Its sensor was mounted on a 3.6-m-long boom that was directed generally antisunward (Figures 1.2 and 1.3). The instrument had both a coarse range, ±51,300 nT full scale (1.6-nT resolution), for preflight testing, and a fine range, ±1530 nT full scale (0.047-nT resolution), for operation near Mercury. A magnetic cleanliness program followed during the design and construction of the spacecraft minimized variable and static spacecraft-generated fields at the sensor. Analog signals from the three instrument axes were lowpass filtered (10-Hz cutoff) and sampled simultaneously by three 20-bit analog-to-digital converters every 50 ms. To accommodate variable telemetry rates, MAG provided 11 output rates from 0.01 s−1 to 20 s−1. The instrument also provided continuous measurement of fluctuations by means of a digital 1–10-Hz bandpass filter. This fluctuation level was used to trigger high-time-resolution sampling in 8-min segments to record events of interest when continuous high-rate sampling was not possible. Mercury Laser Altimeter The Mercury Laser Altimeter (MLA) (Cavanaugh et al., 2007) (Figure 1.4) measured the round-trip time of flight of transmitted laser pulses reflected from the surface of Mercury



which, in combination with the spacecraft orbit position and pointing data, gave a high-precision measurement of surface topography referenced to Mercury’s center of mass. The laser transmitter was a diode-pumped Nd:YAG slab laser with passive Q-switching. The transmitter emitted 5-ns-wide pulses at an 8-Hz rate with 20 mJ of energy at a near-infrared wavelength of 1064 nm. The receiver consisted of four refractive telescopes and four equal-length optical fibers to couple the received optical signal onto a single silicon avalanche photodiode. The timing of laser pulses was measured with a set of time-to-digital converters and counters and a crystal oscillator operating at a frequency that was monitored regularly from Earth. MLA sampled the planet’s surface to within a 1-m range error when the line-of-sight range to Mercury was less than 1500 km under spacecraft nadir pointing or the slant range was less than ~1000 km at off-nadir angles up to ~40°. Mercury Atmospheric and Surface Composition Spectrometer MESSENGER’s Mercury Atmospheric and Surface Composition Spectrometer (MASCS) (McClintock and Lankton, 2007) consisted of a small Cassegrain telescope with 257-mm effective focal length and a 50-mm aperture that simultaneously fed an Ultraviolet and Visible Spectrometer (UVVS) and a Visible and Infrared Spectrograph (VIRS) (Figure 1.4). UVVS was a 125-mm-focal-length, scanning grating, Ebert–Fastie monochromator equipped with three photomultiplier tube detectors that covered far-ultraviolet (115–180 nm), middle-ultraviolet (160–320 nm), and visible (250–600 nm) wavelength ranges with an average spectral resolution of 0.6 nm. It was designed to measure profiles with altitude of known exospheric species, to search for previously undetected exospheric species, and to observe Mercury’s surface in the far and middle ultraviolet at a spatial scale of 10 km or smaller. VIRS was a fixed concave grating spectrograph with a 210-mm focal length equipped with a beam splitter that simultaneously dispersed the spectrum onto a 512-element silicon visible-wavelength photodiode array (300–1050 nm) and a 256-element indiumgallium-arsenide infrared-wavelength photodiode array (850–1450 nm). The VIRS was designed to map surface reflectance with 5-nm spectral resolution in the wavelength range 300–1450 nm. Energetic Particle and Plasma Spectrometer The Energetic Particle and Plasma Spectrometer (EPPS) instrument on MESSENGER consisted of two sensors (Andrews et al., 2007), an Energetic Particle Spectrometer (EPS) and a Fast Imaging Plasma Spectrometer (FIPS) (Figure 1.4). The EPS was a hockey-puck-sized energy by time-of-flight spectrometer designed to measure in situ the energy, angular, and compositional distributions of the high-energy components of electrons (>20 keV) and ions (>5 keV/nucleon) near Mercury. The FIPS measured the energy, angular, and compositional distributions of the lowenergy components of the ion distributions (95% variance reduction) yield a Te of about 150 km, with a range of 110–180 km. T15 posited that the present distribution of insolation-driven subsurface temperature anomalies is temporally part of a quasisteady-state condition that was achieved following the last time that Mercury acquired a 3:2 spin–orbit resonance (SOR). Essentially, the assumption is that the 3:2 SOR thermal deformation, once established, does not adjust to a thickening lithosphere; it is “frozen in” (see Section, though T15 did not specifically use this terminology. So when in Mercury’s history was the elastic lithosphere about 150 km thick? T15 addressed this question by making the assumption that the long-

wavelength curvature in lithospheric deformation is small, which implies that the thicknesses of the elastic and mechanical lithospheres are nearly equal. As discussed in Section, Tm is defined by a threshold creep stress (15 MPa as given by T15) that is highly temperature dependent and can be predicted from the output of a thermal model. The result (Figure 3.18) is that the 150-km thickness is achieved about 1.5 Gyr into the evolution of Mercury, and the last 3:2 SOR reacquisition would have occurred approximately 0.5 Gyr earlier, or at about 3.5 Ga, to allow for the diffusion of the temperature perturbations into the interior.

3.7.7 Tosi et al. (2015) Model: Why Does It Work and What Are the Issues? Why Does the Model Work? There are three main contributions to the elastic shell deformation in the T15 model: (i) buoyancy forces associated with thermal density anomalies in the lithosphere generated by Tθφ; (ii) thermal expansion and contraction associated with those anomalies; and (iii) radial displacement in the sub-lithosphere region, largely from thermal strain but with a minor contribution from CMB shape. The last contribution is likely the smallest of the three, partly because the temperature differences between the thermal poles decrease approximately linearly with depth and vanish on the CMB. In general, the geoid and shape signals are dominated by lithospheric processes, as indicated, for example, by the similarity in model predictions in the convective and conductive regimes. The shape can be taken as a top load on an elastic lithosphere (Section 3.7.4), and we examine the resulting state of compensation with the T15 model parameters. The departure from mass balance is embodied in the dimensionless “degree of

3.7 Degree-2 Shape and Geoid

cold pole

hot pole

hot pole

cold pole

Figure 3.19. Schematic view of the elastic lithosphere component of the T15 model in Mercury’s equatorial plane. Hot (cold) poles have surface temperatures higher (lower) than the planetary mean. The temperature anomaly lessens with depth, and the integrated buoyancy and radial strain lead to an elevation increase (decrease) in hot (cold) pole regions. Horizontal thermal strains are accommodated by upwarping (downwarping) in hot (cold) pole regions, and the thermoelastic stresses support at least half of the high-admittance shape (magenta dashed curve), which is an apparent top load.

compensation” (DC) introduced by Turcotte et al. (1981). Given the T15 best-fitting range of Te, 110–180 km, and their Young’s modulus for the mantle of 175 GPa, the degree of compensation for top loading by the shape is quite low, about 0.15 (Figure 3.15b). This result implies, inter alia, minor mass cancellation of the shape due to deformation of the CrMB in response to the load; this outcome is consistent with the relative insensitivity of T15 model fits to variation in Hcr. Although the DC results in Figure 3.15b are specifically for top loading, they do suggest, at least qualitatively, that internal buoyant loading within a lithosphere 150 km thick will have a muted response at the surface because of the membrane resistance of the elastic shell. The radially integrated thermal buoyancy across the entire silicate shell for a conduction solution is about 50% (with opposite sign) of the buoyancies stored in the shapes, ρcrustg0c20 and ρcrustg0c22. Membrane resistance means that the 50% fraction has to be taken as a generous upper bound of the buoyancy contribution to shape. With Te = 150 km, radial thermal expansion beneath and mostly within the lithosphere can account for on the order of 10% of both the c20 and c22 terms, with the major contribution provided by the lithosphere. Buoyancy contributions likely contribute at about the same level. The sum of thermal buoyancy, compositional buoyancy (CrMB), and strain probably accounts for no more than half of the shape. The remainder of the shape can be attributed to confined horizontal (θ, φ) thermoelastic strain. The T15 model produces a scenario that is geophysically comparable to top loading, in the sense that long-wavelength shape is supported partially by thermoelastic membrane and bending stresses in the lithospheric shell (Figure 3.19). The T15 mechanism does not


require the emplacement of material at Mercury’s surface or near subsurface. Rather, this process creates its own top load and is akin to warping, where long-wavelength shape is produced through uneven thermal expansion in a spherical shell and supported largely by thermoelastic stresses, as noted. What about the geoid? We noted earlier that the observed d2 geoid is about a third of that due to shape alone. To first order this ratio can be achieved largely with the density anomalies associated with the thermal buoyancy providing partial compensation of the shape, as the two masses have opposite signs. To a lesser extent, perturbation of the CrMB according to the 0.15 DC value also contributes to the compensation. The model presented by T15 represents the only hypothesis advanced to date to explain the d2 correlation of shape and surface temperature. The model also provides the necessary partially compensated d2 surface load that reproduces both the shape and the geoid. So it looks as though solar energy in concert with the 3:2 SOR is responsible for Mercury’s marked departure from a d2 hydrostatic state, and this mechanism differs thoroughly from that responsible for the Moon’s departure from a hydrostatic state. However, there are issues with the T15 model that are discussed in the next four subsections. Late 3:2 SOR Acquisition? Mercury should have entered a 3:2 SOR early in its history (Correia and Laskar, 2004, 2009; Makarov, 2012), but it is also possible that at later times impacts of sufficient energy temporarily disrupted the 3:2 SOR (Correia and Laskar, 2012; Noyelles et al., 2014), and it is the last of these events that set the thermal environment in the T15 model. However, a late acquisition of a 3:2 SOR is in conflict with geological observations. Tectonic mapping indicates that tidal spindown could have operated at the same time as, but did not postdate, global contraction from secular cooling (Klimczak et al., 2015); superposition relations show that tectonic deformation arising from global contraction was underway by about 3.6 Ga (see Section 10.8.1). Moreover, global contraction was likely operating for some time before then, because the tectonic manifestation of this process (i.e., faulting) was preceded by elastic deformation (Section 10.8.1). Thus, given that the strain pattern predicted by tidal spindown does not dominate Mercury’s tectonic fabric, the acquisition of Mercury’s current SOR must have occurred considerably before 3.6 Ga, a conclusion that is at odds with the T15 model. Variations in Eccentricity The Tosi et al. (2015) model neglects the consequences of the strong variations in Mercury’s orbital eccentricity (Figure 3.17b). However, the quality of the model fit was judged by the reduction in the joint error variance of shape and geoid for all admissible d2 and d4 terms; this comparison is dominated by the two zonal (m = 0) terms, which are insensitive to eccentricity variations (Figure 3.17a). Nevertheless, the equatorial d2 shape should be constantly changing, following orbital eccentricity changes with a diffusion time lag.


Mercury’s Crust and Lithosphere: Structure and Mechanics Relationship of Te to Tm The basic result of T15 depends on the assumption that Te = Tm, i.e., there has been very little deformation at the top (via frictional sliding) and bottom (by viscous creep) of the mechanical lithosphere. Given that the model deformation is largely occurring at wavelengths corresponding to d2 and to a lesser extent d4, curvature should be small and setting Te = Tm is seemingly justified. At these same wavelengths, bending stresses are much smaller than membrane stresses. Whereas the bending stress profile passes from extensional to compressive with depth (or vice versa), the membrane stress profile is uniform with depth, leading to a different style of plate deformation. Furthermore, there is abundant evidence that long-wavelength strain has been localized (see Chapter 10), so curvature could have been sufficiently large locally to negate the idea that Te = Tm. Additional processes such as impact cratering and volcanism can relieve lithospheric stresses over geological time. The net result is that the relationship between Tm and Te is somewhat ambiguous. The estimation of subsurface mechanical parameters from surface profiles over faults (Watters et al., 2002; Nimmo and Watters, 2004) had been limited with Mariner 10 data to a single structure, Discovery Rupes (see Section With highly accurate profiles from MESSENGER MLA data, Peterson et al. (2017) extended this methodology to 31 fault structures. The elastic dislocation model that was employed to estimate fault parameters (dip, minimum and maximum fault depth) produced solutions at each of these 31 faults that were a good match to the MLA profiles. The maximum fault depth was equated to the BDT depth, and the temperature there, ΨBDT, was determined from the employed yield strength envelope. ΨBDT was linearly extrapolated to a depth at which the temperature generated the T15 threshold creep stress of 15 MPa to provide an estimate of Tm. The three regions studied (northern smooth plains, intercrater plains, and circum-Caloris plains) all have Tm mean values at the assumed ~4 Ga age that are distinctly smaller than the T15 thermal model result (Peterson et al., 2017). Differences in Tm estimates among the three regions may reflect differences in their volcanic histories. We do not compare the Te estimates because they are based on diametrically opposed assumptions. The T15 approach was based on the assumption that stresses in the lithosphere do not reach the yield stress (i.e., the strength envelope) anywhere, and the Peterson et al. (2017) approach is predicated implicitly on the premise that the stresses are saturated at the yield stress everywhere in the lithosphere because the maximum depth of faulting is equated to the BDT depth. The mechanical thickness, Tm, is a stand-in for heat flux from Mercury’s interior, so differences in the Tm results between T15 and Peterson et al. (2017) would at face value (and lateral heterogeneities aside) indicate distinctly different planets; but of course there is only one Mercury. Age of Thermal Structure Is the density configuration of the T15 model really frozen at ~3 Ga? Certainly, estimates of Te as low as 30 km at ~4 Ga (e.g., James et al., 2016) imply, unsurprisingly, that stresses can be

locked-in on Mercury. In the Tosi et al. (2015) model, thermoelastic stresses are largely responsible for elastic lithosphere deformation and the resulting shape. The corresponding geoid signal is generated mostly by the shape, with incomplete cancellation dominated by the density anomalies associated with the thermal buoyancy distribution. As the elastic lithosphere cools and thickens, the thermoelastic strains change. Density boundaries, such as the CrMB, are “frozen” and do not adjust to accommodate the changing mechanical state because viscosities are too high. In contrast to the flexural problem (Section, thermoelastic processes do not require substantial viscous flow of the mantle to track the equilibrium stress state of the thickening lithosphere. That is, this state can be achieved largely by elastic strain, and by contrast the flexural problem also requires (implicitly) viscous strain, which is unavailable. Viscous strain is never seen explicitly in the classical flexural equilibrium solution because such strain is part of a transient phase. We argue that the T15 estimate of Te and geoid is in fact germane to the current era on Mercury, but this issue can be resolved by engaging a viscoelastic model in place of the elastic model. The lack of a rheological memory would remove the requirement of a late final acquisition of the 3:2 SOR with a seemingly over-thick lithosphere. Note also that the notion that the equatorial shape ellipticity (c22) tracks changes in orbital eccentricity (Section is also predicated on an absence of rheological memory. Assigning Te to the present era is possibly more consistent with estimates of that quantity that are based on data, e.g., ~30 km at ~4 Ga (Figure 3.18). Extrapolating the 30-km value by following the same rate of elastic lithosphere thickening as that predicted by the T15 thermal model leads to a current value of about 125 km for Te, which falls within the estimated range of T15. The thickening rate for Te should be more reliable than specific numbers. Remaining unsolved is the discrepancy between estimates of Te from gravity and shape data and estimates of Te from thermal models, but reconsidering the assumption that Te = Tm would be a good place to start to resolve this difference. 3.7.8 An Alternative Model We have already discussed another model (in Section 3.7.4 and shown in Figure 3.15a) that might depend on the surface temperature distribution. This is a top-loading model with an elastic lid and partial compensation at the CrMB. This model satisfies (i) the crustal thickness result of Padovan et al. (2015), (ii) the elastic thickness estimate of James et al. (2016), and (iii) the Z20 gravity/shape admittance. The challenge to this model is to identify the load implied by the c20 and c22 shape terms. One possibility is early crustal thickening that was sensitive to the insolation-driven thermal anomalies. Convection models (e.g., Tosi et al., 2015) show that temperature differences between the pole and the equator persist deep into the mantle and in a convection regime will generate more melt at the equator than at a spin pole (Figure 3.20). Under this scenario, melt would be brought upward into the crust by buoyancy of the melt itself, providing early crust that thickens toward the equator. Pressurerelease partial melting of upward-advecting solid mantle material would add to the crustal volume (see Section 3.4.1). As

3.8 Final Thoughts 4. If so, what is its mean thickness?

0 T > Tsolidus 100 Depth (km)



Estimates of mean crustal thickness are converging to a value less than 40–50 km (Section 3.4.4). More generally, there is good evidence for a strong density boundary within 50 km of the surface, and the most straightforward interpretation is that it is the crust–mantle boundary. 5. Is Mercury’s shape isostatically compensated?


400 300






Temperature (K)

Figure 3.20. Solutions after 1 Gyr for the three-dimensional thermal evolution model of Tosi et al. (2015). Profiles are for the minimum (blue), average (black), and maximum (red) mantle temperatures; the differences are due largely to the insolation-driven surface temperature variations. The green region has temperatures above the solidus temperature, which is from Hirschmann (2000) and represents a quadratic fit to data from 63 sets of peridotite melting experiments.

noted earlier, partial melts would tend to remain buoyant as they traversed the mantle (Vander Kaaden and McCubbin, 2015), maximizing the mass difference between the equatorial and polar regions. The difficulty with this model may be that partial melting differences between the temperature poles are not large, and it may take an inordinate amount of time to build up the shape variations.

3.8 F INAL THOUGHTS Following the end of the MESSENGER mission, we can evaluate our understanding of Mercury’s crust and lithosphere from a geophysical perspective. We do so by asking a series of questions and supplying what we hope are objective answers. 1. How well do we know the shape of Mercury? The answer depends on which is more important to a specific end user (see Section 3.2), accuracy or resolution. Stereoderived digital elevation models have 250-m/pixel grids but suffer from long-wavelength biases. MLA grids in the northern hemisphere have high accuracy but lower resolution than the stereo. Spherical harmonic representations of shape beyond about degree 4 generally need to be localized to the northern hemisphere to be useful for modeling, though the global power spectral density can be used to about degree 8. 2. How well do we know the gravity field of Mercury? The degree strength map (Figure 3.4) answers this question in full. 3. Does Mercury have a crust? Yes, there is geochemical evidence for both primary and secondary crustal material (Section 3.4.1). Geophysical evidence is provided in the next answer.

For the most part, the answer is no (Section 3.5.3). Crustal thickness variations derived from Bouguer δg signals are generally not isostatic. This result may be due, inter alia, to the contribution of dynamic pressure from mantle flow, or to the presence of a substantial elastic lithosphere (Te ≥ 30 km) early in Mercury’s history. 6. Do we know the thickness of the elastic lithosphere and how it evolved with time? The answer is no. Elastic lithosphere evolution can be linked to thermal model evolution if one is willing to equate Te with Tm, the mechanical lithosphere thickness. Thermal models tend to produce thick lithospheres (e.g., >100 km at 4 Ga). Spectral analyses (admittance, correlation) yield Te values equal to about one-third of the values from thermal modeling (Section The assumption that Te = Tm is very likely too simplified and, furthermore, studies that use gravity and shape data to estimate Te have been extremely limited to date. 7. What is the cause of the non-hydrostatic state of Mercury’s second-degree shape and geoid (Section 3.7)? A successful model appears to require a top load with substantial support by elastic membrane and flexural stresses, and variations in the load must be closely related to Mercury’s principal axes. The insolation pattern resulting from Mercury’s 3:2 spin–orbit resonance and near-zero obliquity, as a source of buoyancy anomalies and thermoelastic stresses in the lithosphere (Tosi et al., 2015), is a leading candidate. The top load in this case is a warping of the surface in response to spatially varying thermoelastic stresses. Among the challenges to this model is the timing of the acquisition of the subsurface temperature distribution that would create the necessary conditions in buoyancy and stress. A second (and untested) mechanism is that the same insolation pattern causes variations in partial melting in the mantle that migrate upward to create crustal loads in phase with surface temperature (and thus with the principal axes). 8. What work is needed? There are several routes for conducting additional work with MESSENGER data that would advance our understanding of the issues discussed in this chapter: a. Develop alternative methodologies for relating elastic thickness to thermal evolution models. b. Conduct hydrocode modeling of basin formation on Mercury in order to characterize lithospheric thickness as a function of basin age. c. Take advantage of the highest-resolution MESSENGER gravity data, acquired during the last year of the mission, to address problems at shorter length scales than before.


Mercury’s Crust and Lithosphere: Structure and Mechanics

d. Investigate the origin and evolution of fold-and-thrust belts (Section 3.6.4) with an approach that integrates structural geology and geophysical modeling. e. Look for tectonic evidence of changes in the equatorial shape in response to variations in orbital eccentricity. f. Develop new models for the origin and support of the northern rise.

ACKNOWLEDG E MENTS We thank Walter Kiefer, Nicola Tosi, and Sean Solomon for extremely constructive reviews. We also greatly appreciate the leadership of David Smith and Maria Zuber in the delivery, analysis, and interpretation of MESSENGER altimetry and gravity field measurements.

R EF E R EN CE S Albert, R. A. and Phillips, R. J. (2000). Paleoflexure. Geophys. Res. Lett., 27, 2385–2388. Anderson, F. S. and Smrekar, S. E. (2006). Global mapping of crustal and lithospheric thickness on Venus. J. Geophys. Res., 111, E08006, doi:10.1029/2004JE002395. Anderson, J. D., Colombo, G., Esposito, P. B., Lau, E. L. and Trager, G. B. (1987). The mass, gravity field and ephemeris of Mercury. Icarus, 71, 337–349. Anderson, J. D., Jurgens, R. F., Lau, E. L., Slade M. A., III and Schubert, G. (1996). Shape and orientation of Mercury from radar ranging data. Icarus, 124, 690–697. Asmar, S. W. and Armstrong, J. W. (2005). Spacecraft Doppler tracking: Noise budget and accuracy achievable in precision radio science observations. Radio Sci., 40, RS2001, doi:10.1029/ 2004RS003101. Balcerski, J. A., Hauck, S. A., II, Sun, P., Klimczak, C., Byrne, P. K., Phillips, R. J. and Solomon, S. C. (2013). New constraints on timing and mechanisms of regional tectonism from Mercury’s tilted craters. Lunar Planet. Sci., 44, abstract 2444. Banerdt, W. B. (1986). Support of long-wavelength loads on Venus and implications for internal structure. J. Geophys. Res., 91, 403–419. Becker, K. J., Weller, L. A., Edmundson, K. L., Becker, T. L., Robinson, M. S., Enns, A. C. and Solomon, S. C. (2012). Global controlled mosaic of Mercury from MESSENGER orbital images. Lunar Planet. Sci., 43, abstract 2654. Becker, K. J., Robinson, M. S., Becker, T. L., Weller, L. A., Edmundson, K. L., Neumann, G. A., Perry, M. E. and Solomon, S. C. (2016). First global digital elevation model of Mercury. Lunar Planet. Sci., 47, abstract 2959. Bindschadler, D. L., Schubert, G. and Ford, P. G. (1994). Venus’ center of figure–center of mass offset, Icarus, 111, 417–432. Brace, W. F. and Kohlstedt, D. L. (1980). Limits on lithospheric stress imposed by laboratory measurements. J. Geophys. Res., 85, 6248–6252. Brown, S. M. and Elkins-Tanton, L. T. (2009). Compositions of Mercury’s earliest crust from magma ocean models. Earth Planet. Sci. Lett., 286, 446–455, doi:10.1016/j.epsl.2009.07.010. Burke, K. C., Şengör, A. M. C. and Francis, P. W. (1984). Maxwell Montes in Ishtar: A collisional plateau on Venus? Lunar Planet. Sci., 15, 104–105. Byerlee, J. D. (1978). Friction of rocks. Pure Appl. Geophys., 116, 615–626.

Byrne, P. K., Klimczak, C., Williams, D. A., Hurwitz, D. M., Solomon, S. C., Head, J. W., Preusker, F. and Oberst, J. (2013). An assemblage of lava flow features on Mercury. J. Geophys. Res. Planets, 118, 1303–1322. Byrne, P. K., Klimczak, C., Şengör, A. M. C., Solomon, S. C., Watters, T. R. and Hauck II, S. A. (2014). Mercury’s global contraction much greater than earlier estimates. Nature Geosci., 7, 301–307. Byrne, P. K., Klimczak, C., McGovern, P. J., Mazarico, E., James, P. B., Neumann, G. A., Zuber, M. T. and Solomon, S. C. (2015). Deep-seated thrust faults bound the Mare Crisium lunar mascon. Earth Planet. Sci. Lett., 427, 183–190. Cavanaugh, J. F., Smith, J. C., Sun, X., Bartels, A. E., RamosIzquierdo, L., Krebs, D. J., McGarry, J. F., Trunzo, R., NovoGradac, A.-M., Britt, J. L., Karsh, J. L., Katz, R. B., Lukemire, A. T., Symkiewicz, R., Berry, D. L., Swinski, J. P., Neumann, G. A., Zuber, M. T. and Smith, D. E. (2007). The Mercury Laser Altimeter instrument for the MESSENGER mission. Space Sci. Rev., 131, 451–480. Chandrasekhar, S. (1987). Ellipsoidal Figures of Equilibrium. New York: Dover, 264 pp. Chao, B. F. (2006). Earth’s oblateness and its temporal variations. Comptes Rendus Geoscience, 338, 1123–1129. Charlier, B., Grove, T. L. and Zuber, M. T. (2013). Phase equilibria of ultramafic compositions on Mercury and the origin of the compositional dichotomy. Earth Planet. Sci. Lett., 363, 50–60. Correia, A. C. M. and Laskar, J. (2004). Mercury’s capture into the 3/2 spin-orbit resonance as a result of its chaotic dynamics. Nature, 429, 848–850. Correia, A. C. M. and Laskar, J. (2009). Mercury’s capture into the 3/2 spin–orbit resonance including the effect of core–mantle friction. Icarus, 201, 1–11. Correia, A. C. M. and Laskar, J. (2012). Impact cratering on Mercury: Consequences for the spin evolution. Astrophys. J., 751, L43, 5 pp. Courtney, R. C. and Beaumont, C. (1983). Thermally-activated creep and flexure of the oceanic lithosphere. Nature, 305, 201–204, doi:10.1038/305201a0. Denevi, B. W., Robinson, M. S., Solomon, S. C., Murchie, S. L., Blewett, D. T., Domingue, D. L., McCoy, T. J., Ernst, C. M., Head, J. W., Watters, T. R. and Chabot, N. L. (2009). The evolution of Mercury’s crust: A global perspective from MESSENGER. Science, 324, 613–618, doi:10.1126/science.1172226. Denevi, B. W., Ernst, C. M., Meyer, H. M., Robinson, M. S., Murchie, S. L., Whitten, J. L., Head, J. W., Watters, T. R., Solomon, S. C., Ostrach, L. R., Chapman, C. R., Byrne, P. K., Klimczak, C. and Peplowski, P. N. (2013). The distribution and origin of smooth plains on Mercury. J. Geophys. Res. Planets, 118, 891–907, doi:10.1002/jgre.20075. Dombard, A. J., Hauck, S. A., II, Solomon, S. C. and Phillips, R. J. (2001). Potential for long-wavelength folding on Mercury. Lunar Planet. Sci., 32, abstract 2035. Elgner, S., Stark, A., Oberst, J., Perry, M. E., Zuber, M. T., Robinson, M. S. and Solomon, S. C. (2014). Mercury‫׳‬s global shape and topography from MESSENGER limb images. Planet. Space Sci., 103, 299–308. Elkins-Tanton, L. T. (2012). Magma oceans in the inner solar system. Annu. Rev. Earth Planet. Sci., 40, 113–139. Ernst, C. M., Murchie, S. L., Barnouin, O. S., Robinson, M. S., Denevi, B. W., Blewett, D. T., Head, J. W., Izenberg, N. R., Solomon, S. C. and Roberts, J. H. (2010). Exposure of spectrally distinct material by impact craters on Mercury: Implications for global stratigraphy. Icarus, 209, 210–223, doi:10.1016/j.icarus.2010.05.022. Evans, L. G., Peplowski, P. N., Rhodes, E. A., Lawrence, D. J., McCoy, T. J., Nittler, L. R., Solomon, S. C., Sprague, A. L., StockstillCahill, K. R., Starr, R. D., Weider, S. Z., Boynton, W. V., Hamara,

References D. K. and Goldsten J. O. (2012). Major-element abundances on the surface of Mercury: Results from the MESSENGER GammaRay Spectrometer. J. Geophys. Res., 117, E00L07, doi:10.1029/ 2012JE004178. Fjeldbo, G.A., Kliore, A., Sweetnam, D., Esposito, P., Seidel, B. and Howard, T. (1976). The occultation of Mariner 10 by Mercury. Icarus, 27, 439–444. Folkner, W. M., Yoder, C. F., Yuan, D. N., Standish, E. M. and Preston, R. A. (1997). Interior structure and seasonal mass redistribution of Mars from radio tracking of Mars Pathfinder. Science, 278, 1749–1752. Garrick-Bethell, I., Wisdom, J. and Zuber, M. T. (2006). Evidence for a past high eccentricity lunar orbit. Science, 313, 652–655. Garrick-Bethell, I., Perera, V., Nimmo, F. and Zuber, M. T. (2014). The tidal-rotational shape of the Moon and evidence for polar wander. Nature, 512, 181–184. Gaskell, R. W., Barnouin-Jha, O. S., Scheeres, D. J., Konopliv, A. S., Mukia, T., Abe, S., Saito, J., Ishiguro, M., Kubota, T., Hashimoto, T., Kawaguhi, J., Yoshikawa, M. S., Kominato, T., Hirata, N. and Demura, H. (2008). Characterizing and navigating small bodies with imaging data. Meteorit. Planet. Sci., 43, 1049–1061. Grott, M. and Breuer, D. (2008). The evolution of the Martian elastic lithosphere and implications for crustal and mantle rheology. Icarus, 193, 503–515. Guest, J. E. (1971). Centres of igneous activity in the maria. In Geology and Physics of the Moon, ed. G. Fielder. New York: Elsevier, pp. 41–53. Hager, B. H. and Clayton, R. W. (1989). Constraints on the structure of mantle convection using seismic observations, flow models, and the geoid. In Mantle Convection: Plate Tectonics and Global Dynamics, ed. R. W. Peltier. New York: Gordon and Breach Science, pp. 657–763. Harmon, J. K., Campbell, D. B., Bindschadler, D. L., Head, J. W. and Shapiro, I. I. (1986). Radar altimetry of Mercury: A preliminary analysis. J. Geophys. Res., 91, 385–401. Hauck, S. A., II, Dombard, A. J., Phillips, R. J. and Solomon, S. C. (2004). Internal and tectonic evolution of Mercury. Earth Planet. Sci. Lett., 222, 713–728. Hauck, S. A., II, Margot, J.-L., Solomon, S. C., Phillips, R. J., Johnson, C. L., Lemoine, F. G., Mazarico, E., McCoy, T. J., Padovan, S., Peale, S. J., Perry, M. E., Smith, D. E. and Zuber, M. T. (2013). The curious case of Mercury’s internal structure. J. Geophys. Res. Planets, 118, 1204–1220. Hauck, S. A., II, Byrne, P. K., Denevi, B. W., Grott, M., McCoy, T. and Stanley, S. (2015). Mercury’s global evolution: New views from MESSENGER. Presented at 2015 Fall Meeting, American Geophysical Union, abstract P53A-2105, San Francisco, CA, 14–18 December. Head, J. W., Chapman, C. R., Strom, R. G., Fassett, C. I., Denevi, B. W., Blewett, D. T., Ernst, C. M., Watters, T. R., Solomon, S. C., Murchie, S. L., Prockter, L. M., Chabot, N. L., Gillis-Davis, J. J., Whitten, J. L., Goudge, T. A., Baker, D. M. H., Hurwitz, D. M., Ostrach, L. R., Xiao, Z., Merline, W. J., Kerber, L., Dickson, J. L., Oberst, J., Byrne, P. K., Klimczak, C. and Nittler, L. R. (2011). Flood volcanism in the northern high latitudes of Mercury revealed by MESSENGER. Science, 333, 1853–1856. Herrick, R. R. and Phillips, R. J. (1992). Geological correlations with the interior density structure of Venus. J. Geophys. Res., 97, 16,017–16,034. Hirschmann, M. M. (2000). Mantle solidus: Experimental constraints and the effects of peridotite composition. Geochem. Geophys. Geosyst., 1, 1042, doi:10.1029/2000GC000070. Iess, L., Asmar, S. and Tortora, P. (2009). MORE: An advanced tracking experiment for the exploration of Mercury with the mission BepiColombo. Acta Astronautica, 65, 666–675.


James, P. B., Byrne, P. K., Solomon, S. C., Zuber, M. T. and Phillips, R. J. (2014). Surface strains associated with the evolution of Mercury’s domical swells. Presented at 2014 Fall Meeting, American Geophysical Union, abstract P21C-393, San Francisco, CA, 15–19 December. James, P. B., Mazarico, E., Genova, A., Smith, D. E., Neumann, G. A. and Solomon, S. C. (2015a). Mercury’s lithospheric thickness and crustal density, as inferred from MESSENGER observations. Presented at 2015 Fall Meeting, American Geophysical Union, abstract P53A-2102, San Francisco, CA, 14–18 December. James, P. B., Zuber, M. T., Phillips, R. J. and Solomon, S. C. (2015b). Support of long-wavelength topography on Mercury inferred from MESSENGER measurements of gravity and topography. J. Geophys. Res. Planets, 120, 287–310. James, P. B., Phillips, R. J., Grott, M., Hauck, S. A., II and Solomon, S. C. (2016). The thickness of Mercury’s lithosphere inferred from MESSENGER gravity and topography. Lunar Planet. Sci., 47, abstract 1992. Jeffreys, H. (1970). The Earth, Its Origin, History and Physical Constitution, 5th edn. Cambridge: Cambridge University Press. Johnson, B. C., Blair, D. M., Collins, G. S., Melosh, H. J., Freed, A. M., Taylor, G. J., Head, J. W., Wieczorek, M. A., Andrews-Hanna, J. C., Nimmo, F., Keane, J. T., Miljković, K., Soderblom, J. M. and Zuber, M. T. (2016). Formation of the Orientale lunar multiring basin. Science, 354, 441–444. Kaula, W. M. (1966). Theory of Satellite Geodesy. Waltham, MA: Blaisdell. Kaula, W. M. (1979). The moment of inertia of Mars. Geophys. Res. Lett., 6, 194–196. Keane, J. T. and Matsuyama, I. (2014). Evidence for lunar true polar wander and a past low-eccentricity, synchronous lunar orbit. Geophys. Res. Lett., 41, 6610–6619. Kegege, O., Fuentes, M., Meyer N. and Sil, A. (2012). Three-dimensional analysis of Deep Space Network antenna coverage. 2012 IEEE Aerospace Conference, 9 pp., Big Sky, Montana, 4–10 March, doi:10.1109/AERO.2012.6187124. Khan, A., Mosegaard, K. and Rasmussen, K. L. (2000). A new seismic velocity model for the Moon from a Monte Carlo inversion of the Apollo lunar seismic data. Geophys. Res. Lett., 27, 1591–1594. King, S. D. (2008). Pattern of lobate scarps on Mercury’s surface reproduced by a model of mantle convection. Nature Geosci., 1, 229–232. Klimczak, C., Watters, T. R., Ernst, C. M., Freed, A. M., Byrne, P. K., Solomon, S. C., Blair, D. M. and Head, J. W. (2012). Deformation associated with ghost craters and basins in volcanic smooth plains on Mercury: Strain analysis and implications for plains evolution. J. Geophys. Res., 117, E00L03, doi:10.1029/2012JE004100. Klimczak, C., Ernst, C. M., Byrne, P. K., Solomon, S. C., Watters, T. R., Murchie, S. L., Preusker, F. and Balcerski, J. A. (2013). Insights into the subsurface structure of the Caloris basin, Mercury, from assessments of mechanical layering and changes in long-wavelength topography. J. Geophys. Res. Planets, 118, 2030–2044. Klimczak, C., Byrne, P. K. and Solomon, S. C. (2015). A rock-mechanical assessment of Mercury’s global tectonic fabric. Earth Planet. Sci. Lett., 416, 82–90, doi:10.1016/ j.epsl.2015.02.003. Kohlstedt, D. L. and Mackwell, S. J. (2010). Strength and deformation of planetary lithospheres. In Planetary Tectonics, ed. T. R. Watters and R. A. Schultz. New York: Cambridge University Press, pp. 397–456. Kohlstedt, D. L., Evans, B. and Mackwell, S. J. (1995). Strength of the lithosphere: Constraints imposed by laboratory experiments. J. Geophys. Res., 100, 17,587–17,602. Konopliv, A. S., Banerdt, W. B. and Sjogren, W. L. (1999). Venus gravity: 180th degree and order model. Icarus, 139, 3–18.


Mercury’s Crust and Lithosphere: Structure and Mechanics

Lago, B. and Cazenave, A. (1981). State of stress in the oceanic lithosphere in response to loading. Geophys. J. Roy. Astron. Soc., 64, 785–799. Laplace, P.-S. (1878). Oeuvres Complètes de Laplace. Paris: GauthiersVillars. Laskar, J. (1988). Secular evolution of the Solar System over 10 million years. Astron. Astrophys., 198, 341–362. Lemoine, F. G. R., Smith, D. E., Zuber, M. T., Neumann, G. A. and Rowlands, D. D. (1997). A 70th degree lunar gravity model (GLGM-2) from Clementine and other tracking data. J. Geophys. Res., 102, 16,339–16,359. Lemoine, F. G., Smith, D. E., Rowlands, D. D., Zuber, M. T., Neumann, G. A., Chinn, D. S. and Pavlis, D. E. (2001). An improved solution of the gravity field of Mars (GMM-2B) from Mars Global Surveyor. J. Geophys. Res., 106, 23,359–23,376. Lemoine, F. G., Goossens, S., Sabaka, T. J., Nicholas, J. B., Mazarico, E., Rowlands, D. D., Loomis, B. D., Chinn, D. S., Caprette, D. S., Neumann, G. A., Smith, D. E. and Zuber, M. T. (2013). Highdegree gravity models from GRAIL primary mission data. J. Geophys. Res. Planets, 118, 1676–1698. Makarov, V. V. (2012). Conditions of passage and entrapment of terrestrial planets in spin-orbit resonances. Astrophys. J., 752, article 73, doi:10.1088/0004-637X/752/1/73. Margot, J.-L., Peale, S. J., Solomon, S. C., Hauck, S. A., II, Ghigo, F. D., Jurgens, R. F., Yseboodt, M., Giorgini, J. D., Padovan, S. and Campbell, D. B. (2012). Mercury’s moment of inertia from spin and gravity data. J. Geophys. Res., 117, E00L09, doi:10.1029/ 2012JE004161. Matsuyama, I. and Nimmo, F. (2009). Gravity and tectonic patterns of Mercury: Effect of tidal deformation, spin-orbit resonance, nonzero eccentricity, despinning, and reorientation. J. Geophys. Res., 114, E01010, doi:10.1029/2008JE003252. Mazarico, E., Zuber, M. T., Lemoine, F. G. and Smith, D. E. (2008). Observation of atmospheric tides in the Martian exosphere using Mars Reconnaissance Orbiter radio tracking data. Geophys. Res. Lett., 35, L09202, doi:10.1029/2008GL033388. Mazarico, E., Genova, A., Goossens, S., Lemoine, F. G., Neumann, G. A., Zuber, M. T., Smith, D. E. and Solomon, S. C. (2014). The gravity field, orientation, and ephemeris of Mercury from MESSENGER observations after three years in orbit. J. Geophys. Res. Planets, 119, 2417–2436. Mazarico, E., Genova, A., Goossens, S., Lemoine, F. G., Smith, D. E., Zuber, M. T., Neumann, G. A. and Solomon, S. C. (2016). The gravity field of Mercury after MESSENGER. Lunar Planet. Sci., 47, abstract 2022. McAdams, J. V., Bryan, C. G., Bushman, S. S., Calloway, A. B., Carranza, E., Flanigan, S. H., Kirk, M. N., Korth, H., Moessner, D. P., O’Shaughnessy, D. J. and Williams, K. E. (2015). Engineering MESSENGER’s grand finale at Mercury: The lowaltitude hover campaign. Astrodynamics Specialist Conference, American Astronautical Society, paper AAS 15–634, 20 pp., Vail, CO, 9–13 August. McAdoo, D. C. and Sandwell, D. T. (1985). Folding of oceanic lithosphere. J. Geophys. Res., 90, 8563–8569. McCauley, J. F. (1967). Geologic Map of the Hevelius Region of the Moon, Map I-1491, Miscellaneous Investigations Series. Denver, CO: U.S. Geological Survey. McGovern, P. J., Solomon, S. C., Smith, D. E., Zuber, M. T., Simons, M., Wieczorek, M. A., Phillips, R. J., Neumann, G. A., Aharonson, O. and Head, J. W. (2002). Localized gravity/topography admittance and correlation spectra on Mars: Implications for regional and global evolution. J. Geophys. Res., 107, 5136, doi:10.1029/2002JE001854. McGovern, P. J., Solomon, S. C., Smith, D. E., Zuber, M. T., Simons, M., Wieczorek, M. A., Phillips, R. J., Neumann, G. A.,

Aharonson, O. and Head, J. W. (2004). Correction to “Localized gravity/topography admittance and correlation spectra on Mars: Implications for regional and global evolution”. J. Geophys. Res., 109, E07007, doi:10.1029/2004JE002286. McKenzie, D. (2003). Estimating Te in the presence of internal loads. J. Geophys. Res., 108 (B9), 2438, doi:10.1029/2002JB001766. McKenzie, D. and Bowin, C. (1976). The relationship between bathymetry and gravity in the Atlantic Ocean. J. Geophys. Res., 81, 1903–1915. McNutt, M. K. (1984). Lithospheric flexure and thermal anomalies. J. Geophys. Res., 89, 11,180–11,194. Melosh, H. J. (2011). Planetary Surface Processes, Cambridge Planetary Science Series, Cambridge: Cambridge University Press. Melosh, H. J. and McKinnon, W. B. (1988). The tectonics of Mercury. In Mercury, ed. F. Vilas, C. R. Chapman and M. S. Matthews. Tucson, AZ: University of Arizona Press, pp. 374–400. Meyer, J., Elkins-Tanton, L. and Wisdom, J. (2010). Coupled thermal– orbital evolution of the early Moon. Icarus, 208, 1–10. Michel, N. C., Hauck S. A., II, Solomon, S. C., Phillips, R. J., Roberts, J. H. and Zuber, M. T. (2013). Thermal evolution of Mercury as constrained by MESSENGER observations. J. Geophys. Res. Planets, 118, 1033–1044. Muller, P. M. and Sjogren, W. L. (1968). Mascons: Lunar mass concentrations. Science, 161, 680–684. Murchie, S. L., Klima, R. L., Denevi, B. W., Ernst, C. M., Keller, M. R., Domingue, D. L., Blewett, D. T., Chabot, N. L., Hash, C. D., Malaret, E., Izenberg, N. R., Vilas, F., Nittler, L. R., Gillis-Davis, J. J., Head, J. W. and Solomon, S. C. (2015). Orbital multispectral mapping of Mercury with the MESSENGER Mercury Dual Imaging System: Evidence for the origins of plains units and low-reflectance material. Icarus, 254, 287–305, doi:10.1016/j. icarus.2015.03.027. Neumann, G. A., Perry, M. E., Mazarico, E., Ernst, C. M., Zuber, M. T., Smith, D. E., Becker, K. J., Gaskell, R. E., Head, J. W., Robinson, M. S. and Solomon, S. C. (2016). Mercury shape model from laser altimetry and planetary comparisons. Lunar Planet. Sci., 47, abstract 2087. Nimmo, F. and Watters, T. R. (2004). Depth of faulting on Mercury: Implications for heat flux and crustal and effective elastic thickness. Geophys. Res. Lett., 31, L02701, doi:10.1029/2003GL018847. Nittler, L. R., Starr, R. D., Weider, S. Z., McCoy, T. J., Boynton, W. V., Ebel, D. S., Ernst, C. M., Evans, L. G., Goldsten, J. O., Hamara, D. K., Lawrence, D. J., McNutt, R. L., Jr., Schlemm, C. E., II, Solomon, S. C. and Sprague, A. L. (2011). The major-element composition of Mercury’s surface from MESSENGER X-ray spectrometry. Science, 333, 1847–1850, doi:10.1126/science.1211567. Noyelles, B., Frouard, J., Makarov, V. V. and Efroimsky, M. (2014). Spin–orbit evolution of Mercury revisited. Icarus, 241, 26–44. Oberst, J., Preusker, F., Phillips, R. J., Watters, T. R., Head, J. W., Zuber, M. T. and Solomon, S. C. (2010). The morphology of Mercury’s Caloris basin as seen in MESSENGER stereo topographic models. Icarus, 209, 230–238. Oberst, J., Elgner, S., Turner, F. S., Perry, M. E., Gaskell, R. W., Zuber, M. T., Robinson, M. S. and Solomon, S. C. (2011). Radius and limb topography of Mercury obtained from images acquired during the MESSENGER flybys. Planet. Space Sci., 59, 1918–1924. Ostrach, L. R., Robinson, M. S., Whitten, J. L., Fassett, C. I., Strom, R. G., Head, J. W. and Solomon, S. C. (2015). Extent, age, and resurfacing history of the northern smooth plains on Mercury from MESSENGER observations. Icarus, 250, 602–622. Padovan, S., Wieczorek, M. A., Margot, J.-L., Tosi, N. and Solomon, S. C. (2015). Thickness of the crust of Mercury from geoid-totopography ratios. Geophys. Res. Lett., 42, 1029–1038.

References Page, B. R., Bryan, C. G., Williams, K. E., Taylor, A. H. and Williams, B. G. (2014). Tuning the MESSENGER state estimation filter for controlled descent to Mercury impact. Astrodynamics Specialist Conference, American Institute of Aeronautics and Astronautics/ American Astronautical Society, paper AIAA-2014–4129, 16 pp., San Diego, CA, 4–7 August. Pavlis, D. E., Wimert, J. and McCarthy, J. J. (2013). GEODYN II System Description, Volumes 1–5, contractor report. Greenbelt, MD: SGT Inc. Peplowski, P. N., Klima, R. L., Lawrence, D. J., Ernst, C. M., Denevi,B. W., Frank, E. A., Goldsten, J. O., Murchie, S. L., Nittler, L.R. and Solomon, S. C. (2016). Remote sensing evidence for an ancient carbon-bearing crust on Mercury. Nature Geosci., 9, 273–276, doi:10.1038/ngeo2669. Perry, M. E., Kahan, D. S., Barnouin, O. S., Ernst, C. M., Solomon, S. C., Zuber, M. T., Smith, D. E., Phillips, R. J., Srinivasan, D. K., Oberst, J. and Asmar, S. W. (2011). Measurement of the radius of Mercury by radio occultation during the MESSENGER flybys. Planet. Space Sci., 59, 1925–1931. Perry, M. E., Neumann, G. A., Phillips, R. J., Barnouin, O. S., Ernst, C. M., Kahan, D. S., Solomon, S. C., Zuber, M. T., Smith, D. E., Hauck, S. A., II, Peale, S. J., Margot, J.-L., Mazarico, E., Johnson, C. L., Gaskell, R. W., Roberts, J. H., McNutt, R. L., Jr. and Oberst, J. (2015). The low-degree shape of Mercury. Geophys. Res. Lett., 42, 6951–6958, doi:10.1002/2015GL065101. Peterson, G. A., Johnson, C. L., Byrne, P. K., Phillips, R. J. and Neumann, G. A. (2017). Depth of faulting in Mercury’s northern hemisphere from thrust fault morphology. Lunar Planet. Sci., 48, abstract 2315. Phillips, R. J. and Lambeck, K. (1980). Gravity fields of the terrestrial planets: Long-wavelength anomalies and tectonics. Rev. Geophys. Space Phys., 18, 27–76. Phillips, R. J., Johnson, C. L., Mackwell, S. J., Morgan, P., Sandwell, D. T. and Zuber, M. T. (1997). Lithospheric mechanics and dynamics of Venus. In Venus II, ed. S. W. Bougher, D. M. Hunten and R. J. Phillips. Tucson, AZ: University of Arizona Press, pp. 1163–1204. Phillips, R. J., Johnson, C. L., Perry, M. E., Hauck, S. A., II, James, P. B., Mazarico, E., Lemoine, F. G., Neumann, G., Peale, S. J., Siegler, M. A., Smith, D. E., Solomon, S. C. and Zuber, M. T. (2014). Mercury’s 2nd-degree shape and geoid: Lunar comparisons and thermal anomalies. Lunar Planet. Sci., 45, abstract 2634. Poblet, J. and Lisle, R. J. (2011). Kinematic evolution and structural styles of fold-and-thrust belts. In Kinematic Evolution and Structural Styles of Fold-and-Thrust Belts, ed. J. Poblet and R. J. Lisle, Special Publication 349. London: Geological Society, pp. 1–24. Preusker, F., Oberst, J., Head, J. W., Watters, T. R., Robinson, M. S., Zuber, M. T. and Solomon, S. C. (2011). Stereo topographic models of Mercury after three MESSENGER flybys. Planet. Space Sci., 59, 1910–1917, doi:10.1016/j.pss.2011.07.005. Qin, C., Zhong, S. and Phillips, R. J. (2018). Formation of the lunar fossil bulges and its implication for the early Earth and Moon. Geophys. Res. Lett., 45, 1286–1296, doi:10.1002/2017GL076278. Riner, M. A., Lucey, P. G., Desch, S. J. and McCubbin, F. M. (2009). Nature of opaque components on Mercury: Insights into a Mercurian magma ocean. Geophys. Res. Lett., 36, L02201, doi:10.1029/ 2008GL036128. Robinson, M. S., Murchie, S. L., Blewett, D. T., Domingue, D. L., Hawkins, S. E., III, Head, J. W., Holsclaw, G. M., McClintock, W. E., McCoy, T. J., McNutt, R. L., Jr., Prockter, L. M., Solomon, S. C. and Watters T. R. (2008). Reflectance and color variations on Mercury: Regolith processes and compositional heterogeneity. Science, 321, 66–69. Sedgwick, W. F. (1898). On the figure of the Moon. Messenger Math., 27, 171–173.


Seidelmann, P. K., Abalakin, V. K., Bursa, M., Davies, M. E., de Bergh, C., Lieske, J. H., Oberst, J., Simon, J. L., Standish, E. M., Stooke, P. and Thomas, P. C. (2002). Report of the IAU/IAG Working Group on Cartographic Coordinates and Rotational Elements of the Planets and Satellites: 2000. Celest. Mech. Dyn. Astron., 82, 83–110. Şengör, A. M. C., Natal’in, B. A. and Burtman, V. S. (1993). Evolution of the Altaid tectonic collage and Palaeozoic crustal growth in Eurasia. Nature, 364, 299–307. Shearer, C. K., Hess, P. C, Wieczorek, M. A., Pritchard, M. E., Parmentier, E. M., Borg, L. E., Longhi, J., Elkins-Tanton, L. T., Neal, C. R., Antonenko, I., Canup, R. M., Halliday, A. N., Grove, T. L., Hager, B. H., Lee, D.-C. and Wiechert, U. (2006). Thermal and magmatic evolution of the Moon. Rev. Mineral. Geochem., 60, 365–518. Siegler, M. A., Bills, B. G. and Paige, D. A. (2013). Orbital eccentricity driven temperature variation at Mercury’s poles. J. Geophys. Res. Planets, 118, 930–937. Smith, D. E., Zuber, M. T., Solomon, S. C., Phillips, R. J., Head, J. W., Garvin, J. B., Banerdt, W. B., Muhleman, D. O., Pettengill, G. H., Neumann, G. A., Lemoine, F. G., Abshire, J. B., Aharonson, O., Brown, C. D., Hauck, S. A., II, Ivanov, A. B., McGovern, P. J., Zwally, H. J. and Duxbury, T. C. (1999). The global topography of Mars and implications for surface evolution. Science, 284, 1495–1503. Smith, D. E., Zuber, M. T., Sun, X., Neumann, G. A., Cavanaugh, J. F., McGarry, J. F. and Zagwodzki, T. W. (2006). Two-way laser link over interplanetary distance. Science, 311, 53. Smith, D. E., Zuber, M. T., Phillips, R. J., Solomon, S. C., Neumann, G. A., Lemoine, F. G., Torrence, M., Peale, S. J., Margot, J.-L., Barmouin-Jha, O., Head, J. W. and Talpe, M. (2010a). The equatorial shape and gravity field of Mercury from MESSENGER flybys 1 and 2. Icarus, 209, 247–255. Smith, D. E., Zuber, M. T., Neumann, G. A. Lemoine, F. G., Mazarico, E., Torrence, M. H., McGarry, J. F., Rowlands, D. D., Head, J. W., Duxbury, T. H., Aharonson, O., Lucey, P. G., Robinson, M. S., Barnouin, O. S., Cavanaugh, J. F., Sun, X. L., Liva, P., Mao, D. D., Smith, J. C. and Bartles, A. E. (2010b). Initial observations from the Lunar Orbiter Laser Altimeter (LOLA). Geophys. Res. Lett., 37, L18204, doi:10.1029/2010GL043751. Smith, D. E., Zuber, M. T., Phillips, R. J., Solomon, S. C., Hauck, S. A., II, Lemoine, F. G., Mazarico, E., Neumann, G. A., Peale, S. J., Margot, J. L., Johnson, C. L., Torrence, M. H., Perry, M. E., Rowlands, D. D., Goossens, S., Head, J. W. and Taylor, A. H. (2012). Gravity field and internal structure of Mercury from MESSENGER. Science, 336, 214–217. Spudis, P. D. (1996). The Once and Future Moon. Washington, DC: Smithsonian Institution Press, pp. 117–118, Spudis, P. D., McGovern, P. J. and Kiefer, W. S. (2013). Large shield volcanoes on the Moon. J. Geophys. Res. Planets, 118, 1063–1081. Srinivasan, D. K., Perry, M. E., Fielhauer, K. B., Smith, D. E. and Zuber, M. T. (2007). The radio frequency subsystem and radio science on the MESSENGER mission. Space Sci. Rev., 131, 557– 571. Stacey, F. D. and Davis, P. M. (2008). Physics of the Earth, 4th edn. Cambridge: Cambridge University Press. Strom, R. G., Trask, N. J. and Guest, J. E. (1975). Tectonism and volcanism on Mercury. J. Geophys. Res., 80, 2478–2507. Suess, E. (1909). Das Antlitz der Erde, Vol. III.2, ed. E. Suess. Leipzig: F. Tempsky. Tosi, N., Grott, M., Plesa, A. C. and Breuer, D. (2013). Thermochemical evolution of Mercury’s interior. J. Geophys. Res. Planets, 118, 2474–2487. Tosi, N., Čadek, O., Běhounková, M., Káňová, M., Plesa, A. C., Grott, M., Breuer, D., Padovan, S. and Wieczorek, M. A. (2015). Mercury’s


Mercury’s Crust and Lithosphere: Structure and Mechanics

low-degree geoid and topography controlled by insolation-driven elastic deformation. Geophys. Res. Lett., 42, 7327–7335. Turcotte, D. L., Willemann, R. J., Haxby, W. F. and Norberry, J. (1981). Role of membrane stress in the support of planetary topography. J. Geophys. Res., 86, 3951–3959. Vander Kaaden, K. E. and McCubbin F. M. (2015). Exotic crust formation on Mercury: Consequences of a shallow, FeO-poor mantle. J. Geophys. Res. Planets, 120, 195–209. Vasavada, A. R., Paige, D. A. and Wood, S. E. (1999). Near-surface temperatures on Mercury and the Moon and the stability of polar ice deposits. Icarus, 141, 179–193. Watters, T. R., Schultz, R. A., Robinson, M. S. and Cook, A. C. (2002). The mechanical and thermal structure of Mercury’s early lithosphere. Geophys. Res. Lett., 29, 1542, doi:10.029/2001GL014308. Watts, A. B. (1978). An analysis of isostasy in the world’s oceans 1. Hawaiian-Emperor seamount chain. J. Geophys. Res., 83, 5989– 6004. Watts, A. B. (2001) Isostasy and Flexure of the Lithosphere. Cambridge: Cambridge University Press. Wieczorek, M. A. and Phillips, R. J. (1998). Potential anomalies on a sphere: Applications to the thickness of the lunar crust. J. Geophys. Res., 103, 1715–1724. Wieczorek, M. A. and Simons, F. J. (2005). Localized spectral analysis on the sphere. Geophys. J. Int., 162, 655–675. Wieczorek, M. A., Correia, A. C. M., Le Feuvre, M., Laskar, J. and Rambaux, N. (2011). Mercury’s spin–orbit resonance explained by initial retrograde and subsequent synchronous rotation. Nature Geosci., 5, 18–21. Wieczorek, M. A., Neumann, G. A., Nimmo, F., Kiefer, W. S., Taylor, G. J., Melosh, H. J., Phillips, R. J., Solomon, S. C., Andrews-Hanna, J. C., Asmar, S. W., Konopliv, A. S., Lemoine, F. G., Smith, D. E., Watkins, M. M., Williams, J. G. and Zuber, M. T. (2013). The crust of the Moon as seen by GRAIL. Science, 339, 671–675. Williams, J. G., Boggs, D. H., Yoder, C. F., Ratcliff, J. T. and Dickey, J. O. (2001). Lunar rotational dissipation in solid body and molten core. J. Geophys. Res., 106, 27,933–27,968. Williams, J.-P., Ruiz, J., Rosenburg, M. A., Aharonson, O. and Phillips, R. J. (2011). Insolation driven variations of Mercury’s

lithospheric strength. J. Geophys. Res., 116, E01008, doi:10.1029/2001JE003655. Yseboodt, M. and Margot, J.-L. (2006). Evolution of Mercury’s obliquity. Icarus, 181, 327–337. Zuber, M. T. and Parmentier, E. M. (1995). Formation of fold-andthrust belts on Venus by thick-skinned deformation. Nature, 377, 704–707. Zuber, M. T., Smith, D. E., Solomon, S. C., Muhleman, D. O., Head, J. W., Garvin, J. B., Abshire, J. B. and Bufton, J. L. (1992). The Mars Observer Laser Altimeter investigation. J. Geophys. Res., 97, 7781–7797. Zuber, M. T., Aharonson, O., Aurnou, J. M., Cheng, A. F., Hauck, S. A., II, Heimpel, M. H., Neumann, G. A., Peale, S. J., Phillips, R. J., Smith, D. E., Solomon, S. C. and Stanley, S. (2007). The geophysics of Mercury: Current status and anticipated insights from the MESSENGER mission. Space Sci. Rev., 131, 105–132. Zuber, M. T., Montési, L. G. J., Farmer, G. T., Hauck, S. A., II, Ritzer, A., Phillips, R. J., Solomon, S. C., Smith, D. E., Talpe, M. J., Head, J. W., III, Neumann, G. A., Watters, T. R. and Johnson, C. L. (2010). Accommodation of lithospheric shortening on Mercury from altimetric profiles of ridges and lobate scarps measured during MESSENGER flybys 1 and 2. Icarus, 209, 247–255. Zuber, M. T., Smith, D. E., Phillips, R. J., Solomon, S. C., Neumann, G. A., Hauck, S. A., II, Peale, S. J., Barnouin, O. S., Head, J. W., Johnson, C. L., Lemoine, F. G., Mazarico, E., Sun, X., Torrence, M. H., Freed, A. M., Klimczak, C., Margot, J. L., Oberst, J., Perry, M. E., McNutt, R. L., Jr., Balcerski, J. A., Michel, N., Talpe, M. J. and Yang, D. (2012). Topography of the northern hemisphere of Mercury from MESSENGER laser altimetry. Science, 336, 217–220. Zuber, M. T., Smith, D. E., Watkins, M. M., Asmar, S. W., Konopliv, A. S., Lemoine, F. G., Melosh, H. J., Neumann, G. A., Phillips, R. J., Solomon, S. C., Wieczorek, M. A., Williams, J. G., Goossens, S. J., Kruizinga, G., Mazarico, E., Park, R. S. and Yuan, D. N. (2013). Gravity field of the Moon from the Gravity Recovery and Interior Laboratory (GRAIL) mission. Science, 339, 668–671.


a wide range of problems, but Mercury’s unusual insolation and thermal patterns violate our symmetry assumptions. These assumptions must be lifted for certain applications that require precise temperature distributions. Our primary objective is to provide a family of simplified models of Mercury’s interior that satisfy the currently available observational constraints. A secondary objective is to select, among these models, a recommended model that matches all available constraints. This model may be considered a preliminary reference Mercury model (PRMM), evoking a distant connection with its venerable Earth analog (Dziewonski and Anderson, 1981).

4 .1 I N TR O D UCT I O N 4.1.1 Importance of Planetary Interiors We seek to understand the interior structures of planetary bodies because the interiors affect planetary properties and processes in several fundamental ways. First, a knowledge of the interior informs us about a planet’s makeup and enables us to test hypotheses related to planet formation. Second, interior properties dictate the thermal evolution of planetary bodies and, consequently, the history of volcanism and tectonics on these bodies. Many geological features are the surface expression of processes that take place below the surface. Third, the structure of the interior and the nature of the interactions among inner core, outer core, and mantle have a profound influence on the evolution of the spin state and the response of the planet to external forces and torques. These processes dictate the planet’s tectonic and insolation regimes and also affect its overall shape. Finally, interior properties control the generation of planetary magnetic fields, and, therefore, the development of magnetospheres. Four of the six primary science objectives of the MESSENGER mission (Solomon et al., 2001; Chapter 1) rely on an understanding of the planet’s interior structure. These four mission objectives pertain to the high density of Mercury, its geological history, the nature of its magnetic field, and the structure of its core.

4.1.3 Available Observational Constraints All of our knowledge about Mercury’s interior comes from Earthbased observations, three Mariner 10 flybys, three MESSENGER flybys, and the four-year orbital phase of the MESSENGER mission. In the absence of seismological data, our information about the interior comes primarily from geodesy, the study of the gravity field, shape, and spin state of the planet, including solidbody tides. We also draw on constraints derived from the surface expression of global contraction and observations of surface composition, with the caveat that the composition at depth may be substantially different from that inferred for surface material. The structure of the magnetic field and its dynamo origin can also be used to inform interior models.

4.1.2 Objectives

4.1.4 Outline

An ideal representation of a planetary interior would include the description of physical and chemical quantities at every location within the volume of the planetary body at every point in time. Here, we focus on a description of Mercury’s interior at the current epoch. For a description of the evolution of the state of the planet over geological time, see Chapter 19. Because our ability to specify properties throughout the planetary volume is limited, we simplify the problem by assuming axial or spherical symmetry. Specifically, we seek self-consistent depth profiles of density, pressure, and temperature, informed by observational constraints (radius, mass, moment of inertia, composition). The solution requires the use of equations of state and assumptions about material properties, both guided by laboratory data. We compute the bulk modulus and thermal expansion coefficient as part of the estimation process, and we use the profiles to compute other rheological properties, such as viscosity and additional elastic moduli. Finally, we use our models to numerically evaluate the planet’s tidal response and compare it with observational data. Our models of the interior structure are relevant to

The primary observational constraints (Sections 4.2–4.4) are used to develop two- and three-layer structural models (Section 4.5). We then add compositional constraints (Section 4.6) and develop multi-layer models (Section 4.7). We examine the tidal response of the planet (Section 4.8) and the influence of an inner core (Section 4.9). We conclude with a discussion of a representative interior model (Section 4.10) and its implications (Section 4.11).

4.2 R O T AT I O N AL DY NA MI C S In his classic 1976 paper, Stanton J. Peale described the effects of a molten core on the dynamics of Mercury’s rotation and proposed an ingenious method for measuring the size and state of the core (Peale, 1976). Most of our knowledge about Mercury’s interior structure can be traced to Peale’s ideas and to the powerful connection between dynamics and geophysics. We review aspects of Mercury’s rotational dynamics that are relevant to determining its interior structure. Peale (1988) provided a more extensive review.



Mercury’s Internal Structure

4.2.1 Spin–Orbit Resonance Radar observations by Pettengill and Dyce (1965) revealed that the spin period of Mercury differs from its orbital period. To explain the radar results, Colombo (1965) correctly hypothesized that Mercury rotates on its spin axis three times for every two revolutions around the Sun. Mercury is the only known planetary body to exhibit a 3:2 spin–orbit resonance (Colombo, 1966; Goldreich and Peale, 1966).

Spin axis




plane Orbital plane

4.2.2 Physical Librations Peale’s observational procedure allows the detection of a molten core by measuring deviations from the mean resonant spin rate of the planet. As Mercury follows its eccentric orbit, it experiences periodically reversing torques due to the gravitational influence of the Sun on the asymmetric shape of the planet. The torques affect the rotational angular momentum and cause small deviations of the spin frequency from its resonant value of 3/2 times the mean orbital frequency. The resulting oscillations in longitude are called physical librations, not to be confused with optical librations, which are the torque-free oscillations of the long axis of a uniformly spinning body about the line connecting it to a central body. Because the forcing and rotational response occur with a period P ~ 88 days dictated by Mercury’s orbital motion, these librations have been referred to as forced librations. This terminology is not universally accepted (e.g., Bois, 1995) and loses meaning when the amount of angular momentum exchanged between spin and orbit is not negligible (e.g., Naidu and Margot, 2015). We will instead refer to these librations as 88-day librations, in part to distinguish them from librations with longer periods. The amplitude ϕ0 of the 88-day librations for a solid Mercury can be written as (Peale, 1972, 1988):   3 ðB  AÞ 959 4 1  11e2 þ e þ ... ; ð4:1Þ ϕ0 ¼ 2 C 48 where A < B < C are the principal moments of inertia and e is the orbital eccentricity, currently ~0.2056 (e.g., Stark et al., 2015b). This equation encapsulates the fact that the gravitational torques are proportional to the difference in equatorial moments of inertia ðB  AÞ. The polar moment of inertia C appears in the denominator as it represents a measure of the resistance to changes in rotational motion. If the mantle is decoupled from a molten core that does not participate in the 88-day librations, then the moment of inertia in the denominator must be replaced by Cmþcr , the value appropriate for the mantle and crust. Peale (1976) noted that Cmþcr =C ≃ 0:5, suggesting that a measurement of the amplitude of the 88-day librations can be used to determine the state of the core if ðB  AÞ is known. This result holds over a wide range of core–mantle coupling behaviors (Peale et al., 2002; Rambaux et al., 2007). 4.2.3 Cassini State Peale (1969, 1988) formulated general equations for the motion of the rotational axis of a triaxial body under the influence of gravitational torques. He wrote these equations

Figure 4.1. Geometry of Cassini state 1: the three vectors representing spin axis orientation (black), normal to the orbital plane (blue), and normal to the Laplace plane (red) remain coplanar as the orbit precesses around the Laplace plane with a ~300,000-year period. The inclination of Mercury’s orbit with respect to the Laplace plane is represented by the angle ι, which is shown to scale. The tilt of Mercury’s spin axis with respect to the orbit normal is the obliquity θ, which is shown with an exaggeration factor of 100 for clarity.

in the context of an orbit that precesses at a fixed rate around a reference plane called the Laplace plane, extending and refining earlier work by Colombo (1966). These equations generalize Cassini’s laws and describe the dynamics of the Moon, Mercury, Galilean satellites, and other bodies. In the case of Mercury, the gravitational torques are due to the Sun, and the ~300,000-year precession of the orbit is due to the effect of external perturbers, primarily Jupiter, Venus, Saturn, and Earth. On the basis of these theoretical calculations, Peale (1969, 1988) predicted that tidal evolution would carry Mercury to a Cassini state, in which the spin axis orientation, orbit normal, and normal to the Laplace plane remain coplanar (Figure 4.1). Specifically, he predicted that Mercury would reach Cassini state 1, with an obliquity near zero degrees. Numerical simulations (Bills and Comstock, 2005; Yseboodt and Margot, 2006; Peale, 2006; Bois and Rambaux, 2007) and analytical calculations (D’Hoedt and Lemaître, 2008) support these predictions. In a Cassini state, the obliquity has evolved to a value where the spin precession period matches the orbit precession period (Gladman et al., 1996). Because the spin precession period and the gravitational torques depend on moment of inertia differences, there is a powerful relationship between the obliquity of a body in a Cassini state and its moments of inertia. Peale (1976, 1988) wrote:     CA BA þ K2 ðθÞ ¼ K3 ðθÞ; ð4:2Þ K1 ðθÞ C C where K1 ; K2 ; K3 are functions of the obliquity θ that involve the orbital eccentricity, inclination with respect to the Laplace plane, mean motion, spin rate, and precession rate. In this equation, the appropriate moment of inertia in the denominator is that of the entire planet, even if the core is molten, because it

4.3 Gravity Constraints is hypothesized that the core follows the mantle on the ~300,000-year timescale of the orbital precession. If we can confirm that Mercury is in a Cassini state, a measurement of the obliquity becomes extremely valuable: it provides a direct constraint on moment of inertia differences and, in combination with degree-2 gravity information, on the polar moment of inertia. A free precession of the spin axis about the Cassini state could, in principle, compromise the determination of the obliquity. However, such free precession would require a recent excitation because the corresponding damping timescale is ~105 yr (Peale, 2005).

4.2.4 Polar Moment of Inertia Absent seismological data, the polar moment of inertia is arguably the most important quantity needed to quantify the interior structure of a planetary body. Peale (1976, 1988) showed that it is possible to measure the polar moment of inertia C by combining the obliquity with two quantities related to the gravity field. The gravity field of a body of mass M and radius R can be described with spherical harmonics (e.g., Kaula, 2000). The second-degree coefficients C20 and C22 in the spherical harmonic expansion are related to the moments of inertia, as follows: C20 ¼ 

C22 ¼

ðC  ðA þ BÞ=2Þ ; MR2

ðB  AÞ : 4MR2



Combining equations (4.2), (4.3), and (4.4), we find C K1 ðθÞ K2 ðθÞ þ 4C22 ; ¼ ðC20 þ 2C22 Þ 2 MR K3 ðθÞ K3 ðθÞ


which provides a direct relationship between the obliquity, gravity harmonics, and polar moment of inertia for bodies in Cassini state 1. To complete Peale’s argument, we determine the polar moment of inertia of the core, which can be done if the core is molten and does not participate in the 88-day librations. To do so, we write the identity     Cmþcr Cmþcr B  A MR2 ¼ ; ð4:6Þ MR2 C BA C which yields the moment of inertia of the mantle and crust Cmþcr and, therefore, the moment of inertia of the core Cc ¼ C  Cmþcr . Two spin-state quantities and two gravity quantities provide all the information necessary to determine these values. A measurement of the libration amplitude ϕ0 provides a direct estimate of the first factor on the right-hand side of equation (4.6) via equation (4.1). A measurement of the gravitational harmonic C22 provides a direct estimate of the second factor. Measurements of the obliquity, C20, and C22 yield an estimate of the third factor via equation (4.5). The four quantities ϕ0 ; θ; C20 ; and C22 identified by Peale (1976, 1988) thus provide a powerful probe of the interior structure of the planet.


4.2.5 Orbital Precession Implementing Peale’s procedure requires precise knowledge of Mercury’s orbital configuration. Whereas the mean motion and orbital eccentricity have been determined from centuries of observations, relatively little attention had been paid to the orientation of the Laplace plane and the orbital precession rate. Yseboodt and Margot (2006) used a Hamiltonian approach and numerical fits to ephemeris data to determine these ancillary quantities. They showed that the Laplace plane orientation varies due to planetary perturbations on ~10 kyr timescales, and they defined an instantaneous Laplace plane valid at the current epoch for the purpose of identifying the position of the Cassini state and interpreting spin-gravity data. Yseboodt and Margot (2006) gave the coordinates of the normal to the instantaneous Laplace plane in ecliptic and equatorial coordinates at epoch J2000.0 as λinst ¼ 66:6°; βinst ¼ 86:725°;


RAinst ¼ 273:72°; DEC inst ¼ 69:53°;


where λ is ecliptic longitude, β is ecliptic latitude, RA is right ascension, and DEC is declination. The uncertainty in the determination is of order 1°, but the orientation of the narrow error ellipse is such that it can affect the interpretation of the spinstate data only at a level that is well below that due to measurement uncertainties. The inclination of Mercury’s orbit with respect to the instantaneous Laplace plane and the orbit precession rate about that plane ˙ ¼ 0:110°=century, at the current epoch are ι ¼ 8:6° and Ω respectively (Yseboodt and Margot, 2006). We will use both of these quantities to estimate Mercury’s interior structure in Sections 4.5 and 4.7. Stark et al. (2015b) performed an independent analysis and confirmed the values of Yseboodt and Margot (2006), including the orientation of the instantaneous ˙ Laplace plane, the inclination ι, and the precession rate Ω. D’Hoedt et al. (2009) used a Hamiltonian approach and found an instantaneous Laplace plane orientation that differs from our preferred value by 1.4°.

4.3 G RA VITY C ONS TR AIN TS 4.3.1 Methods We are interested in measuring the masses and sizes of planetary bodies because bulk density is a fundamental indicator of composition. In multi-planet systems, masses can be estimated by observing the effects of mutual orbital perturbations, manifested as variations in orbital elements or variations in transit times. Another common mass measurement technique is to determine the orbits of natural satellites. The most precise mass estimates are obtained by radiometric tracking of a spacecraft while it is in close proximity to the body of interest, typically by using the onboard telecommunications system and a network of ground-based radio telescopes. The geodetic observations are then used to obtain a spherical harmonic expansion of the gravity field and to reconstruct the spacecraft trajectory with high fidelity. In addition to providing


Mercury’s Internal Structure

high-precision mass estimates, this technique enables the measurement of the spherical harmonic coefficients C20 and C22, which provide important constraints on interior structure (Section 4.2.4). In the following sections, we describe gravity results obtained from tracking the Mariner 10 spacecraft at a frequency of 2.3 GHz (S-band) during three flybys in 1974–1975 and the MESSENGER spacecraft at frequencies of 7.2 GHz uplink and 8.4 GHz downlink (X-band) during the flybys and orbital phase of the mission. 4.3.2 Mass and Density Results The mass, size, and density of Mercury were known with remarkable precision prior to the exploration of the planet by spacecraft. After adding radar measurements to two centuries of optical observations, Ash et al. (1971) fit planetary ephemerides and determined Mercury’s mass to 0.25% fractional uncertainty. They found a value of 6,025,000 ± 15,000 in inverse solar masses, i.e., M ¼ ð3:300  0:008Þ  1023 kg, which is almost identical to the modern estimate. Using this measurement and the radar estimate of the average equatorial radius that was available at the time, R ¼ ð2439  1Þ km, it was apparent that Mercury’s bulk density was anomalously high, with ρ ¼ ð5430  15Þ kg m3 . On the basis of their density calculation, Ash et al. (1971) concluded that Mercury must be substantially richer in heavy elements than Earth. The pre-Mariner 10 estimates of mass, size, and density remain in excellent agreement with the MESSENGER results, but spacecraft data have enabled a reduction in uncertainties by a factor of ~50. Howard et al. (1974) analyzed the tracking data from the first flyby of Mercury by Mariner 10 and obtained a gravitational parameter GM ¼ ð2:2032  0:0002Þ  1013 m3 s2 , where G is the gravitational constant. Analysis of data from all three Mariner 10 flybys yielded GM ¼ ð2:203209 0:000091Þ  1013 m3 s2 (Anderson et al., 1987). From more than three years of orbital tracking data of MESSENGER, Mazarico et al. (2014) obtained GM ¼ ð2:203187080 0:000000086Þ  1013 m3 s2 , estimated from a gravity field solution to degree and order 50. An independent analysis to degree and order 40 by Verma and Margot (2016) yielded GM ¼ ð2:203187404  0:000000090Þ  1013 m3 s2 . When translating the MESSENGER values to a mass estimate, the majority of the uncertainty comes from the 5  105 uncertainty in the gravitational constant. With G ¼ ð6:67408 0:00031Þ  1011 m3 kg1 s2 (Mohr et al., 2016), the current best estimate of the mass of Mercury is M ¼ ð3:301110  0:00015Þ  1023 kg:


From a combination of laser altimetry (Zuber et al., 2012) and radio occultation data, Perry et al. (2015) determined Mercury’s average radius to be R ¼ ð2439:36  0:02Þ km;


although the stated radius uncertainty may be optimistic given the sparse sampling of the southern hemisphere. The corresponding bulk density is

ρ ¼ ð5429:30  0:28Þ kg m3 :


Mercury’s bulk density is similar to that of Earth, ρ ⊕ ¼ 5514 kg m3 , despite the different sizes of the two bodies. The pressure P at the center of a homogeneous sphere scales as P ∝ ρ2 R2 , so materials in Earth’s interior are more compressed (i.e., denser) than those in Mercury’s interior. If we assume that both planets are made of a combination of a light component (i.e., silicates) and a heavy component (i.e., metals), we can infer from their similar densities and differing sizes that Mercury has a larger metallic component, as recognized by Ash et al. (1971). 4.3.3 C20 and C22 Results The first measurements of the C20 and C22 gravity coefficients were obtained from Mariner 10 data recorded during one equatorial flyby with ~700 km minimum altitude and one polar flyby with ~300 km minimum altitude. Anderson et al. (1987) determined C20 ¼ ð6:0  2:0Þ  105 and C22 ¼ ð1:0  0:5Þ 105 . These values have large fractional uncertainties because there were only two favorable flybys, but the values are consistent with the most recent MESSENGER results (Mazarico et al., 2014; Verma and Margot, 2016). With the normalization that is commonly used in geodetic studies (Kaula, 2000, p. 7), C 20 ¼ the Mariner 10 values can also be expressed p as ffiffiffiffiffiffiffiffiffiffi pffiffiffi C20 = 5 ¼ ð2:68  0:9Þ  105 and C 22 ¼ C22 = 5=12 ¼ ð1:55  0:8Þ  105 , where the overbar indicates normalized coefficients. The next opportunity for measurements arose from the three MESSENGER flybys of Mercury in 2008–2009. However, the equatorial geometry of these flybys did not provide adequate leverage to measure C20 accurately. Because the Mariner 10 tracking data have been lost, it was not possible to perform a joint solution including both equatorial and polar flybys. For these reasons, Smith et al. C 20 ¼ (2010) cautioned that their recovery of ð0:86  0:30Þ  105 might not be reliable. However, the equatorial geometry was suitable for an accurate estimate of C 22 ¼ ð1:26  0:12Þ  105 . Data acquired during the orbital phase of the MESSENGER mission provided significantly better sensitivity and lower uncertainties. Smith et al. (2012) analyzed the first six months of data (>300 orbits) and found C 20 ¼ ð2:25  0:01Þ  105 and C 22 ¼ ð1:25  0:01Þ  105 , where the error bars represent a calibrated uncertainty that is about 10 times the formal uncertainty of the fit. An independent analysis of the same data by Genova et al. (2013) confirmed these results. More recently, Mazarico et al. (2014) analyzed three years of data (2275 orbits) and estimated a gravity field solution to degree and order 50. This solution yielded an order-of-magnitude improvement in the calibrated uncertainties in C20 and C22: C 20 ¼ ð2:2505  0:001Þ  105 and C 22 ¼ ð1:2454  0:001Þ  105 . An independent analysis by Verma and Margot (2016) confirmed these values to better than 0.4%. The unnormalized quantities that we use in equations (4.3)– (4.6) are based on the Mazarico et al. (2014) values:

4.4 Spin-State Constraints C20 ¼ ð5:0323  0:0022Þ  105 and C22 ¼ ð0:8039  0:0006Þ  105 . The J2 =C22 ¼ C20 =C22 value of 6.26 is distinct from the equilibrium value of 7.86 for a body in a 3:2 spin– orbit resonance with the current value of the orbital eccentricity (Matsuyama and Nimmo, 2009), indicating that Mercury is not in hydrostatic equilibrium. It has been proposed that the values of the low-degree gravity coefficients can be explained by deep density anomalies induced by uneven insolation at the surface (Tosi et al., 2015). 4.3.4 Results for k2 In addition to the static gravity field, Mazarico et al. (2014) also solved for the time-variable degree-2 potential that captures the tidal forcing due to the Sun. The tidal forcing is parameterized by the Love number k2 (Section 4.8.1). Mazarico et al. (2014) obtained an estimate of k2 ¼ 0:451  0:014. However, because of potential mismodeling and systematic effects in the analysis, they could not rule out a wider range of values (0.43–0.50). The preferred value of Verma and Margot (2016) is k2 ¼ 0:464  0:023. They, too, encountered a wider range of best-fit values (0.420–0.465) in various trials. The weighted mean of these two estimates is k2 ¼ 0:455  0:012. These estimates are within the expected range from theoretical studies (Van Hoolst and Jacobs, 2003; Van Hoolst et al., 2007; Rivoldini et al., 2009) and from predictions of interior models informed by MESSENGER data and Earth-based radar data (Padovan et al., 2014).

