Life Sciences and Space Research. 15. Sāo Paulo, S.P., Brazil - June 1974: Proceedings of Open Meetings of Working Groups on Physical Sciences of the Seventeenth Plenary Meeting of COSPAR [Reprint 2021 ed.] 9783112482124, 9783112482117

Citation preview

SPACE R E S E A R C H XV

COSPAR

SPACE RESEARCH XV Proceedings of Open Meetings of Working Groups on Physical Sciences of the Seventeenth Plenary Meeting of C O S P A R Sao Paulo, S. P., Brazil — June 1974

Organized by

T H E C O M M I T T E E ON S P A C E R E S E A R C H -

COSPAR

and

THE " I N S T I T U T O DE P E S Q U I S A S E S P A C I A I S OF B R A Z I L Edited by

M. J . R Y C R O F T

AKADEMIE-VERLAG • B E R L I N 1975

INPE"

Executive Editor: Dr. A. C. Stickland

Library of Congress Catalog Card Number 60-50878

© Akademie-Verlag, Berlin 1975 All Rights Reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photo copying, recording or otherwise, without the prior permission of the Copyright owner. 202 • 100/473/75 Gesamtherstellung: YEB Druckhaus „Maxim Gorki", 74 Altenburg Bestellnummer: 762 127 7 (3059/XV) • LSV: 1435, 1495 Printed in GDR

FOREWORD I t was a matter of great satisfaction to the COSPAR community to have the opportunity of meeting in South America for the second time. The large and impressive efforts made in Brazil to develop the techniques of space research and to use them to the benefit of this large country were known to space scientists, and were an important reason for thankfully accepting the invitation to have the seventeenth Plenary and associated meetings in Brazil. Various scientific activities took place, partly in parallel, during the greater part of the month of June 1974. An innovation for COSPAR was the Workshop on remote sensing, mainly set up for Developing Countries. This event, lasting for two weeks, took place in Sào José dos Campos, the seat of the Brazilian Space Research Institute (INPE), and was most successful. During the Workshop about thirty scientists and engineers from developing countries were trained in the interpretation of Earth pictures obtained from satellites by the technique of remote sensing. Later, they participated with other scientists in a Seminar on the same subject. The deliberations at the Workshop and the Seminar on Space Applications of Direct Interest to Developing Countries have in the meantime been published by INPE, under the editorship of Dr. F. de Mendonga, the host of the COSPAR meeting. The COSPAR meeting was also preceded by the COSPAR-IAU-IUTAM Symposium on Satellite Dynamics. Like those of the three earlier symposia in this series the proceedings of this Symposium will be published by Springer Verlag (FRG). The scientific editor of this colume is Prof. G. E. 0 . Giacaglia. As in past years, the proceedings of the open meetings of COSPAR Working Group 5, on Space Biology, will be published in a separate volume, Life Sciences and Space Research. The scientific editor is Prof. P. H. A. Sneath. The present volume, Space Reseach XV, contains the scientific papers presented at the open meetings of the six other COSPAR Working Groups, together with the three Annual Reviews which are all of great and topical interest, viz. 'the ISAGEX Campaign' by J . Kovalevsky ; 'the International Reference Ionosphere' by K. Rawer, and 'Meteorology of the Stratosphere and Mesosphere" by K. Labitzke. The scientific editor is Dr. M. J. Rycroft. Altogether, the various publications issued after the very successful meetings in Brazil witness in an impressive way the many aspects of the intense and highlevel scientific activity of these meetings. Our sincere thanks go to the various

VI

Foreword

scientific editors named above, to Dr. A. C. Stickland, the executive editor of the present volume, of the volume on Life Sciences and of the Symposia on Satellite Dynamics, and also to the scientists who contributed to the volumes, and the Chairmen who so ably chaired the scientific sessions. We are also grateful to Akademie Verlag, Berlin (DDR) the publisher of the present volume. Cornelis de Jager President of COSPAR.

PREFACE To most of us the future seems unsure. But then it always has been; and we who have seen great changes must have great hopes. John Masefield, Grace Before Ploughing I n his message to the Seventeenth Plenary Meeting of COSPAR, held in Sao Paulo, Brazil, during June 1974, the Secretary-General of the United Nations said: "During the last decade and a half, we have witnessed outstanding scientific feats and rapid developments in space technology, all of which bear testimony to the resourcefulness of the space scientists in achieving elaborate accomplishments. As one of the rallying points for these outstanding scientists, COSPAR has for many years played a decisive role in promoting international scientific collaboration in the peaceful uses of outer space ... At the same time we are conscious of the great need to utilize the tools of space research to meet some of the most pressing problems of our times. I t is in such vital areas as enhancing and conserving our environment, in meeting the worldwide shortage of food, and understanding the problems of world population that space technology, particularly in the fields of communications, meteorology and remote sensing, can be utilized to great advantage". I n this spirit, many scientists from many different countries participated in the Meeting. This volume contains some of the papers in the Annual Reviews of Space Research series, some of the invited review papers, and some of the papers contributed to both the Special Topics and the Latest Significant Results sessions of the seven Working Groups of COSPAR. Dealing with the Physical Sciences, this work is composed of eight main sections: Geodesy Remote sensing Upper atmosphere Ionosphere Magnetosphere Cosmic dust Moon Astronomy

VIII

Preface

Papers on Life Sciences are published in a companion volume. The papers, or just résumés of them, presented at some of the Specialized Symposia held in conjunction with COSPAR are published independently. As the Scientific Editor, I most grateful for all the help with refereeing the papers that I have received from the Chairmen and Reporters of the different Sessions of the Working Groups. I am also deeply indebted to Dr. A. C. Stickland who, with unfailing good humour in the face of adversity, so successfully performs the task of being the Executive Editor of these COSPAR volumes. Both of us thank Prof. C. de Jager, President of COSPAR, Mr. Z. Niemirowicz, Executive Secretary of the COSPAR Secreatariat, and other members of the Secretariat for the assistance that they render us, not only before the Meeting, but also during it and afterwards. M. J . Rycroft

Space Research X V — Akademie-Verlag, Berlin 1975

CONTENTS L I S T Foreword Preface

V VII

Geodesy ISAOEX

(International

Satellite Oeodesy

Experiment)

J . KOVALEVSKY

Results of the " I S A G E X " Campaign

3

A . G . MASSEVICH, N . P . E R P Y L E V a n d T . W . K A S I M E N K O

Some Results on the Arctic-Antarctic Project based on Observations made during 1970-1971

G . KARSKY, J . KOSTELECKY, V . SKOUPY a n d I . SYNEK

The Determination of the Coordinates of Station 1147 at Ondrejov

W . EHRNSPERQER

Geometric Adjustment of Western European Satellite Triangulation (Solution 1974)

17 21 25

New Techniques and Results J . M . MORAN

Geodetic and Astrometric Results of Very Long Baseline Interferometric Measurements of Natural Radio Sources

N . J . MOHANA R A O , G . S . S . N U N E S a n d S . AKANTHAKRXSHNAN

Satellite Range Measurements — Tropospheric Refraction Effects

33 49

The Oeopotential G . BALMINO a n d C H . R E I G B E R

13th-0rder Harmonics in the Geopotential from an Analysis of Four Resonant Satellites

W . BENNIKG

Analysis of Satellite Altimetry Data

53 59

Selenodesy M . MOUTSOULAS

Reference Points for Selenodetic Control

65

E . S . B A R K E R , 0 . CALAME, J . D . MULHOLLAND a n d P . J . S H E L U S

Improved Coordinates for Lunokhod 2 based on Observations from McDonald Observatory

71

X

Contents List

Remote Sensing Mesopause A . A . B U Z N I K O V , K . Y A . K O N D R A T Y E V , A . I . L A Z A R E V a n d O . I . SMOKTY

Optical Characteristics of the Mesopause and t h e Lower Thermosphere on the Nightside of t h e E a r t h

77

Meteorology K . LABITZKE

Review on Investigations in the Field: Meteorology of the Stratosphere Mesosphere

and 81

I . HAUPT a n d U . KATERGIANNAXIS

J.

Local Circulations in the European Area seen b y Weather Satellites and results of a more quantitative Interpretation of these Satellite Data W . W A T E R S , D . H . S T A E L I N , K . F . K U N Z I , R . L . P E T T Y J O H N and R . K . L . PoosMicrowave Remote Sensing of Atmospheric Temperatures from the Nimbus 5 Satellite

109 117

T . H . YONDER HAAR, D . REYNOLDS a n d L . LILIE

Direct Readout Meteorological Satellite D a t a Processing with a Low-Cost, ComputerLinked System

123

Upper Atmosphere Stratospheric and Mesospheric

Meteorology

J . P . BALUTEAU a n d E . BUSSOLETTI

High Resolution Spectra of t h e Stratosphere between 30 and 200 c m - 1

131

V . F . L O G I N O V a n d G . I . SUKHOMAZOVA

Zonal Wind in the Stratosphere and Solar Activity

139

F . G . FINGER a n d M . E . GELMAN

Some Results of t h e WMO (CIMO) Rocketsonde Intercomparisons — Phase I I S . S . GAIGEROV,

B . P . ZAICHIKOV,

M . Y A . KALIKHMAN,

L . M . KOLOMIITSEVA,

. .

143

D. A.

T A R A S E N K O , V . V . F E D O R O V a n d L . Y . SHCHERBAKOVA

Analysis of Large-Scale Processes in t h e Upper Atmosphere based on Global Meridional Wind and Temperature Cross Sections S . S . GAIGEROV,

M . Y A . KALIKHMAN,

V . V . FEDOROV,

B . P . ZAICHIKOV,

151

N . F . NOVI-

KOVA, D . A . T A R A S E N K O a n d L . V . SHCHERBAKOVA

Results of t h e Empirical Investigation of t h e relation between Temperature a n d Wind Variations in t h e Mesosphere

157

R . A . BRITVINA, V . G . KIDIYABOVA, D . A . TARASENKO a n d I . A . SHERBA

Statistical Analysis of Meteorological Parameters in the Stratosphere and Mesosphere based on d a t a from t h e Eastern Meridional Rocket Network

161

Y u . P . KOSHELKOV

Meridional Distribution of Zonal Wind in t h e Upper Atmosphere of t h e Southern Hemisphere

167

A . E . COLE

Periodic Oscillations in Stratosphere and Mesosphere Atmospheric

173

Tides

G . V . GROVES

Propagating Modes of the 24-hourly Atmospheric Tide derived from Natal (6° S) Grenade Experiments a n d Global Barometric Oscillations

181

XI

Contents Liât M . G L A S S a n d J . L . FELLOTJS

The Eight-hourly (ter-diurnal) Component of Atmospheric Tides

191

A . S . B U T K O , I . N . IVANOVA, G . A . K O K E N a n d A . A . K U M I N O V

Some Results on t h e Daytime Measurements of Temperature and Wind in the Mesosphere a n d Lower Thermosphere Major Neutral

199

Constituents

C. WULF-MATHIES, P . BLÜM a n d H . THINKS

Local Composition Changes in t h e Thermosphere a t High Latitudes during Moderate Geomagnetic Conditions

203

P . BLUM, C. WULF-MATHIES a n d H . TBINKS

Interpretation of Local Thermospheric Disturbances of Composition observed b y ESRO 4 in t h e Polar Region

209

G . W . P R Ö L S S , K . H . F R I C K E a n d U . VON Z A H N

Observations during a Magnetic Storm in late October 1973

215

G . SCHMIDTKE, CHR. M Ü N T H E R a n d K . R A W E R

Variations of Atomic Oxygen Densities in the Thermosphere

221

V . V . M I K H N E V I C H , A . A . P O K H U N K O V , E . N . G O L U B E V , YTJ. F . IVANOV a n d S . V . G O R BUNOV

Neutral Atmosphere Variations according t o Measurements made in Tropical and Middle Latitudes

227

Y u . M . ZHUCHENKO, V . A . LIPOVETSKY, L . S . NOVIKOV, V . F . TULINOV, G . F . TULINOV a n d V . M. FEIGIN

Simultaneous Rocket Measurements of Structure Parameters of Polar Thermosphere and Auroral Radiation Minor Neutral

233

Constituents

S. P . P E R O V a n d A . S. RAKHMANOV

Atomic Oxygen Concentration Measurements a t altitudes of 75—95 km

237

T . TOHMATSU a n d N . I W A G A M I

Measurement of Nitric Oxide Distribution in t h e Upper Atmosphere V . N . BALABANOVA,

K . D . BYCHKOVA,

V . N . LEBEDINETS,

V . P . MARTYNENKO

241 and

A . A . POKHUNKOV

Experimental D a t a on Atomic Nitrogen Variations in t h e Upper Atmosphere after Sunset

247

Y . SAHAI, A . DRESCHER, H . LAUCHE a n d N . R . TEIXEIRA

First Results of 6300 Á Nightglow Measurements aboard a Rocket launched from Natal, Brazil

251

A . MONFILS a n d J . C. GERARD

Preliminary Results of Observations of Atmospheric Ultraviolet Twilight Emissions b y the T D 1-A Satellite

257

G . M . MARTYNKEVICH

Preliminary Results on Suprathermal H and He Atoms in the Middle Latitudes Lower Thermosphere during a Magnetic Disturbance Low Latitude

263

Observations

J . N . D E S A I , P . D . B H A V S A R , R . RAGHAVARAO a n d M . S . N A R A Y A N A N

Winds and Diffusion in the Upper Atmosphere observed b y a Sodium Vapour Trail released over Thumba Thermospheric

267

Models

G . M . KEATING, E . J . P R I O R , D . S. MCDOUGAL a n d J . NICHOLSON I I I

A Critical Evaluation of the OGO 6 Helium Model

273

Contents List

X I I

C. WULF-MATHIES, E . J . PRIOR a n d G . M . KEATING

Annual and Semiannual Density Variations in the E a r t h ' s Exosphere

279

W . KÖHNLEIN, H . TRINKS a n d H . VOLLAND

The O to N 2 Density Ratio in the Thermosphere derived from ESRO 4 D a t a

. . .

287

Intereomparison of Different Measuring Techniques in the Upper Atmosphere: The International Reference Ionosphere

295

Ionosphere International

Reference

Ionosphere

K . RAWER

K . SPENNER a n d H . W O L F

Comparison between Electron Density and Temperature during Daytime

321

S. FUKAO a n d K . MAEDA

Daytime Electron Density Profiles of the E and F l Regions

327

Solar Radiation and the Ionosphere P . CHAKRABARTY a n d A . P . M I T R A

Solar EUV Flux Models consistent with Ionospheric Ion Composition Observations

335

G . SCHMIDTKE, K . R A W E R , W . FISCHER a n d C. REBSTOCK

Absolute EUV Photon Fluxes of Aeronomic Interest F Region and Mid-latitude

345 Trough

Y u . A . R O M A N O V S K Y , V . V . KATYUSILTNA a n d V . G . I S T O M I N

Mass Spectrometer Measurements of t h e F 2 Region Ion Composition from the Satellite Cosmos 274

351

M . N . VLASOV a n d Y u . A . R O M A N O V S K Y

On the Intensities of 6300 A, 5577 A and 5200 A Emissions from Ion Composition Measurements in the F 2 Region Electron Density and Temperature

357

Variations

K . SPENNER

Quiet and Disturbed Electron Temperature and Density a t Different Latitudes during Daytime

363

Y u . K . CHASOVITIN a n d N . M . K L Y U E V A

Electron Temperature Variations a t 100—200 km from Probe Measurements

. . .

369

A . DUMBS a n d W . NOACK

Very Low Energy Electron Spectra in the Auroral Zone

379

R . R A G H A V A R A O a n d M . R . SIVABAMAN

Formation of Ionization Ledges in t h e Equatorial Topside Ionosphere Eclipse

385

studies

A . D . DANILOV, U . F . IVANOV, G . S. IVANOV-KIIOLODNY, T . V . KAZATCHEVSKAYA, V . K . SEMENOV, V . V . SELANTIEV, Y u . K . CHASOVITIN a n d V . G . K H R Y U X I N

Measurements of Ionospheric Parameters a t 100—170 km during a total Solar Eclipse Oyroharmonic

393

Resonances

J . BITOT™

Theoretical Interpretation of Gyroharmonic Resonances observed during Rocket Experiments

399

XIII

Contents List Ionospheric

Irregularities

K . B I B L , W . P F I S T E R , B . W . REINTSCH a n d G . S . SALES

Velocities of Small and Medium Scale Ionospheric Irregularities deduced from Doppler and Arrival Angle Measurements V. H . R i o s , A. M . SAUVAGE and J . R . MAKZANO Analysis of Scintillation Phenomena produced over Tucuman, for the 1965—1968 Period Very Low Frequency

405 413

Radiophysics

R . A . USHER a n d M . J . RYCROFT

The Refraction of V L F Waves b y a Sporadic E Layer D. NUNN Rigorous Computation of the Radiation Fields of V L F / E L F Antennas

419 425

Magnetosphere High Latitude

Phenomena

D . E . PAGE a n d K . - P . WENZEL

Review of Selected Scientific Results of H E O S 2 I . Y A . KOVALSKAYA,

M . I . PANASYUK,

433

S. P . RYUMIN,

E . N . SOSKOVETS,

S . K . STOL-

BOUSHKIN, L . V . TVERSKAYA a n d O . V . KHOROSHEVA

Features of the Intensity Variations and Spectrum of Low Energy Protons in the Polar Cusp F . CAMBOU,

O . L . VAISBERG,

H . ESPAGNE,

V . V . TEMNY,

C. D'USTOK

and

455

G . N . ZA-

STENKER

Characteristics of Interplanetary Plasma near the E a r t h observed during t h e Solar Events of August 1972

461

Electric Fields U . FAHLESON, C . - G . FALTHAMMAE a n d A . P E D E R S E N

Electric Field and Plasma Measurements in the Auroral Ionosphere prior to a Magnetic Substorm

471

M . SYLVAIN, D . R O U X , A . BERTHELIER, C. GUERIN a n d F . S. MOZER

Simultaneous Observations of the Motion of Large-scale Electron Density Irregularities and of Ionospheric Electric Field near the Polar Border of the Southern Auroral Zone Magnetic Disturbance

477

Effects

A . S . K O V T Y U K H , M . I . P A N A S Y U K a n d E . N . SOSKOVETS

Strong Pitch-Angle Diffusion of Protons during the Magnetic Storm of 19 March 1973 Active F . CAMBOU,

V . S . DOKOTJKINE,

O . K . NAZARENKO,

Experiments

V . N . IVCHENKO,

A . T . NESMYANOVICH,

485

G . G . MANAGADZE,

A . K H . PYATSI,

V . V . MIGULIK,

R . Z . SAGDEEV

and

I. A.

ZHULIN

The Zarnitza Rocket Experiment on Electron Injection

491

Magnetospheres of Other Planets H . S. BRIDGE,

A . J . LAZARUS,

K . W . OGILVIE,

J . D . SCUDDER,

R . E . HARTLE,

J. R.

A S B R I D G E , S . J . B A M E , W . C . F E L D M A N , G . L . SISOOE a n d C . M . Y E A T E S

Preliminary Report of Results from the Plasma Science Experiment on Mariner 10 .

501

XIV

Contents List

A . MOGRO-CAMPERO, R . W . F I L L I U S a n d C . E . M C I L W A I N

Electrons and Protons in Jupiter's Radiation Belts

521

Cosmic Dust Fluxes of Cosmic Dust DAVID W . HUGHES

Cosmic Dust Influx to the Earth

531

C. L . HEMENWAY, D . S. HALLGREN a n d C. D . TACKETT

Near-Earth Cosmic Dust Fluxes obtained from Skylab Experiments

541

J . M . ALVAREZ, D . H . H U M E S , W . H . K I N A R D a n d R . L . O ' N E A L

The Interplanetary and Near Jovian Dust Environment: Some Experimental Results

549

J . A . M . MCDONNELL a n d O . E . BERG

Bounds for the Interstellar to Solar System Microparticle Flux Ratio over the Mass Range 10" 11 —10" 13 g

555

DAVID W . HUGHES

The Cometary Contribution to Cosmic Dust

565

Zodiacal Light A . - C . LEVASSEUR a n d J . E . BLAMONT

The Distribution of Dust along the Earth's Orbit deduced from Satellite Measurements of Zodiacal Light

573

Optical Scattering P . L É N A , D . H A L L , A . SOUFFLOT a n d Y . VIALA

The Solar Corona as observed during the 30 J u n e 1973 Solar Eclipse on board Concorde 001

579

Moon Abundance of Elements G . E . K O C H A R O V , S . V . VICTOROV, V . P . K O V A L E V , G . A . M A T V E E v a n d V . I . C H E S N O K O V

Chemical Composition Variations of the Lunar Surface in the Contact Zone "MareHighland"

587

Y u . A . S U R K O V , G . M . K O L E S O V a n d F . F . KXRNOZOV

Abundance of Some Elements in the Regolith of the Maria and Continental Regions of the Moon L . L . KASHKAROV,

A . K . LAVRUKHINA,

L . I . GENAEVA,

M . K . ANTOSHIN

and

593

G. V.

S P I VAK

Track Investigations of Lunar Soil returned by Luna 20 Physical A . K . LEONOVICH,

V . V . GROMOV,

601

Properties

P . S . SEMYONOV,

V . N . PENETRIGOV

and

V. V.

SHVARYOV

Luna 16 and 20 Investigations of the Physical and Mechanical Properties of Lunar Soil N . N . K R O U P E N I O , A . G . B A L O , E . G . R U Z S K I I , V . A . L A D Y G H I N , V . Y . CHERKASOV

607

and

V . S. FOMIN

Results of Radar Experiments performed aboard the Luna 19 and 20 Automatic Stations

617

XV

Contents List Magnetic

Field

C . T . RTJSSBLL, P . J . C O L E M A N , J R a n d G . S C H U B E R T

The Lunar Magnetic Field

621 Mars

W . K . HARTMANN

Problems of Martian Paleoclimatology

629

Astronomy Sun G . E . K O C H A R O V , V . S . V I C T O R O V a n d V . I . CHESNOKOV

Investigation of Solar X-rays from the Lunar Surface, carried out on Lunokhod 2 . . Y u . I . GRINEVA,

V. I. KAREV,

V . V . KORNEEV,

V . V . KRTJTOV,

633

S . L . MANDEL'STAM,

U . I . SAFRONOVA, A . M . U R N O V , L . A . V A I N S T E I N a n d I . A . ZHITNEK

Investigation of Solar Flare X-ray Spectra in the Hydrogen-like Fe Ion Region . . I . L . BEIGMAN, Y U . I . GRINEVA, V . V . KORNEEV, V . V . KRUTOV,

637

S . L . MANDEL'STAM,

L . A . VAINSTEIN, B . N . VASILYEV a n d I . A . ZHITNIK

The Solar Subflare X-ray Spectrum

641

J . D . BOHLIN, N . R . SHEELEY a n d R . TOUSEY

Structure of the Sun's Polar Cap a t Wavelengths 240—600 A

651

Ultraviolet A . CUCCHIARO, C . J A M A R a n d D . M A C A U - H E R C O T

Ultraviolet Spectra of t h e Wolf-Rayet Stars H D 50896 and H D 191765 from the T D 1 -A Sky-Survey Telescope

657

X-ray J . H . PARKINSON, J . L . CULHANE, F . J . H A W K I N S a n d P . W . SANFORD

X - R a y Astronomy with Copernicus

663 High energy

K . PINKAU

Gamma R a y Astronomy (0.1 —1000 MeV)

681

C. E . FICHTEL

High Energy Gamma R a y Astronomy

699

R . COWSIK

Geometry of Inverse Compton Gamma-Ray Sources Interstellar

715

medium

J . L . BERTATJX, J . E . B L A M O N T , N . T A B A R I E , N . N . D E M E N T E V A , V . G . K U R T a n d A . S . SMIRNOV

Measurement of the Temperature of the Interstellar Medium around the Heliosphere, obtained with the Mars 7 Interplanetary Probe

721

H . J . FAHR

Thermal Behaviour of t h e Neutral Interstellar Gas within t h e Solar System . . . .

727

Papers presented a t t h e Sao Paulo Meeting 1974 and published elsewhere

733

Index of Authors

735

Space Research XV — Akademie-Verlag, Berlin 1975

R E S U L T S OF T H E " I S A G E X " CAMPAIGN J . KOVAIEVSKY

Groupe de Recherches de Géodesie Spatiale, Meudon, France During the International Satellite Geodetic Experiment (ISAGEX) Campaign, organized in 1971 with the sponsorship of COSPAR, a great number of observations of positions of artificial satellites was gathered by about 50 stations. These data are described and analysed and their value in geodetic and geodynamic research is estimated. Examples of their use are given. Future programmes will benefit from the experience obtained. In conclusion, a synthetic view of the future trends of research in the field is proposed.

1. The "ISAGEX" Campaign In 1969, COSPAR decided to sponsor the largest international observation campaign for geodetic purposes ever organized [1]. This was the International SAtellite Geodesy Experiment (ISAGEX), including almost all countries having some precise satellite tracking instrumentation. The decision to undertake this organization was taken in order to take advantage of several technical developments that could contribute to the improvement of the scientific objectives already achieved in space geodesy. These were: (i) The rapid development of laser ranging techniques to a precision better than 2 metres (before 1970 the actual accuracy was 5—10 metres). (ii) The launching by the French agency ONES, of the first low inclination satellite (14°). The absence of such satellites introduced an important inaccuracy in the system of zonal harmonics, and such a satellite was badly needed. Since PEOLE had laser retroreflectors this was a particularly favourable situation. (iii) The construction of many new photographic satellite cameras, especially the Soviet AFU-75. I t was expected that this campaign would not only permit improvement of existing results in space geodesy (earth potential models, geodetic links between stations) but would also serve as a transition towards the new "Space Earth Physics" era by permitting some geodynamical studies before the launch of specialized satellites. Such an ambitious undertaking had to be carefully planned. I t was put under the overall technical leadership of ONES, the project manager being G. Brachet who prepared the operational plan [2]. An international scientific committee was set up in order to prepare the scientific objectives and distribute the data. The stations were directed by and reported to five sub-centres: 1*

J . KOVALEVSKY

5

Results of the "ISAGEX" Campaign

saturation periods are not equivalent, some satellites being more difficult to observe than others. The best data have been gathered on GEOS-A and GEOS-B, with, for the latter, very good photographic observations of flashes (Fig. 2). Table 1 Satellite Reported Midas BEB BEC GEOS-A Pageos DI C DI D GEOS-B PEOLE Total

Optical passes

Lase passes Useful

Reported

—

—

206 316 485 323

158 264 366 173

628 240 365 1186 493 476 654 886 224

2426

1920

5152

—

76 296 724

-

57 230 672

Useful reduced 211 95 76 439 76 165 136 433 176 1807

Unfortunately PEOLE, visible only from a minority of stations, was not sufficiently observed (Fig. 3). I t is generally accepted that, for dynamical geodesy, one needs two- or three-week periods with, at least, 5 or 6 passes observed every day. Under such constraints, 14 arcs observed during saturation periods are good. Actually, four consecutive months of GEOS-A and a similar observational period of GEOS-B have a sufficient number of observations to permit a treatment during long periods required for geodynamic purposes (Fig. 2).

2. Data Evaluation Observations are subject to many systematic or accidental errors and it is necessary to check all the data and remove the spurious observations. This is done in several steps, all reduction corrections having been applied. As an example, we give the procedure adopted by GRGS for laser observations: (i) For a given pass, a fourth- to sixth-order polynomial is used to fit the observations, and permits the removal of the erroneous data which appear with residuals with respect to the smoothed curves of the order of 100 m. (ii) Then an orbit is computed from the remaining observations with a crude theory (Brouwer analytical theory to J 6 ) and taking certain elements as unknown (e.g. eccentricity, mean anomaly, inclination or the longitude of the node). Residuals of the order of 10 m are then well separated and the corresponding observations can be removed. (iii) After the second iteration the remaining residuals deviate from zero by quantities of the order of the noise. This noise represents the internal consistency of the data. During the ISAGEX programme it was as follows: NASA lasers, 0.50 m (ths same network gave 2 m in

6

J. Kovalevsky Smithsonian Astrophysical Observatory, SAO (USA) Goddard Space Flight Center, GSFC (USA) Astronomical Council of the Academy of Sciences (USSR) Ondrejov Observatory (Czechoslovakia) Centre d'Opération du CNES (France)

The problem of ephemerides was a crucial one. Many of the lasers could measure distances blindly, the satellite not being visible. Therefore, ephemerides precise to about two minutes of arc had to be provided to stations. A large number of stations, including the US and French minitrack systems, eight Astro-Soviet NAFA-25 cameras and most of the stations involved in the main observational programmes, participated in a large programme of quick-look observational data collection. These observations were sent by the fastest channels to GSFC and

Pig. 1. Stations that have participated in ISAGEX: 1, Baker-Nunn cameras; 2, other cameras; 3, lasers; 4, laser and camera.

SAO, and ephemerides were computed [3] and distributed to sub-centres which, in turn, derived the working ephemerides for their stations. Over 50 stations have participated partly or fully in the experiment. This number includes nine laser ranging stations (USA, France, Greece), 16 Baker-Nunn and 14 Soviet AFU-75 cameras (Fig. 1); for details see [4]. After two months of a pre-ISAGEX campaign (September—November 1970) to test the information flow and other organizational problems, the actual ISAGEX Campaign started on 5 January 1971 and lasted 34 weeks (until 31 August 1971). The targets of observation were the seven satellites equipped with laser retroreflectors and two large passive satellites (Midas and Pageos). Seven three-week saturation periods were chosen during which three satellites were to be observed whenever possible by all stations. But many stations continued to observe in between. Table 1 gives the number of passes observed. To each pass correspond several photographic positions and from five to three hundred individual distances. However, these observations are not evenly distributed in time. Even various

Results of the "ISAGEX" Campaign

7

1969); an example is shown in Fig. 4; GRGS lasers, 1.50 m; SAO lasers (including Greece), 2 m. This procedure is not applicable to passes with a small number of observations as is the case for all optically observed passes and for some laser passes: it is then !2.

12 . .

'D-

10-.

BGB

8-

8-

6-

6-

4.

PEOLE

4"•

2

24

LLLLL 9

llll

14

II I II I

15

20

APRIL

25

MAY

12-.

12-"

BEB

10--

10-'

8- •

8 - •

6 - •

6--

4" '

4"'

2- •

2" 4

rflul 9 9 APRIL

14 1?

PEOLE

II II 15

, I

20

25

MAY

Fig. 3. Number of passes of difficult satellites (BEB and PEOLE) observed during a typical saturation period: above, photographic observations; below, laser observations.

necessary to compare observations pertaining to several passes. There may also be systematic errors on ranges obtained by a given laser (timing error, calibration error). Hence, the following procedure is to be applied in order to remove remaining systematically or accidentally erroneous observations: All available observations in a four-day interval are used to improve a preliminary orbit using a complete model of Earth Potential (e.g. modified Standard Earth I I or GEM 4) and a full model of luni-solar perturbations as well as perturbations due to radiation pressure, atmospheric drag, etc. Preliminary values of station positions are also included. The residuals that are obtained are larger than the noise, but it is easy to remove observations having large residuals. At the same time, the remaining

8 .