4 . 4 SP I N - S T A T E CO NS T R A I N T S Most of the quantities necessary to implement Peale’s method of probing Mercury’s interior were known when he wrote his paper in 1976. The mass, size, and density had been determined to ρm . We use equations (4.13) and (4.15) to derive the analytical

Peale’s formalism (Section 4.2.4) enabled a determination of Mercury’s polar moment of inertia. Margot et al. (2012) combined measurements of the obliquity and librations with gravity e ¼ 0:346  0:014. Stark et al. (2015a) also data and found C e ¼ 0:346  0:011. A uniform measured θ and ϕ0 , and found C e ¼ 0:4, and a body with a density profile density sphere has C e < 0:4. The Moon, with that increases with depth has C e C≃ 0:393 (Williams et al., 1996), is nearly homogeneous, e ¼ 0:3307 (Williams, 1994), has a whereas the Earth, with C substantial concentration of dense material near the center. e value suggests the presence of a dense Likewise, Mercury’s C metallic core. The moment of inertia of Mercury’s mantle and crust is also available from spin and gravity data (Equation 4.6). Margot et al. (2012) found Cmþcr =C ¼ 0:431  0:025 and Stark et al. (2015a) found Cmþcr =C ¼ 0:421  0:021. Weighted means of the Margot et al. (2012) and Stark et al. (2015a) results provide the most reliable estimates to date of the moments of inertia. We find e ¼ C ¼ 0:346  0:009; C MR2


Cmþcr ¼ 0:425  0:016: C


An error budget similar to that computed by Peale (1981, 1988) demonstrates that the dominant sources of uncertainties in the moment of inertia values can be attributed to spin quantities. Uncertainties arising from gravitational harmonics,

4.5 Two- and Three-Layer Structural Models tides, and orbital elements are at least an order of magnitude smaller (Noyelles and Lhotka, 2013; Baland et al., 2017). Further improvements to our knowledge of Mercury’s moments of inertia therefore require better estimates of obliquity and libration amplitude. Such improved estimates may also enable a determination of the tidal quality factor Q (Baland et al., 2017). 4.5.3 Two-Layer Model Results Using equations (4.16)–(4.18) and estimates of bulk density e (4.19), and Cmþcr =C (4.20), we infer (4.11), C Rc =R ¼ 0:8209; i:e:; Rc ¼ 2002 km;

ð4:21Þ 3

ρc =ρ ¼ 1:3344; i:e:; ρc ¼ 7245 kg m ;


ρm =ρ ¼ 0:5861; i:e:; ρm ¼ 3182 kg m3 :


The results obtained with the two-layer model are within one standard deviation of the results of more elaborate, multi-layer models that take into account mineralogical, geochemical, and rheological constraints on the composition and physical properties of the interior (Hauck et al., 2013; Rivoldini and Van Hoolst, 2013; Section 4.7). Figure 4.6 illustrates the consistency of the two-layer solution (star) and of the multi-layer models of Hauck et al. (2013) (error bars). The two-layer model results are also consistent with results from multi-layer models that consider the total contraction of the planet (Knibbe and van Westrenen, 2015). All points shown on Figure 4.6 are consistent with Mercury’s bulk density ρ. Knowledge of the normalized moment of inertia e restricts acceptable models to a black, constant-C e curve. The C resulting degeneracy corresponds to the underdetermined system of equations (4.13) and (4.15). Knowledge of the moment of inertia of the mantle further restricts acceptable models to the




e ¼ 0:346 black curve (not blue curve. The intersection of the C shown) and of the Cmþcr =C ¼ 0:431 blue curve yields the twolayer model solution. e and Cmþcr =C) can be used Although three observables (ρ, C, to reliably estimate the parameters of a two-layer model (core size, core density, and mantle density), they provide no information about additional phenomena related to the origin, evolution, and present physical state of the planet (e.g., mineralogical composition of the mantle, composition of the core, presence of a solid inner core). Additional insight can be obtained with more elaborate three-layer and multi-layer models. 4.5.4 Three-Layer Models We now consider a three-layer model with core, mantle, and crust of density ρcr . We express the core and mantle radii as fractions of the planetary radius, α ¼ Rc =R and β ¼ Rm =R. With this notation, we can write the bulk density, the polar moment of inertia, and the moment of inertia of the outer solid shell as follows: ρ ¼ ρc α3 þ ρm ðβ3  α3 Þ þ ρcr ð1  β3 Þ;


  e ¼ 2 ρc α5 þ ρm ðβ5  α5 Þ þ ρcr ð1  β5 Þ ; C 5 ρ ρ ρ


Cm þcr ρm ðβ5  α5 Þ þ ρc ð1  β5 Þ ¼ : 5 C ρc α þ ρm ðβ5  α5 Þ þ ρc ð1  β5 Þ


This system of equations has five unknowns and three observables. If we assume a crustal thickness value hcr (i.e., β) and a crustal density value ρcr , the system of equations (4.24)–(4.26) can be solved. The thickness of the crust of Mercury has been estimated from the combined analysis of gravity and topography data (Mazarico et al., 2014; Padovan et al., 2015; James


~ C = 0.39



0.70 0.30



0.4 0.40 0.30

0.2 0.24

Cm /C = 0.431 ~ C = 0.346, Cm /C = 0.431



Hauck et al. (2013)

0.10 1.0



ρc /ρ



Rc /R

ρm /ρ

0.60 0.27


Figure 4.6. Mantle density versus core density showing the consistency of the two-layer model results (star) with those of more elaborate, multilayer models (error bars). The position of the star e ¼ 0:346 and corresponds to values of C Cmþcr =C ¼ 0:431 (Margot et al., 2012). Error bars correspond to the one-standard-deviation intervals for ρc =ρ and ρm =ρ obtained by Hauck et al. (2013). The background color map indicates the value Rc =R in the two-layer model. Black curves illustrate models with various values of the normalized e The blue curve traces the moment of inertia C. locus of two-layer models with Cmþcr =C ¼ 0:431.


Mercury’s Internal Structure

et al., 2015). The density of the crust ρcr can be estimated from the measured composition of the surface of Mercury (e.g., Padovan et al., 2015). We use the results of Padovan et al. (2015) and consider two end-member cases: a crust that is low-density and thin (ρcr ¼ 2700 kg m3 , hcr ¼ 17 km) and a crust that is highdensity and thick (ρcr ¼ 3100 kg m3 , hcr ¼ 53 km). Compared with the two-layer model, the inferred radius of the core is almost unaffected by the inclusion of the crust, and the densities of the mantle and core change by less than 1%. This result can be explained by the small volume of the crust and the fact that its density is lower than that of the underlying layers. Consequently, the presence of the crust does not change the e and Cmþcr =C appreciably. values of ρ; C, Another possible three-layer model includes a solid inner core, a liquid outer core, and a mantle. However, the composition of the core is not well constrained, and the system of equations (4.24)–(4.26) cannot be solved. To make further progress, we build multi-layer models (Section 4.7) that include additional, indirect constraints from the observed composition of the surface (Section 4.6) and from assumptions about interior properties guided by laboratory experiments. We then incorporate constraints that arise from the measurement of planetary tides (Section 4.8).

4 .6 C O MP O S I TI O NA L C O N S T RAI N T S Measurements of the surface chemistry of Mercury by the MESSENGER spacecraft have provided important information on the composition of the interior (e.g., Chapter 2). Observations by the X-Ray Spectrometer (XRS) and GammaRay and Neutron Spectrometer (GRNS) instruments have demonstrated that Mercury’s surface has a low (150,000 K. Unfortunately, the uncertainties in the cross sections are unclear from Huebner et al. (1992). The Mg ionization rate also shows a discrepancy between the two databases used by Huebner and Mukherjee (2015), although it is only ~15%. Kinetic Escape An atom will escape from the gravity well of a planetary body when its velocity exceeds the escape velocity:

v2esc GMm ; ¼ rkT v2m


and is generally referred to as the escape parameter (essentially the gravitational potential energy in units of kT). It is useful to look at gravitational escape as a function of planetary mass, temperature, and mass of the atomic species as shown in Figure 15.6. Even for water, the fraction of released vapor that escapes at 3000 K is less than 25% on the initial trajectory. Therefore, for metals (Na, Mg), escape or permanent loss is expected to result from photoionization or some process that produces a very hot vapor, such as sputtering or the high-velocity tail of the impact vapor. For Na, radiation pressure enhances escape at some true anomaly angles, but radiation pressure is insignificant for Mg. Nevertheless escaping Mg has been observed at Mercury, which necessitates an energetic source process. The extreme temperature of the Ca exosphere (Burger et al., 2014) implies that the Ca is escaping; however, there may be an unobserved molecular component that condenses back to the surface (Killen, 2015).

15. 3 ME S S EN G E R O BS ER VA T IO NS AND MO DELS 15.3.1 Sodium The sodium tail – the component of the exosphere that is escaping anti-sunward primarily due to radiation-pressureinduced acceleration and first imaged by Potter et al. (2002) – was imaged by UVVS during each of MESSENGER’s three


Models of Mercury’s Exosphere Gravitational Escape Fraction, Test Mass = 18 AMU Ceres



0. 5 0. 2 0. 5 1 0. 01


0. 99

Source Temperature (K)

1.0 0.8


5 0. 2 0. 5 1 0. 01





0. 9



0.2 0.0

10 10



Gravitational Escape Fraction

Phobos 10000


Figure 15.6. (a) The gravitational escape fraction for a test mass of 18 AMU (color coded) as a function of source temperature and object radius, for a bulk density of 3 g cm-3. It can be seen that water is either completely lost or retained except for a narrow band as a function of source temperature and object size (or equivalently mass). (b) The source temperature (color coded) for 50% gravitational escape for species mass (AMU) versus object radius, for a bulk density of 3 g cm-3 (adapted from Killen et al., 2014). For Mercury, with a radius of 2439 km, the temperature for 50% gravitational escape probability is above that expected for impact vaporization temperatures (~3000 K) or photon-stimulated desorption (1200 K) for all species except H, H2, He, OH, and H2O.

Object Radius (km) Temperature of 50% Gravitational Escape Ceres Moon Earth






103 102










00 0









Species Mass (AMU)

106 80




100 1000 Object Radius (km)

50% Escape Temperature (K)

Phobos 100


Mercury flybys. The tail exhibited a nearly uniform north–south emission strength during the first and second flybys (Burger et al., 2010). Scaled to the same levels, the tail was practically non-existent during the third flyby. Although the radiation pressure for an atom at rest with respect to Mercury is near its peak for TAA 292° and only beginning to decrease for TAA 326° (see Figure 15.3), there is a negative feedback in the radiation pressure for true anomaly angles >180° such that the effective radiation pressure at TAA 326° is quite weak (e.g., Potter et al., 2007; Cassidy et al., 2015). UVVS observations of the sodium tail confirm the seasonal variation expected from the radiation pressure (Figure 15.7). McClintock et al. (2009) used a Monte Carlo model (see Burger et al., 2010) to demonstrate that the Na observed during the second MESSENGER flyby was not consistent with a spherically symmetric source of Na but instead showed a northern enhancement. On the third flyby there was no apparent northern enhancement in Na relative to east or west, and only a very slight excess of Na over the south pole relative to the north pole (Vervack et al., 2010). This pattern is in contrast to the large excesses in both Ca and Mg over the north pole relative to the south pole during the third flyby, indicating very different source processes. High-latitude enhancement in the Na surface content (Peplowski et al., 2014) must be considered in future simulations of the Na exosphere, especially for models at high latitudes for which the Na content of the regolith is double that at the equator.

The observations during the orbital phase of MESSENGER were of several general types, termed dayside limb scans, nightside tail sweeps, ride-alongs, and stares. Because of orbital constraints, most of the limb scans have tangent points near the equator or at low latitudes (Figure 14.9). Notable exceptions are the limb scans approximately perpendicular to the spin axis at the south pole. In spite of limited spatial coverage, the UVVS provided unprecedented spatial resolution and observation cadence for more than 16 Mercury years of near-daily observations. The UVVS observations analyzed by Cassidy et al. (2015) showed year-to-year repeatability: at a given local time and TAA the emissions were nearly identical from one Mercury year to the next. Cassidy et al. (2015) interpreted the UVVS Na limb scan data with a simple model to estimate the temperature and density of the near-surface exosphere within approximately 50–1500 km of the surface. The model accounts for the effects of radiation acceleration and includes single scattering with a uniform phase function. Cassidy et al. (2015) derived a temperature of ~1200 K with some local-time variation (Figure 15.8), including observations over the south polar region. They found that the tangent column density at 300 km above the subsolar point was a factor of 3 higher at aphelion than at perihelion. This difference may be accounted for by the fact that radiation pressure reduces the scale height at perihelion, so this measurement does not give a direct comparison with total abundance. Ground-based observations taken over

15.3 MESSENGER Observations and Models


Figure 15.7. Illustration of “seasonal” variations in Mercury’s exospheric emission. Observations of Mercury’s sodium tail were taken at the indicated true anomalies. Compare with Figure 15.3 for radiation pressure. Note that the shadow region antisunward of Mercury prevents resonant emission, so the “streamers” are not related to density. The emission scales as the inverse square of heliocentric distance, so the column abundance scales as the product of the emission and the squared heliocentric distance, a factor of 2.3 increase from aphelion to perihelion.

a period of seven years reported by Potter et al. (2007) showed on average about the same disk-averaged column abundance at aphelion and perihelion, but with a spread of a factor of 3 at aphelion and a much smaller spread at perihelion. These results suggest that photon-stimulated desorption is the primary source process. However, a high-energy process must also be present to produce both seasonal and episodically variable effects seen especially in ground-based data, and to populate the tail (Schmidt, 2012). Vaporization by micrometeoroid impact has been shown to be seasonally repeatable (Killen and Hahn, 2015) and thus represents a seasonal high-energy source that can potentially populate the tail. Impacts by larger meteoroids are rare, and the effects are short-lived (Mangano et al., 2007). Sputtering would produce a high-energy, episodically variable component, but observations of ion sputtering by the UVVS instrument were precluded by instrument noise produced by high-energy charged particles, especially during solar energetic particle (SEP) events. Given that impact vaporization is almost certainly the most important source of the calcium exosphere (see Section 15.3.2; Killen and Hahn, 2015), it must play a role in imparting Na to the exosphere as well. The Na content, as well as that of most of the other species seen in the exosphere, is spatially variable in the surface grains (Peplowski et al., 2014), ranging from 2.8 wt% at low northern latitudes (0°–15° N) to 4.9 wt% at high northern latitudes (80°–90°N). Given an average Na surface concentration of 2.8 wt%, and the impact vaporization rate from Cintala (1992), the Na impact

vaporization rate should be 1.0 × 106 cm−2 s−1 at aphelion and 2.4 × 106 cm−2 s−1 at perihelion, corresponding to global rates of 7.5 × 1023 s−1 at aphelion to 1.8 × 1024 s−1 at perihelion. These values represent 4% of the rate needed to populate the exosphere for a one-bounce model (i.e., the atoms stick on re-contact with the surface) or all of the required rate if the lifetime is the photoionization lifetime (i.e., the atoms do not stick on re-contact with the surface). Assumptions concerning the gas–surface interaction are therefore critical to conclusions concerning the importance of the various source processes. In addition, pre-MESENGER models all used lunar composition and their results must be scaled. For instance, Mouawad et al. (2011) assumed a Na fraction of 0.5 wt% (i.e., a lunar composition) in deriving a Na impact vaporization rate of 3.5 × 105 cm−2 s−1, but they gave an upper limit on the impact vaporization rate from observational data of 2.1 × 106 cm−2 s−1, consistent with the impact vaporization rate that would be derived if scaled to MESSENGER surface composition measurements. If the total PSD desorption rate is 3.5 × 1024 s−1 (Burger et al., 2010) and the total impact vaporization rate is 1.8 × 1024 s−1, then impact vaporization represents about 35% of the total ejection of Na at Mercury rather than the 2% estimated by Mouawad et al. (2011) on the basis of an assumed lunar composition. If this is the case, then some thermal accommodation must be taking place to maintain the exosphere at 1200–1400 K. A PSD desorption rate of 3.5 × 1024 s−1 was derived by Burger et al. (2010) from the PSD cross section of


Models of Mercury’s Exosphere

Figure 15.8. Examples of fits to near-equatorial dayside limb scans (local times indicated) and the south pole observations for Na (Cassidy et al., 2015). Only the low-altitude component of the profile was fit. Data are represented by crosses, and the fit is shown by a black dashed line. The blue line is a Chamberlain fit with the optical depth correction. The resulting temperature and surface density used for the fits are indicated by blue text. Observations were taken on 6 June 2012, except for the south pole observation, which took place on 17 October 2011.

Yakshinskiy and Madey (1999), 3 × 10−21 cm2, and a total surface number density of 7.5 × 1014 cm−2 for the first two MESSENGER flybys at a heliocentric distance of 0.35 AU. However, Burger et al. (2010) assumed a lunar abundance of Na, 0.5 wt%, and a surface density of 7.5 × 1014 cm−2. Scaling the PSD source rate to the measured Na wt% would imply a PSD desorption rate of ~1.9 × 1025 s−1. The simulation of Schmidt (2012) yielded a total source rate from impact vaporization in the range 1.8 × 106 to 3.6 × 106 cm−2 s−1 Na atoms in order to obtain the Na escape rate required to match the tail observations, given that the escape is due to an impact vaporization source. Because the Burger (2010) impact vaporization rate was derived from first principles for a lunar composition (Na wt% = 0.5), scaling that figure by a factor of 5.6 to the Na mass fraction observed by MESSENGER (Na wt% = 2.8) would give an impact vaporization rate of 2 × 106 Na atoms cm−2 s−1, consistent with the Schmidt (2012) rate derived by matching the observed Na escaping down the tail. Mura (2012) found that for a Weibull distribution of velocities (their “fast” distribution) for PSD, up to one-third of the sodium particles escape if the characteristic energy β = 0.086 eV, whereas for the Maxwellian case the loss rate is smaller by a factor of 3 but still accounts for 10% of the source.

Thus, the high-velocity portion of the PSD source distribution can populate Mercury’s tail if a high-energy tail to the distribution is assumed. Schmidt et al. (2012) also performed threedimensional time-dependent modeling of Mercury’s extended sodium tail, considering the effects of orbital motion, gas–surface interaction, variable source rates, and spatially nonuniform distributions of the sources (see Section 15.2.1). They concluded that either a combination of a slow impact vaporization source (3000 K) and a “fast” PSD source or a combination of a slow PSD source and a fast (5000 K) impact vaporization source can result in a ~20% loss of the released Na atoms down the tail, depending on orbital phase. They estimated that a loss rate of ~1024 Na atoms s−1 is required to populate the tail on average, and that a sputter source can supply at most 25% of this required rate except in exceptional circumstances. If the average loss rate of Na atoms is 3.5 ×1023 Na atoms s−1, as derived by Leblanc and Johnson (2003), this average represents a loss rate of 26% of the impact vaporization source. Schmidt et al. (2012) estimated that about 15% of the atoms derived from a slow impact vaporization source or 30% of those from a fast impact vaporization source would escape. Thus the derived loss rate is closer to the fast impact vaporization loss rate than the slow rate, but a combination of impact vaporization and PSD would also fit the observations.

15.3 MESSENGER Observations and Models


Figure 15.9. Simulated images of Ca emission in Mercury’s dayside equatorial plane at two locations in Mercury’s orbit were produced by interpolating between observations obtained at the white points approximately perpendicular to the plane of these images (Burger et al., 2014). In each image, dawn is to the left and the Sun is down. The white points represent positions where the UVVS line of sight crosses the equatorial plane; the altitude sampling is higher in the right panel. The images reflect large-scale local-time variations (small-scale variations in the images are not real). Although the magnitude of the emission varies with Mercury true anomaly, Ca is always brightest in the dawn hemisphere, usually, but not always, peaking at dawn.

The robust nature of the year-to-year repeatability and nearly constant near-surface exospheric temperature were unexpected. These findings are surprising in light of the published groundbased observations, which have suggested that sodium is ejected from the surface by a complex mixture of processes. However, because of the geometrical limitations of the UVVS limb scans, with a tangent point almost always at low latitudes on the dayside, a global model is precluded from these data.

15.3.2 Calcium Observations Prior to MESSENGER, ground-based observations identified Ca mainly over the poles and in the anti-sunward direction (Killen et al., 2005). Significant Ca emission (>2 kR) was seen up to two planetary radii behind the planet. The lines of sight for ground-based observations are limited in their ability to separate source regions in the east–west (i.e., dawn–dusk) direction, and scattering from the surface precluded observations near the dayside. During MESSENGER’s Mercury flybys, UVVS observations were able to isolate dawn from dusk, and it was seen that the source of Ca is highly concentrated in the dawn equatorial region (Figure 15.9) (McClintock et al., 2009; Vervack et al., 2010; Burger et al., 2012). The temperature was also determined to be extremely high on the basis of the scale height of Ca observed above Mercury’s poles and by the fact that Ca extends into the tail region despite the short photoionization lifetime (99% of the radiation acceleration is due to the 422.8 nm line). A linear interpolation between tabulated values was used to calculate the g-value at the heliocentric velocity of the atom. Atomic species were tracked by the integrator until they were lost by photoionization, gravitational escape outside the region of interest (~10 RM), or sticking to the surface. Charge exchange and electron-impact ionization do not contribute significantly to the loss because of the low plasma densities. Similarly, the only emission mechanism considered was resonant scattering of sunlight. However, the ability to include plasma effects on loss and emission was incorporated in the model. It was assumed that Ca atoms that return to the surface stick with 100% efficiency, although the sticking efficiency and thermal accommodation coefficients are adjustable parameters in the model (Burger et al., 2010; Mouawad et al., 2011). Burger et al. (2012, 2014) assumed Maxwellian flux distributions as a means of estimating the mean energy of the

outward-directed Ca. Burger et al. (2014) also demonstrated that Gaussian flux distributions are consistent with the data when the mean velocity is greater than Mercury’s escape velocity. This distribution approximates a dissociation source: dissociating Ca-bearing molecules give the fresh neutral Ca atom an energy boost from the excess energy of the dissociating photon, electron, or ion (e.g., Killen, 2015, 2016). If atomic Ca is produced from the dissociation of Ca-bearing molecules near the surface, the initial speed distribution will be approximately Gaussian, in the form 2

fv ¼ eðvvp Þ =2η ; 2


where vp is the most probable speed and η is the width of the distribution. Electron-stimulated desorption (ESD) has also been suggested as a possible source process for Mercury’s exosphere (McLain et al., 2011; Schriver et al., 2011). Burger et al. (2012, 2014) found that ESD is an unlikely Ca source, however, owing to the year-toyear stability of the Ca exosphere compared with the solar-wind interaction with Mercury’s magnetosphere and the lack of evidence for precipitating electrons at dawn. Killen (2016) concluded that electron impact dissociation (EID) cannot be responsible for the energization of the Ca due to the small cross section for EID. The atomic calcium in Mercury’s exosphere is quickly ionized by solar UV radiation. Burger et al. (2012, 2014) assumed that the photoionization lifetime varied between 23 min at perihelion and 52 min at aphelion, although there is a large uncertainty in these values, as discussed in Section Burger et al. (2012, 2014) took a data-fitting approach in their study of Ca observations by MASCS. The goal of these studies was to simulate the observed data from the MESSENGER flybys and orbital phase to determine the spatial and energy

15.3 MESSENGER Observations and Models distributions of the escaping Ca. Other authors (e.g., Killen and Hahn, 2015; Christou et al., 2015; Killen, 2015, 2016) considered the physical mechanisms responsible for producing the modeled source distributions, concentrating on impact vaporization. A process-driven procedure can be used to determine the expected distributions that MESSENGER would observe for a given process. Given that the high latitudes were not well observed by the UVVS instrument, the complete mix of Ca source mechanisms is currently somewhat uncertain. Ion sputtering or electronstimulated desorption were considered unimportant (Burger et al., 2012, 2014) because the magnetosphere is more variable than the data, and ion precipitation is not expected to peak at dawn (Kallio et al., 2008; Benna et al., 2010). Pfleger et al. (2015) computed three-dimensional exosphere models for sputtering by solar-wind H+ and He++ and concluded that solar-wind-sputtered Ca could provide a minor population with respect to the MESSENGER observations. The average dayside number density of Ca produced by solar-wind sputtering was estimated to be less than 1 cm−3, which would probably not be measured by the UVVS instrument but would not be insignificant relative to the 1–4 cm−3 peak Ca density reported by Burger et al. (2014). Pfleger et al. (2015) concluded that a dayside Ca surface density comparable to the long-term observations could be expected from extreme solar events. Unfortunately, the UVVS instrument either went into safe hold or was overwhelmed by noise during extreme solar wind events and could not measure the response of the exosphere to these solar events. The most promising hypothesis for the production of the Ca exosphere is that the primary source is micrometeoroid impact vaporization, which produces Ca-bearing molecules or ions that quickly dissociate to produce energetic Ca atoms (Killen, 2015). This idea was first proposed by Killen et al. (2005) and was considered by Burger et al. (2014). There are several features that make this proposal attractive. First, recent radar observations and models of micrometeoroids at Earth show that there is a strong dawn enhancement in the impactor flux (Janches et al., 2006; Pifko et al., 2013). Horanyi et al. (2015) reported a dawn– dusk asymmetry in dust at the Moon from LADEE dust observations. They indicated that the dust flux at dawn can be up to a factor of 6 larger than at dusk, in contrast to estimates of a factor of 3 dawn-to-dusk ratio of dust influx at Earth (Pifko et al., 2013). Calcium-bearing molecules are more likely to be produced in impact vapor plumes than atomic Ca (Berezhnoy and Klumov, 2008), and the molecules expected to be produced (CaO, CaOH, and/or Ca(OH)2, depending on the temperature of the vapor plume) quickly dissociate to release atomic Ca (Berezhnoy and Klumov, 2008; Berezhnoy, 2013). Killen (2015) estimated that the most likely mechanism for creating high-energy atomic Ca is dissociative ionization of precursor molecules. Killen and Hahn (2015) showed that the total Ca source rate is consistent with the dust flux expected as Mercury traverses the interplanetary dust disk and, in addition, intersects a cometary dust stream near true anomaly 25° (Figure 15.11), possibly associated with comet 2P/Encke (Christou et al., 2015). They did not estimate probable dawn–dusk asymmetries. The exospheric Ca source is unlikely to be adsorbed Cabearing material on the nightside that thermally desorbs as it enters sunlight, because the models of Burger et al. (2012)


required a pre-dawn Ca source and because the temperature required to vaporize Ca is over 3000 K. Pfleger et al. (2015) concluded that sputtering could be responsible for an important fraction of the Ca exosphere, slightly below detectability by the UVVS instrument. This finding might explain the Ca observed over Mercury’s poles in ground-based spectra (e.g., Killen et al., 2005) and possibly the enhancement of Ca over the poles during the third MESSENGER flyby (Vervack et al., 2010). 15.3.3 Magnesium The discovery measurements of magnesium (Mg) obtained during MESSENGER’s second Mercury flyby were modeled to constrain the source and loss processes for this species (Killen et al., 2010; Sarantos et al., 2011). These measurements provided a unique opportunity for constraining the portion of the exosphere produced by energetic processes, because they probed ejecta at very large downtail distances that could not be probed again after MESSENGER’s orbit insertion. The flyby data were matched to Chamberlain models (Killen et al., 2010; Sarantos et al., 2011) and non-uniform sputtering models (Sarantos et al., 2011) in which transport was computed from Liouville’s theorem (i.e., gravity forces were included but not radiation pressure or losses to photoionization). Sarantos et al. (2011) found that the distribution of magnesium with tail distance is suggestive of possibly two energetic ejection processes, because the superposition of a hot source (THOT ≥ 20,000 K) with a cooler source (TCOOL ≤ 5000 K) improved the description of emission detected in the near and far tail (Figure 15.12). The more energetic component populating the tail required higher rates by a factor of at least 5 than ion sputtering could provide under the solar-wind conditions prevailing during the flyby (Sarantos et al., 2011). This conclusion is subject to some uncertainty, however, both from modeling access of the solar wind to the surface and from the fact that the g-value was used for atoms at rest with respect to Mercury. Dayside limb scans and near-tail measurements during MESSENGER’s orbital phase provide evidence for a nonuniform Mg source. Analysis of individual limb scans with Chamberlain fits provides convincing evidence for a dawn–dusk asymmetry with peak dayside abundances at local times of 8–10 h (Merkel et al., 2017). Given only limb scan data, the necessity for two temperature components on the dayside is unclear, as these data can be fit with a single source at 5000 K on most days. The source of exospheric Mg appears to originate near dawn or in the mid-morning (Figure 15.13). The main conclusion from Sarantos et al. (2011) was that the total amount of magnesium at altitudes exceeding 100 km is consistent with predictions from impact vaporization models if micrometeoroid impacts eject both Mg atoms and Mg-bearing molecules (e.g., MgO, MgS) with molecular dissociation lifetimes of no more than 2 min (Figure 15.14). It is conceivable that reactions taking place during a fireball expansion result in at least half of the magnesium being bound in molecules (Berezhnoy and Klumov, 2008). Longer dissociation lifetimes would require a mix of hot atoms from dissociating molecules and fast atoms from sputtering. Observations taken at low


Models of Mercury’s Exosphere

RATEVAP (106 cm–2s–1)

1.5 Dust Disk Plus Comet Dust only: Omega = 290; i = 10 – green Dust + Stream nu = 25, sigma = 15 – red Data – black Vapor Fraction 0.055



0.0 –200




True Anomaly Angle

2 < |z| < 4

2 χ / ν = 5.5



Intensity (R)


1 < |z| < 2


n0 = 9.1 cm , T = 20,000 K χ2 / ν = 5.7






0 1



0 x105 10

Maxwellian T20000 N


Jul8 5

0 –50

S 0 –150 –100 –50



100 150

Longitude (º)

101 103 Intensity (R)

x105 3

Maxwellian T3000




Latitude (º)



Data n0 = 9.6 cm–3, T = 50,000 K

Latitude (º)

Intensity (R)


Radiance (R)

Tail Measurements


Radiance (R)

Figure 15.11. (Left) Derived source rate of Ca from the first nine Mercury years of UVVS observations plotted as a function of Mercury true anomaly angle, along with the modeled hot component of impact vaporization by the flux from interplanetary dust disk onto Mercury (green) with the addition of dust from the comet 2P/Encke centered near 25° (red) (Killen and Hahn, 2015). Away from 0–70° TAA the red line follows the dust-disk-only curve. (Right) Artist’s conception of comet dust impacting Mercury (not to scale).

|z| < 0.5

n0 = 55.3 cm–3, T = 3000 K χ2 / ν = 18










Figure 15.13. The derived Mg flux from the surface (atoms cm−2 s−1) on 8 July 2012, under the assumption of a Maxwellian velocity distribution for released ejecta. The data are consistent with a warm component at 3000 K covering the dawn hemisphere and a hot component at about 20,000 K in the morning quadrant. The white dot denotes the subsolar point, and the dawn terminator is at −90° longitude (vertical dotted line to the left). Other fits are possible, including a single source at 5000 K. Adapted from Sarantos et al. (2012).

Distance Along Tail (RM)

Figure 15.12. Magnesium 285-nm cross-tail line-of-sight radiances (red circles) measured down the tail to 14 Mercury radii (RM) are compared at vertical distances from the equator of (a) (2–4) RM (b) (1–2) RM, and (c) (−0.5–0.5) RM. Models based on source temperatures 50,000 K (magenta) and 20,000 K (blue) are indistinguishable, and both fit the data. The model prediction for a 3000 K source lies well below the data for all vertical distances. From Sarantos et al. (2011).

spacecraft altitudes just nightward of the dawn terminator region and before crossing to the dayside could be interpreted as indicative of thermally accommodated particles (Sarantos et al., 2011). Data from the third MESSENGER flyby, which could provide equally important constraints for hot processes, have not been modeled. Both Mg and Ca exhibit a very hot component, but that for Ca (~70,000 K) (Burger et al., 2014) is

more energetic than that for Mg (~20,000 K). However, this temperature was derived under the assumption of a short photoionization lifetime. Both Ca and Mg require some process in addition to impact vaporization to provide the additional energy, and their observed energies are consistent with the dissociation of a precursor molecule (e.g., Killen, 2015) or sputtering. Presumably a sputtered component should be seen at high latitudes (Pfleger et al., 2015), in contrast to the observed distributions peaking at the equator (Sarantos et al., 2012). There is no evidence for a 3000–5000 K component for the Ca exosphere as there is for Mg. Sarantos et al. (2012) fit limb scans and tail data from the first three Mercury years after MESSENGER’s orbit insertion by employing a Monte Carlo model that included the effects of photoionization, radiation pressure, and velocity-dependent

15.3 MESSENGER Observations and Models Model Sputtering

Model Impact Vaporization

Model MgO Photolysis

1.0 1.25 1.5 1.75 2.0

Fantail Measurements







Angle Relative to Equatorial Dawn (º)

g-values. They used combinations of ejection processes with Maxwellian and sputtered initial distributions. The threedimensional output from each tested source process was discretized in a way that enabled searches for non-uniform ejection in a systematic manner. A Monte Carlo method was used to eject test particles within each of 2000 surface patches for each assumed source temperature. These particles, which were traced until photoionization or contact with the surface, represent the mapping of a “unit flux” moving from one surface element into the three-dimensional (3D) volume. Output from the modeled surface elements was saved separately, and a linear combination of the modeled intensities from each was fit to the emission data using a penalized least-squares regression method. The retrieved unknown variables are the fluxes leaving each surface patch. Such a method was used to estimate the best spatial release pattern that fit the data each day under the assumption of a given velocity distribution function for released ejecta.


2 1 0 –1


20 10 0 –10

–2 –3

Figure 15.14. Fantail measurements of the Mg emission during the second MESSENGER flyby (McClintock et al., 2009) are best fit with a model of MgO dissociation (Sarantos et al., 2011). The addition of smaller contributions from impact vaporization and sputtering yields a better total fit to the details of the observations (red circles with error bars).

Ca+ Rayleighs/nm



log10 Radiance (R)


Mercury Radii

Intensity (R)



Model Sum




–1 –2 –3 –4

Mercury Radii

–20 392




Wavelength (nm)

Figure 15.15. (Left) This pseudo-image of the Ca+ emission detected during the third MESSENGER flyby was generated by projecting the observed column emissions onto a plane containing the Sun–Mercury line and the planet’s spin axis and interpolating to fill in unobserved regions. (Right) The red spectrum represents the average of all the Ca+ emission-line observations (one-standard-deviation error bars are shown) during the third flyby; the green line is a Gaussian fit to the average Ca+ line. Adapted from Vervack et al. (2010).

the same altitude (1630 km above the limb) as the UVVS observation. The lifetime of Ca against photoionization is short, ~1500 s (Huebner and Mukherjee, 2015), so most exospheric Ca+ ions are formed within 2.5 RM and the Ca+ pickup process occurs within Mercury’s magnetosphere. The Ca+ produced over a relatively large volume might be concentrated by magnetospheric convection into the near-tail equatorial region. Vervack et al. (2010) estimated that although Ca+ would fill the entire width of the magnetotail, only about 65% of the Ca+ was in sunlight where it would have been observable by the UVVS. 15.3.5 Weakly Emitting or Less Abundant Species

15.3.4 Ionized Calcium Ionized calcium (Ca+) was observed close to the equatorial plane in a relatively small region approximately (1–2) RM tailward of the planet on the third MESSENGER flyby (Vervack et al., 2010), as shown in Figure 15.15. Similar line-of-sight column densities for Ca+ and Ca were measured, an unlikely phenomenon if Ca+ is produced locally by ionization of Ca because of large differences in their respective observed velocities (Ca, several km s−1; Ca+ up to hundreds of km s−1). The observations occurred tailward of the magnetospheric X-line while the spacecraft was inbound and lend support for the view that a magnetospheric convection pattern led to a concentration of Ca+ in the region behind the X-line before the Ca+ ions were ejected down the tail. UVVS intensities compared favorably with those reported by Bida and Killen (2017) from observations at the Keck 1 telescope of a region (1.7–2.0) RM from planet center (Section This observation corresponds to about a factor of 5 greater than the three-standard-deviation abundance measured on 15 May 2008 and 3 May 2009 (Bida and Killen, 2011, 2017). Bida and Killen (2017) obtained an upper limit to the Ca+ column equal to 3.9 × 106 cm−2 at approximately

Although it was expected that emission from several weaker or less abundant species such as Al, Fe, and Ca+ would be detected with some regularity in Mercury’s exosphere, such an expectation was not met. The MESSENGER UVVS observations in particular are interesting in that no emission from Al, Ca+, or Mn was detected during the orbital phase until late in the mission (the final Earth year in orbit) (Vervack et al., 2011, 2015, 2016). These species have been elusive both from the ground and from MESSENGER observations, pointing to a high level of variability. Even more interesting is that all of the detections of these species have been confined to a particular spatial and temporal pattern. Spatially, all the detections of these three species by UVVS during the orbital phase of MESSENGER occurred in the pre-dawn nightside region (local times around 2–5 h). Also, they were detected only during the outbound leg of Mercury’s orbit, between true anomaly angles of 0° and 70° (Vervack et al., 2016). Aluminum was detected from groundbased observations with the Keck I telescope at TAA 103° and 117°, farther from perihelion but still on the outbound leg of the orbit (Bida and Killen, 2017).


Models of Mercury’s Exosphere

The location and timing of these detections is highly suggestive of a connection to the comet 2P/Encke dust stream that was proposed by Killen and Hahn (2015) to explain the spike in the Ca emission over similar true anomaly angles (see Section Dust from Encke impacting the dawn side of the planet may have led to an enhancement in these species to levels that were detectable by UVVS. It is also possible that some of the material may be cometary in origin. The marked difference in the observed profile for Mn (Figure 14.45) compared with those for Al and Ca+ may indicate that Mn perhaps derives from the cometary dust rather than the surface of Mercury itself. The Encke stream may also explain why these species were not detected by UVVS earlier in the mission, despite observations at the same general location and time. Due to the ecccentricity of Mercury’s orbit, Mercury’s terminator moves backward (toward the nightside) from a TAA of about 340° to TAA 20°. A delayed source as a result of this rock-back of the terminator pre-perihelion would exhibit the same source rate every Mercury year, counter to the observations. Comet streams at Earth are responsible for meteor showers, and occasionally there are larger “clumps” of dust that lead to more spectacular meteor storms. Such behavior is not uncommon with the Taurids at Earth, which also derive from comet Encke (Jenniskens, 2006). A similar change in the amount of dust that is impacting Mercury in a given year (or years) could account for the onset of detection of these species by UVVS. Furthermore, it may not be coincidental that Encke had a very close passage to Mercury in 2013, and these detections began three Mercury years later.

1 5. 4 C ON CLUS I ONS AND UN AN SWERED QUESTIONS Observations of exospheric Ca and Mg show that their respective exospheres are very hot, much hotter than can be explained by impact vaporization. Although sputtering can certainly produce vapor that is as hot as that measured for these species, Mercury’s magnetosphere generally shields the surface from direct penetration by solar wind except at the cusps (Raines et al., 2013). It is unlikely, on the grounds of both measured sputtering yields and inferred ion fluxes to the surface, that the Ca and Mg exospheres are wholly produced by a sputter source. Killen et al. (2005) suggested that the hot vapor could be produced by dissociation of a molecular precursor. Subsequently Berezhnoy (2013) calculated the equilibrium fraction of various atomic and molecular species produced by impact vaporization and concluded that CaO or Ca(OH)2 would be the most likely calciumbearing species in the fireball at the quenching temperatures of greater or less than ~3750 K, respectively. Killen (2015) looked at the likely energy of various dissociation mechanisms and concluded that dissociative ionization would be the most likely candidate to produce escaping Ca. There has been no work on the production of hot Mg, which appears to have either a mixture of 3000 K and 20,000 K components, or a single 5000 K component (Sarantos et al., 2012;

Merkel et al., 2017). Further analysis of the UVVS Mg observations is required to determine the true energy distribution for Mg and the likely source processes. Further work is also needed to determine the effect of high-mass and highenergy ions that may penetrate to the surface more widely. The enhanced abundance of Ca seen following perihelion, at a TAA of about 25°–30°, has been attributed to enhanced impact vaporization by a meteoroid shower due to comet 2P/Encke, which is known to cross Mercury’s orbit (Killen and Hahn, 2015; Christou et al., 2015). Observations of other species, such as Al, Mn, and Ca+, may also be enhanced by this cometary dust stream’s impact on Mercury. The primarily dawnward location of the Mg source suggests that hot Mg atoms could be produced by the same physical process as the hot Ca. In order for micrometeoroid impacts to be responsible for producing gaseous Mg and Ca around Mercury, models of micrometeoroid precipitation onto Mercury’s surface must account for the production of molecules with a dawn–dusk asymmetry. It is likely that the Mg and Ca sources centered near dawn are correlated with impacting dust peaking in Mercury’s ram direction, as seen at the Moon (Szalay and Horanyi, 2016). Because the ram direction moves with TAA (Figure 15.16), the dawn sources may shift to slightly pre- or post-dawn locations over the course of a Mercury year. The combined source fluxes of Mg and Ca inferred from orbital phase data locally approach 2 × 106 atoms cm−2 s−1, consistent with the flyby results, and may be provided by impacts as previously surmised. Despite targeting observations at wavelengths near the oxygen 130.4 nm triplet for over 16 Mercury years, no detection was found in the UVVS observations. However, the analysis is difficult owing to scattered solar O emission from the dayside. If the impact vapor produces oxides rather than atomic oxygen and if the oxide is dissociated, an extremely hot O corona would be produced. A very tenuous oxygen corona would by its nature be difficult to observe. However, MESSENGER’s Fast Imaging Plasma Spectrometer observed a group of ions, including possible constituents O, OH, and H2O. If a mass spectrometer with higher mass resolution were flown on a future mission, then it would be possible to separate these components and to determine the true oxygen abundance at the spacecraft altitude. UVVS observations of sodium did not uncover conclusive evidence for a sputtered component, in contrast to expectations from ground-based observations (e.g., Potter et al., 2006; Mangano et al., 2015; Pfleger et al., 2015). One possible explanation is that most of the data over the dayside, and almost all of the data that were analyzed by Cassidy et al. (2015), were taken at low latitudes. UVVS limb scans were most often taken tangent to the equatorial limb while looking from south to north; thus, they would have missed a highlatitude source near the cusp region for the most part. Future observations that fill this gap in coverage would be worthwhile. High northern latitudes should be targeted in the future to determine whether meteoroid showers expected to result from comets Bradfield (at TAA 130°) and Tempel–Tuttle lead to enhancements in the region of impact. Concurrent measurement of the dust flux onto Mercury, and especially its spatial distribution with respect to leading and trailing hemispheres and expected meteoroid streams, is highly desirable.



similar to that proposed for K by Peplowski et al. (2012), who found a correlation between the distribution of measured surficial K abundances and models of the maximum temperature. Cassidy et al. (2016), on the other hand, found a correlation between Na abundance in the dawn exosphere and the equatorial temperature at Mercury’s hot and cold poles. Other species, notably Mg and Ca, have highly heterogeneous distributions of surface composition (Weider et al., 2015).

REFER E NC ES Figure 15.16. The ram point (local time of the orbital velocity vector) moves as a function of true anomaly angle because of the non-uniform relationship between orbital velocity and rotational velocity as a result of Mercury’s eccentric orbit.

Mercury’s surface composition is highly non-uniform (Weider et al., 2015). A correlation has been found between the local surface composition and the Mg exosphere (Merkel et al., 2018). Further investigations should seek correlations between the exosphere and underlying surface, especially between regions for which there is a large difference in composition. Although there have been ground-based detections of potassium in Mercury’s exosphere from observations of the strong 766.3 nm D2 line, this wavelength was outside the range of UVVS. Searches for K with UVVS had to use the weaker lines at 404.5 and 404.8 nm (with a combined g-value of 5. The large variations in the measured ratios are not surprising in light of the large variation in the surficial K abundance, reported by Peplowski et al. (2012) to be between 300 and 2400 ppm, whereas the Na wt% varies by only about a factor of ~2 from 2.6 wt% in the equatorial regions to 5 wt% at high northern latitudes (Peplowski et al., 2014). A factor of approximately 2.3 was found between the surface Na/Si weight ratios in the northern polar and equatorial regions, respectively, from MESSENGER Gamma-Ray Spectrometer (GRS) data (Peplowski et al., 2014). Peplowski et al. sought to determine whether the sodium could have been redistributed through desorption in equatorial regions and redeposition and cold trapping near the poles. If redistribution is not an important process, then the composition of the polar regions may represent the original Na abundance in that region. Invoking the assumption that the Na content is lowest in regions that experience maximum nearsurface temperatures above 400 K and from temperature maps at the surface and at a depth of 7 cm (Vasavada et al., 1999), Peplowski et al. (2014) concluded that the measured Na distribution supports the hypothesis that the equatorial Na abundances are consistent with depletion via thermal modification

Benna, M., Anderson, B. J., Baker, D. N., Boardsen, S. A., Gloeckler, G., Gold, R. E., Ho, G. C., Killen, R. M., Korth, H., Krimigis, S. M., Purucker, M. E., McNutt, R. L., Raines, J. M., McClintock, W. E., Sarantos, M., Slavin, J. A., Solomon, S. C. and Zurbuchen, T. H. (2010). Modeling of the magnetosphere of Mercury at the time of the first MESSENGER flyby, Icarus, 209, 3–10, doi:10.1016/j.icarus.2009.11.036. Berezhnoy, A. A. (2013). Chemistry of impact events on the Moon. Icarus, 226, 205–211, doi:10.1016/j.icarus.2013.05.030. Berezhnoy, A. A. and Klumov, B. A. (2008). Impacts as sources of the exosphere on Mercury. Icarus, 195, 511–522, doi:10.1016/j. icarus.2008.01.005. Bida, T. A. and Killen, R. M. (2011). Observations of Al, Fe and Ca+ in Mercury’s exosphere. EPSC-DPS Joint Meeting Abstracts and Program, 6, EPSC-DPS2011-1621. European Planetary Science Congress–Division for Planetary Sciences Joint Meeting, Nantes, France, October 2–7. Bida, T. A. and Killen, R. M. (2017). Observations of the minor species Al and Fe in Mercury’s exosphere. Icarus, 289, 227–238, doi:10.1016/j.icarus.2016.10.019. Bida, T. A., Killen, R. M. and Morgan, T. H. (2000). Discovery of calcium in Mercury’s atmosphere. Nature, 404, 159–161. Borin, P., Bruno, M., Cremonese, G. and Marzari, F. (2010). Estimate of the neutral atoms’ contribution to the Mercury exosphere caused by a new flux of micrometeoroids. Astron. Astrophys., 517, A89, doi:10.1051/0004-6361/201014312. Burger, M. H., Killen, R. M., Vervack, R. J., Jr., Bradley, E. T., McClintock, W. E., Sarantos, M., Benna, M. and Mouawad, N. (2010). Monte Carlo modeling of sodium in Mercury’s exosphere during the first two MESSENGER flybys. Icarus, 209, 63–74. Burger, M. H., Killen, R. M., McClintock, W. E., Vervack, R. J., Jr., Merkel, A. W., Sprague, A. L. and Sarantos, M. (2012). Modeling MESSENGER observations of calcium in Mercury’s exosphere. J. Geophys. Res., 117, E0L11B, doi:10.1029/ 2012JE004158. Burger, M. H., Killen, R. M., McClintock, W. E., Merkel, A. W., Vervack, R. J., Jr., Cassidy, T. A. and Sarantos, M. (2014). Seasonal variations in Mercury’s dayside calcium exosphere. Icarus, 238, 51–58, doi:10.1016/j.icarus.2014.04.049. Cassidy, T. A. and Johnson, R. E. (2005). Monte Carlo model of sputtering and other ejection processes within a regolith. Icarus, 176, 499–507. Cassidy, T. A., Merkel, A. W., Burger, M. H., Sarantos, M., Killen, R. M., McClintock, W. E. and Vervack, R. J., Jr. (2015). Mercury’s seasonal sodium exosphere: MESSENGER orbital observations. Icarus, 248, 547–559, doi:10.1016/j.icarus.2014.10.037. Cassidy, T. A., McClintock, W. E., Killen, R. M., Sarantos, M., Merkel, A. W., Vervack, R. J., Jr. and Burger, M. H. (2016). A cold-pole enhancement in Mercury’s sodium exosphere. Geophys. Res. Lett., 43, 11,121–11,128.


Models of Mercury’s Exosphere

Christou, A. A., Killen, R. M. and Burger, M. H. (2015). The meteoroid stream of comet Encke at Mercury: Implications for MErcury Surface, Space ENvironment, GEochemistry, and Ranging observations of the exosphere. Geophys. Res. Lett., 42, 7311–7318, doi:10.1002/2015GL065361. Cintala, M. J. (1992). Impact-induced thermal effects in the lunar and Mercurian regoliths. J. Geophys. Res., 97, 947–973. Combi, M. R., DiSanti, M. A. and Fink, U. (1997). The spatial distribution of gaseous atomic sodium in the comae of comets: Evidence for direct nucleus and extended plasma sources. Icarus, 130, 336–354, doi:10.1006/icar.1997.5832. Cremonese, G. and Verani, S. (1997). High resolution observations of the sodium emission from the Moon. Adv. Space Res., 19, 1561–1569, doi:10.1016/S0273-1177(97)00369-4. Crider, D. H. and Vondrak, R. R. (2003). Space weathering effects on lunar cold trap deposits. J. Geophys. Res., 108 (E7), 5079, doi:10.1029/2002JE002030. Domingue, D. L., Chapman, C. R., Killen, R. M., Zurbuchen, T. H., Gilbert, J. A., Sarantos, M., Benna, M., Slavin, J. A., Schriver, D., Trávníček, P. M., Orlando, T. M., Sprague, A. L., Blewett, D. T., Gillis-Davis, J. J., Feldman, W. C., Lawrence, D. J., Ho, G. C., Ebel, D. S., Nittler, L. R., Vilas, F., Pieters, C. M., Solomon, S. C., Johnson, C. L., Winslow, R. M., Helbert, J., Peplowski, P. N., Weider, S. Z., Mouawad, N., Izenberg, N. R. and McClintock, W. E. (2014). Mercury’s weather-beaten surface: Understanding Mercury in the context of lunar and asteroidal space weathering studies. Space Sci. Rev., 181, 121–214, doi:10.1007/ s11214-014-0039-5. Doressoundiram, A., Leblanc, F., Foellmi, C., Gicquel, A., Cremonese, G., Donati, J.-F. and Veillet, C. (2010). Spatial variations of the sodium/potassium ratio in Mercury’s exosphere uncovered by high-resolution spectroscopy. Icarus, 207, 1–8, doi:10.1016/j.icarus.2009.11.020. Dzioba, S. and Kelly, R. (1980). On the kinetic energies of sputtered excited particles II. Theory and applications to group IIA fluorides. Surf. Sci., 100, 119–134. Farrell, W. M., Hurley, D. M. and Zimmerman, M. I. (2015). Solar wind implantation into lunar regolith: Hydrogen retention in a surface with defects. Icarus, 255, 116–126, doi:10.1016/j. icarus.2014.09.014. Farrell, W. M., Stubbs, T. J., Vondrak, R. R., Delory, G. T. and Halekas, J. S. (2007). Complex electric fields near the lunar terminator: The near-surface wake and accelerated dust. Geophys. Res. Lett., 34, L14201, doi:10.1029/2007GL029312. Gerasimov, M. V., Dikov, Yu. P., Yakolev, O. I., Wlotzka, F. and Distler, V. V. (1999). The volatility of microelements during impact vaporization of meteorites. Meteorit. Planet. Sci., 34, Suppl., A42. Hofer, W. O. (1991). Angular, energy, and mass distribution of sputtered particles. In Sputtering by Particle Bombardment III, ed. R. Behrisch and K. Wittmaack, Topics in Applied Physics, Vol. 64. Berlin: Springer-Verlag, pp. 15–90. Holmlid, L. (2006). The alkali metal atmospheres on the Moon and Mercury: Explaining the stable exospheres by heavy Rydberg matter clusters. Planet. Space Sci., 54, 101–112, doi:10.1016/j. pss.2005.10.005. Horanyi, M., Szalay, J. R., Kempf, S., Schmidt, J., Grun, E., Srama, R. and Sternovsky, Z. (2015). A permanent, asymmetric dust cloud around the Moon. Nature, 522, 324–326, doi:10.1038/nature14479. Horanyi, M., Szalay, J. R., Grun, E., Glenar, D., Wang, X. and Zakharov, A. (2016). The dust environment of the Moon. New Views of the Moon, abstract 6005. Houston, TX: Lunar and Planetary Institute.

Hornung, K., Malama, Y. G. and Kestenboim, K. S. (2000). Impact vaporization and ionization of cosmic dust particles. Astrophys. Space Sci., 274, 355–363. Huebner, W. F. and Mukherjee, J. (2015). Photoionization and photodissociation rates in solar and blackbody radiation fields. Planet. Space Sci., 106, 11–45, doi:10.1016/j.pss.2014.11.022. Huebner, W. F., Keady, J. J. and Lyon, S. P. (1992). Solar photo rates for planetary atmospheres and atmospheric pollutants. Astrophys. Space Sci., 195, 1–294. Hunten, D. M. and Sprague, A. L. (1997). Origin and character of the lunar and mercurian atmospheres. Adv. Space Res., 19, 1551–1560, doi:10.1111/j.1945-5100.2002.tb00888.x. Hunten, D. M. and Sprague, A. L. (2002). Diurnal variation of sodium and potassium at Mercury. Meteorit. Planet. Sci., 37, 1191–1195, doi:10.1111/j.1945-5100.2002.tb00888.x. Hunten, D. M., Morgan, T. H. and Shemansky, D. E. (1988). The Mercury atmosphere. In Mercury, ed. F. Vilas, C. R. Chapman and M. S. Matthews. Tucson, AZ: University of Arizona Press, pp. 562–612. Ip, W.-H. (1986). The sodium exosphere and magnetosphere of Mercury. Geophys. Res. Lett., 13, 423–426. Janches, D., Heinselman, C. J., Chau, J. L., Chandran, A. and Woodman, R. (2006). Modeling the global micrometeor input function in the upper atmosphere observed by high power and large aperture radars. J. Geophys. Res., 111, A07317, doi:10.1029/ 2006JA011628. Jenniskens, P. (2006). Meteor Showers and Their Parent Comets. Cambridge: Cambridge University Press, 752 pp. Johnson, R. E. (1990). Energetic Charged-Particle Interactions with Atmospheres and Surfaces. Berlin: Springer. Johnson, R. E., Leblanc, F., Yakshinskiy, B. V. and Madey, T. E. (2002). Energy distributions for desorption of sodium and potassium from ice: The Na/K ratio at Europa. Icarus, 156, 136–142. Jurac, S., Baragiola, R. A., Johnson, R. E. and Sittler, E. C. (1995). Charging of ice grains by low-energy plasmas: Application to Saturn’s E-ring. J. Geophys. Res., 100, 14,821–14,831. Kallio, E, Wurz, P., Killen, R. M., McKenna-Lawlor, S., Milillo, A., Mura, A., Massetti, S., Orsini, S., Lammer, H., Janhunen, P. and Ip, W.-H. (2008). On the impact of multiply charged heavy solar wind ions on the surface of Mercury, the Moon and Ceres. Planet. Space Sci., 56, 1506–1516. Keller, L. P. and McKay, D. S. (1997). The nature and origin of rims on lunar soil grains. Geochim. Cosmochim. Acta, 61, 2331–2341, doi:10.1016/S0016-7037(97)00085-9. Killen, R. M. (2015). Processes producing the extremely hot Ca and Mg exospheres at Mercury. Presented at 2015 Fall Meeting, American Geophysical Union, abstract P53A–2090, San Francisco, CA, 14–18 December. Killen, R. M. (2016). Pathways for energization of Ca and Mg in Mercury’s exosphere. Icarus, 268, 32–36, doi:10.1016/j. icarus.2015.12.035. Killen, R. M. and Hahn, J. M. (2015). Impact vaporization as a possible source of Mercury’s calcium exosphere. Icarus, 250, 230–237, doi:10.1016/j.icarus.2014.11.035. Killen, R. M., Sarantos, M., Potter, A. E. and Reiff, P. (2004). Source rates and ion recycling rates for Na and K in Mercury’s atmosphere. Icarus, 171, 1–19. Killen, R. M., Bida, T. A. and Morgan, T. H. (2005). The calcium exosphere of Mercury, Icarus, 173, 300–311. Killen, R. M., Cremonese, G., Lammer, H., Orsini, S., Potter, A. E., Sprague, A. L., Wurz, P., Khodachenko, M. L., Lichtenegger, H. I. M., Milillo, A. and Mura, A. (2007).

References Processes that promote and deplete the exosphere of Mercury. Space Sci. Rev., 132, 433–509, doi:10.1007/s11214-007-9232-0. Killen, R. M., Shemansky, D. E. and Mouawad, N. (2009). Expected emission from Mercury’s exospheric species, and their UVvisible signatures. Astrophys. J. Suppl. Ser., 181, 351–359. Killen, R. M., Potter, A. E., Vervack, R. J., Jr., Bradley, E. T., McClintock, W. E., Anderson, C. M. and Burger, M. H. (2010). Observations of metallic species in Mercury’s exosphere. Icarus, 209, 75–87, doi:10.1016/j.icarus.2010.02.018. Killen, R. M., Burger, M. H., Hurley, D. M., Sarantos, M. and Farrell, W. M. (2014). Exospheres from asteroids to planets. In Asteroids, Comets, Meteors: Book of Abstracts, Helsinki, Finland, 2014. Helsinki: University of Helsinki, p. 285. Kurosawa, K., Sugita, S., Kadono, T., Shigemori, K., Hironaka, Y., Ozaki, N., Shiroshita, A., Cho, Y., Sakaiya, T., Fujioka, S., Tachibana, S., Vinci, T., Kodama, R. and Matsui, T. (2010). Roles of shock-induced ionization due to > 10 km/s impacts on evolution of silicate vapor clouds. Lunar Planet. Sci., 41, abstract 1785. Leblanc, F. and Johnson, R. E. (2003). Mercury’s sodium exosphere. Icarus, 164, 261–281, doi:10.1016/S0019-1035(03)00147-7. Leblanc, F. and Johnson, R. E. (2010). Mercury exosphere I. Global circulation model of its sodium component. Icarus, 209, 280–300. Leblanc, F., Delcourt, D. and Johnson, R. E. (2003). Mercury’s sodium exosphere: Magnetospheric ion recycling. J. Geophys. Res., 108 (E12), 5136, doi:10.1029/2003JE002151. Madey, T. E., Yakshinskiy, B. V. and Ageev, V. N. (1998). Desorption of alkali atoms and ions from oxide surfaces: Relevance to origins of Na and K in atmospheres of Mercury and the Moon. J. Geophys. Res., 103, 5873–5887. Madey, T. E., Johnson, R. E. and Orlando, T. M. (2002). Far-out surface science: Radiation-induced surface processes in the Solar System. Surf. Sci., 500, 838–858. Mangano, V., Milillo, A., Mura, A., Orsini, S., DeAngelis, E., DiLellis, A. M. and Wurz, P. (2007). The contribution of impulsive meteoritic impact vapourization to the Hermean exosphere. Planet. Space Sci., 55, 1541–1556. Mangano, V., Massetti, S., Milillo, A., Plainaki, C., Orsini, S., Rispoli, R. and Leblanc, F. (2015). THEMIS Na exosphere observations of Mercury and their correlation with in-situ magnetic field measurements by MESSENGER. Planet. Space Sci., 115, 102–109, doi:10.1016/j.pss.2015.04.001. Marchi, S., Morbidelli, A. and Cremonese, G. (2005). Flux of meteoroid impacts on Mercury. Astron. Astrophys., 431, 1123–1127. McClintock, W. E., Vervack, R. J., Jr., Bradley, E. T., Killen, R. M., Mouawad, N., Sprague, A. L., Burger, M. H., Solomon, S. C. and Izenberg, N. R. (2009). Mercury’s exosphere during MESSENGER’s second flyby: Detection of magnesium and distinct distributions of neutral species. Science, 324, 610–613. McCord, T. B., Taylor, L. A., Combe, J.-P., Kramer, G., Pieters, C. M., Sunshine, J. M. and Clark, R. N. (2011). Sources and physical processes responsible for OH/H2O in the lunar soil as revealed by the Moon Mineralogy Mapper (M3). J. Geophys. Res., 116, E00G05, doi:10.1029/2010JE003711. McGrath, M. A., Johnson, R. E. and Lanzerotti, L. J. (1986). Sputtering of sodium on the planet Mercury. Nature, 323, 694–696. McLain, J. L., Sprague, A. L., Grieves, G. A., Schriver, D., Trávníček, P. M. and Orlando, T. M. (2011). Electron-stimulated desorption of silicates: A potential source for ions in Mercury’s


space environment. J. Geophys. Res., 116, E03007, doi:10.1029/ 2010JE003714. Merkel, A. W., Cassidy, T. A., Vervack, R. J., Jr., McClintock, W. E., Sarantos, M., Burger, M. H. and Killen, R. M. (2017). Seasonal variations of Mercury’s magnesium dayside exosphere from MESSENGER observations. Icarus, 281, 46–54, doi:10.1016/j. icarus.2016.08.032. Merkel, A. W., Vervack, R. J., Jr., Killen, R. M., Cassidy, T. A., McClintock, W. E., Nittler, L. R. and Burger, M. H. (2018). Evidence connecting Mercury’s magnesium exosphere to its magnesium-rich surface terrane. Geophys. Res. Lett., 45, 6790–6797, doi:10.1029/2018GL078407. Morgan, T. H., Zook, H. and Potter, A. E. (1988). Impact-driven supply of sodium and potassium to the atmosphere of Mercury. Icarus, 75, 156–170. Mouawad, N., Burger, M. H., Killen, R. M., Potter, A. E., McClintock, W. E., Vervack, R. J., Bradley, E. T., Benna, M. and Naidu, S. (2011). Constraints on Mercury’s Na exosphere: Combined MESSENGER and ground-based data. Icarus, 211, 21–36, doi:10.1016/j.icarus.2010.10.019. Mura, A. (2012). Loss rates and time scales for sodium at Mercury. Planet. Space Sci., 63–64, 2–7, doi:10.1016/j. pss.2011.08.012. Mura, A., Milillo, A., Orsini, S. and Massetti, S. (2007). Numerical and analytical model of Mercury’s exosphere: Dependence on surface and external conditions. Planet. Space Sci., 55, 1569–1583. Mura, A., Wurz, P., Lichtenegger, H. I. M., Schleicher, H., Lammer, H., Delcourt, D., Milillo, A., Orsini, S., Massetti, S. and Khodachenko, M. L. (2009). The sodium exosphere of Mercury: Comparison between observations during Mercury’s transit and model results. Icarus, 200, 1–11, doi:10.1016/j. icarus.2008.11.014 Noble, S. K., Keller, L. P. and Pieters, C. (2005). Evidence of space weathering in regolith breccias I: Lunar regolith breccias. Meteorit. Planet. Sci., 40, 397–408. Peplowski, P. N., Lawrence, D. J., Rhodes, E. A., Sprague, A. L., McCoy, T. J., Denevi, B. W., Evans, L. G., Head, J. W., Nittler, L. R., Solomon, S. C., Stockstill-Cahill, K. R. and Weider, S. Z. (2012). Variations in the abundances of potassium and thorium on the surface of Mercury: Results from the MESSENGER Gamma-Ray Spectrometer. J. Geophys. Res., 117, E00L04, doi:10.1029/2012JE004141. Peplowski, P. N., Evans, L. G., Stockstill-Cahill, K. R., Lawrence, D. J., Goldsten, J. O., McCoy, T. J., Nittler, L. R., Solomon, S. C., Sprague, A. L., Starr, R. D. and Weider, S. Z. (2014). Enhanced sodium abundance in Mercury’s north polar region revealed by the MESSENGER Gamma-Ray Spectrometer. Icarus, 228, 86–95, doi:10.1016/j.icarus.2013.09.007. Pfleger, M., Lichtenegger, H. I. M., Wurz, P., Lammer, H., Kallio, E., Alho, M.., Mura, A., McKenna-Lawler, S. and Martin-Fernandez, J. A. (2015). 3D-modeling of Mercury’s solar wind sputtered surface-exosphere environment. Planet. Space Sci., 115, 90–101, doi:10.1016/j.pss.2015.04.016. Pifko, S., Janches, D., Close, S., Sparks, J., Nakamura, T. and Nesvorny, D. (2013). The Meteoroid Input Function and predictions of mid-latitude meteor observations by the MU radar. Icarus, 223, 444–459, doi:10.1016/j.icarus.2012.12.014. Poston, M. J., Grieves, G. A., Aleksandrov, A. B., Hibbitts, C. A., Dyar, M. D. and Orlando, T. M. (2015). Temperature programmed desorption studies of water interactions with Apollo


Models of Mercury’s Exosphere

lunar samples 12001 and 72501. Icarus, 255, 24–29, doi:10.1016/j.icarus.2014.09.049 Potter, A. E. (1995). Chemical sputtering could produce sodium vapor and ice on Mercury. Geophys. Res. Lett., 22, 3289–3292. Potter, A. E., Killen, R. M. and Morgan, T. H. (2002). The sodium tail of Mercury. Meteorit. Planet. Sci., 37, 1165–1172. Potter, A. E., Killen, R. M. and Sarantos, M. (2006). Spatial distribution of sodium on Mercury. Icarus, 181, 1–12, doi:10.1016/j. icarus.2005.10.026. Potter, A. E., Killen, R. M. and Morgan, T. H. (2007). Solar radiation acceleration effects on Mercury’s sodium emission. Icarus, 186, 571–580, doi:10.1016/j.icarus.2006.09.025. Press, W. H., Teukolsky, S. A., Vetterling, W. T. and Flannery, B. P. (2007). Numerical Recipes: The Art of Scientific Computing, 3rd edn. Cambridge: Cambridge University Press. Raines, J. M., Gershman, D. J., Zurbuchen, T. H., Sarantos, M., Slavin, J. A., Gilbert, J. A., Korth, H., Anderson, B. J., Gloeckler, G., Krimigis, S. M., McNutt, R. L., Jr. and Solomon, S. C. (2013). Distribution and compositional variations of plasma ions in Mercury’s space environment: The first three Mercury years of MESSENGER observations. J. Geophys. Res. Space Physics, 118, 1604–1619, doi:10.1029/ 2012JA018073. Roth, J. (1983). Chemical sputtering. In Sputtering by Particle Bombardment II, ed. R. Behrish, Topics in Applied Physics, Vol. 52. Berlin: Springer-Verlag, pp. 91–146, doi:10.1007/ 3-540-12593-0_3. Sarantos, M., Killen, R. M., McClintock, W. E., Bradley, E. T., Vervack, R. J., Jr., Benna, M. and Slavin, J. A. (2011). Limits to Mercury’s magnesium exosphere from MESSENGER second flyby observations. Planet. Space Sci., 59, 1992–2003, doi:10.1016/j.pss.2011.05.002. Sarantos, M., Killen, R. M., McClintock, W. E., Vervack, R. J., Jr., Merkel, A. W., Burger, M. H., Cassidy, T. A., Slavin, J. A., Sprague, A. L. and Solomon, S. C. (2012). Mercury’s Mg exosphere from MESSENGER data. EPSC Abstracts, 7, abstract EPSC2012-707-1. European Planetary Science Congress, Madrid, Spain, 23–28 September. Schmidt, C. A., Baumgardner, J., Mendillo, M. and Wilson, J. K. (2012). Escape rates and variability constraints for high-energy sodium sources at Mercury. J. Geophys. Res., 117, A03301, doi:10.1029/2011JA017217. Schriver, D., Trávníček, P., Ashour-Abdalla, M., Richard, R. L., Hellinger, P., Slavin, J. A., Anderson, B. J., Baker, D. N., Benna, M., Boardsen, S. A., Gold, R. E., Ho, G. C., Korth, H., Krimigis, S. M., McClintock, W. E., McLain, J. L., Orlando, T. M., Sarantos, M., Sprague, A. L. and Starr, R. D. (2011). Electron transport and precipitation at Mercury during the MESSENGER flybys: Implications for electron-stimulated desorption. Planet. Space Sci., 59, 2026–2036, doi:10.1016/j.pss.2011.03.008. Schultz, P. (1996). Effect of impact angle on vaporization. J. Geophys. Res., 101, 21,117–21,136, doi:10.1029/96JE02266. Sieveka, E. M. and Johnson, R. E. (1984). Ejection of atoms and molecules from Io by plasma-ion impact. Astrophys. J., 287, 418–426, doi:10.1086/162701. Sigmund, P. (1969). Theory of sputtering I: Sputtering yields of amorphous and polycrystalline targets. Phys. Rev., 184, 383–416. Smith, R. S. and Kay, B. D. (1997). Adsorption, desorption, and crystallization kinetics in nanoscale water films. Recent Res. Dev. Phys. Chem., 1, 209–219. Smyth, W. H. (1983). Io’s sodium cloud: Explanation of the east–west asymmetries. II. Astrophys. J., 264, 708–725.

Smyth, W. H. and Marconi, M. L. (1995a). Theoretical overview and modeling of the sodium and potassium atmospheres of the Moon. Astrophys. J., 443, 371–392. Smyth, W. H. and Marconi, M. L. (1995b). Theoretical overview and modeling of the sodium and potassium atmospheres of Mercury. Astrophys. J., 441, 839–864. Sneh, O., Cameron, M. A. and George, S. M. (1996). Adsorption and desorption kinetics of H2O on a fully hydroxylated SiO2 surface. Surf. Sci., 364, 61–78. Starukhina, L. V. and Shkuratov, Y. G. (2000). The lunar poles: Water ice or chemically trapped hydrogen?, Icarus, 147, 585–587, doi:10.1006/icar.2000.6476. Szalay, J. R. and Horanyi, M. (2016). Annual variation and synodic modulation of the sporadic meteoroid flux to the Moon. Geophys Res. Lett., 42, 10,580–10.584, doi:10.1002/2015GL066908. Vasavada, A. R., Paige, D. A. and Wood, S. E. (1999). Nearsurface temperatures on Mercury and the Moon and the stability of polar ice deposits. Icarus, 141, 179–193, doi:10.1006/ icar.1999.6175. Verner, D. A. and Yakovlev, D. G. (1995). Analytic FITS for partial photoionization cross sections. Astron. Astrophys. Suppl. Ser., 109, 125–135. Verner, D. A., Yakovlev, D. G., Band, I. M. and Trzhaskovskaya, M. B. (1993). Subshell photoionization cross sections and ionization energies of atoms and ions from He to Zn. Atom. Data Nucl. Data Tables, 55, 233–280. Verner, D. A., Ferland, G. J., Korista, K. T. and Yakovlev, D. G. (1996). Atomic data for astrophysics. II. New analytic FITS for photoionization cross sections of atoms and ions. Astrophys. J., 465, 487–498. Vervack, R. J., Jr., McClintock, W. E., Killen, R. M., Sprague, A. L., Anderson, B. J., Burger, M. H., Bradley, E. T., Mouawad, N., Solomon, S. C. and Izenberg, N. R. (2010). Mercury’s complex exosphere: Results from MESSENGER’s third flyby. Science, 329, 672–675. Vervack, R. J., Jr., Killen, R. M., Sprague, A. L., Burger, M. H., Merkel, A. W. and Sarantos, M. (2011). Early MESSENGER results for less abundant or weakly emitting species in Mercury’s exosphere. EPSC-DPS Joint Meeting Abstracts and Program, 6, abstract EPSC-DPS2011-1131. European Planetary Science Congress–Division for Planetary Sciences Joint Meeting, Nantes, France, 2–7 October. Vervack, R. J., Jr., McClintock, W. E., Killen, R. M., Merkel, A. W., Burger, M. H., Sarantos, M. and Cassidy, T. A. (2015). Mercury’s exosphere: New detections, discoveries, and insights. Abstracts, 47th Division for Planetary Sciences Annual Meeting, abstract 107.01. National Harbor, MD, 8–13 November, pp. 19–20. Vervack, R. J., Jr., Killen, R. M., McClintock, W. E., Merkel, A. W., Burger, M. H., Cassidy, T. A. and Sarantos, M. (2016). New discoveries from MESSENGER and insights into Mercury’s exosphere. Geophys. Res. Lett., 43, 11,545–11,551. Weider, S. Z., Nittler, L. R., Starr, R. D., Crapster-Pregont, E. J., Peplowski, P. N., Denevi, B. W., Head, J. W., Byrne, P. K., Hauck, S. A., Ebel, D. S. and Solomon, S. C. (2015). Evidence for geochemical terranes on Mercury: Global mapping of major elements with MESSENGER’s X-Ray Spectrometer. Earth Planet. Sci. Lett., 416, 109–120, doi:10.1016/j. epsl.2015.01.023. Wurz, P. and Lammer, H. (2003). Monte-Carlo simulation of Mercury’s exosphere. Icarus, 164, 1–13. Wurz, P., Whitby, J. A., Rohner, U., Martin-Fernandez, J. A., Lammer, H. and Kolb, C. (2010). Self-consistent modelling of

References Mercury’s exosphere by sputtering, micro-meteorite impact and photon-stimulated desorption. Planet. Space Sci., 58, 1599–1616, doi:10.1016/j.pss.2010.08.003. Yakshinskiy, B. V. and Madey, T. E. (1999). Photon-stimulated desorption as a substantial source of sodium in the lunar atmosphere. Nature, 400, 642–644. Yakshinskiy, B. V. and Madey, T. E. (2000). Desorption induced by electronic transitions of Na from SiO2: Relevance to tenuous planetary atmospheres. Surf. Sci., 451, 160–165.


Yakshinskiy, B. V. and Madey, T. E. (2004). Photon-stimulated desorption of Na from a lunar sample: Temperature-dependent effects. Icarus, 168, 53–59. Yakshinskiy, B. V. and Madey, T. E. (2005). Temperature-dependent DIET of alkalis from SiO2 films: Comparison with a lunar sample. Surf. Sci., 593, 202–209. Yakshinskiy, B. V., Madey, T. E. and Ageev, V. N. (2000). Thermal desorption of sodium atoms from thin SiO2 films. Surface Rev. Lett., 7, 75–87.