J 3

"

J. Kovalevsky . Residuals

(n)

-2.9 •-

-4.5.. 1 .533

2*. 07 3

2'. 6)3

3*. 153

3*. 693 4'.233 Ti«»io>

4*. 773

s'.313

s'.853

6*. 393

Fig. 4 a. Residuals of a typical GSFC laser pass in 1969 (see also Fig. 4b)

observations and the parameters, referred to a given dynamical model, are ready for refined analysis. These contain the geodetic information. This is clearly shown in Fig. 5, where the discrepancy between the residuals obtained from observations of French lasers in San Fernando and Haute-Provence are due to an inaccuracy in the assumed respective positions of the stations. The actual precision of the data is difficult to assess from these residuals, because they are strongly correlated with the dynamical model used to compute the orbit. The figures obtained are, therefore, a pessimistic representation of the value of the data. An example of such an analysis is a global station coordinate solution by Marsh, Douglas and Klosko based on camera and laser data, some of which were collected during IS AGEX [5]. From about 32 000 camera direction observations and 7000 laser distances, the following rms of fit was obtained: in right ascension and declination, ± 1 . 6 " ; in laser distances, ± 4 . 6 m. This is significantly larger than the internal precision of the data as far as laser distances are concerned. For the photographic data, this number corresponds exactly to an estimate made by Lambeck from the analysis of the observational and reduction procedures. For AFU-75 cameras, the rms error seems to be of the order of 4", except when GEOS2 flashes were observed. Then the accuracy seems to be better than 1".

6*. 93?

9

Results of the "ISAGEX" Campaign i , Residuals 1.0 . (m)

- 1 . 0 •-

o.poo

1

.265

1

.530

1

.795

1

1.060

1

1.325

1

1.590

1

1.855

!

2.120

1

2.385

1—

2.650

Fig. 4b Pig. 4. Comparison of residuals of a GSFC laser during a typical pass in 1969 (see Fig. 4a) and b, during ISAGEX showing the improvement in precision.

I n conclusion, one can agree with the assumed accuracy of ISAGEX observations as taken for weighting the data for the SAO Standard Earth I I I [6]. Lasers. Observations before 1970: SAO and GSFC: ± 5 m . GRGS: ± 1 0 m . Observations during ISAGEX: All lasers: ± 2 m . Optical

data.

Baker-Nunn smoothed data: ± 2 " . Other ISAGEX optical data range from 1" to 5" according to the type of observations and the type of camera as shown by various investigators and, for instance, by Marsh, Douglas and Klosko [7] or Massevich, Erpylev and Kasimenko [8]. Special care was taken to provide instructions for the proper use of these observations [9].

10

J . KOVALEVSKY

a •

a

i •

•

«•

a

% ' , a• * 1

a

«« »

ii

° •

• SAN FERNANDO X HAUTE PROVENCE » a . . • • ' •

•

•

a •

*

•

•

* M rS nw il iac

• •

••

•

•

il I

2

il 3 4 5 6 JUNE 27,1971 I hr. 50min«

7

Fig. 5. Laser residuals of the same pass of GEOS-A observed by the French lasers in HauteProvence and San Fernando yield information on the respective positions of the stations. Abscissa scale is number of minutes after 1 hr 50 min.

3. The Use of IS AGE X Data These conclusions already show the main contribution of ISAGEX: viz. about 200000 new individual laser distances, grouped in about 2000 passes with an accuracy 2 or 5 times better than that of the previous data. This is a scientific achievement that is unique and has not been repeated since. A catalogue of the observations collected in the data bank will be published soon.* One should not underestimate the other data, in particular photographic observations from many stations, but the latter contribution is not essentially larger or more precise than the previously collected observations of the 1960's. However, they usefully complete the laser data in the same periods of observation, and this is very important for the implementation of precise orbit determinations for geodetic and geodynamic studies. We have seen] that ISAGEX gave 14 arcs useful for such global solutions, but these arcs do not supersede analogous arcs already obtained in earlier campaigns (SAO observation compaigns, GEOS campaigns, D1 Mediterranean and European programmes, etc.). Some such arcs are even more densely observed than the best ISAGEX arcs (but with less accurate laser data, if any). This is why ISAGEX data are not to be considered alone for global work, but as an addition to existing data. An example of the improvement permitted by ISAGEX data is the construc* Note added after the GOSPAR Meeting: all data reduced during the I S A G E X programme and collected by the ISAGEX data bank are now available to all scientists. The data bank is located at GRGS/CNES in Toulouse, France.

Results of the "ISAGEX" Campaign

11

tion of the Standard Earth I I I in 1973 by G. M. Gaposchkin and his colleagues of the SAO [6]. The difference between Standard Earth I I [10] and I I I is not only the inclusion of ISAGEX data : in particular, the gravimetric data have also been improved; but the comparison between the two models gives fairly good indications on the contribution of the ISAGEX data: (i) Using the model S. E. I l l ,

Fig. 6. Improvement in geoid heights from SAO Standard Earth II (above) to Standard Earth III (below).

orbital positions of satellites can be reproduced to 4—10 m (rms) accuracy, twice as good as for S. E. I I . (ii) The geoid obtained has an accuracy of ¿ 3 m in geoid height and ¿ 8 mgal for the whole earth (correspondingly 5 m and probably 15 mgal for S. E. II). The improvement comes from the addition of low inclination satellite data, better ranges and better surface gravity data. However, the general features of the geoid are not changed (Fig. 6). (iii) The position of 90 satellite tracking sites in a uniform system were obtained (46 stations in S. E. II) with an uncertainty in the range 2—8 m (5—10 m in S. E. I I ) ; the ISAGEX laser stations are the best determined.

12

J. Kovalevsky

One can say, then, that ISAGEX has been a major factor in the recent improvement of global E a r t h Models and will continue to be so for further work in the field. As far as the geometric geodetic ties are concerned, the number of simultaneous observations obtained during ISAGEX was very small and they are only a contribution to longer systematic programmes like Arcant [8]. The positions of several stations were also determined by semi-analytical methods [5]. Other current work has also benefited from ISAGEX observations: for example, the continuing work of King-Hele to improve the system of zonal harmonics has greatly benefited from observations of P E O L E during ISAGEX [11]. An analysis of 15 arcs of DI-B, GEOS-2, B. E. C. and ANNA I B , one third of which comes from ISAGEX data, enabled Balmino and Reigber to study the resonant terms (arcsec)

*

1970

I

(m)

197!

Pig. 7. Variations of latitude obtained by GSFC after 17 months of observation of BEC by laser, compared with the smoothed BIH value.

of the 13th order for various degrees [12]. An analogous study has just been published by D. G. King-Hele for order 15, but he uses for this much lower satellites which are sensitive to this type of resonance [13]. One of the objectives of ISAGEX was to test some methods of determining quantities related to Solid Earth Physics. A good example is the determination of polar motion using laser observations of a satellite. When four consecutive passes of BE-C can be observed by the GSFC laser system, it is possible to determine exactly the inclination of the satellite and the latitude of the station. Although most of these events happened outside the ISAGEX and pre-ISAGEX period, this result by R. Kolenkiewicz, D . E . S m i t h and P . J . D u n n [14—16] can still be related to the experiment. Seventeen months of observations showed t h a t an accuracy of better than 0.05" (1.4 m) can be obtained for the latitude (Fig. 7). This is an encouraging result, but does not compete with the determination made by Anderle [17] at the NWL from TRANSIT system observations ( ± 0 . 4 m at present). The analysis of the variations in inclination observed during the same campaign also gave very significant results for earth tides which produce periodic perturbations on a satellite (Fig. 8). This confirms some other results obtained earlier by other authors from satellite results. A value of k = 0.25 for the second order Love number was obtained, which is different from the value obtained from earth-based observations (see, for instance [18]). Recently, Lambeck, Cazenave and Balmino [19] showed that this is due to the neglect of ocean tides. Their

13

Results of the "ISAGEX" Campaign (arcsec) r

2.5

(m) 80

'

60 40

20

0

J

A

S

0

N

D

I J

F

M

A

M

1970

J 1971

J

A

S

O

N

0

Pig. 8. Tidal perturbations upon the inclination of BEC computed with k2 = 0.245 and a phase angle q> = 3.2° compared with values deduced from laser observations.

analysis of the effects of various tides on a satellite [20] shows that the orbital perturbations are generally very small (a few tenths of a second of arc for most of the elements and satellites, when there are no resonance effects). More numerous and more accurate observations are still needed to proceed to a more detailed analysis of global earth and ocean tides by space techniques.

4. Future Prospects ISAGEX can be considered as the transition between two epochs. I t is the largest, but also the last, experiment in what is called "space geodesy". Although quite sizable, the improvement to the earth models is not as large as could be hoped and the effort to continue on the same trend may not worth while. On the other hand, earth physics can benefit very much from systematic precise observations of satellites as already shown by the first results on polar motion or earth tides, but the necessary accuracy means that the new techniques must be introduced together with a new and considerable improvement in instrumentation. The main scientific objectives for Earth Physics for future years may be defined as follows: (1) Earth potential. At present, spherical harmonics up to approximately order 10-10 and some resonant harmonics are derived from satellite tracking, others (up to 18-18) by analysis of gravity data. The study of the structure of the mantle necessitates knowing harmonics up to 30-30 and the investigation of some particularly important regions of contact between the lithosphere and the mantle might require knowledge of harmonics of order up to 50 or 60, or of the equivalent global representation of the geoid (for trenches, ridges or between plates). (2) Rotation of the earth. The rotation of the earth is now determined with an accuracy of 1 millisecond for a 5-day mean and space geodesy techniques still cannot compete with astronomical. For the motion of the pole, a resolution of 30 cm every two days is obtained through the observations of satellites. The problem is to understand the mechanism and the origin of the energy for the observed variations. Long-term effects will be subject to studies through the motion of the moon (lunar-laser), but the accuracy and time resolution must be improved by

14

J . KOVALEVSKY

another factor of 5 to 10 in order to permit studies of possible correlations with geodynamic triggers or other phenomena. (3) Earth tides. The global results on the earth, obtainable only by satellites, are only starting to appear and much better tracking data are necessary for refined analysis. These studies touch on the problem of energy dissipation within the earth. Important geophysical results are to be expected from the knowledge of this dissipation in the solid earth. This implies a much better knowledge of oceanic and atmospheric tides and brings global oceanography into the picture. (4) Plate tectonics. The relative motion of plates, from 1 to 10 cm per year, has not yet been measured directly. I t is necessary to improve the geodetic ties between fixed stations by another factor of 10 at least, to obtain significant results for a few stations. But it is very important to assess whether this motion is smooth or proceeds by jumps. I n order to obtain the results needed, it is necessary to improve greatly the existing precision of instrumentation. The goals are: (i) Lasers: To get to 3 cm accuracy in a few years. The recent improvements are quite satisfactory for such a prospect. (ii) Long base interferometry: To develop this technique and to obtain a 0.01" level of precision. (iii) Doppler and radio-range observation technique: To improve the accuracy by a factor of 100, especially in using satellite tracking techniques for which there is no ionospheric correction (satellite-to-satellite tracking); continuous lasers could give even better results. Table 2 Date of Launch Name

Country

Characteristics

1974

Timation III

USA

1974

Starlette

France

1974

GEOS-C

USA

1975

D5-B

France

1976

LAGEOS

USA

1977

Sea Sat 1

USA

1977

Dialogue

France

1980

Grav Sat Geo Pause Sea Sat 2 Small satellite Starlette type GEOLE

USA USA USA USA?

High altitude satellite with laserreflectors and precise Doppler system. Small dense sphere covered with reflectors for geodynamical and earth tides studies. 1-m altimeter and satellite to satellite tracking with ATS F laser reflectors. Low perigee satellite with microaccelerometer. Dense sphere covered with reflectors at high altitude for kinematic studies of the earth. 25-cm altimeter for potential and oceanographic studies. Data collection satellite for local geodetic purposes and testing the GEOSsystem. Refined gravimetric satellite (gradiometer and satellite to satellite tracking). 10 cm altimetric satellite. Dense spheres for geodynamical studies.

Space Shuttle ?

France or ESA

General localization system, with applications for kinematic studies of the earth.

Results of the "ISAGEX" Campaign

15

(iv) Altimetry: This technique, to be tested next year by GEOS-C, should make it possible at the end of the decade, to measure altitudes of a satellite above the sea to 10 or 20 cm. (v) Other techniques such as gradiometer or data collecting satellite (GEOLE) will also contribute to the objectives. The conjunction of the scientific objectives and the technical progress constitute the future Solid Earth Physics programme to be expected in future years. Such a programme has already been formalized by the US under the name of Earth and Ocean Physics Application Program (EOPAP) and also by France under its national programme as shown in Table 2. If this programme is actually performed, we may expect in the next decade that we shall understand much better than we do now the dynamical behaviour of the solid earth, since most of the present theories have not yet received quantitative confirmation. For this, a very large cooperative effort will have to be made from earth-wide networks. The example given by the cooperation that took place during ISAGEX will have to be followed in order to obtain the maximum results from the considerable efforts that are to be deployed during the next few years in Earth Physics. References [1] COSPAR, Decision No. 2, Prague (1969). [2] G. BRÄCHET, ISAGEX Experiment Plan, ISAGEX/7/CNES-GB/ZK/0.631/MT/CB (1970). [ 3 ] R . W . AGREEN, J . G. IIARSH, J . P . MURRAY a n d M . L . SANDSON, G o d d a r d S p a c e F l i g h t Center R e p . X - 5 5 3 - 7 2 - 2 3 (1972).

[4] GRGS, ISAGEX, Report of the observation Campaign, ISAGEX/14/CNES-GB/GR/ 1—77/CB/GRGS (1971). [5] J. G. MARSH, B. C. DOUGLAS and S. M. KLOSKO, Goddard Space Flight Center Rep. X - 5 9 2 - 7 3 - 1 7 1 (1973).

[6] E. M. GAPOSCHKIN, Ed., Smithson. Astrophys. Obs. Spec. Rep. No. 353 (1974). [7] J. G. MARSH, B. C. DOUGLAS and S. M. KLOSKO, Goddard Space Flight Center Report X - 5 9 0 - 7 3 - 3 4 0 (1973). [ 8 ] A . G. MASSEVTCH, N . P . ERPYLEV a n d T . V . KASIMENKO, S p a c e R e s e a r c h X V , 1 7 ( 1 9 7 5 ) .

[9] G. BRÄCHET, Data Handling Booklet, ISAGEX/IG/CNES (1972). [10] E. M. GAPOSCHKIN and K. LAMBECK, Smithson. Astrophys. Obs. Spec. Rep. No. 315 (1970). [ 1 1 ] D . G. K I N G - H E L E a n d G. E . COOK, R o y a l A i r c r a f t E s t . T e c h . R e p . 7 3 1 5 3 ( 1 9 7 3 ) . [ 1 2 ] G. BALMINO a n d CH. REIGBER, S p a c e R e s e a r c h X V , 5 3 ( 1 9 7 5 ) . [ 1 3 ] D . G. K I N G - H E L E , D . M . C. WALKER a n d R . H . GOODING, N a t u r e 2 4 9 , 7 4 8 ( 1 9 7 4 ) .

[14] R. KOLENKIEWICZ, D. E. SMITH and P. J. DUNN, Goddard Space Flight Center Rep. X-592-73-236 (1973). [ 1 5 ] D . E . SMITH, R . KOLENKIEWICZ, P . J . D U N N a n d H . H . PLOTKIN, S c i e n c e 1 7 8 , 4 0 5 ( 1 9 7 2 ) . [ 1 6 ] D . E . SMITH, R . KOLENKIEWICZ a n d P . J . D U N N , N a t u r e 2 4 4 , 4 9 8 ( 1 9 7 3 ) .