16 Structure and Configuration of Mercury’s Magnetosphere H AJ E KORTH, BRIA N J . AND ERS ON, CATHERIN E L . J OHN SON , J AMES A. SLA VIN, JIM M . RA INES, AN D TH OMA S H. ZU RBUCHEN

1 6 . 1 I N T R O D U C T IO N

revealed the presence of a weak, global-scale planetary magnetic field that was represented by a dipole with a southwarddirected planetary moment, like at Earth, but with a surface field weaker by a factor of ~1000 (Ness et al., 1974, 1975). Mercury’s internal field produces a magnetosphere, albeit much smaller than Earth’s, with a magnetotail, a magnetopause, and a bow shock, as illustrated in Figure 16.1. The boundary enveloping the planetary magnetic field lines and separating them from the IMF lines is the magnetopause. The magnetosphere is an obstacle to the supersonic flow of the solar wind, and consequently a bow shock wave forms upstream of the magnetopause. At the dayside magnetopause, magnetic reconnection opens magnetic flux that is subsequently transported anti-sunward by the solar wind to form a quasi-cylindrical structure on the nightside of the planet, with northern and southern lobes that contain magnetic flux linked to the northern and southern polar regions of the planet, respectively (Russell et al., 1988). The plasma sheet, the layer of plasma separating the northern and southern lobes of the magnetotail, exhibits a plasma β value that is greater than unity. It is composed of closed magnetic flux closer to the planet but open magnetic flux farther downstream, beyond the distance at which the magnetic flux from the two lobes reconnects (Slavin et al., 2007). Estimates of Mercury’s planetary dipole moment were made from the Mariner 10 observations under the assumption that the dipole is centered on the planet. The non-dipole structure was only loosely constrained because the limited spatial distribution of the observations led to high correlations among the spherical harmonic coefficients describing the model field (Connerney and Ness, 1988; Chapter 5). In the absence of more extensive observations of the magnetosphere and its boundaries, our understanding of the magnetospheric structure remained largely conceptual, and models for Mercury’s magnetospheric field were either highly simplified (Whang, 1977; Grosser et al., 2004) or scaled-Earth models (Jackson and Beard, 1977; Luhmann et al., 1998; Korth et al., 2004). Mariner 10 also obtained observations of Mercury’s plasma environment by measuring low-energy (~100 eV) electrons in the magnetosheath and electrons with energies up to the instrument limit of 1 keV in the plasma sheet and its boundary layer (Ogilvie et al., 1974). The electron observations indicated that the solar wind is the primary source for the plasma sheet. No charged particle observations were obtained in the energy range 1−100 keV, and the population of energetic particles with energies >100 keV (Simpson et al., 1974) may have been overestimated because of an instrumental effect (Armstrong et al.,

Of the terrestrial planets, only Mercury and Earth possess global magnetic fields and, in consequence, magnetospheres. Mercury’s magnetosphere is substantially different from Earth’s in a number of key respects. In the inner heliosphere, the solar wind subjects Mercury’s magnetosphere to a much higher ram pressure, 10 to 30 nPa, than the ~2 nPa at Earth, and the magnitude of the interplanetary magnetic field (IMF) is much higher, ~30 nT, than at Earth (~5 nT). Also, the IMF is predominantly radially (sunward or anti-sunward) directed at Mercury, so that its bow-shock normal is more likely to be quasi-parallel to the solar wind flow than that of Earth. Moreover, the lower Alfvén Mach number – the ratio of plasma speed to Alfvén speed – leads to a somewhat weaker bow shock and lower magnetosheath plasma β – the ratio of the plasma thermal pressure to the magnetic pressure – at Mercury. These environmental differences have profound consequences for both the structure and dynamics of Mercury’s magnetosphere. In addition, the relatively small planetary moment implies that Mercury’s magnetosphere is much smaller than Earth’s, so the characteristic timescales for convection and wave transits through Mercury’s magnetosphere are nearly two orders of magnitude shorter than at Earth. The volume fraction of Mercury’s magnetosphere occupied by the planet itself is about a factor of 500 larger at Mercury than at Earth, implying that plasma dynamics within Mercury’s magnetosphere is substantially different and reflects the direct interaction with the planetary surface. Furthermore, owing to Mercury’s large iron core, the effects of magnetic induction within the core on the magnetosphere are clearly evident at Mercury. Finally, Mercury possesses no permanent atmosphere of appreciable density and, hence, supports no ionosphere, so the inner boundary of the magnetosphere is fundamentally different from Earth’s. Given the breadth of contrasts between Earth’s and Mercury’s magnetospheres, the quantitative comparison of the two systems affords a critical test of our understanding of the physics of planetary magnetospheres. The first in situ observations of Mercury and its space environment made four decades ago by the Mariner 10 spacecraft revealed that the innermost planet has a magnetic field that is sufficiently strong to stand off the solar wind and form a magnetosphere. Mariner 10 executed three flybys of Mercury, two of which passed through the magnetosphere – one near the equatorial plane and one approximately over the northern pole. The magnetic field data acquired during these flybys


16.2 Mercury’s Solar Wind Environment


Figure 16.1. Schematic view of Mercury’s magnetosphere with boundaries, major regions, magnetic field orientation (green curves and arrows), and MESSENGER orbit (dashed red curve) identified. Adapted from Zurbuchen et al. (2011).

1975). Unfortunately, the plasma ion instrument failed in flight, so Mariner 10 was unable to return information about ions at Mercury. Many new insights into Mercury’s magnetosphere were enabled by data returned from the MESSENGER spacecraft. MESSENGER completed three equatorial flybys of Mercury during the 6.6-year mission cruise phase and conducted orbital observations for more than four Earth years (more than 16 Mercury years) (see Chapter 1). The extensive magnetic field and particle observations accumulated in orbit allowed detailed characterization of the structure and configuration of Mercury’s magnetosphere. MESSENGER magnetic field observations were used to determine definitively the orientation, intensity, and location of the internal planetary magnetic moment by first fitting the magnetic equator crossing points (Anderson et al., 2011b) to yield the position and orientation of an equivalent planetary magnetic dipole relative to the body center and then using these constraints in the evaluation of the magnetic moment (Johnson et al., 2012). (See also Chapter 5 for an extensive discussion of Mercury’s internally generated field.) The result is that the dipole is aligned to within 0.6° ± 0.1° of the planetary rotation axis and is offset 484 ± 4 km northward, along the rotation axis, and with these specifications an internal dipole moment of 190 ± 10 nT RM3, where RM is Mercury’s mean radius (2439.4 km), was determined (Chapter 5). These specifications for the internally generated magnetic field formed the basis for subsequent exploration of Mercury’s magnetosphere. MESSENGER observations were used to establish the configuration of the magnetopause, bow shock, cross-tail current sheet, and field-aligned or Birkeland currents. Plasma observations were used to determine the distribution and composition of plasma in the magnetosphere. This suite of analyses revealed processes unique to Mercury among magnetospheres in our solar system. In this chapter, we review our understanding of the geometry and dominant physical processes of Mercury’s

magnetosphere inferred from MESSENGER data. We first review the solar wind environment in the inner solar system, because such an overview provides the context that governs the magnetospheric geometry and because the solar wind is the source of most of the plasma in Mercury’s magnetosphere. We consider the shape and location of the magnetospheric boundaries and discuss the fundamental regions and configuration of the magnetosphere. We then describe the magnetospheric current systems and present state-of-the-art models of the magnetospheric magnetic field that combine mathematical descriptions of most of these current systems and the planetary dipole. We conclude by considering the plasma environment in Mercury’s magnetosphere, discussing the sources and losses of plasma and describing the processes by which plasma is transported and heated.

16. 2 ME RC UR Y’S SOLAR WIND EN VIR ONM ENT The solar wind near Mercury is composed of fast and slow solar wind, and these components are identified by their distinct origins, composition, and dynamic properties. The fast solar wind at Mercury’s heliocentric latitudes originates in so-called coronal holes from which outflowing plasma expands into space on solar magnetic field lines (Zurbuchen, 2007). It has a typical velocity of ~700 km s−1 and tends to be time-stationary in composition; the plasma has substantial non-thermal He and heavy-ion components that have velocities exceeding those of pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi the protons by about the local Alfvén speed, vA ¼ B2 =ðμ0 ρÞ, where B is the magnetic field magnitude, μ0 is the permeability of free space, and ρ is the plasma mass density (von Steiger et al., 1995, 2000). At Mercury’s heliocentric distance, the nonthermal He and heavy ions flow up to ~150 km s−1 faster than the protons. The ion temperatures in the fast solar wind are not


Structure of the Magnetosphere

equal but are proportional to the ion mass, leading to heavy-ion temperatures of up to ~106 K (Marsch et al., 1982). MESSENGER observations showed that nearly 40% of the thermal pressure and up to 20% of the momentum flux is carried by alpha particles and heavy ions (Gershman et al., 2012). The dynamic pressure of the fast solar wind is higher and the Mach number lower than expected from adiabatic propagation of the much more thermalized and equilibrated solar wind observed near Earth. The slow solar wind originates at coronal hole boundaries (Neugebauer et al., 1998), in active regions (Neugebauer et al., 2002), and in coronal streamers (Gosling et al., 1981). In the latter, the solar magnetic field is nominally perpendicular to the outward flow, resulting in a lower speed (vsw ~ 350 km s−1), higher density, and greater thermalization (Zurbuchen, 2007). In contrast to the fast solar wind, the density, flow speed, number flux, and temperature of the slow solar wind stream are highly variable on timescales of hours to days (Gosling, 1997). The momentum and Mach number in the slow wind are dominated by protons, and different ion species have similar temperatures and flow speeds (von Steiger and Zurbuchen, 2006). The slow solar wind also hosts the heliospheric current sheet that separates the northern and southern magnetic polarity regions of the heliosphere (Smith, 2001).

1 6. 3 S HA PE A ND LO CA TIO N OF M A G N E TO S P H E R I C B O U N D A R I E S The boundaries of the magnetosphere are the bow shock and the magnetopause. At the bow shock, the plasma and magnetic field are compressed as the particles’ bulk flow speed is substantially reduced (Spreiter et al., 1966a). The flow energy is converted to thermal energy so that the plasma in the magnetosheath located immediately downstream of the shock is denser and hotter than in the solar wind. The location and shape of the bow shock are controlled by the solar wind momentum, solar wind Mach number, and, to a lesser extent, the IMF direction (Fairfield et al., 2001). The magnetopause position is determined to first order by the balance of the dynamic pressure exerted by the solar wind ram flow, which is partially converted to thermal pressure at the bow shock, and the magnetic pressure of the planetary field (Schield, 1969). The bow shock and magnetopause locations vary substantially in response to dynamics arising from boundary waves and reconnection (Chapter 17). Reconnection of interplanetary and planetary magnetic field lines can modify the bow shock and magnetopause locations by eroding the dayside magnetopause and loading the magnetosphere with magnetic flux on the field lines opened by reconnection, in turn leading to flaring of the magnetopause. Because Mercury can be embedded in either the fast or slow solar wind, the degree of compression of the magnetosphere and imposed solar wind plasma varies markedly with time. This variability, together with that introduced by the influence of the IMF orientation and magnitude on the boundaries, leads to correspondingly large variations in the magnetopause and bow shock locations. The extensive observations provided by the orbital phase of the MESSENGER mission

were therefore important to constraining the average boundary positions. The MESSENGER spacecraft crossed the bow shock and magnetopause twice on every orbit, allowing a quantitative characterization of their location and shape throughout the range of heliocentric distances spanned by Mercury’s orbit. Both the bow shock and the magnetopause have been mapped in detail, using MESSENGER observations to document their morphology and provide insight into the physical processes that lead to their formation and influence their dynamics. The bow shock and magnetopause shape were characterized by Winslow et al. (2013) from magnetic field observations during three Mercury years that spanned a broad range of solar wind and IMF conditions. Figure 16.2 shows an example for a magnetosphere transit from the dayside to the nightside. Observations are shown in the Mercury solar orbital (MSO) coordinate system, in which +X points toward the Sun, +Y is the direction opposite to the planet’s orbital motion and orthogonal to +X (toward dusk), and +Z completes the right-handed system. The inbound (outbound) bow shock encounter is marked by a sharp increase (decrease) in B, which was most readily observed when the subsolar shock normal was quasi-perpendicular to the IMF, as was the case during the interval shown. In addition, there was a pronounced increase in the 1–10-Hz bandpass amplitude, denoted BAC (Anderson et al., 2007). For quasi-parallel shocks, the bow shock was often quite broad and was conservatively bracketed in time by noting the outermost and innermost excursions in B. Magnetopause crossings were most reliably identified by the rotation of the magnetic field across the magnetopause current layer. The example in Figure 16.2 illustrates a high-shear magnetopause marked by the rotation signature in BY and BZ. The magnetopause also exhibited an increase in BAC, as was almost always the case even when the magnetic shear was smaller. Because shear angles between the magnetospheric and magnetosheath magnetic fields were often X(RM) > –2.0 μ = 8.44 σ = 7.09 M = 7.47 –2.0 > X(RM) > –2.3 σ = 6.24 M = 5.81 μ = 6.79 –2.3 > X(RM) > –2.6 σ = 5.52 M = 6.93 μ = 6.63


25 15 5 –5

BZ (nT)

–1.4 > X(RM) > –1.7

Normalized Probability

–1.4 > X (RM) > –1.7



25 15 5 –5

BZ (nT)


Structure of the Magnetosphere

25 15 5 –5

BZ (nT)


25 15 5 –5

= 100 x σM

Slope = –2.81 ± 0.34

–1.7 > X (RM) > –2.0 Slope = –2.11 ± 0.42

–2.0 > X (RM) > –2.3 Slope = –0.74 ± 0.43

–2.3 > X (RM) > –2.6 0.0

–30 –20 –10



20 30 BZ (nT)





Slope = –0.57 ± 1.19









1.5 DUSK

Figure 16.5. (a) Distribution of the BZ component in MSM coordinates of the plasma sheet magnetic field at four 0.3-RM-wide ranges from X = −1.4 RM to −2.6 RM. Mean (μ), median (M), and standard deviation (σ) for each range are provided. (b) The average BZ magnetic field component as a function of the cross-tail location, −1.6 RM < Y < 1.6 RM, for four downstream distances. For each distance range, the standard error of the mean, σM, is displayed after being multiplied by 100, and the slope of a linear fit to the data is given. Quadratic fits to the strong BZ in the top panel (closest to Mercury) and the weak BZ in the bottom panel (farthest from Mercury) are displayed in red.

the near-Mercury neutral line forming at distances of X ~ −1.5 RM to −2.5 RM (Slavin et al., 2009a, 2010, 2012; DiBraccio et al., 2015b). 16.4.2 Plasma Sheet The plasma sheet hosts the majority of the plasma in Mercury’s magnetosphere. The transport of flux into the tail and toward the plasma sheet in the lobes corresponds to a dawn-to-dusk electric field, E, and this electric field transports plasma on open magnetic flux tubes in the tail lobes via drift in the E × B direction. MESSENGER acquired plasma observations with the Fast Imaging Plasma Spectrometer (FIPS) (Andrews et al., 2007), which measured protons and heavy-ion species in the energy range from 46 eV e−1 to 13 keV e−1, where e is the electron charge, with a 1.4π sr field of view. FIPS observations during MESSENGER’s first Mercury flyby on 15 January 2008 showed that Mercury’s plasma sheet is Earth-like with respect to proton densities, pressures, and plasma β (Raines et al., 2011). Near the planet, the plasma sheet extends to low altitudes just poleward of the boundary between open and closed magnetic field lines located at mid latitudes (Korth et al., 2014; Sun et al., 2015). The sources of the plasma in the plasma sheet are the solar wind and the planet, and charged particles populate this region via the processes described in Section 16.6.3. The average distribution functions for both solar wind and planetary ions in Mercury’s pre-midnight plasma sheet are well described by hot Maxwell– Boltzmann distributions, and the plasma bulk properties are shown in Figure 16.6 (Gershman et al., 2014). Densities and temperatures of the H+-dominated plasma sheet are in the ranges ~1–10 cm−3 and ~5–30 MK, respectively, and the plasma sheet maintains a thermal pressure of ~1 nPa. The plasma-sheet

density decreases with increasing solar wind velocity, vsw, whereas the temperature increases with vsw. The average bulk properties of other ion species in the plasma sheet relative to H+ (n = 7.8 cm−3, T = 9.3 MK) are shown in Figure 16.7. The dominant planetary ion species are Na+-group ions [mass per charge (m/q) ratio = 21−30], which exhibit number densities ~10% of those of H+ on average, followed by He2+ and O+-group ions (m/q = 16−20) with number densities 3.5% and 1.5%, respectively. The average temperature of the planetary ion species is a factor of ~1.5 larger than that of the protons observed nearby. These values imply that planetary ions could contribute ~15% to the plasma thermal pressure and ~50% to the mass density in the nightside plasma sheet. Whereas solar wind ions (i.e., H+, He2+, O6+) show mass-proportional temperatures, the temperatures of planetary ions are approximately equal. The latter characteristic may be additional evidence of non-adiabatic particle motion in Mercury’s magnetosphere, because it is consistent with the ion motion induced by a potential drop, rather than an E × B drift, as would be expected for energetic heavy ions gyrating with a large radius in a weak magnetic field (see Section 16.6.2). The spatial distribution of plasma-sheet protons in the magnetic equatorial plane has been determined from FIPS observations and inferred independently from Magnetometer data. Although MESSENGER’s orbit line of apsides is substantially inclined with respect to Mercury’s equatorial plane, the distribution of plasma at the magnetic equator may be obtained by mapping observations along magnetic field lines. Statistical mapping of the plasma-sheet proton flux (Figure 16.8a) reveals the existence of a plasma enhancement within a toroidal section, which is centered at the magnetic equator near local midnight and extends on the nightside from dusk to dawn (Korth et al., 2014). An enhanced plasma population is also found near the

16.4 Regions and Configuration of the Magnetosphere 102 (a)


Subsolar Magnetosheath








3.2 1.6

Plasma Sheet
















ni / nH+

nH+ (cm–3)


Solar Wind




0.8 0.2





TH+ (MK)

(b) Ti / TH+ = mi



6+ 5+

Ti / TH+

nH+ (cm–3)

101 He2+

+ O group



He+ Ti = TH+

5 40 (c)

TH+ (MK)

10 15 20 Ion Mass (amu)


Figure 16.7. Average (a) density ni and (b) temperature Ti of ion species relative to H+. Dashed lines corresponding to Ti = TH+ and Ti/TH+ = mi are also shown. Adapted from Gershman et al. (2014).





+ Na group



500 600 VSW (km/s)



Figure 16.6. (a) Orbit-averaged density and temperature of H+ in Mercury’s pre-midnight/dusk-side plasma sheet at heliocentric distance R ≈ 0.35 AU for 113 orbits. Lines of constant pressure (in nPa) are dashed. Ranges of H+ density and temperature observed elsewhere in Mercury’s magnetosphere are indicated with red, yellow, and green boxes for the solar wind (Gershman et al., 2012; Baker et al., 2013), subsolar magnetosheath (Gershman et al., 2013), and northern magnetospheric cusp (Raines et al., 2014), respectively. (b) Plasmasheet density versus upstream solar wind speed. (c) Plasma temperature versus upstream solar wind speed. Black squares denote bin averages, vertical error bars denote the standard deviation of values in each bin, and horizontal error bars correspond to the bin size. Adapted from Gershman et al. (2014).

magnetopause flanks, indicating entry of magnetosheath plasma into the low-latitude boundary layer of the magnetosphere. The plasma bulk properties are derived from the measured fluxes (Gershman et al., 2012), and regions showing higher fluxes also exhibit higher plasma pressures (Korth et al., 2014). Consistent with the distribution of protons, the average densities of all planetary ions within 30° of the planetary equator are

depressed at the subsolar point relative to the dawn and dusk terminators (Raines et al., 2013). The effect is largest for Na+group ions, which are 49% lower in density at the subsolar point than at the terminators. The observations show that dense plasma does not form a closed distribution around the planet, likely because of the dynamic solar wind and IMF conditions, which prevent the formation of drift paths that close around the planet. Enhancements of the plasma pressure in the plasma sheet were independently determined from diamagnetic depressions of the background magnetic field (Korth et al., 2012). These decreases in the magnetic field magnitude arise from reductions in magnetic pressure in the presence of an increase in plasma thermal pressure to maintain total pressure balance. The magnetic-field technique complements observations from FIPS because the FIPS instrument observes only 35% of the full solid angle. Using the paraboloid magnetic field model described in Section 16.5.3 to represent the average magnetospheric magnetic field, Korth et al. (2012) inferred the plasma pressure enhancements from the magnetic field perturbations with respect to that baseline. The resulting distribution of the plasma pressure enhancement (Figure 16.8b) qualitatively reproduces that of the proton flux in the inner magnetosphere (Figure 16.8a), where the pressures are large and localized. The magnitude of the average magnetic pressure deficit normalized for heliocentric distance is 1.45 nPa and exhibits a weak, 0.05nPa h−1, dusk-to-dawn gradient with local time. The magnitudes of the pressure agree with estimates derived from plasma observations on average but can deviate for individual events by factors of up to ~3 (Korth et al., 2014).


Structure of the Magnetosphere

Figure 16.8. Distributions of (a) the mean proton flux observed by FIPS and (b) the mean magnetic pressure deficit determined from Magnetometer data mapped to Mercury’s magnetic equatorial plane and normalized to a heliocentric distance of 0.39 AU. The circle denotes the planet, the Sun is to the right, and the magnetopause of the magnetic field model is represented by the solid black curves. Adapted from Korth et al. (2014).

16.4.3 Cusp In a vacuum representation, the cusps are singularities in the magnetic field immediately inside the magnetopause. The magnetic field vanishes at the cusps, and all magnetic field lines immediately inside the current layer thread the cusps (Olson, 1984). The Chapman–Ferraro currents (Chapman and Ferraro, 1930, 1931) that form the dayside magnetopause flow around the cusps, and for Mercury the current loops are counterclockwise in the north and clockwise in the south when viewed from the Sun (Olson, 1984). In Mercury’s magnetosphere, the cusps are regions of plasma exchange between the magnetosheath and the magnetosphere, for two reasons. First, because the magnetic field is weak near the cusps, the magnetic pressure cannot stand off the thermal pressure of the magnetosheath, so magnetosheath plasma has ready access to magnetospheric field lines. The cusps are therefore regions of high thermal pressure and high magnetosheath densities near and even within the magnetopause. Second, because magnetic fields on the dayside magnetopause map to the cusps, fields that are reconnected with the magnetosheath field on the dayside map near the cusp and its projection to the planetary surface. Reconnection at the dayside magnetopause therefore injects plasma into the cusps and drives convection of magnetic flux through the cusps and into the polar caps, which are the regions threaded by field lines connected to the tail lobes. Open field lines allow access of plasma from both magnetosheath and planetary sources, so that the plasma population in the vicinity of the cusp is a mixture of these plasmas. In addition, because Mercury’s atmosphere is tenuous, the cusp plasma can interact directly with the surface, causing space weathering and providing a source of sputtered neutral atoms and ions to the exosphere and magnetosphere.

The location of Mercury’s northern cusp at dayside high latitudes was first studied statistically by Winslow et al. (2012) with MESSENGER data acquired during the first and second Mercury years in orbit. Using the method summarized in Section 16.4.2, these authors inferred plasma pressure enhancements in the cusp from diamagnetic depressions in the magnetic field magnitude. The cusp was found on average to span a region extending 11° in latitude and 4.5 h in local time at spacecraft altitudes (Figure 16.9). The bounds of the northern cusp are 55.8°N and 83.6°N MSO latitude and 7.2 h and 15.9 h local time, and the cusp is approximately symmetric about noon (Figure 16.9). Because the MESSENGER orbit was eccentric and periapsis during this phase of the mission was on the descending latitude portion of the orbit, the cusp was encountered at lower altitudes on the descending than on the ascending orbit leg. Consistent with the expected shift in cusp latitude closer to the magnetopause, the high-altitude cusp was observed on average a few degrees equatorward of that seen at lower altitude. Winslow et al. (2014) independently determined the extent of the northern cusp using proton reflectometry to measure the proton loss cone indicating persistent ion precipitation to the surface. Tracing the locations of FIPS observations in the cusp taken from 7 June 2011 to 7 June 2012 along magnetic field lines to the surface with a paraboloid magnetic field model (Section 16.5.3) showed that the cusp is centered on noon at 76.4°N latitude and extends 15.6° in latitude and 7.5 h in local time. The cusp location coincides with a region where high fluxes of planetary Na+-group and O+-group ions peaking near noon and 60°N latitude are observed (Zurbuchen et al., 2011). Because of the northward offset of the planetary dipole, the extent of the cusp in the southern hemisphere is expected to be larger than in the north. However, as a result of MESSENGER’s highly eccentric orbit, the spacecraft was located outside the

16.4 Regions and Configuration of the Magnetosphere


Figure 16.9. Stereographic projections of the pressure deficit along profiles across the cusp shown in MSO coordinates corrected for aberration of the magnetosphere resulting from Mercury’s finite orbital velocity. During portions of MESSENGER’s first Mercury year in orbit (MSO1), the Magnetometer was off when the spacecraft experienced long eclipses or was close to the planet, resulting in gaps in data coverage (between ~10 h and ~12 h local time) for the descending tracks. Complete coverage was obtained during MESSENGER’s second Mercury year in orbit (MSO2). Projections span local times from 6.7 h to 17.3 h and latitudes 55°N to the pole. The color bar is saturated so that observed, but localized, pressure deficits greater in magnitude than −3 nPa are shown in red. Adapted from Winslow et al. (2012).

Na group 90





–3 30

2000 3000






500 1000 2000 3000

–4 –5


Latitude (deg)

magnetosphere when traversing the dayside high-latitude region in the southern hemisphere, so the southern cusp could not be mapped with the techniques mentioned above. The plasma populations of the northern cusp were observed by the FIPS sensor. On most orbits that crossed the cusp, MESSENGER traversed the cusp at altitudes ranging from 200 to 600 km, whereas the sampling altitude decreased to as low as 11 km during the final phase of the mission (Chapter 1). Consistent with the location of the cusp identified above (Figure 16.9), overall lower fluxes of cusp ions were observed during times when the spacecraft was in a dawn–dusk orbit. The FIPS observations showed that protons are the dominant ion species in the cusp, followed by alpha particles (He2+) and Na+-group ions (Raines et al., 2014). Other planetary ions, e.g., O+-group ions and He+, were also observed but with lower fluxes. The distribution of Na+-group ions, shown in Figure 16.10, exhibits a peak in the flux within the northern cusp. Protons enter the cusp from the solar wind directly from the dayside reconnection region or drift toward the planet along newly reconnected field lines as they convect across the polar cap. Some protons have sufficient energy parallel to the magnetic field to overcome the mirror force of the planetary magnetic field (which increases in strength toward the surface) and impact the surface. The remaining protons mirror and flow upward along the field lines away from the planet and into the magnetotail. Precipitating energetic protons impacting at the surface can lead to ion and neutral sputtering, a process that results in release of



–90 0

6 12 18 Local Time (h)

Figure 16.10. Distribution of Na+-group ions as a function of latitude and local time sampled along the MESSENGER trajectory during 130 orbits in 2011. Horizontal lines denote spacecraft altitude in km. Adapted from Zurbuchen et al. (2011).

neutral atoms (~90−99% by number density) and ions (~1−10% by number density) into the exosphere and the magnetosphere (Killen et al., 2007). Similarly, precipitating solar wind electrons contribute to the exospheric and magnetospheric populations through electron-stimulated desorption. Evidence of the precipitation process leading to depletion of flux in the upward direction is shown in Figure 16.11 (Raines et al., 2014). Figure 16.11a


Structure of the Magnetosphere Na+ -group

H+ 10–9.5



10 W⊥ (keV/e)

W⊥ (keV/e)





–0.1 0.1 W|| (keV/e)

10 Precipitating H+


Loss cone





f// –3 n = 21.4 cm , T = 5.75 MK

10–10 10–12 10–14







W/q (keV/e)



(d) Phase Space Density (s3 m–6)


(c) Phase Space Density (s3 m–6)









0.1 W|| (keV/e)


Phase Space Density (s3 m–6)



f⊥ f//

10–8 Upwelling Na+


n = 1.03 cm , T = 11.4 MK

10–10 10–12 10–14 10–16







W/q (keV/e)

Figure 16.11. (Top) Pitch-angle distributions for (a) protons and (b) Na+-group ions summed over 77 selected cusp crossings resolved by energy parallel, W||, and perpendicular, W⊥, to the local magnetic field. Positive (negative) W|| values correspond to precipitating (upwelling) particles. (Bottom) Phase-space density versus energy per charge W/q for directions parallel (f||, blue squares to the right of the dotted vertical line), antiparallel (−f||, blue squares to the left of the dotted vertical line), and perpendicular (f⊥, red triangles) to the magnetic field for (c) protons and (d) Na+-group ions. The black curves indicate the best-fit Maxwell–Boltzmann distributions, and uncertainties in the counting statistics are shown as vertical bars. Adapted from Raines et al. (2014).

shows the proton distribution as a function of the pitch angle, pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi α ¼ tan1 ð W? =Wjj Þ, where W|| and W⊥ are the particle energies parallel and perpendicular to the magnetic field, respectively. Positive (negative) W|| values correspond to precipitating (upwelling) particles. Figure 16.11a shows that the precipitating proton flux is larger than that of the upwelling population, implying that protons are lost to the surface. This loss is also evident in the distribution of the phase-space density (Figure 16.11c) and is labeled as the loss cone. Loss processes are discussed in more detail in Section 16.6.3. The energy of the Na+-group ions in the cusp typically ranges from about 800 eV to a few keV, with an average of 2.7 keV and a detected maximum of 13 keV limited by the measurement range of the FIPS sensor. Planetary ions produced locally in the cusp initially have much lower energies, 0.1−10 eV, depending on their source (Raines et al., 2014, and references therein), implying that some mechanism accelerates the Na+-group ions to their observed energies. Several mechanisms that have been shown to operate at Earth, e.g., the “cleft ion fountain” (Lockwood et al., 1985) and wave–particle interactions (Ashour-Abdalla et al., 1981), may also be acting at Mercury but are expected to yield energies of only tens of eV, much lower than observed. Raines et al. (2014) proposed that the keV-energy Na+-group ions in the cusp are created when neutral Na atoms, which have drifted beyond the magnetopause boundary, are photoionized and accelerated as they are picked up in the convection of newly reconnected magnetic field lines, which flow anti-sunward over the polar cap. The local Alfvén speed, hundreds of kilometers per second, is consistent with acceleration of Na+-group ions to the

observed energies because, for a Na+ ion, an energy of 2.7 keV corresponds to a speed of 210 km s−1. A population of Na+-group ions moving upward away from the planet was also observed in the cusp (Figures 16.11b and 16.11d, Raines et al., 2014) and provides the first evidence for the predicted effect of ion precipitation and interaction with surface material (Killen et al., 2007). The upwelling component is visible in the energy-resolved pitch-angle distribution (Figure 16.11b) as a narrow band of enhanced flux in the 160° to 180° pitch-angle bin (the bin closest to the −W|| direction) and as a bump in the phase space density in the one-dimensional anti-sunward cut (Figure 16.11d). The energy of these upwelling ions ranges from 100 to 300 eV, implying that they were accelerated by factors of 10−100 after generation at the surface. 16.4.4 Dayside Boundary Layer The first MESSENGER flyby of Mercury revealed a striking transition near the morning magnetopause during which the magnetic field intensity decreased by nearly a factor of 2 without a significant change in orientation, as shown in Figure 16.12 (Anderson et al., 2011a). This transition occurred within 200−300 km of the magnetopause and was attributed by Anderson et al. (2011a) to a boundary layer (BL in Figure 16.12) of plasma of solar wind origin just inside the magnetopause, consistent with results from hybrid simulations of Mercury’s magnetosphere (Trávníček et al., 2007). The mechanism by which this plasma is transported across the magnetopause is not

16.4 Regions and Configuration of the Magnetosphere


Figure 16.12. Magnetic field and proton data for the outbound magnetosphere crossing of MESSENGER’s first Mercury flyby. The inner edge of the boundary layer and the magnetopause are labeled BL and MP, respectively. From top to bottom, the panels show the magnetic field magnitude; the proton phase-space density (PSD) relative to the maximum in the interval; the proton counts in each 8-s integration; the polar (θ) and azimuthal (φ) direction angles of the magnetic field, where θ = 0° is northward and φ = 0° is sunward; and the amplitude of the magnetic fluctuations in the passband 1 Hz to 10 Hz. Adapted from Anderson et al. (2011a).

completely understood, but transport in the cusp region via finitegyroradius effects may play a role (Section 16.6.3). Alternatively, Slavin et al. (2008) noted that the thickness of the layer corresponds approximately to the gyroradius of Na+ picked up in the solar wind, and they proposed that the layer is due to the Na+ pressure. If this is the case, then the boundary layer should appear in the morning for southward IMF and in the afternoon for northward IMF (Anderson et al., 2011a). Subsequent analysis of FIPS and Magnetometer data from Mercury orbit (Liljeblad et al., 2015) confirmed the presence of a dayside boundary layer at morning local times with a thickness commensurate with the initial estimates. However, Liljeblad et al. (2015) showed that the boundary layer is persistently present in the morning independent of the IMF north–south polarity, implying that Na+ pickup ions do not govern its formation. Additionally, the localtime asymmetry in the boundary layer, which is present in the morning but absent for local times after noon, may contribute to the prevalence on the dusk flank of Kelvin–Helmholtz (KH) waves (e.g., Liljeblad et al., 2014) arising from plasma flow shears across the magnetopause boundary, because the KH instability threshold is lower for thinner boundary layers and hence for greater velocity shear.

16.4.5 Plasma Depletion Layer The magnetosphere is directly impacted by the shocked solar wind in the magnetosheath (Figure 16.1) rather than by the solar wind itself, and this distinction is important for the dynamics of Mercury’s magnetosphere. Of particular significance is the formation of a plasma depletion layer (PDL), with decreased plasma density and increased magnetosheath magnetic field relative to the plasma immediately downstream of the shock (Zwan and Wolf, 1976; Denton and Lyon, 1996). The physical processes downstream of the bow shock that determine whether a PDL will form are governed by the solar wind environment, so we review the salient features of the solar wind at Mercury that distinguish it from nominal conditions at Earth. To date the most complete observations of the solar wind and IMF at Mercury orbit were made by the Helios spacecraft (Marsch et al., 1982). Some of the derived solar wind parameters that are most important for Mercury’s magnetosphere are illustrated in Figure 16.13 (Sarantos and Slavin, 2009). At 1 AU, the sonic and Alfvénic Mach numbers are generally in the range ~7−10, which places Earth, and the more distant planets, in the hypersonic and hyper-Alfvénic flow regime (Marsch et al., 1982). For these upstream conditions, the bow-shock jump conditions are near their


Structure of the Magnetosphere

Figure 16.13. Relative probability of occurrence of solar wind properties observed by Helios at heliocentric distances near Mercury perihelion (0.31 −0.35 AU). (a) The relationship between the Alfvén and sonic Mach numbers, MA and MS, respectively, governs the strength of the bow shock and conditions in the magnetosheath. Lines of constant plasma β are indicated. (b) Variation of MA with dynamic pressure, psw. Lines of constant IMF magnitude BSW are indicated. Adapted from Sarantos and Slavin (2009).

asymptotic limit (e.g., Spreiter et al., 1966b) and relatively insensitive to changing Mach number. In contrast, the Mach numbers at Mercury are much lower, ~2−5, as shown in Figure 16.13. The upstream conditions at Mercury therefore fall into the low-Machnumber regime in which the bow-shock jump conditions and its standoff distance are quite sensitive to changes in the solar wind and IMF. The six black lines shown in Figure 16.13 indicate relatively low solar wind β and strong IMF magnetic fields at Mercury’s orbit. Figure 16.13b displays the positive correlation between MA and Psw. Whereas typical solar wind dynamic pressures near Earth are ~1−2 nPa, Helios observed a peak in Psw in the ~10–20-nPa range near Mercury. The orbital phase of the MESSENGER mission was marked by generally weak solar activity, and Baker et al. (2013) used ENLIL solar wind simulations (Toth and Odstrcil, 1996; Odstrcil, 2003) to infer the solar wind conditions during the MESSENGER orbital mission phase. They determined typical dynamic pressure values of ~5−15 nPa, placing the solar wind conditions at Mercury in the low-MA regime. One of the most important effects of the low Alfvénic Mach numbers at Mercury is the formation of a PDL. The development of PDLs just upstream of planetary “obstacles” (i.e., a magnetopause or ionopause) was predicted from magnetohydrodynamic theory (Zwan and Wolf, 1976) and confirmed at Earth (Anderson et al., 1991; Fuselier et al., 1991; Phan et al., 1994). The effect of high or low Alfvénic Mach number on conditions in the magnetosheath is illustrated in Figure 16.14 (Gershman et al., 2013). As shown, the region of sub-Alfvénic flow expands greatly in the right-hand panel where the upstream Alfvén Mach number is low (Gershman et al., 2013). In fact, the degree of depletion as measured by plasma β and the extent of the PDL grow as the inverse square of MA (Zwan and Wolf, 1976). Gershman et al. (2013) conducted a detailed examination of the degree of depletion and the thickness of the PDL at Mercury as functions of MA and plasma β. Figure 16.15 illustrates their results for the PDL thickness as a function of the ratio between β just outside the magnetopause in the magnetosheath, βMP, to β just downstream of the bow shock, βMS. Although there is substantial scatter, the thickness of the PDL increases with decreasing βMP/βMS, in qualitative agreement with the

Subsonic MSH Sub-Alfvénic MSH





|B| n v vA vS (a) High Upstream MA

vA vS (b) Low Upstream MA

Figure 16.14. Comparison of schematic radial profiles in plasma speed and density and magnetic field intensity from the magnetosphere to the magnetosheath and the upstream solar wind for (a) high-solar-wind MA and (b) low-solar-wind MA; a portion of a typical MESSENGER orbit is shown as a red curve ending in a symbol of the spacecraft. With decreasing MA, a larger fraction of the subsolar subsonic magnetosheath (red shading) shows v < vA, i.e., is sub-Alfvénic, as indicated by the blue shaded regions in the cross sections. In addition, a thicker region of magnetic flux pileup is evident by an increase in B and a decrease in plasma density, n. The Alfvén speed (vA) and sound speed (vS) are shown as dashed blue and red lines, respectively, and the magnetopause and bow shock are labeled MP and BS, respectively. Adapted from Gershman et al. (2013).

predictions of Zwan and Wolf (1976). The characteristic length scale for the depletion layer at Mercury is 335 ± 49 km or ~ 0.1 RSS. As discussed in Chapter 17, one important consequence of a PDL is that the jump in magnetic field intensity across the

16.5 Current Systems and Magnetic Field Models


0.707 exp(–D/335 km)


θBN < 45º

βMP / βBS

θBN > 45º



103 D (km)

Figure 16.15. Depletion ratio βMP/βBS as a function of measured PDL thickness, D, for all orbits with βMP/βBS < 1/√2, regardless of upstream conditions. Events are classified by the shock angle, θBN, between the magnetic field and the shock normal direction. Those associated with quasi-parallel, θBN < 45°, and quasi-perpendicular, θBN > 45°, shocks are shown as yellow squares and blue circles, respectively. A best-fit exponential relationship (red line) is shown to match the data well. Adapted from Gershman et al. (2013).

magnetopause is less than the jump without plasma depletion (Phan et al., 1996; Anderson et al., 1997a). This pattern is particularly true during the impact of a coronal mass ejection (CME) onto the magnetosphere when the upstream Alfvénic Mach number at Mercury can approach unity (Sarantos and Slavin, 2009). Analyses of such impacts at Mercury by Slavin et al. (2014) indicated very high reconnection rates at the dayside magnetopause and extremely high numbers of flux transfer events, deep magnetospheric cusps because of intense plasma injections, and large numbers of flux ropes in the plasma sheet, even when the angle between the magnetosheath and magnetospheric magnetic fields at the magnetopause was substantially less than 90°. Further, DiBraccio et al. (2013) found that the magnetopause reconnection rate at Mercury is relatively insensitive to IMF orientation but increases as βMP decreases. Such a dependence, but over a smaller range of plasma β, has been observed at Earth (Scurry et al., 1994; Farrugia et al., 1995). For extreme conditions at Earth, Anderson et al. (1997a) showed that the reconnection rate should vary as the square of MA and suggested that the relative increase in magnetic field in the PDL facilitates near-subsolar reconnection for a broad range of magnetosheath field orientations relative to the magnetospheric field if component reconnection occurs (Sonnerup, 1984). In this way, it appears that the occurrence of subsolar reconnection for nearly all IMF orientations is likely a consequence of the prevalence of a PDL at Mercury.

1 6. 5 CU RREN T S YSTEMS AN D MA GN ETIC FIELD MODELS 16.5.1 Cross-Tail Current The stretching of the magnetic field in the magnetotail (Figure 16.1) implies a dawn-to-dusk current centered at the magnetic equator. Orbits from the first three Mercury years of

Figure 16.16. Stacks of measurements from 79 deep-tail current-sheet crossings with (red) and without (black) the dipole field removed for (a) tilt, θ = cos−1(Bρ/BρZ) in degrees, where BρZ is the magnitude of B projected into the ρ−Z plane, and (b) field magnitude (B) in nT. Each orbit is aligned on its equator crossing, ZEq, before stacking. (c) Stacks for 47 near-tail orbits (see text for selection criteria) showing B, BZ, and θ after removal of the dipole field. Adapted from Johnson et al. (2012).

observations with magnetic equator crossings within 3 h of local midnight were used to constrain the current-sheet halfthickness, DD, in the deep tail and the distance from the planetary spin axis to the inner edge of the current sheet, R2, in the near tail (Johnson et al., 2012). For the deep tail, the current sheet was indicated by a rotation in the field direction from dominantly anti-sunward in the southern tail lobe to dominantly sunward in the northern tail lobe. A mean thickness for the current sheet in this region was obtained from a superposed epoch analysis. The magnetic equator crossing for each orbit was assigned as the reference time, and the tilt of the field in the ρ–Z plane was calculated from θρZ ¼ cos1 ðBρ =BρZ Þ, where BρZ is the magnitude of B projected onto the ρ–Z plane; θρ–Z values close to 0° and 180° indicate anti-sunward and sunward field directions, respectively. The angles θρ–Z and the magnetic field magnitudes for the 79 selected deep-tail orbits were averaged, and the results are shown in Figures 16.16a and 16.16b. The rotation of the field direction from the southern to the northern lobe is clear (black curve), and after removal of the dipole field (red curve) the field is almost purely anti-sunward in the southern lobe and sunward in the northern lobe. The depression in field magnitude


Structure of the Magnetosphere

associated with the plasma sheet (Korth et al., 2011) is centered on the field reversal, although the magnetic depression is broader than the field rotation, indicating that the plasma sheet is thicker than the current sheet on average. The field rotation is 95% complete within 140 km (0.09 RM) of the current-sheet center, and this observation was used to define the current-sheet half-thickness in the far tail, DD = 0.09 RM. Orbits that sampled the near-tail region provide information on how close, on average, the current sheet comes to the planet (R2). These near-tail trajectories generally traversed the equator planetward of the cross-tail current. Depressions in the field magnitude, indicating spacecraft encounters with the plasma sheet, occurred on some but not all crossings of the near-tail region. Near-tail orbits with stronger plasma-sheet signatures were selected by evaluating the minimum value of the magnetic field strength near the equator crossing. Orbits with the strongest plasma-sheet signatures and deepest magnetic field minima should be those that passed closest to the current sheet. Superposed epoch averages were obtained for the 25% of the orbits with the lowest minimum field magnitudes (Figure 16.16c). The vector dipole field was subtracted to assess the field properties from external currents only, and averages of the residual field magnitude (B), Z-component (BZ), and polar angle (θρ–Z) were taken. The near-tail field, even for these orbits, was quite different from the far-tail field. First, θρ–Z came only within ~30° of the 180° or 0° direction, indicating that the orbits were, on average, planetward of the current-sheet and tail lobes. Second, the magnitude of the external field (after removal of the dipole field) increased rather than decreased near the equator crossing and was almost entirely in the −BZ direction, i.e., southward, at the equator. This behavior indicates that the external field in this region was dominated by the fringing field of the crosstail current planetward of the tail current sheet. Thus, these orbits passed close to but not through the current sheet. 1.41 RM was estimated Therefore, a lower bound on R2pof ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi from the mean radial distance ( X 2 þ Y 2 ) to these equator crossings. 16.5.2 Birkeland Currents The interaction of the solar wind and IMF with Mercury’s magnetic field drives the magnetic convection cycle, also known as the Dungey cycle (Dungey, 1963; Slavin et al., 2007), and signatures of magnetic reconnection at the magnetopause, in the cusp, and in the magnetotail all confirm this process at Mercury. This convection implies the imposition of an electrical potential on the open magnetic flux of Mercury’s magnetosphere, which is estimated to be in the range 15−30 kV (DiBraccio et al., 2013, 2015b). At Earth, the Dungey cycle is responsible for the field-aligned or Birkeland current system that conveys stress between the ionosphere and magnetosphere, and closure of these currents drives Joule dissipation (Cowley, 2000; Richmond and Thayer, 2000). The average Birkeland currents at Earth for southward IMF consist of two concentric upward/downward pairs of currents appearing approximately in arcs at constant magnetic latitude centered at dawn and dusk. The poleward pair of currents, denoted Region 1, is upward at dusk and downward at dawn, whereas the second pair, denoted

Region 2, is within ~5° latitude of Region 1 and has the opposite polarity, downward at dusk and upward at dawn (Iijima and Potemra, 1976; Anderson et al., 2008). Whether Mercury supports steady-state Birkeland currents without a conducting ionosphere was not known, and a variety of suggestions had been made for their configuration and closure at Mercury (Glassmeier, 2000; Ip and Kopp, 2004; Janhunen and Kallio, 2004). MESSENGER observations revealed that Birkeland currents corresponding to the terrestrial Region 1 polarity are present at Mercury (Anderson et al., 2014). To identify the signals of these currents, the magnetic residuals within Mercury’s magnetosphere over the northern hemisphere were calculated by removing a model field, Bm, that includes both an internal field, Bint, represented as an axially aligned, offset dipole (Anderson et al., 2012; Johnson et al., 2012), and an external field, Bext, accounting for magnetopause and magnetotail currents (Korth et al., 2015; and Section 16.5.3). Writing the total model field as Bm = Bint + Bext and the observed magnetic field as Bobs, the residuals are δB = Bobs − Bm. The δB vector components perpendicular to Bm projected onto the MSO X–Y plane as viewed from above the north pole and plotted along the orbit trajectory are shown for two sets of three sequential orbits in Figure 16.17 (Anderson et al., 2014). The sets of orbits are ~90 days apart, corresponding to slightly more than one Mercury year and ~1.5 planetary spin periods. In the upper and lower sets of orbits, the planetary orientation is nearly opposite: relative to its orientation in the upper set of orbits, the planet rotated ~200° counterclockwise for the bottom sets of orbits. The residuals poleward of 60°N are consistently sunward regardless of the planet’s orientation, and their magnitude varies by a factor of 2 or more from one orbit to the next. These features indicate that the signals reflect currents of external origin that vary in intensity on timescales comparable with or shorter than the MESSENGER orbit period. The signals were interpreted by Anderson et al. (2014) as fields associated with Birkeland currents flowing between altitudes above and below the spacecraft: an upward current in the evening and a downward current in the morning, corresponding to the poleward Region 1 currents documented at Earth. To derive field-aligned current densities, the portion of δB parallel to Bm was subtracted to obtain the transverse residual, δB⊥ (Anderson et al., 2014). These data were averaged to obtain maps of δB⊥ for each year of orbit operations, binned by magnetic disturbance level (Anderson et al., 2013). The fieldaligned current density threading the mean “orbit surface” for each year was calculated from the curl of δB⊥ and mapped to the planetary surface along Bm to obtain the surface radial current density, jrS. Figure 16.18 shows maps of jrS versus latitude and local time for low, moderate, and high levels of magnetic disturbance for the first three years of orbit operations obtained from data acquired during the descending orbit segments (Anderson et al., 2016). Results for ascending orbit segments are essentially the same. The increase of current with activity is clear, as is the basic structure of an upward current in the dusk sector and a downward current in the dawn sector. The total average currents range from just under 20 kA for quiet conditions to nearly 40 kA during disturbed conditions (Anderson et

16.5 Current Systems and Magnetic Field Models


Figure 16.17. Magnetic perturbations recorded by MESSENGER as viewed from above Mercury’s north pole with the Sun to the right during 21–22 January 2012 (upper panels) and 21–22 April 2012 (lower panels). The spacecraft trajectory is shown in red below 1000-km altitude. Magnetic residuals perpendicular to the total model field and projected onto the X–Y plane, δBXY, are plotted at 12-s intervals and shown by colored lines originating at the observation point. The directions correspond to the δBXY direction, and the color and length indicate |δBXY| (see color bar and blue reference arrow at upper left). Start and end times are given by day of year and UTC. Adapted from Anderson et al. (2014).

Birkeland Currents at Mercury vs Magnetic Disturbance Level +Y′



nA/m2 200 100 0 –100 –200 +X′ 60º



VSolar Wind Dist. level: 0–20%



Figure 16.18. Average Birkeland current densities in aberrated MSO coordinates determined from descending orbit segments of MESSENGER Magnetometer data from 23 March 2011 through 31 March 2014 and mapped to the planetary surface. Upward currents are indicated in red and downward currents in blue, with contours every 20 nA m−2 and current densities as indicated by the color bar. From left to right, panels show current density for the quietest 20% of orbits, middle disturbance level (40−60%), and most magnetically disturbed orbits (80−100%), where the disturbance level is the magnetic activity index for Mercury of Anderson et al. (2013). From Anderson et al. (2016).

al., 2014), a factor of ~100 lower than the Birkeland currents at Earth (Richmond and Thayer, 2000). Significantly, there is no evidence of a Region 2 current system at Mercury, consistent with the expectation that Mercury’s magnetosphere does not support a ring current plasma population, which is central to the Region 2 currents at Earth (e.g., Slavin et al., 2007).

Although the convection potential applied to Mercury’s magnetosphere, ~30 kV, is lower than that at Earth, which is typically ~100 kV during active conditions, the decrement is only a factor of 3, whereas the currents are weaker by a factor of 100, indicating a much greater electrical resistance to their closure at Mercury. Anderson et al. (2014) used a simple single spherical


Structure of the Magnetosphere

shell model for interior electrical conductivity to show that the observed current distribution and estimated total potential are consistent with current closure through the planet, as proposed by Janhunen and Kallio (2004) for a nominal conductivity structure (e.g., Verhoeven et al., 2009). Anderson et al. (2016) extended this work to allow a general radially varying conductivity and analyzed a range of possible conductivity–depth profiles to show that the initial result holds generally, indicating that the current most likely does close through the planet and between 50% and 90% of the current closes through the core. In addition to core induction effects on the magnetosphere (Johnson et al., 2016), this coupling between Mercury and its magnetosphere is unique in the solar system in that the solid planet itself is an integral part of the magnetospheric electrodynamic system. 16.5.3 Magnetic Field Models A consequence of Mercury’s weak internal magnetic field is that external fields generated by magnetospheric current systems contribute substantially to the observed magnetic field throughout the small magnetosphere. Accurate modeling of the magnetospheric magnetic field is therefore important to analyzing planetary fields arising from the internal dynamo, crustal magnetization, and induction effects. Because current systems are present throughout the magnetosphere, spatially distributed observations are required to characterize them. MESSENGER’s sampling of the magnetic field in the inner magnetosphere was sufficiently dense to allow the magnetic field at spacecraft altitudes less than ~800 km to be predicted to better than 10% accuracy (Johnson et al., 2012; Korth et al., 2015). Two models were developed over the course of the MESSENGER mission, with the first derived early in the orbital mission phase to enable other analyses and the latter taking advantage of the full mission data. The first magnetospheric model developed was the paraboloid model (Alexeev et al., 2010; Johnson et al., 2012). The paraboloid model represents the magnetospheric magnetic field as the sum of the contributions from sources internal and external to the planet. It includes the internal magnetic field of the offset dipole, Bint, and external magnetic fields, Bt and Bcf, generated by cross-tail and magnetopause currents, respectively. The total magnetic field, Bm ¼ Bint þ Bt þ Bcf ;

radius, rMP, to the Shue et al. (1997) magnetopause (Section 16.3). Finally, the properties of Mercury’s axisymmetric internal dipole field, which is offset by 0.196 RM to the north (Anderson et al., 2011b, 2012), are described in Chapter 5. Using the above parameters as a priori constraints in the paraboloid model, the dipole moment, μ ¼ 190 nT R3M , was obtained by minimizing the root mean square (rms) value of the residual magnetic field, sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi i N h X δB ¼ ðBX  Bm;X Þ2 þ ðBY  Bm;Y Þ2 þ ðBZ  Bm;Z Þ2 =N ; i¼1

ð16:5Þ between the components Bi of the magnetic field observed within the magnetosphere and the components Bm,i of the model field obtained for the set of N observations (Johnson et al., 2012). Magnetic field lines of this model are shown as blue dashed lines in Figure 16.19, and the gray solid and dashed ellipses indicate the observational sampling of the MESSENGER orbit. The most serious limitation of the paraboloid model is the shape of the model magnetopause (Figure 16.19), which deviates substantially from the observed magnetopause and flares too much with anti-sunward distance from the planet. The result is that the predicted contribution from magnetopause currents is too small at high northern latitudes. Less serious but also problematic is the sharp inner edge of the cross-tail current sheet in this model, which yields magnetic islands at which magnetic field lines loop back on themselves rather than map to the planet. The rigid analytical formalism of the paraboloid model allows neither modifications to resolve these discrepancies nor


is confined within a magnetopause prescribed by a paraboloid of revolution. All except one of the parameters describing the model can be established directly from MESSENGER observations. The parameters describe the magnetopause shape, geometry of the cross-tail current sheet, magnetic flux in the magnetotail, and dipole location and orientation. In this model, the magnetopause is represented by a paraboloid [Section 16.3, equation (16.3)]. The cross-tail current sheet of the model is parameterized by its thickness, 0.09 RM, and the location of its inner edge, 1.41 RM (Section 16.5.1). The mag2 =2 ¼ 2:6 MWb, was estimated from the netic flux, F ¼ BπrMP average magnetic field magnitude, B, in the magnetotail and the

Figure 16.19. Magnetic field lines of the paraboloid (dashed blue) and KT14 (solid black) models in the MSM X−Z plane confined within the observed average magnetopause (red), modeled after Shue et al. (1997) using best-fit parameters determined from MESSENGER Magnetometer observations. The planet is shown as a circle with dayside and nightside in orange and black, respectively, and MESSENGER orbits 2723 and 2822 are shown as solid and dashed gray lines, respectively. The model magnetopause for the paraboloid model is indicated by the dashed red line.

16.5 Current Systems and Magnetic Field Models extensions to include additional current systems, such as the Birkeland currents discussed above. The extensive orbital observations provided the basis to model Mercury’s magnetospheric magnetic field while allowing for an arbitrary magnetopause shape, a representation of the magnetic field that is continuous in space, in particular in the near-tail region on the nightside, and use of a modular, expandable model framework. The development of this model, termed KT14 (Korth et al., 2015), followed the data-based approach to generate magnetic shielding fields used for Earth’s magnetosphere (e.g., Tsyganenko, 2013). In the absence of reconnection, these shielding fields eliminate the normal component of magnetic field, Bn, at the magnetopause and separate interplanetary and planetary magnetic field lines. The numerical approach to minimizing Bn can be applied to determine the shielding field and, hence, the magnetopause currents for an arbitrary magnetopause shape. The shielding field is defined as Bcf ¼ rU;


where the scalar potential, U, is represented by Cartesian harmonic basis functions in the MSM frame: U¼


aik exp

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  p2i þ p2k X cosðpi YÞ sinðpk ZÞ:



The N2 linear coefficients aik and N non-linear coefficients pi are obtained by minimizing the RMS residual of the magnetic field component normal to the boundary surface: vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi u M uX ð16:8Þ σ ¼ t ½ðBj  rUÞ∙ n=M; j¼1

where M is the number of data points at the boundary surface. The boundary surface is given by the Shue et al. (1997) magnetopause with best-fit parameterization (Winslow et al., 2013) (Section 16.3). The shielding field is defined separately for each magnetospheric source field, thus ensuring modular extensibility of the model. To provide a spatially continuous representation, the magnetic field of the cross-tail current sheet in the KT14 model is composed of two distinct contributions. In the far magnetotail, ≳10 RM anti-sunward, the current density at the magnetic equator is held constant with radial distance but decreases to the north and south and vanishes at the northern and southern edges of the current sheet. Closer to the planet on the nightside, ~(1.5–5) RM from Mercury’s center, the current density is higher, corresponding to higher magnetic field strength in the lobes. The resulting current distribution is represented with a disk-shaped current sheet (Tsyganenko and Peredo, 1994) in the near magnetotail and a sheet-shaped current farther tailward. The disk current, I, is expressed as a piecewise continuous function of radial distance, pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ρ ¼ X2 þ Y 2 : IðρÞ ¼ 0

  π ρ  ρ1 IðρÞ ¼ Im sin ρ2  ρ1 ρ 2  ρ 2 IðρÞ ¼ Im exp L 2

ρ < ρ1 ; ρ1 < ρ < ρ 2 ; ρ > ρ2 :



In equation (16.9), Im is the peak current, ρ1 and ρ2 are the cylindrical radial distances to the inner edge of the current sheet and the peak current, respectively, and the e-folding scale L imposes a decrease of the current for large ρ. This form provides a smooth inner edge that avoids magnetic islands. Consistent with the paraboloid model, the inner edge of the current sheet is ρ1 = 1.41 RM, and ρ2 = 1.56 RM and L = 1.43 RM were chosen a posteriori to yield a gradient near the inner edge as steep as could be fit using the procedure below and a decay to Im/10 at a radial distance of 5 RM. These geometrical approximations were introduced to avoid an unphysically sharp inner edge of the cross-tail current. A magnetic vector potential, A, was then sought, so that the magnetic field of the disk current, Bd ¼ t1 r  A;


corresponds to that of the current represented by equation (16.9), where the amplitude parameter t1 is proportional to the strength of the cross-tail current. For distances >5 RM tailward, the current disk merges into a sheet current having a vector potential of the form A = (0, AY, 0), where z AY ¼ 2 t2 ln cosh ; ð16:11Þ d and where t2 is the current amplitude, d is the half-thickness of the current sheet, and z is the distance to the current sheet. The disk and the sheet currents are both centered on the magnetic equator and widen infinitely toward the dayside magnetosphere, where the cross-tail current vanishes. As in the paraboloid model, the magnetic field is composed of the sum of internal and external contributions. The KT14 model parameters for the dipole moment and offset and the nominal current-sheet half-thickness are identical to those of the paraboloid model. The amplitude parameters of the disk and sheet currents, t1 and t2, respectively, were obtained by minimization of the RMS residual of the magnetic field observed within the magnetosphere and corrected for aberration. The expansion of the tail current-sheet thickness in MSO X and Y directions was set empirically to match the boundary between open and closed magnetic field lines inferred from plasma observations (Korth et al., 2014). The model was developed from a data set acquired over seven Mercury years with an updated value of the subsolar standoff distance of RSS = 1.42 RM (Winslow et al., 2013). The magnetic field lines of this model are shown in Figure 16.19, with solid traces indicating the confinement within the observed magnetopause shape. The KT14 model RMS residual is δB = 24.8 nT, and the resulting model satisfies ∇∙B = 0 in the magnetosphere. Figure 16.20 shows the spatial distributions of magnetic field residuals in the radial (a and d), southward (b and e), and eastward (c and f) components with respect to the KT14 model. For the analysis period, the ascending (a–c) and descending (d–f) orbit legs correspond to higher- and lower-altitude observations, respectively. The magnetic field residuals exhibit several systematic features. The overall structures of the residuals obtained from the KT14 and paraboloid models were found to be very similar (Korth et al., 2015), reflecting in part the limitations of the MESSENGER data distribution in prescribing the nightside external fields. The largest


Structure of the Magnetosphere

Figure 16.20. Magnetic field residuals with respect to the KT14 model in the (a and d) radial, (b and e) southward, and (c and f) eastward directions for ascending (a–c) and descending (d–f) orbit legs averaged over the period 24 March 2011 to 23 November 2012. Grid lines are labeled in aberrated local time and MSM latitude, and the color bar denotes the magnitude of the residuals. In each panel, the latitude range over which MESSENGER crossed the magnetopause on the dayside is delineated by the dashed lines, and the average crossing latitude is indicated by the solid line. Adapted from Korth et al. (2015).

KT14 residuals are in δBr and δBθ near the dayside magnetopause and may originate from a dayside boundary layer (Section 16.4.4). Similar residuals associated with the presence of plasmas are observed in the northern cusp and in the plasma sheet near the magnetic equator. Because neither the KT14 model nor the paraboloid model include the diamagnetic effects of the plasma populations, these models cannot account for local magnetic field variations associated with the maintenance of total pressure balance in these regions. Second, large-scale residuals in δBϕ, which are directed eastward at dawn and westward at dusk, are found on both the ascending and descending orbit legs and extend from ~30°N to the north pole. Their distribution is consistent with the existence of steady field-aligned Birkeland currents (Section 16.5.2). Third, residuals in the downward (−δBr) and poleward (−δBθ) directions are observed in a partial ring structure extending between latitudes 30°N and 60°N; the structure is especially pronounced on the descending orbit legs. These residuals indicate additional magnetic field sources not represented by either model.

1 6. 6 P L A S MA 16.6.1 Sources Mercury is embedded in the most dense and most dynamic solar wind of any planet in our solar system, and

MESSENGER provided critical insights regarding both the dynamics of solar plasma acceleration and the effects of these unique solar wind conditions on Mercury’s magnetosphere. At the Sun, energy transfer from the solar magnetic field accelerates solar plasma within the magnetically dominant solar atmosphere, the solar corona, to produce the solar wind that fills the entire heliosphere. The solar wind outflow reaches supersonic and super-Alfvénic speeds at heliocentric radial distances between (2−10) RS and (10−20) RS (where RS = 695,700 km is the nominal solar radius), respectively. In situ observations in the inner heliosphere by MESSENGER provided evidence that this heating process is both transient in time and highly spatially dependent, especially near Mercury’s orbit (Gershman et al., 2012). With the exception of He ions, the fast solar wind associated with the solar poles and coronal holes is nearly photospheric in ion composition. Slow and transient streams from CMEs are enriched relative to the photospheric composition in ions with a first ionization potential (FIP) 1 MK. CME-associated plasmas contain both the ions with the

16.6 Plasma hottest electron temperatures (up to 3 MK) and also those with the coolest temperatures, associated with erupting solar filaments ( 30, such as S+, and doubly charged ions in the range 4.5 ≤ m/q ≤ 12 were detected in the plasma sheet at a radial distance between 1.6 RM and 2.2 RM during MESSENGER’s first flyby of Mercury (Zurbuchen et al., 2008). The distribution of planetary ions varies substantially by region, as shown in Figure 16.22. Enhancements of these ion species are evident at both dawn and dusk, although the variations with altitude are different.

Figure 16.23 compares the abundances of major ion species in distinct locations within the magnetosphere. Notable findings are that, in the northern cusp, the abundance of Na+-group ions exceeds that of solar wind alpha particles by a factor of 2 on average (Figure 16.23a), and, in the central plasma sheet (Figure 16.23b), planetary ions can contribute up to 50% of the mass density and 15% of the plasma thermal pressure (Gershman et al., 2014). Compared with the post-midnight sector, Na+-group ions in the pre-midnight plasma sheet, with typical energies of a few keV, are substantially enhanced over a broad altitude range, ~1500−6000 km (Figure 16.22, plasma population 4) (Raines et al., 2013; Gershman et al., 2014). This observation is consistent with test particle simulations by Delcourt (2013), which showed that Na+ and O+ ions are preferentially transported into this altitude range on the pre-midnight side of the central plasma sheet by the mechanism detailed in Section 16.6.3. There also appears to be a dependence of ion abundance on solar wind conditions as manifested in a seasonal enhancement in both Na+-group and O+-group ions (Raines et al., 2013). Both species show enhancements by a factor of more than 2 for measurements acquired at Mercury true anomalies of 120° and 315°. The extrema in ion abundance do not strictly correlate with Mercury’s heliocentric distance, which is known to modulate this planet’s solar wind environment (Korth et al., 2012), so an as yet unidentified physical process must govern the seasonal variation of these ion populations. He+ ions do not show such enhancement, presumably because these ions are partially sourced by the solar wind.

16.6 Plasma The enhancement of Na+-group ions in the dawn-side magnetosphere (Figure 16.22b, plasma population 3) is collocated with a persistent source of high-energy neutral Ca observed by MESSENGER’s Ultraviolet and Visible Spectrometer (McClintock and Lankton, 2007) in the exosphere (Burger et al., 2012), although a direct link with this source has not been established. Similarly, seasonal enhancements in the ion population noted above do not correlate with those of the exospheric neutral atoms, which exhibit much lower variability (Cassidy et al., 2015). To examine the relationship between the neutral and ionized components, the e-folding heights of planetary ions were compared with the scale heights of exospheric neutral atoms, and the former were typically found to be much larger (Raines et al., 2013). The e-folding distance of Na+-group ions is a factor of 5–10 greater than the Na scale height (Cassidy et al., 2015). This difference implies that unlike Na, which is cool and gravitationally bound to the surface, the Na+ ions are more energetic and may escape to high altitudes, e.g., into the magnetosphere. Electrons Magnetospheric electrons with energies >35 keV were routinely observed by the Energetic Particle Spectrometer (EPS) (Andrews et al., 2007), beginning shortly after Mercury orbit insertion (Ho et al., 2011b). The energetic electrons were registered as bursts lasting from seconds to hours with typical energies up to 100 keV, although energies >200 keV were measured occasionally. The events were most often observed close to the planet in the northern hemisphere from the magnetic equator to the pole, but they were present over a broad range of local times (Ho et al., 2012). The pitch-angle distributions of the electrons suggest that the bulk of the population does not execute complete drift paths around the planet (Ho et al., 2011b) and thus does not form long-lived


radiation belts such as those found at Earth. As discussed in Section 16.6.3, low fluxes of ~200 keV electrons detectable by the Gamma-Ray Spectrometer (GRS) sensor (Goldsten et al., 2007) did on occasion exhibit periodicities consistent with the azimuthal drift time of electrons around the planet (drift echoes) but with lifetimes of not more than approximately five complete drifts around the planet (Baker et al., 2016). Interestingly, no energetic ions with energies >25 keV were detected above the EPS detector intensity threshold even though 1-keV ions are prevalent in Mercury’s magnetosphere (Ho et al., 2011b). The energy range of the EPS sensor was designed to sense energies observed by Mariner 10 (>35 keV). Fortuitously, three other sensors of the MESSENGER payload – the XRay Spectrometer (XRS) (Schlemm et al., 2007), the Neutron Spectrometer (NS) (Goldsten et al., 2007), and the GRS – all returned signals from orbit about Mercury attributable to energetic electrons with energies that EPS was not designed to measure. The XRS instrument observed photons resulting from low-energy (~10 keV) electrons impinging on its detectors approximately uniformly throughout the magnetosphere (Ho et al., 2011a). Although the energy spectra could not be measured directly by XRS, simulation of the XRS detector response indicates an energy spectrum that peaks in the range 0.7–1.0 keV and has a flux consistent with that observed by EPS at 45 keV. The spatial distribution of the electrons inferred from XRS observations is shown in Figure 16.24 (Ho et al., 2016). The events spanned all local times, with the highest concentration near dawn and dusk, and, similarly to the proton distribution (Figure 16.21), were clustered within a narrow latitude band located at higher latitudes on the dayside than on the nightside. The GRS and NS detectors responded to bremsstrahlung photons produced when energetic electrons impacted materials nearby and were found to be substantially more sensitive to this

Figure 16.24. Distribution of suprathermal electron events detected by the MESSENGER X-Ray Spectrometer (XRS) organized by latitude and local time. Adapted from Ho et al. (2016).


Structure of the Magnetosphere

population than the EPS instrument, owing to their comparatively large geometric factors for responding to electron impacts. These sensors regularly detected energetic electrons with energies from tens to hundreds of keV, mostly within the magnetosphere and predominantly on closed magnetic field lines (Lawrence et al., 2015). A subset of these energetic electron events showed periodicities in their occurrence rates of a few minutes, consistent with Mercury’s Dungey cycle period (Section 16.6.3), indicating that they may be related to reconnection or magnetospheric convection. 16.6.3 Transport and Heating Interaction of Mercury’s planetary magnetic field with the solar wind and IMF gives rise to circulation of magnetic flux (Dungey cycle, Section 16.5.2) and the corresponding convection electric field, E, and plasma circulation drifts (e.g., Hughes, 1996). This circulation was directly observed by MESSENGER (DiBraccio et al., 2015a) in the plasma mantle, a persistent layer of tailwardflowing magnetosheath-like plasma inside of and adjacent to the magnetopause (Rosenbauer et al., 1975), and is similar to that observed at Earth but is one to two orders of magnitude faster, i.e., with cycle times of minutes instead of hours at Mercury (Slavin et al., 2009a, 2010). In the region of the magnetosphere dominated by the dipole field, ions and electrons gyrate about magnetic field lines of force and bounce along magnetic field lines between the northern and southern hemisphere while they also drift from the nightside to the dayside. The actual drift trajectories depend on the strength of the convection electric field and on the particle’s energy, charge, and pitch angle. Because of Mercury’s weak planetary field, the region of quasidipolar magnetic field configuration is very close to the planet, within 1 RM, and everywhere else the magnetic field gradient and curvature drifts are negligible relative to convective E × B drift motion. Within ~1 RM altitude, the gradient-curvature drift might be important and, in principle, a very narrow annulus near the magnetic equator close to the planet might support drift motions that close on themselves around the planet. The character of charged particle motion in electromagnetic fields depends on the ratio between the gradient scale of the magnetic pfield ffiffiffiffiffiffiffiffiffiffiffiffiffi and the particle gyroradius, rg ¼ mv? =ðjqjBÞ ¼ 2mW?=ðjqjBÞ, where m is the mass, v⊥ is the particle velocity perpendicular to the magnetic field, W? ¼ mv2? =2, and q is the particle charge. When the magnetic field varies on scales much greater than rg, particles conserve the magnetic moment of their gyromotion, known as the first adiabatic invariant, μ ¼ ðmv? Þ=ð2BÞ, and the particle transport can be described in the guiding center approximation, by which one considers the drift motion only of the gyrocenter subject to the E × B and gradient-curvature drifts. In the magnetotail, the minimum radius of curvature of the magnetic field on a given field line, Rc, occurs at the cross-tail current sheet within the plasma sheet (Sections 16.4.2 and 16.5.1). This location is also where the magnetic field is a minimum and rg is a maximum. For this reason, the magnetotail equator is a localized region of nonadiabatic motion within which the guiding center treatment breaks down, and one must evaluate particle motions from the Lorentz force explicitly.

Non-adiabatic behavior is most important for the heavy ions, e.g., Na+, because, pffiffiffiffi for a fixed energy, the gyroradius is proportionalpffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi to m. For a given field line, the parameter κ ¼ Rc =rg;max , where rg,max is the maximum gyroradius on the field line, is a useful measure of non-adiabatic behavior (Büchner and Zelenyi, 1989). The motion is generally adiabatic for κ > 3, whereas for κ < 1 particles experience meandering motions about the field minimum (Speiser, 1965; Delcourt and Martin, 1994). For intermediate κ, particles have their magnetic moments quasi-randomly altered with each crossing of the magnetic equator, a process referred to as μ-scattering (Birmingham, 1984; Anderson et al., 1997b; Delcourt et al., 2003). For Mercury, the guiding center drift approximation is violated throughout most of the magnetosphere, and stochastic processes must be considered in plasma-sheet transport and heating (Korth et al., 2012). Korth et al. (2011) showed that plasma-sheet protons with energies in the range 0.5–5 keV have 1 < κ < 3, and their motion is non-adiabatic throughout much of the nightside. Heavy ions are more strongly nonadiabatic because their gyroradii are larger by a factor of ~20 than those of protons, so that Na+ motions will be fully chaotic with κ < 1. Furthermore, non-adiabatic transport extends closer to the surface for higher ion energies. For 5-keV protons, the first adiabatic invariant fails to be conserved even within 500-km altitude at midnight, implying that wave–particle interactions are not required to scatter protons into the loss cone because μ-scattering occurs simply by virtue of nonadiabatic motions (e.g., Anderson et al., 1997b). By contrast, electrons should remain adiabatic even to fairly high energies, e.g., hundreds of keV. At Mercury, bombardment of the surface by charged particles is expected to yield low-energy heavy ions of planetary origin (O+, Na+, or even Ca+) near the surface. Remarkably, a lowenergy, 5-keV Na+ ions in the plasma sheet, with a strong bias to the evening side as observed (Figure 16.22), is additional evidence for this process because such an evening bias is a prediction of the mechanism. In the magnetotail, plasma drifts from the lobes into the magnetic equatorial plane to form the plasma sheet.

Planetward of the magnetotail reconnection site, the return convection transports plasma sunward toward the planet. At Earth, the magnetosphere is ~10 times larger than the planet and the region within ~10 RE is quasi-dipolar, so that gradient curvature drifts grow to dominate the E × B convection as plasma convects Earthward from the tail. As a consequence, plasma is diverted around the Earth in the distance range 5 RE to 8 RE. In addition, Earth’s rotation rate is 59 times greater than Mercury’s, so the corotation electric field is much stronger at Earth and dominates the drift for particles with energies of ~1 keV and lower within ~6 RE. At Mercury, the cross-tail current is very close to the planetary surface, the region of quasi-dipolar field is within 200 particles cm−2 s−1 sr−1 keV−1 at 45 keV) during the first year of MESSENGER observations are projected onto two orthogonal planes in MSO coordinates. The circles denote the location of MESSENGER at the time of the peak intensity during each event. Circle size increases with event intensity, and the color represents the power-law index fit to the energy spectrum. The average magnetopause and bow shock surfaces are shown in red and blue, respectively; the planetary outline is shown in black. From Ho et al. (2012).

are compressionally dominant. Transverse-dominant waves are also observed, but these are typically seen farther from the magnetic equator. The wave power maximizes at the equator and peaks in the occurrence frequency in the pre-midnight and post-midnight sectors, with a clear predominance for dusk-side waves (Boardsen et al., 2012). Boardsen et al. (2015) interpreted these characteristics as a form of ion-Bernstein-mode wave. As these waves propagate back and forth across the magnetic equator, the polarization and transmission properties change with plasma β, and the waves cycle between two different branches of the instability, one with high compression at the

magnetic equator, and one for which the compression maximum occurs off, and symmetric to, the equatorial plane (Boardsen et al., 2015). These properties agree well with what is seen in the observations. In addition to the narrow-band harmonic waves, there is also evidence for quasi-periodic waves at frequencies of ~0.1 Hz, i.e., below the typical ion cyclotron frequency. These waves are likely externally driven, as they have been observed in association with both Kelvin–Helmholtz activity on the magnetopause and large-amplitude magnetosheath waves downstream of the quasi-parallel bow shock (Sundberg et al., 2012a, 2013).

Mercury’s Dynamic Magnetosphere

Electron Flux (particles/cm2/s/sr/keV)


Time (UTC) Distance (RM) Local Time Latitude

36 – 57 keV 57 – 89 keV 89 – 140 keV




101 02:30 2.1 32 12:31 10.895



02:45 1.555 12:45 37.127


3 Electron Fluence (particles/cm2/sr/keV)

2 03:00 ZMSO (RM)

1 0 –1


–3 –2



03:15 1.279 00:15 23:592

22 December 2011 02:50–03:05 UTC




102 –3



Electron Event


03:00 1.132 18:16 83.812











7 8 9 100


Energy (keV)

Figure 17.25. The intensity versus time, event-averaged energy spectrum, and location for a pair of two closely spaced high-latitude energetic electron bursts on 22 December 2011. From Ho et al. (2012).

17.4 ENERGETIC ELECTRON BURSTS After Mercury’s global planetary magnetic field, arguably the most surprising Mariner 10 mission discovery was that this small magnetosphere is a source of intense bursts of energetic (>35 keV) charged particles (Simpson et al., 1974; Eraker and Simpson, 1986). Unfortunately, instrumental effects made an unambiguous determination of species, flux, and energy spectrum for the Mariner 10 events impossible (Armstrong et al., 1975; Christon et al., 1979). For this reason, definitive measurement of the properties and acceleration processes for the energetic particles in Mercury’s magnetosphere has long been a priority for the planetary magnetosphere community (Sundberg and Slavin, 2015; Seki et al., 2015; Raines et al., 2015). Data from the MESSENGER EPS have shown that these energetic particle bursts are composed entirely of electrons (Ho et al., 2011). EPS made measurements of these electrons from ~30 to 300 keV energy during its 3-s scans. The durations of the energetic electron bursts ranged from several minutes to

nearly an hour, and they tended to be centered on the time of the spacecraft’s closest approach to the planet. The energy, E, of these electrons sometimes exceeded 200 keV, but usually the energy distributions exhibited a cutoff near E = 100 keV. However, no ions with energies >35 keV were detected by EPS anywhere in Mercury’s magnetosphere, and no evidence of a stably trapped high-energy charged particle population was found. This latter result was not a surprise, because Mercury’s magnetic field is weak, and the dayside magnetosphere is too small, to support the stably trapped radiation belts found at Earth (e.g., Shriver et al., 2011a; Walsh et al., 2013). An energy–time spectrogram for the EPS electron measurements covering one month early in the orbital phase of the MESSENGER mission is shown in Figure 17.23 (Ho et al., 2012). Electron bursts were detected during each of the 12-h orbits from 22 September to 22 October 2011. Two SEP events (on 22 September and 4 October) occurred during this interval, and they are identified by their characteristic high energies (>100 keV) and several-day durations. Some of the events reported by Ho et al. (2012) occurred near the magnetic equator,

17.4 Energetic Electron Bursts


Figure 17.26. An electron burst event near local midnight on 21 November 2013 displayed in (a) GRS, (b) NS, and (c)–(f) Magnetometer measurements in MSM coordinates. A ~1-minute-long growth phase denotes loading of the tail, and a sudden increase in the BZ magnetic field coordinate (vertical dashed line) marks the onset of a dipolarization event in the near tail. From Baker et al. (2016).

but most were concentrated at high northern latitudes, near spacecraft periapsis. As shown in Figure 17.24, these electron events appeared mainly around local midnight near the magnetic equator and at high latitudes near the magnetospheric cusp region on open as well as closed magnetic field lines. Two examples of the energetic electron bursts measured by EPS observed on 22 December 2011 are displayed in Figure 17.25 (Ho et al., 2012). Typically, these events were

brief, lasting a few seconds to a few minutes. They have pancake pitch angle distributions and power-law energy spectra for energies above ~50 keV. The first event in Figure 17.25 was detected at ~02:55 UTC and lasted ~1 min. The second burst began at 03:01 UTC and continued for about 3 min. Multiple events are often seen during a single orbit, as in this example. The energy spectrum shown bends over at energies below ~50 keV, and it is well fit with a kappa function. However, these


Mercury’s Dynamic Magnetosphere

events are usually characterized by power law indices of ~1.5–4.5 (Ho et al., 2012). The Gamma-Ray and Neutron Spectrometer (GRNS), with separate Gamma-Ray Spectrometer (GRS) and Neutron Spectrometer (NS) sensors, was sensitive to electrons with E > 50 keV, and it made measurements with temporal resolutions from seconds to 10 milliseconds (Goldsten et al., 2007). Lawrence et al. (2015) conducted a survey of energetic electron burst properties as observed at Mercury with the NS sensor covering the period from orbit insertion through 31 December 2013. The NS had a much larger geometric factor than EPS and a time resolution of 1 s or 20 s. It responded primarily to incident electrons in the energy range ~20–40 keV. The Lawrence et al. (2015) study identified 2711 electron events and surveyed their temporal, spatial, and spectral behavior. The duration of the events ranged from tens of seconds to nearly 20 min, and the events were classified as “bursty” (large amplitude variation with time during an event) or “smooth.” Almost all events were detected inside Mercury’s magnetosphere on closed field lines. The bursty events were observed most frequently near dawn, whereas the smooth events were detected most frequently in the midnight sector at lower latitudes. Some of the NS events exhibited periodicities from hundreds of seconds to tens of milliseconds. Lawrence et al. (2015) attributed the short-period variations to particle dynamics such as north–south bouncing within dipolar magnetic flux tubes, whereas the longer few-minute variations were interpreted as substorm injection events in Mercury’s magnetic tail, similar to the Mariner 10 burst events (e.g., Siscoe et al., 1975; Baker et al., 1986; Christon et al., 1987; Ip, 1987; Delcourt et al., 2005). The GRNS gamma-ray sensor was cryocooled and measured gamma rays in the energy range ~50 keV to 10 MeV (Goldsten et al., 2007). It was used primarily for Mercury surface composition measurements. The GRS operated in this mode until June 2012 when the cryocooler failed after reaching its expected lifetime. The GRS sensor incorporated a borated plastic anticoincidence shield (ACS) surrounding its germanium detector. The ACS was sensitive to electrons with energies from ~50 keV to several hundred keV. Beginning on 25 February 2013, the germanium detector’s telemetry was re-allocated to the ACS system so that it could provide near-continuous measurements of energetic electrons with 10-ms time resolution. Baker et al. (2016), building on the initial analyses of the high-resolution GRS data by Lawrence et al. (2015), conducted a detailed examination of intense electron bursts in the high-time-resolution ACS data collected from 1 March 2013 to October 2014. Energetic electrons measured with FIPS during SEP events were also observed in closed-field regions of the magnetotail and were consistent with the average spatial distribution of events from GRS (Gershman et al., 2015b). The acceleration of electrons at Earth is often closely associated with the formation of reconnection X-lines in the cross-tail current sheet (see the review by Birn et al., 2012). The magnetic flux loading and unloading of the tail lobes, the dipolarization of the plasma sheet magnetic field as it piles up at the inner edge of the tail, and the formation of flux ropes in the cross-tail current layer at Earth are all frequently observed at Mercury. Inductive electric fields resulting from the rapid

reconfiguration of the magnetic field at X-lines, and Fermi acceleration in flux ropes as they contract, can readily accelerate electrons to energies of hundreds of keV on very short timescales (Sarris et al., 1976; Baker and Stone, 1977; Richardson et al., 1996; Øieroset et al., 2002; Hoshino, 2005; Drake et al., 2006; Ashour-Abdalla et al., 2011; Birn et al., 2012). Given the high frequency and intensity of reconnection at Mercury (Slavin et al., 2009a, 2010b; DiBraccio et al., 2013, 2015a), these processes are the most likely acceleration mechanisms for energetic electrons at Mercury. An example of one of the energetic (~100–200 keV) electron injection events in the GRS and NS data studied by Baker et al. (2016) is displayed in Figure 17.26. Sharp enhancements in NS and GRS count rates occurred at ~18:01:36 UTC on 21 November 2013. The MESSENGER spacecraft was located at a planetocentric distance of 1.90 RM in the post-midnight region at ~00:54 local time. Figure 17.26a shows that there was a pulse of energetic electrons with complex structure lasting until ~18:01:45 UTC. The magnetic field associated with this electron event showed “tail-like” stretching (Sundberg et al., 2012b; Sun et al., 2015a) prior to the particle injection, with the BX component strengthening and the BZ component diminishing between ~18:01:00 and 18:01:35 UTC. At the time of the energetic electron flux injection (18:01:36 UTC), the BX component decreased and the BZ component increased. The elevation angle, θ, of the magnetic field to the magnetic equator is shown in the bottom panel. Its sudden increase marks the arrival of a dipolarization front, and a pulse of energetic electrons was observed as in the Mariner 10 events (Baker et al., 1986; Christon et al., 1987; Delcourt et al., 2005). Baker et al. (2016) found that the most intense energetic electron bursts detected by MESSENGER appeared to be produced in the midnight sector of Mercury’s magnetosphere. The data show that the accelerated electrons were frequently observed on closed, rapidly reconfiguring magnetic field lines during substorm-like events. Further, the high-time-resolution GRS ACS (10-ms sampling) electron data show that injected electrons can in some instances complete one or more drifts around the planet on closed, quasi-trapped paths, creating Earthlike drift-echo events (Baker et al., 1996; Schriver et al., 2011a). However, it is also important to bear in mind the compactness of Mercury’s dayside magnetosphere. In most cases, it is likely that substorm-injected electrons will not be able to execute even one complete azimuthal drift around the planet before being lost. In fact, relatively few instances of electron burst events in the dusk or pre-midnight sectors were found (Baker et al., 2016), consistent with most of the injected electron bursts not having executed complete drifts about Mercury. Overall, the EPS and GRNS data clearly indicate that Earth-like substorm energetic electron injection events take place at Mercury (Ho et al., 2012; Lawrence et al., 2015; Baker et al., 1996). Indeed, these results support the hypothesis proposed by Baker et al. (1987) that Mercury might possess an auroral oval analog whereby an annular region of the surface centered on the boundary between open and closed magnetic field lines is warmed by the precipitation of energetic particles injected during substorms. In point of fact, Lindsay et al. (2016) recently used MESSENGER’s XRS measurements to construct a statistical image of X-ray fluorescence emissions from the surface mostly

17.5 MESSENGER’s Answers resulting from the precipitation of magnetospheric electrons with energies of ~1–10 keV. These emissions originate on the nightside of Mercury near the boundary between open and closed magnetic field lines at ~50°N and ~20°S, and they resemble the auroral ovals observed at Earth and the other planets possessing intrinsic magnetic fields and dense neutral atmospheres (Lindsay et al., 2016).

1 7 . 5 M E S S E N G E R’S ANS WERS To summarize the new insights that MESSENGER has provided into the dynamics of Mercury’s magnetosphere, we return to the science questions listed in the introduction. For each question set, brief answers are provided here, to the extent that they can be confidently inferred from the MESSENGER data. We then close by considering, very briefly, how the observations from the upcoming BepiColombo mission to Mercury are likely to lead to major new advances in our understanding of Mercury’s magnetosphere. Question set 1: How do magnetosheath conditions at Mercury differ from what is found at the other planets? Answer: The strong interplanetary magnetic fields in the inner heliosphere result in low MA, especially during coronal mass ejections. Under these conditions the bow shock is weak, the plasma β in the inner magnetosheath is very low, and a thick plasma depletion layer forms adjacent to the dayside magnetopause. Question set 2: How do these conditions in Mercury’s magnetosheath contribute to the dynamic nature of its magnetosphere? How does magnetopause reconnection at Mercury differ from that at Earth? Are flux transfer events (FTEs) a major driver of magnetospheric convection at Mercury? Answers: The intensity of the magnetic fields on the two sides of the magnetopause current layer tend to be comparable at Mercury due to the formation of strong plasma depletion layers. Under these conditions reconnection is said to be “symmetric,” in contrast with the typical situation at Earth where the field intensity inside the magnetosphere is usually much stronger than in the magnetosheath. As expected for symmetric conditions, reconnection at Mercury occurs for all non-zero magnetic shear angles and appears to be several times faster than at Earth. Flux transfer events occur much more frequently, and they carry relatively more magnetic flux than at Earth. At Mercury they are estimated to account for a third of the total magnetic flux transferred from the dayside magnetosphere into the magnetotail and a similar fraction of the dawn-to-dusk magnetospheric electric field. Question set 3: Does reconnection ever erode the dayside magnetosphere to the point that the subsolar region of the surface is exposed to direct solar wind impact? To what extent do induction currents driven in Mercury’s interior limit the solar wind flux to the surface? Do FTEs contribute significantly to the solar wind flux reaching the surface? Answers: During some CME events the magnetopause has been observed to be eroded and/or compressed to altitudes of


75% of mercurian lavas, on the basis of experiments.

interior lies between a low of IW−6.3 and a high of IW−2.6, that is, 10−2.6 to 10−6.3 below IW. The high end of this range is an extremely generous upper limit (cf. Zolotov et al., 2013). The results from MESSENGER also stimulated other experiments that led to similar findings. Namur et al. (2016) parameterized S and metal solubility in magma compositions representative of mercurian lavas, and given mantle/core equilibrium they calculated f(O2) = IW−5.4 ± 0.4 (10−5.0 to 10−5.8 below IW; Figure 8.2). Earth, on the other hand, was not as reduced during its accretion (e.g., Frost et al., 2008), and modern mid-ocean ridge basalts record upper mantle f(O2) at 102 above IW (IW+2), near the quartz–fayalite–magnetite (QFM) buffer (Cottrell and Kelley, 2011; Figure 8.2). The mantle source of venusian lavas is inferred to have fO2 similar to Earth’s upper mantle (Wadhwa, 2008). The known or inferred f(O2) values for a broad range of solar system materials are summarized in Figure 18.2. Mercury is the most reduced planet and more reduced than all measured early solar system materials except, possibly, the enstatite chondrite and enstatite achondrite meteorites (aubrites) and a small class of Ca-, Al-rich inclusions (CAIs) in some chondritic meteorites (Figure 18.2; Beckett, 1986). Recent geochemical thermodynamic modeling suggests that, as Earth and Mars accreted, their oxidation states progressively increased (Righter et al., 2008., 2016; Wood et al., 2009; Badro et al., 2015; Rubie et al., 2015). Earth’s lower mantle, Mars, and Vesta (represented by diogenite meteorites) all record mantle f(O2) near the iron–wüstite buffer curve (Figure 18.2) (Ghosal et al., 1998; Wadhwa, 2001; Frost et al., 2008; Szymanski et al., 2010; Tuff et al., 2013). The FeO content of silicates and the Si content of metals

18.2 How Anomalous Is Mercury? are proxies for the f(O2) during their formation. The silicates in pallasite meteorites record f(O2) similar to that in Earth’s lower mantle (Figure 18.2), with olivine, (Mg,Fe)2SiO4, ranging from 10 to 20 mol% Fe2SiO4 (11–19 wt% FeO) (Righter et al., 1990) and very low Si in metal. The most Si-rich iron meteorite, Horse Creek, contains 2.5 wt% Si in metal (Buchwald, 1975). Thus, the meteorites that represent cores or lower mantles of early differentiated planetesimals all record higher f(O2) than Mercury. Intrinsic fugacities, inferred from thermodynamic analyses of measured mineral compositions, have been calculated for some chondrite classes. The intrinsic f(O2) measured for the ordinary chondrites least equilibrated on their parent bodies (LL3, H4) range as low as IW−1; however, the equilibrated enstatite chondrite Hvittis (EL6) reaches IW−3 (Figure 18.2; Brett and Sato, 1984). The reduced, metal-rich CH and CB chondrites contain primarily 1600 K) elements Th or U reflect differences in the abundances of moderately volatile elements between planetary bodies. These elements are all highly incompatible in crystals that remain behind when partial melts rise to planetary surfaces, so their ratios are preserved in volcanic rocks that result from mantle melting processes. However, if some U and Th were incorporated into the core, which is possible under highly reduced conditions, the total mercurian inventory of K may be lower than that inferred from surface measurements of K/U and K/Th, under the assumption of chondritic mantle K, U, and Th (Malavergne et al., 2010; McCubbin et al., 2012). Furthermore, the crustal Cl/K ratio is close to chondritic (Evans et al., 2015; Chapter 2), as is the crustal Na/Si ratio (Evans et al., 2012; Peplowski et al., 2014). Overall, Cl and Na appear to be present in nearchondritic abundances. Mercury’s abundances of moderately volatile K, Na, and Cl relative to refractory Mg are inferred to be most similar to those measured for the EH enstatite chondrites (Figure 18.4). In summary, Mercury is anomalously enriched in Fe, S, and possibly also Si, relative to the other terrestrial planets (Figure 18.3). Its apparent bulk Na and possibly Cl compositions are also enriched above chondritic, whereas K is depleted, but less so than for Earth (Figure 18.3). A model for Mercury’s origin must elevate Si-rich Fe metal compared with Mg-silicates and retain volatile S and Na at near-chondritic values relative to Mg,

1800 1600 1400 1200

Relative Atomic Abundance /Cl/Mg



Ca V


Ga K Na Ge


1 Hf

Yb Ce Sr Ba



Mn Cu


Zn Cs

Ag Sb Cl

As Re Pt Pd







600 K












Silicate Bi Earth litho chalc sid






log10(T at 50% condensation, K at P tot = 10 Pa)

Figure 18.4. Volatile depletion of solar system reservoirs. Bulk silicate Earth (BSE, large circles), LL (diamonds), CO (triangles), CV (squares), CR (asterisks), CM (crosses), and EH (small circles) chondrite atomic abundances of selected elements, normalized to Mg (star) and CI chondrite (dotted line at a relative atomic abundance of 1) are plotted against the temperature T50% at which they are 50% condensed from a vapor of solar composition at 10 Pa (10−4 bar) total pressure (Lodders, 2003). Earth data are from McDonough (2014); CI chondrite from Lodders et al. (2009); CM, CO, CV, EH, LL from Wasson and Kallemeyn (1988); and CR from Lodders and Fegley (1998). Trend lines for bulk silicate Earth (dashed), CV, CM, and EH are estimated. In accord with Goldschmidt’s (1937) geochemical classification, bulk silicate Earth’s lithophile, chalcophile, and siderophile elements are distinguished by shading.

while maintaining abundance ratios of K, Ca, and Al that are similar to those for the other terrestrial planets. A reasonable inference from these observations is that the formation processes that led to the depletion of moderately volatile elements in planets compared with chondrites were decoupled from the origin of the large metal fraction of Mercury. 18.2.4 Surface Reflectance The surface reflectance of airless bodies depends upon chemical composition and regolith maturity. Mercury has a lower global reflectance than the Moon, but matures at a rate that is more rapid by as much as a factor of 4 (Robinson et al., 2008; Braden and Robinson, 2013; Chapter 8). Comparison of immature surfaces indicates that the difference in reflectance is likely due primarily to differences in the compositions of those surfaces. Darkening agents on the Moon include Fe- and Ti-bearing phases, but Mercury’s surface is depleted in both elements. Low-reflectance material (LRM) on Mercury is up to 30% lower in reflectance than the global mean, but LRM is not enriched in either Fe or Ti, and Fe concentration does not correlate with reflectance (Weider et al., 2012, 2014; Murchie et al., 2015). The LRM likely represents excavated mid to lower crustal material (Ernst et al., 2010), so the darkening agent may represent a major component of the silicate portion of Mercury. Gamma-Ray Spectrometer (GRS) measurements are consistent with 0–4.1 wt% C on Mercury’s surface (Peplowski et al., 2015, 2016; Chapter 2). High thermal-neutron count rates measured by the MESSENGER Neutron Spectrometer (NS) correlate with LRM, consistent with LRM C abundances 1–3 wt% greater than in surrounding higher-reflectance material (Peplowski et al.,

18.3 Planet Formation in Disks 2016). Experiments on materials analogous to Mercury’s mantle indicate that the only major mineral that would be buoyant in a Mercury magma ocean would be graphite (Vander Kaaden and McCubbin, 2015). Peplowski et al. (2016) inferred that the LRM in particular, and the volcanic upper crust generally, samples remnants of an early graphite-rich flotation crust subsequently mixed and modified by impacts and magmatic intrusions and later excavated by large craters and/or assimilated into later volcanic magmas (Chapter 6). Attribution of the low reflectance of Mercury to elevated elemental carbon is consistent with the high S/Si ratio in surface materials. While graphite is a common mineral in enstatite chondrites, those meteorites contain only ~10% of the C measured in CI chondrites. 18.2.5 Summary The anomalously large Si-bearing Fe core, oxidation state, volatile enrichment, and reflectance of Mercury all demand explanation, but they are also clues to Mercury’s formation. The Si enrichment of the core follows from the observed oxidation state, which probably also controls S distribution in the planet. In the meteorite record there exist rare materials that are similarly iron-rich, or similarly reduced, but not both. The EH enstatite chondrites are reduced, enriched in Si relative to Mg, Al, and Ca, and similarly volatile-rich. Although the EH chondrites contain Cl and K in high abundance, inferences from MESSENGER data for Mercury may represent lower bounds, subject to further experimental partitioning studies (McCubbin et al., 2012). The low reflectance of Mercury’s surface may be closely related to its reduced chemistry if the planet formed with a substantial carbon content, as suggested by MESSENGER measurements (Peplowski et al., 2016).

18.3 PLANET FORMATION I N D ISKS 18.3.1 Theoretical Considerations The eighteenth century concept of the solar nebula (Kant, 1755) has evolved into modern astrophysical disk theory that treats three main stages of planet formation: (1) mineral dust concentrates in the midplane of the solar nebula and accretes to form multi-kilometer-sized planetesimals, (2) the largest planetesimals grow in annuli by runaway and preferential accretion to the largest bodies (oligarchic growth) to form Moon- to Mars-mass planetary embryos, and (3) the final terrestrial planets form by energetic, stochastic collisions between embryos driven by gravitational interactions (Safronov, 1972; Morbidelli et al., 2012). Stage 1 is poorly understood and is thought to be rapid (~105 yr), stage 2 (~106 yr) forms embryos with characteristic spacing, with gas dissipating in a few million years, and stage 3 (~108 yr) establishes the final radius, mass, and composition of each of the terrestrial planets through violent collisions between planetesimals and embryos from disparate solar radii (Chambers, 2004, 2009a). The challenge lies in understanding the details of how these processes occurred. The cornerstone for chemical models for planet formation is the chondritic model for condensates in the solar nebula. The variations in the volatile compositions of the classes of


chondritic meteorites are thought to reflect differences in the pressure–temperature conditions of equilibration of component solids with the gas in the solar nebula, overprinted with variable abundances of volatile ices (Figure 18.4; Davis, 2006). How these conditions varied over time with distance from the Sun and height in the disk is not known, but the conditions are probably recorded in the oldest, most volatile-rich meteorites, from bodies that did not differentiate into cores and mantles (Alexander et al., 2001). These meteorites come from asteroids and were accreted from high-temperature, 100–1000-micrometer-size chondrules and Ca-, Al-rich inclusions that were once free-floating nebular solids. In the chondrites least altered by water and heat on their parent bodies, these objects are surrounded by a fine-grained matrix containing presolar interstellar grains, organic matter, amorphous particles, and other materials (Alexander et al., 2007). The physical origin of meteoritic assemblages of chondrules and matrix materials remains elusive (e.g., Ebel et al., 2016). Although separation of metal and silicate among different components is observable at the millimeter scale, and there is significant variation in the metal content of chondrite groups, only one group (CH chondrites) has a metal/silicate ratio similar to that of Mercury (Figure 18.3). Understanding the origin of Mercury is significantly handicapped by the lack of certainty about the formation of terrestrial planets in general. The physics of growth from dust to Mars-mass bodies is particularly poorly constrained. As a result, the chemical evolution of the precursor materials of planets cannot be robustly predicted, and certainly not as a function of time and distance from the Sun. In the past several years, new ideas have challenged traditional models of the orderly growth of planets. For example, planetesimals may rapidly grow into embryos by so-called “pebble accretion.” In this model, the accretion efficiency of centimeter- to sub-meter-sized “pebbles” is greatly enhanced by Stokes drag in the atmospheres around growing embryos (Lambrechts and Johansen, 2012; Johansen et al., 2014). Calculations of embryo growth by pebble accretion have successfully led to systems that resemble the solar system’s outer planets (e.g., Chambers, 2014; Levison et al., 2015a) and the terrestrial planets (Levison et al., 2015b). However, neither the physical origin nor the chemical nature of pebbles is constrained, because the models require only that pebbles are objects with a favorable Stokes number, such that the gas-drag stopping time is comparable to the time it takes for the pebble to cross the embryo’s region of gravitational influence (Hill radius). In most models, pebbles are larger than the mostly sub-millimeter-sized chondrules found in meteorites (Friedrich et al., 2015), which would be strongly coupled to the gas. Thus, the relationship between pebbles and meteorites is currently unknown, and the chemical relationship between meteorites and growing embryos in this model is unexplored. Central to the accretion of the terrestrial planets are the motions of the giant planets. The Nice model describing the outward migration of the giant planets from an earlier, more compact configuration is now a widely accepted basis for scenarios describing the early history of our solar system (e.g., Tsiganis et al., 2005; Levison et al., 2007; Morbidelli et al., 2007; Batygin and Brown, 2010). The possibility of inward and outward migration, as proposed in the Grand Tack model

The Elusive Origin of Mercury

18.3.2 Observations Protoplanetary disks are rotationally supported structures of gas and dust around young stars (Williams and Cieza, 2011; Armitage, 2011). Disks are observed around many T-Tauri type stars, actively accreting low-mass (1 m) in Keplerian orbits experienced a headwind due to the radiation pressure experienced by the gas. This gas drag caused meter-sized grains to drift sunward faster than the gas. Larger and/or denser boulders experienced slower orbital decay. Weidenschilling proposed that precursor multimeter-size solids (boulders) with differing Fe/Si ratios could dynamically interact in such a way that silicates were preferentially removed into the Sun by gas drag. Hubbard (2014) explored a “magnetic erosion” model that combined the magnetic attraction of metallic grains and the differential comminution of silicates and metal by collision. He called on particular conditions at the outer edge of the inner disk to enhance magnetization of metal-rich grains, thus enhancing both the collisional removal of non-magnetic silicate and the rapid collisional growth of large metal grains to sizes


beyond the meter barrier. The magnetic field requirement restricts this mechanism to the inner disk. All three of these models are consistent with Mercury’s anomalous density and Earth-like mantle abundances, and they all enrich Mercury in metal that is commonly associated with sulfide, consistent with Mercury’s likely S enrichment. None of these models, however, addresses the extremely reduced nature of Mercury. They all call on special astrophysical conditions in the innermost nebula, and none has been explored with fully three-dimensional chemical–physical models. If precursor metal is assumed to carry sulfur, then Mgsilicates and refractory elements might be assumed to be in large grains (chondrules, CAIs), whereas K, Na, and Cl would have concentrated in smaller grains, the matrix mineral dust of chondrites. MESSENGER results for K indicate that Mercury’s volatility curve is not steeper than Earth’s (Figure 18.4). The fates of the volatiles in these scenarios are difficult to predict. Several recent astrophysical models have yielded results consistent with Mercury’s metal enrichment and reduced chemistry for disks around stars similar to the Sun. Pasek et al. (2005) used a two-dimensional steady-state α-disk solution to yield time– temperature–pressure histories in the disk midplane at various solar radii along with chemistry codes to compute condensation fronts. With a model for diffusive transport driven by condensation fronts in the inner disk, they then calculated condensation histories at progressively more O-depleted inner radii. They found that the effects of water depletion on nebular S speciation formed reduced enstatite-chondrite-like rocks at Mercury-like astrocentric radii. Moriarty et al. (2014) applied a disk model (Chambers, 2009b) to calculate disk temperature, pressure, and density over time. They then calculated equilibrium chemistry at steps in time and radius, removing gas and dust into planetesimals growing at a prescribed rate and decoupled from the gas. They also accounted for the effects of radial gas movement (with perfectly coupled dust) on the chemical inventory. Finally, their planetesimals were input into an N-body simulation of late-stage planet formation. Their model produced C-rich, short-period planetesimals around stars with C/O ratios slightly above the solar ratio. Pignatale et al. (2016) used a disk model (D’Alessio et al., 1999) to prescribe temperature and pressure in a two-dimensional disk and then calculated condensation under each of those conditions. They then applied dust-settling and radial-migration models to calculate redistributions of material. Their model produced sulfide- and enstatite-rich zones within 1 AU of the young Sun. Although feedback between chemistry and dynamics is highly limited or missing from efforts to date to couple chemistry and dynamics, model results consistently predict reduced, metal-enriched inner disks, the probable feeding zone for planet Mercury. C-Rich Condensation The gross separation of inner volatile-poor planets and outer volatile-rich planets immediately suggests a nebular “snow line,” a complex time-dependent surface inside of which water remained in the vapor phase. In the presence of free oxygen, a similar “C-line” would have marked the locus


The Elusive Origin of Mercury

inside of which graphite would have oxidized to CO and/or CO2. The time–temperature and oxygen distribution histories of the inner disk are not known, nor is the time-dependent accretion flux of interstellar material onto the innermost disk well quantified. The balance between C and O as a function of time and radius in the terrestrial planet-forming region may, therefore, have been quite heterogeneous. Ebel and Alexander (2011) investigated the consequences of carbon enrichment and oxygen depletion on the stability of minerals in a cooling H2-rich vapor. They noted that the most abundant interplanetary dust particles (IDPs) in the present solar system are 50–1000-μm sized, anhydrous, porous, chondritic “C-IDPs.” The C-IDPs are aggregates containing highly primitive sub-micrometer silicates, metal, sulfide and presolar grains all attached together by poorly graphitized carbon (Messenger et al., 2003; Busemann et al., 2009; Bradley, 2014). Their original C content has been diminished by precapture stratospheric entry. Ebel and Alexander (2011) explored the consequences of equilibrium condensation in systems enriched in a C-enriched, O-depleted analog C-IDP dust. At high (1000×) enrichments in such a dust (relative to H2), condensates at 1650 K (at 10 Pa total pressure) have atomic Fe/Si reaching 50% of the estimate for bulk Mercury, because Si remains in the vapor to low temperatures (4.53 Ga) global differentiation of the silicate Earth. Science, 309, 576–581. Boynton, W. V., Sprague, A. L., Solomon, S. C., Starr, R. D., Evans, L. G., Feldman, W. C., Trombka, J. I. and Rhodes, E. A. (2007). MESSENGER and the chemistry of Mercury’s surface. Space Sci. Rev., 131, 85–104. Braden, S. E. and Robinson, M. S. (2013). Relative rates of optical maturation of regolith on Mercury and the Moon. J. Geophys. Res. Planets, 118, 1903–1914, doi:10.1002/jgre.20143 Bradley, J. P. (2014). Early solar nebula grains – interplanetary dust particles. In Meteorites and Cosmochemical Processes, ed. A. M. Davis, Treatise on Geochemistry, 2nd edn, Vol. 1, ed. H. D. Holland and K. Turekian. Amsterdam, Oxford: Elsevier, pp. 287–308. Brett, R. and Sato, M. (1984). Intrinsic oxygen fugacity measurements on seven chondrites, a pallasite, and a tektite and the redox state of meteorite parent bodies. Geochim. Cosmochim. Acta, 48, 111–120. Brownlee, D. (2014). The Stardust mission: Analyzing samples from the edge of the solar system. Annu. Rev. Earth Planet. Sci., 42, 179–205. Buchwald, V. F. (1975). Handbook of Iron Meteorites, Their History, Distribution, Composition and Structure. Berkeley, CA: University of California Press, 1426 pp. Burbine, T. H., Meibom, A. and Binzel, R. P. (1996). Mantle material in the main belt: Battered to bits? Meteorit. Planet. Sci., 31, 607–620. Burkhardt, C. (2014). Isotopic composition of the Moon and the lunar isotopic crisis. In Encyclopedia of Lunar Science. Springer (online), doi:10.1007/SpringerReference_440362. Burkhardt, C., Kleine, T., Bourdon, B., Palme, H., Zipfel, J., Friedrich, J. and Ebel, D. S. (2008). Hf-W systematics of Ca-Al-rich inclusions from carbonaceous chondrites: Dating the age of the solar system and core formation in asteroids. Geochim. Cosmochim. Acta, 72, 6177–6197. Burkhardt C., Borg, L. E., Brennecka, G. A., Shollenberger, Q. R., Dauphas, N. and Kleine T. (2016). A nucleosynthetic origin for the Earth’s anomalous 142Nd composition. Nature, 537, 394–398. Busemann, H., Nguyen, A. N., Cody, G. D., Hoppe, P., Kilcoyne, A. L. D., Stroud, R. M., Zega, T. J. and Nittler, L. R. (2009). Ultraprimitive interplanetary dust particles from the comet 26P/GriggSkjellerup dust stream collection. Earth Planet. Sci. Lett., 288, 44–57. Cameron, A. G. W. (1985). The partial volatilization of Mercury. Icarus, 64, 285–294. Cameron, A. G. W., Fegley, B., Benz, W. and Slattery, W. L. (1988). The strange density of Mercury: Theoretical considerations. In Mercury, ed. F. Vilas, C. R. Chapman and M. S. Matthews. Tucson, AZ: University of Arizona Press, pp. 692–708. Canup, R. M. (2004). Dynamics of lunar formation. Annu. Rev. Astron. Astrophys., 42, 441–475. Canup, R. M. (2008). Accretion of the Earth. Phil. Trans. Roy. Soc. London A, 366, 4061–4075. Canup, R. M. (2012). Forming a Moon with an Earth-like composition via a giant impact. Science, 338, 1052–1055. Canup, R. M. and Asphaug, E. (2001). Origin of the Moon in a giant impact near the end of the Earth’s formation. Nature, 412, 708–712.

References Canup, R. M., Visscher, C., Salmon, J. and Fegley, B., Jr. (2015). Lunar volatile depletion due to incomplete accretion within an impactgenerated disk. Nature Geosci., 8, 918–921, doi:10.1038/ ngeo2574. Carry, B. (2012). Density of asteroids. Planet. Space Sci., 73, 98–118. Carter, P. J., Leinhardt, Z. M., Elliott, T., Walter, M. J. and Stewart, S. T. (2015). Compositional evolution during rocky protoplanet accretion. Astrophys. J., 813, 72–91. Chambers, J. E. (2004). Planetary accretion in the inner Solar System. Earth Planet. Sci. Lett., 223, 241–252. Chambers, J. E. (2009a). Planetary migration: What does it mean for planet formation? Annu. Rev. Earth Planet. Sci., 37, 321–344. Chambers, J. E. (2009b). An analytical model for the evolution of a viscous, irradiated disk. Astrophys. J., 705, 1206–1214. Chambers, J. E. (2013). Late-stage planetary accretion including hitand-run collisions and fragmentation. Icarus, 224, 43–56. Chambers, J. E. (2014). Giant planet formation with pebble accretion. Icarus, 233, 83–100. Chapman, C. R. (1988). Mercury: Introduction to an end-member planet. In Mercury, ed. F. Vilas, C. R. Chapman and M. S. Matthews. Tucson, AZ: University of Arizona Press, pp. 1–23. Chen, J. and Kipping, D. (2016). Probabilistic forecasting of the masses and radii of other worlds. Astrophys. J., 834, 17–30. Clayton, R. N. and Mayeda, T. K. (1996). Oxygen isotope studies of achondrites. Geochim. Cosmochim. Acta, 60, 1999–2017. Consolmagno, G. J. and Britt, D. T. (2004). Meteoritical evidence and constraints on asteroid impacts and disruption. Planet. Space Sci., 52, 1119–1128. Cottrell, E. and Kelley, K. A. (2011). The oxidation state of Fe in MORB glasses and the oxygen fugacity of the upper mantle. Earth Planet. Sci. Lett., 305, 270–282. Ćuk, M. and Stewart, S. T. (2012). Making the Moon from a fastspinning Earth: A giant impact followed by resonant despinning. Science, 338, 1047–1052, doi:10.1126/science.1225542. Cuzzi, J. N., Hogan, R. C. and Shariff, K. (2008). Toward planetesimals: Dense chondrule clumps in the protoplanetary nebula. Astrophys. J., 687, 1432–1447. Dahl, T. W. and Stevenson, D. J. (2010). Turbulent mixing of metal and silicate during planet accretion – And interpretation of the Hf–W chronometer. Earth Planet. Sci. Lett., 295, 177–186. D’Alessio, P., Calvet, N., Hartmann, L., Lizano, S. and Cantó, J. (1999). Accretion disks around young objects. II. Tests of wellmixed models with ISM dust. Astrophys. J., 527, 893–909. Dauphas, N. and Morbidelli, A. (2014). Geochemical and planetary dynamical views on the origin of Earth’s atmosphere and oceans. In The Atmosphere – History, ed. J. Farquhar, Treatise on Geochemistry, 2nd edn, Vol. 6, ed. H. D. Holland and K. Turekian. Amsterdam, Oxford: Elsevier, pp. 1–35. Davis, A. M. (2006). Volatile evolution and loss. In Meteorites and the Early Solar System II, ed. D. Lauretta and H. Y. McSween, Jr. Tucson, AZ: University of Arizona Press, pp. 295–307. Domingue, D. L., Chapman, C. R., Killen, R. M., Zurbuchen, T. H., Gilbert, J. A., Sarantos, M., Benna, M., Slavin, J. A., Schriver, D., Trávníček, P. M., Orlando, T. M., Sprague, A. L., Blewett, D. T., Gillis-Davis, J. J., Feldman, W. C., Lawrence, D. J., Ho, G. C., Ebel, D. S., Nittler, L. R., Vilas, F., Pieters, C. M., Solomon, S. C., Johnson, C. L., Winslow, R. M., Helbert, J., Peplowski, P. N., Weider, S. Z., Mouawad, N., Izenberg, N. R. and McClintock, W. E. (2014). Mercury’s weather-beaten surface: Understanding Mercury in the context of lunar and asteroidal space weathering studies. Space Sci. Rev., 181, 121–214, doi:10.1007/s11214-014– 0039-5. Dressing, C. D. and 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. Astrophys. J., 807, 45–68. Dwyer, C. A., Nimmo, F. and Chambers, J. E. (2015). Bulk chemical and Hf–W isotopic consequences of incomplete accretion during planet formation. Icarus, 245, 145–152. Ebel, D. S. (2001). Vapor/liquid/solid equilibria when chondrites collide. Meteorit. Planet. Sci. Suppl., 36, A52–A53. Ebel, D. S. (2006). Condensation of rocky material in astrophysical environments. In Meteorites and the Early Solar System II, ed. D. Lauretta and H. Y. McSween, Jr. Tucson, AZ: University of Arizona Press, pp. 253–277. Ebel, D. S. and Alexander, C. M. O’D. (2011). Equilibrium condensation from chondritic porous IDP enriched vapor: Implications for Mercury and enstatite chondrite origins. Planet. Space. Sci., 59, 1888–1894. Ebel, D. S. and Grossman, L. (2000). Condensation in dust-enriched systems. Geochim. Cosmochim. Acta, 64, 339–366. Ebel, D. S. and Sack, R. O. (2013). Djerfisherite: Nebular source of refractory potassium. Contrib. Mineral. Petrol., 166, 923–934. Ebel, D. S., Brunner, C., Leftwich, K., Erb, I., Lu, M., Konrad, K., Rodriguez, H., Friedrich, J. M. and Weisberg, M. K. (2016). Abundance, composition and size of inclusions and matrix in CV and CO chondrites. Geochim. Cosmochim. Acta, 172, 322– 356, doi:10.1016/j.gca.2015.10.007. Elkins-Tanton, L. T. (2013). Planetary science: Occam’s origin of the Moon. Nature Geosci., 6, 996–998. Ernst, C. M., Murchie, S. L., Barnouin, O. S., Robinson, M. L., Denevi, B. W., Blewett, D. T., Head, J. W., Izenberg, N. R., Solomon, S. C. and Roberts, J. H. (2010). Exposure of spectrally distinct material by impact craters on Mercury: Implications for global stratigraphy. Icarus, 209, 210–223, doi:10.1016/j.icarus.2010.05.022. Espaillat, D., Muzerolle, J., Najita, J., Andrews, S., Zhu, Z., Calvet, N., Kraus, S., Hashimoto, J., Kraus, A. and D’Alessio, P. (2014). An observational perspective of transitional disks. In Protostars and Planets VI, ed. H. Beuther, R. Klessen, C. Dullemond and Th. Henning. Tucson, AZ: University of Arizona Press, pp. 497–520. Evans, L. G., Peplowski, P. N., Rhodes, E. A., Lawrence, D. J., McCoy, T. J., Nittler, L. R., Solomon, S. C., Sprague, A. L., StockstillCahill, K. R., Starr, R. D., Weider, S. Z., Boynton, W. F., Hamara, D. K. and Goldsten, J. O. (2012). Major-element abundances on the surface of Mercury: Results from the MESSENGER GammaRay Spectrometer. J. Geophys. Res., 117, E00L07, doi:10.1029/ 2012JE004178. Evans, L. G., Peplowski, P. N., McCubbin, F. M., McCoy, T. J., Nittler, L. R., Zolotov, M. Yu., Ebel, D. S., Lawrence, D. J., Starr, R. D., Weider, S. Z. and Solomon, S. C. (2015). Chlorine on the surface of Mercury: MESSENGER gamma-ray measurements and implications for the planet’s formation and evolution. Icarus, 257, 417–427. Fabrycky, D. C., Lissauer, J. J., Ragozzine, D., Fowe, J. F., Steffen, J. H., Agol, E., Barclay, T., Batalha, N., Borucki, W. and Ciardi, D. R. (2014). Architecture of Kepler’s multi-transiting systems. II. New investigations with twice as many candidates. Astrophys. J., 790, 146–157. Fang, J. and Margot, J. L. (2012). Architecture of planetary systems based on Kepler data: Number of planets and coplanarity. Astrophys. J., 761, 92–105. Fedkin, A. V., Grossman, L., Humayun, M., Simon, S. B. and Campbell, A. J. (2015). Condensates from vapor made by impacts between metal-, silicate-rich bodies: Comparison with metal and chondrule in CB chondrites. Geochim. Cosmochim. Acta, 164, 236–261. Fegley, B., Jr. and Cameron, A. G. W. (1987). A vaporization model for iron/silicate fractionation in the Mercury protoplanet. Earth Planet. Sci. Lett., 82, 207–222.


The Elusive Origin of Mercury

Feigelson, E. D. (2010). X-ray insights into star and planet formation. Proc. Natl. Acad. Sci., 107, 7153–7157. Friedrich, J. M., Weisberg, M. K., Ebel, D. S., Biltz, A. E., Corbett, B. M., Iotzov, I. V., Khan, W. S. and Wolman, M. D. (2015). Chondrule size and density in all meteorite groups: A compilation and evaluation of current knowledge. Chemie der Erde, 75, 419– 443, doi:10.1016/j.chemer.2014.08.003. Frost, D. J., Mann, U., Asahara, Y. and Rubie, D. C. (2008). The redox state of the mantle during and just after core formation. Phil. Trans. Roy. Soc. London A, 366, 4315–4337. Genda, H., Kokubo, E. and Ida, S. (2011). Merging criteria for giant impacts of protoplanets. Astrophys. J., 744, 137–144. Ghosal, S., Sack, R. O., Ghiorso, M. S. and Lipschutz, M. E. (1998). Evidence for a reduced, Fe-depleted martian mantle source region of shergottites. Contrib. Mineral. Petrol., 130, 346–357. Gladman, B. and Coffey, J. (2009). Mercurian impact ejecta: Meteorites and mantle. Meteorit. Planet. Sci., 44, 285–291. Goldschmidt, V. M. (1937). The principles of distribution of chemical elements in minerals and rocks. The seventh Hugo Müller Lecture, delivered before the Chemical Society on March 17th, 1937. J. Chem. Soc., 1937, 655–673. Grossman, J. N. (1996). Chemical fractionations of chondrites: Signatures of events before chondrule formation. In Chondrules and the Protoplanetary Disk, ed. R. H. Hewins, R. H. Jones and E. R. D. Scott. New York: Cambridge University Press, pp. 243–253. Grossman, L. and Larimer, J. W. (1974). Early chemical history of the solar system. Rev. Geophys. Space Phys., 12, 71–101. Haisch, K. E., Lada, E. A. and Lada, C. J. (2001). Circumstellar disks in the IC 348 cluster. Astrophys. J., 121, 2065–2074. Halliday, A. N. (2013). The origins of volatiles in the terrestrial planets. Geochim. Cosmochim. Acta, 105, 146–171. Hansen, B. M. S. (2009). Formation of the terrestrial planets from a narrow annulus. Astrophys. J., 703, 1131–1140. Hartmann, L. (2009). The star-jet-disk system and angular momentum transfer. In Protostellar Jets in Context, ed. K. R. Tsinganos and M. Stute. Berlin: Springer, pp. 23–32. Hauck S. A., II, Margot, J.-L., Solomon, S. C., Phillips, R. J., Johnson, C. L., Lemoine, F. G., Mazarico, E., McCoy, T. J., Padovan, S., Peale, S. J., Perry, M. E., Smith, D. E. and Zuber, M. T. (2013). The curious case of Mercury’s internal structure. J. Geophys. Res. Planets, 118, 1204–1220. Hewins, R. H. and Ulmer, G. C. (1984). Intrinsic oxygen fugacities of diogenites and mesosiderite clasts. Geochim. Cosmochim. Acta, 48, 1555–1560. Hubbard, A. (2014). Explaining Mercury’s density through magnetic erosion. Icarus, 241, 329–335. Izenberg, N. R., Klima, R. L., Murchie, S. L., Blewett, D. T., Holsclaw, G. M., McClintock, W. E., Malaret, E., Mauceri, C., Vilas, F., Sprague, A. L., Helbert, J., Domingue, D. L., Head, J. W., III, Goudge, T. A., Solomon, S. C., Hibbitts, C. A. and Dyar, M. D. (2014). The low-iron, reduced surface of Mercury as seen in spectral reflectance by MESSENGER. Icarus, 228, 364–374. Jarosewich, E. (1990). Chemical analyses of meteorites: A compilation of stony and iron meteorite analyses. Meteoritics, 25, 323–337. Johansen, A., Blum, J., Tanaka, H., Ormel, C., Bizzarro, M. and Rickman, H. (2014). The multifaceted planetesimal formation process. In Protostars and Planets VI, ed. H. Beuther, R. Klessen, C. Dullemond and Th. Henning. Tucson, AZ: University of Arizona Press, pp. 547–570. Kant, I. (1755). Universal Natural History and Theory of the Heavens. In Kant’s Critical Religion (2000), translated by S. Palmquist. Aldershot: Ashgate, 320 pp.

Kargel, J. S. and Lewis, J. S. (1993). The composition and early evolution of Earth. Icarus, 105, 1–25. Keil, K. (1968). Mineralogical and chemical relationships among enstatite chondrites. J. Geophys. Res., 73, 6945–6976. Kleine, T., Mezger, K., Palme, H. and Münker, C. (2004). The W isotope evolution of the bulk silicate Earth: Constraints on the timing and mechanisms of core formation and accretion. Earth Planet. Sci. Lett., 228, 109–123. Kleine, T., Touboul, M., Bourdon, B., Nimmo, F., Mezger, K., Palme, H., Jacobsen, S. B., Yin, Q.-Z. and Halliday, A. N. (2009). Hf-W chronology of the accretion and early evolution of asteroids and terrestrial planets. Geochim. Cosmochim. Acta, 73, 5150–5188. Klima, R. L., Izenberg, N. R., Murchie, S., Meyer, H. M., StockstillCahill, K. R., Blewett, D. T., D’Amore, M., Denevi, B. W., Ernst, C. M., Helbert, J., McCoy, T., Sprague, A. L., Vilas, F., Weider, S. Z. and Solomon, S. C. (2013). Constraining the ferrous iron content of silicate minerals in Mercury’s crust. Lunar Planet. Sci., 44, abstract 1602. Kokubo, E. and Genda, H. (2010). Formation of terrestrial planets from protoplanets under a realistic accretion condition. Astrophys. J. Lett., 714, L21–L25. Kokubo, E. and Ida, S. (1996). On runaway growth of planetesimals. Icarus, 123, 180–191. Kokubo, E. and Ida, S. (1998). Oligarchic growth of protoplanets. Icarus, 131, 171–178. Krauss, O. and Wurm, G. (2005). Photophoresis and the pile-up of dust in young circumstellar disks. Astrophys. J., 630, 1088–1092. Krot, A. N., Amelin, Y., Cassen, P. and Meibom, A. (2005). Young chondrules in CB chondrites from a giant impact in the early Solar System. Nature, 436, 989–992. Krot, A. N., Amelin, Y., Bland, P., Ciesla, F. J., Connelly, H. J., Jr., Davis, A. M., Huss, G. R., Hutcheon, I. D., Makide, K., Nagashima, K., Nyquist, L. E., Russell, S. S., Scott, E. R. D., Thrane, K., Yurimoto, H. and Yin, Q.-Z. (2009). Origin and chronology of chondritic components: A review. Geochim. Cosmochim. Acta, 73, 4963–4997. Lambrechts, M. and Johansen, A. (2012). Rapid growth of gas-giant cores by pebble accretion. Astron. Astrophys., 544, A32–A45. Larimer, J. W. and Anders, E. (1967). Chemical fractionation in meteorites: II. Abundance patterns and their interpretation. Geochim. Cosmochim. Acta, 31, 1239–1270. Leinhardt, Z. M. and Stewart, S. T. (2011). Collisions between gravitydominated bodies. I. Outcome regimes and scaling laws. Astrophys. J., 745, 79–106. Leinhardt, Z. M., Dobinson, J., Carter, P. J. and Lines, S. (2015). Numerically predicted indirect signatures of terrestrial planet formation. Astrophys. J., 806, 23–32. Levison, H. F., Morbidelli, A., Gomes, R. and Backman, D. (2007). Planet migration in planetesimal disks. In Protostars and Planets V, ed. B. Reipurth, D. Jewitt and K. Keil. Tucson, AZ: University of Arizona Press, pp. 669–684. Levison, H. F., Kretke, K. A. and Duncan, M. J. (2015a). Growing the gas-giant planets by the gradual accumulation of pebbles. Nature, 524, 322–324. Levison, H. F., Kretke, K. A., Walsh, K. J. and Bottke W. F. (2015b). Growing the terrrestrial planets from the gradual accumulation of submeter-sized objects. Proc. Natl. Acad. Sci., 112, 14,180– 14,185. Lewis, J. S. (1972). Metal/silicate fractionation in the solar system. Earth Planet. Sci. Lett., 15, 286–290. Lewis, J. S. (1973). Chemistry of the planets. Annu. Rev. Phys. Chem., 24, 339–352.

References Lewis, J. S. (1988). Origin and composition of Mercury. In Mercury, ed. F. Vilas, C. R. Chapman and M. S. Matthews. Tucson, AZ: University of Arizona Press, pp. 651–666. Lissauer, J. J., Fabrycky, D. C., Ford, E. B., Borucki, W. J., Fressin, F., Marcy, G. W., Rorosz, J. A., Rowe, J. F., Torres, G., Welsh, W. F., Batalha, N. M., Bryson, S. T., Buchhave, L. A., Caldwell, D. A., Cartre, J. A., Charbonneau, D., Christiansen, J. L., Cochran, W. D., Desert, J.-M., Dunham, E. W., Fanelli, M. N., Fortney, J. J., Gautier, T. N., III, Geary, J. C., Gilliland, R. L., Haas, M. R., Hall, J. R., Holman, M. J., Coch, D. G., Latham, D. W., Lopez, E., McCauliff, S., Miller, N., Morehead, R. C., Quintana, E. V., Ragozzine, D., Sasselov, D., Short, D. R. and Stefffen, J. H. (2011). A closely packed system of low-mass, low-density planets transiting Kepler-11. Nature, 470, 53–58. Lock S. J., Stewart, S. T., Petaev, M. I., Leinhardt, Z. M., Mace, M., Jacobsen, S. B. and Ćuk, M. (2016). A new model for lunar origin: Equilibration with Earth beyond the hot spin stability limit. Lunar Planet. Sci., 47, abstract 2881. Lodders, K. (2003). Solar system abundances and condensation temperatures of the elements. Astrophys. J., 591, 1220–1247. Lodders, K. and Fegley, B., Jr. (1997). An oxygen isotope model for the composition of Mars. Icarus, 126, 373–394. Lodders, K. and Fegley, B., Jr. (1998). The Planetary Scientist’s Companion. New York: Oxford University Press. Lodders, K., Palme, H. and Gail, H. P. (2009). Abundances of the elements in the solar system. In Landolt-Börnstein, New Series, Vol. VI/4B, ed. J. E. Trümper. Berlin, Heidelberg, New York: Springer-Verlag, pp. 560–630. Loesche, C., Wurm, G., Kelling, T., Teiser, J. and Ebel, D. S. (2016). The motion of chondrules and other particles in a protoplanetary disk with temperature fluctuations. Mon. Not. Roy. Astron. Soc., 463, 4167–4174. Macke, R. J. (2012). Survey of meteorite physical properties: Density, porosity and magnetic susceptibility. Ph.D. thesis, University of Central Florida, Orlando, FL. Mahoney, T. J. (2014). Mercury, A Compendium of the Astronomical Lexicon, Part A: Gazetteer and Atlas of Astronomy, Vol. I: The Terrestrial Planets, Part 1. New York: Springer, doi:10.1007/ 978–1-4614–7951-2. Malavergne, V., Toplis, M. J., Berthet, S. and Jones J. (2010). Highly reducing conditions during core formation on Mercury: Implications for internal structure and the origin of a magnetic field. Icarus, 206, 199–209, doi:10.1016/j.icarus.2009.09.001. Marcus, R. A., Stewart, S. T., Sasselov, D. and Hernquist, L. (2009). Collisional stripping and disruption of super-Earths. Astrophys. J. Lett., 700, L118–L122. Marcus, R. A., Sasselov, D., Hernquist, L. and Stewart, S. T. (2010a). Minimum radii of super-Earths: Constraints from giant impacts. Astrophys. J. Lett., 712, L73–L76. Marcus, R. A., Sasselov, D., Stewart, S. T. and Hernquist, L. (2010b). Water/icy super-Earths: Giant impacts and maximum water content. Astrophys. J. Lett., 719, L45–L49. McClure, M. K., D’Alessio, P., Calvet, N., Espaillat, C., Hartmann, L., Sargent, B., Watson, D. M., Ingleby, L. and Hernández, J. (2013). Curved walls: Grain growth, settling, and composition patterns in T Tauri disk dust sublimation fronts. Astrophys. J., 775, 114–124. McCubbin, F. M., Riner, M. A., Vander Kaaden, K. E. and Burkemper L. K. (2012). Is Mercury a volatile-rich planet? Geophys. Res. Lett., 39, L09202, doi:10.1029/2012GL051711. McDonough W. F. (2014). Compositional model for the Earth’s core. In The Mantle and Core, ed. R. W. Carlson, Treatise on Geochemistry, 2nd edn, Vol. 3, ed. H. D. Holland and K. Turekian. Amsterdam, Oxford: Elsevier, pp. 559–577.


McNally, C. P. and McClure, M. K. (2017). Photophoretic levitation and trapping of dust in the inner regions of protoplanetary disks. Astrophys. J., 834, 48–60. McNally, C. P., Hubbard, A., Mac Low, M.-M., Ebel, D. S. and D’Alessio, P. (2013). Mineral processing by short circuits in protoplanetary disks. Astrophys. J., 767, L2–L7. Melosh, H. J. (2014), New approaches to the Moon’s isotopic crisis. Phil. Trans. Roy. Soc. London A, 372, 20130168. Messenger, S., Keller, L. P., Stadermann, F. J., Walker, R. M. and Zinner, E. (2003). Samples of stars beyond the Solar System: Silicate grains in interplanetary dust. Science, 300, 105–108. Mitchell, D. and de Pater, I. (1994). Microwave imaging of Mercury’s thermal emission at wavelengths from 0.3 to 20.5 cm. Icarus, 110, 2–32. Morbidelli, A. and Raymond, S. N. (2016). Challenges in planet formation. J. Geophys. Res. Planets, 121, 1962–1980, doi:10.1002/ 2016JE005088. Morbidelli, A., Tsiganis, K., Crida, A., Levison, H. F. and Gomes, R. (2007). Dynamics of the giant planets of the Solar System in the gaseous protoplanetary disk and their relationship to the current orbital architecture. Astron. J., 134, 1790–1798. Morbidelli, A., Lunine, J. I., O’Brien, D. P., Raymond, S. N. and Walsh, K.J. (2012). Building terrestrial planets. Annu. Rev. Earth Planet. Sci., 40, 251–275. Morgan, J. W. and Anders, E. (1980). Chemical composition of Earth, Venus, and Mercury. Proc. Natl. Acad. Sci., 77, 6973–6977. Moriarty, J., Madhusudhan, N. and Fischer, D. (2014) Chemistry in an evolving protoplanetary disk: Effects on terrestrial planet composition. Astrophys. J., 787, 81–91. Morishima, R., Golabek, G. J. and Samuel, H. (2013). N-body simulations of oligarchic growth of Mars: Implications for Hf–W chronology. Earth Planet. Sci. Lett., 366, 6–16. Murchie, S. L., Klima, R. L., Denevi, B. W., Ernst, C. M., Keller, M. R., Domingue, D. L., Blewett, D. T., Chabot, N. L., Hash, C. D., Malaret, E., Izenberg, N. R., Vilas, F., Nittler, L. R., Gillis-Davis, J. J., Head, J. W. and Solomon, S. C. (2015). Orbital multispectral mapping of Mercury with the MESSENGER Mercury Dual Imaging System: Evidence for the origins of plains units and low-reflectance material. Icarus, 254, 287–305. Nakajima, M. and Stevenson, D. J. (2015). Melting and mixing states of the Earth’s mantle after the Moon-forming impact. Earth Planet. Sci. Lett., 427, 286–295. Namur, O., Charlier, B., Holtz, F., Cartier, C. and McCammon, C. (2016). Sulfur solubility in reduced mafic silicate melts: Implications for the speciation and distribution of sulfur on Mercury. Earth Planet. Sci. Lett., 448, 102–114. Nittler, L. R., Starr, R. D., Weider, S. Z., McCoy, T. J., Boynton, W. V., Ebel, D. S., Ernst, C. M., Evans, L. G., Goldsten, J. O., Hamara, D. K., Lawrence, D. J., McNutt, R. L., Jr., Schlemm, C. E., II, Solomon, S. C. and Sprague, A. L. (2011). The major-element composition of Mercury’s surface from MESSENGER X-ray spectrometry. Science, 333, 1847–1850. O’Brien, D. P., Morbidelli A. and Levison, H. F. (2006). Terrestrial planet formation with strong dynamical friction. Icarus, 184, 39–58. O’Neill, H. St. C. and Palme H. (2008). Collisional erosion and the nonchondritic composition of the terrestrial planets. Phil. Trans. Roy. Soc. London A, 366, 4205–4238. Ostro, S. J., Hudson, R. S., Nolan, M. C., Margo, J-L., Scheeres, D. J., Campbell, D. B., Magri, C., Giorgini, J. D. and Yeomans, D. K. (2000). Radar observations of asteroid 216 Kleopatra. Science, 288, 836–839. doi:10.1126/science.288.5467.836.


The Elusive Origin of Mercury

Pasek, M. A., Milsom, J. A., Ciesla, F. J., Lauretta, D. S., Sharp C. M. and Lunine, J. I. (2005). Sulfur chemistry with time-varying oxygen abundance during Solar System formation. Icarus, 175, 1–14. Peplowski, P. N., Evans, L. G., Hauck, S. A., II, McCoy, T. J., Boynton, W. V., Gillis-Davis, J.-J., Ebel, D. S., Goldsten, J. O., Hamara, D. K., Lawrence, D. J., McNutt, R. L., Jr., Nittler, L. R., Solomon, S. C., Rhodes, E. A., Sprague, A. L., Starr, R. D. and StockstillCahill, K. R. (2011). Radioactive elements on Mercury’s surface from MESSENGER: Implications for the planet’s formation and evolution. Science, 333, 1850–1852. Peplowski, P. N., Lawrence, D. J., Rhodes, E. A., Sprague, A. L., McCoy, T. J., Denevi, B. W., Evans, L. G., Head, J. W., Nittler, L. R., Solomon, S. C., Stockstill-Cahill, K. R. and Weider, S. Z. (2012). Variations in the abundances of potassium and thorium on the surface of Mercury: Results from the MESSENGER GammaRay Spectrometer. J. Geophys. Res., 117, E00L04, doi:10.1029/ 2012JE004141. Peplowski, P. N., Evans, L. G., Stockstill-Cahill, K. R., Lawrence, D. J., Goldsten, J. O., McCoy, T. J., Nittler, L. R., Solomon, S. C., Sprague, A. L., Starr, R. D. and Weider, S. Z. (2014). Enhanced sodium abundance in Mercury’s north polar region revealed by the MESSENGER Gamma-Ray Spectrometer. Icarus, 228, 86–95. Peplowski, P. N., Lawrence, D. J., Evans, L. G., Klima, R. L., Blewett, D. T., Goldsten, J. O., Murchie, S. L., McCoy, T. J., Nittler, L. R., Solomon, S. C., Starr, R. D. and Weider, S. Z. (2015). Constraints on the abundance of carbon in near-surface materials on Mercury: Results from the MESSENGER gamma-ray spectrometer. Planet. Space Sci., 108, 98–107. Peplowski, P. N., Klima, R. L., Lawrence, D. J., Ernst, C. M., Denevi, B. W., Frank, E. A., Goldsten, J. O., Murchie, S. L., Nittler, L. R. and Solomon, S. C. (2016). Remote sensing evidence for an ancient carbon-bearing crust on Mercury, Nature Geosci., 9, 273–276, doi:10.1038/ngeo2669. Pignatale, F. C., Liffman, K., Maddison, S. T. and Brooks, G. (2016). 2D condensation model for the inner Solar Nebula: An enstatiterich environment. Mon. Not. Roy. Astron. Soc., 457, 1359–1370. Quintana, E. V., Barclay, T., Borucki, W., Rowe, J. F. and Chambers, J. E. (2016). Giant impacts on Earth-like worlds. Astrophys. J., 821, 126–139. Raymond, S. N., O’Brien, D. P., Morbidelli, A. and Kaib, N. A. (2009). Building the terrestrial planets: Constrained accretion in the inner Solar System. Icarus, 203, 644–662. Reipurth, B. and Bally, J. (2001). Herbig-Haro flows: Probes of early stellar evolution. Annu. Rev. Astron. Astrophys., 49, 195–236. Reufer, A., Meier, M. M. M., Benz, W. and Weiler, R. (2012). A hitand-run giant impact scenario. Icarus, 221, 296–299. Righter, K. (2003). Metal-silicate partitioning of siderophile elements and core formation in the early Earth. Annu. Rev. Earth Planet. Sci., 31, 135–174. Righter, K., Arculus, R. J., Delano, J. W. and Paslick C. (1990). Electrochemical measurements and thermodynamic calculations of redox equilibria in pallasite meteorites: Implications for the eucrite parent body. Geochim. Cosmochim. Acta, 54, 1803–1815. Righter, K., Drake, M. J. and Scott, E. (2006). Compositional relationships between meteorites and terrestrial planets. In Meteorites and the Early Solar System II, ed. D. Lauretta and H. Y. McSween, Jr. Tucson, AZ: University of Arizona Press, pp. 803–828. Righter, K., Humayun, M. and Danielson, L. (2008). Partitioning of palladium at high pressures and temperatures during core formation. Nature Geosci., 1, 321–323, doi:10.1038/ngeo180. Righter, K., Sutton, S. R., Danielson, L., Pando, K. and Newville, M. (2016). Redox variations in the inner solar system with new constraints from vanadium XANES in spinels. Amer. Mineral., 101, 1928–1942.

Ringwood, A. E. and Kesson, S. E. (1977). Basaltic magmatism and the bulk composition of the Moon. Moon, 16, 425–464. Rizo, H., Walker, R. J., Carlson, R. W., Horan, M. F., Mukhopadhyay, S., Manthos, V., Francis, D. and Jackson, M. G. (2016). Preservation of Earth-forming events in the tungsten isotopic composition of modern flood basalts. Science, 352, 809–812. Robinson, M. S. and Taylor, G. J. (2001). Ferrous oxide in Mercury’s crust and mantle. Meteorit. Planet. Sci., 36, 841–847. Robinson, M. S., Murchie, S. L., Blewett, D. T., Domingue, D. L., Hawkins, S. E., III, Head, J. W., Holsclaw, G. M., McClintock, W. E., McCoy, T. J., McNutt, R. L., Jr., Prockter, L. M., Solomon, S. C. and Watters, T. R. (2008). Reflectance and color variations on Mercury: Regolith processes and compositional heterogeneity. Science, 321, 66–69, doi:10.1126/science.1160080. Rubie, D. C., Jacobson, S. A., Morbidelli, A., O’Brien, D. P., Young, E. D., de Vries, J., Nimmo, F., Palme, H. and Frost, D. J. (2015). Accretion and differentiation of the terrestrial planets with implications for the compositions of early-formed Solar System bodies and accretion of water. Icarus, 248, 89–108. Rudge, J. F., Kleine, T. and Bourdon, B. (2010). Broad bounds on Earth’s accretion and core formation constrained by geochemical models. Nature Geosci., 3, 439–443. Russell, S. S., Hartmann, L., Cuzzi, J. N., Krot, A. N., Gounelle, M. and Weidenschilling, S. J. (2006). Timescales of the solar protoplanetary disk. In Meteorites and the Early Solar System II, ed. D. Lauretta and H. Y. McSween, Jr. Tucson, AZ: University of Arizona Press, pp. 233–251. Safronov, V. S. (1972). Evolution of the Protoplanetary Cloud and Formation of the Earth and Planets. Tech. Transl. F-677. Washington, DC: NASA. Sarid, G., Stewart, S. T. and Leinhardt, Z. M. (2014). Mercury, the impactor. Lunar Planet. Sci., 46, abstract 2723. Schönbächler, M., Carlson, R. W., Horan M. F., Mock T. D. and Hauri, E. H. (2010). Heterogeneous accretion and the moderately volatile element budget of Earth. Science, 328, 884–887. Simon, S. B., Sutton, S. R. and Grossman, L. (2007). Valence of titanium and vanadium in pyroxene in refractory inclusion interiors and rims. Geochim. Cosmochim. Acta, 71, 3098–3118. Smith, J. V. (1979). Mineralogy of the planets: A voyage in space and time. Mineral. Mag., 43, 1–89. Sneden, S., Lawler, J. E., Cowan, J. J., Ivans, I. I. and Den Hartog, E. A. (2009). Astrophys. J. Suppl. Ser., 182, 80–96. Solomon, S. C., McNutt, R. L., Jr., Gold, R. E., Acuña, M. H., Baker, D. N., Boynton, W. V., Chapman, C. R., Cheng, A. F., Gloeckler, G., Head, J. W., III, Krimigis, S. M., McClintock, W. E., Murchie, S. L., Peale, S. J., Phillips, R. J., Robinson, M. S., Slavin, J. A., Smith, D. E., Strom, R. G., Trombka, J. I. and Zuber, M T. (2001). The MESSENGER mission to Mercury: Scientific objectives and implementation. Planet. Space Sci., 49, 1445–1465. Solomon, S. C., McNutt, R. L., Jr., Gold, R. E. and Domingue, D. L. (2007). MESSENGER mission overview. Space Sci. Rev., 131, 3–39. Sprague, A. L. and Roush, T. L. (1998). Comparison of laboratory emission spectra with Mercury telescopic data. Icarus, 133, 174–183. Stewart, S. T. and Leinhardt, Z. M. (2012). Collisions between gravitydominated bodies. II. The diversity of impact outcomes during the end stage of planet formation. Astrophys. J., 751, 32–49. Stewart, S. T., Leinhardt, Z. M. and Humayun, M. (2013). Giant impacts, volatile loss, and the K/Th ratios on the Moon, Earth, and Mercury. Lunar Planet. Sci., 44, abstract 2306. Stewart, S. T., Lock, S. J., Petaev, M. I., Jacobsen, S. B., Sarid, G., Leinhardt, Z. M., Mukhopadhyay, S. and Humayun, M. (2016). Mercury impact origin hypothesis survives the volatile crisis: Implications for terrestrial planet formation. Lunar Planet. Sci., 47, abstract 2954.

References Svetsov, V. (2011). Cratering erosion of planetary embryos. Icarus, 214, 316–326. Szymanski, A., Brenker, F. E., Palme, H. and El Goresy, A. (2010). High oxidation state during formation of Martian nakhlites. Meteorit. Planet. Sci., 45, 21–31. Tsiganis, K., Gomes, R., Morbidelli, A. and Levison, H. F. (2005). Origin of the orbital architecture of the giant planets of the Solar System. Nature, 435, 459–461. Tuff, J., Wade, J. and Wood, B. J. (2013). Volcanism on Mars controlled by early oxidation of the upper mantle. Nature, 498, 342–345. Urey, H. (1950). The origin and development of the Earth and other terrestrial planets. Geochim. Cosmochim. Acta, 1, 209–277. Vander Kaaden, K. E. and McCubbin, F. M. (2015). Exotic crust formation on Mercury: Consequences of a shallow, FeO-poor mantle. J. Geophys. Res. Planets, 120, 195–209, doi:10.1002/ 2014je004733. Vilas, F. (1988). Surface composition of Mercury from reflectance spectrophotometry. In Mercury, ed. F. Vilas, C. R. Chapman and M. S. Matthews. Tucson, AZ: University of Arizona Press, pp. 59–76. Vityazev, A. V., Pechernikova, G. V. and Safronov, V. S. (1988). Formation of Mercury and removal of its silicate shell. In Mercury, ed. F. Vilas, C. Chapman and M. S. Matthews. Tucson, AZ: University of Arizona Press, pp. 667–669. Wadhwa, M. (2001). Redox state of Mars’ upper mantle and crust from Eu anomalies in shergottite pyroxenes. Science, 291, 1527–1530. Wadhwa, M. (2008). Redox conditions on small bodies, the Moon and Mars. Rev. Mineral. Geochem., 68, 493–510. Wagner, F. W., Sohl, F., Hussmann, H., Grott, M. and Rauer, H. (2011). Interior structure models of solid exoplanets using material laws in the infinite pressure limit. Icarus, 214, 366–376. Walsh, K. J., Morbidelli, A., Raymond, S. N., O’Brien, D. P. and Mandell, A. M. (2011). A low mass for Mars from Jupiter’s early gas-driven migration. Nature, 475, 206–209, doi:10.1038/ nature10201. Wänke, H. (1981) Constitution of terrestrial planets. Phil. Trans. Roy. Soc. London A, 303, 287–302. Warell, J. and Blewett, D. T. (2004). Properties of the Hermean regolith: V. New optical reflectance spectra, comparison with lunar anorthosites, and mineralogical modelling. Icarus, 168, 257–276. Wasson, J. T. and Kallemeyn, G. W. (1988). Composition of chondrites. Phil. Trans. Roy. Soc. London A, 325, 535–544. Wasson, J. T. and Wai, C. M. (1970). Composition of the metal, schreibersite and perryite of enstatite achondrites and the origin of enstatite chondrites and achondrites. Geochim. Cosmochim. Acta, 34, 169–184. Weidenschilling, S. J. (1978). Iron/silicate fractionation and the origin of Mercury. Icarus, 35, 99–111. Weider, S. Z., Nittler, L. R., Starr, R. D., McCoy, T. J., StockstillCahill, K. R., Byrne, P. K., Denevi, B. W., Head, J. W. and Solomon, S. C. (2012). Chemical heterogeneity on Mercury’s


surface revealed by the MESSENGER X-Ray Spectrometer. J. Geophys. Res., 117, E00L05, doi:10.1029/2012JE004153. Weider, S. Z., Nittler, L. R., Starr, R. D., McCoy, T. J. and Solomon, S. C. (2014). Variations in the abundance of iron on Mercury’s surface from MESSENGER X-Ray Spectrometer observations. Icarus, 235, 170–186. Weider, S. Z., Nittler, L. R., Starr, R. D., Crapster-Pregont, E. J., Peplowski, P. N., Denevi, B. W., Head, J. W., Byrne, P. K., Hauck, S. A., II, Ebel, D. S. and Solomon, S. C. (2015). Evidence for geochemical terranes on Mercury: Global mapping of major elements with MESSENGER’s X-Ray Spectrometer. Earth Planet. Sci. Lett., 416, 109–120. Weisberg, M. K. and Kimura, M. (2012). The unequilibrated enstatite chondrites. Chemie der Erde, 72, 101–115. Weisberg, M. K., Prinz, M. and Nehru, C. E. (1988). Petrology of ALH85085: A chondrite with unique characteristics. Earth Planet. Sci. Lett., 91, 19–32. Weisberg, M. K., Prinz, M., Clayton, R. N., Mayeda, T. K., Sugiura, N., Zashu, S. and Ebihara, M. (2001). A new metal-rich chondrite grouplet. Meteorit. Planet. Sci., 36, 3401–3418. Weisberg, M. K., Ebel, D. S., Nakashima, D., Kita, N. T. and Humayun, M. (2015). Petrology and geochemistry of chondrules and metal in NWA 5492 and GRO 95551: A new type of metal-rich chondrite. Geochim. Cosmochim. Acta, 167, 269–285. Weisberg, M. K., Bigolski, J., Ebel, D. S. and Walker, D. (2016). Calcium-aluminum-rich (CAI) and sodium-aluminum-rich (NAI) inclusions in the PAT 91546 CH chondrite. Lunar Planet. Sci., 47, abstract 2152. Wetherill, G. W. (1994). Provenance of the terrestrial planets. Geochim. Cosmochim. Acta, 58, 4513–4520. Williams, J. P. and Cieza, L. A. (2011). Protoplanetary disks and their evolution. Annu. Rev. Astron. Astrophys., 49, 67–118. Wood, B. J., Walter, M. J. and Wade, J. (2006). Accretion of the Earth and segregation of its core. Nature, 441, 825–833. Wood, B. J., Wade, J. and Kilburn, M. (2009). Core formation and the oxidation state of the Earth: Additional constraints from Nb, V and Cr partitioning. Geochim. Cosmochim. Acta, 72, 1415–1426. Wurm, G., Trieloff, M. and Rauer, H. (2013). Photophoretic separation of metals and silicates: The formation of Mercury like planets and metal depletion in chondrites. Astrophys. J., 769, 78–85. Yang J., Goldstein J. I. and Scott E. R. D. (2007). Iron meteorite evidence for early formation and catastrophic disruption of protoplanets. Nature, 446, 888–891. Zeng, L., Sasselov, D. and Jacobsen, S. (2016). Mass–radius relation for rocky planets based on PREM. Astrophys. J., 819, 127–131. Zolotov, M. Yu., Sprague, A. L., Hauck, S. A., II, Nittler, L. R., Solomon, S. C. and Weider, S. Z. (2013). The redox state, FeO content, and origin of sulfur-rich magmas on Mercury. J. Geophys. Res. Planets, 118, 138–146, doi:10.1029/ 2012JE004274.


abundance of sulfur (Nittler et al., 2011). Furthermore, MESSENGER observations showed the planet to be unexpectedly volatile rich, including considerable abundances of the heat-producing elements potassium, thorium, and uranium (Peplowski et al., 2011). The chemically reduced interior has major implications for the composition of Mercury’s core, its structure, and how the magnetic field is generated, as does the newly constrained understanding of the abundance of heatproducing elements, which control the rate at which the planet cooled and its ability to generate magma. In order to better understand how Mercury evolved over the past 4.5 billion years we synthesize observations by MESSENGER that elucidate the primary processes that have governed its history. We begin by outlining results from MESSENGER that clarify both how the crust of the planet formed and the history of the crust and lithosphere, including constraints from observations of surface geochemistry, the record of volcanism and tectonics, and the structure of the crust. Then we focus on observations that provide information on the state, structure, and behavior of the deeper interior. In tandem, we investigate the thermochemical evolution of the interior of Mercury subject to the constraints provided by MESSENGER’s observations. Finally, we discuss the implications of these results for the history of the planet and outline prospects for future progress on understanding how the whole of Mercury has evolved.

1 9 . 1 I N T R O D U C T IO N From formation to quiescence, the history of a planet is the consequence of an intricate set of relationships between processes that both shape the surface and operate through the entirety of the planet (Kaula, 1975). MESSENGER, which completed the first orbital investigation of Mercury in April 2015 (Chapter 1), has revealed that planet to be as rich an example of that intricacy as any of the major bodies of the inner solar system. Mercury has long been known as a planet of enigmas, from its 3:2 spin–orbit resonance with the Sun, to its global contraction, to its unexpected magnetic field (Solomon, 2003). Now, MESSENGER has unveiled the majority of the planet that was previously unseen (Chapters 6, 9–13), characterized the large-scale chemical composition and heterogeneity of the surface (Chapters 2, 7–8), determined Mercury’s shape, gravity, and rotational state (Chapters 3–4), and revealed unknown structure and ancient activity of the magnetic field (Chapter 5). The broad set of observations of Mercury’s surface and interior by MESSENGER places fundamental constraints on the processes governing the planet’s evolution. Although few of these observations individually lead to unique conclusions about the history of the innermost planet, taken as a whole, and in combination with an understanding of the processes that operate on and within planets in general, they provide an important picture of how Mercury evolved. At its most basic level, a planet seen today is the consequence of how material and heat have been transported on and to its surface and within the interior. Mercury’s early history was marked by both intense bombardment and widespread volcanism (Chapters 6, 9, 11). Generally overprinting this record of crustal growth and reworking is a global set of tectonic features, predominantly shortening in nature and indicative of substantial contraction of Mercury, formed largely since the end of the period of heaviest bombardment of the planet (Chapter 10). MESSENGER’s observations of remanent crustal magnetism during its final year in orbit revealed that Mercury possessed an internal magnetic field early in the planet’s history (Chapter 5). This result indicates that within the first several hundred million years of Mercury’s history, the deep interior where the magnetic field was generated was vigorously active. Each of these findings is set against the backdrop of a geochemically diverse and quite surprising surface and, by inference, interior composition (Chapters 2, 7). Indeed, MESSENGER found Mercury to be the most chemically reduced terrestrial planet on the basis of its low surface abundance of iron and relatively large surface

1 9 . 2 E A R L I E S T H I S T O R Y O F TH E C R U S T 19.2.1 Geological Constraints The geologic record of Mercury’s earliest crust – the outermost, petrologically distinct layer of the silicate portion of the planet derived from melting of the mantle (e.g., Brown and ElkinsTanton, 2009; Namur et al., 2016; Namur and Charlier, 2017; Chapter 3) – is largely obscured by resurfacing by both impacts and volcanism (e.g., Trask and Guest, 1975; Spudis and Guest, 1988; Strom and Neukum, 1988; Denevi et al., 2009). Indeed, the most heavily cratered terrain has been estimated to have an age of 4.0–4.1 Gyr (Marchi et al., 2013). However, despite the fact that there are no areas of the crust that can be quantifiably ascribed to the first ~500 Myr of Mercury’s history, important clues to the nature and origin of the crust are found in several areas that appear to have undergone only minimal resurfacing as well as in material exposed from depth by large impact events (Chapter 6).