[17] R. J. ANDERLE, Determination of Polar Motion from Satellite Observations, Geophysical Surveys 1, 147 (1973). [18] P. MELCHIOR, Physique et Dynamique Planétaires, Vander, Louvain (1972). [ 1 9 ] K . LAMBECK, A . CAZENAVE a n d G. BALMINO, i n : 1 s t I n t . S y m p . o n t h e U s e of A r t i f i c i a l

Satellites for Geodesy and Geodynamics, Lagonissi 1973. [ 2 0 ] K . LAMBECK a n d A . CAZENAVE, B u l l e t i n G R G S N o . 7 ( 1 9 7 3 ) .

Space Research X V — Akademie-Verlag, Berlin 1975

SOME

RESULTS

BASED

ON T H E

ARCTIC-ANTARCTIC

ON O B S E R V A T I O N S

MADE

DURING

PROJECT 1970-1971

A . G . MASSEVICH, N . P . EKPYLEV a n d T . W . KASIMENKO Astronomical Council, USSR Academy of Sciences, Moscow, USSR

The geodetic vector traverse "Arctic-Antarctic" and the "East—West" traverse of the Project "Large Arc" consists of two parts: the determination of the directions of the geodetic vectors using photographic satellite tracking; and the determination of the lengths of these vectors using both photographic and laser-range observations. The analysis of photographic observations giving several directions in the northern part of the traverse "Arctic-Antarctic" has been completed. Some calculations have been made to determine preliminary values of the directions Riga—Cairo, Zvenigorod—Cairo, U z h g o r o d Cairo, Helsinki—Riga, Mirny—Kerguelen) using observations obtained during 1970—1971. The root mean square errors of these values are from 1.4" to 2.6". The equipment for laser-range observations of satellites of the USSR, Czechoslovakia, GDR, Poland and Hungary will be placed at the Cairo station in 1974 to take part in an observational campaign to determine the geodetic vectors Cairo—Uzhgorod, Cairo—Khartoum, Cairo—Afgoi as well as Cairo—Fort Lami. A new code for simultaneous determination of directions and ranges on base of photographic and laser tracking data is used.

The aim of the project "Arctic-Antarctic", started by USSR scientists in 1969 and adopted as an international project by COSPAR in 1971, is to determine, by means of simultaneous photographic and laser-range measurements of geodetic satellites, the directions and distances between stations situated along the meridian of the earth. According to estimates made by Zhongolovich [1] based on data available for San Fernando—Haute Provence), the accuracy of the distance determinations between separate stations will be 1:500000 and for the whole arc 1:700000. The participating stations are equipped with photographic cameras AFU [2] and laser devices (LSD) "Crypton" [3] built as part of the Intercosmos cooperation between Socialist countries. The main and secondary directions and distances are determined for the participating stations (see Fig. 1): Zvenigorod, Riga, Uzhgorod (USSR), Sofia (Bulgaria), Poznan (Poland), Ondfejov (Czechoslovakia) and for a number of cooperative stations by virtue of a bilateral agreement with the USSR Academy of Sciences: Cairo (Egypt), Khartoum (Sudan), Fort-Lami (Chad), Bamako (Mali), Afgoi (Somalia), Kerguelen (France), observations are starting in Barentsburg (Norway). Cooperation has also been established with the Finnish station in Helsinki. All these stations cover fairly well the northern part of the traverse Arctic2

Space Research XV

18

A . G . MASSEVICH, N . P . E R P Y L E V a n d T . W .

KASIMENKO

Antarctic. There are still problems with the southern part; particularly there is a gap between the stations Afgoi and Kerguelen covering an arc of about 56°. In 1973 we succeeded in obtaining 22 simultaneous groups for the PAGEOS satellite from Afgoi and Kerguelen; this allows preliminary determination of directions.

During 1974—1975 further attempts will be made. However, with such a large arc there are not many possibilities of obtaining a sufficient number and a favourable distribution of simultaneous groups. Also, this arc is too large for laser-range measurements simultaneous with photographic tracking. Another station in the area of Madagascar, Mauriki or Reunion is highly desirable. The position of the Kerguelen station has recently been determined using photographic tracking data of Midas 4 obtained in March 1971 to an accuracy of ± 3 5 m by Gajazov [4] to be X = 1407234 m; Y = 3917788 m; Z = - 4 8 1 6 0 0 6 m. A routine of improvement of the orbital elements and station coordinates, similar to that used for the "SAO Standard Earth", has been used [5]. Tracking data from stations Riga, Uzhgorod, Naini Tal, Addis Ababa, Oliphantsfontein

19

Arctic-Antarctic Project 1970—1971 Observations

and Kerguelen obtained in March 1971 during the ISAGEX campaign have been reduced. Simultaneously preliminary work on the "East—West" traverse has begun. Tracking stations inYuzhno-Sakhalinsk (USSR), Ulan Bator (Mongolia), Sant-Jago de Cuba (Cuba), Kourou (French Guiana), La Paz (Bolivia) are participating in this project. An AFU camera is also mounted in Tokyo (Japan). Negotiations on the building of new cooperative tracking stations in Ecuador and India are proceeding. Almost all these stations have participated in the ISAGEX programme. Using data obtained during the ISAGEX campaign several directions of the northern part of the Arctic-Antarctic project have been determined. Some of them are shown in Table 1 (

• ^ c o c o © © c o t - » © i > CO IO H oo CO ^

CO CO iA CO co co © 00 l A CO os o t*» co co © oo CO lA TU ^ I f l CO H Oí 00 kA IA CO Oi 00 kA CO co CO Oí *; o co co t co > 00 »A co N «O H cor-c-Oico^Hi-icooìoo co © © t -• co " ' Tt> co ~ ' H M W t»

oo © © GO oò Tt N © CO co

N t^ (O rt oo © t-

oo lA co CO vA tOí coo

co 00 00 co o ^H o 00 o í CO o í (M Oí co o -H CO O -H O O o o o

00 co N cô O GO Oí « o co H r- H

iA OI oj Oí

CO lA oo O O! cô ^ O O CO co co H © H TC kÄ

t> CO cô CO C0 co o ®

OÌ © oo 00 ^ C0 cô Oí Oí ' t T-H 00 oo N H N © lA lA

« 00 tH ® 00 H ® ** lA

® 00 C0 N o CO 00 Iß oí kA

H IM iA W h IM N N co CO

o o © Iii œ Oí í» O t— lA

co O í - « J < O í t - - ^ H t - - ^ C O O í © O í © C © O í C © i A i A I > ® C O © H W T H N W H 5 ® ^ « C M O > 0 ) C O W H ® « © , ' , ® 0 ^ co H Oit-COCOOir-Oí©©iA©COCOt"-COCOCOOílAiAOí N , , , co C O © C O C O ( M C O C O ' ^ © C O k A ' A i A C O t - C 0 T i < C 0 - ^ 0 0 ^H co C O C O m O r f M M N W N i f l W C ì O ì i O W H W N W O I

I

I I

I

CO C Í CO H O i © C O O O C O i H C O c O C O O i - H - H O i © C O C O O i C O C O i - I O O C O C O l A C O CO © CO eo Oí VA CO F - ••H TH CO H H N C 0 H W O i Ä C C O O i / 5 H H ' ^ 0 5 ' H i - i 0 5 ® M 0 0 t » 0 0 N CO © Oí t«. CO CO H w ^ o o o o c o i A o o © o > t » o o a o w c o i A c o c o > A c o c o c o i A c o o > i A © O © N N H O C O W N C C H H ' ^ ' t C O H O i n N O l C O C O N r t O i f l l A © CO Oi W O O i A C J i H l A k A l A O O C O O v H t ^ C O C O e o c O O O O O O O O i O i i H C O 00 H CO I f l eo ^ co

® N CO iH oo co H l > CO © lA © © ^ f-î ^ © »A O N co co CO iA © co G H •rH CO co co co • I

© © © VA © co o> kA co

^ O 0 0 N l O H N a O N ® N N ® ® N « © N 0 0 ® O N N N N Ö 5 O l f t 0 0 H 0 > O ® CO©OSlAlAlAI>CO©t-i>t^COCOCO I 001> I H H Î I IHTJ«TÍ «O W vA^HCO©COMCOiA'^'^CO-*} ) 1

© © © © •»$< © CO VA co o co H C-i l ¿ IO N N

©© ^ © 00 © co H O

© © co O OS CO ©

© CO CO co

o 00 CO co

© tVA co

© © lA ©

© VA kA ©

í i A C O i - i A ^ O -í ^ M ' c o e o e o t - r - c o * 1-1 1-1 1-1 1—I I I I H I

co © © CO Oí »A 00 t~» Oí co t > co o co c ò o i rtàoio VA ©

© © © © © © l A CO TH © Oí co co co

© © lA oo

© I> CO co

. . . . .

© © © © © © Oí t Tf co © l > Oí 00 VA CO lA © CO co co lA CO © Oí co l> ©

'S >> .S A

ft •

£«

s

ft

© co © © O 00 © Oí

o H s

•

S

I

a ftp S o © © © © © © © © © © ! — < © © © ! — ! © 3 © © © © © © © © 0 © © © © © © © © © O O © © © © © © © © O © © iH © © 00 © © © © © © © © © © © © © © o © ^HC0C0C0kAiAkAiAiACOCOC0COCOCO00XO0000C00i0í©©©iHCxrC0C0-^TfTí Ö CT I

I

CT

I

I I I I

I

-H

IO CO0¿TÍCTCOTjH>Ó'^HCO>O0'r-H0Ó«O

CO o

t^-^OOOOTjloOlO^OOOOOnHCT ®OSTÍttÍÓO!rt(NNr!oo()GlÍri CT -H « CT CT I I I I I I I I I 1 1

o ö CT

COcÓcÓC¿4-THCDCTOOOOCTOSOOCOt-;OOTH N-JrtioœoirioTÎMMdoi® I I

ce m

CT CO 00 CO 00 co >Ó CT -rt Oí CT

1

O COTjHOOOOOOl>COCTCOCO©COCOCO © i> t^O¿TjH-rHC>0¿C¿>0COCOI>C5TÍl> t^

A M

tí tí tí tí © © © ©

ü Oü O tí Ph.g dn ¿ Oh ¿4 e OOO O -4c3 CO o m o o o o CT CO « ÍD TiJ in M o¿ io o ri ò « CT "»M ^ CO I I !

CT Ö IN rt ^ CT CT 1 I I I I

1

© © © t©HfrH -H -43 frH -43 í--443

0

co 00 o co 00 © IO io 4-1 TH CT^ 05 OS 00 tH o 4-H os 05 io TH 00 00 có O 4-H có CT 1 1 1 1 1 1

có I

H J.

© -4h3

tí ©

CO © >0 CT © •»Í coCTio oó có

CT COCT05 © CO •«Í lH CTC5 TH O OS CT 4-( CT 00 o o O o ai COCTtÍ TÍ oí

4M CO CO es o co •

10

00

O CO OCT!> CT •»H 4H CO CO «5 CT 4-1 4-1 CO

CD TÍ

œ H >• b i --tí s ft ft CS çS ^ fcu ÖD bO O O o o PM PM

-tí j a o o ® 5 H H

O O O 5H fe fa 3 "3 O o ft s o •s o o

m m m

T3 « 'B S

c8 -Q

bp J, tao 'U -tì Í t í o r , 2 iE S © £ S

Ö c5 O 9 Hl H H H tí û5 PL, ^ « B § tí « § tí 1-5 g § S I ? » tí £ o H M O H « ¡ B o f l H h O Î Î H

S O

m

Geodetic and Astrometrio Results of VLBI Measurements

39

T a b l e 2a Source Positions from SBI, VLBI, and Photography Source

Declination (1950)

Right ascension (1950)

Code

03M6 m 29.568 s 29.568 29.558 29.565 29.571 29.536 29.58 29.562 29.548

± 0.007s ± 0.001 ± 0.004 ± 0.007 ± 0.010 ± 0.03 ±0.04 ± 0.003 ± 0.014

41°19'51.81" 51.81 51.91 51.95 51.82 51.67 52.1 51.99 52.19

± 0.02" ± 0.05 ± 0.2 ± 0.1 ± 0.15 ± 0.23 ±0.4 ± 0.04 ± 0.12

A B C D E F G I K

NRA0140

03 33 22.409 22.405 22.42 22.40 22.325

± 0.007 ± 0.002 ±0.03 ± 0.01 ± 0.04

32 08 36.52 36.55 36.7 36.8 36.1

± 0.02 ± 0.07 ± 0.4 ±0.2 ±0.5

A B G H J

CTA26

03 36 58.956 ± 0.023 58.953 ± 0.002

- 1 56 16.4 ± 0.3 16.8 ± 0 . 2

A B

3C120

04 30 31.603 ± 0.007 31.605 ± 0.001 31.605 ± 0.005 31.599 ± 0.01 31.605 ± 0.007 31.57 ± 0.03 31.60 ± 0 . 0 1 5

5 14 59.3 ± 0 . 1 59.4 ± 0 . 2 59.2 ± 0 . 1 59.9 ± 0 . 3 60.2 ± 0 . 3 60.1 ± 0 . 4 58.9 ± 0.3

A B C D E G J

OJ287

08 51 57.256 57.250 57.253 57.258 57.243 57.29

20 17 58.40 ± 0."03 58.41 ± 0.05 58.36 ± 0.1 58.7 ± 0.3 58.50 ± 0.23 58.3 ± 0 . 4

A B C E F G

4C39.25

09 23 55.326 ± 0.007 55.317 ± 0.003 55.319 ± 0.004 55.325 ± 0.007 55.295 ± 0.05 55.33 ± 0.04 55.29 ± 0 . 0 1

39 15 23.60 23.57 23.57 23.65 23.72 23.5 24.3

± 0.02 ± 0.03 ± 0.04 ± 0.1 ± 0.2 ±0.4 ±0.2

A B C D F G H

3C345

16 41 17.618 17.610 17.609 17.617

39 54 10.74 10.84 10.87 10.86

± ± ± ±

A B C D

3C84

± 0.007 ± 0.001 ± 0.005 ± 0.012 ± 0.02 ±0.03

± ± ± ±

0.007 0.002 0.004 0.010

0.02 0.04 0.07 0.20

a These coordinates are not corrected for elliptical aberration, which is the normal astrometric convention. The lists A, B, C, D and E definitely follow this convention. No reference to the contrary can be found in the other lists. The elliptical aberration corrections are given by Rogers et al. [23].

40

J . M . MORAN

Table 2 (Cont.) Source

Right ascension (1950)

Declination (1950)

Code

3C345

16 h 41 m 17.613 s ± 0.010 17.64 ± 0 . 0 4 17.610 ± 0 . 0 1 0 17.603 ± 0.002 17.56 ± 0.03 17.606 ± 0.013

39°54'11.15"± 0.15" 10.6 ± 0 . 4 11.3 ± 0 . 2 10.89 ± 0.03 10.7 ± 0.3 10.72 ± 0.12

E G H I J K

PKS2134 ± 00

21 34

5.223 5.208 5.205 5.212 5.23

± ± ± ± ±

0.007 0.001 0.005 0.010 0.03

00 28 25.6 25.0 25.2 28.6 25.7

±0.3 ±0.1 ± 0.2 ± 4.4 ± 0.4

A B C E G

VR042.22.01

22 00 39.377 39.367 39.374 39.379 39.365 39.387 39.31 39.37 39.362

± ± ± ± ± ± ± ± ±

0.007 0.003 0.007 0.02 0.010 0.02 0.04 0.01 0.007

42 02 8.48 8.55 8.45 8.33 8.52 8.47 9.0 8.8 8.69

± 0.01 ± 0.05 ±0.1 ± 0.25 ± 0.15 ± 0.23 ±0.4 ± 0.2 ± 0.04

A B C D E F G H I

3C454.3

22 51 29.533 29.524 29.519 29.515 29.524 29.54 29.54 29.510 29.485 29.533

± ± ± ± ± ± ± ± ± ±

0.007 0.005 0.009 0.01 0.010 0.02 0.03 0.006 0.035 0.011

15 52 54.15 54.30 54.29 54.26 54.36 54.37 54.90 54.54 54.45 54.98

± 0.02 ±0.1 ± Q.03 ± 0.2 ± 0.2 ± 0.23 ± 0.4 ± 0.09 ± 0.40 ± 0.15

A B C D E F G I J K

the various experiments [21 — 30], labeled A through K for easy reference. The most accurate positions available are the preliminary ones from J P L [21] and MIT/GSFC [22], lists A and B. All the positions of the ten sources common to lists A and B are given in Table 2. Fig. 1 shows the various measurements of the position of 3C454.3. Fig. 2 shows the difference between the J P L and MIT/GSFC measurements of eight sources. The sources PKS 2134 + 00 and CTA 26 were excluded from the comparison because they have low declinations, which were poorly determined in the J P L fringe-rate analysis. The rms deviation for the eight sources is 0.08 arcsec in declination and 0.09 arcsec in right ascension. The mean differences in the coordinates are 0.02 i 0.08 arcsec in declination and —0.07 i 0.05 arcsec in right ascension. The offset in right ascension is probably significant and is due to the difference in convention for the origin of right ascension. The rms deviation for the six sources common to lists B and I is 0.16 arcsec in both coordinates. There is a significant offset in both coordinates between

Geodetic and Astrometric Results of VLBI Measurements

41

the lists: 0.12 ± 0.10 arcsec in declination and —0.14 ^ 0.07 arcsec in right ascension. There are some problems in comparing the lists of source positions because of different conventions. The lists are all related to the F K 4 optical catalog in some way, however. One problem in the lists presented here involves the origin of right 0.2

1 1 MIT/GSFC-JPL

o >

3C84

n

1 3C454.3 •

3CI20•

0.1 ?