19.2 Earliest History of the Crust Spectral units termed low-reflectance material (LRM) (Robinson et al., 2008; Denevi et al., 2009; Murchie et al., 2015; Klima et al., 2016) appear to be one key to our understanding of Mercury’s crust. With a reflectance of just 4–5% at 550 nm (Chapter 8), LRM is ~30% darker than Mercury’s average surface and is found concentrated in the ejecta of large impact craters (Denevi et al., 2009; Ernst et al., 2010; Klima et al., 2016). The reflectance and spectral properties of the LRM are consistent with the deposits having a graphite component (Murchie et al., 2015). Furthermore, increases in thermal neutron count rates associated with LRM deposits suggest a carbon abundance that is 1–3 wt% higher than that of surrounding terrain (Peplowski et al., 2015a, 2016). These observations are consistent with the hypothesis that Mercury developed a carbon-rich flotation crust due to buoyancy of graphite in an early magma ocean (Vander Kaaden and McCubbin, 2015). Any early crust, particularly one as thin as a graphite-rich crust might have been, was surely disrupted heavily by impacts, modified by magmatic intrusions, and buried by volcanic deposits. Therefore, the modern distribution of this primordial material on the surface is limited, as it has been substantially mixed and diluted with other materials. By this reasoning, LRM is the material with the greatest concentration of carbon in a C-rich crust (Peplowski et al., 2015a). The depth of origin of LRM, calculated from the excavation depth of impact craters, is often several to tens of kilometers (Denevi et al., 2009; Ernst et al., 2010, 2015; Peplowski et al., 2015a). These depth estimates provide lower bounds to the depth of burial by impact and volcanic deposits subsequent to the formation of the original flotation crust. In some of the most heavily cratered terrains, the overall surface is relatively low in reflectance and all impact craters in the region expose LRM, suggesting that these regions may have experienced less resurfacing than average (Chapter 6). However, in other large regions, no LRM is found in any crater smaller than ~150 km in diameter, suggesting burial by at least 8 km of volcanic material (Chapter 6). RiveraValentin and Barr (2014) explored impact redistribution models for an impactor population consistent with Mercury’s cratering record and found that the LRM is consistent with a darkening agent approximately 30 km deep, which would be within the lowermost crust or upper mantle (James et al., 2015; Padovan et al., 2015). Concentration of a darkening agent, such as graphite, from a layer deep within the crust may also imply that volcanism was substantial and occurred with a flux much greater than that of impact redistribution of upper crustal material in the period before the onset of the late heavy bombardment (LHB). Otherwise, the darkening agent would have been efficiently mixed throughout the crust and unlikely to display variations associated with exhumation from depth. 19.2.2 Geochemical State of the Crust and Mantle The composition and chemical diversity of the surface of Mercury provide important insights into the nature, origin, and evolution of the crust and mantle. Given that Mercury is strongly differentiated with an uncommonly low silicate-to-


metal ratio (Chapters 2, 4), understanding the mechanisms that may be responsible for Mercury’s crustal formation has been a long-standing question. At their most basic, models for the formation of the crust include partial melting of an undifferentiated, chondritic-like mantle; formation as the uppermost layer of a solidifying magma ocean; or products of remelting of a magma ocean. Geochemical observations and the relative ages of the surface units of Mercury argue against an undifferentiated mantle as the source region for melts erupted onto the surface. Melting of enstatite chondrites has been investigated experimentally and modeled from phase equilibria to understand both the origin of the highly reduced aubrite parent body (McCoy et al., 1999) and Mercury (Burbine et al., 2002; Malavergne et al., 2010). An undifferentiated chondritic mantle would produce sodium-rich melts at low degrees of partial melting, consistent with the composition of the northern smooth plains (NSP) (Vander Kaaden and McCubbin, 2016). However, the high Mg/Si and low Al/Si ratios observed for Mercury’s average surface composition require relatively high degrees of partial melting (Burbine et al., 2002; Nittler et al., 2011). Further, the formation of the high-sodium flood basalts of the NSP relatively late in the history of Mercury would require a fertile mantle source that had not experienced earlier partial melting. Finally, the highly differentiated nature of Mercury, including the presence of a large core, argues against preservation of a wholly undifferentiated mantle. A widely accepted model of the mantle and crust suggests that Mercury once had a magma ocean responsible for an initial stage of silicate differentiation. Prior to MESSENGER’s orbit insertion and the early geochemical measurements of the surface of Mercury, the nature of the crust and the bulk composition of the surface and planet were poorly constrained, although the surface was known to be FeO-poor and the bulk composition of the planet to be rich in iron metal (Taylor and Scott, 2003, and references therein). This uncertainty led to a range of magma ocean models producing either a plagioclase flotation crust or a low-FeO magmatic crust, depending on the bulk composition of the magma ocean (Brown and Elkins-Tanton, 2009; Riner et al., 2009). Some of these petrologic models produce gravitationally unstable mantles that would experience overturn, similar to that posited for the lunar mantle. With the realization that the crust of Mercury is neither a plagioclase-rich flotation crust nor chemically homogeneous, models emerged that considered a magma ocean with subsequent remelting (Charlier et al., 2013; Vander Kaaden and McCubbin, 2015, 2016). Charlier et al. (2013) suggested that compositional heterogeneity observed during early MESSENGER orbital observations could have been the result of melting of different layers within the mantle during convection and adiabatic pressure-release melting, even in the absence of mantle overturn. Vander Kaaden and McCubbin (2015) strengthened the argument against a significant primary flotation crust experimentally by demonstrating that graphite is the only phase that would be buoyant in a Mercury magma ocean. The equivalent thickness of such a graphite layer is directly dependent on the concentration of carbon in the silicate portion of the planet. Should Mercury have a bulk silicate carbon content similar to those of Earth, Mars, or the Moon, that layer


Mercury’s Global Evolution

might be up to ~100-m thick. However, if Mercury’s carbon content is more similar to that of chondritic materials, a graphite crust could range in thickness from as little as 100 m to more than 10 km, with the largest values for bulk silicate compositions similar to carbonaceous chondrites (Vander Kaaden and McCubbin, 2015). These authors further noted that, unlike the case in many other planetary bodies, partial melts derived from mantle melting on Mercury are buoyant throughout the mantle and would rise to the surface without stalling at some neutral buoyancy depth. Thus, the crust of Mercury is likely composed of an impact-gardened mixture of primary crust formed during a magma ocean stage and subsequent volcanic deposits. Vander Kaaden and McCubbin (2016) further refined this idea by noting that a crystallizing magma ocean without buoyant silicate phases would concentrate incompatible elements, including volatiles, near the surface of the planet. Thus, remelting of shallow cumulates can produce volatile-rich compositions, like the NSP, even at high degrees of partial melting.

19.3 H I STORY OF T HE CRUS T AND LITHO SPHERE Geological observations provide compelling evidence that Mercury’s crust is largely volcanic in origin and has experienced widespread tectonic deformation. The accumulated, observable history of Mercury’s crust and lithosphere contains fundamental clues to the processes that shaped the surface of the planet, and, importantly, the time progression of these processes. Whereas the earliest history of the planet may have included a magma ocean and the generation of a thin and rather exotic flotation crust, it is the subsequent history that is more discernable. MESSENGER’s collected geophysical, geological, and geochemical observations of Mercury provide important insights into both the planet’s integrated history and many discrete events, of variable duration, that reflect its evolutionary path. 19.3.1 Crustal Thickness In addition to the geochemical and geological markers of crustal formation, Mercury’s gravity field and topography provide important clues to the nature and formation of the crust (Perry et al., 2015; Tosi et al., 2015; Chapter 3). Mercury’s crust is the product of the combined processes of crystallization of any magma ocean and upward transport of mantle partial melts integrated over the course of the planet’s history. Therefore, knowledge of the thickness of the crust is a crucial indicator of the efficiency and pattern of igneous differentiation of the planet, which in turn depend strongly on Mercury’s internal activity. Orbital observations of Mercury’s gravity field by MESSENGER provided the first detailed measurements of its mass distribution. MESSENGER’s eccentric orbit (Chapter 1), with the periapsis located at a high northern latitude, resulted in gravity field measurements that have the highest spatial resolution in the north and that resolve only

much longer wavelengths in the southern hemisphere (Smith et al., 2012; Mazarico et al., 2014; Verma and Margot, 2016). Focusing on the higher-resolution information in the northern hemisphere, several estimates of the thickness of the crust have been calculated (Smith et al., 2012; James et al., 2015; Padovan et al., 2015). The most recent models place the average crustal thickness of the northern hemisphere at 35 ± 18 km on the basis of geoid-to-topography ratios (GTRs) (Padovan et al., 2015) and place a minimum on the average thickness of 38 km with a model that accounts for both crustal and mantle sources of compensation (James et al., 2015). Density differences between the crust and mantle are a major source of uncertainty in crustal thickness models. Padovan et al. (2015) considered a range of crustal densities from 2700 to 3100 kg m−3, with the upper bound consistent with grain densities they inferred from MESSENGER elemental compositions, the lower bound the result of including 12% porosity throughout the crust, as has been inferred for the Moon (Wieczorek et al., 2013), and a mantle density of 3300 kg m−3. This range overlaps independent estimates of the grain densities calculated from experimental determinations of the modal mineralogy consistent with the range of surface compositions across Mercury (Namur and Charlier, 2017). Similarly, the inversion approach of James et al. (2015) was for a nominal crustal density of 3200 kg m−3 and a mantle density of 3400 kg m−3. Generally speaking, the small difference in grain density between the crust and mantle, approximately 200 kg m−3, is a reflection of the inferred low iron content of Mercury’s silicate layers. This density difference is also important for crustal flow models, as the driving stress for any topographic relaxation via lower-crustal flow scales directly with the density contrast (e.g., Nimmo and Stevenson, 2001), so a small density contrast implies less lower-crustal flow. Potentially of greater importance is that the inferred crustal thickness values when compared with the thickness of the mantle imply that Mercury has experienced the most efficient extraction of crust among the terrestrial bodies. Indeed, Mercury’s crust represents approximately 10% of all silicate material on the planet (James et al., 2015; Padovan et al., 2015). Such efficient extraction is likely the result of relatively high degrees of partial melting, consistent with geochemical observations of the surface and inferences for the interior (Chapters 2, 7). Compared with Mercury’s global shape as derived from laser altimetry and radio occultation measurements, the geoid has a spectral power of only ~1% that of the shape at spherical harmonic degree and order 2, which indicates that topographic variations on Mercury at the longest wavelengths are largely isostatically compensated (Perry et al., 2015). Should the variations at degree and order 2 be compensated just by variations in the thickness of the crust, this difference would imply a ~24 km pole-to-equator change in crustal thickness. However, other mechanisms such as variations in density due to temperature or composition may contribute to the compensation, potentially reducing any long-wavelength crustal thickness variation (Perry et al., 2015; Tosi et al., 2015; Chapter 3). Regardless, a substantial latitudinal variation in the crustal thickness of Mercury would be an

19.3 History of the Crust and Lithosphere Smooth plains with crater-based ages Smooth plains without ages Pyroclastic vents 60°N


60°S 270°E



Figure 19.1. Overview of major geological features on Mercury. Top: Smooth plains are in purple; the darker units have estimated ages whereas the lighter-shade units are too small for reliable crater-based ages. Mapped units are from Denevi et al. (2013a) and Byrne et al. (2016). Locations of pyroclastic vents are from the compilation of Thomas et al. (2014). Bottom: Compilation of tectonic structures, with shortening structures outlined in shades of blue and extensional structures in orange. Structures in light blue are associated with smooth plains units as outlined in the top map in purple. The shortening structures are from Byrne et al. (2014), and the extensional structures are from Klimczak et al. (2012), Ferrari et al. (2014), and Chapter 10.

90°E 60°N


60°S Smooth-plains-related shortening structures Non-smooth-plains shortening structures Extensional structures

important, if as yet poorly understood, constraint on crustal production (Chapter 3). 19.3.2 Surface History One of the more direct measures of the evolution of a planet’s crust is the geological history of its surface. To first order, Mercury’s surface can be classified into units of either smooth plains or intercrater plains (Chapter 6). The former type of unit is texturally smooth and relatively sparsely cratered, displays sharp boundaries with adjacent regions, and is level to gently sloped over baselines of ~100–200 km (Trask and Guest, 1975; Denevi et al., 2013a; Chapter 6). These smooth plains units occupy about 27% of the planet’s surface (Figure 19.1) and are predominantly located in the northern hemisphere in the NSP and within and adjacent to the Caloris basin. The remainder of the surface is largely dominated by intercrater plains, which are characterized by gently rolling terrain with gradational boundaries and a greater density of secondary craters 5– 10 km in diameter than on smooth plains (Trask and Guest, 1975; Denevi et al., 2013a). The intercrater plains are situated between individual large (>30 km) craters and clusters of such craters, which generally superpose the plains and are the source of the secondary craters. As the density of superposed impact

craters appears to be the main distinction between the varieties of plains (Byrne et al., 2016), their main difference likely reflects a range in age rather than specific lithological or rheological differences (Murray et al., 1975; Strom, 1977; Spudis and Guest, 1988; Denevi et al., 2009; Whitten et al., 2014). Little evidence remains of an older, more heavily cratered surface apart from several regions that have undergone only partial resurfacing or portions of basin massifs that predate the intercrater plains (Chapter 6). Observations of Mercury have established that the planet has been heavily shaped by volcanic activity. For example, the majority of smooth plains units are interpreted as effusive volcanic deposits, on the basis of their distinct unit boundaries, embayment relations with surrounding topography, the presence of buried “ghost craters” within these units, spectral differences from neighboring terrain, and deposits located far from any large basins (Murray et al., 1974, 1975; Strom et al., 1975; Spudis and Guest, 1988; Robinson and Lucey, 1997; Head et al., 2008, 2011; Murchie et al., 2008; Robinson et al., 2008; Denevi et al., 2009, 2013a; Chapter 11). A number of other volcanic landforms formed by effusive activity have also been reported across the planet, including a small shield volcano, lobate flow margins, and lava-sculpted valles (Head et al., 2008,


Mercury’s Global Evolution

2011; Byrne et al., 2013; Hurwitz et al., 2013). Landforms attributed to explosive volcanism (e.g., Kerber et al., 2009; Thomas et al., 2014), often in close spatial proximity to smooth plains, have also been identified. The major smooth plains deposits on Mercury have crater densities that vary by up to a factor of 5 for craters larger than 10 km. However, because of the inferred rapid decline in cratering during their formation, their derived model ages are the same, within statistical error, for any of the published model production function (PF) chronologies for Mercury (Strom and Neukum, 1988; Neukum et al., 2001; Marchi et al., 2009; Le Feuvre and Wieczorek, 2011), though differences among the model chronologies are greater for lower-density (younger) deposits (Chapter 9). Crater size–frequency analyses have shown that the NSP, the single largest smooth plains deposit on the planet (Chapter 6), as well as the plains interior to the Caloris and the Rembrandt impact basins, were emplaced around 3.8–3.7 Ga (Fassett et al., 2009; Head et al., 2011; Strom et al., 2011; Denevi et al., 2013a; Ferrari et al., 2014; Ostrach et al., 2015; Chapter 9). The areal densities of impact craters for two additional large smooth plains deposits on Mercury, those near the Faulkner crater and the Rachmaninoff basin, are comparable to the densities for the NSP and Caloris interior plains (Fassett et al., 2009; Denevi et al., 2013a; Whitten et al., 2014; Ostrach et al., 2015), implying that these other units are similar in age. Crater density measurements for several additional, smaller smooth plains deposits yield ages of ~3.8– 3.5 Ga for these sites (Byrne et al., 2016) with the crater model production function of Le Feuvre and Wieczorek (2011). Only one definitively volcanic smooth plains deposit has been identified on the planet with a substantially younger age than those above. Situated within the inner peak ring of the Rachmaninoff impact basin, this deposit is considerably smaller than other plains units for which ages have been determined (Prockter et al., 2010; Marchi et al., 2011). The distribution of the model ages of smooth plains units (in particular the units shaded dark purple in Figure 19.1), which are stratigraphically the youngest effusive volcanic features on Mercury, suggest therefore that flood volcanism was largely completed by ~3.5 Ga (Byrne et al., 2016). Similar to the smooth plains, intercrater plains units have a range of crater areal densities, the lowest values of which overlap the highest corresponding values for smooth plains units (Whitten et al., 2014; Byrne et al., 2016). The model ages of the intercrater plains are ~4.1–3.9 Ga (e.g., Whitten et al., 2014; Chapter 9). Notably, nowhere on Mercury is as heavily cratered as the lunar highlands (Strom, 1977; Strom et al., 2008; Fassett et al., 2011; Marchi et al., 2011), and the most heavily cratered regions on Mercury have been dated at just 4.1–4.0 Ga (Marchi et al., 2013) with the chronology of Marchi et al. (2009). These model age results suggest that little remains of the geologic record of the earliest ~500 Myr of Mercury’s surface history (Chapters 6, 9). The origin of Mercury’s intercrater plains is less certain than that of the smooth plains, but they may also be dominantly products of volcanism. The main line of evidence lies in their age: model ages of 4.1–4.0 Ga imply major resurfacing of the earliest crust, with volcanism being a likely major cause (Head et al., 2011; Denevi et al., 2013a; Whitten et al., 2014;

Chapters 6, 11). For example, Whitten et al. (2014) showed that cratering of smooth plains, particularly by secondaries from nearby primary craters, renders those smooth deposits texturally similar to intercrater plains. Large regions within the intercrater plains have also been interpreted as volcanic in origin on the basis of a substantial deficit of the most degraded class of craters, as well as stratigraphic and color relationships that are analogous to volcanic smooth plains deposits (Denevi et al., 2013b; Chapter 6). Although discrete volcanic landforms may not have survived the history of impact bombardment of Mercury prior to the emplacement of the smooth plains, thermochemical evolution models of the planet imply that voluminous and widespread effusive volcanic activity operated for at least the planet’s first half-billion years (Michel et al., 2013; Tosi et al., 2013). If so, then the intercrater plains we observe today are likely just older smooth plains deposits (e.g., Strom, 1977; Spudis and Guest, 1988; Denevi et al., 2009; Whitten et al., 2014). This inference is consistent with the observed compositional heterogeneity on Mercury, where differences in composition do not always follow morphologic boundaries, and where smooth and intercrater plains can share similar compositions (Weider et al., 2015). The cessation of large-scale effusive volcanism on Mercury, as seen in the smooth plains and the older intercrater plains, effectively heralded the end of the crust-building phase of Mercury’s evolution, but volcanic activity in some form continued thereafter. For example, the identification of irregular pits across Mercury, often characterized by a lack of a raised rim, scalloped edges, and diffuse-edged deposits with a distinct reddish color, provides evidence for explosive volcanism having occurred on the planet (Head et al., 2008; Murchie et al., 2008; Kerber et al., 2009; Chapter 11). Some of these pyroclastic deposits may be as young as ~1 Ga (Thomas et al., 2014). Many of Mercury’s explosive volcanic landforms and deposits are spatially associated with areas of pre-existing crustal weaknesses, including the surface breaks of thrust faults underlying lobate scarps and within the heavily fractured central peaks and peak rings of craters (Figure 19.1) (Kerber et al., 2011; Thomas et al., 2014; Chapter 10). Additionally, the areal extents of pyroclastic deposits are far less than those of effusive volcanic deposits. Although widely distributed, the role of explosive volcanism in the building and resurfacing of Mercury’s crust was negligible compared with the contribution from effusive volcanism. The history of Mercury’s surface is recorded as much in its tectonic landforms as in its volcanic ones. Indeed, the surface of Mercury is replete with tectonic features, including landforms termed “wrinkle ridges” and “lobate scarps” (see the bottom panel of Figure 19.1), interpreted to have accommodated crustal shortening in response to global contraction (Strom et al., 1975). The number and structural relief of this ensemble of landforms correspond to a decrease in planetary radius of at least 5 to 7 km (Byrne et al., 2014; Chapter 10). These figures are in stark contrast with earlier estimates from more limited Mariner 10 data and early flyby data from MESSENGER, which suggested that perhaps no more than 2 km of contraction was likely (Strom et al., 1975; Watters et al., 1998, 2009). Importantly, crater and thrust fault superposition relations indicate that global contraction was underway by around the time that widespread effusive volcanism came

19.3 History of the Crust and Lithosphere 90


Figure 19.2. Overview of geochemically distinct terranes on Mercury (Chapters 2, 7). From Patrick Peplowski.





–90 –180

Geochemical Terranes: Caloris Interior Plains Northern High Mg Low Fast









East Longitude

to an end (Banks et al., 2015; Byrne et al., 2016). Observations of craters that formed during the Calorian system (Spudis and Guest, 1988; Chapter 9) and superpose scarps show that shortening of Mercury’s surface on at least a regional scale had begun at some time before ~3.6 Ga (Banks et al., 2015). Further, the discovery with MESSENGER low-altitude image data of a population of lobate scarps at least an order of magnitude smaller than previously recognized (Watters et al., 2015b), and the stratigraphic relationships between such scarps and impact craters with a range of degradation states, is suggestive that tectonic accommodation of global contraction persisted over most of Mercury’s history (Banks et al., 2015). Observations made with MESSENGER data have helped characterize the resurfacing mechanisms and history of the innermost planet. Voluminous magma genesis within Mercury’s interior likely resulted in globally extensive effusive volcanism that persisted for at least several hundred million years. This volcanic activity, together with an increase in the impact flux at the start of the LHB, has obscured the geological record of the first ~500 Myr of Mercury’s surface history. With a reduction in magma genesis as a result of secular cooling and with the horizontal compressive state in Mercury’s lithosphere resulting from global contraction, widespread effusive volcanism began to wane, with eruptive volumes decreasing with time, before ultimately ending by about 3.5 Ga. Explosive volcanism endured for far longer, but the vast majority of Mercury’s crust was in place prior to 4 Ga, and smooth plains formation constituted the tapering end of the planet’s crust-building phase. 19.3.3 Chemical and Petrological Constraints on Crustal Formation Observations by MESSENGER’s suite of geochemical sensors have provided important insight into the composition of the planet, the makeup of the crust, and how it formed (Chapters 2, 7). In particular, the X-Ray Spectrometer (XRS), Gamma-Ray Spectrometer (GRS), and Neutron Spectrometer (NS) provided spatially resolved surface abundances of U, K, and Th, as well as Si-normalized elemental abundances for Na, Mg, Al, S, Cl, Ca, Ti, Cr, Mn, Fe, and O. On a global scale, XRS measurements (Nittler et al., 2011) indicate that the surface

of Mercury exhibits a high Mg/Si ratio (0.33–0.67), which is intermediate between those of terrestrial oceanic and lunar mare basalts and highly magnesian komatiites. Mercury’s surface also exhibits lower Al/Si and Ca/Si ratios than typical terrestrial or lunar basalts. Most surprising, high S/Si ratios (0.05–0.15) suggest abundances of the moderately volatile element S up to ~4 wt%. Observations from the GRS further argue against a volatile-depleted composition for Mercury. For example, Mercury’s K/Th ratio is comparable with that of other terrestrial planets and is much higher than observed in the volatile-depleted lunar crust (Peplowski et al., 2011). Moreover, large ratios of Na/Si (0.12) and Cl/Si (0.0057) are also observed (Evans et al., 2012, 2015). Together, these observations suggest a magnesium-rich, iron-poor crust formed under chemically reducing conditions, yet not depleted in volatiles as had been predicted for an iron-rich planet so close to the Sun (e.g., Taylor and Scott, 2003). The surface of Mercury exhibits considerable chemical and, therefore by extension, mineralogical diversity. This diversity is best documented in the northern hemisphere, where high-spatial-resolution measurements allow us to distinguish discrete geochemical terranes (Figure 19.2; Chapters 2, 7). These include the Northern Geochemical Terrane, the Caloris Interior Plains Terrane, the High-Magnesium Terrane, and the “Low-Fast” Terrane (so named because it has a low count rate for fast neutrons). Among these terranes, the Northern Geochemical Terrane and the Low-Fast Terrane are present largely, though not exclusively, within the NSP. The Caloris Interior Plains Terrane corresponds spatially to the boundaries of the smooth plains within the Caloris impact basin. In contrast, the High-Magnesium Terrane is geochemically coherent in a number of features but exhibits no clear correlation with spectral or morphometric features across the entirety of the region. However, while the crustal thickness within the majority of the High-Magnesium Terrane is similar to the average for the northern hemisphere, the northern and eastern boundaries are approximately coincident with areas that transition from average to thicker-than-average crust (Chapter 3). In contrast to the well-resolved XRS measurements in the northern hemisphere, the large XRS footprints in the southern hemisphere yield only a single hemispheric average composition.


Mercury’s Global Evolution

Chemical compositions derived from the four distinct, northern hemisphere geochemical terranes range in composition from basaltic andesite to trachyte on the basis of their total alkali content (Na and K) compared with silica, but ultimately all share a boninite classification due to their high MgO (> 8 wt%) and low TiO2 (38 km, depending on the method employed (James et al., 2015; Padovan et al., 2015; Chapter 3). That the average density of the outermost solid shell of the planet is greater than expected for iron-poor silicate materials, together with estimates of the composition of Mercury’s core inferred from the strongly chemically reducing conditions discovered at the surface, has led to the consideration of a solid iron sulfide layer at the top of the core (Smith et al., 2012; Hauck et al., 2013; Padovan et al., 2014; Chapter 4). Given that both Si and S should have partitioned into the core (Chapter 2; Section 19.4.1), at the modest pressures prevalent at the top of the core, melting Fe–S–Si can yield two immiscible liquids (one Fe–S rich and the other Fe–Si rich) over a broad range of bulk compositions (Morard and Katsura, 2010). This behavior would lead to segregation of the S-bearing liquids to the shallowest portions of the liquid core, including the core–mantle boundary. Recent metal–silicate partitioning experiments at 100 kPa (1 bar) pressure, however, suggest that the range of potential core S and Si contents consistent with the surface S content may not lead to core compositions that permit immiscibility and compositional segregation (Chabot et al., 2014) (see also Chapter 4). Additional experimental work at higher pressures and varying silicate compositions are necessary to fully test the importance of liquid immiscibility in Mercury’s core and the possibility of a solid FeS layer. However, measurement of induced magnetic fields at Mercury has led to estimates of the depth to the top of the core (Johnson et al., 2016; Chapter 5) that are consistent with internal structure models. Taken together, the consistency between the internal structure models that give an estimate of the depth to the top of the liquid outer core, and the induced magnetic field analyses that yield the depth to the top of an electrically conducting layer, indicates that any FeS layer, if present, is limited in thickness. Similarly, an Fe-rich solid inner core may also be present, though constraints on its size are sparse. Internal structure models consistent with the gravity field and rotational state of Mercury are generally limited in their ability to resolve the inner core (Chapter 4), though there does appear to be a slight tendency toward relatively modest inner core radii (Hauck et al., 2013; Dumberry and Rivoldini, 2015), perhaps smaller than half the total core radius. Recent work on the dynamic coupling of the rotation of the inner core to the outer, librating solid shell of the planet indicates that, for inner cores larger than ~30% of the radius of the planet, it is necessary to know the size of the inner core in addition to the gravity field and rotation data in order to infer the moments of inertia of the planet (Peale et al., 2016; Chapter 4). Although at present it is not possible to determine independently the size of the inner core, models with inner cores larger than 30% of the radius of the planet tend to have silicate layer densities less than the densities of magnesian olivine and pyroxene, the likely dominant constituents of Mercury’s mantle. Thus, Mercury’s inner core, if present, is unlikely to have a radius more than 30% of the planet’s radius. 19.4.3 Magnetic Field Mercury’s magnetic field observations demonstrate that a global-scale field is presently being generated by a core dynamo

19.4 Knowledge of the Interior (Chapter 5). Initial data from Mariner 10, along with the more recent MESSENGER mission measurements, show that Mercury’s dynamo-generated field is relatively weak and dominated by an axially aligned dipole. The dipole dominance of the field suggests, at first glance, that Mercury’s dynamo may be quite Earth-like in its morphology, although a suite of characteristics of Mercury’s field suggest that it has distinctive properties. The weak intensity of the field challenges our understanding of how Mercury’s magnetic field is generated. Both energy- and force-balance arguments suggest that Mercury’s observed magnetic field should be at least two orders of magnitude stronger than the field measured by Mariner 10 and MESSENGER. Although the dipole is the largest harmonic in the field, the quadrupole component is relatively large, at approximately 40% of the dipole strength. This quadrupolar component – equivalent to an offset of the dipole from Mercury’s center – is larger than observed for other planets with dipole-dominated fields. Indeed, it is larger than those of other planets even when corrections are made for the relatively shorter distance from the surface to the core–mantle boundary (CMB) at Mercury, with a proportionately smaller attenuation of the quadrupole component with distance from the dynamo region. The multipolar terms beyond the quadrupole, though, are quite small. Furthermore, a property that has not received much attention to date is the axisymmetry of the dipole and quadrupole components. With the possible exception of Saturn, no other planet has a field as axisymmetric as Mercury. The combination of these three characteristics requires alterations to dynamo scenarios previously proposed for Mercury. The weakness of Mercury’s field was the first puzzle to be confronted, and several solutions have been suggested (e.g., Wicht and Heyner, 2014). For example, numerical dynamo models with very large inner cores (Stanley et al., 2005) or with very small inner cores (Heimpel et al., 2005) could produce relatively weak fields. However, current compositional, thermal, and structural models for Mercury’s core suggest that the inner core is unlikely to be sufficiently large to satisfy the large inner core models, even if the size of the inner core is weakly constrained at best (see Section 19.4.2). Another explanation for the relative weakness of Mercury’s field is that the outer portion of the core may be stably stratified, an idea consistent with the small magnitudes of the terms beyond the quadrupole in the field’s multipolar expansion. This stratification could be thermal (the result of subadiabatic heat flux at the CMB) or compositional (due to light element segregation) in origin. Such a stably

stratified layer may attenuate the field intensity observed at the surface (Christensen, 2006; Christensen and Wicht, 2008), although double-diffusive convection in the stable layer may hinder the attenuation (Manglik et al., 2010). A third suggestion is that feedback between currents generated in Mercury’s magnetosphere and those in Mercury’s core may result in a weakfield state (Glassmeier et al., 2007; Heyner et al., 2011). A fourth possibility is that, if S is the principal light element in the core, temperatures may drop below the melting temperature near the top and the middle of the core in regions often termed Fe-snow zones, where Fe would crystallize and then sink through the core; this situation contrasts with that of Earth, where crystallization first occurs at the center of the planet (Chen et al., 2008). A proposed consequence of such topdown crystallization in Mercury’s core is that there could be two separate regions of dynamo generation and that the dipole components oppose each other, yielding a weak net external field (Vilim et al., 2010). Although these scenarios offer promising avenues for understanding the weakness of Mercury’s field, they must also explain the other characteristics of the magnetic field observed by MESSENGER. None of these proposed mechanisms, by themselves, have yet been shown to lead naturally to magnetic fields with large quadrupole components and very axisymmetric fields. The combination of a large quadrupole component and an axisymmetric dipole component is particularly challenging because dynamo theory demonstrates that when a fluid velocity mode excites the generation of the axial quadrupole component, it will also excite the non-axisymmetric dipole component (Bullard and Gellman, 1954). Special circumstances may therefore apply in order to dampen only one of these magnetic modes. Two recent studies have had some success in this vein. Cao et al. (2014) imposed a north–south symmetric thermal perturbation at the CMB in a numerical dynamo model (resulting in higher heat flux at the CMB equator: see Figure 19.3) along with volumetric heat sources throughout the core. Their model matched the dipole–quadrupole dominance and axisymmetry in Mercury observations, but it did not reproduce the relatively low strength of Mercury’s field. The likelihood that such a thermal perturbation is present at Mercury’s CMB is also unclear. In contrast, a numerical dynamo model by Tian et al. (2015) instead imposed a north–south antisymmetric thermal perturbation (i.e., of spherical harmonic degree 1) at the CMB (Figure 19.3), resulting in higher heat flux in the northern hemisphere. In addition, a thin, stably stratified layer was imposed at

Relative magnitude

1 0.5

Figure 19.3. Relative variations in the imposed heat flux along the core–mantle boundary in MESSENGER-era dynamo models. The work of Cao et al. (2014) invoked a core heat flux that is higher at, and symmetric about, the equator, whereas Tian et al. (2015) assumed a core heat flux that is greater in the northern hemisphere than in the southern hemisphere.

Cao et al., 2014

0 −0.5 Tian et al., 2015 −1 −90


−30 0 30 Latitude (degrees)





Mercury’s Global Evolution

the top of the core in this model. This combination of properties resulted in a magnetic field that reproduced the dipole–quadrupole dominance, the axisymmetry, and the weakness of Mercury’s field. The north–south antisymmetric thermal perturbation in this model was justified on the basis of the concentration of smooth plains in Mercury’s northern hemisphere (Head et al., 2011). Recent work by Philpott et al. (2014) also suggested that there has been little to no secular variation in the large-scale magnetic field components between the time of the Mariner 10 flybys (1974–1975) and the four years that MESSSENGER was in orbit about Mercury. A study by Stanley and Bloxham (2016) of the saturnian dynamo suggests that if Mercury possesses a stably stratified layer at the top of the core, and if the magnetic field is very axisymmetric, then very slow secular variation of the field is a natural result. This correspondence between slow secular variation and a stably stratified layer may help to explain the lack of observed secular variation in Mercury’s magnetic field. 19.4.4 Core Properties The relative dominance of Mercury’s core as a fraction of the planet’s mass and volume (Chapter 4) underscores the influence of the core in the planet’s overall evolution. The basic properties of the core, and particularly its thermodynamic attributes, are critical for understanding how it has evolved. Siegfried and Solomon (1974) utilized a thermal conduction model for heat transport through the planet, in concert with knowledge of the thermodynamic properties of iron for the core, to investigate the thermal history and core crystallization of Mercury. More recent approaches have generally considered heat transport through the mantle via convection and various alloys of iron and sulfur for the core (e.g., Schubert et al., 1988; Hauck et al., 2004; Grott et al., 2011; Tosi et al., 2013). Over the past two decades, knowledge of the behavior of a variety of potential core-forming materials has grown considerably. The pressures within Mercury’s core, ~5–40 GPa (Hauck et al., 2013), are directly accessible in laboratory experiments. Of particular interest are the temperature and pressure dependencies of the properties of iron alloys, including the thermal conductivity, thermal expansivity, and melting behavior. It is well-known that Fe–S alloys have the peculiar behavior that their eutectic melting temperature decreases with increasing pressure (e.g., Fei et al., 1997, 2000; Li et al., 2001; Chudinovskikh and Boehler, 2007; Stewart et al., 2007; Chen et al., 2008) up to 14 GPa, with shifts in the eutectic composition toward more Fe-rich compositions at pressures up to at least 40 GPa (Stewart et al., 2007). Iron–silicon alloys, which may be present in the core as a consequence of Mercury’s chemically reduced conditions (see Section 19.4.1), behave differently from alloys of iron and sulfur. The primary distinctions are that the presence of silicon results in a smaller melting point depression than with S and the Fe–Si alloys show a strong solid solution (Kuwayama and Hirose, 2004), particularly when compared with the limited solubility of S in solid Fe, even at high pressure (Li et al., 2001). Furthermore, temperature differences between the liquidus and solidus are –5 km

Effusive volcanism ?

30 25

Explosive volcanism

20 15

Accumulation of visible contraction

10 5 0





Time before present (Ga) 0




Length of era of mantle convection (Gyr)

Figure 19.5. Results of Monte Carlo simulations of Mercury’s thermal evolution for the duration of mantle convection. A total of 351 (blue) out of 2000 models from the simulations are consistent with the constraints posed by Mercury’s magmatic evolution, global contraction, and magnetic field generation. The histogram shows the fraction of models in which mantle convection stopped at a given time. About 40% of the successful models convect to the present. Models shown in blue predict a reduction in planetary radius of between 5 and 7 km. This result should be compared with the models in orange, in which global contraction less than 5 km occurs but which otherwise satisfy the constraints, indicating the sensitivity of the inference on the longevity of mantle convection to the total observed radial contraction. Because of the uncertainty in core composition (Section 19.4.4), contraction from inner core growth is neglected in these calculations. Note that the convention for global contraction here is a negative change in radius.

activity within younger impact basins (Prockter et al., 2010; Denevi et al., 2013a; Byrne et al., 2016; Chapter 11). Explosive volcanism continued for a longer time period than did widespread plains volcanism (Kerber et al., 2009; Thomas et al., 2014). The global contraction accumulated on shortening tectonic landforms records planetary cooling from the end of the LHB to the present (Chapter 10). Following the approach of Tosi et al. (2013), the range of models that satisfy the following major constraints can be determined. Successful models must (1) produce at least 5 km of crust by partial melting of the mantle, which is a minimal requirement for producing the intercrater and smooth plains, (2) show 5–7 km of global contraction following the end of the late heavy bombardment, and (3) exhibit heat flow from the core that would permit, though not require, the generation of a magnetic field. It is worth noting that the choice of the thickness of extracted crust has little influence on the results, as long as some crust is produced. Furthermore, the requirement on heat flow from the core serves to reject those models that would preclude a thermally driven core dynamo during the earliest evolution, but it is not particularly restrictive later in the planet’s history, as core heat flux is generally small after 4 Gyr of evolution. Models that satisfy all of these constraints show some common trends. Slow cooling of the planet is required, and model mantle reference viscosities at 1600 K range from 1020 to 1022 Pa s. Additionally, most models also show an early phase

Figure 19.6. Schematic timeline of major processes in Mercury’s evolution. Evidence of the planet’s history during the first ~500 Myr has been erased by effusive volcanism and impact bombardment, as indicated by the gray shading.

of mantle heating, whereas the core cools monotonically throughout evolution. Also, although up to 100 km of crust can be produced, most models produce less than 75 km. Furthermore, surface heat flow declines from about 30 mW/ m2 at the beginning of evolution to ~10 mW/m2 today, consistent with an estimate derived from tectonic modeling (EgeaGonzález et al., 2012). Finally, and most interestingly, the ratio of the concentration of heat-producing elements in the crust to that in the primordial mantle is found to be between 2 and 4.5, which is similar to the results obtained by Tosi et al. (2013) for a core radius of 1940 km. On the other hand, the initial mantle temperature in the models is poorly constrained and can range from 1600 to 1900 K, similar to the range in initial core temperature. A typical thermochemical evolution model that satisfies the above constraints is shown in Figure 19.7, where the core and mantle temperature; the core, mantle, and surface heat flow; the radius change from thermal contraction and mantle differentiation; and the crustal thickness, stagnant lid thickness, and extent of the partial melt zone are shown as functions of time. In this model, the initial mantle temperature is 1700 K, the initial core temperature is 1875 K, the crustal thermal conductivity is 2.5 W m−1 K−1, a poorly conducting regolith layer, 5 km thick and with a thermal conductivity of 0.2 W m−1 K−1 is included, and the mantle viscosity is 1020.5 Pa s. With surface abundances of radiogenic elements of 1288 ppm 40K, 155 ppb 232 Th, and 90 ppb 238U (Peplowski et al., 2012) and a crustal enrichment factor of 3.5, this typical model has bulk silicate concentrations of heat-producing elements of 368 ppm 40K, 44 ppb 232Th, and 25 ppb 238U, similar to values for Earth and Mars. Following the late heavy bombardment (i.e., at ~3.8 Ga), the model monotonically cools at a rate of 40 K Gyr−1, with the core and mantle cooling at the same rate. Global crustal production ceases around 2.5 Ga (though is largely complete nearly 1 Gyr earlier), and a total of 25 km of crust is produced, resulting in a final crustal thickness of 30 km. Total radial contraction is just short of 7 km, with continuous accumulation of contraction following the late heavy bombardment. It is worth noting that care must be taken when interpreting the timing of crustal production from such one-dimensional

19.5 Thermochemical Models of Interior Evolution 3.0


0 4.5







T (K)





30 20

1700 Tm

q (mW m–2)

4.5 1900 (a)



ΔR (km)


0 0




100 Stagnant Lid


Melt Zone




Depth (km)

1600 5


Figure 19.7. Representative thermochemical evolution model for Mercury with parameters as discussed in the text. (a) Evolution of mantle temperature Tm and CMB temperature Tc; (b) evolution of the surface heat flux qs, mantle heat flux qm, and core heat flux qc; (c) evolution of the planetary radius change from thermal expansion and contraction of the mantle and core, Rth, from mantle differentiation, Rmd, and from the sum of the two contributions, RP; (d) evolution of the thickness of the secondary crust, of the stagnant lid, and of the region in which partial melting occurs.

ΔRP –10 4.5



Time before present (Ga)

0 4.5





Time before present (Ga)

models, as this timing can differ considerably from that determined with fully dynamical two- or three-dimensional models, which generally have crustal production concentrated earlier in the planet’s evolution but result in similar total crustal thickness values (e.g., Tosi et al., 2013). Given the uncertainties associated with the state and composition of Mercury’s core, the model shown in Figure 19.7 focuses on the most robust aspects of the core and considers only thermal contraction of the core; it does not take into account contraction by core solidification. Although, for a given amount of inner core growth, this solidification could be a major contribution to planetary contraction for an Fe–FeS core composition (e.g., Solomon, 1976; Schubert et al., 1988; Knibbe and van Westrenen, 2015), it would be less so if Si were the major alloying light element in the core (Fei et al., 2011) as the density difference between solid and liquid would be smaller because of the very small difference in Si content between solid and liquid (Kuwayama and Hirose, 2004). However, the melting behavior of core material is an important factor in core contraction arising from crystallization: S-bearing cores would experience less inner core growth due to their stronger melting point depression relative to Si-bearing alloys. The true contribution of core freezing to global contraction will likely fall between these two limiting cases, but this effect is difficult to quantify without further data on the equation of state of the Fe–S–Si system. More importantly, it is clear that there is little room for a large contribution to the observed global contraction from core crystallization. The solidification of a large volume fraction of the core would lead to significantly more total contraction than that from thermal contraction alone, e.g., crystallization of >2.5% the volume of the core (equivalent to an inner core >30% of the radius of the core) would lead to at least 2 km of additional contraction (Grott et al., 2011). Thus, the contribution of core crystallization is likely limited, as fewer models would be permitted because they would exceed the 7 km of radial contraction accommodated by tectonic deformation

and even the 9 km inferred for total planetary contraction that includes the elastic accommodation of radial contraction prior to the formation of major faults (Chapter 10). This result implies that either core solidification was close to complete by the end of the late heavy bombardment, or that only a small inner core started freezing in the recent past. Because of indications that the inner core is likely to be small (Chapter 4), the latter scenario is more likely. A three-dimensional view of the thermal evolution of a model with the same properties as discussed above is shown in Figure 19.8. Additionally, the surface temperature variation imposed by Mercury’s 3:2 spin–orbit resonance is taken into account (Chapter 4). The model is similar to that presented by Tosi et al. (2015), in which chemical composition is tracked with a particle tracer technique (Plesa et al., 2013), and uses the same initial conditions as the model shown in Figure 19.7. Figure 19.8a shows the variation in the average annual surface temperature, which ranges from 260 K to 430 K between the poles and the equatorial regions. The mantle convection pattern shown in Figure 19.8b reflects this type of temperature distribution, with downwellings (blue) more focused near the polar regions. As a result of the small thickness of Mercury’s mantle, the convective pattern shows only small-scale up- and downwellings, and the more linear structures found in earlier simulations of mantle convection with a mantle thickness of 600 km (King, 2008) are not reproduced. Toward the end of the model run, mantle convection ceases, resulting in a conductive temperature profile in the mantle (Figure 19.8 c). In this model, modern mantle temperatures reflect the forcing imposed by the insolation pattern. However, it should be noted that it takes a few hundred million years for the perturbation from insolation to diffuse to any meaningful depth. Therefore, the full extent of the temperature forcing will be reflected in the deep interior only if the 3:2 spin–orbit resonance has been stable for an extended period of time (Correia and Laskar, 2004; Noyelles et al., 2014).


Mercury’s Global Evolution

Figure 19.8. (a) Distribution of Mercury’s average near-surface temperature according to the model of Vasavada et al. (1999). Hot equatorial poles are located at 0° and 180°E longitude, whereas cold poles are located at ±90°E. (b) Interior temperature anomalies after 1 Gyr of evolution when the mantle was still convecting. The color scale refers to the two mantle slices passing through the 0° and 90° meridional planes (the x–z and y–z planes, respectively), on top of which streamlines are plotted. Within the sector of mantle between the two meridional planes, blue volumes mark the locations of downwelling flow in which temperatures are 4–5% colder than average at that depth, and the red surface is the shallowest surface of the volume of upwelling material in which temperatures are 1–2% hotter than average at that depth. (c) Interior temperature anomalies at present after the mantle transitioned to a conductive state, shown on the 0° meridional plane (x–z). Figure courtesy of Nicola Tosi.

Although the general picture of Mercury’s thermochemical evolution is consistent with the constraints provided by MESSENGER observations, details of the models may change as more data are analyzed and further data are eventually provided by new missions such as BepiColombo (Chapter 20). In particular, the amount of radial contraction documented in shortening tectonic structures has been continuously refined (Strom et al., 1975; Watters et al., 2009; Di Achille et al., 2012; Byrne et al., 2014), resulting in less stringent constraints on Mercury’s thermal evolution. Current best estimates for the total radial contraction accumulated by brittle structures since the late heavy bombardment range from 5 to 7 km (Byrne et al., 2014) but may be as large as ~9 km when elastic deformation is considered, or less than 5 km if the dip angles of the thrust faults are uniformly and surprisingly steep (Chapter 10). Importantly, larger values (>7 km) of contraction would allow for lower mantle viscosities and thus more efficient mantle convection. Alternatively, such greater contraction could also allow for a larger contribution of core solidification to the total contraction of Mercury, depending on core composition, or more likely some combination of increased cooling and core solidification. 19.5.4 Other Factors Influencing Mercury’s Thermochemical Evolution One of the factors not considered in the above models is the potential presence of heat-producing elements in Mercury’s core. At the low oxygen fugacities inferred from the high S abundance and low FeO content in Mercury’s crust (Zolotov et al., 2013), lithophile elements such as K, Th, and U can become more siderophile (Malavergne et al., 2010). McCubbin et al. (2012) estimated that up to 10% of the total inventory of U and potentially Th could have partitioned into

the core, thus providing an additional heat source that could slow global contraction. However, the differences in global contraction between models with and without heat-producing elements in the core have been found to be minor (Tosi et al., 2013), as the total inventory of heat-producing elements in the interior is only weakly affected. Partitioning of U and Th into the core tends to increase the heat flux out of the core and can extend the period during which a thermal-buoyancy-generated dynamo can operate by as much as 100 Myr. In addition to the production of partial melt in the interior, Mercury’s surface compressive stress state has likely been an important factor controlling effusive volcanism. On a contracting planet such as Mercury, extrusive volcanism may be substantially inhibited as magma pathways to the surface are shut off by maximum compressive stresses in the horizontal direction (Chapter 11). Therefore, the longevity of volcanism as observed on the surface may not be a direct indicator of the timing of melt production in the deep interior. On the other hand, local factors such as variations in the thickness of an insulating crust and/or regolith layer, which would have a lower thermal conductivity than the mantle (Section 19.5.3), largely due to higher porosities (Schumacher and Breuer, 2006), are usually not fully taken into account in thermochemical evolution models. Therefore, local volcanism may be ongoing even if global models, particularly one-dimensional models, do not predict the production of partial melt at a given time. Another energy source not treated in the above discussion is impact heating, which would be expected to contribute to the global energy balance mainly during the early phases of Mercury’s evolution. Impact heating associated with the formation of the Caloris impact basin was modeled by Roberts and Barnouin (2012), who showed that impact heat can alter mantle dynamics. In addition to the production of melt at the impact site itself, partial melting may be induced even far from the impact. Thus, the smooth plains within

19.5 Thermochemical Models of Interior Evolution and adjacent to the Caloris basin could be at least in part the consequence of the impact itself, the heat for which was stored in the mantle over an extended period of time. On the other hand, the influence of isolated impacts on the global evolution of the planet is relatively small (Roberts and Barnouin, 2012), and the conclusions drawn from the simpler models discussed above remain essentially unchanged. 19.5.5 Core Evolution MESSENGER’s unveiling of Mercury’s internal structure and the geometry and history of its internal magnetic field underscore the important role of the metallic core on the planet’s evolution. Taken in concert with the growing understanding of the properties of materials at the conditions of Mercury’s core (Section 19.4.4), which indicate the potential importance of zones of top-down crystallization and liquid–liquid immiscibility, it is clear that core evolution in Mercury differed from that in Earth. Ultimately, models of core evolution on Mercury must account for the planet’s magnetic field structure and history (Chapter 5), match the internal structure (Chapter 4), and be consistent with the magnitude of the planet’s contraction (Chapter 10). The driving mechanisms of core evolution are cooling and the chemical differentiation that results from crystallization as the core cools below its melting temperature. The rate of core cooling depends strongly on how the mantle is cooling, as all of the heat from the core must pass through the mantle on its way to the planet’s surface. Early in the planet’s history, core cooling may have been relatively rapid (Figure 19.7), especially if the planet was hot, because high internal temperatures would reduce the viscosity of the mantle and make it easier to remove heat quickly by convection. Of course, just as the cooling of the mantle slows as its initial store of heat of formation is lost and heat production follows the decay of radioactive elements, the cooling of the core slows as well. The rate of cooling of the core is important because a source of convection is necessary to drive the motions in the electrically conductive liquid metal that generate the magnetic field. A minimum condition for thermal convection throughout the entire core is that the heat flux through the CMB must exceed that which can be conducted along the adiabat. Given a thermal conductivity of 40 W m−1 K−1, previous workers (Hauck et al., 2004; Tosi et al., 2013) found the minimum core heat flux for thermal convection to be in the range of 12–19 mW m−2 for a range of possible thermal expansivity values. Such core heat fluxes were exceeded only early in Mercury’s history. The more recent, higher estimates of the thermal conductivity of pure iron at pressures near that of Mercury’s CMB (Deng et al., 2013b) of 40–120 W m−1 K−1 could increase this minimum heat flux by up to a factor of 3. Such high thermal conductivities would limit thermally driven core convection to a very short time period following planet formation. However, the presence of light alloying elements tends to decrease the thermal conductivity; for example, as little as 9 wt% Si reduces the thermal conductivity of the Fe alloy to 41–60 W m−1 K−1 (Seagle et al., 2013) at Earth’s core conditions. As Mercury’s core likely hosts considerable abundances of light elements (Section 19.4.1), the


earlier value adopted for thermal conductivity may not be far off, though the uncertainty may be considerable. Although it is possible that Mercury’s early magnetic field (Chapter 5) was driven by thermal convection, the present-day field is likely dominated by flows driven by compositional buoyancy. The simplest mechanism for generating compositional buoyancy is crystallization of a core alloy in a situation where the compositional difference between the precipitating solid and residual liquid is large, such as has been previously described in the Fe–S system. Sulfur-bearing systems are the best-studied analog for Mercury because of the broad literature on Fe–S melting and because S has such a large meltingpoint depression even at high pressure (e.g., Fei et al., 1997). The consequence of the decreasing melting temperatures and eutectic S contents with increasing pressure (Section 19.4.4) is that, if the core is composed of an Fe–S alloy, then it is likely that the crystallization of core material at these pressures began at the top, rather than the bottom, of the core (Hauck et al., 2006; Stewart et al., 2007; Chen et al., 2008; Williams, 2009). An interesting consequence of the combination of the shifts in eutectic temperature and compositions, which vary with pressure, is that two radially separated regions of the core may experience such top-down crystallization, also termed Fe snow (Chen et al., 2008). Both pre- and post-MESSENGER models (Chen et al., 2008; Dumberry and Rivoldini, 2015) of an Fe–S core indicate multiple modes of crystallization, including bottom-up (like Earth) and top-down (Fe snow). In such a system, at low S contents of ~5 wt% or less and with small inner cores, Dumberry and Rivoldini (2015) found that bottom-up crystallization would be expected. However, those workers did not model the nonideal mixing behavior observed at 14 GPa in the Fe–S system (Chen et al., 2008), which essentially requires a zone of Fe snow between 10 and 14 GPa at even very small S contents because the decrease in melting temperature is so large. With larger S contents or with larger inner core sizes, various top-down crystallization regimes are possible, whether there is a layer of crystallizing material overlying a layer in which the Fe snow re-melts, whether the crystallizing material simply falls to the top of the growing inner core (Hauck et al., 2006; Dumberry and Rivoldini, 2015), or whether there is a second layer of top-down crystallization (Chen et al., 2008). Top-down crystallization is a consequence of a situation in which the melting temperature increases as a function of depth more slowly than the actual temperature (Hauck et al., 2006; Williams, 2009). In the Fe–S system there is a marked decrease in the eutectic melting temperature with increasing pressure, as well as a reduction in the S content of the eutectic with increasing pressure, both of which lead to melting temperatures decreasing with depth for a wide range of bulk compositions. Measurements of the density and sound velocity of Fe–S liquids at high pressure also indicate that S tends to result in larger adiabatic temperature gradients relative to pure Fe liquids, enhancing this effect and extending to even lower S contents (Jing et al., 2014). As a result of the small melting-point reduction in Fe–S alloy cores with low abundances of S, such systems tend to have large inner cores, which in turn tends to concentrate S in the outer core because of the low solubility of S in solid Fe. As a consequence, Fe–S-dominated cores are likely to have


Mercury’s Global Evolution

experienced Fe snow regardless of their composition. However, such large inner cores are not favored in structural models constrained by Mercury’s rotational dynamics (Chapter 4). Even though our understanding of the evolutionary paths of Mercury’s core under scenarios in which S is the sole light element is becoming more mature, it is also clear that other light elements in addition to, or instead of, S are likely to be present in the core (Section 19.4.1; Chapter 2). As noted above, carbon is generally a siderophile element, but it has been suggested that C is present as graphite in the mantle and that graphite may have formed an early flotation crust on the planet (Vander Kaaden and McCubbin, 2015), an idea that is consistent with spectral reflectance and neutron spectroscopy observations of the surface (Murchie et al., 2015; Peplowski et al., 2015a, 2016). Consequently, if the core and mantle formed in equilibrium then the core may be saturated in C, although the total amount would be small as the maximum solubility of C in Fe is ~4 wt% and that value decreases with increasing pressure (Lord et al., 2009). This value would be larger if Fe3C were present, but the density and compressibility of C-bearing alloys are such that it would be difficult for C to be the sole light element in Mercury’s core. However, the consequences of even some C being present might be important. For example, the decreasing amount of C in eutectic melts with increasing pressure in the Fe–Fe3C system is consistent with top-down crystallization, even without S. In contrast, the presence of silicon, which is likely because of the planet’s strongly reducing conditions (see Sections 19.4.1 and 19.4.4), has rather different implications for the evolution of the core. The melting behavior of Si-bearing Fe alloys at conditions appropriate to Mercury is more poorly known than for alloys with S or even C. The phase diagram of Fe–FeSi at 21 GPa determined experimentally by Kuwayama and Hirose (2004) is instructive, as they found that the eutectic point is at both a higher temperature and a larger Si abundance than at 100 kPa (1 bar). They also found, as noted above, that the difference in composition between the coexisting solid and liquid phases at temperatures between the solidus and liquidus on the Fe side of the eutectic is very small: there is a maximum of ~2 wt% Si between the solid and liquid phases. An important consequence of this behavior is an Earth-like bottom-up crystallization of the core, but with residual liquids left by crystallizing of Fe–Si core material that would be only marginally less dense than surrounding material, limiting the buoyancy available to drive convection were the core sufficiently chemically reduced that Si were the only light alloying element present. Perhaps most critical to understanding the evolution of Mercury’s core is the behavior of Fe alloys with combinations of S, Si, and possibly C. Despite the fact that the thermodynamic properties of multi-component Fe alloys are less well known than for the binary systems, the data that are available suggest interesting evolutionary paths for Mercury’s core. For example, liquid immiscibility, such as displayed in both Fe–S–C (e.g., Dasgupta et al., 2009) and Fe–S–Si liquids (Section 19.4.1), has potential consequences for compositional segregation within the outer core. Fe–S–C immiscibility would have an influence within only a relatively thin layer near Mercury’s CMB because immiscible behavior occurs only at pressures less than 6 GPa (Dasgupta et al., 2009), which is close to the possible CMB

pressure (Chapter 4). However, immiscibility in the Fe–S–Si system would extend deeper within Mercury’s outer core (Section 19.4.4). Such segregation, if present, likely developed early in the planet’s history during metal–silicate differentiation and immediately thereafter. For bulk core compositions near the miscibility limit, however, there is a possibility that the progressive crystallization of an Fe–Si-rich solid and resultant increase in concentration of S in the liquid would drive Mercury’s core into a liquid immiscibility state later in its evolution. For this situation to occur, however, relatively large inner core growth would be required to substantially change the outer core composition, an outcome that is inconsistent with models of Mercury’s thermal contraction discussed above and estimates of the planet’s internal structure (Chapter 4). A relative lack of experimental data limits firm conclusions about the crystallization behavior in an Fe–S–Si core. Recent experimental results on the Fe–S–Si–C system (Martin et al., 2015) indicate eutectic melting temperatures similar to those of the Fe–S–C system at ~4–15 GPa, with minimal pressure dependence of the eutectic. Top-down crystallization would be favored in that system. However, data on the pressure dependence of melting in the Fe–S–Si system are not available at present. While the melting behavior of the Fe–S and Fe–S–Si–C systems suggest that top-down crystallization is likely, the Fe– Si system appears more consistent with a bottom-up crystallization sequence; whether the effects of alloying with S or Si would dominate that behavior is unclear without further data. Determination of melting behavior in the Fe–S–Si system, and of the thermodynamic properties that control the adiabatic temperature gradient, are crucially needed in order to understand more fully the crystallization of Mercury’s core.

1 9. 6 DISC USS I ON MESSENGER observations have substantially altered our understanding of how Mercury has evolved over its history, but several crucial questions remain open. In particular, we are at a relatively early stage in understanding the connection between the dynamics of the mantle and the production of the crust and the generation of the magnetic field. We next discuss these issues in more detail, focusing on open questions that may be addressed through a combination of analysis of MESSENGER data, modeling, and the acquisition of new observations from BepiColombo and other future missions. 19.6.1 Crustal Production and Mantle Dynamics Global crustal production through time is a primary indicator of the evolution of a planet – that of its crust and of the interior from which the crust was derived. For planets without crustal recycling, the crust represents a nearly complete time history of intrusive and extrusive volcanism. This history, even when known only to first order, places powerful constraints on our understanding of the evolution of the interior (e.g., Hauck and Phillips, 2002). On Mercury, the clearest constraints on crustal formation are that the best estimate of its average thickness is approximately 35 km (James et al., 2015; Padovan et al., 2015; Chapter 3) and that the tail end of the era of effusive volcanism

19.6 Discussion postdates the Caloris impact by perhaps a few hundred million years at most (Byrne et al., 2016; Chapters 6, 11). Intercrater plains, also interpreted to be dominantly volcanic in origin, are more areally extensive than the smooth plains and in places are as old as 4.1–4.0 Ga (Chapter 6). The first ~500 Myr of Mercury’s surface history is also veiled by the overprinting of the late heavy bombardment. Regardless, MESSENGER observations have revealed that Mercury has experienced the most efficient and complete differentiation of mantle and crust among the terrestrial planets, and that this crust was largely built up by successive episodes of effusive volcanism that were likely largely complete within the first 1 Gyr of planet history. Given that Mercury has such a thin mantle, prone to relatively sluggish mantle flow and even the cessation of mantle convection entirely, it is remarkable that generation of the crust could have been so efficient and rapid – particularly in light of the idea that crustal products of a magma ocean may have been only meters thick (e.g., Vander Kaaden and McCubbin, 2015), leaving virtually all of the crust to be produced by serial magmatism. However, because of the low FeO content and modest pressures in Mercury’s mantle, the partial melts produced throughout the mantle would be exceptionally buoyant and less susceptible to stalling during ascent (Vander Kaaden and McCubbin, 2015), perhaps facilitating such efficient crustal formation. The heterogeneity of Mercury’s crustal production as observed in its geochemical diversity (e.g., Weider et al., 2015; Chapters 2, 7), and the spatial distribution of smooth plains volcanism, also provide important clues to the history and dynamics of the interior. Indeed, observations by MESSENGER’s suite of geochemical sensors indicate both a range of crustal compositions that point to partial melting from multiple sources (Charlier et al., 2013), and a spatial heterogeneity in compositions that does not always follow geomorphological unit boundaries (Peplowski et al., 2015b; Weider et al., 2015). Interestingly, in a manner similar to the Moon’s spatial dichotomy in mare volcanism between its near and far sides, and the asymmetric concentration of volcanism on Mars near the Tharsis province, there is a distinctive spatial difference in the abundance of smooth plains units between Mercury’s northern and southern hemispheres (Chapters 6, 11). The largest expanses of smooth plains reside at high northern latitudes and within and around the Caloris basin, which is also located in the northern hemisphere. Smaller areas of smooth plains are found generally in proximity to impact basins, with little difference in areal coverage between the hemispheres (Chapter 6). Consequently, the processes responsible for the formation of smooth plains in the Caloris region and the northern volcanic plains may be different from those that yielded the isolated, small smooth plains units distributed more evenly throughout the northern and southern hemispheres. Any hemispherical differences in the earlier volcanic activity that produced the intercrater plains are not clear at this time, though some regions also appear to be associated with impact basins (Denevi et al., 2013b). Although MESSENGER provided global geochemical coverage of Mercury, the spacecraft’s highly eccentric orbit and high northern periapsis resulted in measurements only at low spatial resolution in the southern hemisphere. That these measurements cannot resolve distinct geochemical terranes in the southern hemisphere limits our understanding of the global


evolution of Mercury. The planned orbit for the Mercury Planetary Orbiter on the BepiColombo mission (Chapter 20) will yield higher-resolution southern hemisphere data and may resolve additional geochemical terranes analogous to those observed by MESSENGER in the northern hemisphere. These heterogeneities in the geochemical and volcanic character of the surface are largely connected to the thermal and chemical properties of the mantle. Mercury’s thin mantle yielded a style of mantle convection that was both relatively sluggish and strongly spatially restricted, because the size of individual convective cells would have been on the order of the thickness of the mantle itself. Thus, the large expanses of volcanism in the northern hemisphere require conditions that either permit extraordinarily voluminous magma production from spatially restricted upwellings or conditions that substantially altered the mantle flow dynamics from that expected on the basis of Mercury’s mantle thickness. One such mechanism for altering those dynamics is a large impact, such as that which formed the Caloris basin. Indeed, the large thermal perturbation imparted by shock heating from the Caloris impact event may have led to substantial heating of the shallow mantle beneath the impact, but it might also have enhanced some nearby, preexisting mantle upwellings that generated magma distal from the impact site (Roberts and Barnouin, 2012). Such a mechanism could have been a major contributor to the generation of the Caloris-centric volcanism, but the northern volcanic plains do not appear to host such a large impact capable of triggering such volcanism, even though Caloris and the northern volcanic plains have indistinguishable crater size–frequency distributions and thus ages (e.g., Ostrach et al., 2015). On the other hand, both the broad geochemical heterogeneity across the surface and the smaller, more distributed areas of smooth plains on Mercury could be direct consequences of the small, spatially restricted upwellings and inefficient mixing in a mantle of small thickness. This fluid dynamic behavior of the mantle could act to preserve large-scale geochemical heterogeneities, yet also focus volcanism in locally restricted areas. An important question regarding the era of dwindling effusive volcanism is the relative importance of the pattern of convection (e.g., small yet relatively abundant upwellings) to the total cooling of Mercury that led to a strongly compressive stress state, one that tended to favor intrusive over extrusive volcanic activity. 19.6.2 Evolution of the Core and Magnetic Field The operation of an internally generated planetary magnetic field is a fundamental indicator of the dynamical behavior of the deep interior of a planet. MESSENGER observations of Mercury’s magnetic field have provided important constraints on the character of field generation at present as well as early in the planet’s history. Orbital measurements of the geochemical character of the surface materials, as well as gravity and rotational state determinations by MESSENGER, also help to characterize the core. However, these new observations raise a number of interesting questions about the behavior of the interior over the history of the planet. In particular, the mechanism of magnetic field generation may require a number of special conditions in order to produce a weak, axisymmetric field with a large dipole offset. Further, the magnetic field, with remanent


Mercury’s Global Evolution

crustal magnetism indicating an ancient field in addition to the modern field, places limits on the rate of cooling over the planet’s history. Although explaining Mercury’s weak magnetic field has long been a challenge (e.g., Heimpel et al., 2005; Stanley et al., 2005; Christensen, 2006), it is the combination of the weakness of the field and its axial alignment and asymmetry about the equator that makes understanding the dynamo mechanism even more intriguing. A common thread in many models of Mercury’s magnetic field is the presence of a layer stable against convection (e.g., Christensen, 2006; Vilim et al., 2010; Tian et al., 2015). If such a layer is present, most likely at the top of the fluid core, then the heat flux out of the core may be less than what can be conducted along the adiabatic temperature gradient. In addition, compositional stratification may also be present. As discussed above, it is quite likely that there is a thermal component to the stability of such a layer, as thermal history calculations generally predict a subadiabatic heat flux at present. Furthermore, many potential core alloy compositions favor top-down crystallization regimes that lead to compositionally stratified layers. Thus, it seems likely that Mercury’s core contains a stable layer that plays a role in the strength and geometry of the planet’s magnetic field. Yet the presence of a stable layer alone appears insufficient for explaining the strength and geometry of Mercury’s magnetic field. To that end, recent models have included additional heterogeneity capable of further influencing magnetic field character (e.g., Figure 19.3). In particular, both Cao et al. (2014) and Tian et al. (2015) imposed laterally variable heat flux conditions at the CMB. Cao et al. (2014) utilized a heat flux pattern symmetric about the equator similar to the latitudinal variation in surface temperature consistent with Mercury’s small axial tilt. Should the mantle be in a conductive, rather than convective state, then surface temperature variations at the surface may also have a signature at the CMB if enough time has passed since the end of the convective motions. Cao et al. (2014) investigated models with the highest or lowest heat flow at the equator, and they found that models with higher heat flow near the equator were better able to stabilize fields with geometries similar to those observed by MESSENGER. However, the mechanism for inducing larger heat fluxes along the equator, rather than at the poles, is unclear. Diffusion of surface temperatures to the CMB would result in relatively lower mantle temperatures near the poles, and therefore larger temperature differences and heat fluxes across the CMB there, rather than at the equator. As demonstrated in Figure 19.8, the limited thickness of the mantle seems to preclude long-wavelength convective patterns, so a degree-2 style of mantle convection appears unlikely as well. Therefore, some other mechanism for inducing a symmetric equator-to-pole variation in heat flux appears necessary for this mode of dynamo generation to operate. Alternatively, Tian et al. (2015) imposed an asymmetric heat flux boundary condition along the CMB, with a higher heat flux out of the core near the north pole of Mercury (Figure 19.3). Those authors postulated that the higher heat flux there is a remnant of the magmatism that produced the NSP. As discussed in the previous section, there is a notable spatial dichotomy in the distribution of the youngest smooth plains on Mercury, with

the largest expanses in the northern hemisphere (e.g., Ostrach et al., 2015; Chapters 6, 11). However, as those volcanic deposits were emplaced at 3.8–3.7 Ga, the thermal conditions that generated them are likely no longer present. Furthermore, smaller though still extensive ( >105 km2 area) (Byrne et al., 2016) smooth plains units, the youngest effusive volcanic deposits on Mercury, are relatively well distributed between the northern and southern hemispheres, exclusive of the NSP and the plains associated with Caloris. Thus, it is worth considering whether the mechanisms for the large volcanic deposits and the smaller, more evenly distributed smooth plains deposits are the same (including whether some of the smaller deposits are even volcanic). Whereas the relatively larger concentrations of K at high northern latitudes on Mercury (Peplowski et al., 2012; Chapter 7) might argue for a mantle source more enriched in heat-producing elements, such enhanced heat production would in fact lead to smaller temperature contrasts and a lower heat flux across the CMB. Interestingly, the K enhancement at high northern latitudes does not respect the morphologic boundaries of the northern plains, nor are the lavas in Caloris so enriched. However, if the generation of the NSP substantially depleted the mantle at high northern latitudes of heat-producing elements compared with the rest of the planet, then core heat fluxes might be somewhat higher there due to the cooler mantle temperatures. The relatively limited amount of lateral mixing of the mantle expected under low-Rayleigh-number convection, coupled with the small scale of convection, could act to preserve such heterogeneity. It is worth noting that MESSENGER gravity and topography data indicate that the domical rise within the northern volcanic plains is substantively compensated within ~100 km of the CMB (James et al., 2015). James et al. (2015) investigated a variety of mechanisms for the source of this compensation, including relief along a compositional interface (e.g., between the silicate mantle and a possible solid FeS layer at the top of the core) as well as other density variations. Variations in the thickness of an FeS layer would also result in changes in the thermal conductivity profile above the liquid core, leading to lateral differences in heat flux. A variety of compositions or viscosities at that depth may also have induced additional thermal heterogeneity, though the impact of such variations relative to the remainder of the planet remains to be investigated. It is clear that heterogeneity within Mercury’s mantle may influence the mechanisms by which the planet’s magnetic field is generated, though more work – and the need for further observations – remains. Indeed, any geochemical and petrologic heterogeneity (Chapters 2, 7) inherited from Mercury’s earliest history may have substantially influenced the planet’s history; yet, as less is known about the geochemical and geophysical character of the entire southern hemisphere than the north, we have much more to learn about the distribution of any heterogeneous properties of the interior. Mercury’s internal structure and chemical makeup strongly influence the manner by which the planet’s core, and therefore its magnetic field, has evolved. The discovery of Mercury’s remanent crustal magnetism (Johnson et al., 2015; Chapter 5) in crust that was last emplaced before ~3.7 Ga raises the question of how a planet cooling as modestly as suggested by its

19.7 Conclusions record of global contraction could have hosted both a relatively protracted period of early magnetic field generation and a modern field. A purely thermally generated dynamo that spans both time periods is unlikely, as the thermal history models indicate that core heat flux dropped below the critical value for convection early in the planet’s history and remains so. Indeed, early-onset thermal dynamos would tend to be short-lived, as evidenced by Figure 19.7 and previous modeling efforts (Hauck et al., 2004; Grott et al., 2011; Tosi et al., 2013). Although much shorter than the upper bound of ~800 Myr implied by the surface age of the crust in areas of remanent magnetism, such shorter-duration dynamos are potentially consistent with observations, as the column of crust hosting the remanence may predate the surface age. Models with longer-lived supercritical core heat fluxes are also possible. Under that scenario, the simplest explanation for the modern magnetic field is that it restarted comparatively recently as a result of the onset of core crystallization and perhaps even inner core growth. Alternatively, core crystallization that operated throughout the past 3.7 Gyr would account for both the ancient and modern fields. This mechanism is possible, yet would likely result in solidification of a substantial fraction of the core and greater contraction of the planet than has been documented so far. A large inner core does not appear to be compatible with the planet’s internal structure (Hauck et al., 2013; Dumberry and Rivoldini, 2015; Peale et al., 2016) nor with magnetic field generation, as compositional gradients imposed by top-down crystallization, coupled with a large inner core, may serve to stabilize the entire core against convection (Dumberry and Rivoldini, 2015; Rückriemen et al., 2015). Thus, a full understanding of the operation and evolution of Mercury’s magnetic field depends on characterizing the age distribution of remanent crustal magnetism and understanding how core evolution, including the effects of core chemistry, was coupled to mantle convection and cooling through time.

1 9. 7 CON CL US I O NS MESSENGER has been instrumental in unveiling key elements of the global evolution of Mercury. From firmly establishing the occurrence of volcanism and its distribution in space and time, to substantively resolving the long-standing paradox between predicted and observed values for Mercury’s global contraction and cooling, MESSENGER has brought new insight to fundamental questions about the planet that stood for nearly four decades. In turn, and as with all new missions of discovery, MESSENGER has raised new questions about how Mercury has operated over its history. With Mercury’s remarkably thin mantle, which is incapable of significantly homogenizing its chemical character by mantle convection, it is clear that chemical heterogeneity has played an important role in the planet’s history. The weak, axially aligned, and northward offset geometry of the internally generated magnetic field may be a distinct manifestation


of internal heterogeneity. However, it is the discovery of Mercury’s ancient magnetic field, recorded in the crustal rocks, that may hold some of the deepest clues to the planet’s internal evolution.

REFER E NC ES Anderson, B. J., Johnson, C. L., Korth, H., Purucker, M. E., Winslow, R. M., Slavin, J. A., Solomon, S. C., McNutt, R. L., Jr., Raines, J. M. and Zurbuchen, T. H. (2011). The global magnetic field of Mercury from MESSENGER orbital observations. Science, 333, 1859–1862, doi:10.1126/science.1211001. Anderson, B. J., Johnson, C. L., Korth, H., Winslow, R. M., Borovsky, J. E., Purucker, M. E., Slavin, J. A., Solomon, S. C., Zuber, M. T. and McNutt, R. L., Jr. (2012). Low-degree structure in Mercury’s planetary magnetic field. J. Geophys. Res., 117, E00L12, doi:10.1029/2012je004159. Anderson, J. D., Jurgens, R. F., Lau, E. L., Slade, M. A., III and Schubert, G. (1996). Shape and orientation of Mercury from radar ranging data. Icarus, 124, 690–697, doi:10.1006/icar.1996.0242. Banks, M. E., Xiao, Z., Watters, T. R., Strom, R. G., Braden, S. E., Chapman, C. R., Solomon, S. C., Klimczak, C. and Byrne, P. K. (2015). Duration of activity on lobate-scarp thrust faults on Mercury. J. Geophys. Res. Planets, 120, 1751–1762, doi:10.1002/ 2015je004828. Berthet, S., Malavergne, V. and Righter, K. (2009). Melting of the Indarch meteorite (EH4 chondrite) at 1 GPa and variable oxygen fugacity: Implications for early planetary differentiation processes. Geochim. Cosmochim. Acta, 73, 6402–6420, doi:10.1016/j.gca.2009.07.030. Beuthe, M. (2010). East–west faults due to planetary contraction. Icarus, 209, 795–817, doi:10.1016/j.icarus.2010.04.019. Blair, D. M., Freed, A. M., Byrne, P. K., Klimczak, C., Prockter, L. M., Ernst, C. M., Solomon, S. C., Melosh, H. J. and Zuber, M. T. (2013). The origin of graben and ridges in Rachmaninoff, Raditladi, and Mozart basins, Mercury. J. Geophys. Res. Planets, 118, 47–58, doi:10.1029/2012JE004198. Breuer, D., Hauck, S. A., II, Buske, M., Pauer, M. and Spohn, T. (2007). Interior evolution of Mercury. Space Sci. Rev., 132, 229–260, doi:10.1007/s11214-007-9228-9. Brown, S. M. and Elkins-Tanton, L. T. (2009). Compositions of Mercury’s earliest crust from magma ocean models. Earth Planet. Sci. Lett., 286, 446–455, doi:10.1016/j.epsl.2009.07.010. Bullard, E. and Gellman, H. (1954). Homogeneous dynamos and terrestrial magnetism. Phil. Trans. Roy. Soc. London A, 247, 213– 278, doi:10.1098/rsta.1954.0018. Burbine, T. H., McCoy, T. J., Nittler, L. R., Benedix, G. K., Cloutis, E. A. and Dickinson, T. L. (2002). Spectra of extremely reduced assemblages: Implications for Mercury. Meteorit. Planet. Sci., 37, 1233–1244, doi:10.1111/j.1945-5100.2002.tb00892.x. Byrne, P. K., Klimczak, C., Williams, D. A., Hurwitz, D. M., Solomon, S. C., Head, J. W., Preusker, F. and Oberst, J. (2013). An assemblage of lava flow features on Mercury. J. Geophys. Res. Planets, 118, 1303–1322, doi:10.1002/jgre.20052. Byrne, P. K., Klimczak, C., Şengör, A. M. C., Solomon, S. C., Watters, T. R. and Hauck, S. A., II (2014). Mercury’s global contraction much greater than earlier estimates. Nature Geosci., 7, 301–307, doi:10.1038/ngeo2097. Byrne, P. K., Ostrach, L. R., Fassett, C. I., Chapman, C. R., Denevi, B. W., Evans, A. J., Klimczak, C., Banks, M. E., Head, J. W. and Solomon, S. C. (2016). Widespread effusive volcanism on Mercury likely ended by about 3.5 Ga. Geophys. Res. Lett., 43, 7408–7416, doi:10.1002/2016GL069412.