1

>VR042.22.01 .NRAO 140 .0J287 .4C39.25

oo




® 0 o> O
- 0 . 5

£

J = 21, = 6, orbit 522, 20 Jan., 0115). The atomic oxygen density increased continuously during this interval, then slightly decreased, but remained above its normal value even until orbit 532, 20 Jan., 1706, when the magnetic disturbance had become quite small.* Thus O-densities follow the geomagnetic disturbance with a time lag, i.e. the integrated disturbance effect is important. The atomic oxygen density increases first at greater altitudes near 400 km then also at altitudes down to 200 km. The relative importance of the disturbance is greater at the top level.

References [ 1 ] G . SCHMIDTKE, K . R A W E R , T H . FISCHER a n d W . LOTZE, S p a c e R e s e a r c h X I I , 169 (1974).

[2] S. CHAPMAN, Proc. Phys. Soc. 43, 26 (1931). [3] W. SWIDER JR, Planet. Space Sei. 12, 761 (1964). [4] L. G. JACCHIA, Smithson. Astrophys. Obs. Spec. Rep. No. 332 (1971). [ 5 ] H . G . M A Y R a n d H . VOLLAND, J . G e o p h y s . R e s . 78, 2251 (1973) a n d p r i v a t e c o m m u n i -

cation by H. Volland. [ 6 ] D . R . TAEUSCH, G . R . CARIGNAN a n d C. A . REBER, J . G e o p h y s . R e s . 76, 8318 (1971).

[7] U. VON ZAHN, Kleinheubacher Berichte, 17, 113 (1973). * A complete series of even orbit profiles from orbit 516 to orbit 532 is available from the author on request.

15

Space Research X V

Space Research XV — Akademie-Verlag, Berlin 1975

N E U T R A L A T M O S P H E R E VARIATIONS ACCORDING TO M E A S U R E M E N T S MADE IN TROPICAL AND M I D D L E L A T I T U D E S V . V . M I K H N E V I C H , A . A . P O K H U N K O V , E . N . GOLTJBEV, Y u . F . IVANOV a n d S . V . GORBUNOV

Hydrometeorological Service of the USSR, Moscow, USSR

Experimental data on the variations of neutral atmospheric composition, pressure and temperature obtained by MR-12 meteorological rockets launched from a research ship in summer 1973 at equatorial latitudes are compared with those obtained on the same days from the rocket site Volgograd. The ratio 0 2 /N 2 lies within ± 2 5 % of the Jacchia 1971 model values. All experimental data of the ratio 0 / 0 2 lie below the corresponding model values.

1. Introduction To reveal both the role of corpuscular radiation in the upper atmosphere and its latitudunal variation a combined rocket experiment was carried out in June to July 1973. The experiment also included simultaneous measurements of neutral atmospheric composition, density, temperature, corpuscular radiation and electron concentrations. The conditions of the experiment were chosen in order to minimize the effects of changes of solar radiation on the upper atmosphere. Since the experiment was carried out close to summer solstice, the site of rocket launchings from the scientific research ship Professor Vise was chosen as the Atlantic Ocean at 2° south, so that the noon solar zenith angle was 26°, which was almost equal to that of the rocket launching site Volgograd (latitude 49° N). One launching was also made from a point with coordinates 49° N 30° W, the geographical latitude of which corresponds to that of Volgograd, but the geomagnetic latitude is much higher. In Table 1 are given data on rockets, launchings and the geophysical conditions under which the experiments were carried out. In this paper the results of composition measurements carried out by a MX6407P mass spectrometer using procedures from [1] are considered as well as measurements of density and temperature by means of ionization pressure gauges [2]. 2. Measurement Results and Discussion As is seen in Fig. 1, in launchings 1 and 2 on 12 June, carried out at equal solar zenith angles a considerable difference in the relative concentration of atomic and molecular oxygen at heights above 110 km was revealed. Comparing measurements made at different latitudes by day (flights 4 and 7) one can see that, near 15*

228

V. V. M i k h n e v i c h , A. A. P o k h u n k o v et al. 0> C3 T> Oí Oí os CO co CO CO CO S t> r-; CO t> TH có có co co [ os 05 Oí Oí 05 05 05 00 TÍ 00 00 00 o o O TÍ Cl co co tí s q o 1-3

H w in

E) = N0 exp

(-EIE0)

for the energy range 0.5—50 keV. For the second launch, the downleg showed a marked increase of energetic electrons (10 _1 —10 2 keV). Electrons (few tens of hundreds keV) observed by Geiger counters SBT-9 (with an effective threshold energy for electrons of 40keV), show the absorption of radiation between 70 and 110 km. The energy spectrum becomes harder, as a rule, in this energy region and is approximated by 2V(2g E) a E~y,

where

y ~ 1.5—3.

This agrees with the measurements (Fig. 2) [7] carried out in February 1972.

3. Discussion Comparing the results obtained in polar regions during 1970 and 1972, it should first be noted that good agreement exists between the two groups of data. Secondly, unlike the 1970 experiment, smaller temperature and pressure variations

Rocket Measurements of Polar Thermosphere and Auroral Radiation

235

Table 1 Observed Quantities, March—April 1972 No.

1

Date, time (LT) K index (Heiss Isl.) Kp index

4 March, 1922 4 March, 2248 4 4 2 1

T (°K)

675 745 800

720 780 825

1.8 X 10"6 1.2 X 10- 6 8.4 X 10"'

2.1 x lO"6 1.4 X lO"6 9.8 X 10"'

4 6.8 X 106

Ascent 4.5 3.3 X 10'

Descent 4 5 X 108

4.3X10-2

2.4X10-1

3 . 2 x 1 0 ° 7.7 X 10"3

155 km 165 km 175 km

P (torr) 155 km 165 km 175 km Electron fluxes ( 1 4 0 - 1 8 0 km alt.) Ee = 0 . 5 - 5 0 keV E0 (keV) N0 (el cm s"1) Energy input (erg c m - 2 s _ 1 )

9 March, 0310 5 April, 1231 3 5 1_ 4

4 1.2 X 10«

4.4 1.7x10' 1.2 X 10- 1

are observed at an altitude of 165 km in the 1972 series. This is in agreement both with the electron intensity variations measured and changes of the local geomagnetic field (Table 1). On the ascending and descending parts of the trajectory of the second rocket the temperatures at an altitude of 155 km are 675 °K and 720 °K respectively. Data on the electron energy spectrum in the interval of 0.5—50 keV make it possible to derive the high-altitude profile of energy input SQjdh to the polar thermosphere from the electrons. Taking into consideration only this source and neglecting thermal conductivity processes, the time rates of change of temperature

Energy

Ee(keV)

Fig. 2. Energy spectra of energetic electrons in the upper atmosphere in polar regions

236

YTJ. M . ZHUCHENKO, V . A . LIPOVETSKY e t a l .

A

B

C

140 0.5

0.75 .HT

2

3 dT 5 dt • » " ' ' Heat rate (°Ks~1)

6

7 dT

Fig. 3. Heating rate of the polar thermosphere by electrons recorded in launchings 1 (curve A) and 2 (curves B and C).

T of the polar atmosphere at heights of 140—170 km were calculated using 8T 1 8Q HT ~ CpQ(h) ~8h' Using experimentally measured atmospheric parameters, the results of such calculations are shown in Fig. 3 for launchings No. 1 (curve A) and No. 2 (B, ascent, C, descent). The polar thermosphere heating rate increases from launching 1 to launching 2 (3.5 hours later) up to 25°K/hour at altitudes of 150—170 km. I t is concluded that the experimental investigations show that preceipitating electrons contribute appreciably to the heating of the polar thermosphere during periods of geomagnetic disturbance. References [1] S. AKASOFTT, Polar and Magnetospheric Substorms, D. Reidel Publ. Co., Dordrecht, Holland 1968. [2] V. F. TULINOV et al., Kosm. Issled. 1 1 , 704 (1973). [3] Bjuleten, Rezultaty raketnogo zondirovanijy atmosfery. o. Heisa 1970, No. 1, 1971. [4] Yu. M. ZHUCHENKO et al., Trudy CAO No. I l l , 137 (1972). [ 5 ] V . F. TULINOV et al., Trudy CAO No. 107 (1973). [ 6 ] L . C. NOVIKOV a n d G . F . TULINOV, T r u d y I P G , N o . 1 7 , 3 6 .

[7] V. F.

TULINOV

et al., Trudy IPG, No. 2 (1973).

Space Research XV — Akademie-Verlag, Berlin 1975

ATOMIC OXYGEN CONCENTRATION M E A S U R E M E N T S AT A L T I T U D E S OF 7 5 - 9 5 km S. P . PEROV a n d A . S. RAKHMANOV

Hydrometeorological Service of the USSR, Moscow, USSR

Results of atomic oxygen concentration measurements, made from seven rockets launched from different sites at altitudes between 75 and 95 km are given. Values range from 2 x 1011 to 3 x 1012 cm - 3 , and tend to increase with increasing height, except at high latitudes (Heiss Island) where the upper limit can be placed at 8 X10 10 cm"', The different values could be due to different conditions of geomagnetic activity.

1. Experimental Technique Measuring the atomic oxygen concentration was first conducted on a meteorological rocket M-100 using the surface heat effect of recombining O atoms on a fine wire (a resistance thermometer) [1]. In subsequent work the gauge construction was modified. In particular, a wire was used with an Ag—Pd coating having high catalytic activity for oxygen atoms. The device is mounted on the front part of the M-100 meteorological rocket nose cone in place of the standard pressure gauge. In flight the filament and the tube wall temperatures are measured, as is the pressure, from the heat deposited in the gauge, which works in the Pitot tube regime. The heating of the filament is a function of altitude; it reaches tens of degrees at altitudes greater than 90 km, where the wire is in conditions of minimal heat exchange and where the atomic oxygen concentration is a maximum. Reliable recording of the heat effect of recombining O atoms begins at altitudes of 70 to 75 km. The data obtained in the experiments are processed on the basis of freemolecular flow in cylindrical tubes [2].

2. Observations The measuring device was used in a set of rocket experiments carried out over Heiss Island (80°37' N 58°03' E), in the middle latitudes of USSR over Volgograd (48°41' N 44°21' E) and over Kerguelen Island (49°20' S 70°15' E). The data obtained on the rocket launched from the Soviet research ship Yu. M. Sholcalsky in the Indian ocean (50° S 65° E) near Kerguelen Island are also analysed. The results are shown in Fig. 1.

238

S. P . PEROV a n d A . S. RAXHMANOV

Smallest values of atomic oxygen concentration were recorded over Heiss Island in November 1972. Curve 1 is an upper limit to possible O-atom concentrations in the atmosphere, since the heating of the detecting elements was not more than 1°, the sensitivity limit of the apparatus. Interesting results were obtained in the first Soviet-French joint expedition on rocket sounding from

Fig. 1. Atomic oxygen concentration as a function of altitude on the basis of the experiments with heat gauges. Curves: 1, two flights over Heiss, 9 November 1972 at 0916 GMT and 15 November 1972 at 0926 GMT; 2, Volgograd, 27 February 1973 at 0800 GMT; 3, Indian Ocean (50° S 65° E), 26 February 1970 at 2100 GMT; 4, Kerguelen, 25 April 1973 at 1456 GMT; 5, Kerguelen, 27 April 1973 at 1458 GMT; 6, Kerguelen, 18 April 1973 at 1500 GMT.

Kerguelen. Curves 3—5 are in satisfactory agreement with one another as well as with the theoretical calculations [3]. Maximum values of atomic oxygen concentration were obtained in the flight on 18 April 1973 when considerable geomagnetic and ionospheric disturbances were recorded. I t is worth noting that a strong temperature anomaly was observed in the atmosphere at the time: a sharp temperature rise in the region of the mesopause and a sharp temperature drop (by 20°—30°) in the middle of the mesosphere (60—65 km) were recorded in the two flights on 18 and 20 April. There was also recorded a temperature drop of 15° in the upper stratosphere near 45 km altitude and a rise in the stratopause level by several kilometres. I n interpreting these results, account should be taken of the fact that Kerguelen Island is situated at the low latitude border of the auroral zone in the southern hemisphere where geomagnetic activity affects the atmosphere much more strongly than in other regions, and that the phenomena observed took place in autumn when the atmosphere is unstable and is greatly affected by geomagnetic disturbances.

Atomic Oxygen Concentration at Altitudes of 75—95 km

23»

3. Conclusions (i) The heat gauge used is a simple and reliable device for measuring atomic oxygen concentrations at heights of 75—100km aboard rockets; (ii) the data obtained give high 0-atom concentrations of the order of 10 12 cm - 3 ; (iii) some flights revealed a tendency for the O-atom concentration to increase at altitudes below 85 km ; this conclusion is, however, preliminary ; (iv) the smallest values of O-atom concentration were obtained in the northern polar zone in winter. References [1] S. P. PEROV and A. V. FEDYNSKY, paper presented at COSPAR Meeting, Tokyo 1968. [ 2 ] A . I . IVANOVSKY, T r u d y C A O , 7 2 , 3 6 ( 1 9 6 6 ) .

[3] V. N. DOULKIN et al., paper presented at COSPAR Meeting, Constance 1973.

Space Research X V — Akademie-Verlag, Berlin 1975

MEASUREMENT OF NITRIC O X I D E DISTRIBUTION IN THE U P P E R ATMOSPHERE T . TOHMATSU a n d N . IWAGAMI

Geophysics Research Laboratory, University of Tokyo, Tokyo, Japan

An ultraviolet photometer, designed to be used on board small sounding rockets for measurements of the nitric oxide gamma-bands in the day airglow, was launched on 19 August 1973 at Uchinoura, Japan. The density of neutral nitric oxide was deduced in the altitude range between 80 and 120 km. The aeronomic implication of the present experimental results is discussed in terms of the currently accepted photochemical-dynamical model of the upper atmosphere.

1. Introduction The main purpose of this paper is to report the results of a recent rocket experiment on atmospheric nitric oxide densities in the altitude range between 80 and 120 k m using a newly designed ultraviolet photometer. Nitric oxide (NO) is one of the most important minor constituents in the lower ionosphere. Production and loss of ionization in the D and E regions of the ionosphere are related to the photochemistry of NO. The abundance of neutral NO in the upper atmosphere can be determined from t h e observations of the NO day airglow [1—5]. I n particular, the gamma-bands of NO can be observed as the brightest spectroscopic feature of the dayglow in the middle ultraviolet. They are thought to arise from resonance fluorescence of atmospheric NO by solar middle ultraviolet radiation; specific fluorescence rates have been calculated by Pearce [6]. The brightest emission is the (1,0) band near 2148 A which has an apparent emission rate exceeding 1 k R at the base of the mesosphere. Whenever gamma-band dayglow observations of the atmosphere are made, the procedure for eliminating t h e spectral background always poses a serious problem from the experimental point of view. Rayleigh scattering of solar radiation by atmospheric molecules is the m a j o r source of background spectra a t altitudes below 90 km. Besides Rayleigh scattering an additional background exists, according to the present experimental results, which has been tentatively identified to be fluorescence of the Schumann-Runge bands of molecular oxygen [7]. F o r controlling the spectral background of the gamma-bands the scanning spectrometer has been widely used in previous dayglow experiments [8]. However, in order to gather sufficient light, it is always necessary to use this instrument at a moderate resolving power of 3 to 10 A corresponding to the rather wide slitwidth of 2—5 mm. Despite this merit of eliminating the background, the scanning 1 (j Space Research XV

242

T . TOHMATSTJ a n d N .

IWAGAMI

spectrometer generally requires large optical elements to compensate for its poor light-gathering property, especially for diffuse, faint sources like the dayglow. Instead of a spectrometer, a special filter photometer (NOSAC, nitric oxide self-absorption cell photometer) has been developed [9] for use in gamma-band measurements. The NOSAC makes use of the self-absorption effect of NO gas in a quartz cell, which serves as a sharp rejection filter for discriminating the gamma-band emissions from a continuous background. Because of its relatively high efficiency for transmission of light, NOSAC can be used for the detection of weak gamma-band emissions at high altitudes. I t is also rugged and compact enough to be used with small sounding rockets. 2. Instrumentation The NOSAC optical system consists of a solar baffle, a metallic honeycomb collimator which has an almost circular field of view of 10° in diameter and effective transmittance of 95%, a multilayered interference filter having a peak transmittance of 15% at 2150 A and a half bandwidth of about 120 A and two optical cells. Each cell is 30 mm in diameter and 20 mm in path-length; one (CI) is unfilled, but the other (C2) is filled with pure NO under a pressure of 200 torr. These cells can be alternately placed in the optical path of the system automatically every 0.8 second by an electric motor-driven mechanism. The second cell (C2) works as a rejection filter which very sharply cuts off the gamma-band emissions terminating at the ground vibrational level (v" = 0). This rejecting action is demonstrated by the spectra in Fig. 1, where the upper curve is the

WITH NO CELL WITHOUT INTERFERENCE FILTER

Fig. 1. Laboratory NO fluorescence spectra: upper curve, spectrum obtained without the NO cell; Lower curve, spectrum obtained with the NO cell.

Measurement of Nitric Oxide Distribution in Upper Atmosphere

243

spectrum of an artificially produced NO fluorescence detected by a laboratory monochromator in the absence of C2, and the lower curve is the spectrum taken under the same conditions, but with C2 placed before the entrance slit of the monochromator. One can notice that the (0,0) (1,0) and (2,0) bands are completely absent in the spectrum taken with C2 along the optical path, while the other features (v" =f= 0) remain unchanged. The NOSAC radiation sensor is an 11-stage photomultiplier which has a Cs-Te solar blind photocathode with a maximum quantum efficiency of about 10% at 2200 A. The photocurrent is amplified and telemetered to the ground station. When cell CI is in position, the output is the resultant of (i) the signal due to the gamma-bands from all combinations of the upper and lower vibrational levels and (ii) the background signal. On the other hand, when cell C2 is in position, the output is the resultant of (i) the signal due to the gamma-bands terminating at the ground vibrational level and (ii) the background signal. Consequently, a simple difference of successive output signals, first with CI in position and then with C2, should represent the intensity of the gamma-bands of the v" = 0 series. Because of the distributions of the spectral sensitivity of the photomultiplier and the bandwidth of the filter, the observed intensity is a weighted sum of the intensities of the (0,0) 2262 A, (1,0) 2148 A and (2,0) 2047 A bands. I t is assumed that the intensity ratios of these vibrational bands are constant and can be represented by the specific fluorescence rates themselves. This assumption is valid so long as the atmosphere is optically thin to the gamma-band dayglow. The (1,0) band constitutes the most significant part of the signal, being 88% of the total. 3. Experiment The rocket experiment with NOSAC was carried out at 1855 JST (ground sunset) on 19 August 1973 at Uchinoura, Japan (31°15'N 131°02' E). The photometer, mounted atop a single-stage rocket, had its optical axis of NOSAC aligned parallel to the spin axis of the rocket such that its field of view was always in a forward direction irrespective of the spinning motion of the rocket. The observation of dayglow started some 60 seconds after launch when the nose cone of the rocket was released. An altitude range between 80 and 120 km was covered in this experiment. The flight was normal as far as could be determined from the data of aspect magnetometer, the precession motion of the rocket being negligibly small around the spin axis directed at an azimuthal angle of 140° reckoned eastward from north. Although the rocket flew about 150 km towards the southeast, the solar zenith angle remained almost constant at 90° during the dayglow observation. 4. Results Fig. 2 illustrates the altitude profile of the emission rate of the NO gamma-bands observed in the ascending part of the flight. The intensity scale is given in equivalent emission rate of the (1,0) band. Also, the scale for the columnar density of NO is given assuming that the specific fluorescence rate of the (1,0) band is g10 = 3.93 X 10 -6 s _1 . Because the observation was carried out at a zenith angle of 20°, a reduction factor of 0.94 has been applied to all observed data to convert them 16*

244

T . TOHMATSU a n d N .

IWAGAMI

to equivalent zenithal values. The attenuation of the glow by atmospheric ozone and molecular oxygen was considered, but it was found to be negligibly small compared with the accuracy of measurement. A correction is needed for the attenuation of solar middle ultraviolet radiation by NO itself; according to NO Columnar Density (cm 2 )

Fig. 2. The NO gamma-band apparent emission rate.

theoretical calculations this correction is about 5% at the peak altitude of the NO density, i.e. around the 110 km level. However, this correction is within the magnitude of experimental error and thus was ignored. There is considerable scatter in the data points in Fig. 2 which is apparently due to light from the horizon of the earth. Because of this scatter, some smoothing procedure was necessary to evaluate the NO density from the vertical emission rate profile shown. Two smooth profiles were drawn by a least squares method using the data points in Fig. 2. Curve 1 is drawn assuming that a polynomial of order 3 with respect to altitude best fits all data points. Curve 2, on the other hand, is drawn for a polynomial of order 5, imposing a physical assumption that NO above 115 km is in diffusive equilibrium so that its density above that altitude should obey an exponential law with the scale height of NO determined by atmospheric temperature. The scale height was chosen to be 8.5 km corresponding to a temperature of 300 K, and the columnar density of NO above 115 km was determined to be 6.3 X 1013 molecules cm -2 from the data obtained near apogee. The vertical profile of NO density can be obtained by differentiating these smoothed curves with respect to altitude. The results are shown in Fig. 3, together with the model NO distributions calculated by Ogawa and Shimazaki [10]. The family of theoretical profiles was drawn for various values of the parameters related to the production of excited atomic nitrogen in the lower thermosphere. The parameter rjR represents the efficiency of the production of N( 2 D) (written N * ) in the dissociative recombinations: NO+ + e = N * + 0 ;

N2+ + e = N + N * ,

Measurement of Nitric Oxide Distribution in Upper Atmosphere

245

and r/j the same quantity for the photodissociation of N 2 , N 2 + hv = N + N * . The experimental profiles seem to suggest a high probability of the production of excited nitrogen atoms in the dissociative recombinations as well as in the photodissociation of N 2 .

140

\K% \ \\oiv\\l 0

N0 + e = 0 + N ( ! 0 ) NÎ+e=N+Nl20)

V

120

y 80

60

y)

y

100

v|f

V

1

\

s)/) y NO

—

Model B Obs.

r

1=1.0 0.5 0.1 Ni + hv = N + N ( ! D )

40

104

10s

10"

10'

10'

10*

Number Density (cm -3 )

Fig. 3. The observed NO density profiles compared with the theoretical profiles calculated by Ogawa and Shimazaki [10]. References [1] [2] [3] [4] [5] [6] [7] [8]

C. A. BARTH, J . Geophys. R e s . 6 9 , 3301 (1964). C. A. BABTH, P l a n e t . S p a c e Sei. 1 4 , 6 2 3 (1966). J . B . PEARCE, J . Geophys. R e s . 74, 853 (1969). L . G. MEIRA, J B , J . Geophys. R e s . 76, 202 (1971). D . W . RUSH, J . Geophys. R e s . 78, 5 6 7 6 (1973). J . B . PEABCE, J . Quant. Spectrosc. R a d . T r a n s f . 9 , 1592 (1969). R . D . HUDSON a n d S . H . MAHLE, J . Geophys. R e s . 77, 2 9 0 2 (1972). W . G. FASTIE, J . Quant. Spectrosc. R a d . T r a n s f . 3 , 507 (1963).

[9] T. TOHMATSU and N. IWAGAMI, Bull. Inst. Space Aeronaut. Sei., Univ. Tokyo 10, 537

(1974). [10] T . OGAWA and T . SHIMAZAKI, to be published in J . Geophys. R e s . , 1975.

Space Research XV — Akademie-Verlag, Berlin 1975

E X P E R I M E N T A L DATA ON ATOMIC NITROGEN VARIATIONS IN T H E UPPER ATMOSPHERE A F T E R SUNSET V. N. Balabaitova, K. D. Bychkova, V. N. Lebedinets, V. P. M a r t y n e n k o and A. A. Pokhunkov Hydrometeorological Service of the USSR, Moscow, USSR

Values of atomic nitrogen concentration at 140 km altitude after sunset (zenith angles 100° —150°) obtained at middle latitudes by the ethylene luminous cloud method indicate that the mechanism of N disappearance at night is more complex than predicted by theoretical calculations. It is shown that 1 h 15 min after sunset the N concentration at 140 km is 1010 atom c m - 3 and close to the daytime N concentration. Eight hours after sunset the N concentration has considerably decreased and is not observed by the luminous cloud method nor by mass spectrometers. It is concluded that shortly after sunset either there are reasonably intense sources of atomic nitrogen or the loss processes are rather slow.

1. Introduction Atomic nitrogen is an important component of the reactions characterizing ion and neutral composition of the upper atmosphere. Experimental data show that the atomic nitrogen concentration is equal to some per cent and may reach 80% of that of N2 [1—5], Nitrogen is prominent in night-time conditions when, according to present concepts [6], the main aeronomic processes are dissociative recombination NO+ + e

N + O

and interaction of nitrogen atoms with molecular oxygen and nitric oxide N + 0 2 ^ N 0 + 0;

N + N 0 ^ N 2 + 0.

Experimental data on nitrogen concentrations in the upper atmosphere for daytime and night-time conditions are lacking. Theoretical calculations give a wide range of nitrogen concentration values at night [7—10]; this seems to be explained by lack of data on the sources of atomic nitrogen and the rate of N loss processes after sunset. According to [8, 9] the atomic nitrogen content at night is 103 atom cm -3 in the 100—200 km height region. I t was found [7] that one hour after sunset at 140 km the atomic nitrogen concentration decreases by a factor of 104. Using mass spectrometers [1] the ratio [N]/[N2] is shown to be 0.5% in the 120—210 km height region (32°24' N, 20 July 1967, 0200 MST). In this case [N] is 3 X 108 atom cm - 3 at 140 km.

248

V. N. Balabanova, K. D. B y c h k o v a et al.

2. Observations The experiment carried out on 31 October 1972 at 0130 LT (48°41'N) was aimed at simultaneous determination of atomic nitrogenconcen trations at night by mass spectrometers, with a sensitivity limit of 108 atom cm - 3 (a radiofrequency mass spectrometer MX-6407P), and the ethylene luminous cloud method [11], the sensitivity limit of which is 106—107 atom cm -3 . Nitrogen atoms were not found by both methods during this time, i.e. 8 h 17 min after sunset (solar zenith angle % = 145°). [N] sS 107 atom cm -3 . N concentrations observed by the ethylene luminous cloud method 1 h 17 min after sunset were 8 x 109 atom cm - 3 at 140 km {•/ = 99°30'), i.e. some orders of magnitude higher than theoretical values given in [7]. On the other hand, near twilight this [N] value is close to the atomic nitrogen concentration (109—1010 atom cm"3) at this height during the daytime, as measured by mass spectrometers [1, 3—5]. Experimental [N] values obtained after sunset are given in Table 1; theoretical values of the night-time N concentration and experimental data on daytime N concentrations are given for comparison. Table 1 Data on N Concentration at 140 km at Different Times of the Day [N] (atom cm - 3 ) after sunset Experimental data

Theoretical data, references

Daytime [N] (atom cm"3) (latitude) [references]

Data, local time, latitude

Time after sunset

[N] (atom c m [references]

1 July 1969 2121 48°41' N

01 h 15 min

7 X 109 [*]

< 105 [7], ^ 103 [8] 3 X 109 48° 41' N) [3] 103 [9], 107 [10] 1.8 X 109 (48°41' N) [5]

8 July 1969 2120 48°41'N

01 h 17 min

8 x 109 [*]

1.2 X10 10 (48°41' N) [4] 3 X10 9 (32°20' N) [1]

31 October 1972 0130 48°41' N

08 h 17 min

< 10' [*]

20 July 1967 0200 32°20'N

04 h 56 min

< 3 x 108 [1]

0 [7], g 103 [8] 10 3 [9], 107 [10]

[*] our data

Table 1 shows that data on N concentrations obtained by the ethylene luminous cloud method (1010 atom cm - 3 at twilight) are not beyond the limits of scatter in the mass-spectrometer measurement data for daytime conditions. If such a scatter in the data on [N] corresponds to an actual variation of atomic nitrogen concentration in the upper atmosphere, the data given in Table 1 confirm strong solar zenith angle dependence of [N].

Atomic Nitrogen Variations in Upper Atmosphere after Sunset

249

A strong dependence of atomic nitrogen concentration on ionospheric disturbances and solar activity has been considered [4, 5]. One can note a correlation between atomic nitrogen concentration and precipitating charged particles and electric fields [5], which can be irregular but rather powerful additional sources of atomic components of the upper atmosphere. Of particular interest for aeronomy would be a series of additional experiments aimed at investigating the recognized strong variability of data on atomic nitrogen concentrations, its time dependence and the causes of such variability. I n such experiments the atomic nitrogen concentration should be measured simultaneously by different methods, for example by mass spectrometers and by the luminous cloud method in order to decrease the effects of measurement errors. References [ 1 ] D . KRANKOWSKY, W . T . KASPRZAK a n d A . O. NIER, J . G e o p h y s . R e s . 7 3 , 2 3 ( 1 9 6 8 ) .

[2] S. N. GHOSH et al., J. Geophys. Res. 73, 13 (1968). [ 3 ] A . A . POKHUNKOV, S p a c e R e s e a r c h X I I , 6 5 7 ( 1 9 7 2 ) . [ 4 ] G. M . MARTYNKEVICH a n d E . D . BYURO, S p a c e R e s e a r c h X I I , 6 8 1 ( 1 9 7 2 ) .