Mercury’s Global Evolution

Cao, H., Aurnou, J. M., Wicht, J., Dietrich, W., Soderlund, K. M. and Russell, C. T. (2014). A dynamo explanation for Mercury’s anomalous magnetic field. Geophys. Res. Lett., 41, 4127–4134, doi:10.1002/2014gl060196. Chabot, N. L., Wollack, E. A., Klima, R. L. and Minitti, M. E. (2014). Experimental constraints on Mercury’s core composition. Earth Planet. Sci. Lett., 390, 199–208, doi:10.1016/j.epsl.2014.01.004. Charlier, B., Grove, T. L. and Zuber, M. T. (2013). Phase equilibria of ultramafic compositions on Mercury and the origin of the compositional dichotomy. Earth Planet. Sci. Lett., 363, 50–60, doi:10.1016/ j.epsl.2012.12.021. Chen, B., Li, J. and Hauck, S. A., II (2008). Non-ideal liquidus curve in the Fe–S system and Mercury’s snowing core. Geophys. Res. Lett., 35, L07201, doi:10.1029/2008gl033311. Christensen, U. R. (2006). A deep dynamo generating Mercury’s magnetic field. Nature, 444, 1056–1058, doi:10.1038/nature05342. Christensen, U. R. and Wicht, J. (2008). Models of magnetic field generation in partly stable planetary cores: Applications to Mercury and Saturn. Icarus, 196, 16–34, doi:10.1016/j. icarus.2008.02.013. Chudinovskikh, L. and Boehler, R. (2007). Eutectic melting in the system Fe–S to 44 GPa. Earth Planet. Sci. Lett., 257, 97–103, doi:10.1016/j.epsl.2007.02.024. Correia, A. C. M. and Laskar, J. (2004). Mercury’s capture into the 3/2 spin-orbit resonance as a result of its chaotic dynamics. Nature, 429, 848–850, doi:10.1038/nature02609. Dasgupta, R., Buono, A., Whelan, G. and Walker, D. (2009). Highpressure melting relations in Fe–C–S systems: Implications for formation, evolution, and structure of metallic cores in planetary bodies. Geochim. Cosmochim. Acta, 73, 6678–6691, doi:10.1016/ j.gca.2009.08.001. de Koker, N., Steinle-Neumann, G. and Vlcek, V. (2012). Electrical resistivity and thermal conductivity of liquid Fe alloys at high P and T, and heat flux in Earth’s core. Proc. Natl. Acad. Sci., 109, 4070–4073, doi:10.1073/pnas.1111841109. Denevi, B. W., Robinson, M. S., Solomon, S. C., Murchie, S. L., Blewett, D. T., Domingue, D. L., McCoy, T. J., Ernst, C. M., Head, J. W., Watters, T. R. and Chabot, N. L. (2009). The evolution of Mercury’s crust: A global perspective from MESSENGER. Science, 324, 613– 618, doi:10.1126/science.1172226. Denevi, B. W., Ernst, C. M., Meyer, H. M., Robinson, M. S., Murchie, S. L., Whitten, J. L., Head, J. W., Watters, T. R., Solomon, S. C., Ostrach, L. R., Chapman, C. R., Byrne, P. K., Klimczak, C. and Peplowski, P. N. (2013a). The distribution and origin of smooth plains on Mercury. J. Geophys. Res. Planets, 118, 891–907, doi:10.1002/jgre.20075. Denevi, B. W., Ernst, C. M., Whitten, J. L., Head, J. W., Murchie, S. L., Watters, T. R., Byrne, P. K., Blewett, D. T., Solomon, S. C. and Fassett, C. I. (2013b). The volcanic origin of a region of intercrater plains on Mercury. Lunar Planet. Sci., 44, abstract 1218. Deng, L., Fei, Y., Liu, X., Gong, Z. and Shahar, A. (2013a). Effect of carbon, sulfur and silicon on iron melting at high pressure: Implications for composition and evolution of the planetary terrestrial cores. Geochim. Cosmochim. Acta, 114, 220–233, doi:10.1016/j.gca.2013.01.023. Deng, L., Seagle, C., Fei, Y. and Shahar, A. (2013b). High pressure and temperature electrical resistivity of iron and implications for planetary cores. Geophys. Res. Lett., 40, 33–37, doi:10.1029/ 2012GL054347. Di Achille, G., Popa, C., Massironi, M., Epifani, E. M., Zusi, M., Cremonese, G. and Palumbo, P. (2012). Mercury’s radius change estimates revisited using MESSENGER data. Icarus, 221, 456–460. Dombard, A. and Hauck, S. A., II (2008). Despinning plus global contraction and the orientation of lobate scarps on Mercury: Predictions

for MESSENGER. Icarus, 198, 274–276, doi:10.1016/j. icarus.2008.06.008. Dumberry, M. and Rivoldini, A. (2015). Mercury’s inner core size and core-crystallization regime. Icarus, 248, 254–268, doi:10.1016/j. icarus.2014.10.038. Egea-González, I., Ruiz, J., Fernández, C., Williams, J.-P., Márquez, Á. and Lara, L. M. (2012). Depth of faulting and ancient heat flows in the Kuiper region of Mercury from lobate scarp topography. Planet. Space Sci., 60, 193–198, doi:10.1016/j.pss.2011.08.003. Ernst, C. M., Murchie, S. L., Barnouin, O. S., Robinson, M. S., Denevi, B. W., Blewett, D. T., Head, J. W., Izenberg, N. R., Solomon, S. C. and Roberts, J. H. (2010). Exposure of spectrally distinct material by impact craters on Mercury: Implications for global stratigraphy. Icarus, 209, 210–223, doi:10.1016/j.icarus.2010.05.022. Ernst, C. M., Denevi, B. W., Barnouin, O. S., Klimczak, C., Chabot, N. L., Head, J. W., Murchie, S. L., Neumann, G. A., Prockter, L. M., Robinson, M. S., Solomon, S. C. and Watters, T. R. (2015). Stratigraphy of the Caloris basin, Mercury: Implications for volcanic history and basin impact melt. Icarus, 250, 413–429, doi:10.1016/j.icarus.2014.11.003. Evans, L. G., Peplowski, P. N., Rhodes, E. A., Lawrence, D. J., McCoy, T. J., Nittler, L. R., Solomon, S. C., Sprague, A. L., StockstillCahill, K. R., Starr, R. D., Weider, S. Z., Boynton, W. V., Hamara, D. K. and Goldsten, J. O. (2012). Major-element abundances on the surface of Mercury: Results from the MESSENGER GammaRay Spectrometer. J. Geophys. Res., 117, E00L07, doi:10.1029/ 2012je004178. Evans, L. G., Peplowski, P. N., McCubbin, F. M., McCoy, T. J., Nittler, L. R., Zolotov, M. Yu., Ebel, D. S., Lawrence, D. J., Starr, R. D., Weider, S. Z. and Solomon, S. C. (2015). Chlorine on the surface of Mercury: MESSENGER gamma-ray measurements and implications for the planet’s formation and evolution. Icarus, 257, 417– 427, doi:10.1016/j.icarus.2015.04.039. Fassett, C. I., Head, J. W., Blewett, D. T., Chapman, C. R., Dickson, J. L., Murchie, S. L., Solomon, S. C. and Watters, T. R. (2009). Caloris impact basin: Exterior geomorphology, stratigraphy, morphometry, radial sculpture, and smooth plains deposits. Earth Planet. Sci. Lett., 285, 297–308, doi:10.1016/j.epsl.2009.05.022. Fassett, C. I., Kadish, S. J., Head, J. W., Solomon, S. C. and Strom, R. G. (2011). The global population of large craters on Mercury and comparison with the Moon. Geophys. Res. Lett., 38, L10202, doi:10.1029/2011gl047294. Fei, Y., Bertka, C. M. and Finger, L. W. (1997). High-pressure iron sulfur compound, Fe3S2, and melting relations in the Fe–FeS system. Science, 275, 1621–1623, doi:10.1126/science.275.5306.1621. Fei, Y., Li, J., Bertka, C. M. and Prewitt, C. T. (2000). Structure type and bulk modulus of Fe3S, a new iron-sulfur compound. Amer. Mineral., 85, 1830–1833, doi:10.2138/am-2000-11-1229. Fei, Y., Hillgren, V. J., Shahar, A. and Solomon, S. C. (2011). On the silicon content of Mercury’s core and implications for core mineralogy, structure, and density. Lunar Planet. Sci., 42, abstract 1949. Ferrari, S., Massironi, M., Marchi, S., Byrne, P. K., Klimczak, C., Martellato, E. and Cremonese, G. (2014). Age relationships of the Rembrandt basin and Enterprise Rupes, Mercury. In Volcanism and Tectonism Across the Solar System, ed. T. Platz, M. Massironi, P. K. Byrne and H. Hiesinger, Special Publication 401. London: Geological Society, pp. 159–172, doi:10.1144/ SP401.20. Fischer, R. A., Campbell, A. J., Reaman, D. M., Miller, N. A., Heinz, D. L., Dera, P. and Prakapenka, V. B. (2013). Phase relations in the Fe–FeSi system at high pressures and temperatures. Earth Planet. Sci. Lett., 373, 54–64, doi:10.1016/j.epsl.2013.04.035. Freed, A. M., Solomon, S. C., Watters, T. R., Phillips, R. J. and Zuber, M. T. (2009). Could Pantheon Fossae be the result of the

References Apollodorus crater-forming impact within the Caloris basin, Mercury? Earth Planet. Sci. Lett., 285, 320–327, doi:10.1016/j. epsl.2009.02.038. Freed, A. M., Blair, D. M., Watters, T. R., Klimczak, C., Byrne, P. K., Solomon, S. C., Zuber, M. T. and Melosh, H. J. (2012). On the origin of graben and ridges within and near volcanically buried craters and basins in Mercury’s northern plains. J. Geophys. Res., 117, E00L06, doi:10.1029/2012je004119. Glassmeier, K.-H., Auster, H.-U. and Motschmann, U. (2007). A feedback dynamo generating Mercury’s magnetic field. Geophys. Res. Lett., 34, L22201, doi:10.1029/2007gl031662. Gomi, H., Ohta, K., Hirose, K., Labrosse, S., Caracas, R., Verstraete, M. J. and Hernlund, J. W. (2013). The high conductivity of iron and thermal evolution of the Earth’s core. Phys. Earth Planet. Inter., 224, 88–103, doi:10.1016/j.pepi.2013.07.010. Grasset, O. and Parmentier, E. M. (1998). Thermal convection in a volumetrically heated, infinite Prandtl number fluid with strongly temperature-dependent viscosity: Implications for planetary thermal evolution. J. Geophys. Res., 103, 18,171–18,181, doi:10.1029/ 98JB01492. Grott, M., Breuer, D. and Laneuville, M. (2011). Thermo-chemical evolution and global contraction of Mercury. Earth Planet. Sci. Lett., 307, 135–146, doi:10.1016/j.epsl.2011.04.040. Hauck, S. A., II and Phillips, R. J. (2002). Thermal and crustal evolution of Mars. J. Geophys. Res., 107, 5052, doi:10.1029/ 2001JE001801. Hauck, S. A., II, Dombard, A. J., Phillips, R. J. and Solomon, S. C. (2004). Internal and tectonic evolution of Mercury. Earth Planet. Sci. Lett., 222, 713–728, doi:10.1016/j.epsl.2004.03.037. Hauck, S. A., II, Aurnou, J. M. and Dombard, A. J. (2006). Sulfur’s impact on core evolution and magnetic field generation on Ganymede. J. Geophys. Res., 111, E09008, doi:10.1029/2005je002557. Hauck, S. A., II, Margot, J.-L., Solomon, S. C., Phillips, R. J., Johnson, C. L., Lemoine, F. G., Mazarico, E., McCoy, T. J., Padovan, S., Peale, S. J., Perry, M. E., Smith, D. E. and Zuber, M. T. (2013). The curious case of Mercury’s internal structure. J. Geophys. Res. Planets, 118, 1204–1220, doi:10.1002/jgre.20091. Head, J. W., Murchie, S. L., Prockter, L. M., Robinson, M. S., Solomon, S. C., Strom, R. G., Chapman, C. R., Watters, T. R., McClintock, W. E., Blewett, D. T. and Gillis-Davis, J. J. (2008). Volcanism on Mercury: Evidence from the first MESSENGER Flyby. Science, 321, 69–72, doi:10.1126/science.1159256. Head, J. W., Chapman, C. R., Strom, R. G., Fassett, C. I., Denevi, B. W., Blewett, D. T., Ernst, C. M., Watters, T. R., Solomon, S. C., Murchie, S. L., Prockter, L. M., Chabot, N. L., Gillis-Davis, J. J., Whitten, J. L., Goudge, T. A., Baker, D. M. H., Hurwitz, D. M., Ostrach, L. R., Xiao, Z., Merline, W. J., Kerber, L., Dickson, J. L., Oberst, J., Byrne, P. K., Klimczak, C. and Nittler, L. R. (2011). Flood volcanism in the northern high latitudes of Mercury revealed by MESSENGER. Science, 333, 1853–1856, doi:10.1126/science.1211997. Heimpel, M. H., Aurnou, J. M., Al-Shamali, F. M. and Gomez Perez, N. (2005). A numerical study of dynamo action as a function of spherical shell geometry. Earth Planet. Sci. Lett., 236, 542–557, doi:10.1016/j.epsl.2005.04.032. Herzberg, C. T., Raterron, P. and Zhang, J. (2000). New experimental observations on the anhydrous solidus for peridotite KLB-1. Geophys. Geochem. Geosyst., 1, doi:10.1029/2000GC000089. Heyner, D., Wicht, J., Gómez-Pérez, N., Schmitt, D., Auster, H.-U. and Glassmeier, K.-H. (2011). Evidence from numerical experiments for a feedback dynamo generating Mercury’s magnetic field. Science, 334, 1690–1693, doi:10.1126/science.1207290. Hurwitz, D. M., Head, J. W., Byrne, P. K., Xiao, Z., Solomon, S. C., Zuber, M. T., Smith, D. E. and Neumann, G. A. (2013).


Investigating the origin of candidate lava channels on Mercury with MESSENGER data: Theory and observations. J. Geophys. Res. Planets, 118, 471–486, doi:10.1029/2012je004103. James, P. B., Zuber, M. T., Phillips, R. J. and Solomon, S. C. (2015). Support of long-wavelength topography on Mercury inferred from MESSENGER measurements of gravity and topography. J. Geophys. Res. Planets, 120, 287–310, doi:10.1002/2014je004713. Jing, Z., Wang, Y., Kono, Y., Yu, T., Sakamaki, T., Park, C., Rivers, M. L., Sutton, S. R. and Shen, G. (2014). Sound velocity of Fe–S liquids at high pressure: Implications for the Moon’s molten outer core. Earth Planet. Sci. Lett., 396, 78–87, doi:10.1016/j.epsl.2014.04.015. Johnson, C. L., Phillips, R. J., Purucker, M. E., Anderson, B. J., Byrne, P. K., Denevi, B. W., Feinberg, J. M., Hauck, S. A., Head, J. W., Korth, H., James, P. B., Mazarico, E., Neumann, G. A., Philpott, L. C., Siegler, M. A., Tsyganenko, N. A. and Solomon, S. C. (2015). Low-altitude magnetic field measurements by MESSENGER reveal Mercury’s ancient crustal field. Science, 348, 892–895, doi:10.1126/science.aaa8720. Johnson, C. L., Philpott, L. C., Anderson, B. J., Korth, H., Hauck, S. A., II, Heyner, D., Phillips, R. J., Winslow R. M. and Solomon S. C. (2016). MESSENGER observations of induced magnetic fields in Mercury’s core, Geophys. Res. Lett., 43, 2436–2444, doi:10.1002/ 2015GL067370. Kaula, W. M. (1975). The seven ages of a planet. Icarus, 26, 1–15, doi:10.1016/0019-1035(75)90138-4. Keil, K. (1989). Enstatite meteorites and their parent bodies. Meteoritics, 24, 195–208. Kennedy, P. J., Freed, A. M. and Solomon, S. C. (2008). Mechanisms of faulting in and around Caloris basin, Mercury. J. Geophys. Res., 113, E08004, doi:10.1029/2007JE002992. Kerber, L., Head, J. W., Solomon, S. C., Murchie, S. L., Blewett, D. T. and Wilson, L. (2009). Explosive volcanic eruptions on Mercury: Eruption conditions, magma volatile content, and implications for interior volatile abundances. Earth Planet. Sci. Lett., 285, 263– 271, doi:10.1016/j.epsl.2009.04.037. Kerber, L., Head, J. W., Blewett, D. T., Solomon, S. C., Wilson, L., Murchie, S. L., Robinson, M. S., Denevi, B. W. and Domingue, D. L. (2011). The global distribution of pyroclastic deposits on Mercury: The view from MESSENGER flybys 1–3. Planet. Space Sci., 59, 1895–1909, doi:10.1016/j.pss.2011.03.020. King, S. D. (2008). Pattern of lobate scarps on Mercury’s surface reproduced by a model of mantle convection. Nature Geosci., 1, 229–232, doi:10.1038/ngeo152. Kirk, R. L. and Stevenson, D. J. (1989). The competition between thermal contraction and differentiation in the stress history of the Moon. J. Geophys. Res., 94, 12133–12144, doi:10.1029/ JB094iB09p12133. Klima, R. L., Blewett, D. T., Denevi, B. W. Ernst, C. M., Frank, E. A., Head, J. W., III, Izenberg, N. R., Murchie, S. L., Nittler, L. R., Peplowski, P. N. and Solomon, S. C. (2016). Global distribution and spectral properties of low-reflectance material on Mercury. Lunar Planet. Sci., 47, abstract 1195. Klimczak, C. (2015). Limits on the brittle strength of planetary lithospheres undergoing global contraction. J. Geophys. Res. Planets, 120, 2135–2151, doi:10.1002/2015je004851. Klimczak, C., Watters, T. R., Ernst, C. M., Freed, A. M., Byrne, P. K., Solomon, S. C., Blair, D. M. and Head, J. W. (2012). Deformation associated with ghost craters and basins in volcanic smooth plains on Mercury: Strain analysis and implications for plains evolution. J. Geophys. Res., 117, E00L03, doi:10.1029/2012je004100. Knibbe, J. S. and van Westrenen, W. (2015). The interior configuration of planet Mercury constrained by moment of inertia and planetary contraction. J. Geophys. Res. Planets, 120, 1904–1923, doi:10.1002/2015JE004908.


Mercury’s Global Evolution

Konôpková, Z., McWilliams, R. S., Gómez-Pérez, N. and Goncharov, A. F. (2016). Direct measurement of thermal conductivity in solid iron at planetary core conditions. Nature, 534, 99–101, doi:10.1038/nature18009. Kuwayama, Y. and Hirose, K. (2004). Phase relations in the system Fe– FeSi at 21 GPa. Amer. Mineral., 89, 273–276, doi:10.2138/am2004-2-303. Le Feuvre, M. and Wieczorek, M. A. (2011). Nonuniform cratering of the Moon and a revised crater chronology of the inner Solar System. Icarus, 214, 1–20, doi:10.1016/j.icarus.2011.03.010. Li, J., Fei, Y., Mao, H. K., Hirose, K. and Shieh, S. R. (2001). Sulfur in the Earth’s inner core. Earth Planet. Sci. Lett., 193, 509–514, doi:10.1016/S0012-821X(01)00521-0. Lord, O. T., Walter, M. J., Dasgupta, R., Walker, D. and Clark, S. M. (2009). Melting in the Fe–C system to 70 GPa. Earth Planet. Sci. Lett., 284, 157–167, doi:10.1016/j.epsl.2009.04.017. Malavergne, V., Toplis, M. J., Berthet, S. and Jones, J. (2010). Highly reducing conditions during core formation on Mercury: Implications for internal structure and the origin of a magnetic field. Icarus, 206, 199–209, doi:10.1016/j.icarus.2009.09.001. Manglik, A., Wicht, J. and Christensen, U. R. (2010). A dynamo model with double diffusive convection for Mercury’s core. Earth Planet. Sci. Lett., 289, 619–628, doi:10.1016/j.epsl.2009.12.007. Marchi, S., Mottola, S., Cremonese, G., Massironi, M. and Martellato, E. (2009). A new chronology for the Moon and Mercury. Astron. J., 137, 4936–4948, doi:10.1088/0004-6256/137/6/4936. Marchi, S., Massironi, M., Cremonese, G., Martellato, E., Giacomini, L. and Prockter, L. (2011). The effects of the target material properties and layering on the crater chronology: The case of Raditladi and Rachmaninoff basins on Mercury. Planet. Space Sci., 59, 1968–1980, doi:10.1016/j.pss.2011.06.007. Marchi, S., Chapman, C. R., Fassett, C. I., Head, J. W., Bottke, W. F. and Strom, R. G. (2013). Global resurfacing of Mercury 4.0–4.1 billion years ago by heavy bombardment and volcanism. Nature, 499, 59–61, doi:10.1038/nature12280. Margot, J. L., Peale, S. J., Jurgens, R. F., Slade, M. A. and Holin, I. V. (2007). Large longitude libration of Mercury reveals a molten core. Science, 316, 710–714, doi:10.1126/science.1140514. Margot, J.-L., Peale, S. J., Solomon, S. C., Hauck, S. A., II, Ghigo, F. D., Jurgens, R. F., Padovan, S. and Campbell, D. B. (2012). Mercury’s moment of inertia from spin and gravity data. J. Geophys. Res., 117, E00L09, doi:10.1029/2012JE004161. Martin, A. M., Van Orman, J., Hauck, S. A., II, Chen, B., II, Sun, N., Moore, R. D. and Han, J. (2014). In situ determination of the eutectic melting temperature of Fe–FeS–Fe3C between 4.5 and 24.5 GPa and implications for Mercury’s core. Lunar Planet. Sci., 45, abstract 2854. Martin, A. M., Van Orman, J., Hauck, S. A., II, Sun, N., Yu, T. and Wang, Y. (2015). Role of sulfur, silicon and carbon on the crystallization processes in Mercury’s core inferred from insitu melting experiments between 4.5 and 15.5 GPa. Lunar Planet. Sci., 46, abstract 2627. Matsuyama, I. and Nimmo, F. (2009). Gravity and tectonic patterns of Mercury: Effect of tidal deformation, spin-orbit resonance, nonzero eccentricity, despinning, and reorientation. J. Geophys. Res., 114, E01010, doi:10.1029/2008je003252. Mazarico, E., Genova, A., Goossens, S., Lemoine, F. G., Neumann, G. A., Zuber, M. T., Smith, D. E. and Solomon, S. C. (2014). The gravity field, orientation, and ephemeris of Mercury from MESSENGER observations after three years in orbit. J. Geophys. Res. Planets, 119, 2417–2436, doi:10.1002/ 2014je004675. McCoy, T. J., Dickinson, T. L. and Lofgren, G. E. (1999). Partial melting of the Indarch (EH4) meteorite: A textural, chemical, and phase

relations view of melting and melt migration. Meteorit. Planet. Sci., 34, 735–746, doi:10.1111/j.1945-5100.1999.tb01386.x. McCubbin, F. M., Riner, M. A., Vander Kaaden, K. E. and Burkemper, L. K. (2012). Is Mercury a volatile-rich planet? Geophys. Res. Lett., 39, L09202, doi:10.1029/2012gl051711. Melosh, H. J. (1977). Global tectonics of a despun planet. Icarus, 31, 221–243, doi:10.1016/0019-1035(77)90035-5. Melosh, H. J. and Dzurisin, D. (1978). Mercurian global tectonics: A consequence of tidal despinning? Icarus, 35, 227–236, doi:10.1016/0019-1035(78)90007-6. Melosh, H. J. and McKinnon, W. B. (1988). The tectonics of Mercury. In Mercury, ed. F. Vilas, C. R. Chapman and M. S. Matthews. Tucson, AZ: University of Arizona Press, pp. 374–400. Michel, N. C., Hauck, S. A., II, Solomon, S. C., Phillips, R. J., Roberts, J. H. and Zuber, M. T. (2013). Thermal evolution of Mercury as constrained by MESSENGER observations. J. Geophys. Res. Planets, 118, 1033–1044, doi:10.1002/jgre.20049. Morard, G. and Katsura, T. (2010). Pressure–temperature cartography of Fe–S–Si immiscible system. Geochim. Cosmochim. Acta, 74, 3659–3667, doi:10.1016/j.gca.2010.03.025. Morard, G., Andrault, D., Guignot, N., Siebert, J., Garbarino, G. and Antonangeli, D. (2011). Melting of Fe–Ni–Si and Fe–Ni–S alloys at megabar pressures: Implications for the core–mantle boundary temperature. Phys. Chem. Minerals, 38, 767–776, doi:10.1007/ s00269-011-0449-9. Moresi, L. N. and Solomatov, V. S. (1995). Numerical investigation of 2D convection with extremely large viscosity variations. Phys. Fluids, 7, 2154–2162, doi:10.1063/1.868465. Murchie, S. L., Watters, T. R., Robinson, M. S., Head, J. W., Strom, R. G., Chapman, C. R., Solomon, S. C., McClintock, W. E., Prockter, L. M., Domingue, D. L. and Blewett, D. T. (2008). Geology of the Caloris basin, Mercury: A view from MESSENGER. Science, 321, 73–76, doi:10.1126/science.1159261. Murchie, S. L., Klima, R. L., Denevi, B. W., Ernst, C. M., Keller, M. R., Domingue, D. L., Blewett, D. T., Chabot, N. L., Hash, C. D., Malaret, E., Izenberg, N. R., Vilas, F., Nittler, L. R., Gillis-Davis, J. J., Head, J. W. and Solomon, S. C. (2015). Orbital multispectral mapping of Mercury with the MESSENGER Mercury Dual Imaging System: Evidence for the origins of plains units and low-reflectance material. Icarus, 254, 287–305, doi:10.1016/j. icarus.2015.03.027. Murray, B. C., Belton, M. J. S., Danielson, G. E., Davies, M. E., Gault, D. E., Hapke, B., O’Leary, B., Strom, R. G., Suomi, V. and Trask, N. (1974). Mercury’s surface: Preliminary description and interpretation from Mariner 10 pictures. Science, 185, 169–179. Murray, B. C., Strom, R. G., Trask, N. J. and Gault, D. E. (1975). Surface history of Mercury: Implications for terrestrial planets. J. Geophys. Res., 80, 2508–2514, doi:10.1029/JB080i017p0 2508. Namur, O. and Charlier, B. (2017). Silicate mineralogy at the surface of Mercury. Nature Geosci., 10, 9–13, doi:10.1038/ngeo2860. Namur, O., Collinet, M., Charlier, B., Grove T. L., Holtz F. and McCammon, C. (2016). Melting processes and mantle sources of lavas on Mercury. Earth Planet. Sci. Lett., 439, 117–128, doi:10.1016/j.epsl.2016.01.030. Ness, N. F. (1979). The magnetic field of Mercury. Phys. Earth Planet. Inter., 20, 209–217, doi:10.1016/0031-9201(79)90044-X. Ness, N. F., Behannon, K. W., Lepping, R. P. and Whang, Y. C. (1975). The magnetic field of Mercury. I. J. Geophys. Res., 80, 2708– 2716, doi:10.1029/JA080i019p02708. Ness, N. F., Behannon, K. W., Lepping, R. P. and Whang, Y. C. (1976). Observations of Mercury’s magnetic field. Icarus, 28, 479–488, doi:10.1016/0019-1035(76)90121-4.

References Neukum, G., Ivanov, B. A. and Hartmann, W. K. (2001). Cratering records in the inner solar system in relation to the lunar reference system. Space Sci. Rev., 96, 55–86. Nimmo, F. and Stevenson, D. J. (2001). Estimates of Martian crustal thickness from viscous relaxation of topography. J. Geophys. Res., 106, 5085–5098, doi:10.1029/2000JE001331. Nittler, L. R., Starr, R. D., Weider, S. Z., McCoy, T. J., Boynton, W. V., Ebel, D. S., Ernst, C. M., Evans, L. G., Goldsten, J. O., Hamara, D. K., Lawrence, D. J., McNutt, R. L., Jr., Schlemm, C. E., II, Solomon, S. C. and Sprague, A. L. (2011). The major-element composition of Mercury’s surface from MESSENGER X-ray spectrometry. Science, 333, 1847–1850, doi:10.1126/science.1211567. Noyelles, B., Frouard, J., Makarov, V. V. and Efroimsky, M. (2014). Spin-orbit evolution of Mercury revisited. Icarus, 241, 26–44, doi:10.1016/j.icarus.2014.05.045. Ostrach, L. R., Robinson, M. S., Whitten, J. L., Fassett, C. I., Strom, R. G., Head, J. W. and Solomon, S. C. (2015). Extent, age, and resurfacing history of the northern smooth plains on Mercury from MESSENGER observations. Icarus, 250, 602–622, doi:10.1016/j.icarus.2014.11.010. Padovan, S., Margot, J.-L., Hauck, S. A., II, Moore, W. B. and Solomon, S. C. (2014). The tides of Mercury and possible implications for its interior structure. J. Geophys. Res. Planets, 119, 850–866, doi:10.1002/2013je004459. Padovan, S., Wieczorek, M. A., Margot, J.-L., Tosi, N. and Solomon, S. C. (2015). Thickness of the crust of Mercury from geoid-totopography ratios. Geophys. Res. Lett., 42, 1029–1038, doi:10.1002/2014gl062487. Peale, S. J. (1988). The rotational dynamics of Mercury and the state of its core. In Mercury, ed. F. Vilas, C. R. Chapman and M. S. Matthews. Tucson, AZ: University of Arizona Press, pp. 461–493. Peale, S. J., Phillips, R. J., Solomon, S. C., Smith, D. E. and Zuber, M. T. (2002). A procedure for determining the nature of Mercury’s core. Meteorit. Planet. Sci., 37, 1269–1283, doi:10.1111/j.19455100.2002.tb00895.x. Peale, S. J., Margot, J.-L., Hauck, S. A., II and Solomon, S. C. (2016). Consequences of a solid inner core on Mercury’s spin configuration. Icarus, 264, 443–455, doi:10.1016/j.icarus.2015.09.024. Peplowski, P. N., Evans, L. G., Hauck, S. A., II, McCoy, T. J., Boynton, W. V., Gillis-Davis, J. J., Ebel, D. S., Goldsten, J. O., Hamara, D. K., Lawrence, D. J., McNutt, R. L., Jr., Nittler, L. R., Solomon, S. C., Rhodes, E. A., Sprague, A. L., Starr, R. D. and Stockstill-Cahill, K. R. (2011). Radioactive elements on Mercury’s surface from MESSENGER: Implications for the planet’s formation and evolution. Science, 333, 1850–1852, doi:10.1126/science.1211576. Peplowski, P. N., Lawrence, D. J., Rhodes, E. A., Sprague, A. L., McCoy, T. J., Denevi, B. W., Evans, L. G., Head, J. W., Nittler, L. R., Solomon, S. C., Stockstill-Cahill, K. R. and Weider, S. Z. (2012). Variations in the abundances of potassium and thorium on the surface of Mercury: Results from the MESSENGER GammaRay Spectrometer. J. Geophys. Res., 117, E00L04, doi:10.1029/ 2012JE004141. Peplowski, P. N., Lawrence, D. J., Evans, L. G., Klima, R. L., Blewett, D. T., Goldsten, J. O., Murchie, S. L., McCoy, T. J., Nittler, L. R., Solomon, S. C., Starr, R. D. and Weider, S. Z. (2015a). Constraints on the abundance of carbon in near-surface materials on Mercury: Results from the MESSENGER Gamma-Ray Spectrometer. Planet. Space Sci., 108, 98–107, doi:10.1016/j.pss.2015.01.008. Peplowski, P. N., Lawrence, D. J., Feldman, W. C., Goldsten, J. O., Bazell, D., Evans, L. G., Head, J. W., Nittler, L. R., Solomon, S. C. and Weider, S. Z. (2015b). Geochemical terranes of Mercury’s northern hemisphere as revealed by MESSENGER neutron measurements. Icarus, 253, 346–363, doi:10.1016/j.icarus. 2015.02.002.


Peplowski, P. N., Klima, R. L., Lawrence, D. J., Ernst, C. M., Denevi, B. W., Frank, E. A., Goldsten, J. O., Murchie, S. L., Nittler, L. R. and Solomon, S. C. (2016), Remote sensing evidence for an ancient carbon-bearing crust on Mercury, Nature Geosci., 9, 273–276, doi:10.1038/ngeo2669. Perry, M. E., Neumann, G. A., Phillips, R. J., Barnouin, O. S., Ernst, C. M., Kahan, D. S., Solomon, S. C., Zuber, M. T., Smith, D. E., Hauck, S. A., II, Peale, S. J., Margot, J.-L., Mazarico, E., Johnson, C. L., Gaskell, R. W., Roberts, J. H., McNutt, R. L., Jr. and Oberst, J. (2015). The low-degree shape of Mercury. Geophys. Res. Lett., 42, 6951–6958, doi:10.1002/2015gl065101. Philpott, L. C., Johnson, C. L., Winslow, R. M., Anderson, B. J., Korth, H., Purucker, M. E. and Solomon, S. C. (2014). Constraints on the secular variation of Mercury’s magnetic field from the combined analysis of MESSENGER and Mariner 10 data. Geophys. Res. Lett., 41, 6627–6634, doi:10.1002/2014gl061401. Plesa, A. C., Tosi, N. and Hüttig, C. (2013). Thermo-chemical convection in planetary mantles: Advection methods and magma ocean overturn simulations. In Integrated Information and Computing Systems for Natural, Spatial, and Social Sciences, ed. R. ClausPeter. Hershey, PA: IGI Global, pp. 302–323, doi:10.4018/978-14666-2190-9.ch015. Pozzo, M., Davies, C., Gubbins, D. and Alfe, D. (2012). Thermal and electrical conductivity of iron at Earth’s core conditions. Nature, 485, 355–358, doi:10.1038/nature11031. Prockter, L. M., Ernst, C. M., Denevi, B. W., Chapman, C. R., Head, J. W., Fassett, C. I., Merline, W. J., Solomon, S. C., Watters, T. R., Strom, R. G., Cremonese, G., Marchi, S. and Massironi, M. (2010). Evidence for young volcanism on Mercury from the third MESSENGER flyby. Science, 329, 668–671, doi:10.1126/ science.1188186. Raghavan, V. (1988). Phase Diagrams of Ternary Iron Alloys, Part 2: Ternary Systems Containing Iron and Sulphur. Calcutta: Indian Institute of Metals. Redmond, H. L. and King, S. D. (2007). Does mantle convection currently exist on Mercury? Phys. Earth Planet. Inter., 164, 221–231, doi:10.1016/j.pepi.2007.07.004. Reese, C. C., Solomatov, V. S. and Moresi, L. N. (1998). Heat transport efficiency for stagnant lid convection with dislocation viscosity: Application to Venus and Mars. J. Geophys. Res., 103, 13643– 13658, doi:10.1029/98JE01047. Riner, M. A., Lucey, P. G., Desch, S. J. and McCubbin, F. M. (2009). Nature of opaque components on Mercury: Insights into a Mercurian magma ocean. Geophys. Res. Lett., 36, L02201, doi:10.1029/2008GL036128. Rivera-Valentin, E. G. and Barr, A. C. (2014). Impact-induced compositional variations on Mercury. Earth Planet. Sci. Lett., 391, 234– 242, doi:10.1016/j.epsl.2014.02.003. Rivoldini, A. and Van Hoolst, T. (2013). The interior structure of Mercury constrained by the low-degree gravity field and the rotation of Mercury. Earth Planet. Sci. Lett., 377–378, 62–72, doi:10.1016/j.epsl.2013.07.021. Roberts, J. H. and Barnouin, O. S. (2012). The effect of the Caloris impact on the mantle dynamics and volcanism of Mercury. J. Geophys. Res., 117, E02007, doi:10.1029/2011JE003876. Robinson, M. S. and Lucey, P. G. (1997). Recalibrated Mariner 10 color mosaics: Implications for Mercurian volcanism. Science, 275, 197–200, doi:10.1126/science.275.5297.197. Robinson, M. S., Murchie, S. L., Blewett, D. T., Domingue, D. L., Hawkins, S. E., Head, J. W., Holsclaw, G. M., McClintock, W. E., McCoy, T. J., McNutt, R. L., Prockter, L. M., Solomon, S. C. and Watters, T. R. (2008). Reflectance and color variations on Mercury: Regolith processes and compositional heterogeneity. Science, 321, 66–69, doi:10.1126/science.1160080.


Mercury’s Global Evolution

Rückriemen, T., Breuer, D. and Spohn, T. (2015). The Fe snow regime in Ganymede’s core: A deep-seated dynamo below a stable snow zone. J. Geophys. Res. Planets, 120, 1095–1118, doi:10.1002/ 2014JE004781. Schubert, G., Ross, M. N., Stevenson, D. J. and Spohn, T. (1988). Mercury’s thermal history and the generation of its magnetic field. In Mercury, ed. F. Vilas, C. R. Chapman and M. S. Matthews. Tucson, AZ: University of Arizona Press, pp. 429–460. Schumacher, S. and Breuer, D. (2006). Influence of a variable thermal conductivity on the thermochemical evolution of Mars. J. Geophys. Res., 111, E02006, doi:10.1029/2005JE002429. Seagle, C. T., Cottrell, E., Fei, Y., Hummer, D. R. and Prakapenka, V. B. (2013). Electrical and thermal transport properties of iron and iron–silicon alloy at high pressure. Geophys. Res. Lett., 40, 5377– 5381, doi:10.1002/2013gl057930. Siegfried, R. W., II and Solomon, S. C. (1974). Mercury: Internal structure and thermal evolution. Icarus, 23, 192–205, doi:10.1016/0019-1035(74)90005-0. Smith, D. E., Zuber, M. T., Lemoine, F. G., Solomon, S. C., Hauck, S. A., II, Lemoine, F. G., Mazarico, E., Phillips, R. J., Neumann, G. A., Peale, S. J., Margot, J.-L., Johnson, C. L., Torrence, M. H., Perry, M. E., Rowlands, D. D., Goossens, S., Head, J. W. and Taylor, A. H. (2012). Gravity field and internal structure of Mercury from MESSENGER. Science, 336, 214–217, doi:10.1126/science.1218809. Solomon, S. C. (1976). Some aspects of core formation in Mercury. Icarus, 28, 509–521, doi:10.1016/0019-1035(76)90124-X. Solomon, S. C. (1977). The relationship between crustal tectonics and internal evolution in the Moon and Mercury. Phys. Earth Planet. Inter., 15, 135–145, doi:10.1016/0031-9201(77)90026-7. Solomon, S. C. (2003). Mercury: The enigmatic innermost planet. Earth Planet. Sci. Lett., 216, 441–455, doi:10.1016/s0012-821x (03)00546-6. Spudis, P. D. and Guest, J. E. (1988). Stratigraphy and geologic history of Mercury. In Mercury, ed. F. Vilas, C. R. Chapman and M. S. Matthews. Tucson, AZ: University of Arizona Press, pp. 118–164. Stanley, S. and Bloxham, J. (2016). On the secular variation of Saturn’s magnetic field. Phys. Earth Planet. Inter., 250, 31–34, doi:10.1016/j.pepi.2015.11.002. Stanley, S., Bloxham, J., Hutchison, W. and Zuber, M. (2005). Thin shell dynamo models consistent with Mercury’s weak observed magnetic field. Earth Planet. Sci. Lett., 234, 27–38, doi:10.1016/j. epsl.2005.02.040. Stevenson, D. J., Spohn, T. and Schubert, G. (1983). Magnetism and thermal evolution of the terrestrial planets. Icarus, 54, 466–489, doi:10.1016/0019-1035(83)90241-5. Stewart, A. J., Schmidt, M. W., van Westrenen, W. and Liebske, C. (2007). Mars: A new core-crystallization regime. Science, 316, 1323–1325, doi:10.1126/science.1140549. Stockstill-Cahill, K. R., McCoy, T. J., Nittler, L. R., Weider, S. Z. and Hauck, S. A., II (2012). Magnesium-rich crustal compositions on Mercury: Implications for magmatism from petrologic modeling. J. Geophys. Res., 117, E00L15, doi:10.1029/2012JE004140. Strom, R. G. (1977). Origin and relative age of lunar and Mercurian intercrater plains. Phys. Earth Planet. Inter., 15, 156–172, doi:10.1016/0031-9201(77)90028-0. Strom, R. G. and Neukum, G. (1988). The cratering record on Mercury and the origin of impacting objects. In Mercury, ed. F. Vilas, C. R. Chapman and M. S. Matthews. Tucson, AZ: University of Arizona Press, pp. 336–373. Strom, R. G., Trask, N. J. and Guest, J. E. (1975). Tectonism and volcanism on Mercury. J. Geophys. Res., 80, 2478–2507, doi:10.1029/JB080i017p02478. Strom, R. G., Chapman, C. R., Merline, W. J., Solomon, S. C. and Head, J. W. (2008). Mercury cratering record viewed from

MESSENGER’s first flyby. Science, 321, 79–81, doi:10.1126/ science.1159317. Strom, R. G., Banks, M. E., Chapman, C. R., Fassett, C. I., Forde, J. A., Head, J. W., III, Merline, W. J., Prockter, L. M. and Solomon, S. C. (2011). Mercury crater statistics from MESSENGER flybys: Implications for stratigraphy and resurfacing history. Planet. Space Sci., 59, 1960–1967, doi:10.1016/j.pss.2011.03.018. Taylor, G. J. and Scott, E. R. D. (2003). Mercury. In Meteorites, Comets and Planets, ed. A. M. Davis, Treatise on Geochemistry, Vol. 1, ed. H. D. Holland and K. K. Turekian. Oxford: Pergamon, pp. 477–485, doi:10.1016/B0-08-043751-6/01071-9. Thomas, R. J., Rothery, D. A., Conway, S. J. and Anand, M. (2014). Mechanisms of explosive volcanism on Mercury: Implications from its global distribution and morphology. J. Geophys. Res. Planets, 119, 2239–2254, doi:10.1002/2014je004692. Tian, Z., Zuber, M. T. and Stanley, S. (2015). Magnetic field modeling for Mercury using dynamo models with a stable layer and laterally variable heat flux. Icarus, 260, 263–268, doi:10.1016/j. icarus.2015.07.019. Tosi, N., Grott, M., Plesa, A. C. and Breuer, D. (2013). Thermochemical evolution of Mercury’s interior. J. Geophys. Res. Planets, 118, 2474–2487, doi:10.1002/jgre.20168. Tosi, N., Čadek, O., Běhounková, M., Káňová, M., Plesa, A. C., Grott, M., Breuer, D., Padovan, S. and Wieczorek, M. A. (2015). Mercury’s low-degree geoid and topography controlled by insolation-driven elastic deformation. Geophys. Res. Lett., 42, 7327– 7335, doi:10.1002/2015gl065314. Trask, N. J. and Guest, J. E. (1975). Preliminary geologic terrain map of Mercury. J. Geophys. Res., 80, 2461–2477, doi:10.1029/ JB080i017p02461. Vander Kaaden, K. E. and McCubbin, F. M. (2015). Exotic crust formation on Mercury: Consequences of a shallow, FeO-poor mantle. J. Geophys. Res. Planets, 120, 195–209, doi:10.1002/ 2014je004733. Vander Kaaden, K. E. and McCubbin, F. M. (2016). The origin of boninites on Mercury: An experimental study of the northern volcanic plains lavas. Geochim. Cosmochim. Acta, 173, 246– 263, doi:10.1016/j.gca.2015.10.016. Vander Kaaden, K. E., McCubbin, F. M., Nittler, L. R., Peplowski, P. N., Weider, S. Z., Frank, E. A. and McCoy, T. J. (2017). Geochemistry, mineralogy, and petrology of boninitic and komatiitic rocks on the mercurian surface: Insights into the mercurian mantle. Icarus, 285, 155–168, doi:10.1016/j.icarus.2016.11.041. Vasavada, A. R., Paige, D. A. and Wood, S. E. (1999). Near-surface temperatures on Mercury and the Moon and the stability of polar ice deposits. Icarus, 141, 179–193, doi:10.1006/icar.1999.6175. Verma, A. and J.-L. Margot (2016). Mercury’s gravity, tides, and spin from MESSENGER radio science data, J. Geophys. Res. Planets, 121, 1627–1640, doi:10.1002/2016JE005037. Vilim, R., Stanley, S. and Hauck, S. A., II (2010). Iron snow zones as a mechanism for generating Mercury’s weak observed magnetic field. J. Geophys. Res., 115, E11003, doi:10.1029/2009JE003528. Watters, T. R. and Nimmo, F. (2010). Tectonism on Mercury. In Planetary Tectonics, ed. T. R. Watters and R. A. Schultz. Cambridge: Cambridge University Press, pp. 15–80. Watters, T. R., Robinson, M. S. and Cook, A. C. (1998). Topography of lobate scarps on Mercury: New constraints on the planet’s contraction. Geology, 26, 991–994. Watters, T. R., Solomon, S. C., Robinson, M. S., Head, J. W., André, S. L., Hauck, S. A., II and Murchie, S. L. (2009). The tectonics of Mercury: The view after MESSENGER’s first flyby. Earth Planet. Sci. Lett., 285, 283–296, doi:10.1016/j.epsl.2009.01.025. Watters, T. R., Selvans, M. M., Banks, M. E., Hauck, S. A., II, Becker, K. J. and Robinson, M. S. (2015a). Distribution of large-scale

References contractional tectonic landforms on Mercury: Implications for the origin of global stresses. Geophys. Res. Lett., 42, 3755–3763, doi:10.1002/2015gl063570. Watters, T. R., Solomon, S. C., Daud, K., Banks, M. E., Selvans, M. M., Robinson, M. S., Murchie, S. L., Chabot, N. L., Denevi, B. W., Ernst, C. M., Chapman, C. R., Fassett, C. I., Klimczak, C., Byrne, P. K. and Blewett, D. T. (2015b). Small thrust fault scarps on Mercury revealed in low-alitude MESSENGER images. Lunar Planet. Sci., 46, abstract 2240. Weider, S. Z., Nittler, L. R., Starr, R. D., McCoy, T. J., StockstillCahill, K. R., Byrne, P. K., Denevi, B. W., Head, J. W. and Solomon, S. C. (2012). Chemical heterogeneity on Mercury’s surface revealed by the MESSENGER X-Ray Spectrometer. J. Geophys. Res., 117, E00L05, doi:10.1029/2012je004153. Weider, S. Z., Nittler, L. R., Starr, R. D., Crapster-Pregont, E. J., Peplowski, P. N., Denevi, B. W., Head, J. W., Byrne, P. K., Hauck, S. A., II, Ebel, D. S. and Solomon, S. C. (2015). Evidence for geochemical terranes on Mercury: Global mapping of major elements with MESSENGER’s X-Ray Spectrometer. Earth Planet. Sci. Lett., 416, 109–120, doi:10.1016/j.epsl.2015.01.023. Whitten, J. L., Head, J. W., Denevi, B. W. and Solomon, S. C. (2014). Intercrater plains on Mercury: Insights into unit definition, characterization, and origin from MESSENGER datasets. Icarus, 241, 97–113, doi:10.1016/j.icarus.2014.06.013.


Wicht, J. and Heyner, D. (2014). Mercury’s magnetic field in the MESSENGER era. In Planetary Geodesy and Remote Sensing, ed. S. Jin. New York: CRC Press, pp. 223–262. Wieczorek, M. A., Neumann, G. A., Nimmo, F., Kiefer, W. S., Taylor, G. J., Melosh, H. J., Phillips, R. J., Solomon, S. C., AndrewsHanna, J. C., Asmar, S. W., Konopliv, A. S., Lemoine, F. G., Smith, D. E., Watkins, M. M., Williams, J. G. and Zuber, M. T. (2013). The crust of the Moon as seen by GRAIL. Science, 339, 671–675, doi:10.1126/science.1231530. Williams, J.-P., Ruiz, J., Rosenburg, M. A., Aharonson, O. and Phillips, R. J. (2011). Insolation driven variations of Mercury’s lithospheric strength. J. Geophys. Res., 116, E01008, doi:10.1029/ 2010JE003655. Williams, Q. (2009). Bottom-up versus top-down solidification of the cores of small solar system bodies: Constraints on paradoxical cores. Earth Planet. Sci. Lett., 284, 564–569, doi:10.1016/j. epsl.2009.05.019. Zhang, P., Cohen, R. E. and Haule, K. (2015). Effects of electron correlations on transport properties of iron at Earth’s core conditions. Nature, 517, 605–607, doi:10.1038/nature14090. Zolotov, M. Yu., Sprague, A. L., Hauck, S. A., II, Nittler, L. R., Solomon, S. C. and Weider, S. Z. (2013). The redox state, FeO content, and origin of sulfur-rich magmas on Mercury. J. Geophys. Res. Planets, 118, 138–146, doi:10.1029/2012je004274.


2 0 . 1 I N T R O D U C T IO N

risk of irrevocably destroying it by the introduction of terrestrial organisms is negligibly small. These examples do not by any means exhaust the list of specific priorities; nevertheless the subject as a whole is still largely open.

Mercury has now been explored by two spacecraft, Mariner 10, which flew by the planet three times in 1974–1975, and MErcury Surface, Space ENvironment, GEochemistry, and Ranging (MESSENGER), which in 2015 completed four years of orbital observations, the results of which are discussed in detail throughout this book. In this chapter we consider Mercury exploration going forward in the broad context of the history of Mercury exploration and its link to scientific policy and progress in space technology. The MESSENGER scientific and technical achievements are briefly reviewed in light of the consensus science policies for Mercury to set the stage for an overview of the next mission to Mercury, the dual-spacecraft BepiColombo mission, developed jointly by the European Space Agency (ESA) and the Japan Aerospace Exploration Agency (JAXA) and scheduled for launch in October 2018. The prospects for more ambitious missions to land on the surface or even return samples from the planet are then discussed, highlighting the key science questions that only such missions could address and considering their feasibility given the broad trends of spaceflight technology.

By the time of the next comprehensive review of the future of space research by the Space Science Board (1966), Mars had become the primary scientific priority in planetary research, but a priority scheme for all of the planets was also laid out. Mercury was ranked sixth in importance, just after “comets and asteroids” and before Pluto (seventh) and dust (eighth and last). Advances in our scientific knowledge of Mercury were viewed as following from future Earth-based observations, a “close fly-by mission,” a lander, and a geodetic satellite: Mercury is an object unique in the solar system, and should be included in a comprehensive program of planetary exploration. On the one hand, it can be classed as a terrestrial planet; on the other, it can be classed with the major satellites such as the Moon, Io, Triton, etc. From either point of view, comparative study of surfaces, atmospheres, and interiors is relevant to the more general problem of the evolution and origin of the planets and the solar system as a whole. It is worth noting that it was at this time that the Space Science Board advocated the use of large launch vehicles, the Saturn IB and Saturn V, both being developed to support the Apollo lunar landing program, to support large scientific payloads to the surface of Mars. Indeed, “In light of the excellent progress on Saturn V, we recommend that the Office of Space Science and Applications (OSSA) and the Office of Manned Space Flight (OMSF) jointly undertake a study of the early use of Saturn V for exploration of the planets with special emphasis on a Martian capsule landing in the early 1970’s” (Space Science Board, 1966). The contemplated large surface laboratory mission to Mars was part of the first “Voyager” missions1 (Cortright, 1968), but this mission as well as potentially larger missions were subsequently ruled out due to cost realities (Ezell and Ezell, 1984). The proposal of a biological laboratory to the martian surface was down-scoped and eventually flown on the less capable Titan Centaur in the form of the independent launches of Viking 1 and 2 landers and orbiters.

2 0. 2 M E RC UR Y AS A PLAN ETA RY EXPLOR ATION T ARG ET We begin by considering the history of US science policy with respect to Mercury exploration. The initial assessments of priorities for solar system exploration were made in the early 1960s (e.g., Space Science Board, 1962). Lunar and planetary priorities were focused on the Moon and the needs of the manned Apollo Moon-landing program. With respect to planetary targets, Mars had already been called out as a priority, because of its perceived potential for exobiological activity and for its possibility for eventual human exploration. Mercury was not mentioned explicitly in the report of the Space Science Board (1962) of the National Research Council: Lunar specialists emphasized that certain kinds of scientific data about the Moon must be obtained relatively early, not because they are of greater scientific importance but because they are required for the proper execution of the Apollo mission. A comprehensive recommendation cutting across all other interests is that, in the early exploration of Mars, biological and biochemical studies must have the right-ofway until we find either that there is no life on Mars or that the



Although initially billed as “a program of planetary exploration to be carried out with automated spacecraft during the 1970’s,” with “Mars and Venus will be the primary objects for investigation during that time period,” the only real mission planed was a dual launch with a Saturn V, and with a projected cost of $2.2 billion in April 1967, the program was deleted from the 1968 budget.

20.2 Mercury as a Planetary Exploration Target In 1968 the Space Science Board established a special panel to examine the comprehensive scientific exploration of the solar system. Of the principal recommendations (Space Science Board, 1968), item 3 (d) was “We accord next priorities (in descending order) to a Mariner-class Venus-Mercury fly-by in 1973 or 1975, a multiple dropsonde mission to Venus in 1975, and a major lander on Mars, perhaps in 1975 (page 6).” (In this context, “dropsonde” denoted a battery-powered payload that would return data during its descent through the venusian atmosphere and possibly from the surface.) That Mercury was locked into a 3:2 spin–orbit resonance rather than synchronous rotation had been demonstrated only three years earlier by Earth-based radar observations. It had also only just been realized that a gravity assist at Venus could enable a mission to the innermost planet with a far less powerful launch vehicle than would be required for a spacecraft capable of deceleration at Mercury after a direct trajectory from Earth (“Exploration of Mercury would not otherwise be possible without employing a very much larger booster.”) The study noted that such a mission could be carried out in 1973 or 1975, and that the next opportunity would not be until the following decade. The presence of an atmosphere on Mercury was considered doubtful but unknown. Earth-orbital telescopic observations were also considered problematic because of thermal constraints driven by Mercury’s small elongation. (A footnote says with respect to high-resolution telescopic observations: “The planet Mercury is excluded because of the difficulties in arriving at a thermal design for an Earth-orbital telescope pointing within 20° of the Sun. These difficulties are unlikely to be solved in the next few years.” Indeed, this is still the case as Mercury is excluded from Hubble Space Telescope observations because of its proximity to the Sun.) The report therefore rated atmospheric study of Mercury rather low: “Atmospheric investigations of Mercury should not command a high priority in the near future. Fly-by missions to the planet may take place for other reasons, however. If so, radiooccultation and flourescence [sic] measurements would be of interest.” The 1968 report had indeed laid some interesting groundwork for future observations (“In the absence of other information we may take this figure of 1 km as the upper limit of resolution needed to detect past volcanic activity on Mercury or Mars. Other processes require even better resolution.” “It is also important to attempt at least an exploratory, preliminary examination of all the terrestrial planets. In practical terms, this means that we place considerable value upon a photographic reconnaissance of Mercury and radar examination of Venus.”). The summary of knowledge was given as follows: Mercury is the smallest and most dense of the terrestrial planets. It is also the closest to the Sun. Albedo and radar reflection suggest a surface resembling the lunar maria. Only indistinct surface markings are visible from Earth. The great range in temperature between subsolar and midnight positions, the large solar radiation flux, and the probable lack of an atmosphere must all influence the nature of the surface in a major way. Yet the remoteness of Mercury from


Earth has served to lessen its attractiveness to those interested in planetary exploration. Recently, however, an upsurge in interest has been sparked by the realization that in 1973 or 1975 a spacecraft launched by a modest-sized vehicle can take advantage of the Venus gravitational field to gain acceleration and thereby enter a trajectory that will send it past Mercury. With respect to a “Venus-Mercury flyby” mission, the 1968 report continued: The Venus swing-by mission proposed for 1973 provides the first opportunity to examine Mercury. As a prime objective, we recommend photographing the planet with a resolution of about 2 km. Similar photography of Venus during the fly-by portion of the orbit may reveal cloud patterns indicative of the atmospheric circulation system. Additional experiments for the Mercury encounter should include a magnetometer to determine if Mercury has a magnetic field and some form of emission line photometer to determine if Mercury has an atmosphere. If the trajectory permits occultation, a second test of the existence of an atmosphere can be achieved from the S-band radio links. If imagery of Mercury at 2-km resolution is indeed obtainable from a Pioneer class Venus-Mercury fly-by, it becomes a most significant experiment from the viewpoint of planetary surfaces. The Sun may have a profound effect on its nearest neighbor so that an unusual balance of internal and external activity is evidenced by surface topography. The best way to underscore the scientific value of a limited imagery mission for Mercury is to recall the major changes in our thinking about Mars produced by the 4-km resolution Mariner 4 pictures. The short lead time, the low cost of a Pioneer mission, and the value to our national prestige of a planetary first are strong arguments supporting the basic scientific value of the mission. Further, the 1968 perspective already included the notions of orbiters and landers: From the point of view of planetary dynamics, Mercury is perhaps the most important object in the solar system. Being closest to the Sun, it is the most sensitive detector of departures from the laws proposed to account for planetary orbital motions. Its spin is also unusual, being coupled to its orbital motion in a three-halves resonance state. In view of its unusually high density, the interior of Mercury is also of special interest. A space-probe fly-by of Mercury could provide important information on both its dynamics and its interior. From photographs we may obtain the precise orientation of Mercury. Combined with similar pictures from later fly-bys, the vital knowledge of the direction of Mercury’s spin axis and the fractional difference in its equatorial moments of inertia can be determined. Search for magnetic field strengths and determination of the electromagnetic radiation from Mercury’s surface, as well as photographs of the surface, will provide important data on its interior structure. The radius, mass, and hence, density, and the orbit can also be refined from the fly-by data. An orbiter is required to


Future Missions: Mercury after MESSENGER

determine the detailed gravitational field of Mercury and a lander to study the interior by monitoring seismic activity. From this early vantage point, perhaps the most interesting comment with respect to Mercury is actually aimed at the future exploration of Mars: With respect to its interior and dynamical properties, Mars is of no greater intrinsic interest than Venus or Mercury, perhaps less. However, most of the appropriate measurements are much more easily made for Mars; we address ourselves to the determination of the dynamics and interior of Mars with special cognizance of this fact and of the unique biological and geological interest in the planet. The science questions raised and means of addressing them, notably by orbiters and landers all were – and remain – relevant to Mercury. The significant difference between obtaining new scientific knowledge at Mercury and Mars from the perspective of the late 1960s was, of course, the difference in difficulty to reach these different planets. The 1968 study cemented the ground for the subsequent Mariner 10 mission to Mercury, making use of a Venus gravity assist to enable the mission on an available launch vehicle (the Atlas Centaur) and using resonant flybys, following the first Mercury encounter on 29 March 1974, to make two more visits to the planet (21 September 1974 and 16 March 1975). Owing to the novel Venus flyby trajectory and motivated by interest in conducting reconnaissance in the near term, the Mariner Venus–Mercury mission was developed and renamed Mariner 10 following its successful launch. The observations of Mercury during the three Mariner 10 flybys provided tantalizing clues that greatly increased interest in the planet. The Mariner 10 mission was extremely successful (Balogh et al., 2007a), but it was also clear that substantial further progress in the scientific exploration of Mercury required an orbiter. Orbiting Mercury, however, was recognized as far more difficult than “simple” flybys because of the need to insert a spacecraft into orbit in Mercury’s relatively weak gravity field deep within that of the Sun, while also dealing with the harsh thermal environment close to the Sun and near Mercury’s hot dayside surface (Balogh et al., 2007b). The next study assessing solar system exploration priorities was a NASA internal study conducted by a Terrestrial Bodies Science Working Group chartered in 1976 (Toksöz et al., 1977). This study advocated a Mercury polar orbiter “with a subsatellite and/or surface lander” having a launch in (calendar year) 1986 and operations in 1988 (the same year as a “Mars Surface Sample Return (MSSR) with mobility”). Low-altitude (≤500 km) circular orbits of the planet enabled by “low-thrust propulsion system such as solar sail or ion drive to meet the science objectives” were advocated. The proposed science was similar to that recommended a year later, but the NASA study also provided details for scientific instruments, which perhaps not surprisingly targeted observational objectives very similar to those achieved by the MESSENGER payload (Solomon et al., 2007; Chapter 1) and closely aligned with the instrumentation to be flown on BepiColombo (discussed below). The most comprehensive study articulating NASA’s overall strategy for solar system exploration in the late 1970s was that

of the Committee on Planetary and Lunar Exploration, or COMPLEX (1978). The basic scheme advocated by COMPLEX was to follow flyby reconnaissance with exploration by orbiters and entry probes (for planets with atmospheres) and then with intensive study, e.g., by soft landers conducting in situ investigations, and this strategy has held well. As technologies for remote automation and mobility have advanced, the use of surface rovers has been favored, when practical, over stationary landers for many targets. In the COMPLEX report, low-thrust solar electric propulsion (SEP) was advocated for providing the means to place a spacecraft into orbit around Mercury: Advances in the exploration of Mercury, however, are predicated on observations and measurements from a circular orbit of the planet, which the present U.S. launch capability cannot currently provide. Steps should be made to prepare for the investigation of Mercury after definition of an adequate propulsion capability and in advance of availability of the system. Sample return in general was advocated as an extremely important part of the intensive study portfolio of approaches. COMPLEX did not discuss such a mission for Mercury, no doubt due to the difficulties at that time just of achieving an orbital mission. “In general, sample return from any extraterrestrial body cannot be considered as a viable mission option ab initio.” Careful deliberations to establish the scientific need and context were required to establish the specific targets. For the period under consideration, primary objectives for Mercury included determining the chemical composition of the surface; ascertaining the structure and state of the interior; and extending coverage and resolution of orbital imaging. All of these objectives were viewed as achievable with an orbiter. Exploration of the magnetosphere and magnetic field, measuring the planetary heat flow, and providing global gravity and topography of the planet were all relegated to “secondary planetary objectives.” The committee cautioned that in the absence of new propulsion technology “undertaking the next investigations of Mercury must remain indeterminate” and that, even with the advent of such new propulsive means, the exploration of Mercury should “not inhibit or detrimentally affect the primary emphasis on the triad Earth-Mars-Venus” [emphasis in original].

2 0 . 3 B R E A K T H R O U G H F O R ME R C U R Y O R B I T A L MI S S I O N S In the absence of in-space propulsion breakthroughs, the problem of placing a spacecraft in orbit about Mercury remained a major impediment to efforts to motivate an orbital mission to the planet. Direct trajectories from Earth required too large an excess velocity at Mercury approach for chemical propulsion systems to compensate. The approach speed at Mercury would be ~9 km/s, and a braking burn that imparted a velocity change (ΔV) of at least ~7.5 km/s would be needed to enter orbit. In the face of this technical obstacle, investigation of Mercury from orbit was simply not feasible.

20.3 Breakthrough for Mercury Orbital Missions


20.3.1 Multiple Gravity-Assist Solution

20.3.2 A Focused Mercury Orbital Mission: MESSENGER

It was the 1985 discovery by Chen-wan Yen of multiple Venus and Mercury gravity assist combinations that, at the expense of longer mission flight times, enabled Mercury approach speeds within the capabilities of chemical propulsion (Yen, 1985, 1989; Stern and Vilas, 1988; Belcher et al., 1991; Balogh et al., 2007b). From that time on, all serious proposals for orbital missions to Mercury relied at some level on multiple gravity assists at Venus and Mercury to enable orbit insertion. This was the case for multiple proposals in both Europe and the United States, including several mission proposals to the NASA Discovery Program and the MESSENGER mission as implemented. The ESA–JAXA BepiColombo mission is no exception. With its diverse payload and dual-spacecraft implementation, it uses both planetary gravity assists and solar electric propulsion to reach Mercury orbit with a large payload mass. Following the discovery of this multiple planetary gravity assist technique to enable a Mercury orbiter, NASA chartered a Mercury Orbiter Science Working Team (MeO SWT) to consider an orbital mission to Mercury (Belcher et al., 1991). The approach envisioned two identical spacecraft with 11 instruments (Table 20.1), each focused on a combination of magnetospheric and planetary science objectives and employing a single Titan IV–Centaur launch. The “Dual Orbiter” concept of the MeO SWT represented a nearly optimal mission concept for Mercury exploration, but the projected high cost relative to NASA’s established priority of Mercury relative to Mars and Venus (COMPLEX, 1978) precluded its implementation. Even so, the MeO SWT concept provided the justification for Mercury mission proposals to the NASA Discovery Program in the 1990s (Nelson et al., 1995; McNutt et al., 2006).

On 29 January 1996, Charles Elachi, Director of the Jet Propulsion Laboratory (JPL), issued a “Dear Colleague” letter to assemble a roadmap development team to prepare input for a “Mission to the Solar System” roadmap. A community meeting at Caltech on 5–6 March 1996 began the process that led to the final Roadmap report (Gulkis et al., 1998). The roadmap included Mercury Orbiter and Mercury Magnetospheric MultiSatellites missions. At the same time, a workshop was set up to begin deliberations on a roadmap focused on NASA’s Office of Space Science (OSS) Solar Connections theme, with its meeting scheduled at the Johns Hopkins University Applied Physics Laboratory (APL) on 10–12 April 1996. By 29 March 1996, the program for the workshop had been set to begin with four talks on potential future missions to Mercury and on Mercury magnetospheric science on the first day (10 April 1996). In March 1996 discussions and initial analysis at APL focused on how the eight key science questions for a Mercury orbiter mission suggested some years earlier (Stern and Vilas, 1988) might be addressed within the Discovery Program constraints to identify a feasible suite of payload instruments and corresponding measurement requirements. Preliminary analysis was made of an “Orbiting Mercury Observatory” (OMO), and on 2 April 1996 Sean Solomon, then the Director of the Department of Terrestrial Magnetism of the Carnegie Institution of Washington, agreed to lead the effort. The need for substantial propulsion to achieve the required ΔV to enter orbit about Mercury, together with a “focused” (i.e., low-mass) payload to allow for as high a mass ratio of fuel to payload as possible, were the immediate concerns of the core team of scientists, engineers, and managers assembled to develop the mission concept. Initial estimates were made of component masses. Combinations of gravity assist, chemical propulsion, and SEP were all considered, along with assessments of the technological readiness of the various approaches. Mission tour, science questions, payload, and instrument requirements were all discussed. An initial 11-instrument payload, with a “floor” of six instruments, was identified, and an initial science team was assembled. The mission acronym “MESSENGER” was adopted in May 1996. Use of an SEP system was ruled out because of the lack of technological maturity, and the use of an all-chemical system with “staging” at Venus was considered as well in July 1996 but was rejected as not viable. The review cycle of the initial MESSENGER proposal yielded critical guidance for work on key technology issues, chief among which were the development and qualification of a solar array design that could withstand Mercury’s thermal environment. The early decisions on science and basic approach held as the mission analyses continued through a second, and successful, proposal round in 1998–1999 (McNutt et al., 2006). The MESSENGER mission, spacecraft, and payload have been described in some detail, both in their initially accepted configurations (Gold et al., 2001; Solomon et al., 2001; Santo et al., 2001) and in the configuration as built and flown (Anderson et al., 2007; Andrews et al., 2007; Cavanaugh et al., 2007; Goldsten et al., 2007; Hawkins et al., 2007; Leary et al., 2007; McAdams et al., 2007; McClintock and Lankton, 2007; Schlemm et al., 2007; Solomon et al., 2007; Srinivasan et al., 2007) as

Table 20.1. Notional instruments for each of the MeO SWT Dual Orbiters (Belcher et al., 1991). Item


1 2 3 4 5

DC Electric Field Analyzer Energetic Particle Detector Fast Electron Analyzer Fast Ion Analyzer Gamma/X-Ray Spectrometer Ion Composition Plasma Analyzer Solar Wind Analyzer Line-Scan Imaging (and Thermoelectric Cooler) Magnetometer Radio/Plasma Wave Analyzer Solar Neutron Analyzer

6 7 8 9 10 11 Total

Mass (kg)

Power (W)

18.2 15.0 4.0 4.0 17.0

7.0 15.0 5.0 5.0 14.3



10.0 5.1

10.0 11.0

5.3 7.2

5.3 7.2

10.0 105.8

10.0 101.8


Future Missions: Mercury after MESSENGER

Figure 20.1. The last image downlinked from MESSENGER prior to impact on Mercury. Central latitude and longitude are 72.0°N and 223.8°E. The image was acquired on 30 April 2015 at 11:07:43 UTC during orbit 4104 (the last complete orbit of Mercury by the spacecraft). The area imaged is about 1 km across, and the resolution is 2.1 m/pixel.

summarized also in Chapter 1. Moreover, the progress of the mission has been chronicled from its beginning through launch (McNutt et al., 2006), the Venus flybys (McNutt et al., 2008), the first Mercury flybys (McNutt et al., 2010), its primary orbital mission (Bedini et al., 2012; McNutt et al., 2014), and its first extended mission (XM1) from 18 March 2012 to 17 March 2013 (McNutt et al., 2012). In addition, more than 400 articles have been published by the MESSENGER team in the scientific and engineering literature. A second extended mission (XM2) was proposed in February 2013 to last through a planned impact of MESSENGER onto Mercury in March 2015 after propellant for orbit corrections was exhausted. A final extension (XM2′) was proposed on 7 November 2014 and accepted 12 days later to take full advantage of the very low periapsis passages of the final orbits of the mission, allowing for unprecedented, and not likely to be repeated, close-up measurements of the surface and near-surface environment (Chapters 1 and 5). Out of propellant to raise its periapsis altitude further, MESSENGER impacted Mercury’s surface at an estimated spacecraft event time of 3:26 pm Eastern Daylight Time (EDT) (3:34 pm EDT ground receipt time) on 30 April 2015. The impact occurred 3922 days after launch from Cape Canaveral Air Force Station and 1504 days following Mercury orbit insertion, after starting orbit 4105 about the innermost planet of the solar system. The last image downlinked from the spacecraft is shown in Figure 20.1. 20.3.3 Pointing the Way to Deeper Discoveries The progression in capability from Mariner 10 to MESSENGER follows the paradigm of planetary exploration outlined above of initial reconnaissance followed by more in-

depth exploration. Mariner 10 was a low-budget mission relative to other options available, and this choice limited the scope of the scientific payload. Nonetheless, imaging of almost half the planet, discovery of the magnetic field and magnetosphere, discovery of energetic particles, and detection of the planetary exosphere provided many of the insights required to optimize the MESSENGER payload, both in function and in capabilities, given the corresponding limitations on budget and mass for a mission in NASA’s Discovery Program. Primary among these limitations were the inability to include several candidate instruments, such as a plasma wave detector, a thermal emission spectrometer, and electric field probes, and the need to choose one hemisphere for a more detailed investigation than the other, in order to accommodate thermal design constraints. With bodyfixed instruments and only moderate downlink rates, pointing scenarios and data return were judged to be adequate for an initial orbital survey but less than ideal for some science questions that were deferred to future missions. The concept for the ideal orbital mission to Mercury articulated by the MeO SWT was never implemented by NASA because of the higher priority of other exploration targets. However, the priority for a dual-orbiter mission to Mercury was higher in Europe and Japan, enabling implementation of the BepiColombo mission as described in detail below. There is a striking correspondence between the MeO SWT strawman payload (Table 20.1) and that of the BepiColombo spacecraft (see Section 20.5). A dualspacecraft mission with a more extensive payload, however, comes at substantially higher cost than MESSENGER. The total cost for MESSENGER through all mission extensions was ~$500 million in real-year dollars, compared with the projected costs for the BepiColombo mission, which are currently ~$1.3 billion for ESA and $140 million for JAXA. Although not initially planned in such a way (Balogh et al., 2007b), the sequenced approach of the NASA Discoveryclass mission MESSENGER arriving at Mercury nearly 15 years earlier than BepiColombo yielded many benefits from a scientific viewpoint. MESSENGER provided answers to the key science questions that it targeted, but it also turned up a number of unexpected results that raise new questions to be addressed by BepiColombo. Moreover, the vast amount of new information about Mercury from the MESSENGER mission is helping the BepiColombo mission with targeted investigations and mission operations planning.

20.4 SCIENTIFIC AND TECHNICAL A DV AN CE S O F M E S S E NG ER : FA CILITATING FU TUR E M E RCU RY E X P LO R A TI O N The MESSENGER mission led to important achievements in both science and engineering that profoundly advanced our understanding of Mercury and point to new directions for future exploration. The scientific advances are summarized by mission phase in Chapter 1 and covered in detail in the other chapters of this volume. Although MESSENGER answered the scientific questions that framed its primary and extended missions, this

20.4 Scientific and Technical Advances of MESSENGER deeper understanding has led to new mysteries. As the first extended mission built on the results from the primary mission, and the second extension on those of the first, the BepiColombo mission to Mercury will build on the results from MESSENGER. In addition to the scientific direction and focus that MESSENGER provides, there were a number of engineering achievements that are relevant to future exploration. To set the context for the description of BepiColombo, it is useful to review the MESSENGER results and achievements from the perspective of the technical opportunities they enable and new questions they raise. MESSENGER and BepiColombo have been and remain complementary missions (McNutt et al., 2004). Aside from repeating some measurements more than a decade later, thereby providing a longer temporal baseline for a variety of measurements and observations, BepiColombo will collect a number of high-resolution observations and make many entirely new observations. Compared with MESSENGER, the southern hemisphere will be observed from markedly lower altitudes, and the two BepiColombo spacecraft will perform simultaneous measurements of the magnetic field and its dynamic response to changes in solar activity from two spacecraft in different orbits. BepiColombo contains payload instruments not included on MESSENGER, such as a thermal infrared spectrometer, a full complement of plasma physics instrumentation, and a tripleband, radio-science instrument with an on-board accelerometer to obtain high-precision measurements to determine the orbit and test gravitational theory. Indeed, that the BepiColombo mission, nearly ready for launch, is well suited to resolve many of the new mysteries revealed by the discoveries from the MESSENGER mission makes this an especially exciting time for Mercury science. 20.4.1 Surface Composition Elemental composition mapping from MESSENGER was largely limited to the northern hemisphere, and most maps are at low (hundreds of kilometers) spatial resolution because of limited count rates in the measuring instruments and the eccentricity of MESSENGER’s orbit (Chapter 2). Acquiring compositional maps at higher spatial resolution is key to understanding processes responsible for different geologic units and other surface features discovered with MESSENGER imaging (Chapters 6, 7, 8, and 11). The compositions of the hollows (Chapter 12) and polar deposits (Chapter 13) are particularly interesting and could not be well resolved in the MESSENGER data. Resolving compositional variations may also allow determination of unique signatures that might establish whether there are any samples from Mercury in meteorite collections on Earth. More globally, the composition derived from MESSENGER revealed a low oxygen-to-silicon ratio of surface materials, a characteristic not understood; the anomaly could be a surface processing effect or may reflect the bulk properties of the surface materials (Chapter 7). MESSENGER showed, too, that carbon in the form of graphite is likely the dominant darkening agent on Mercury’s surface (Chapters 7 and 8), but improved measurements of the surface carbon abundance and its variation with geochemical terrane and morphological and spectral unit would be valuable. The nature of the high-magnesium region


(Chapters 2 and 7), mechanisms of source processes for neutral and ionized species in the exosphere and magnetosphere (Chapters 14, 15, 16, and 17), and details of the space weathering of Mercury’s surface (Chapter 8) are also open issues, as are the identification and roles of various minor exospheric species (Chapters 14 and 15). 20.4.2 Interior Structure The radio science and altimetry data from MESSENGER have yielded significant new data and constraints on the geophysical properties of Mercury that in turn raise many new questions (Chapters 3 and 4). MESSENGER observations of degree-2 coefficients in the spherical harmonic expansion of Mercury’s gravity field, C20, and C22, together with Earth-based radar observations of the libration amplitude and obliquity, have been used to determine the planet’s interior structure (Chapter 4). These results constrain the core radius to be 2000–2050 km and the outer silicate shell thickness to be ~400 km. However, owing to MESSENGER’s eccentric orbit, the gravity field in the northern hemisphere was determined to much higher resolution than that in the southern hemisphere (Chapter 3), so determination of the distribution of mascons and other gravity anomalies in the south must await comparable coverage in the southern hemisphere. Tracking of the BepiColombo spacecraft will help to refine the gravity field of Mercury and will yield refinements to the size and physical state of its core. The mission will provide additional constraints on models of the planet’s internal structure and test theories of gravity with unprecedented accuracy, e.g., by obtaining full global coverage of topography because of the more nearly circular orbit of the BepiColombo Mercury Planetary Orbiter (see Section 20.5.4) and the shift of its periapsis over the course of the mission. In addition, the equatorial asymmetry in the global magnetic field identified by MESSENGER, the crustal magnetic field of Mercury, discovered during MESSENGER’s second extended mission, and the nature of how field-aligned currents at Mercury close through the planet’s interior (Chapter 5) bear further investigation with lower altitude observations in the southern hemisphere and simultaneous observations at high altitudes, in the solar wind, and near the planet. 20.4.3 Volcanic Processes Not only did the MESSENGER observations provide definitive confirmation that volcanism played a substantial role in Mercury’s surface evolution, but they revealed that multiple volcanic processes have been active (Chapters 6 and 11). The initial characterization of the diverse forms of volcanism on Mercury motivates many new questions. The relationship between volcanism and mantle processes, the variation of magma composition with time, the hemispheric difference in the areal density of impact basins, and the hemispheric asymmetry in the distribution of smooth plains remain to be fully explained (Chapters 6 and 11). Specifically, it is not known how pyroclastic and effusive eruptions are related. Understanding relationships between volcanism and global contraction may allow greater understanding of the planet’s thermal evolution.


Future Missions: Mercury after MESSENGER

Further insights into these questions can be attained with highresolution, global measurements from BepiColombo instrumentation. BepiColombo can contribute to the search for additional basins and improve our understanding of spatial and temporal variations in volcanic eruptive style. BepiColombo high-resolution imaging and topography can be helpful, particularly in the southern hemisphere where imaging resolution by MESSENGER was poorer than in the north. Measurements to determine relations among fault activity, plains emplacement, and craters of differing ages can also be provided by the higherresolution compositional, altimetric, gravity, and spectroscopic observations from BepiColombo.

most scenarios previously advanced for planet formation (Chapters 18 and 19). The orbits and instrumentation of BepiColombo should allow major advances in understanding the physical processes and geological implications of the polar deposits. Most obviously, the BepiColombo Mercury Planetary Orbiter (see Section 20.5.4) will enable study of the characteristics of volatile emplacement in the south polar region of Mercury as well as the north polar region. The expanded instrumentation on BepiColombo will enable combined observations to investigate the chemical and physical characteristics of some of these deposits.

20.4.4 Mineralogy

20.4.6 Hollows

One of the chief results from MESSENGER is that Mercury’s surface has been subjected to intense space weathering, which has resulted in substantial muting of spectroscopic absorption signatures from near-ultraviolet to near-infrared wavelengths characteristic of different minerals (Chapters 6 and 7). The muted spectral signatures are in part the result of the low iron abundance in Mercury’s silicate fraction (Chapter 2) and in part the result of higher rates of space weathering on Mercury than on the Moon or other airless bodies (Domingue et al., 2014; Chapter 8). Characterizing the surface mineralogy of Mercury has therefore proved to be more challenging than anticipated, and the mineralogy of Mercury’s surface has largely been inferred to date from elemental composition and petrological considerations (Chapter 7). Broad spectral variations between geologic units are present, however, and raise key questions that the instrumentation carried by BepiColombo can help answer. Spectral units are distinguishable on the basis of color and overall reflectance (Chapter 8), and crater ejecta, hollows, and volcanic deposits often differ in spectral character from their surroundings (Chapters 9, 11, and 12). Understanding the physical processes that produced these differences requires characterization of the mineralogical variations at scales as small as individual geologic features, which proved a challenge to MESSENGER measurements. BepiColombo will make more sensitive spectral observations and measurements over a greater range of wavelengths (e.g., mid-infrared) than MESSENGER and is therefore likely to improve our understanding of surface mineralogy and its variation among spectral and geological units.

The discovery of hollows – rimless depressions commonly associated with bright haloes – on Mercury’s surface was a complete surprise (Chapter 12). Hollow formation is thought to involve a volatile wasting process that may have occurred via sublimation, space weathering, pyroclastic volcanism, or outgassing. Understanding the physical processes that produce the hollows and linking this information to the geologic processes that bring volatile material to the surface requires compositional and mineralogical characterization on spatial scales at least as small as clusters of hollows, and the BepiColombo instruments are likely to provide invaluable observations to constrain the physical processes and implications of these remarkable formations.

20.4.5 Polar Deposits MESSENGER confirmed earlier suggestions from Earth-based radar imaging that Mercury’s polar deposits consist predominantly of water ice and discovered that most of these deposits are overlain by a layer of dark volatile material several tens of centimeters in thickness (Chapter 13). The formation dynamics and sources of the polar deposits are not known; nor is the composition of the dark surficial layer, the nature of its interactions with the underlying water ice, or the extent to which the trapping of water ice and other frozen volatiles is sporadic or ongoing. Moreover, MESSENGER showed that Mercury’s crustal material is generally richer in volatiles than previously thought, and the compositions are not consistent with

20.4.7 Neutral Particle Environment: Exosphere and Dust MESSENGER was not instrumented to survey the dust environment of Mercury, although the altitude distribution and temporal variation of the density of neutral atoms in the exosphere point to micrometeoroid impacts as an important source process (Chapters 14 and 15). Hence, direct measurements of the dust environment as planned for BepiColombo should confirm these inferences of the dynamic sources for the exosphere. The discovery of hollows and confirmation of water-ice deposits in the polar regions by MESSENGER suggest that sensitive neutral and ion instruments might be able to probe material sputtered from these regions, and the higher-sensitivity neutral particle spectrometers carried by BepiColombo may allow specific inferences on the material in hollows and polar deposits. 20.4.8 Magnetosphere The extensive survey of the magnetic and ion plasma dynamics of Mercury’s magnetosphere conducted by MESSENGER revealed a remarkably dynamic and active space environment with substantial acceleration and precipitation to the surface of energetic particles (Chapters 16 and 17). These results make it even more imperative to measure the thermal electron population, plasma convection dynamics, and electromagnetic wave environment, which MESSENGER was not instrumented to resolve but which are key targets of BepiColombo instrumentation. Moreover, the remarkable

20.5 BepiColombo: The Next Step dynamics of the magnetosphere raises fundamental questions regarding internal magnetospheric processes versus those driven directly by the interaction with the solar wind, a topic that will be better addressed with the two-point magnetic field observations that are planned with BepiColombo. Characterizing the magnetosphere is of primary importance to deriving the structure of Mercury’s internally generated magnetic field (Chapters 6 and 16). There are several ways in which BepiColombo can advance quantitative determination of the magnetospheric magnetic field. Because the loweraltitude (60° away from the thermally critical perihelion conditions is allowed. The final Mercury orbit insertion maneuvers will be performed by chemical propulsion engines integrated into the MPO. The MPO chemical propulsion system will provide the required thrust for a firm Mercury capture and orbit injection, using a set of four 22 N thrusters. The velocity changes needed to insert the MMO and MPO into their nominal orbits are 325 m/s and 620 m/s, respectively. The arrival conditions are constrained such that the MPO and MMO operational orbits will have their lines of apsides close to the Mercury equator and their periapses on the antisolar side at Mercury perihelion, which yields a more benign thermal environment than other geometries. The polar orbits are optimal to obtaining global coverage of the measurements. The baselined lifetime of the MPO and MMO in Mercury orbit is one Earth year (about four Mercury years, or two Mercury solar days). A mission extension by another Earth year is optional and has been factored into the consumables and expected radiation damage. Orbit maintenance is not required over the planned operational lifetimes of the MPO and MMO. 20.5.4 Mercury Planetary Orbiter The BepiColombo MPO accommodates 11 scientific instruments (Table 20.2) and has a box-like shape with a size of 3.7 × 2.2 × 1.7 m (Figures 20.2 and 20.3). The entire MPO totals up to 1230 kg of dry mass, including 85 kg of science payload. The structure allows most units and payloads to be mounted on the MPO outer face, ensuring good accessibility during integration. The primary structure carries a thin cage frame to which the high-temperature multi-layer insulation (MLI) is fixed. In the center of the MPO are two tanks that carry the chemical propellant. The MPO is designed to take scientific measurements in all parts of the orbit throughout the Mercury year, implying that most of the apertures of the remote sensing instruments are continuously nadir pointing. As a consequence, five out of six spacecraft faces may be illuminated by the Sun at some point. This total leaves only one spacecraft side for a radiator to dump excess heat into space and to avoid solar exposure of the radiator. A further consequence is that a spacecraft flip-over maneuver is needed twice per Mercury year. One of the biggest challenges in exploring Mercury is the planet’s thermal environment. Not only is the intensity of the solar radiation up to 11 times higher than at Earth, up to 14,000 W/m2, but the dayside of Mercury also reflects about 600 W/m2 of sunlight and emits infrared radiation of as much as ~5400 W/m2 at the MPO orbital altitude. This environment imposes strong requirements on the spacecraft design, particularly elements that are exposed to the Sun and Mercury, such as the solar array, mechanisms, antennas, multi-layer insulation, thermal coatings, and radiator. The development of these elements and the solar electric propulsion system have been the main cost drivers for the MPO. In addition, some late design updates to cope with the harsh environment are responsible for an increase of the overall spacecraft mass by more than a metric ton compared with the initial plans when the mission was adopted.

The MPO design incorporates numerous features to deal with the harsh thermal environment. The outer MLI surface of the MPO has a low solar absorptivity to reflect most of the sunlight. Nevertheless, it heats up to more than 360°C, which prevents the utilization of standard MLI, so a ceramic fabric with titanium layers is used. Low absorptivity is also adopted for the high-gain antenna coating, which, because of its position, is fully exposed to the Sun and the reflected light and infrared radiation from the planet. In addition, the radio science experiment requires a very stable antenna, which is constructed largely of titanium. Although the radiator is not exposed to direct sunlight, it will receive intense reflected light and heat from Mercury. To minimize the influence of this heat flux on the radiator, highly reflective fins (polished and geometrically reflecting outwards) have been mounted to it at an appropriate angle, to minimize absorption of heat radiated from Mercury while allowing radiation to deep space. Inside the spacecraft, temperatures are kept within the standard range (0–40°C), and for specific instruments interface temperatures below −10°C are provided. Thermal transport from spacecraft components to the radiators is accomplished passively or via heat pipes when necessary. The design yields an even temperature distribution within the spacecraft. The average power demand of the MPO in Mercury orbit, when conducting scientific measurements, will be approximately 1300 W, which will be provided by a solar array. The solar array must continuously track the Sun in order to keep the array temperature below 200°C. This requirement will be met by choosing solar incidence angles of up to 80° that yield sufficient power but minimize heating of the solar array. Communication with Earth is ensured via a high-gain antenna, a medium-gain antenna, and two low-gain antennas. The medium-gain antenna is mounted on a boom and provides global coverage, ensuring that contact is maximized with respect to spacecraft attitude and Earth position. The high-gain antenna provides a link with a high data rate for science data transmission and is the basis for the MORE radio science experiment. This link is achieved using X-band uplink for commanding, both Ka and X-band for data downlink, and both Ka and X-band uplink and downlink for radio science. Over one Earth year in Mercury orbit, about 1550 Gbit will be downlinked to Earth. Precise range and range rate (Doppler) measurements are enabled by the dual-frequency uplink and downlink, allowing accurate orbit determination. Spacecraft attitude control is provided by a set of four reaction wheels and small thrusters for angular momentum management. Three star trackers, Sun sensors, and a high-precision gyroscope package are employed as sensors for attitude control. The combination of the star trackers and gyroscopes ensure a precise attitude determination (a few arcseconds), required by several experiments. The Sun sensors are important to minimize the solar irradiation of sensitive surfaces in case of an anomaly that entails a loss of attitude. The control concept ensures that in no case will a sensitive surface be exposed to the Sun for longer than 85 s. The MPO propulsion system consists of four redundant 22 N thrusters pointing from the nadir face of the spacecraft. The thrusters are employed in orbital maneuvers only until final orbit acquisition, after which they will be rendered passive.

20.5 BepiColombo: The Next Step For reaction wheel desaturation and attitude control, four redundant 10 N thrusters are mounted on the radiator. The two propellant tanks in the center of the spacecraft contain pure hydrazine as fuel and nitrogen tetroxide (NTO) as oxidizer. In this dual-mode system the 22 N thrusters are operated in a bipropellant mode using hydrazine and NTO to yield a high specific impulse, whereas the 10 N thrusters operate just with hydrazine, which minimizes contamination by the exhaust plume.


MPO equation of motion in the precise orbit determination (POD) that is the core of the RSE. As a “byproduct,” ISA will deliver a characterization of the MPO dynamics at low frequencies. The ISA accelerometer is able to measure variable accelerations in the frequency band 3 × 10−5 to 0.1 Hz with a sensitivity of 10−8 m/s2 Hz0.5. MPO-MAG

The BepiColombo Laser Altimeter (BELA) (Gunderson and Thomas, 2010) will characterize and measure the figure, topography, and surface morphology of Mercury. It will provide absolute topographic height and position with respect to a Mercury-centered coordinate system. This information will be used to create a digital terrain model that allows quantitative exploration of the planet’s geology and tectonics. In synergy with the stereo camera, BELA will improve knowledge of Mercury’s geology, geomorphology, tectonics, volcanism, and the evolution of the planet. BELA uses a classic approach to laser altimetry. A Nd:YAG laser produces a 50 mJ pulse at 1064-nm wavelength with a duration of 5–8 ns. This beam produces a 60-μrad footprint (20–50 m) on the surface of the planet. The laser light is reflected from the surface, received with a telescope, and fed into the pulse discrimination electronics, which determines the time of flight, the integrated pulse intensity, and the pulse width. Onboard data compression and data storage are essential. BELA requires significant baffling and thermal control and can operate over the dayside and nightside hemispheres, allowing optimum data acquisition. Performance estimates show that data return is expected at altitudes up to 1050 km above the surface with very low probabilities of false detections. Samples will be acquired every 250 m along ground tracks that will be separated by 25 km at the equator. The experiment will provide return pulse intensity and width information, allowing an assessment of surface reflectance and roughness.

The primary objective of the MPO Magnetometer (MPO-MAG) (Glassmeier et al., 2010) is to collect magnetic field measurements to describe Mercury’s planetary magnetic field and its source in great detail. These measurements will improve understanding of the origin, evolution, and current state of the planetary interior. The requirement is to determine all the terms associated with the internal field up to octupole components with high accuracy, using accurate magnetic field measurements on the low portions of the MPO orbit. This campaign will be supported by similar measurements to be made by MMO (Section 20.5.6), to distinguish the effects of the magnetospheric currents on the MPO measurements and to use the MMO measurements directly to augment the database for the determination of the internal field. The secondary objectives of MPO-MAG are related to the interaction of the solar wind with Mercury’s magnetic field and the planet itself. This interaction leads to the formation of highly dynamic global magnetospheric current systems. In particular, measurements close to the planet will allow a determination of the conditions for access of the solar wind to the planetary surface and an assessment of the role and importance of different current systems, including subsurface induction currents sensitive to the conductivity of the interior. These objectives will again be assisted by the planned close association with the magnetic field investigation on the MMO. The MPO-MAG experiment consists of a dual fluxgate magnetometer system that will measure vector magnetic fields from direct current (DC) to 128 Hz frequency within ±2048 nT with a digital resolution better than 60 pT. In order to determine and remove magnetic contamination (alternating current and DC) from the spacecraft, MPO-MAG consists of two sensors, an inboard sensor and an outboard sensor, mounted on a 2.8-m-long boom and separated by 50 cm. ISA MERTIS

The Italian Spring Accelerometer (ISA) (Iafolla et al., 2010) is a three-axis, high-sensitivity accelerometer devoted to measurement of the acceleration related to the non-gravitational perturbations (NGPs) acting on the surface of the MPO spacecraft. The three sensitive elements are based on a mechanical harmonic oscillator with a resonance frequency of 3.5 Hz. The weakest accelerations that ISA is able to measure cause a displacement of the proof masses of about 2×10−11 m. These displacements are detected by means of a capacitive pickup system in a bridge configuration. The NGPs in Mercury orbit are mainly due to the incoming solar visible radiation and visible and infrared radiation from the planet. The scientific objectives of ISA are strongly related to the BepiColombo Radio Science Experiment (RSE) of MORE (Section ISA’s key role is to remove the NGPs from the list of unknowns in the

The goal of the Mercury Thermal Infrared Spectrometer (MERTIS) instrument (Hiesinger and Helbert, 2010) is to provide detailed information about the mineralogical composition of Mercury’s surface material by globally mapping spectral emittance at high spectral resolution. MERTIS will cover a wavelength range from 7 to 14 μm with a spectral resolution of 90 nm, although some binning will be needed to improve the signal-to-noise ratio for identifying subtle spectral features at lower spectral resolution. This resolution will allow the detection and identification of the characteristic features of surface minerals in this spectral region, such as the Christiansen frequencies, Reststrahlen bands, and transparency features. In addition, MERTIS will be able to measure thermophysical properties of the surface, such as thermal inertia and surface texture. MERTIS is an infrared imaging spectrometer and will

20.5.5 Instruments on the Mercury Planetary Orbiter BELA


Future Missions: Mercury after MESSENGER

make use of micro-bolometer technology for which no cooling is required. MERTIS will globally map the planet with a spatial resolution of 500 m. MGNS The scientific goal of the Mercury Gamma-ray and Neutron Spectrometer (MGNS) (Mitrofanov et al., 2010) is to measure the elemental surface and subsurface composition for distinguishable regions over the entire surface of Mercury by measuring (a) the nuclear lines of major elements in Mercury surface material (Na, Fe, Ti, Al, Mg, Si, Ca, O, K, U, and Th), (b) the leakage flux of neutrons, and (c) the lines of naturally radioactive elements, including U, Th, and K. It will also determine the regional distribution of volatile deposits in permanently shadowed polar areas of Mercury and provide a map of column density of these deposits with an accuracy of 0.1 g cm−2 and a surface resolution of about 400 km. MIXS The Mercury Imaging X-ray Spectrometer (MIXS) (Fraser et al., 2010) will measure the planetary X-ray flux from Mercury, stimulated by high-energy solar X-rays and charged particle interactions with the surface of the planet. This information, in combination with simultaneous measurements of the solar X-ray flux with the Solar Intensity X-ray and particle Spectrometer (Section (Huovelin et al., 2010), will allow measurements at high spectral and spatial resolution of the planetary surface composition. In order to achieve its science objectives, MIXS consists of two channels: MIXSC, a collimator providing efficient flux collection over a broad range of energies with a wide field of view for global planetary mapping, and MIXS-T, an imaging telescope with a narrow field of view for high-resolution measurements of the surface. By use of the X-ray fluorescence technique, MIXS will provide a global view of Mercury’s surface composition in both hemispheres. It is expected that MIXS will obtain global elemental abundance maps of key rock-forming elements (e.g., Na, Mg, Al, Si, S, K, Ca, Ti, and Fe) to an accuracy of 10–20%. MIXS will also obtain high-spatial-resolution measurements of the distribution of chemical elements on the local scale. These measurements will enable the composition determination of small surface features (down to the few-kilometer scale) during periods of high solar activity. In addition, the MIXS data set will allow the study of complex interactions among Mercury’s surface, its local environment, and the solar wind, via remote sensing of the energetic electrons and particles that are known to precipitate to the surface and produce X-ray emission (Starr et al., 2012). MORE The Mercury Orbiter Radio Science Experiment (MORE) addresses scientific goals in geodesy, geophysics, and fundamental physics. It will help to determine the gravity field of Mercury as well as the size and physical state of its core. It will provide crucial experimental constraints to model the planet’s internal structure and test theories of gravity with unprecedented accuracy. MORE will also measure the

gravitational oblateness of the Sun and test and characterize the most advanced interplanetary tracking system ever built. Finally, it will assess the performance of the novel tracking system for precise orbit determination and space navigation. These scientific goals will be achieved by means of several data types, generated by MORE itself at the ground station, other onboard instruments (BELA, ISA, and SIMBIO-SYS), and the onboard attitude determination and control system. MORE will also contribute to the determination of Mercury’s obliquity (i.e., the angle that the spin axis makes to the normal to the orbital plane) and the amplitude of its 88-day physical libration in longitude. These two quantities, together with the coefficients of the second-degree harmonics of the gravity field, will more precisely constrain the outer radius of the planet’s molten core. PHEBUS The Probing of Hermean Exosphere by Ultraviolet Spectroscopy (PHEBUS) experiment (Chassefière et al., 2010) is an ultraviolet spectrometer devoted to the characterization (structure, composition, and dynamics) of Mercury’s exosphere and to the understanding of the coupled surface–exosphere–magnetosphere system. The spectral range of PHEBUS spans the major resonance lines of most detected or expected species. One of the key objectives is to produce an average exosphere, i.e., the altitude profile of density for key atmospheric species at different distances from the Sun, and to quantify further north–south and east–west asymmetries (Sprague et al., 1997; Potter et al., 1999). An aim of the experiment is to produce such maps every one-eighth of a Mercury year, that is, on a timescale of 10 Earth days. Mercury’s exosphere is expected to vary rapidly in response to solar wind variations, and therefore it is important to provide partial maps of the exosphere on timescales of less than a few hours. The polar orbit of the spacecraft will allow the exosphere to be monitored at all latitudes but only within a narrow longitudinal region along the orbit, with the restriction that only regions of the exosphere illuminated by the Sun may be observed. The spectral range of PHEBUS is covered by three instruments: an extreme ultraviolet detector (EUV), a far-ultraviolet detector (FUV), and a near-ultraviolet spectrometer (NUV). The EUV detector covers emission lines at 25–155 nm, and the FUV detector does the same at 145–315 nm. The FUV detector is protected by a vacuum cover consisting of a sealed MgF2 window that is transparent above 115 nm, allowing FUV to cover the wavelength range 145–422 nm. The EUV detector must respond to wavelengths shorter than the MgF2 window cutoff, so it has a cover that is opened in flight. The NUV spectrometer monitors the spectrum out to >425 nm to observe exospheric Ca and K emissions at 404 and 422 nm, respectively. The wavelength resolution of all observations will be better than 1 nm. A vertical scanning range covers the altitude range 0–1500 km, with a vertical resolution of about 20 km. SERENA The Search for Exosphere Refilling and Emitted Neutral Abundances (SERENA) experiment (Orsini et al., 2010) will provide information about the global surface–exosphere– magnetosphere system and its interaction with the solar wind.

20.5 BepiColombo: The Next Step The experiment consists of four sensors that can be operated individually: (1) Emitted Low-Energy Neutral Atoms (ELENA) is a 4° × 76° one-dimensional imager (the spacecraft track will provide the second dimension) of energetic neutral particles emitted from the surface of Mercury (with energies from about 50 eV up to 5 keV); (2) Strofio is a mass spectrograph that determines particle mass per charge (mass resolution m/Δm ≥ 60) by a time-of-flight (TOF) technique; (3) Miniature Ion Precipitation Analyser (MIPA) is an ion mass analyzer with a hemispheric field of view for the energy range 10 eV – 15 keV and a time cadence of 22 s per full distribution function to measure ions that precipitate toward the surface; and (4) Planetary Ion Camera (PICAM) is an ion mass spectrometer (mass resolution m/Δm about 50) with a field of view of 0.4π sr, in the energy range from the spacecraft potential up to ~3 keV with 32 energy channels. SERENA will measure in situ both neutral species and ions and, in addition to investigating the planetary response to external forcing, it will complement the MMO for magnetospheric dynamics investigations. The key scientific objectives of SERENA include the identification and localization of source and sink processes of neutral and charged particles as well as estimates of their relative efficiencies. The latter depend on surface composition and external forcing such as solar irradiance, plasma, precipitation, or micrometeoroid impact, and both spatial and temporal variability are expected. Measurement objectives further include the composition and altitude profile of neutral particles and ions in the exosphere for all species, including their energy spectra and spatial distributions. The dynamics of the neutral and ionized exosphere, e.g., circulation from day to night and active to inactive regions, will be investigated as well as atmosphere–magnetosphere exchange and transport processes. SIMBIO-SYS The Spectrometer and Imagers for MPO BepiColomboIntegrated Observatory SYStem (SIMBIO-SYS) instrument suite is an integrated package for imaging and spectroscopic investigation of the surface of Mercury (Flamini et al., 2010). The science goals of SIMBIO-SYS are to examine the surface geology (stratigraphy, geomorphology), volcanism (lava plain emplacement, volcano identification), global tectonics (structural geology, mechanical properties of the lithosphere), surface age (crater population and morphometry, degradation processes), surface composition (maturity and crustal differentiation, weathering, rock-forming mineral abundance determination), and geophysics (libration measurements, internal planet dynamics) of Mercury. It incorporates capabilities to perform global mapping at medium spatial resolution in stereo and color imaging using two pan-chromatic and four broadband filters, respectively, as well as high-spatial-resolution imaging with pan-chromatic and three broadband filters and imaging spectroscopy at visible to near-infrared wavelengths. The instrument suite consists of three units: (a) The Stereo Channel (STC) will provide global color coverage of the surface in full stereo at 60 m/pixel resolution with the aim of defining the main geological units, large-scale tectonic features, impact crater population, and volcanic edifices. The STC design,


composed of two “sub-channels” that utilize the same detector and based on a push-frame acquisition mode, yields good stereo performance with general compactness, saving mass, volume, and power resources (Cremonese et al., 2009; Da Deppo et al., 2010). (b) The High spatial Resolution Imaging Channel (HRIC) will characterize special surface targets with high-resolution images at ground pixel sizes of about 6 m/pixel from 480-km altitude in four different bands. (c) The Visible Infrared Hyperspectral Imager Channel (VIHI) is a hyperspectral imager in the visible to nearinfrared wavelength range (400–2000 nm) that will map the planet to provide the global mineralogical composition of the surface at a spectral resolution of 6.25 nm and at 500 m/pixel size and will give coverage of selected areas with a resolution as good as 125 m (Capaccioni et al., 2010). SIXS The Solar Intensity X-ray Spectrometer (SIXS) experiment (Huovelin et al., 2010) will monitor solar X-rays (SIXS-X) and energetic particles (SIXS-P). The X-ray data are required for a fluorescence analysis of MIXS spectra. Because the intensity and energy spectrum of both X-rays and energetic particles emitted by the Sun are highly variable, simultaneous operation of SIXS and MIXS is a strong requirement. Scientific objectives for SIXS-X are to monitor the solar X-ray corona and solar flares and to determine their temporal variability and spectral classification. Therefore, SIXS-X needs a clear view to the Sun as continuously as possible, whatever the spacecraft attitude. The sensor contains three detectors, each having about a 100°-wide field of view and covering a spectral range of 1–20 keV with about 300-eV resolution. SIXS-P will monitor solar energetic electron and proton fluxes and their variations. The key scientific objective is to study the interaction of this radiation with Mercury’s exosphere, magnetosphere, and surface. SIXS is a key supporting instrument to MIXS, but SIXS can also be operated independently of MIXS (whereas the opposite is not true), as its observations will be desirable for other investigations for which the measurements of solar X-rays and energetic particles are important or necessary inputs. These investigations include exospheric studies with SERENA and PHEBUS on MPO and most studies with the MMO payload. In addition, X-ray observations by SIXS of the side of the Sun not visible to instruments near Earth can be useful to space weather studies at Earth. 20.5.6 Mercury Magnetospheric Orbiter The BepiColombo MMO will be a spin-stabilized spacecraft once it has separated from the MPO following Mercury orbit insertion. The MMO is optimized for in situ measurements of plasma and electromagnetic fields and waves in orbit about Mercury. The nominal spin rate is 15 rpm (or a spin period of 4 s) to meet the scientific requirements. The spin axis is pointed nearly perpendicular to the Mercury orbital plane. The total MMO mass is 255 kg, including 45 kg for the science payload and N2 gas for attitude control after separation. The MMO main structure consists of two decks (upper and lower), a central


Future Missions: Mercury after MESSENGER

cylinder (thrust tube), and four bulkheads. Normal to its short dimension, the spacecraft has an octagonal shape, which can be surrounded by a 1.8-m-diameter circle. The height of the main body is 1.1 m, and the octagonal structure is divided into three parts: upper, middle, and lower sections. The external surface of the upper section is covered in solar cells and optical solar reflectors (OSRs) in a 50:50 ratio, and OSRs are put on the internal surface of the upper section to reduce the cell temperature. The middle section has OSRs on the exterior and MLI on the interior, whereas the lower section is entirely covered with OSRs to reflect the direct solar flux. The instruments are located on the upper and lower decks, which are separated by 37 cm. Most of the scientific instruments (e.g., particle sensors) are mounted on the upper side of the bottom deck, whereas the four deployment units of the electric probe antennas for the Plasma Wave Instrument (PWI) are installed on the lower side of the bottom deck. During the interplanetary cruise phase, the MMO is shielded and thus cannot produce its own power. Therefore, the BepiColombo MPO provides heater power and energy for regular status checks. After MMO separation, the solar cells experience wide temperature variations because of the large range in Mercury’s distance from the Sun. All external surfaces have high electrical conductivity to keep the surface at the same electric potential with respect to the environment, which is essential to measurements of DC electric fields and lowenergy electrons. The MMO is controlled by a combination of passive and active thermal design techniques to maintain the onboard equipment and spacecraft structure within the proper temperature range during all mission phases. The passive control elements are the OSRs, a thermal shield, paints, films, and MLI blankets. The internal surfaces of the upper and lower deck have highemissivity surfaces (black paint) to equalize internal temperature. The external surface of the upper deck is covered by MLI for insulation from the external thermal environment, whereas the external surface of the lower deck is covered by OSRs to give low absorptivity and high emissivity. The octagonal structure (substrate) is insulated from the upper and lower decks with thermal standoffs. Most of the internal components have a surface with high emissivity (black paint) to equalize the internal temperature. The temperatures of the batteries are controlled independently with the aid of radiators and heaters, which are installed on a battery panel. This panel is attached to the bottom of the central cylinder and insulated from the main structure by MLI and thermal standoffs. The radiator has a surface with high emissivity (OSRs). The antenna despun motor (ADM) and gaseous nitrogen (GN2) tank are covered with MLI for insulation from the external thermal environment. The high-gain antenna (HGA) disks are painted white. The 80-cm-diameter HGA is used for the high-rate X-band telemetry (TLM) command (CM) and ranging link, with the use of a 20-W power amplifier. The MMO HGA is pointed toward Earth with the ADM and antenna pointing mechanism (APM) to control elevation angle to between −90° and +15°, depending on the positions of Mercury and Earth. A mediumgain antenna (MGA) is accommodated for emergency TLM/ CM link. The MGA is installed on the lower surface of the MMO and will be deployed after MMO separation. In orbit

about Mercury, the MMO telemetry rate will change as a function of the distance from Earth. The average data rate of the HGA is 16 kbps, which translates into 40 MB/day given a 6-h pass in view of a radio antenna on Earth. A pair of Sun sensors on one side panel will measure the spin-stabilized MMO spacecraft attitude, and a star scanner is attached to the bottom surface. The attitude is controlled by the propulsion system with a cold gas jet. A nutation damper installed inside the central cylinder is used for passive nutation damping. The MMO propulsion system uses a cold-gas jet system, since only attitude control capability is required (i.e., no orbit control function is needed). The system consists of one propellant tank, six 0.2-N nitrogen-gas jet thrusters, valves, piping, and thermal control equipment (heaters and sensors). The four tangential thrusters for roll control are all on side panels, whereas the two axial thrusters are mounted at the bottom of the spacecraft body. The GN2 tank consists of titanium alloy liner and carbon fiber shell. The tank volume is 14.7 liter, and the maximum designed pressure (MDP) is 27.6 MPa. About 4 kg of GN2 will be loaded at the launch site, including residual propellant of 0.25 kg. The downstream part of the propulsion system consists of a fully redundant system.

20.5.7 MMO Sunshield and Interface Structure The MOSIF provides the interface structure between the MPO and the MMO and protects the MMO from the full intensity of the Sun until the probe’s separation, at which time the spacecraft will have reached its operational orbit. The sunshield is a metal truss structure covered with MLI with appropriate thermal finishes inside and out to ensure suitable temperatures for the MMO. The conical shape of the MOSIF – with an opening angle of about 16° – is needed to allow for the lateral velocity and wobble of the MMO generated during its spin-up at separation. 20.5.8 Instruments on the Mercury Magnetospheric Orbiter MDM The main objective of the Mercury Dust Monitor (MDM) (Nogami et al., 2010) is to measure the dust environment at Mercury’s region of the solar system (0.31–0.47 AU). The impact of micrometeoroids may provide an important source process for the planet’s exosphere. At Mercury’s orbit, the main dust components are Keplerian dust particles and betameteoroid particles. The Keplerian dust particles are interplanetary dust particles (IDPs) that originate from asteroids or comets and gradually decrease their solar-centric distance by the Poynting–Robertson effect, whereas beta-meteoroids are dust particles that are on unbound orbits from the direction of the Sun. The MDM system is composed of a 64-cm2 piezoelectric lead zirconate titanate (PZT) sensor unit (MDM-S) attached to the outside of the side panel and the electronics unit (MDM-E) installed inside the spacecraft. This instrument can detect impact momentum, crude direction, and the number density of dust particles in the local environment. The viewing direction covers nearly a hemisphere. The PZT is a very simple

20.5 BepiColombo: The Next Step device that can withstand high temperatures (about 230°C) and does not need any bias voltage or high voltage for operation. MMO-MAG The primary objective of the MMO Magnetometer (MMOMAG) (Baumjohann et al., 2010) is to collect magnetic field measurements to study the variability of Mercury’s magnetosphere and probe the planetary interior. The MMO-MAG consists of fluxgate magnetometers, an outer sensor (MGO-O) that is a so-called digital type and an inner sensor (MGF-I) that is a traditional analog type. Both have their own electronics boards, and both are mounted on a 4.4-m-long boom. The outer sensor is mounted on the tip of the boom, and the inner sensor is mounted 1.6 m from the boom tip. The instrument is designed to measure magnetic fields with an accuracy of about 10 pT, a dynamic range of ±2048 nT, and a sampling rate of up to 128 Hz. The sampling rate of the data transmitted to Earth is flexible and adapted to study each particular process in the different regions of observation. MPPE The magnetic field of Mercury, although weak, is sufficiently strong to stand off the solar wind much of the time and to form a magnetosphere. The Mercury Plasma Particle Experiment (MPPE) is designed to investigate the plasma and particle environment around the planet (Saito et al., 2010). MPPE is a comprehensive instrument suite for measurements of plasma, high-energy particles, and energetic neutral atoms. It consists of seven sensors: two Mercury Electron Analyzers (MEA1 and MEA2) (Sauvaud et al., 2010), the Mercury Ion Analyzer (MIA) (Miyake et al., 2009), the Mercury mass Spectrum Analyzer (MSA) (Delcourt et al., 2009), the High Energy Particle instrument for electrons (HEP-ele), the High Energy Particle instrument for ions (HEP-ion), and the Energetic Neutrals Analyzer (ENA). The first six sensors perform in situ observations and cover the range of particle species and energy of interest from the perspective of space plasma physics. The MEA will provide fast electron measurements at Mercury’s orbit, MIA will provide precise measurements of both solar wind ions and Mercury magnetospheric ions, MSA will provide plasma composition information with high mass resolution, HEP-ele will measure the energy and angular distributions of electrons in the energy range 30–700 keV, and HEP-ion will provide the energy or velocity distribution of ions in the energy range 30–1500 keV. The ENA will image Mercury’s magnetosphere in energetic neutrals created via charge-exchange on the Mercury exosphere and map the plasma precipitation via backscattered neutral particles to investigate the global dynamics of the Mercury magnetosphere and exosphere–magnetosphere interactions. MSASI Direct exposure of Mercury’s rocky surface to the space environment gives the planet distinct characteristics in its atmospheric composition. Its tenuous atmosphere is known to have a substantial sodium component. The Mercury Sodium Atmospheric Spectral Imager (MSASI) (Yoshikawa et al., 2010) is a high-dispersion visible spectrometer working in the


spectral range around the wavelength of the sodium D2 emission (589 nm). A Fabry–Pérot etalon is used to achieve a compact design. A one-degree-of-freedom scanning mirror is employed to obtain full-disk images of the planet. A radiation-tolerant complementary metal-oxide semiconductor (CMOS) device with an image intensifier is used as a photon detector. The Fabry–Pérot interferometer comprises two parallel, flat, transparent plates coated with a film of high reflectivity. Its principal advantage is that its throughput is much higher than that of a prism or grating spectrometer. The MSASI will be the first use of such a device for a planetary mission. The combination of Fabry–Pérot etalon and filter accommodates the mass and power limitations on the instrument and provides high sensitivity (16 counts per 2 ms sampling time per bin per 10 kR, achieving a signal-to-noise ratio of 4) and spectral resolution (0.009 nm or better). PWI Plasma wave and radio wave receivers provide rich information regarding a plasma environment, but no such device flew on either the Mariner 10 or MESSENGER spacecraft. The Plasma Wave Instrument (PWI) on MMO (Kasaba et al., 2010) was designed and developed in a collaboration between Japanese and European scientists. The PWI will provide the first electric field, plasma wave, and radio wave data from the Mercury plasma environment. It will give important information regarding energy exchange processes in the small magnetosphere, where the role of microphysics is particularly important for global dynamics. The PWI consists of three sets of receivers connected to two sets of electric field sensors and two magnetic field sensors. The receivers include an Electric Field Detector (EFD), WaveForm Capture (WFC), and Onboard Frequency Analyzer (OFA), together denoted as EWO for EFD–WFC– OFA; a Spectroscopie des Ondes Radio et du Bruit Electrostatique Thermique (SORBET), which translates to “spectroscopy of radio waves and thermal electrostatic noise;” and the Active Measurement of Mercury’s Plasma (AM2P). The electric field sensors are the Mercury Electric Field In-Situ TOol (MEFISTO) and Wire-Probe anTenna (WPT). The magnetic field sensors are the Low-Frequency Search Coil (LF-SC) and Dual-Band Search Coil (DB-SC). The PWI will observe both waveforms and spectra in the frequency range from DC to 10 MHz for the electric field and from 0.1 Hz to 640 kHz for the magnetic field. 20.5.9 Mercury Transfer Module The Mercury Transfer Module provides the acceleration and braking required during interplanetary cruise to reach eventual capture by Mercury and the large amount of power required by the solar electric propulsion system. The MTM also constitutes the bottom element in the overall spacecraft structure. The MTM is equipped with a bipropellant propulsion system of 10-N thrusters that are used for attitude control activities during cruise. The bipropellant system is also able to provide navigation ΔV maneuvers during cruise. By far the major part of the ΔV required during the cruise


Future Missions: Mercury after MESSENGER

trajectory is delivered by the SEP system, using its four 145mN ion thrusters, which are initially operated singly and later in pairs. The MTM solar arrays use the same hightemperature technologies as developed for the MPO and are rotated away from the Sun for temperature control. At their peak output, the MTM solar array delivers 14 kW, of which 10.6 kW is required by the SEP system. The MTM structure is based on a carbon-fiber-reinforced plastic (CFRP) conical primary structure interfacing with the launch vehicle adapter and the MPO. The mechanical interfaces to MPO are characterized by cup-cone separation systems for in-flight separation 6 years after launch, and they provide the primary load path through the Mercury composite spacecraft (MCS) structure at launch.

2 0 . 6 I N T EN S I V E I N V E S T I G A TI O N S : M ER CU RY LA ND E R/R O V ER 20.6.1 Overview To make further substantive steps forward in our scientific understanding of Mercury, the next mission to Mercury after BepiColombo should be a lander, one that functions as long as possible on the planet’s surface. A lander would provide ground truth on geochemical remote sensing and perform geophysical and chemical measurements. There are several scientifically compelling landing sites that could be selected on the basis of our current understanding. For example, a lander in a permanently shadowed crater near one of the poles could provide answers to questions about the composition and physical characteristics of the volatiles in Mercury’s polar deposits, including the dark material that covers the water ice in most shadowed regions. A landing site near one of the hollows could address the nature of the host material for these features, the volatile material lost during their formation, and their relation to source processes for the exosphere and ions in the magnetosphere. A landing site on volcanic plains or a pyroclastic deposit could ascertain the composition of volcanic material on Mercury. These three types of landing sites are most likely mutually exclusive, but the general questions of planetary heat flow, seismicity, the abundance of graphite as a darkening agent in the crust, and the effects of space weathering could be addressable at any site. As demonstrated by landers and rovers on Mars, the mobility provided by a rover is a distinct advantage to sample better the vicinity of a landing site uncontaminated by the propulsive plume from the landing system. The trade-off of rover complexity and mass versus lander simplicity and limitations would, of course, be subject to detailed scrutiny in the trade-off against transport capabilities and scientific priorities. Thermal constraints would likely lead to a requirement to land at night, soon after “sundown” before the site had cooled to its lowest temperatures (depending upon the thermal inertia of surface layers at the site). A radioisotope power supply, such as the Multi-Mission Radioisotope Thermoelectric Generator (MMRTG) (Hammel et al., 2013) powering the Curiosity rover on Mars (Grotzinger et al., 2012), would be required. Artificial illumination of the work site would be needed, although some illumination, albeit

at a low level, might be provided by emission from sodium atoms in the exosphere or by indirect lighting from illuminated portions of topography in view from the landing site. Whether long-lived solar-powered landers could be implemented in sufficiently favorable thermal locations near Mercury’s poles could be considered as well. Given the large thermal variations even in the polar environment (Paige et al., 2013; Chapter 13), such an approach would require extremely accurate autonomous descent and landing. One could also consider the implementation of a short-lived lander, such as is under study for Europa (Hand et al., 2017). Given the same types of mass constraints that would be encountered for a Europa lander (due to the large amounts of propellant required), a similar lifetime of 20 days or less would be probable for such an implementation. To place a lander or rover on the innermost planet, we may well be at a similar crossroads as faced by those considering early landers on Mars. The early Voyager Program considered the use of the Saturn V launch vehicle to land large spacecraft on Mars (Cortright, 1968), i.e., exploiting large launch vehicles primarily designed to support an expanded human presence in space. Saturn V costs and miniaturization of robotic systems led to the implementation of Mars landers and rovers that could be delivered with smaller launch systems. The much longer flight times to Mercury, coupled with the planet’s depth in the Sun’s gravity well, once again pushes the implementation of current state-of-the-art landers and rovers, e.g., those recently and currently operating on Mars, to extremely large launch vehicles, at this time the Space Launch System (SLS) Block 1B (McNutt and Vernon, 2016), now under development to provide human access with the Orion spacecraft to cislunar space, and, potentially, beyond. 20.6.2 Decadal Survey At the time of the last decadal survey for planetary science (Committee on the Planetary Science Decadal Survey, 2011), MESSENGER was en route to Mercury orbit insertion, although the first flybys of Mercury since those of Mariner 10 had been completed. The orbital phase of MESSENGER’s primary mission remained in the future. BepiColombo was also well into development, but its launch was still several years away. Hence, what was known scientifically about Mercury as described in the decadal survey was at a relatively primitive state compared with what is now known. At the time, measurements of Mercury’s forced libration via Earth-based radar had demonstrated that Mercury has a liquid outer core, but quantification of the extent of the core had to await the better definition of Mercury’s gravitational field, possible only after orbital observations by MESSENGER (Chapters 4 and 19). Of the three main goals for inner planet research in the decadal survey, two are relevant to Mercury, namely “Understand the origin and diversity of terrestrial planets” and “Understand how the evolution of terrestrial planets enables and limits the origin and evolution of life.” The third: “Understand the processes that control climate on Earth-like planets” does not apply to Mercury. The first goal of the decadal survey mentioned above is largely achieved for Mercury by the findings documented in

20.6 Intensive Investigations: Mercury Lander/Rover this book, with more to come from BepiColombo. The second goal is also relevant to Mercury, although less so: “The Moon and Mercury are unlikely to harbor life, but they provide critical records of processes and information about the early solar system when life emerged on Earth.” The objectives under this goal include (Committee on the Planetary Science Decadal Survey, 2011): • Understand the composition and distribution of volatile chemical compounds; • Understand the effects of internal planetary processes on life and habitability; and • Understand the effects of processes external to a planet on life and habitability. MESSENGER results that relate to the first objective concern volatile elements and compounds and their distribution and physical properties, especially the northern polar deposits, where water ice has been confirmed and organic materials have been suggested to make up the dark layer that covers and insulates the water ice in most areas of permanent shadow (Chapter 13). Confirmation and detailed mapping of the corresponding features in the southern polar regions, along with an improved understanding of their nature, are expected outcomes from BepiColombo with its closer view of the southern hemisphere. With respect to the second objective (Committee on the Planetary Science Decadal Survey, 2011): Despite the dearth of spacecraft missions to explore the inner planets in the past decade, there have been several important discoveries about internal processes. Recent flybys of Mercury by MESSENGER have confirmed the dipole field measured by Mariner 10. Flyby data also confirm that Mercury’s plains are volcanic and show that some are far younger than previously had been proposed. Further improvements in our knowledge of Mercury’s internal structure and geologic history are expected after MESSENGER enters its mapping orbit in 2011. Those “further improvements” have indeed materialized during MESSENGER’s orbital mission (Chapters 3, 4, 6–13, 18, and 19). With respect to the third objective, the issues noted in the decadal survey center on volatile influx and escape as well as transport on atmosphere-free bodies, both with (Mercury) and without (the Moon) a global magnetic field. The hollows and the possible presence of organic compounds cold trapped with water ice near the poles of Mercury are both relevant to this objective and were unknown prior to MESSENGER’s orbital mission. Both of these MESSENGER discoveries feed into the identified “Future Directions for Investigations and Measurements” and will have a role in discussions for the next planetary decadal survey. Mercury has been and remains an important part of the study of the inner planets, both in its own right and as it relates to Earth and the origin of life in our solar system. As with all solar system exploration, the advancement of scientific goals at Mercury is a multi-decadal process. As noted in the decadal survey (Committee on the Planetary Science Decadal Survey, 2011):


A series of National Research Council (NRC) reports, culminating in the 2003 planetary science decadal survey, affirm that the exploration of Mercury is central to the scientific understanding of the solar system. The successful achievement of science objectives of the NASA MESSENGER and the European Space Agency-Japan Aerospace Exploration Agency (ESA-JAXA) BepiColombo missions remains a high priority. Given all the advances that will likely come from MESSENGER and BepiColombo, as well as ongoing technology and capability enhancement work, the high priority of Mercury landed science could be revisited at the earliest opportunity in the mid to late years of this decade. With respect to further exploration, in Box 5.1 “Planetary Roadmaps” the decadal survey notes (Committee on the Planetary Science Decadal Survey, 2011): For Mercury, the current MESSENGER mission will provide a wealth of new information that could further redefine our understanding of the planet and modify priorities for future missions. The planned European Space Agency (ESA) BepiColombo mission will augment those data and fill important data gaps. Given these missions, the next logical step for the exploration of Mercury would be a landed mission to perform in situ investigations, such as those delineated in the committee’s study of a Mercury lander concept (Appendixes D and G). Additional Discovery missions and ground-based observations (e.g., at the Arecibo Observatory in Puerto Rico and the National Radio Astronomy Observatory in Green Bank, West Virginia) will be important in addressing data gaps not filled by current and planned missions. Later Mercury missions would likely include the establishment of a geophysical network and sample return. The possibility of follow-up NASA Discovery Program missions to Mercury is also noted, as well as landed seismic networks (Table 11.2 of Committee on the Planetary Science Decadal Survey, 2011). Given the richness and diversity of data and knowledge returned by the MESSENGER mission from Mercury, along with the exploding knowledge of extrasolar planets, many of which orbit their primary star much closer than Mercury orbits our own Sun, it is clear that Mercury is a much higher-priority interest in planetary system studies than had been thought by many at the conclusion of the Mariner 10 flybys. With the upcoming BepiColombo mission, it is also apparent that Mercury science will have a significant role to play in the next planetary decadal survey. In addition to the questions raised by MESSENGER observations of hollows and polar materials noted above, MESSENGER measurements of the magnetic and gravity fields of Mercury at low altitudes near the end of the mission have also posed new intriguing questions for further study (Chapters 3 and 5). Not all additional, in-depth studies would necessarily require taking the arduous path to Mercury’s location. For example, further studies of Mercury’s atmosphere and magnetosphere by remote sensing from Earth orbital, Lagrange point, or even lunar observatories that have low solar elongation capability, perhaps carried out in conjunction with the BepiColombo operations at Mercury, could significantly add to our understanding of Mercury.


Future Missions: Mercury after MESSENGER

Figure 20.5. Concept of a Mercury Lander during local night when the surface temperature is low. The dominant illumination comes from indirect lighting from topographical features in sunlight in view from the landing site. Emissions from exospheric Na (yellow D-line) are shown as faintly visible since the brightness of Na emissions at Mercury is comparable to bright aurorae at Earth (Cassidy et al., 2016). Figure credit: Johns Hopkins University Applied Physics Laboratory.

20.6.3 An Implementation Example A Mercury Lander was studied for technical implementation (Hauck et al., 2010) during the course of the last planetary decadal survey (Committee on the Planetary Science Decadal Survey, 2011). This mission concept was one of 12 studied but not selected by the committee for analysis with the Aerospace Corporation’s cost and technical evaluation methodology, and a notional lander is shown in Figure 20.5. The science objectives for a lander on Mercury, as defined prior to MESSENGER’s orbital mission, included (Hauck et al., 2010): • Characterize major and minor elements of the chemical composition of Mercury’s surface. • Characterize the mineralogy and structural state of the materials at Mercury’s surface. • Investigate the magnitude and time dependence of Mercury’s magnetic field, for at least one location on the surface. • Characterize geologic activity (e.g., volcanism, tectonism, impact cratering) at scales ranging from regional to local. • Determine the rotational state of Mercury. Given the results of MESSENGER, and those anticipated from BepiColombo, the most significant modification would be landing-site selection – possibly concentrating on coldtrapped material in a permanently shadowed crater, as illustrated in Figure 20.5; an example of hollows; or a volcanic deposit well characterized by orbital remote sensing. A potential payload was posed but not analyzed; the focus was placed on “getting there” with a simple lander. Approaches with both ballistic trajectories and chemical propulsion – like MESSENGER – and solar electric propulsion – similar to BepiColombo – were considered, but the former (ballistic) approach was estimated to be lower in cost. In any case, a ~5-year interplanetary cruise time with multiple planetary gravity assists was envisioned. With a reduced 21-kg payload

mass and a landed dry mass of 289 kg, including a 30% reserve margin, the launch mass was 4630 kg (including margins) on an Atlas V 551 launch vehicle. We can compare these characteristics with those of the Curiosity rover currently operating on Mars: 75 kg of instrumentation and a landed mass of 899.2 kg powered with ~110 W of electricity from an MMRTG (Grotzinger et al., 2012). The launch mass was 3893 kg, including a 539-kg fueled stage for cruise and a 2401-kg entry, descent, and landing (EDL) system (aeroshell plus fueled descent stage). So for a Mercury lander about one-third the mass of the Curiosity rover, the launch mass is about 800 kg higher, but this figure is still with a ~5-year interplanetary cruise duration. By using an SLS Block 1B with upper (solid) stages, a direct landing with such a robotic probe on Mercury may be possible, but further study is required.

2 0. 7 MERC UR Y SA MPLE RETUR N? While not an “end game” in planetary exploration, planetary sample returns continue to have the potential for major advances in knowledge of solar system bodies (COMPLEX, 1978). The only real change in viewpoint over the last 40 years regarding such missions has been how technically challenging – and expensive – they actually are. The main challenges of returning a sample from the surface of Mercury are those that have challenged all missions to the innermost planet of the solar system: the location of Mercury deep in the gravity well of the Sun and the associated extreme thermal and radiation environment. As with architectures for sample return missions to Mars, the most tenable robotic approach will be transport of the means of sample collection to the surface of the planet, and, either on that or a subsequent mission, transport to the surface of an ascent vehicle capable of delivering the sample in its sealed transport canister to planetary orbit to await orbital rendezvous, transfer to a return vehicle, and subsequent return of the sample to Earth. In the same way that emplacement of a lander or rover onto the surface of Mercury in a relatively short time will likely require high-capability transport, e.g., an SLS-class chemical rocket, transport of a Mercury Ascent Vehicle (MeAV) to the surface and the subsequent orbital pickup and transport to Earth will likely require two more robotic missions, both of which will also rely on large chemical rockets [compare with the current, notional Mars Sample Return (MSR) architecture (Committee on the Planetary Science Decadal Survey, 2011)]. Once a robotic MSR has been accomplished successfully (and ahead of a human mission to Mars’ surface), some fortuitous similarities between the sizes and surface gravitational acceleration of Mars and Mercury suggest that some commonalities in required flight hardware could be exploited. Operating near Mercury provides thermal challenges not present for a Mars mission, but a simple thermal-shield approach, such as implemented on the Mariner 10, MESSENGER, and BepiColombo missions, can deal with heating by the Sun. In addition, for the surface part of the system, the problems of landing and operating on the nightside of the planet or a permanently shadowed region will need to have already

20.8 A Postscript


been solved for a precursor lander or rover prior to sample return. Operations at the far higher surface temperatures on the sunlit side of the planet would be significantly more challenging technically and are not a pressing science requirement (at least given our current state of knowledge of Mercury). Following Earth and Venus, Mars and Mercury have the two highest surface escape velocities of any of the solid bodies in the solar system, 5.02 km/s and 4.25 km/s, respectively. Although Mercury is smaller than Mars (0.382 versus 0.532 Earth radii, respectively), and less massive (0.055 versus 0.107 Earth masses, respectively), the higher density of Mercury yields a surface gravitational acceleration on that planet nearly equal to that on Mars, 0.38 that at the surface of Earth. Hence, the design of a Mars Ascent Vehicle (MAV) for a sample return will require about the same acceleration, and thus thrust, for a similar-mass MeAV, although the MAV’s resultant parking orbit will reside deeper in Mars’ gravity well than will the MeAV in Mercury’s. The nominally desired 520-km altitude orbit at Mars (Ross et al., 2012) can be translated into an approximate circular orbit at Mercury on the basis of energy considerations. For any planet, the change in energy per mass in going from the planetary surface (radius R) to a circular orbit of altitude h can be estimated as   ΔE 1 GM GM GM 1 1 ¼ þ ¼ 1 2 ð1 þ h=RÞ m 2 ðR þ hÞ R R   1 1 1 ; ¼ ν2esc 1 

demonstrated to date, e.g., on the Deep Space 1 (Rayman and Lehman, 1997; Boice et al., 2000; Rayman and Varghese, 2001) and Dawn (Russell et al., 2004) missions. Solar sails have also been advocated specifically as enabling for Mercury Sample Return (Hughes and McInnes, 2002), but these remain at a very low technology readiness level (TRL) and have been flagged as a low developmental priority for NASA (Steering Committee for NASA Technology Roadmaps, 2012). BepiColombo is an SEP mission but still retains the need for multiple planetary gravity assists, leading to a lengthy transit time to Mercury. Whether future SEP systems can be made more mass efficient to enable faster transit times with very large, SLS-class launch vehicles is an open question. Until that technical and cost issue is settled, the best approach to a sample return mission to Mercury will remain open. A Mercury sample return, built on large launch systems and a successful Mars sample return architecture, is not without challenges but appears technically feasible. Unlike the case at Mars, there is no current advocacy for a human mission to Mercury, and a robotic approach is required. Such a mission, if targeted to a landing site on one of Mercury’s polar deposits, could provide definitive identification of all aspects of the dark surficial materials that insulate water ice in many of Mercury’s permanently shadowed craters. Such identification could well lead to a paradigm shift in our understanding of the transport of organic materials from the outer solar system, the rate and timing of that activity, and its potential connection to the origin(s) of life on Earth.

where m and M are the mass of the spacecraft and planet, respectively, G is the gravitational constant, and vesc is the escape velocity. If we assume for simplicity that the MAV and MeAV vehicles have the same mass (there are trade-offs, of course; the MeAV must deal with a more challenging thermal environment, and the MAV must deal with the atmosphere of Mars), we can equate the energy per mass at each planet that the two vehicles can provide and solve for a corresponding circular orbit altitude at Mercury. For hMars = 520 km, RMars = 3394 km, and RMercury = 2439 km, with the ratio of the square of the escape speeds v2esc,Mercury/v2esc,Mars = 0.717, we obtain hMercury = 3380 km. That is, by using a suitably modified MAV flight system, a Mercury surface sample of the same size could be put into a far higher orbit at Mercury to await pickup. The thermal design would have to be developed, but the relatively high orbit would help to minimize the thermal input from the planet itself, leaving a sunward-pointed thermal shield as the major hardware requirement. The stability of a high circular orbit would need to be assessed, as this would drive the duration of the period over which the sample could be parked in Mercury orbit before retrieval by the Earth-return system. Even more than for a Mercury lander or rover, the use of highperformance launch vehicles for a Mercury Sample Return mission is compelling. Although low-thrust, in-space propulsion systems, notably solar sails and SEP, have been discussed for some time for use on Mercury orbital missions (e.g., Friedlander and Feingold, 1977), only SEP has been



2 ð1 þ h=RÞ

Exploration is never over, nor should it be. New scientific results lead to new insights and new questions. How the current era of solar system exploration will play out is a work in progress and will remain so for many decades to come. In any event, it will only be with a lander on Mercury – and perhaps a returned sample – that some of our current questions about the innermost planet can be definitively answered. The curious 3:2 spin–orbit resonance of Mercury (Goldreich and Peale, 1966; Correia and Laskar, 2004) will not, however, be one of those questions. The origin of this resonance is a different class of problem – and one relating to the origin of the solar system itself – which may be “solved” only with new insight gained from extrasolar planet research (Brown et al., 2014). Additional sample returns following a first will remain even more problematic than for Mars because of the technical difficulties. This challenge will place even more pressure on sampling the “right” spot the first time (COMPLEX, 1978), which has been a topic of extended discussion for Mars (e.g., COMPLEX, 1996), a situation further complicated at Mars with the possibility of planetary back-contamination (Nealson et al., 1997). From the current vantage point, it is clear that human missions to Mercury would be so technically challenging that even their serious consideration would be warranted only by some truly unforeseen development in the future. Hence, robotic sampling of Mercury’s surface – with or without returned


Future Missions: Mercury after MESSENGER

samples – will most likely be the extent of the scientific exploration of Mercury as far as one can foresee. That said, there is still much to do.

R EF E R EN CE S Anderson, B. J., Acuña, M. H., Lohr, D. A., Scheifele, J., Raval, A., Korth, H. and Slavin, J. A. (2007). The Magnetometer instrument on MESSENGER. Space Sci. Rev., 131, 417–450. Anderson, D. J., Colombo, G., Esposito, P. B., Lau, E. L. and Trager, G. B. (1987). The mass, gravity field, and ephemeris of Mercury, Icarus, 71, 337–349, doi:10.1016/0019-1035(87)90033-9. Andrews, G. B., Zurbuchen, T. H., Mauk, B. H., Malcom, H., Fisk, L. A., Gloeckler, G., Ho, G. C., Kelley, J. S., Koehn, P. L., Lefevere, T. W., Livi, S. S., Lundgren, R. A. and Raines, J. M. (2007). The Energetic Particle and Plasma Spectrometer instrument on the MESSENGER spacecraft. Space Sci. Rev., 131, 523–556. Balogh, A., Ksanfomality, L. and von Steiger, R. (2007a). Introduction. Space Sci. Rev., 132, 183–187, doi:10.1007/s11214-007-9293-0. Balogh, A., Grard, R., Solomon, S. C., Schulz, R., Langevin, Y., Kasaba, Y. and Fujimoto, M. (2007b). Missions to Mercury. Space Sci. Rev., 132, 611–645, doi:10.1007/s11214-007-9212-4. Baumjohann, W., Matsuoka, A., Magnes, W., Glassmeier, K.-H., Nakamura, R., Biernat, H., Delva, M., Schwingenschuh, K., Zhang, T., Auster, H.-U., Fornacon, K.-H., Richter, I., Balogh, A., Cargill, P., Carr, C., Dougherty, M., Horbury, T. S., Lucek, E. A., Tohyama, F., Takahashi, T., Tanaka, M., Nagai, T., Tsunakawa, H., Matsushima, M., Kawano, H., Yoshikawa, A., Shibuya, H., Nakagawa, T., Hoshino, M., Tanaka, Y., Kataoka, R., Anderson, B. J., Russell, C. T., Motschmann, U. and Shinohara, M. (2010). Magnetic field investigation of Mercury’s magnetosphere and the inner heliosphere by MMO/MGF. Planet. Space Sci., 58, 279–286, doi:10.1016/j.pss.2008.05.019. Bedini, P. D., Solomon, S. C., Finnegan, E. J., Calloway, A. B., Ensor, S. L., McNutt, R. L., Jr., Anderson, B. J. and Prockter, L. M. (2012). MESSENGER at Mercury: A mid-term report. Acta Astronautica, 81, 369–379, doi:10.1016/j.actaastro.2012.07.011. Beerer, J. (1970). Historical account of return trajectory. Interoffice Memorandum, Beerer to Gordon, 16 July 1970. Pasadena, CA: Jet Propulsion Laboratory. Belcher, J. W., Slavin, J. A., Armstrong, T. P., Farquhar, R. W., Akasofu, S. I., Baker, D. N., Cattell, C. A., Cheng, A. F., Chupp, E. L., Clark, P. E., Davies, M. E., Hones, E. W., Kurth, W. S., Maezawa, J. K., Mariani, F., Marsch, E., Parks, G. K., Shelley, E. G., Siscoe, G. L., Smith, E. J., Strom, R., Trombka, J. I., Williams, D. J. and Yen, C. (1991). Mercury Orbiter: Report of the Science Working Team. NASA Technical Memorandum 4255. Greenbelt, MD: NASA Goddard Space Flight Center, 134 pp. Boice, D. C., Soderblom, L. A., Britt, D. T., Brown, R. H., Sandel, B. R., Yelle, R. V., Buratti, B. J., Hicks, M. D., Nelson, R. M., Rayman, M. D., Oberst, J. and Thomas, N. (2000). The Deep Space 1 encounter with comet 19p/Borrelly. Earth Moon Planets, 89, 301–324. Brown, S. P., Mead, A. J., Forgan, D. H., Raven, J. A. and Cockell, C. S. (2014). Photosynthetic potential of planets in 3:2 spin–orbit resonances. Int. J. Astrobiology, 13, 279–289, doi:10.1017/ S1473550414000068. Capaccioni, F., Sanctis, M. C. D., Filacchione, G., Piccioni, G., Ammannito, E., Tommasi, L., Veltroni, I. F., Cosi, M., Debei, S., Calamai, L. and Flamini, E. (2010). VIS-NIR imaging spectroscopy of Mercury’s surface: SIMBIO-SYS/VIHI experiment onboard the

BepiColombo mission. IEEE Trans. Geosci. Remote Sensing, 48, 3932–3940, doi:10.1109/TGRS.2010.2051676. Cassidy, T. A., McClintock, W. E., Killen, R. M., Sarantos, M., Merkel, A. W., Vervack, R. J., Jr. and Burger, M. H. (2016). A cold-pole enhancement in Mercury’s sodium exosphere. Geophys. Res. Lett., 43, 11,121–11,128, doi:10.1002/ 2016GL071071. Cavanaugh, J. F., Smith, J. C., Sun, X., Bartels, A. E., Ramos-Izquierdo, L., Krebs, D. J., McGarry, J. F., Trunzo, R., Novo-Gradac, A. M., Britt, J. L., Karsh, J., Katz, R. B., Lukemire, A. T., Szymkiewicz, R., Berry, D. L., Swinski, J. P., Neumann, G. A., Zuber, M. T. and Smith, D. E. (2007). The Mercury Laser Altimeter instrument for the MESSENGER mission. Space Sci. Rev., 131, 451–479. Chassefière, E., Maria, J. L., Goutail, J. P., Quémerais, E., Leblanc, F., Okano, S., Yoshikawa, I., Korablev, O., Gnedykh, V., Naletto, G., Nicolosi, P., Pelizzo, M. G., Correia, J. J., Gallet, S., Hourtoule, C., Mine, P. O., Montaron, C., Rouanet, N., Rigal, J. B., Muramaki, G., Yoshioka, K., Kozlov, O., Kottsov, V., Moisseev, P., Semena, N., Bertaux, J. L., Capria, M. T., Clarke, J., Cremonese, G., Delcourt, D., Doressoundiram, A., Erard, S., Gladstone, R., Grande, M., Hunten, D., Ip, W., Izmodenov, V., Jambon, A., Johnson, R., Kallio, E., Killen, R., Lallement, R., Luhmann, J., Mendillo, M., Milillo, A., Palme, H., Potter, A., Sasaki, S., Slater, D., Sprague, A., Stern, A. and Yan, N. (2010). PHEBUS: A double ultraviolet spectrometer to observe Mercury’s exosphere. Planet. Space Sci., 58, 201–223, doi:10.1016/j.pss.2008.05.018. Choo, T. H., Murchie, S. L., Bedini, P. D., Steele, R. J., Skura, J. P., Nguyen, L., Nair, H., Lucks, M., Berman, A. F., McGovern, J. A. and Turner, F. S. (2014). SciBox, an end-to-end automated science planning and commanding system. Acta Astronautica, 93, 490–496. Colombo, G. (1965). Rotational period of the planet Mercury. Nature, 208, 575, doi:10.1038/208575a0. Colombo, G. and Shapiro, I. I. (1966). The rotation of the planet Mercury. Astrophys. J., 145, 295–307, doi:10.1086/148762. Committee on the Planetary Science Decadal Survey (2011). Vision and Voyages for Planetary Science in the Decade 2013–2022. Washington, DC: National Academies Press, 382 pp. COMPLEX (Committee on Planetary and Lunar Exploration) (1978). Strategy for Exploration of the Inner Planets: 1977–1987. Washington, DC: National Academies Press, 53 pp. COMPLEX (Committee on Planetary and Lunar Exploration) (1996). On NASA Mars Sample-Return Mission Options, Letter Report. Washington, DC: National Academies Press, 19 pp. Correia, A. C. M. and Laskar, J. (2004). Mercury’s capture into the 3/2 spin-orbit resonance as a result of its chaotic dynamics. Nature, 429, 848–850. Cortright, E. M. (1968). The Voyager Program. American Astronautical Society Science and Technology Series, 16, 65–92. Cremonese, G., Fantinel, D., Giro, E., Capria, M. T., Deppo, V. D., Naletto, G., Forlani, G., Massironi, M., Giacomini, L., Sgavetti, M., Simioni, E., Bettanini, C., Debei, S., Zaccariotto, M., Borin, P., Marinangeli, L. and Flamini, E. (2009). The stereo camera on the BepiColombo ESA/JAXA mission: A novel approach. In Advances in Geosciences, ed. A. Bhardwaj. Singapore: World Scientific Publishing, pp. 305–322. Da Deppo, V., Naletto, G., Cremonese, G. and Calamai, L. (2010). Optical design of the single-detector planetary stereo camera for the BepiColombo European Space Agency mission to Mercury. Appl. Optics, 49, 2910–2919, doi:10.1364/AO.49.002910. Delcourt, D., Saito, Y., Illiano, J. M., Krupp, N., Berthelier, J. J., Fontaine, D., Fraenz, M., Leblanc, F., Fischer, H., Yokota, S.,

References Michalik, H., Godefroy, M., Saint-Jacques, E., Techer, J. D., Fiethe, B., Covinhes, J., Gastou, J. and Attia, D. (2009). The mass spectrum analyzer (MSA) onboard BEPI COLOMBO MMO: Scientific objectives and prototype results. Adv. Space Res., 43, 869–874, doi:10.1016/j.asr.2008.12.002. Domingue, D. L., Chapman, C. R., Killen, R. M., Zurbuchen, T. H., Gilbert, J. A., Sarantos, M., Benna, M., Slavin, J. A., Schriver, D., Trávníček, P. M., Orlando, T. M., Sprague, A. L., Blewett, D. T., Gillis-Davis, J. J., Feldman, W. C., Lawrence, D. J., Ho, G. C., Ebel, D. S., Nittler, L. R., Vilas, F., Pieters, C. M., Solomon, S. C., Johnson, C. L., Winslow, R. M., Helbert, J., Peplowski, P. N., Weider, S. Z., Mouawad, N., Izenberg N. R. and McClintock, W. E. (2014). Mercury’s weather-beaten surface: Understanding Mercury in the context of lunar and asteroidal space weathering studies. Space Sci. Rev., 181, 121–214. Ezell, E. C. and Ezell, L. N. (1984). Voyager: Perils of advanced planning, 1960–1967. In On Mars: Exploration of the Red Planet 1958–1978, Chapter 4, Special Publication SP-4212. Washington, DC: National Aeronautics and Space Administration, pp. 83–119. Flamini, E., Capaccioni, F., Colangeli, L., Cremonese, G., Doressoundiram, A., Josset, J. L., Langevin, Y., Debei, S., Capria, M. T., De Sanctis, M. C., Marinangeli, L., Massironi, M., Mazzotta Epifani, E., Naletto, G., Palumbo, P., Eng, P., Roig, J. F., Caporali, A., Da Deppo, V., Erard, S., Federico, C., Forni, O., Sgavetti, M., Filacchione, G., Giacomini, L., Marra, G., Martellato, E., Zusi, M., Cosi, M., Bettanini, C., Calamai, L., Zaccariotto, M., Tommasi, L., Dami, M., Ficai Veltroni, J., Poulet, F. and Hello, Y. (2010). SIMBIO-SYS: The spectrometer and imagers integrated observatory system for the BepiColombo planetary orbiter. Planet. Space Sci., 58, 125–143, doi:10.1016/j.pss.2009.06.017. Fox, N. J., Velli, M. C., Bale, S. D., Decker, R., Driesman, A., Howard, R. A., Kasper, J. C., Kinnison, J., Kusterer, M., Lario, D., Lockwood, M. K., McComas, D. J., Raouafi, N. E. and Szabo, A. (2016). The Solar Probe Plus mission: Humanity’s first visit to our star. Space Sci. Rev., 204, 7–48. Fraser, G. W., Carpenter, J. D., Rothery, D. A., Pearson, J. F., Martindale, A., Huovelin, J., Treis, J., Anand, M., Anttila, M., Ashcroft, M., Benkoff, J., Bland, P., Bowyer, A., Bradley, A., Bridges, J., Brown, C., Bulloch, C., Bunce, E. J., Christensen, U., Evans, M., Fairbend, R., Feasey, M., Giannini, F., Hermann, S., Hesse, M., Hilchenbach, M., Jorden, T., Joy, K., Kaipiainen, M., Kitchingman, I., Lechner, P., Lutz, G., Malkki, A., Muinonen, K., Näränen, J., Portin, P., Prydderch, M., Juan, J. S., Sclater, E., Schyns, E., Stevenson, T. J., Strüder, L., Syrjasuo, M., Talboys, D., Thomas, P., Whitford, C. and Whitehead, S. (2010). The mercury imaging X-ray spectrometer (MIXS) on bepicolombo. Planet. Space Sci., 58, 79–95, doi:10.1016/j.pss.2009.05.004. Friedlander, A. L. and Feingold, H. (1977). Mercury Orbiter Transport Study. Report SAI 1-120-580-176, NASA-CR-158658. Schaumburg, IL: Science Applications, Incorporated, 128 pp. Glassmeier, K.-H., Auster, H.-U., Heyner, D., Okrafka, K., Carr, C., Berghofer, G., Anderson, B. J., Balogh, A., Baumjohann, W., Cargill, P., Christensen, U., Delva, M., Dougherty, M., Fornaçon, K.-H., Horbury, T. S., Lucek, E. A., Magnes, W., Mandea, M., Matsuoka, A., Matsushima, M., Motschmann, U., Nakamura, R., Narita, Y., O’Brien, H., Richter, I., Schwingenschuh, K., Shibuya, H., Slavin, J. A., Sotin, C., Stoll, B., Tsunakawa, H., Vennerstrom, S., Vogt, J. and Zhang, T. (2010). The fluxgate magnetometer of the BepiColombo Mercury Planetary Orbiter. Planet. Space Sci., 58, 287–299, doi:10.1016/j.pss.2008.06.018. Gold, R. E., Solomon, S. C., McNutt, R. L., Jr., Santo, A. G., Abshire, J. B., Acuña, M. H., Afzal, R. S., Anderson, B. J.,


Andrews, G. B., Bedini, P. D., Cain, J., Cheng, A. F., Evans, L. G., Feldman, W. C., Follas, R. B., Gloeckler, G., Goldsten, J. O., Hawkins, S. E., III, Izenberg, N. R., Jaskulek, S. E., Ketchum, E. A., Lankton, M. R., Lohr, D. A., Mauk, B. H., McClintock, W. E., Murchie, S. L., Schlemm, C. E., II, Smith, D. E., Starr, R. D. and Zurbuchen, T. H. (2001). The MESSENGER mission to Mercury: Scientific payload. Planet. Space Sci., 49, 1467–1479. Goldreich, P. and Peale, S. J. (1966). Spin-orbit coupling in the solar system, Astron. J., 71, 425–438. Goldsten, J. O., Rhodes, E. A., Boynton, W. V., Feldman, W. C., Lawrence, D. J., Trombka, J. I., Smith, D. M., Evans, L. G., White, J., Madden, N. W., Berg, P. C., Murphy, G. A., Gurnee, R. S., Strohbehn, K., Williams, B. D., Schaefer, E. D., Monaco, C. A., Cork, C. P., Del Eckels, J., Miller, W. O., Burks, M. T., Hagler, L. B., Deteresa, S. J. and Witte, M. C (2007). The MESSENGER Gamma-Ray and Neutron Spectrometer. Space Sci. Rev., 131, 339–391. Grotzinger, J. P., Crisp, J., Vasavada, A. R., Anderson, R. C., Baker, C. J., Barry, R., Blake, D. F., Conrad, P., Edgett, K. S., Ferdowski, B., Gellert, R., Gilbert, J. B., Golombek, M., GómezElvira, J., Hassler, D. M., Jandura, L., Litvak, M., Mahaffy, P., Maki, J., Meyer, M., Malin, M. C., Mitrofanov, I., Simmonds, J. J., Vaniman, D., Welch, R. V. and Wiens, R. C. (2012). Mars Science Laboratory mission and science investigation. Space Sci. Rev., 170, 5–56, doi:10.1007/s11214012-9892-2. Gulkis, S., Stetson, D. S. and Stofan, E. R. (1998). Mission to the Solar System: Exploration and Discovery – A Mission and Technology Roadmap. Publication 97–12. Pasadena, CA: Jet Propulsion Laboratory, 65 pp. Gunderson, K. and Thomas, N. (2010). BELA receiver performance modeling over the BepiColombo mission lifetime. Planet. Space Sci., 58, 309–318, doi:10.1016/j.pss.2009.08.006. Hammel, T. E., Bennett, R., Sievers, R. K., Keyser, S., Otting, W. and Gard, L. (2013). Multi-Mission Radioisotope Thermoelectric Generator (MMRTG) performance data and application to life modeling. 11th International Energy Conversion Engineering Conference, American Institute of Aeronautics and Astronautics, 8 pp., doi:10.2514/6.2013-3925. Hand, K. P., Murray, A. E., Garvin, J. B., Brinckerhoff, W. B., Christner, B. C., Edgett, K. S., Ehlmann, B. L., German, C. R., Hayes, A. G., Hoehler, T. M., Horst, S. M., Lunine, J. I., Nealson, K. H., Paranicas, C., Schmidt, B. E., Smith, D. E., Rhoden, A. R., Russell, M. J., Templeton, A. S., Willis, P. A., Yingst, R. A., Phillips, C. B., Cable, M. L., Craft, K. L., Hofmann, A. E., Nordheim, T. A., Pappalardo, R. P. and the Project Engineering Team (2017). Report of the Europa Lander Science Definition Team. Pasadena, CA: Jet Propulsion Laboratory, 264 pp. Hauck, S. A., II, Eng, D. A., Treiman, A., Tahu, G., Lindstrom, K., Blewett, D., Seifert, H., Stambaugh, K., Chavers, G., Oleson, S., Mcguire, M., Guo, Y., Dankanich, J., Dong, C., Burke, L., Wolfarth, L., Hahn, M., Drexler, J., Holdridge, M., Cockrell, J., Miller, T., Trinh, H., Fittje, J., Verhey, T., Gyekenyesi, J., Ercol, J., Abel, E., Colozza, T., Sequeira, B., Warner, J., Fraeman, M., Williams, G., Schmitz, P., Lowery, E., Landis, G., Hojniki, J., Adams, D., Martini, M., Williams, S. and Drexler, J. (2010). Mercury Lander Mission Concept Study. Laurel, MD: The Johns Hopkins University Applied Physics Laboratory, 132 pp. Hawkins, S. E., III, Boldt, J. D., Darlington, E. H., Espiritu, R., Gold, R. E., Gotwols, B., Grey, M. P., Hash, C. D., Hayes, J. R., Jaskulek, S. E., Kardian, C. J., Keller, M. R., Malaret, E. R., Murchie, S. L., Murphy, P. K., Peacock, K., Prockter, L. M.,


Future Missions: Mercury after MESSENGER

Reiter, R. A., Robinson, M. S., Schaefer, E. D., Shelton, R. G., Sterner, R. E., Taylor, H. W., Watters, T. R. and Williams, B. D. (2007). The Mercury Dual Imaging System on the MESSENGER spacecraft. Space Sci. Rev., 131, 247–338. Hiesinger, H., Helbert, J. and MERTIS Co-I Team (2010). The Mercury Radiometer and Thermal Infrared Spectrometer (MERTIS) for the BepiColombo mission. Planet. Space Sci., 58, 144–165, doi:10.1016/j.pss.2008.09.019. Hughes, G. and McInnes, C. (2002). Mercury sample return missions using solar sail propulsion. 34th Scientific Assembly of the Committee on Space Research and 53rd International Astronautical Congress of the International Astronautical Federation, World Space Congress 2002, paper IAC-02-W.2.08. Houston, TX, 11 pp. Huovelin, J., Vainio, R., Andersson, H., Valtonen, E., Alha, L., Mälkki, A. Grande, M., Fraser, G. W., Kato, M., Koskinen, H., Muinonen, K., Näränen, J., Schmidt, W., Syrjäsuo, M., Anttila, M., Vihavainen, T., Kiuru, E., Roos, M., Peltonen, J., Lehti, J., Talvioja, M., Portin, P. and Prydderch, M. (2010). Solar Intensity X-ray and particle Spectrometer (SIXS). Planet. Space Sci., 58, 96–107, doi:10.1016/j.pss.2008.11.007. Iafolla, V., Fiorenza, E., Lefevre, C., Morbidini, A., Nozzoli, S., Peron, R., Persichini, M., Reale, A. and Santoli, F. (2010). Italian Spring Accelerometer (ISA): A fundamental support to BepiColombo radio science experiments. Planet. Space Sci., 58, 300–308, doi:10.1016/j.pss.2009.04.005. Jehn, R., Campagnola, S., Garcia, D. and Kemble, S. (2004), Lowthrust approach and gravitational capture at Mercury. 18th International Symposium on Space Flight Dynamics. Munich, Germany, 6 pp. Kasaba, Y., Bougeret, J.-L., Blomberg, L. G., Kojima, H. Yagitani, S., Moncuquet, M., Trotignon, J.-G., Chanteur, G., Kumamoto, A., Kasahara, Y., Lichtenberger, J., Omura, Y., Ishisaka, K. and Matsumoto, H. (2010). The Plasma Wave Investigation (PWI) onboard the BepiColombo/MMO: First measurement of electric fields, electromagnetic waves, and radio waves around Mercury. Planet. Space Sci., 58, 238–278, doi:10.1016/j.pss.2008.07.017. Kirk, M. N., Flanigan, S. H., O’Shaughnessy, D. J., Bushman, S. S. and Rosendall, P. E. (2015). MESSENGER maneuver performance during the low-altitude hover camaign. Astrodynamics Specialist Conference, American Astronautical Society, paper AAS 15–652. Vail, CO, 19 pp. Langevin, Y. (2000). Chemical and solar electric propulsion options for a cornerstone mission to Mercury. Acta Astronautica, 47, 443–452, doi:10.1016/S0094-5765(00)00084-9. Leary, J. C., Conde, R. F., Dakermanji, G., Engelbrecht, C. S., Ercol, C. J., Fielhauer, K. B., Grant, D. G., Hartka, T. J., Hill, T. A., Jaskulek, S. E., Mirantes, M. A., Mosher, L. E., Paul, M. V., Persons, D. F., Rodberg, E. H., Srinivasan, D. K., Vaughan, R. M. and Wiley, S. R. (2007). The MESSENGER spacecraft. Space Sci. Rev., 131, 187–217. Lewis, R. A., Pérez Luna, J., Coombs, N. and Guarducci, F. (2015). Qualification of the T6 thruster for BepiColombo. Joint Conference of the 30th International Symposium on Space Technology and Science, 34th International Electric Propulsion Conference, and 6th Nano-satellite Symposium. Hyogo-Kobe, Japan, 10 pp. McAdams, J. V., Farquhar, R. W., Taylor, A. H. and Williams, B. G. (2007). MESSENGER mission design and navigation. Space Sci. Rev., 131, 219–246. McAdams, J. V., Bryan, C. G., Bushman, S. S., Calloway, A. B., Carranza, E., Flanigan, S. H., Kirk, M. N., Korth, H., Moessner, D. P., O’Shaughnessy, D. J. and Williams, K. E. (2015). Engineering MESSENGER’s grand finale at Mercury – The low-

altitude hover campaign. Astrodynamics Specialist Conference, American Astronautical Society, paper AAS 15–634. Vail, CO, 20 pp. McClintock, W. E. and Lankton, M. R. (2007). The Mercury Atmospheric and Surface Composition Spectrometer for the MESSENGER mission. Space Sci. Rev., 131, 481–521. McNutt, R. L., Jr. and Vernon, S. R. (2016). Enabling solar system science with the Space Launch System (SLS). 67th International Astronautical Congress, International Astronautical Federation. Guadalajara, Mexico, 7 pp. McNutt, R. L., Jr., Solomon, S. C., Grard, R., Novara, M. and Mukai, T. (2004). An international program for Mercury exploration: Synergy of MESSENGER and BepiColombo. Adv. Space Res., 33, 2126–2132, doi:10.1016/s0273-1177(03)00439-3. McNutt, R. L., Jr., Solomon, S. C., Gold, R. E., Leary, J. C. and the MESSENGER Team (2006). The MESSENGER mission to Mercury: Development history and early mission status, Adv. Space Res., 38, 564–571, doi:10.1016/j.asr.2005.05.044. McNutt, R. L., Jr., Solomon, S. C., Grant, D. G., Finnegan, E. J. and Bedini, P. D. (2008). The MESSENGER mission to Mercury: Status after the Venus flybys. Acta Astronautica, 63, 68–73. McNutt, R. L., Jr., Solomon, S. C., Bedini, P. D., Finnegan, E. J. and Grant, D. G. (2010). The MESSENGER mission: Results from the first two Mercury flybys. Acta Astronautica, 67, 681–687, doi:10.1016/j.actaastro.2010.05.020. McNutt, R. L., Jr., Solomon, S. C., Nittler, L. R., Bedini, P. D., Finnegan, E. J., Winters, H. L. and Grant, D. G. (2012). The MESSENGER mission continues: Transition to the extended mission. International Astronautical Congress. Naples, Italy, 15 pp. McNutt, R. L., Jr., Solomon, S. C., Bedini, P. D., Anderson, B. J., Blewett, D. T., Evans, L. G., Gold, R. E., Krimigis, S. M., Murchie, S. L., Nittler, L. R., Phillips, R. J., Prockter, L. M., Slavin, J. A., Zuber, M. T., Finnegan, E. J. and Grant, D. G. (2014). MESSENGER at Mercury: Early orbital operations. Acta Astronautica, 93, 509–515, doi:10.1016/j.actaastro.2012.08.012. Mitrofanov, I. G., Kozyrev, A. S., Konovalov, A., Litvak, M. L., Malakhov, A. A., Mokrousov, M. I., Sanin, A. B., Tret’ykov, V. I., Vostrukhin, A. V., Bobrovnitskij, Y. I., Tomilina, T. M., Gurvits, L. and Owens, A. (2010). The Mercury Gamma and Neutron Spectrometer (MGNS) on board the Planetary Orbiter of the BepiColombo mission. Planet. Space Sci., 58, 116–124, doi:10.1016/j.pss.2009.01.005. Miyake, W., Saito, Y., Harada, M., Saito, M., Hasegawa, H., Ieda, A., Machida, S., Nagai, T., Nagatsuma, T., Seki, K., Shinohara, I. and Terasawa, T. (2009). Mercury Ion Analyzer (MIA) onboard Mercury Magnetospheric Orbiter: MMO. Adv. Space Res., 43, 1986–1992, doi:10.1016/j.asr.2009.03.011. Nealson, K. H., Carr, M. H., Clark, B. C., Doolittle, R. F., Jakosky, B. M., Korwek, E. L., Pace, N. R., Poindexter, J. S., Race, M. S., Reysenbach, A.-L., Schopf, J. W. and Stevens, T. O. (1997). Mars Sample Return: Issues and Recommendations. Washington, DC: National Academies Press, 57 pp. Nelson, R. M., Horn, L. J., Weiss, J. R. and Smythe, W. D. (1995). Hermes Global Orbiter: A Discovery mission in gestation. Acta Astronautica, 35, Suppl. 1, 387–395, doi:10.1016/0094-5765(94) 00204-5. Nogami, K., Fujii, M., Ohashi, H., Miyachi, T., Sasaki, S., Hasegawa, S., Yano, H., Shibata, H., Iwai, T., Minami, S., Takechi, S., Grün, E. and Srama, R. (2010). Development of the Mercury dust monitor (MDM) onboard the BepiColombo mission. Planet. Space Sci., 58, 108–115, doi:10.1016/j. pss.2008.08.016. Novara, M. (2002). The BepiColombo ESA cornerstone mission to Mercury. Acta Astronautica, 51, 387–395, doi:10.1016/S00945765(02)00065-6.