[5] A. A. POKHUNKOV et al., in: The Sun-Atmosphere 71, Hydrometeoizdat, Leningrad 1972. [6] A. D. DANLLOV and M. N. VLASOV, Photochemistry of Ionized and Excited Particles in the Lower Ionosphere, Hydrometeoizdat, Leningrad 1973. [7] C. A. BARTH, Chemical Reactions in the Lower and Upper Atmosphere, Interscience Publ., New York, London 1961. [8] D. P. STROBEL, J. Geophys. Res. 76, 10 (1971). [9] P. P. SATENA, Annl. Geophys. 25, 1 (1969). [10] M. NIKOLE, Aeronomy, Izd. Mir, Moscow 1964. [ 1 1 ] V . N . BALABANOVA, K . D . BYCHKOVA a n d V . P . MARTYNENKO, P r o c . I E M , V o l . 1 , 1 9 7 2

(p. 39).

Space Research X V — Akademie-Verlag, Berlin 1975

F I R S T R E S U L T S OF 6300 À NIGHTGLOW M E A S U R E M E N T S ABOARD A ROCKET L A U N C H E D FROM NATAL, B R A Z I L Y . SAHAP, A . DRESCHER", H . LAUCHE0 a n d N . R . TEIXEIRA* a

Instituto de Pesquisas Espacisais, Sào José dos Campos, SP, Brazil ^Deutsche Forschungs- und Versuchsanstalt für Luft- und Raumfahrt, Oberpfaffenhofen, FRG c Max-Planck Institut für Aeronomie, Lindau/Harz, FRG

The 0 1 6300 A nightglow emission was observed by a rocket-borne photometer launched from Natal (5.9°S 35.2°W), Brazil. The preliminary height profiles of the volume emission rate are presented. These profiles are compared with the calculated emission profile using ionosonde data from Natal and a model atmosphere.

1. Introduction Two identical photometers to measure the red line emission of atomic oxygen at 6300 A were flown on a Black Brant 5 c rocket payload launched at 0516 UT on 16 February 1973 from Natal (5.9° S 35.2° W), Brazil, in conjunction with a pass of the German Aeronomy Satellite AEROS. In this paper the preliminary results of the OI 6300 A emission profiles observed from the rocket are presented; calculated emission profiles for this emission by dissociative recombination are also presented. The electron density profile was obtained from an ionogram taken at Natal, and temperature and density profiles were calculated using the static models of the thermosphere and exosphere [1]. 2. Instrumentation and Flight History Each photometer consisted of a light baffle (to prevent direct light from the moon), an interference filter, a field aperture, a focusing lens, a field of view defining stop, and a selected photomultiplier with a S-20 photocathode, and was calibrated using a 14C radioactivated phosphor source in the laboratory. The aperture was 40 mm in diameter and the field of view 4° (full angle). The interference filter of photometer I had a 25.7 A half-power bandwidth and of photometer I I 10.6 A. The photometers were mounted parallel to the rocket axis, looking forward. The signal was measured by a single-photoelectron-pulse counting technique with low pulse height discrimination and transmitted by PCM telemetry. Besides the photometers, the rocket payload consisted of a neutral mass spectrometer, a mass spectrometer for ions and neutrals, a retarding potential analyser, an electron density experiment (differential Doppler), a low energy electron spectrometer (cylindrical electrostatic analyser, 17—1450 eV) and an attitude control system.

252

Y . SAHAI, A. DRESCHER

et

al.

The flight was towards the east, and reached an apogee of 312 km. The moon was below 30° elevation and at about 290° azimuth. The spin rate of the rocket was 2.6 revolutions per second. The retarding potential analyser and the neutral mass spectrometer gave no useful data. The attitude control system showed disturbed data.

3. Preliminary Data Evaluation The photometer data showed no spin modulation. Therefore, every 64 transmitted values of the counter were averaged to give one signal point, corresponding to a time interval of 1.024 seconds. Both the photometers showed periodic variations in the signal strength in phase with the rocket precession. The first evaluation of the disturbed attitude data showed an elevation variation of the rocket axis between 0° and 40° zenith distance. No corresponding signal modulation (Van Rhijn effect) is apparent from the data. Therefore, the preliminary attitude data were not used and a graphical smoothing procedure was adopted to eliminate signal variations caused by the rocket precession.

4. Results and Discussion Photometer I I data were selected for calculating the integral intensity profiles shown in Fig. 1 because they show less noise and lower background than the photometer I data. The observed integral intensities are 165 rayleighs and

INTEGRAL

INTENSITY OF [OL] 6 3 0 0 A

VOLUME

[RAYLEIGHS]

Fig. 1. Height profiles of the 0 1 6300 A emission;

E M I S S I O N OF [ O I ] 6 3 0 0 Ä [PHOTONS. C M 3 ]

upleg,

down leg.

6300 A Nightglow Measurements aboard Rocket launched from Brazil

253

190 rayleighs for the upleg and downleg respectively. The curves of the integral intensity versus height contain background and dark current. The curves of volume emission rate versus height shown in Fig. 1 were calculated by differentiation of the integral intensity versus height curves. Estimates of the error

LOCAL

TIME

Fig. 2. Ground-based measurements of the 01 6300 A emission. Arrow indicates rocket launch time 0216 LT. zenith intensities 0° zenith distance; zenith intensities 20° east zenith distance; Q, calculated intensity with ionogram data of 0241 LT using T ^ [1]; calculated intensity with same ionogram data using — 900°K.

introduced by the smoothing procedure are 5 — 1 0 % in the integral intensity and 15—30% in the volume emission rate. The height profiles of the volume emission rate show a broad emission layer. The emitting layer had maximum brightness between 255 and 275 km. The discrepancies between the profiles of the upleg and downleg above 304 km are probably due to a horizontal variation in the residual intensity of the OI 6300 A emission above the apogee of the rocket. This may account in part also for the apparent slow cut-off at the lower boundary of the emission layer and larger integral intensity for the downleg profile. The OI 6300 A emission was also monitored with a ground-based tilting filter photometer on the night of 15—16 February 1973 at the launch site. The zenith intensity variation for 0° and 20° east (the direction of rocket launch) zenith distances are presented in Fig. 2. The OI 6300 A emission in the direction of 20° east was about 155 rayleighs at 0516 U T (0216 L T ) and its intensity was increasing; this is also evident from the rocket data. Thus the ground-based measurements are in fair agreement with those from the rocket observations.

254

Y. Sahai, A. D r e s c h e r

e t al.

Figs. 2 and 3 show respectively the calculated 0 1 6300 A emission intensity and volume emission rate due to dissociative recombination of 0 2 + ions. Eq. (1) of [2] was used for calculations with the following rate coefficients: yx = 1.5 X 10~ u cm 3 s- 1 [3]; y2 = 7 X 10"13 cm 3 s" 1 [3]; a = 5 x 10"11 cm 3 s" 1 [ 2 ] ; « , = « j = l x 10"7 cm 3 s _ 1 [4]; k = 0.9 [5]. The bottomside electron density profile (kindly supplied

VOLUME

EMISSION OF [Ol] 6 3 0 0 1 - PHOTONS. C M 3

Fig. 3. Observed and calculated ( T x = 900°K) OI 6300 À emission profiles.

by Dr. A. Ramakrishnan of Institut für Physikalische Weltraumforschung at Freiburg, Germany) was obtained from the ionogram taken at 0541 UT at Natal and a Chapman function was assumed for the topside profile. Exospheric temperature and density profiles were calculated using static models of the thermosphere and exosphere [1]. The calculated intensity was smaller t h a n that observed but showed a better agreement when the exospheric temperature was assumed to be 900 °K, about 140 °K higher than that given by the Jacchia model [1], This indicates that a larger neutral particle density in the F region is required to explain the observations. The calculated volume emission height profile (with T = 900 °K) shown in

6300 A Nightglow Measurements aboard Rocket launched from Brazil

255

Fig. 3 no doubt shows a broad red emission layer similar to that observed from the rocket but the shape at lower emission heights is different. This discrepancy is likely to be better resolved when final rocket data evaluation is completed. 5. Conclusions The height profiles of the volume emission rate show a broad red 6300 A 0 1 emission layer with maximum brightness between 255 and 275 km. The difference in the upleg and downleg profiles may be due to a non-uniform horizontal structure of the emission layer. The observed intensity from the ground-based measurements and that calculated using the dissociative recombination show better agreement when a higher exospheric temperature than that given by the Jacchia model [1] is used. Also, the calculated volume emission rate shows a different shape at lower emission heights than that given by the preliminary rocket data. A better picture of the equatorial F region is likely to emerge when we are able to estimate the OI 6300 A emission rate using on-board ion, electron and neutral densities and to compare it with the final rocket photometer data. This in turn may also provide better estimates of certain reaction rates. References [1] L. G. JACCHIA, Smithson. Astrophys. Obs. Spec. Rep. No. 332 (1971). [2] V . L . PETERSON a n d T. E . VANZAKDT, P l a n e t . S p a c e Sci. 17, 1725 (1969).

[3] E. E. FERGUSON, Annls Geophys. 25, 819 (1969). [4] M. A. BIONDI, Canad. J. Chem. 47, 1711 (1969).

[5] E. C. ZIPF, Bull. Amer. Phys. Soc. 15, 418 (1970).

Space Research XV — Akademie-Verlag, Berlin 1975

PRELIMINARY RESULTS OF OBSERVATIONS OF ATMOSPHERIC ULTRAVIOLET TWILIGHT EMISSIONS BY THE TD1-A SATELLITE A . MONFILS a n d J . C. GÉRARD

Institut d'Astrophysique, Université de Liege, Belgium

Preliminary data collected by the UV telescope on board the ESRO TD1-A satellite in its winter spinning mode are described and analysed. They consist of angular scans parallel to the terminator plane, near the dusk meridian, in four ultraviolet channels covering the spectral range 1 3 5 0 - 2 2 0 0 A. The main twilight emissions are identified as well as the altitude-latitude intensity dependence. The Mg II resonance doublet near 2800 A appears to be present in one of the channels, in agreement with previous high altitude observations. The altitude distribution of magnesium ions obtained by comparison between the northern and southern scans at a given latitude is illustrated.

1. Introduction Satellite TD1-A [1] was launched into a polar sun-synchronous orbit in March 1972. Among the seven experiments on board, S2/S68 [2] was designed to scan the sky and has provided spectrophotometry data for several thousand hot stars. In accordance with the astronomical aims, the telescope is permanently pointing towards the zenith. When extended sources are observed, the instrument works as a sensitive fourchannel photometer in the following spectral bands : Channel A1 A2 A3 A4

2 7 5 0 ± 150 1330-1780 1730-2180 2130-2580

 (at half maximum) A A A

The data recorded during the first scan immediately showed an unexpected signal in channel A l , which appeared sporadically near the evening crossings of the magnetic dip equator. I t was recognized by Boksenberg and Gérard [3] that this signal was to be attributed to a resonance scattering of Mg+ ions lifted up by the well-known "fountain effect" in the ionosphere [4]. Maps of the observations were drawn up by Gérard and Monfils [5], who were able to localize the phenomenon, which is limited to a narrow region north and south of the dip equator. A longitudinal effect has also been detected by the same authors. During the winter period, the spacecraft is kept stabilized in a fast spinning mode about the sun-satellite axis. This mode of observation has been used to scan the plane perpendicular to the spin axis, thus parallel to the terminator. Measure17

Space Research X V

258

A . MONFILS a n d J . C . G E R A R D

ments of the altitude profiles of the emissions situated in the spectral ranges of the four channels, and particularly of the last one, were expected as a function of latitude.: Channel A2 N2 (Lyman-Birge-Hopfield) [0] A 1356 A Channel A3 NO y system (mainly (1,0) band) Channel A4 NO y system Channel A1 Mg I I X 2800 A residual NO (y system) The preliminary results presented here are based on the few "quick look data" available in January 1974. The profiles investigated have been recorded between 60° and 15° geographic north latitude. Owing to the orbital properties of the satellite, the earth below was then in darkness. The intersection of the sunlit limit of the atmosphere and of the scanning plane tends towards zero altitude when the satellite approaches the equator. 2. Description of the Observations In the considerations which follow, emphasis has been put on the mid-latitude scans, because, for the northern ones, the atmosphere is in darkness up to 300 km altitude and in the southernmost ones, the Rayleigh scattering complicates the analysis and cannot be readily subtracted quantitatively. Apart from some high latitude profiles, obviously influenced by auroral activity, all the northern profiles corresponding roughly to the same latitude are almost identical. The same is true for the southern profiles. Fig. 1 shows a typical recording of the southern

Channel

A1 A2 AU

z> o

o

Orbit 3 rd.

100

7

\\

/'

•A A'i I

¡V V

m &-A

! A

'V

W' . 100

10208 S o u t h e r n Horizon Scan

\ /' \\ V \

20'LH

N.V )

\

\ i

200

.'

f /;
-x/

A''

300

'

.

'

./ . ; ,

\

ALTITUDE

¿00 (km)

Pig. 1. Southern horizon scans recorded around 35° N by channels A l , A 2 and A4.

Observations of Atmospheric UV Twilight Emissions by T D 1-A Satellite

259

horizon recorded around 35° N. The value of the impact parameter, which is closely related to the altitude of the emitting layers, has been calculated by simple trigonometric formulae. The most conspicuous feature of channel A l is alsmost certainly due to Rayleigh scattering. I t corresponds to an altitude of ~ 10 km which can hardly be distinguished from the ground owing to the measuring error. The intensity is low: 650 counts, i.e. 30 rayleighs/Á. At higher altitudes a new maximum is observed around 150 km, with an intensity of 0.5 k R situated in the sunlit part of the atmosphere. This may correspond to the signal observed during normal satellite operations, i.e. the Mg I I fluorescence. In the present case, the intensity drops to the background at a little over 250 km and consequently no emission is to be seen by looking upward. The intensity maximum corresponds, through the Van Rhijn effect, to a latitude of approximately 20° N. The two other channels drawn are A2 and A4, showing probably [O] X 1356 Á in the first case, and NO bands in the second. In the first, a distinct maximum at high altitude (200 km) can be seen. This seems too high for the Lyman-BirgeHopfield system, which is the only other obvious possibility. The intensity is low, (250 R at peak), but the maximum is very broad (250 km at half height). As far as channel A4 is concerned, a very broad asymmetrical maximum is observable, with a peak near 200 km, immediately above the limit of the sunlit part of the atmosphere; it is present down to very low altitudes. I t should be pointed out that the latter limit has been computed without taking into account either the refraction effect or any screening heights. Fig. 2 corresponds to an observation which has been made around 25° N, which is 10° closer to the equator than in the case of Fig. 1. I t can be seen that the A l profile shows a high altitude Mg I I emission, although the intensity is not very

Fig. 2. Southern horizon scans by all four channels around 25° N. The similarity of the channels A 3 and A 4 is apparent. 17*

260

A . MONFILS a n d J. C.