References Orsini, S., Livi, S., Torkar, K., Barabash, S., Milillo, A., Wurz, P., Di Lellis, A. M. and Kallio, E. (2010). SERENA: A suite of four instruments (ELENA, STROFIO, PICAM and MIPA) on board BepiColombo-MPO for particle detection in the Hermean environment. Planet. Space Sci., 58, 166–181, doi:10.1016/j. pss.2008.09.012. O’Shaughnessy, D. J., McAdams, J. V., Bedini, P. D., Calloway, A. B., Williams, K. E. and Page, B. R. (2014). MESSENGER’s use of solar sailing for cost and risk reduction. Acta Astronautica, 93, 483–489. Paige, D. A., Siegler, M. A., Harmon, J. K., Neumann, G. A., Mazarico, E. M., Smith, D. E., Zuber, M. T., Harju, E., Delitsky, M. L. and Solomon, S. C. (2013). Thermal stability of volatiles in the north polar region of Mercury. Science, 339, 300–303. Philpott, L. C., Johnson, C. L., Winslow, R. M., Anderson, B. J., Korth, H., Purucker, M. E. and Solomon, S. C. (2014), Constraints on the secular variation of Mercury’s magnetic field from the combined analysis of MESSENGER and Mariner 10 data. Geophys. Res. Lett., 41, 6627–6634, doi:10.1002/ 2014GL061401. Potter, A. E., Killen, R. M. and Morgan, T. H. (1999). Rapid changes in the sodium exosphere of Mercury. Planet. Space Sci., 47, 1441–1448, doi:10.1016/S0032-0633(99)00070–7. Rayman, M. D. and Lehman, D. H. (1997). Deep Space One: NASA’s first deep-space technology validation mission. Acta Astronautica, 41, 289–299. Rayman, M. D. and Varghese, P. (2001). The Deep Space 1 extended mission. Acta Astronautica, 48, 693–705. Ross, D., Russell, J. and Sutter, B. (2012). Mars Ascent Vehicle (MAV): Designing for high heritage and low risk. 2012 IEEE Aerospace Conference. Big Sky, MT, 6 pp., doi:10.1109/ AERO.2012.6187296. Russell, C. T., Coradini, A., Christensen, U., De Sanctis, M. C., Feldman, W. C., Jaumann, R., Keller, H. U., Konopliv, A. S., McCord, T. B., McFadden, L. A., McSween, H. Y., Mottola, S., Neukum, G., Pieters, C. M., Prettyman, T. H., Raymond, C. A., Smith, D. E., Sykes, M. V., Williams, B. G., Wise, J. and Zuber, M. T. (2004). Dawn: A journey in space and time. Planet. Space Sci., 52, 465–489, doi:10.1016/j.pss.2003.06.013. Saito, Y., Sauvaud, J. A., Hirahara, M., Barabash, S., Delcourt, D., Takashima, T. and Asamura, K. (2010). Scientific objectives and instrumentation of Mercury Plasma Particle Experiment (MPPE) onboard MMO. Planet. Space Sci., 58, 182–200, doi:10.1016/j. pss.2008.06.003. Santo, A. G., Gold, R. E., McNutt, R. L., Jr., Solomon, S. C., Ercol, C. J., Farquhar, R. W., Hartka, T. J., Jenkins, J. E., McAdams, J. V., Mosher, L. E., Persons, D. F., Artis, D. A., Bokulic, R. S., Conde, R. F., Dakermanji, G., Goss, M. E., Jr., Haley, D. R., Heeres, K. J., Maurer, R. H., Moore, R. C., Rodberg, E. H., Stern, T. G., Wiley, S. R., Williams, B. G., Yen, C. L. and Peterson, M. R. (2001). The MESSENGER mission to Mercury: Spacecraft and mission design. Planet. Space Sci., 49, 1481–1500. Sauvaud, J. A., Fedorov, A., Aoustin, C., Seran, H. C., Le Comte, E., Petiot, M., Rouzaud, J., Saito, Y., Dandouras, J., Jacquey, C., Louarn, P., Mazelle, C. and Médale, J. L. (2010). The Mercury Electron Analyzers for the Bepi Colombo mission. Adv. Space Res., 46, 1139–1148, doi:10.1016/j.asr.2010.05.022. Schlemm, C. E., II, Starr, R. D., Ho, G. C., Bechtold, K. E., Hamilton, S. A., Boldt, J. D., Boynton, W. V., Bradley, W., Fraeman, M. E., Gold, R. E., Goldsten, J. O., Hayes, J. R., Jaskulek, S. E., Rossano, E., Rumpf, R. A., Schaefer, E. D., Strohbehn, K., Shelton, R. G., Thompson, R. E., Trombka, J. I. and Williams, B. D. (2007). The X-Ray Spectrometer on the MESSENGER spacecraft. Space Sci. Rev., 131, 393–415.


Solomon, S. C., McNutt, R. L., Jr., Gold, R. E., Acuña, M. H., Baker, D. N., Boynton, W. V., Chapman, C. R., Cheng, A. F., Gloeckler, G., Head, J. W., Krimigis, S. M., McClintock, W. E., Murchie, S. L., Peale, S. J., Phillips, R. J., Robinson, M. S., Slavin, J. A., Smith, D. E., Strom, R. G., Trombka, J. I. and Zuber, M. T. (2001). The MESSENGER mission to Mercury: Scientific objectives and implementation. Planet. Space Sci., 49, 1445–1465. Solomon, S. C., McNutt, R. L., Jr., Gold, R. E. and Domingue, D. L. (2007). MESSENGER mission overview. Space Sci. Rev., 131, 3–39, doi:10.1007/s11214-007-9247-6. Space Science Board (1962). A Review of Space Research: The Report of the Summer Study Conducted under the Auspices of the National Academy of Sciences at the State University of Iowa, Iowa City, Iowa, June 17-August 10, 1962. Washington, DC: National Academies Press, 604 pp. Space Science Board (1966). Space Research: Directions for the Future. Washington, DC: National Academies Press, 653 pp. Space Science Board (1968). Planetary Exploration: 1968–1975. Washington, DC: National Academies Press, 30 pp. Sprague, A. L., Kozlowski, R. W. H., Hunten, D. M., Schneider, N. M., Domingue, D. L., Wells, W. K., Schmitt, W. and Fink, U. (1997). Distribution and abundance of sodium in Mercury’s atmosphere, 1985–1988. Icarus, 129, 506–527, doi:10.1006/icar.1997.5784. Srinivasan, D. K., Perry, M. E., Fielhauer, K. B., Smith, D. E. and Zuber, M. T. (2007). The radio frequency subsystem and radio science on the MESSENGER mission. Space Sci. Rev., 131, 557–571. Starr, R. D., Schriver, D., Nittler, L. R., Weider, S. Z., Byrne, P. K., Ho, G. C., Rhodes, E. A., Schlemm, C. E., II, Solomon, S. C. and Trávníček, P. M. (2012). MESSENGER detection of electroninduced X-ray fluorescence from Mercury’s surface. J. Geophys. Res., 117, E00L02, doi:10.1029/2012je004118. Steering Committee for NASA Technology Roadmaps (2012). NASA Space Technology Roadmaps and Priorities – Restoring NASA’s Technological Edge and Paving the Way for a New Era in Space. Washington, DC: National Academies Press, 122 pp., doi:10.17226/13354. Stern, S. A. and Vilas, F. (1988). Future observations of and missions to Mercury. In Mercury, ed. F. Vilas, C. R. Chapman and M. S. Matthews. Tucson, AZ: University of Arizona Press, pp. 24–36. Toksöz, M. N., Malin, M. C., Albee, A. L., Brandt, J. C., Briggs, G. A., Chapman, C. R., Coroniti, F. V., Duke, M. B., Fanale, F. P., Flinn, E. A., Haskin, L. A., Hayes, J. M., Johnson, T. V., Kaula, W. M., Masursky, H., McCord, T. B., Prinn, R., Schopf, J. W., Sonett, C. P., Stewart, A. I., Trombka, J. I., Wood, J. A. and Young, R. E. (1977). Report of the Terrestrial Bodies Science Working Group. Report NASA-CR-155189, Publication 77–51. Pasadena, CA: Jet Propulsion Laboratory, 33 pp. Ward, W. R., Colombo, G. and Franklin, F. A. (1976). Secular resonance, solar spin down, and the orbit of Mercury. Icarus, 28, 441–452, doi:10.1016/0019-1035(76)90117-2. Yen, C.-W. (1985). Ballistic Mercury orbiter mission via Venus and Mercury gravity assists. Astrodynamics Specialist Conference, paper AAS-85-346. Vail, CO, 15 pp. Yen, C.-W. (1989). Ballistic Mercury orbiter mission via Venus and Mercury gravity assists. J. Astron. Sci., 37, 417–432. Yoshikawa, I., Korablev, O., Kameda, S., Rees, D., Nozawa, H., Okano, S., Gnedykh, V., Kottsov, V., Yoshioka, K., Murakami, G., Ezawa, F. and Cremonese, G. (2010). The Mercury sodium atmospheric spectral imager for the MMO spacecraft of Bepi-Colombo. Planet. Space Sci., 58, 224–237, doi:10.1016/j.pss.2008.07.008.


BepiColombo, 46, 109, 134, 136, 138, 279, 314, 315, 366, 403, 463, 487, 488, 535, 544, 546, 547, 548–562, 563, 564, 565 BELA. See BepiColombo: BepiColombo Laser Altimeter BepiColombo Laser Altimeter, 554, 557, 558 gravity assists, 555 gyroscope, 556 HGA. See BepiColombo: high-gain antenna high-gain antenna, 556, 560 ISA. See BepiColombo: Italian Spring Accelerometer Italian Spring Accelerometer, 549, 554, 557, 558 Magnetospheric Orbiter Sunshield and Interface, 552, 553, 555, 560 MDM. See BepiColombo: Mercury Dust Monitor Mercury Dust Monitor, 554, 560–561 Mercury flybys, 555 Mercury Gamma-ray and Neutron Spectrometer, 554, 558 Mercury Imaging X-ray Spectrometer, 558 Mercury Magnetospheric Orbiter, 552, 553, 554, 555, 556, 557, 559–561 Mercury Orbiter Radio Science Experiment, 554, 556–558 Mercury Planetary Orbiter, 366, 549, 550, 551, 552, 553, 554, 555, 556–559, 560, 562 Mercury Plasma Particle Experiment, 554, 561 Mercury Sodium Atmospheric Spectral Imager, 554, 561 Mercury Thermal Infrared Spectrometer, 366, 554, 557–558 Mercury Transfer Module, 552, 553, 555, 561–562 MERTIS. See BepiColombo: Mercury Thermal Infrared Spectrometer MGNS. See BepiColombo: Mercury Gamma-ray and Neutron Spectrometer MMO. See BepiColombo: Mercury Magnetospheric Orbiter MMO Magnetometer, 554, 561 MMO-MAG. See BepiColombo: MMO Magnetometer MORE. See BepiColombo: Mercury Orbiter Radio Science Experiment MOSIF. See BepiColombo: Magnetospheric Orbiter Sunshield and Interface MPO. See BepiColombo: Mercury Planetary Orbiter MPO Magnetometer, 554, 557 MPO periapsis, 549 MPO-MAG. See BepiColombo: MMO Magnetometer MPPE. See BepiColombo: Mercury Plasma Particle Experiment MSASI. See BepiColombo: Mercury Sodium Atmospheric Spectral Imager MTM. See BepiColombo: Mercury Transfer Module MTM thrusters, 561 perihelion, 555 PHEBUS. See BepiColombo: Probing of Hermean Exosphere by Ultraviolet Spectroscopy Plasma Wave Instrument, 556, 559, 561

253 Mathilde, 196 2P/Encke, 392 4 Vesta, 195, 196, 350 433 Eros, 195, 196, 339 activation energy, 409, 412 adiabat, 38 adiabatic decompression melting, 38, 60, 168, 186 adiabatic gradient, 96 admittance, 64, 65, 74, 271 aerodynamic fractionation, 507, 509 Airy isostasy, 64 Al. See aluminum Al exosphere. See aluminum exosphere albedo, 192, 198 compared with other bodies, 196 Alfvén Mach number, 430, 433, 442, 463 aluminum, 36, 38, 147, 177, 178–184, 185, 186, 209, 210 aluminum exosphere, 371, 399–400, 403, 423–424 ground-based observations, 423 andesite, 179, 182, 183 Andrade creep function, 100 Andrade rheological model, 100 anorthosite, 30, 210 anticline, 70, 251 Apollo program, 544 apparent depth of compensation, 74 Arecibo Observatory, 346, 347, 348 Ariane 5, 552, 553, 555 asteroid impacts, 217, 225, 232, 347, 365 asteroids, 30, 40, 191, 195, 225, 233, 235, 365, 506–507, 509 E-belt, 239 main belt, 233, 234, 236 near-Earth, 233, 237, 238, 239 size–frequency distribution, 233, 239 Atlas Centaur, 546 aubrites, 40, 498, 523 average radius, 88 basal décollement, 255 basalt, 36, 60, 61, 145, 206, 210, 261, 262, 263, 413 basin tectonics, 268–270 BCFDs. See bright crater-floor deposits BDT. See brittle–ductile transition bencubbinites, 39, 43, 184, 185, 499, 501, 506 bending moment, 65 bending stress, 64


Index Probing of Hermean Exosphere by Ultraviolet Spectroscopy, 554, 558, 559 propellant, 556, 560 PWI. See BepiColombo: Plasma Wave Instrument reaction wheels, 556 Search for Exosphere Refilling and Emitted Neutral Abundances, 554, 558–559 SERENA. See BepiColombo: Search for Exosphere Refilling and Emitted Neutral Abundances SIMBIO-SYS. See BepiColombo: Spectrometers and Imagers for MPO BepiColombo-Integrated Observatory SYStem SIXS. See BepiColombo: Solar Intensity X-ray Spectrometer solar arrays, 556, 562 Solar Intensity X-ray Spectrometer, 554, 559 Spectrometers and Imagers for MPO BepiColombo-Integrated Observatory SYStem, 554, 558, 559 star trackers, 556 thrusters, 556, 560 Venus flybys, 555 biannual average temperature, 355 biannual maximum surface temperature, 355 bipropellant, 561 Birch–Murnaghan equation of state, 96 bombardment history, 238, 241 boninite, 36, 145, 180, 181, 182, 183, 206, 307, 522 Bouguer correction, 61 Bouguer disturbance, 61 boundary-normal coordinates, 467 bow shock, 430, 431, 432–434, 441–442, 443, 455, 463, 466, 470, 474, 481, 487 quasi-parallel, 481, 482 standoff distance, 442 bright crater-floor deposits, 324–326 brittle deformation, 261–263 brittle–ductile transition, 62, 263 broad rises, 17 bulk composition, 43–45, 497, 499, 517 bulk density, 88, 92, 95 Byerlee’s law, 262 C. See carbon Ca. See calcium Ca exosphere. See calcium exosphere calcium, 33, 36, 38, 147, 177, 178–184, 185, 186, 206, 209, 210, 336–337, 339, 498 calcium exosphere, 371, 390–392, 402, 419–421 dawn enhancement, 392, 402, 419, 420 ground-based observations, 390, 419, 421 MESSENGER observations, 390–392 seasonal variation, 392, 410, 419 source, 410, 417 tail, 392, 419 temperature, 390, 392 calcium sulfide, 208, 337, 339, 340, 341 calderas, 299 Callisto, 338 Calorian period, 157, 159, 166, 169, 238, 272, 310, 312 Calorian System, 157, 159, 166–167 Caloris exterior plains. See circum-Caloris plains Caloris interior plains, 17, 136, 150, 152, 153, 200, 205, 206, 210, 222, 223, 229, 237, 238, 520 canali, 296 Cape Canaveral Air Force Station, 548


carbon, 34, 37, 180, 184, 185, 191, 206–207, 209, 336–337, 339, 340, 499, 500, 508 carbonaceous chondrites, 168, 497, 501 Cassini state, 86–87, 90, 103, 524 cavi, 326 cavus, 326 CB chondrites. See bencubbinites center of figure, 56 center of mass, 56 CH chondrites. See bencubbinites CH4. See methane Chamberlain model, 372, 418, 421 Chapman–Ferraro currents. See magnetopause: current charged particle environment, 18, 430–455, 461–488 chemical convection, 114 chemical sputtering, 379 chlorine, 33, 34, 177, 179, 500, 505 Christiansen frequencies, 557 chromium, 36, 177, 179, 182, 183, 184 CIPW norm, 177–178, 180, 182, 183 circum-Caloris plains, 78, 136, 150, 152, 169, 201, 205, 221, 222, 224, 229, 232, 233, 235, 236, 237, 238, 254, 269, 292, 294, 298, 303, 310 Cl. See chlorine closest-approach altitude, 116, 117 CMB. See core–mantle boundary CME. See coronal mass ejection cold poles, 74, 349 cold traps, 346 collisional erosion, 505 collisions, 501, 503, 505, 506 Colombo, Giuseppe (Bepi), 551 comet impacts, 217, 225, 347, 365 comets, 225, 235 Committee on Planetary and Lunar Exploration, 546, 547, 564, 565 COMPLEX. See Committee on Planetary and Lunar Exploration complex craters. See craters: complex compositional buoyancy, 533 condensation, 44, 501, 506, 508 convecting mantle, 65 convection, 115 core, 3, 14–15, 114, 115, 117, 124, 129, 130, 133, 250, 472, 526–527, 549, 554, 558 convection, 114, 533–534 cooling, 533 density, 93, 94, 97 dynamo, 135, 138, 524, 528 dynamo field, 114, 116, 117, 120, 124, 126, 130, 132, 133, 134, 135, 138 dynamo models, 125, 138 evolution, 533–534, 535–537 field structure, 117, 126, 135 induced fields, 126–129 induction, 126 radius, 93, 96, 98, 99, 104, 108, 115, 116, 128, 129, 138, 549 core composition, 39–43, 499, 523 carbon, 41, 523, 534 hydrogen, 40–41 iron, 94, 115 light elements, 115 oxygen, 41 phosphorus, 40, 523 potassium, uranium, thorium, 43



core composition (cont.) silicon, 41, 42, 94, 523, 534 sulfur, 41, 42, 94, 523, 533 core–mantle boundary, 38, 61, 97, 114, 115, 116, 126, 129, 131, 133, 134, 135, 136, 270, 525 coronal mass ejection, 463, 466, 472, 474, 487 correlation spectra, 65 Cosmic Vision Programme, 552 Coulomb criterion, 262 Cr. See chromium crater degradation, 225, 227, 232, 235 classification system, 154–157 state, 154–157, 311 crater ejecta, 221, 230, 231, 235, 550 blanket, 224, 225, 226, 232, 234 deposits, 223, 224, 225, 226, 229, 230 morphology, 227–229 rays, 168, 198, 224, 225, 227, 229, 230, 332 crater obliteration, 234 crater rays. See crater ejecta: rays crater retention age, 520 crater saturation, 231, 234, 235, 236, 238 cratered plains structures, 253 crater-related structures, 253 craters areal density, 145–151, 156, 157, 159, 161, 167, 168, 169, 222, 223, 224, 231, 232, 234, 236, 237, 238 buried. See ghost craters central peak, 218, 219, 220, 221, 224 class 1, 232 complex, 218, 219, 220, 221, 225, 226, 227, 229, 241, 348, 349 crater chains, 223, 230, 301, 311 depth, 226 depth-to-diameter ratio, 219, 226–227, 229 elliptical, 229 floor area, 219 floor uplift, 220 polygonal, 229 population 1, 232, 233, 239 population 2, 232, 233, 239 primary, 225 secondary, 146, 159, 217, 224, 225, 226, 227, 228, 229, 230, 232, 234, 235, 236, 238, 301, 337 SFD. See craters: size–frequency distribution simple, 220, 226, 227, 346, 348, 349, 355 size–frequency distribution, 150, 156, 225, 226, 231, 232, 233, 234, 235, 236, 237, 238, 239, 292, 309 tectonics, 226 volcanic fill, 219, 221, 226, 229 crater-to-basin transition, 218–220 creep strength, 63 creep stress, 78 C-rich condensation, 507, 509 CrMB. See crust–mantle boundary cross-polar-cap potential, 470 crust, 60, 79, 114, 117, 129, 130 crustal density, 93, 104, 518 crustal electrical conductivity, 138 crustal formation, 168–170, 517, 534 crustal magnetic fields, 23, 115, 116, 117, 120, 126, 130–132, 138, 549 crustal magnetization, 115, 117, 130, 133, 136–138, 536 crustal stratigraphy, 305 crustal structure, 206, 210 crustal thickness, 14, 60–62, 66, 67, 79, 93, 136, 518–519, 524

crust–mantle boundary, 60, 65, 66, 79, 258, 263 crust–mantle density contrast, 61 Curie temperature, 137 Curiosity rover, 562, 564 currents Birkeland, 22, 114, 117, 119, 126, 129, 134, 138, 431, 444–446, 447, 448, 455, 478, 549, 551 cross-tail, 431, 435, 443–444, 446–447, 452, 453, 475 magnetopause. See magnetopause: current magnetotail. See magnetotail: current cusp. See magnetospheric cusp cusp filaments, 465, 466, 469–470, 487 dark spots, 326, 330, 334 dayside boundary layer, 440–441, 448 DC. See degree of compensation Deep Space Network, 56 deformational history, 272–276 degree of compensation, 77 degree-2 geoid, 70–79 degree-2 shape, 70–79 Deimos, 196 DEM. See digital elevation model depth extent of faulting, 65, 78 desorption, 412, 413, 425 diamagnetic depressions, 437, 438 digital elevation model, 55, 90, 250, 288, 355, 557 digital terrain model. See digital elevation model dike propagation, 270 dikes, 221, 301, 313 diopside, 183, 185 dipolarization, 476, 479 events, 477, 478 front, 476, 478, 486, 487 dipole axial, 118, 128, 134 azimuth, 121 equivalent source, 120, 131, 133, 134 field, 117, 119, 120, 121, 125, 129, 137 induction, 129 moment, 115, 120, 121, 123, 124, 125, 128, 430, 431, 446, 447, 453 offset, 123, 125, 431, 432, 438, 444, 446, 447, 453, 454–455 origin, 116 planetary, 121 planetocentric, 118 structure, 120 tilt, 120, 121, 123, 125, 443 direct trajectory, 545 displacement–length scaling, 250, 264 dissociative ionization, 421, 424 djerfisherite, 508 Doppler, 556 downlink, 548, 556 DSN. See Deep Space Network DTM. See digital terrain model ductile deformation, 263 ductile strength, 263 Dungey cycle, 444, 452, 463, 464, 475, 478 dynamic compliance, 100 dynamic pressure, 66 dynamic recrystallization, 263 dynamic viscosity, 63, 100

Index dynamo, 108, 114, 115, 133, 135, 136, 137 field, 115, 136, 137, 138 models, 115, 133, 135, 138, 525 origin, 120 self-sustaining, 114 thermoelectric, 116 early bombardment, 217, 234 early crust, 160–163, 168–170, 205, 516–518 East Kaibab monocline, 251 effective elastic lithosphere, 63 effusive volcanism, 166, 221, 225, 232, 236, 309–311, 519 EID. See electron impact dissociation elastic lithosphere, 62, 66, 76, 78 thickness, 62, 65, 79 elastic–plastic lithosphere, 69 electrical conductance, 129 electrical conductivity, 114, 115, 117, 126, 129, 130, 136, 138 electrical conductivity structure, 126–130 electromagnetic wave. See plasma wave electron impact dissociation, 420 electron reflectometry, 124 electron-stimulated desorption, 410, 411, 420, 421 elemental abundance maps, 32, 34, 164, 178 elemental abundances. See surface composition emergence angle, 192, 197 Encke dust stream, 410, 421, 422, 424 energetic electron bursts, 451, 476, 484–487 energetic electrons, 14, 18, 22, 451–452, 461, 486, 551, 558, 559 energetic neutral atoms, 559, 561 Energetic Particle and Plasma Spectrometer, 8, 372 energetic particle bursts. See energetic electron bursts Energetic Particle Spectrometer, 8, 451, 461, 476, 484 energetic particles, 548, 550, 554, 561 enstatite, 38, 94, 97, 182, 186 enstatite chondrites, 36, 39, 43, 136, 137, 178, 184, 498, 499, 501, 506, 508, 509 EOS. See equation of state epithermal neutrons, 31, 32, 350, 351, 352 EPPS. See Energetic Particle and Plasma Spectrometer EPS. See Energetic Particle Spectrometer equation of state, 94, 95, 531 equator geographic, 118, 119, 122, 123, 125, 133 magnetic, 121, 122, 123, 125, 126, 431, 436, 443, 447, 448, 450, 451, 452, 454, 455 magnetic dip, 122, 123 ER. See electron reflectometry ESA. See European Space Agency escarpments, 249 ESD. See electron-stimulated desorption Europa, 338 Europe spaceport, 555 European Space Agency, 544, 547, 551–552, 563 exobase, 372 exosphere, 3–4, 15, 17–18, 30, 371–403, 407–425, 461, 548, 549, 550, 552, 554, 558, 559, 560, 561, 562 loss processes, 407, 413–415, 421 observational techniques, 372–378 source processes, 407–413, 416, 417, 419, 424 tail, 413, 414 exosphere species, 378–402 aluminum. See aluminum exosphere calcium. See calcium exosphere


helium, 371, 398 hydrogen. See hydrogen exosphere ionized calcium, 371, 401, 403, 423–424 iron, lithium, silicon, 402 magnesium. See magnesium exosphere manganese, 371, 401, 403, 423–424 oxygen, 371, 399, 403, 424 potassium. See potassium exosphere sodium. See sodium exosphere explosive volcanism, 13, 15, 167, 277, 297–299, 311, 312, 520 external fields, 116, 117, 119, 120, 121, 123, 134 extrasolar planets, 502, 508, 509, 563, 565 extreme ultraviolet, 558 failure criteria, 262 far ultraviolet, 364, 558 Fast Imaging Plasma Spectrometer, 8, 372, 436, 437, 438–440, 441, 449, 455, 461, 469, 470, 472, 474, 475, 476 fast neutrons, 31, 32, 36, 176, 182, 183, 350, 351, 521 fault dip angle, 251, 252, 265 fault displacement-gradient fold, 251 fault heave, 265 fault throw, 265 fault-bend fold, 251 fault-propagation fold, 251 Fe. See iron feldspar, 33, 181, 182, 210, 498 Fe–Ni. See iron–nickel FeO. See ferrous iron ferrous iron, 30, 39, 145, 191, 193, 195, 196, 209, 210, 337, 498, 508 FeS. See iron sulfide FeS layer. See iron sulfide layer field-aligned currents. See currents: Birkeland FIPS. See Fast Imaging Plasma Spectrometer flexural stress, 62 flood basalts, 188, 279, 281, 289, 292, 293, 294, 295, 297, 299, 305, 314, 517 flood volcanism, 34, 180, 182, 184, 276–277, 290, 293, 296, 314, 520 floor-fractured craters, 300 flotation crust, 60, 163, 168, 184, 188, 206, 207, 210, 305, 501, 517 fluidized impact ejecta, 287, 303 flux rope, 443, 454, 468, 469–470, 475, 479–480, 486, 487 flux transfer event, 466, 469–470, 474, 479, 487 fO2. See oxygen fugacity fold-and-thrust belt, 17, 69–70, 256 formation, 2, 13, 497–509 forsterite, 38, 39, 97, 186 fossil bulge, 71 fractionation, 497 free precession, 87 free-air gravity anomaly, 58 frictional resistance, 262 FTB. See fold-and-thrust belt FTE. See flux transfer event gabbro, 180, 182, 183, 184 galactic cosmic rays, 31, 350, 363 Gamma-Ray and Neutron Spectrometer, 6–7, 30, 94, 176, 191, 288, 350, 401, 461, 476, 486 Gamma-Ray Spectrometer, 6, 30, 31, 36, 37, 147, 179, 201, 336, 339, 451, 454, 486, 500 anticoincidence shield, 31, 147, 486 Ganymede, 338 gas–surface interaction, 412–413, 417, 418



general relativity, 552 geochemical modeling, 37–39, 201, 206, 207, 209, 498 geochemical terranes, 22, 33, 36, 176–188, 521 Caloris Interior Plains Terrane, 178, 184, 185, 186, 188 High-Magnesium Terrane, 178, 183, 185, 188 Low-Fast Terrane, 178, 183–184, 186 Northern Terrane, 178, 182, 184, 186 geodesy, 85 GEODYN software, 57 geoid, 52, 58, 70 equatorial ellipticity, 73 polar flattening, 71, 73 geoid-to-topography ratio, 62, 518 geological history, 2–3, 13–14, 250, 519–521 geomorphology, 206, 338, 557 ghost craters, 149, 229, 268, 287, 292, 309, 519 giant impact, 41, 44, 503–505, 508 giant planet migration, 233, 234, 238 global contraction, 3, 17, 21, 69, 107, 167, 249, 255, 264–266, 276–277, 278, 311, 313, 520, 522, 527, 549 global evolution, 516–537 Goldstone Deep Space Communications Complex, 90, 346, 347 graben, 221, 222, 223, 224, 226, 229, 230, 249, 251, 258, 268, 301 grain density, 94, 104 Grand Tack model, 501 graphite, 37, 60, 153, 163, 167, 168, 180, 182, 184, 188, 191, 200, 206–207, 209, 210, 305, 336–337, 339, 340, 341, 501, 517 gravitational constant, 88 gravitational parameter, 88 gravitational potential, 56 gravitational torques, 86 gravity anomaly, 58, 59, 68, 271 gravity assist, 545, 546, 547, 564, 565 gravity disturbance, 58 range, 58 gravity field, 56–60, 79, 87, 518 degree strength, 58, 59, 64 equatorial ellipticity, 56, 71 polar flattening, 56, 71 resolution, 58 gravity–shape correlation, 64, 65 GRNS. See Gamma-Ray and Neutron Spectrometer GRS. See Gamma-Ray Spectrometer GTR. See geoid-to-topography ratio gyroradius, 441, 452, 453, 454 H. See hydrogen H exosphere. See hydrogen exosphere hanging wall, 251 harzburgite, 184, 185, 186, 188 He. See helium heat pipes, 556 heat production rate, 101 heat-producing elements, 249, 266, 527, 529, 532 heavily cratered terrain, 37, 39, 40, 101, 104, 131, 161, 180, 182, 226, 227, 232, 233, 234, 235, 236, 237, 238, 253, 302, 303, 311, 336, 499, 516, 517 high-admittance shape, 66, 74 high-energy particles. See energetic particles highly siderophile elements, 217 high-Mg region, 33, 34, 38, 100, 147, 164, 183, 205, 210, 218, 308, 549 high-reflectance red material, 163–164, 166, 168 high-reflectance red plains, 145, 163, 166, 180, 182, 184, 196, 200, 201, 203, 205, 206, 222, 223, 224, 225, 330

high-relief ridge, 251 high-speed stream, 472 high-terrain-bounding structures, 253 hit-and-run collision, 41, 44, 217, 505, 508 Hoek–Brown criterion, 262 hollows, 14, 15, 17, 19, 35, 37, 167, 199, 200, 201, 204, 210, 217, 222, 224, 225, 226, 311, 324–341, 549, 550, 562, 563, 564 color and spectral properties, 335–336 composition, 336–337, 339 definition, 326 evidence for young ages, 332–334 formation, 207–208, 224, 338–341 formation rate, 334 geographic distribution and compositional affinities, 328–332 geological setting, 326 planetary analogs, 338–341 sizes, shapes, depths, 326–327 texture, 335 homologous temperature, 102 horizontal shortening, 249 hot poles, 62, 74, 349 hot-pole longitudes, 328 HRP. See high-reflectance red plains Hubble Space Telescope, 545 hydrazine, 557 hydrogen exosphere, 371, 396–398, 403 temperature, 396 hydrostatic equilibrium, 71, 73 ICAG. See Inter-Agency Consultative Group icy surfaces, 338 illumination bias, 256, 522 ilmenite, 31, 192, 200, 337 IMF. See interplanetary magnetic field immiscible liquids, 524 impact basins, 70, 147, 151, 152, 154–160, 185, 217–225, 268–270, 303 distribution, 218 ejecta, 222, 223, 225 multi-ring basins, 218, 220, 221, 241 peak-ring basins, 219, 220, 221, 222, 223, 224, 225, 241 protobasins, 220, 221, 226 rim-crest diameter, 219, 221 tectonics, 222, 225 volcanic fill, 223, 224 impact cratering, 217–241 cratering rate, 229 impact velocity, 217, 219, 227, 232, 235, 237, 238 PF. See impact cratering: production function production function, 232, 234, 235, 238, 239 target properties, 217, 218, 226, 227, 229, 234, 236, 237, 239 impact gardening, 358 impact heating, 532 impact melt, 156, 157, 167, 199, 218, 219, 220, 221, 222, 223, 224, 225, 230–231, 340 impact vaporization, 384, 385, 390, 392, 396, 403, 407, 409–410, 411, 413, 416, 417, 418, 421, 422, 423, 424 impactor size–frequency distribution, 233 IMPs. See Moon: irregular mare patches in situ modulus of deformation, 261 incidence angle, 192, 197 induced magnetic fields, 22, 524 induced magnetization, 136 induction, 115, 121, 136, 472, 473, 487

Index induction currents, 464, 557 infrared spectrometer, 549 injections energetic electron, 454 plasma, 443 inner core, 94, 95, 98, 103–104, 114, 524 density, 103 growth, 115 radius, 104, 133 insolation models, 348, 355 instantaneous Laplace plane, 87 Institute of Space and Astronautical Science, 552, 553 Inter-Agency Consultative Group, 552 intercrater plains, 78, 132, 133, 137, 145–148, 157–168, 200, 205, 232, 233, 234, 236, 287, 301–305, 310, 315, 336, 499, 519 crater density, 145–148 origin, 147, 169 reflectance, 147 roughness, 146 thickness, 163 interior outgassing, 364 intermediate plains, 163, 200, 201, 203, 205, 206, 209 intermediate plains stratigraphic unit, 150, 151 internal fields, 126, 130 internal structure, 15, 52–109, 114, 117, 499, 523–524 International Mercury Exploration Mission, 552 interplanetary dust, 347, 392, 410, 421, 422, 508, 552, 554, 560 interplanetary magnetic field, 114, 430, 432, 435, 437, 441, 442, 443, 444, 452, 455, 463, 469, 472, 474, 551 interplanetary shock, 464 intrusive magmatism, 298 ion sputtering, 379, 390, 470 ion-enhanced diffusion, 379 IP. See intermediate plains iron, 31, 36, 94, 147, 169, 177, 178–184, 192, 200, 201, 205, 207, 209, 210, 497, 498 alloys, 136 minerals, 137 multidomain, 137 partitioning, 136 iron meteorites, 40, 499, 506 iron snow, 115, 136, 525, 533 iron sulfide, 136, 206, 207, 339, 499 iron sulfide layer, 68, 97, 128, 524 iron–nickel, 40 iron–wüstite buffer, 36, 177 ISAS. See Institute of Space and Astronautical Science isostasy, 52, 62, 73 isostatic compensation, 79, 257, 518 isothermal bulk modulus, 96 Japan Aerospace Exploration Agency, 544, 547, 548, 551, 552, 563 Jet Propulsion Laboratory, 547, 551 joints, 259 JPL. See Jet Propulsion Laboratory K. See potassium K exosphere. See potassium exosphere Ka-band, 556 Kaula rule, 57 Kelvin–Helmholtz instability, 441, 449 waves, 441, 471–472, 487, 488 KH. See Kelvin–Helmholtz


kinetic escape, 415, 416 kīpukas, 296 komatiite, 36, 60, 145, 206, 307, 340 KREEP, 33, 35, 166 KT14. See magnetic field model: KT14 Kuiperian period, 157, 167, 230, 238, 239, 240, 241, 273, 310, 332 Kuiperian System, 157, 159, 167, 332 lag deposit, 339, 340, 341, 346, 355, 359 Laplace plane, 86 large igneous province, 289, 293 laser altimetry, 53–54 late heavy bombardment, 225, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241, 253, 312, 314, 517 lava flow fronts, 293 LBP. See low-reflectance blue plains Legendre polynomial, 53 LHB. See late heavy bombardment lherzolite, 186, 188, 499 libration, 549, 558, 559 88-day, 86 forced, 562 libration amplitude, 86, 89, 91 limb profile, 55, 375, 385 limb scans, 375, 393, 416, 419, 420, 422 dayside, 416, 418, 419, 421, 424 south pole, 416 Liouville’s theorem, 421 liquid immiscibility, 95 liquidus temperature, 177, 180, 181, 182, 183, 184, 185, 186 lithosphere, 62–63, 72, 260–264, 522 lithospheric folding, 69, 271 lithospheric loading, 249, 269 lithospheric state of stress, 262 lithospheric strength envelope, 262 lobate scarp, 62, 65, 217, 221, 223, 251, 252, 253, 311, 520, 521, 522, 523 long-term librations, 91 long-wavelength shape, 67, 77 long-wavelength topographic undulations, 68, 223, 259–260, 271, 275–276 low-latitude boundary layer, 471 low-reflectance blue plains, 200, 201, 203, 205, 206, 330, 339, 340 low-reflectance material, 14, 23, 37, 60, 145, 150, 161, 164, 166, 168–170, 185, 200, 203, 205, 206, 222, 224, 225, 305, 329, 330, 336, 339, 340, 500, 517 composition and formation, 206–207 spectral heterogeneity, 205 LRM. See low-reflectance material Lunar Crater Observation and Sensing Satellite, 363 Lunar Reconnaissance Orbiter Camera, 146 Lyman alpha photodissociation, 362 Mach number, 466 MAG. See Magnetometer magma, 33, 34, 38, 166–170, 184, 186, 209–210, 276–278, 295, 299–301, 307–308, 312–314, 337, 340, 364–365, 498–499, 517, 521, 528, 532, 535, 549 magma ascent, 312–314 magma buoyancy, 60, 518 magma conduits, 277 magma ocean, 37, 60, 163, 168, 184–185, 186, 188, 207, 305, 517 magmatic volatiles, 34, 208–210, 308, 332, 340, 341



magmatism, 137 magnesium, 33, 36, 38, 147, 164, 177, 178–184, 186, 204, 206, 209, 210, 336–337, 339, 499 magnesium exosphere, 371, 392–396, 402, 421–423 morning enhancement, 394, 396, 402 tail, 421, 422 temperature, 393 magnesium sulfide, 208, 336, 337, 339, 340, 341 magnetic carriers, 136 magnetic diffusion time, 135 magnetic dipole, 3, 14 magnetic erosion, 507, 509 magnetic field, 3, 14, 108, 430–455, 461, 524–526, 545, 546, 548, 549, 551, 552, 553, 554, 561, 563, 564 global, 114, 116, 124, 135 planetary, 554, 557 spherical harmonic descriptions, 164 magnetic field lines, 114 magnetic field model, 438, 446–448 KT14, 446, 447–448 paraboloid, 438, 446–448, 449, 455 scaled Earth, 455 magnetic field modeling, 120, 132 magnetic quadrupole, 525 magnetic reconnection, 14, 22, 117, 126, 430, 432, 435, 438, 443, 444, 447, 449, 452, 453, 454, 463–464, 465, 468, 469, 470, 472–474, 475, 476–480, 486, 487–488 magnetotail, 453 rate, 443, 464, 466, 467, 468 X-lines, 435, 453, 469, 473, 475, 476, 479, 480, 486 magnetic remanence, 114 magnetic residuals, 121, 123, 126, 127, 444, 445, 447, 448 magnetization, 114, 115, 116, 120 Magnetometer, 7, 117, 118, 123, 124, 126, 128, 402, 433, 436, 438, 439, 441, 445, 446, 461, 545, 553, 557, 561 magnetopause, 114, 117, 118, 119, 121, 122, 125, 126, 128, 133, 430, 431, 432–434, 437, 438, 440, 441, 442, 443, 444, 447, 448, 449, 452, 453, 454, 455, 463–464, 466, 467, 470, 471, 472, 553 current, 114, 116, 119, 120, 432, 438, 444, 446, 447, 469, 487 field, 119, 123, 124, 128, 129 flaring, 435 shape, 119, 121, 446, 447 standoff distance, 433, 447 subsolar, 126, 128, 473 subsolar distance, 116 magnetosheath, 430, 432, 437, 438, 441, 442, 461, 463, 464, 469, 470, 471, 472, 474, 481, 482, 487 magnetic field, 432, 438, 441, 443 plasma, 437, 438, 452 plasma beta, 430, 442 thermal pressure, 435 thickness, 433 magnetosphere, 22–23, 114, 116, 117, 118, 119, 120, 121, 122, 123, 124, 126, 128, 133, 138, 430–455, 461–488, 546, 548, 549, 550–551, 552, 553, 554, 558, 559, 561, 562, 563 disappearing dayside, 474 magnetospheric activity index, 120 magnetospheric convection, 487 magnetospheric current systems, 431, 446, 447, 455 magnetospheric cusp, 124, 126, 127, 133, 134, 437, 438–441, 443, 444, 448, 450, 452, 453, 455, 464, 466, 470, 487 magnetospheric polar cap, 438, 439, 440, 452, 455 magnetotail, 118, 121, 122, 430, 435, 439, 444, 446, 447, 452, 453, 454, 463, 464, 466, 469, 470, 475, 476–480, 482, 486, 487, 551, 553

current, 119, 120, 443, 444, 447 field, 123, 124 flaring, 433 loading–unloading, 464, 475, 479–480, 487 lobes, 430, 453 structure, 434, 435 X-line, 435, 453 Malaita anticlines, 257 manganese, 36, 177, 179, 182, 183, 184 manganese sulfide, 208 Mansurian period, 157, 159, 167, 230, 238, 239, 240, 273, 310 Mansurian System, 157, 159, 166–167 mantle, 60, 128, 549 conductivity, 129, 136, 138 magnetization, 136 mantle composition, 37–39, 185–186 mantle convection, 60, 69, 76, 105, 186, 188, 249, 270–271, 528–529 mantle overturn, 60 mantle partial melting, 37–39, 60, 61, 62, 78, 79, 94, 102, 164, 166, 169, 170, 185, 186, 187, 188, 305, 517, 518, 527, 530, 531, 535 mantle solidus, 527 mantle stripping, 44, 497, 502, 503–505 Mariner 10, 1, 30, 53, 57, 88, 108, 115, 116, 117, 119, 125, 135, 144, 152, 191, 192, 217, 222, 226, 227, 229, 230, 240, 249–250, 258, 287, 288, 301, 324, 346, 371, 374, 375, 396, 398, 399, 544, 546, 548, 551, 554, 561, 562, 563, 564 Mercury flybys, 115, 116, 117, 125 Venus flyby, 546 Mars, 3, 33, 34, 44, 116, 168, 169, 217, 219, 226, 227, 228, 232, 233, 235, 236, 238, 257, 296, 301, 338, 346, 498, 535, 544, 545, 546, 547, 562, 564, 565 cratering record, 232 craters, 226 degree-2 geoid, 71 degree-2 shape, 71 impactor velocity, 232 moment of inertia, 71 Olympus Mons, 269 rampart craters, 229 Tharsis rise, 71, 271 Mars lander missions, 544, 545, 562 mascons, 14, 56, 58, 69, 549 MASCS. See Mercury Atmospheric and Surface Composition Spectrometer mass, 88, 92 maximum faulting depth, 78 Maxwell rheological model, 100 Maxwell rheology, 63 Maxwell time, 63, 100 Maxwell–Boltzmann distribution, 408, 410 MBAs. See asteroids: main belt MBF. See Mercury body-fixed coordinates MDIS. See Mercury Dual Imaging System mean radius, 53 mechanical equilibrium, 64 mechanical erosion, 296 mechanical lithosphere, 63 thickness, 63 megaregolith, 163, 225, 227, 237 membrane stress, 62, 64 MeO SWT. See Mercury Orbiter Science Working Team Mercury Atmospheric and Surface Composition Spectrometer, 8, 191, 335, 371, 375, 385, 407, 420

Index Mercury body-fixed coordinates, 116, 117, 120, 122, 123, 126, 127, 130, 134, 135 Mercury Dual Imaging System, 6, 55, 60, 144, 191, 193–195, 220, 227, 236, 240, 250, 288, 324, 335, 347 Mercury Exploration Working Group, 552 Mercury lander mission, 315, 544, 545, 546, 562, 564, 565 Mercury Laser Altimeter, 7–8, 53, 90, 134, 219, 223, 226, 227, 250, 288, 348, 352 Mercury Orbiter Science Working Team, 547, 548, 554 Mercury sample return mission, 563, 564–565 Mercury solar magnetospheric coordinates, 117, 118, 432, 433, 435, 436, 446, 447, 453, 466 Mercury solar orbital coordinates, 116, 117, 121, 122, 124, 126, 432, 433, 438, 439, 444, 447, 448, 470 MESSENGER mission Earth flyby, 11, 53 first extended mission, 15–18 first extended mission objectives, 15–16 first extended mission project requirements, 16 key scientific questions, 2–4 low-altitude campaigns, 19 Mercury flybys, 12, 54, 116, 117, 120, 193, 198, 324, 371, 376, 381, 384, 391, 392, 548 Mercury orbit insertion, 551 orbit-correction maneuvers, 131 primary mission, 11–12, 53 project requirements, 4 science data acquisition planning, 8–11 science observation performance, 10–11 science planning, 10 scientific objectives, 4 second extended mission, 18–23 second extended mission objectives, 19 second extended mission project requirements, 19 solar radiation pressure, 551 solar sailing, 551 Venus flybys, 11, 548 MESSENGER spacecraft, 4–5 altitude, 119, 120, 121, 127 apoapsis altitude, 118 attitude control, 5 initial orbit, 8, 12 launch, 1, 11, 117, 548 launch vehicle, 4 Mercury orbit insertion, 1, 12 mission operations constraints, 9 orbit, 53, 350 orbit design, 8 orbit period, 8, 12, 16 orbit-correction maneuver, 8, 12, 16, 19 payload, 5–8 periapsis altitude, 12, 16, 19, 53, 117, 118, 120, 121, 127, 130, 131, 133, 258, 548 periapsis latitude, 12, 16, 19, 53, 57 propellant, 548, 551 science observation constraints, 9 solar arrays, 5, 547 sunshade, 5, 551 surface impact, 19, 548 telecommunications system, 5 metal/silicate fractionation, 506–508 metal/silicate ratio, 2, 13, 217, 523 metasomatism, 315 meteoroid impacts, 407, 409, 417, 424


methane, 339 MEWG. See Mercury Exploration Working Group Mg. See magnesium Mg exosphere. See magnesium exosphere micrometeoroid impacts, 365, 407, 409, 410, 413, 417, 419, 420, 421, 424, 550, 559, 560 microphase iron, 205, 337 mineralogical composition, 554, 557, 559 mineralogy, 178–184, 185–186, 307, 550, 564 MLA. See Mercury Laser Altimeter Mn. See manganese MnS. See manganese sulfide model production function, 310, 520 molecular dissociation, 419, 420, 421, 423, 424 moment of inertia. See polar moment of inertia moment of inertia of core, 87 moment of inertia of mantle and crust, 87, 92, 95 monocline, 251 Monte Carlo modeling, 416, 419, 422 Moon, 33, 34, 37, 146, 152, 168, 169, 191, 192, 195, 196, 210, 217, 225, 226, 227, 229, 232, 234, 326, 332, 500, 504, 535 composition, 237 crater areal density, 169, 238 crater size–frequency distribution, 235, 236 cratering rate, 238, 241 cratering record, 232 craters, 221, 225, 226, 234, 241 degree-2 geoid, 71 highlands, 147, 148, 160–163, 192, 204, 234, 238 Imbrium basin, 156, 222 impact basins, 217, 218, 219 impact flux, 217, 235 impact melt, 231 impactor velocity, 232, 237 impactors, 240 irregular mare patches, 337 lunar light plains, 287 lunar polar deposits, 363–364 Mare Australe, 152 mare basalt, 35, 152, 153, 176, 192, 210, 231, 237, 287, 308, 338, 521 maria, 2, 152, 192, 251, 545 non-hydrostatic shape, 72 Orientale basin, 147, 156, 161, 222 origin, 217 peak-ring basins, 219, 220 resurfacing rate, 235 South Pole–Aitken basin, 217 surface chronology, 237, 238, 239, 241 Timocharis crater, 228 Tsiolkovsky basin, 228 MPF. See model production function MSM. See Mercury solar magnetospheric coordinates MSO. See Mercury solar orbital coordinates multiple saturation point, 38, 186 Na. See sodium Na exosphere. See sodium exosphere NAC. See narrow-angle camera nanophase iron, 205, 337 narrow-angle camera, 6, 55, 220, 250, 288, 324 NASA Discovery Program, 547, 563 National Research Council, 544, 563 NE Rachmaninoff pyroclastic deposit, 35, 199, 209, 337, 340 near-infrared, 550, 559



near-surface thermal models, 355 near-ultraviolet, 550, 558 Neutron Spectrometer, 6, 30, 31, 36, 37, 60, 185, 191, 207, 209, 336, 337, 350, 451, 486, 500 neutron spectroscopy, 31–32, 350 NGP. See non-gravitational perturbation Ni. See nickel Nice model, 234, 501 nickel, 39 non-gravitational perturbation, 557 non-hydrostatic shape, 76, 79 non-volatile elements, 37 norite, 36, 180, 182, 184 normative mineralogy, 177, 183, 184 north polar region, 347, 352 northern rise, 55, 67–68, 259, 271, 276, 536 northern smooth plains, 17, 33, 36, 38, 55, 62, 67, 78, 100, 131, 132, 137, 150, 152, 155, 166, 180, 184, 188, 200, 205, 210, 227, 228, 229, 232, 233, 236, 237, 238, 239, 240, 253, 254, 255, 268, 269, 270, 271, 275, 292, 293, 294, 295, 296, 298, 299, 301, 303, 309, 310, 499, 517, 520 northern terrane, 33, 36 NS. See Neutron Spectrometer NSP. See northern smooth plains nuclear spallation reactions, 350 O. See oxygen obliquity, 63, 76, 79, 86, 89, 90–91, 524, 549, 558 OCM. See MESSENGER spacecraft: orbit-correction maneuver Odin-type plains, 292 offset of centers, 56 oldhamite, 38, 498 olivine, 100, 104, 180, 182, 183, 185, 204, 210, 307, 498 OMCT. See oxygen–metal charge transfer opening-mode fractures, 259 optical maturity, 193, 204, 332 orbit capture, 552 orbit determination, 56, 556, 557, 558 orbit precession period, 86 orbit precession rate, 87 orbital eccentricity, 75, 77, 86, 137 orbital inclination, 86 orbital precession, 86, 87 orbit-plane normal, 86 ordinary chondrites, 195, 339, 341, 499, 506 Orion spacecraft, 562 orthopyroxene, 94, 104 outer core, 86, 91, 95, 114, 115, 136, 523, 562 outer core radius, 94, 97, 104 outflow channel systems, 296 oxygen, 36, 37, 177, 178–184 oxygen fugacity, 36, 38, 41, 42, 43, 44, 94, 97, 176, 177, 178, 206, 308, 498–499, 523 oxygen–metal charge transfer, 191, 195, 200, 203, 205, 210, 299, 336 P. See core composition: phosphorus parameterized convection models, 527 partial melting, 38, 60, 184, 185, 186, 188 PDL. See plasma depletion layer pebble accretion, 501 perihelion, 553, 556 permanently shadowed regions, 32, 35, 346, 347, 562, 564, 565 petrologic modeling, 36, 39, 177–178, 337

phase angle, 192, 197, 198, 335 phase-ratio analysis, 197, 198–200, 335, 341 Phobos, 196 photodesorption, 379 photodissociation, 392 photoionization, 413, 414–415, 420, 421, 422 lifetime, 414, 415, 417, 419, 420, 422, 423 rate, 414, 415 photometric model, 192, 197–200 photon-stimulated desorption, 390, 407, 408, 409, 410–411, 412, 413, 416, 417, 418, 419 photophoresis, 507, 509 physical librations, 86, 108, 524 pickup ions, 441, 449 Pioneer missions, 545 pit complexes, 222 pitch angle, 124 pit-floor craters, 299, 300 plagioclase, 180, 181, 182, 183, 184, 185, 186, 205, 307, 522 planet formation theory, 501–502, 507 planetary accretion, 217, 233, 498, 501, 505, 509 planetary embryos, 501, 505 planetary magnetic fields, 114, 116 planetary reorientation, 271–272 plasma, 119, 123, 124, 131, 133, 135, 430, 431, 432, 436, 445, 447, 448–455, 549, 550, 559, 561 beta, 430, 436, 442, 443, 467, 487 composition, 431, 449–452 convection, 550 cusp, 438, 439 density, 441, 442 electrons, 476 flow, 441, 442 heating, 452–454 losses, 454–455 precipitation, 14, 18 pressure, 126, 437, 438 properties, 436 solar wind, 440 sources, 436, 448–449 temperature, 437 thermal pressure, 430, 436 transport, 438, 440, 452–454 plasma depletion layer, 441–443, 467, 487 plasma sheet, 126, 430, 434–437, 443, 444, 448, 450, 452–453, 455, 466, 475, 476, 487 plasma wave, 548, 550, 551, 559, 561 ion-Bernstein, 483 plasmoid, 435, 454 Poisson’s ratio, 62, 261 polar deposits, 3, 15, 22, 346–366, 549, 550, 558, 562, 563, 565 boundaries, 359 epithermal neutron flux, 350 high-reflectance surfaces, 354 illumination conditions, 347–350 imaging, 357–360 insulating layer, 346, 352, 355, 359 low-reflectance surfaces, 353, 355, 359 low-temperature silicates, 347, 350, 361 neutron spectroscopy, 350–352 organic compounds, 355, 357 relative age, 359 sulfur, 347, 350, 361 surface reflectance, 352–354, 358

Index thermal models, 355 thickness, 350, 358, 363 water ice fraction, 363 polar moment of inertia, 57, 86, 87, 92, 95, 103, 524 ponded lavas, 289 potassium, 31, 33, 36, 166, 177, 178, 500, 505, 507 potassium exosphere, 371, 389–390, 403, 411 ground-based observations, 425 Poynting–Robertson effect, 560 Pratt isostasy, 64 precipitation, 454, 455 ion, 438, 439, 440, 455 Preliminary Reference Mercury Model, 85, 105, 107 pressure-release melting, 78 pre-Tolstojan period, 166, 272, 310 pre-Tolstojan System, 157–159, 160–166, 169 primary crust, 60, 170, 305, 518 principal axes, 75 principal component analysis, 203, 325, 336 principal moments of inertia, 86 principal-axis coordinate system, 52, 71 PRMM, 105. See Preliminary Reference Mercury Model propulsion chemical, 546, 547, 552, 553, 556, 564 solar electric, 546, 547, 552, 555, 556, 561, 562, 564, 565 proton reflectometry, 124–125, 134, 135 protoplanetary disks, 502, 508, 509 PSD. See photon-stimulated desorption PSRs. See permanently shadowed regions pyroclastic deposits, 14, 15, 23, 167, 199, 200, 208, 210, 297, 299, 300, 308, 311, 331, 336, 339, 340, 520, 562 pyroclastic volcanism, 34, 36, 37, 167, 208–210, 337, 549, 550 pyroxene, 94, 102, 104, 177, 180, 182, 183, 184, 185, 186, 188, 192, 204, 210, 307, 498 pyrrhotite, 136, 137 quality factor, 100 quartz, 182, 183 radar backscatter, 346, 363 radar circular polarization ratio, 346, 363 radar-bright deposits. See polar deposits radiation pressure, 372, 407, 413–414, 415, 416, 417, 420, 421, 422 radio frequency, 53 radio occultations, 54–55 radio science, 8, 54, 90, 549, 556 radius change, 264–266 ram point, 425 Rayleigh number, 105, 528 red unit, 200, 201, 203, 205, 208–210 spectral heterogeneity, 205 reducing conditions, 36, 38, 41, 43, 45, 46, 94, 176, 184, 206, 209, 337, 497–499, 507, 509, 521, 523 reference ellipsoid, 58 reference sphere, 52, 58 regolith, 162, 166, 168, 197, 225, 226, 229, 235, 237, 324, 352 conductance, 129, 130 conductivity, 129 grain size, 198 porosity, 198 scattering properties, 198 thickness, 130 regolith gardening, 347


relative plot, 231, 232, 233, 234, 235 remanence crustal magnetization, 115, 137 magnetic field, 115, 137 magnetization, 115, 117, 136, 137 shock magnetization, 115 thermal magnetization, 115, 136 viscous magnetization, 115, 136 remanent. See remanence Reststrahlen bands, 557 resurfacing, 145–148, 163–166 impact, 147 volcanic, 147, 164 reverse fault, 251 RF. See radio frequency rheology, 62–63, 73, 99–100 ridges, 249 rotational bulge, 71 rover mission, 546, 564, 565 R-plot. See relative plot S. See sulfur sample return mission, 546 satellites of Mercury, 240–241 satellite-to-satellite tracking, 56 saturation equilibrium, 161, 234 Saturn V, 544, 562 scarp retreat, 340, 341 SciBox, 9–10 secondary craters. See craters: secondary secondary crust, 60, 170, 305 second-degree gravity coefficients, 56, 73, 87, 88 secular resonance, 234 secular variation, 117, 125–126, 133, 134, 135, 138, 526 self-gravitation, 71 SEP. See solar electric propulsion SFD. See craters: size–frequency distribution SH. See spherical harmonics shape, 52, 53–56, 79 compensation, 52, 63–64 dynamic range, 55 equatorial ellipticity, 55 equatorial shape, 54 polar flattening, 55 shape ellipsoid, 70 shear modulus, 63, 261 shield volcano, 297 shortening strain, 265 Si. See silicon silicon, 36, 38, 147, 178–184, 209, 499 sills, 301, 313 simple craters. See craters: simple simple-to-complex crater transition, 226, 227 single-scattering albedo, 198 sinuous rilles, 296 SLS. See Space Launch System smelting, 37 smooth plains, 2, 131, 132, 133, 149–154, 193, 200, 218, 221, 222, 223, 224, 226, 227, 229, 232, 235, 238–239, 241, 287, 309, 519 color, 152–153 crater density, 151 distribution, 152 formation, 223



smooth plains (cont.) tectonic deformation, 149 thickness, 150 smooth plains structures, 253 snow line, 507 sodium, 33, 38, 177, 178–184, 186, 500 sodium exosphere, 371, 379–389, 402, 407, 408, 410, 412, 413, 415–419, 424, 425 ground-based observations, 379–384, 416, 419, 424 MESSENGER observations, 384–389 seasonal variation, 416, 417 spatial distribution, 379, 384–389 tail, 371, 383, 411, 413, 415, 417, 418 temperature, 385 thermalization, 385 velocity distribution, 381 sodium ions, 466, 472, 475, 487 sodium tail. See sodium exosphere: tail solar cycle maximum, 18 solar cycle minimum, 463 solar energetic particles, 469 solar heating, 332, 339, 341 solar oblateness, 558 solar wind, 114, 116, 117, 119, 126, 430, 434, 435, 441, 442, 444, 452, 463, 464, 549, 551, 552, 554, 557, 558, 561 dynamic pressure, 464, 467, 472–474 environment, 117, 126, 431, 432, 437, 441, 442, 448, 450, 455 fast, 431, 432, 448 flow, 430 ions, 561 Mach number, 432, 433, 442 plasma, 436, 437, 439, 448, 449, 450, 453 plasma beta, 442 ram pressure, 114, 117, 126, 138, 430, 432, 434, 435, 442 slow, 431, 432 speed, 117, 436, 437 velocity, 122 solar wind–surface interactions, 364 solid basal layer, 95 SOR. See spin–orbit resonance source-free region, 118 south polar region, 347, 366 southern smooth plains, 240 Soyuz–Fregat, 552 space environment, 550, 561 Space Launch System, 562 Space Science Board, 544–546 space weathering, 19, 126, 159, 168, 191, 205, 210, 332, 336, 340, 438, 455, 464, 549, 550, 562 spectral absorption, 200, 210 0.6-µm feature, 200, 203, 205, 208, 210 1-µm crystal-field absorption, 191, 192, 193, 200–201 spectral properties, 200–205 relationship to composition, 210 spectral reflectance, 37, 145, 191–210, 324, 335, 337, 340, 500, 508 spectral slope, 192, 193, 199, 200, 201, 204, 324, 336, 337 spectral units, 200–201 spatial distribution, 205 spectral variability, 200–205 spectrally red pitted ground, 332, 335, 339 spherical harmonic coefficients, 53, 60, 524 spherical harmonic expansion, 120, 121, 123, 124, 125, 126, 128, 129, 132, 134, 135, 138 spherical harmonics, 52, 87

spin axis, 86 spin axis orientation, 90 spin precession period, 86 spin rate, 91–92 spin–orbit resonance, 63, 72, 76, 79, 86, 107, 266, 267, 347, 551, 565 sputtering, 408, 409, 413, 415, 417, 419, 422, 423, 424 chemical, 407, 412 ion, 407, 421 models, 421 stagnant lid, 169, 527 standardized reflectance, 192, 196, 198, 201 Steering Committee for Space Science, 552 stereo imaging, 55 stereophotogrammetry, 250, 288 Stokes flow, 65 stratigraphic column, 206 strike-slip deformation, 255 sub-isostatic state, 64 sublimation, 338, 341, 550 sublimation degradation, 338 substorm, 464, 470, 475, 477–478, 479, 480, 486, 487 sulfides, 167, 177, 180, 182, 183, 206, 209, 210, 307, 337, 339, 340 sulfur, 31, 33, 34, 36, 38, 94, 177, 178–184, 185, 206, 208, 209, 210, 336–337, 339, 340, 498, 499 super-isostatic state, 64 surface chronology, 236–239, 240 absolute, 225, 237–239, 240, 241 oldest terrains, 238 relative, 236–237 smooth plains, 238–239 surface composition, 13, 17, 22, 23, 32–37, 145, 153, 176–188, 206, 308, 315, 336, 499–500, 517, 521–522, 558, 559 surface reflectance, 168, 180, 352 surface roughness, 198 surface temperature, 63, 74, 75, 115, 137, 264, 346, 522, 529 synchronous rotation, 267 syntaxis, 257 talus, 333, 341 TAS. See total alkalis versus silica diagram TCR. See traveling compression region tectonic grid, 255 tectonics, 3, 17, 21–22, 218, 222, 241, 249–279, 520, 557, 559, 564 extensional structures, 221, 222, 223, 258–260, 275 shortening structures, 221, 222, 250–258, 272–274 tephra, 297 Th. See thorium thermal conductivity, 101, 526 thermal contraction, 269, 526, 530, 531, 534 thermal desorption, 381, 388, 403, 407–409 thermal erosion, 296 thermal evolution. See thermal history thermal expansion coefficient, 96, 526 thermal fracturing, 341 thermal history, 63, 76, 105, 114, 115, 138, 249, 264, 276, 311–312, 527–534 thermal lithosphere, 63 thermal neutrons, 31–32, 36, 60, 145, 207, 209, 299, 305, 350, 500, 517 thermochemical convection, 114 thermochemical evolution models, 527–534 thermoelastic strain, 77 thermoelastic stress, 77, 78, 79 thorium, 31, 33, 36, 500 thrust duplex, 251

Index thrust fault, 251, 293 thrusters, 556 ion, 555, 562 Ti. See titanium tidal bulge, 71 tidal despinning, 69, 77, 107, 255, 266, 278, 522 tidal forcing, 89, 100 tidal potential, 99 Love number, 57, 89, 99 tidal response, 99–103 tilted crater floors, 68, 223, 259, 275 time-stratigraphic system, 154, 157–168 titanium, 31, 36, 177, 179, 182, 183, 184, 206 Tolstojan period, 166, 272, 310 Tolstojan System, 157–159, 163–166 topography, 13, 52, 224 total alkalis versus silica diagram, 178, 180, 181, 182, 183 total macroscopic neutron absorption cross section, 177 trachyandesite, 145 trachyte, 181 traveling compression region, 469, 470, 479, 480 Triton, 338 TRM. See remanence: thermal magnetization troilite. See iron sulfide true polar wander, 522 U. See uranium Ultraviolet and Visible Spectrometer, 8, 191, 193–195, 336, 375, 377, 390, 392, 396, 401, 407, 408, 411, 415, 416, 417, 419, 421, 422, 423, 424, 425 uncompressed density, 497 unrelaxed rigidity, 100 uplink, 556 uranium, 36, 500 UVVS. See Ultraviolet and Visible Spectrometer valles, 260, 276, 295 velocity change, 546, 556 velocity distribution, 372 exosphere, 407, 409, 410, 411, 413, 415, 422, 423 meteoroid, 409 Venus, 168, 170, 219, 232, 256, 270, 296, 498, 547, 555, 565 vergence, 251 Very Large Array, 346 Viking landers, 544 VIRS. See Visible and Infrared Spectrograph viscosity, 63, 100, 101, 177, 180, 181, 182, 183, 184, 307, 527 Visible and Infrared Spectrograph, 8, 191, 193–195, 335 volatile deposits, 353, 558


volatile elements, 13, 15, 32–36, 362, 499, 504, 509, 522, 563 volatile loss, 332, 338–341 volatile organic compounds, 15, 22, 355, 356, 357, 359, 361, 362, 365, 563 volatile phases, 207–208, 338–341 volatile wasting, 550 volatiles crustal, 550 frozen, 550, 554, 562 volcanic history, 238–239, 308–312 volcanic landforms, 153–154 volcanic pits, 337 volcanic vents, 225, 226, 297, 311 volcanism, 3, 13, 17, 105, 115, 147, 168, 218, 221, 222, 224, 231, 233, 241, 276–277, 287–316, 339, 519, 549–550, 557, 559, 564 early, 241 Voyager Program, 544, 562 VRM. See remanence: viscous magnetization vulcanoids, 225, 240–241 WAC. See wide-angle camera water ice, 3, 15, 22, 346, 348, 351, 361, 550, 562, 563, 565 organic synthesis within, 363 source, 361–366 stability, 346 subsurface, 355 surface exposures, 354, 355, 357 thermal sublimation, 346, 362 wehrlite, 184, 185, 186 Weibull distribution, 409, 410, 412, 418 wide-angle camera, 6, 55, 144, 191, 193–195, 220, 240, 250, 288, 324, 336, 347, 357 wrinkle ridge, 221, 223, 224, 229, 230, 251, 311 X-band, 556, 560 X-ray fluorescence, 31, 558 X-Ray Spectrometer, 7, 30, 36, 37, 94, 147, 164, 176, 179, 191, 201, 207, 209, 288, 336, 339, 401, 451, 461, 476, 498 XRS. See X-Ray Spectrometer Yakima fold belt, 257 Yarkovsky effect, 233, 240 yield strength envelope, 62 yield stress, 78 Young’s modulus, 62, 261 YSE. See yield strength envelope Σa. See total macroscopic neutron absorption cross section


Calvino crater, 204, 206 Calypso Rupes, 274 Caral Vallis, 295 Carnegie Rupes, 59, 131, 252, 273 Catuilla Planum, 152 Chao Meng-Fu crater, 55, 346, 347, 348 Chesterton crater, 349, 354, 355, 359 Copland crater, 328 Cunningham crater, 228, 332, 335

Abedin basin, 221, 228, 275 Ahmad Baba basin, 303, 304 Ailey crater, 167, 168, 332 Akutagawa crater, 207 Alver crater, 310 Amaral crater, 160, 230 Andal–Coleridge basin, 218 Aneirin basin, 159, 309 Angkor Vallis, 295 Antoniadi Dorsum, 252, 256, 257 Apārangi Planitia, 150, 151, 152, 272 Apollodorus crater, 221, 260 Atget crater, 261, 276

Dali basin, 291 de Graft crater, 160, 324, 326 Debussy crater, 160, 168, 230, 231 Degas crater, 160, 167, 168, 259, 260, 275, 301, 326, 332 Derzhavin crater, 207 Derzhavin–Sor Juana basin, 218 Desprez crater, 349, 350 Discovery Rupes, 65, 78, 263 Disney crater, 310 Dominici crater, 200, 326, 332, 336 Dostoevskij basin, 159, 218 Duccio crater, 251, 252

Balanchine crater, 324, 332, 333, 334 Balzac crater, 160, 324, 335 Barma crater, 151, 309 Bartók crater, 160 Bashō crater, 160, 205, 208, 332 Beagle Rupes, 251, 252, 255 Bechet crater, 353 Beckett crater, 300 Beethoven basin, 151, 152, 159, 164, 197, 218, 277, 294, 295, 298, 303, 309, 310, 312 Belgica Rupes, 255 Benoit crater, 300 Blossom Rupes, 255, 272 Borealis basin, 55, 70, 207, 218 Borealis Planitia, 152, 253 Boznańska crater, 326, 328, 329 Budh basin, 55, 70, 218 Budh Planitia, 152, 155, 292 Burke crater, 348, 349

Eastman crater, 290 Egonu crater, 349, 357 Eminescu basin, 199, 324, 326, 335, 341 Endeavour Rupes, 256, 257 Ensor crater, 349, 360 Enterprise Rupes, 70, 252, 254, 255, 268, 269, 272 Erté crater, 160 Faulkner crater, 205, 309, 520 Firdousi crater, 203 Fonteyn crater, 220 Fuller crater, 260, 349, 353, 354, 355, 359, 360 Futabatei crater, 160

Cahokia Vallis, 295 Calder crater, 151 Calder–Hodgkins basin, 159, 164, 165 Caloris basin, 17, 34, 36, 55, 65, 66, 67, 68–69, 70, 131, 132, 136, 137, 138, 145, 150, 151, 152, 154, 155, 156, 158, 159, 161, 162, 166, 182, 185, 188, 199, 205, 217, 218, 220, 221–223, 237, 238, 241, 249, 252, 253, 254, 258, 259, 260, 261, 265, 268, 269, 270, 271, 275, 276, 277, 292, 296, 298, 301, 309, 310, 312, 326, 337, 519, 521, 522, 532, 535 Nervo Formation, 156 Odin Formation, 156 Van Eyck Formation, 156, 159 Caloris Montes, 224 Caloris Planitia, 151, 152, 153, 155, 157, 249, 260, 277, 292, 294, 295, 297, 298, 299, 303, 309, 310, 312

Gaudí crater, 303, 304 Gibran basin, 151, 302 Glinka crater, 277 Goethe basin, 55, 70, 150, 159, 218, 229, 260, 294 Hawthorne–Riemenschneider basin, 218 Hodgkins crater, 151 Hokusai crater, 145, 160, 167, 168, 228, 229, 275, 301, 311, 366 Homer basin, 159, 218 Hopper crater, 200, 324, 325, 326, 336 Jokai crater, 151


Index of Place Names Kandinsky crater, 349, 354, 355, 359 Kertész crater, 324, 326, 340 Kofi basin, 295 Kuiper crater, 145, 159, 160, 166, 167, 168, 230, 324, 332 Kunisada basin, 290 Kuniyoshi crater, 277, 311 Kyosai crater, 326 La Dauphine Rupes, 255 Laxness crater, 349, 353, 354, 360 Lennon–Picasso basin, 159, 165 Lermontov basin, 324, 331, 339 Lugus Planitia, 151, 152 Machaut crater, 231 Mansur crater, 159 Martial crater, 273 Matabei crater, 197 Matisse–Repin basin, 218, 253 Mearcair Planitia, 152, 155 Monk crater, 349, 357 Mozart basin, 155, 221, 222, 224–225, 258, 268, 269, 275, 301 Nabokov crater, 163, 164, 165, 203 Odin Planitia, 55, 70, 152, 155, 156 Otaared Planitia, 152, 165 Paestum Vallis, 295 Pantheon Fossae, 221, 223, 258, 260, 269, 270, 275, 299, 301 Papsukkal Planitia, 151, 152, 153 Paramour Rupes, 255 Picasso crater, 165, 203 Praxiteles basin, 311, 339 Prokofiev crater, 55, 349, 350, 353, 354, 355, 357, 358, 359, 362 Pushkin crater, 151 Qiu Ying crater, 353 Rachmaninoff basin, 35, 36, 55, 62, 70, 145, 154, 167, 199, 201, 205, 221, 222, 223–224, 225, 239, 240, 258, 269, 275, 299, 301, 310, 326, 328, 337, 339, 520 Raditladi basin, 55, 70, 155, 220, 221, 222, 224, 225, 239, 240, 258, 269, 275, 301, 324, 326, 331, 333, 334 Raphael basin, 159, 218 Rembrandt basin, 17, 70, 145, 150, 151, 152, 153, 157, 159, 161, 197, 203, 205, 218, 221, 223, 224, 240, 252, 258, 268, 269, 270, 272, 277, 293, 294, 295, 298, 299, 301, 303, 309, 310, 312 Rudaki crater, 201, 287, 309, 310 Rustaveli basin, 331


Sanai basin, 159 Sander crater, 324, 326, 340 Santa Maria Rupes, 263 Sapkota crater, 348, 349 Scarlatti basin, 339 Seuss crater, 160, 326 Shakespeare basin, 131, 133, 159, 218 Sholem Aleichem basin, 151, 207, 333 Sibelius crater, 231 Sihtu Planitia, 151, 152, 164 Simonides crater, 151 Sobkou basin, 55, 70, 131, 218 Sobkou Planitia, 145, 152, 155, 159, 164, 205, 304 Sor Juana crater, 55, 70, 207 Spitteler crater, 160 Stevenson crater, 230 Stieglitz crater, 228, 303, 304 Stilbon Planitia, 152 Strindberg basin, 131, 133, 303, 304 Suisei Planitia, 131, 132, 133, 136, 152, 155 Sveinsdóttir crater, 229, 251, 252 Theophanes crater, 324, 335 Thoreau crater, 151 Timgad Vallis, 295, 297 Tir Planitia, 70, 151, 152, 155, 292 Tolkien crater, 349, 354, 355, 359 Tolstoj basin, 145, 151, 152, 153, 158, 159, 203, 205, 218, 224, 240, 277, 294, 295, 298, 299, 303, 309, 310, 312 Tryggvadóttir crater, 349, 354, 355, 359 Turgenev crater, 131, 133 Turms Planitia, 151, 152 Tyagaraja crater, 14, 160, 201, 203, 324, 331, 335, 339 Utaridi crater, 151 Utaridi Planitia, 152 Van Eyck Formation, 296 Velázquez basin, 220 Victoria Rupes, 131, 256, 257, 311 Vincente crater, 151 Vincente–Yakovlev basin, 218 Vivaldi basin, 163, 164, 256, 326 Vyāsa basin, 55, 70, 159, 218 Warhol crater, 326, 327 Waters crater, 199 Xiao Zhao crater, 326, 327, 332 Zeami crater, 324, 335