GÉRAKD

high. Due to the much lower shadow height, the Rayleigh scattering is much more intense, peaking to a value of 600 Rayleighs/A for a layer that must be close to the ground if the Van Rhijn effect is responsible for the observed profile, the peak possibly being due to the ground itself. I n order to find the Mg I I fluorescence,

COUNTS

Fig. 3. Successive northward and southward altitude profiles from channel A1 recorded near 38° N.

we have to look at the tail of the curve which indeed shows a counting rate exceeding the 10 counts noise up to more than 450 km. The other channels also show very interesting aspects. A3 and A4 channels exhibit a very similar profile which is readily accounted for by resonance scattering of the NO gamma bands. The peak altitude seems to be around 50 km. Channel A2 shows a peak around 220 km, which corresponds to an intensity of 400raleighs provided the X 1356 is the only contributing emission. The width is the same as for Fig. 1 (250 km), but the intensity is somewhat higher. The effect looked for is, by nature, irregular: only studies extending over many recordings can lead to a complete analysis of the altitude profile and its variations with time and latitude. I t is, however, possible to derive some information on the problem

Observations of Atmospheric UV Twilight Emissions by TD 1-A Satellite

261

from the comparison of consecutive scannings of the northern and southern horizons. Such profile comparisons have been drawn in Figs. 3 and 4. I n the first case, the profiles have been recorded near 38° N. They illustrate the fact that we are very probably observing the expected fluorescence effect: due to the 75 sec-

Fig. 4. Successive northward and southward horizon scans recorded from around 25° N by channel A l .

onds separating the recordings, the shadow height has decreased from 180 km to 130 km, crossing the altitude of the maximum which then appears around 170 km in the sunlit part of the atmosphere. Higher up, there is not much difference. I n Fig. 4, the lower latitude of the observations ( ~ 2 6 ° N) appears much more favourable, although the presence of a low-altitude superposed emission makes it necessary to subtract it before plotting the data. This was done by extrapolating to higher altitudes the exponential decrease. The two profiles obtained by looking north and south show the expected effect: in the northern direction, an important maximum (675 R) is observed around 170 km, and corresponds to a latitude of 45° N. Above 170 km, the decrease is fast, the intensity falling to the noise level at about an altitude of 380 km. The southern profile, on the contrary, shows a lower intensity for the (160 km altitude) maximum (575 R) but the

262

A . MONFILS a n d J . C. GÉRARD

intensity does not drop as fast as a function of altitude as in the first case: an altitude of 450 km is necessary for a drop to the noise level. By easy calculation, it can be seen that the latitude region covered around 350 km is situated around 10° N, which is the geographic latitude of the geomagnetic dip equator, where the ionospheric fountain effect should be maximum. 3. Conclusion By using the fast spinning mode of satellite TD1 several hundreds of horizon profiles have been observed. The preliminary analysis of the quick-look data has shown (i) that they do not correspond to days where the high altitude Mg I I fluorescence effect is noticeable, and (ii) that the same effect as the one invoked for the explanation of the preceding observations is, in fact, likely to be effective. The analysis of the complete data will proceed in the future and will cover the whole of the experimental material. Acknowledgments One of us (J. C. Gérard) is supported by the National Belgian Foundation for Scientific Research (FNRS). We have pleasure in acknowledging with thanks the efficient help given to us by ESOC in coordinating the telemetry operation. References [1] B. TILGNER, ELDO-CECLES/ESRO-CERS Scient. and Tech. Rev. 3, 567 (1971). [ 2 ] A . BOKSENBERG, R . G. EVANS, R . G. FOWLER, I . S. K . GARDNER, L. HOUZIAUX, C. M. HUMPHRIES, C. JAMAR, D . MACAU, D . MALAISE, A . MONFILS, K . NANDY, G. I. THOMPSON, R . WILSON a n d H . WROE, M o n . N o t . R . A s t r . S o c . 1 6 3 , 2 9 1 (1963). [ 3 ] A. BOKSENBERG a n d J . C. GERARD, J . G e o p h y s . R e s . 78, 4 6 4 1 (1973). [ 4 ] W . W . HANSON, D . L. STERLING a n d R . P . WOODMAN, J . G e o p h y s . R e s . 77, 5 5 3 0 ( 1 9 7 2 ) .

[5] J. C. GERARD and A. MONFILS, J. Geophys. Res. 79, 2544 (1974).

Space Research XV — Akademie-Verlag, Berlin 1975

P R E L I M I N A R Y R E S U L T S ON S U P R A T H E R M A L H AND He ATOMS IN T H E M I D D L E L A T I T U D E S LOWER T H E R M O S P H E R E D U R I N G A MAGNETIC D I S T U R B A N C E G . M . MARTYNKEVICH

Hydrometeorological Service of the USSR, Moscow, USSR

An MR-12 rocket, launched from Volgograd on 12 June 1973 during the recovery phase of a geomagnetic storm, carried a time-of-flight mass spectrometer for measurements of neutrals and a double-mode radiofrequency mass spectrometer for measurement of neutrals and positive ions. Suprathermal neutral particles, such as H and He atoms, with energies up to 25 eV were detected. High concentrations of thermalized H and He atoms were also recorded. The results of the measurements are discussed and compared with data obtained earlier by other authors.

I. Introduction An MR-12 rocket was launched on 12 June 1973 at 0554 LT from Volgograd at 48°41' N (geographic) 42°41' N (geomagnetic). The rocket reached a peak altitude of 168 km. This firing formed a part of a series of launchings designed to study the diurnal variability of upper atmosphere parameters. The launching was performed during the recovery phase of a geomagnetic storm, which started on 10 June 1973 with a sudden commencement at 1042 UT [1]. The launch day, 12 June 1973 was one of the most disturbed days of June.

2. Instrumentation Two mass spectrometers, one time-of-flight (TOFMS) [2] and the other radiofrequency (RFMS) [3], were mounted aboard the rocket. The TOFMS registered only neutrals. However, the RFMS operated in two modes, recording alternately positive ions and neutrals. In addition the scan period of the RFMS was shortened up to 1.4 sec without losses in sensitivity. The TOFMS operated well on the upleg (120—168 km) and for a significant portion of the downleg but during re-entry into the dense layers it lost its sensitivity almost entirely. The RFMS operated well in an altitude range of 113—168—70 km, i.e. throughout the whole trajectory. The light mass (1— 4 amu) analyser information of the RFMS was recorded excellently. Difficulties were caused by high values of ion currents of thermal (H, He) and energetic or suprathermal (H E , He E ) hydrogen and helium atoms,

264

G . MARTYNKEVICH

as well as by intense total ion currents in cases of insufficient light mass analyser resolution necessary for a separation of the corresponding peaks, for example He+ and HeE+. 3. Results of Measurements In the positive ion mode, the light mass analyser continuously detected suprathermal ions. At certain (short) time intervals, for example on the downleg in an altitude range of 88—72.6 km, the concentrations of suprathermal ions were

140

-

120

- ^ r J l

80

i

100 E,S

i

50

~~

0

50

WIND COMPONENTS

(msec"1)

1

100 W,N

Fig. 2. Zonal and meridional wind components observed over Thumba at evening twilight of 19 April 1971 by means of sodium trail release.

0

50 WIND

100 SPEED

(msec

-1

)

Fig. 3. Wind speed observed over Thumba at evening twilight of 19 April 1971.

270

J . N . D e s a i , P . D . B h a v s a b e t al. 1

1

ZONAL W N ID — MERD IO INAL W N IO -

—

Si < - .

^

/

ES

'

100

• SO 0 50 WN IO COMPONENTS ( m )

10i0

WN

'

Fig. 4. Zonal and meridional wind components of Pig. 2 plotted with a modified altitude scale

v

=

C dz 80

JH

where H is the pressure scale height.

2.2. Turbulence and Diffusion The trails showed abrupt cessation of turbulence at a height of 104 i 1 km, and a sharp demarcation could be observed between the turbulent and the nonturbulent portions of the trail. Although several vapour release experiments have been made from Thumba, earlier twilight releases were above 105 km and hence the height of the turbopause was not observed. The observations reported here are the first to determine this level over Thumba. The first signs of turbulent brak-up could be noticed 30 seconds after the release at 104 km and the lower portions of the trail became apparently patchy a little later. I n this respect our observations do not seem to support the view [4] that the turbopause is the level above which the time constant of the Kolmogorov microscale of turbulence rapidly increases. Between 30 seconds and 400 seconds after the release, the upper level at which turbulence apparently ceased to exist remained at 104 ± 1 km. High wind shear of about 0.045 sec - 1 was also observed between 105 and 110 km with the wind direction changing from SW to N E . I t appears that, as suggested by Hodges [5], strong wind shears operating in conjunction with gravity waves could have been responsible for producing turbulence. The trail was optically rather thick. Brightness contours were, however, obtained for a portion of the trail at 125 km altitude by microphotometer scan of the trail photographs obtained on a film calibrated by means of a neutral density step wedge. The diffusion coefficient was estimated by plotting the trail width at 1/e (peak brightness) against (time) 2 ; 81) was equal to the slope of this plot. The diffusion coefficient estimated in this way was 1 . 5 x l 0 7 m 2 s _ 1 ( ± 2 5 % ) which is more or less in agreement with an average model value.

Winds and Diffusion in Upper Atmosphere from Na Vapour Trail

271

Acknowledgments The authors wish to express their thanks to NASA, USA, for supplying the rocket and the Department of Space, Government of India, for financial support. Thanks are also due to TERLS, India, for the range support, particularly the group which developed the sodium vapour payload. The authors gratefully remember the encouragement given by Prof. K. R. Ramanathan and (the late) Prof. V. A. Sarabhai. References [1] P. J. SMITH, Planet. Space Sci. 11, 1311 (1963). [2] P . D . BHAVSAR a n d K . RAMANUJA RAO, S p a c e R e s e a r c h V I I I , 6 5 5 (1968).

[3] M. S. NARAYANAN, Ph. D. Thesis, Gujarat University Ahmedabad 1973. [4] D. REES, R . G. ROPER, K. H. LLOYD and C. H. Low, Phil. Trans. Roy. Soc. London A 2 7 1 , 6 3 1 (1972). [5] R . J . HODGES J R , J . G e o p h y s . R e s . 72, 3 4 5 5 (1967).

Space Research XV — Akademie-Verlag, Berlin 1975

A C R I T I C A L E V A L U A T I O N OF T H E OGO 6 H E L I U M M O D E L G . M . KEATING11, E . J . PRIOR", D . S . M C D O U G A L " a n d J . NICHOLSON I I I B a

NASA, Langley Research Center, Hampton, Va, USA b 01d Dominion University, Norfolk, Va, USA

Drag measurements of the Explorer 19, 24 and 39 satellites near 900 km altitude and over the solar cycle through 1973, have been critically compared with the recent thermospheric model based on OGO 6 neutral mass spectrometer measurements from mid-1969 to early 1971. The drag data clearly indicate both a significantly greater exospheric helium and temperature variation over the 11-year solar cycle than given in the model. Assuming the very large poleto-pole temperature variation 400°K) given in the OGO model, the drag data indicate that the winter to summer pole helium concentration ratio is 21 at 450 km compared with the value of 12 indicated by the OGO model. This difference may result from an underestimate of the helium variation in the OGO model or a polar temperature variation somewhat less than that given in the OGO model but significantly larger than that given in the CIRA 72 model. This study gives the first evidence from drag data of a helium diurnal variation near the turbopause with a morning maximum which agrees in phase but not amplitude with that given in the OGO 6 model.

1. Introduction A global model of the neutral thermosphere has recently been constructed based on measurements obtained by means of the neutral particle quadrupole mass spectrometer carried aboard the OGO 6 satellite [1], This OGO 6 model is the first diffusive equilibrium thermospheric model to be based primarily on neutral mass spectrometer data. I t describes longitudinally averaged N 2 , O and He densities for magnetically quiet conditions, and is given in terms of boundary conditions at 120 km and at 450 km, and the exospheric temperature. Previous models like the COSPAR International Reference Atmosphere, 1972 (CIRA 72) [2], based primarily on satellite drag data, infer composition and temperature from density, density scale height and conditions near the turbopause. I n this study the OGO-model densities are compared with those derived from satellite drag measurements obtained by means of three of the 12 foot diameter Air Density Explorer drag satellites (Explorers 19, 24 and 39) to establish how well the model predicts the systematic variations in atmospheric density measured by the drag satellites. The method by which densities are determined from the drag satellites is described in detail in [3—5]. The drag measurements were obtained principally in the helium-rich lower exosphere above 800 km altitude. Attention is focused on the systematic variations of helium concentrations predicted by the model. 18

Space Research XV

274

G . M . K E A T I N G , E . J . PRIOR e t a l .

However, there are a number of reasons to suspect that significant errors could occur in the OGO 6 model. First, a sharp change in calibration of the mass spectrometer occurred immediately after launch which resulted in the sensitivity to various constituents being reduced as follows: mass 4 (helium) by a factor of 4.32, mass 16 by 1.82, mass 28 by 4.70 and mass 32 by 5.03 [6]. Further undetected shifts in sensitivity could have occurred over the satellite lifetime which would cause errors in the representation of systematic variations in the model. Second, considering that most of the mass spectrometer measurements are made near perigee (analogous to a drag satellite) it is unlikely that over the experiment lifetime of less than two years a representative sample of all the global variations indicated by the model could have been obtained. Third, the model which has over 100 coefficients appears to be exceedingly complicated considering the limited data sample. Other potential problems include the accuracy of atomic oxygen measurements and the validity of using molecular nitrogen concentrations as a measure of temperature. Nevertheless, this model attempts to explore quantitatively more variations than previous thermospheric models.

2. Solar Cycle Variation The measurement intervals for the three near-polar satellites used in this study were as follows: Explorer 19 (Dec. 1963-Aug. 1973), Explorer 24 (Dec. 1964 to Aug. 1968) and Explorer 39 (Aug. 1968—Aug. 1973). On the average, measurements were obtained above 800 km. Intervals where atomic hydrogen could significantly affect results were eliminated. Shown in Fig. 1 is the logarithm of the ratio of observed to OGO-model densities for the Explorer 19 satellite from 1964 to 1974. Also shown is .Fio, an index of the long term variation of solar activity. As may be seen, during low solar activity, negative density residuals occur indicating that the OGO model is overestimating atmospheric density. This same trend is evident from the measurements of Explorers 24 and 39. The inadequacy of the model to describe the solar cycle variation

, /L .

200

K

A

>G0—

V

.

100

r

V

50 .5

«

"g{, ,if the OGO 6 Helium Model

275

is understandable considering the time interval over which the OGO 6 measurements were obtained; the bar (Fig. 1) from July 1969 through December 1970 covers 162 of the 166 days of measurements used to generate the model. Since during this period F 1 0 remained relatively constant at approximately F w = 150, a sample was not obtained of conditions during low solar activity. Using all the drag measurements over the solar cycle, improved values of some of the coefficients used in the OGO 6 model [1] for both temperature and helium x in6 1.6 p ^ nu„

1.2-

,™v,„„-3, . .

Z

=

^OGO MODEL 1000

DRAG

|MPR0VED

OGO MODEL

Fig. 2. Solar cycle variation of helium and exospheric temperature (Spring equinox, noon, equator).

were determined (see Table 1). I t is concluded that the exospheric temperature varies much more over the solar cycle than is indicated by the OGO 6 model. This result is shown in the bottom portion of Fig. 2 where during low solar activity exospheric temperatures are 100 ° K lower than indicated by the OGO model. Also the decrease in helium concentrations at 120 km altitude with increasing solar activity is much smaller than is indicated by the OGO model. This effect could be caused by a slight increase in the eddy diffusion coefficient with increasing solar activity. The combined effects of the drag-improved model coefficients result in helium concentrations at 1000 km shown in the upper portion of Fig. 2. The OGO model indicates practically no increase in helium concentrations (for 1000 km, spring equinox, noon, equator, Ap = 4) as F10 increases from 70 to 200, while the dragimproved coefficients indicate a large increase.

3. Latitudinal-Seasonal Variation The pole to pole amplitude of the winter helium bulge determined from drag measurements is very sensitive to the assumed latitudinal-seasonal variation of temperature. As the authors have pointed out in the past [3, 7] if the pole to pole variation of temperature is greater than indicated by drag models like C I R A 72, a larger amplitude winter to summer pole helium variation will be deduced. T o infer helium concentrations from drag measurements, the predicted oxygen concentration is essentially subtracted from the total density. If a larger pole to pole temperature variation is assumed, more atomic oxygen will be subtracted from the total density at the summer pole reducing the inferred helium concentration. 18*

276

G . M . KEATING, E . J . PRIOR e t a l .

The reverse will occur at the winter pole. The total effect will be to increase the amplitude of the drag-inferred winter helium bulge. In addition, if there is a larger seasonal variation of temperature the concentration of helium will decrease more rapidly with altitude in the winter hemisphere since the temperature (and thus the scale height) will be lower in winter and higher in summer. Thus a larger pole to pole temperature variation will result in a more pronounced winter helium bulge flattening with altitude. The winter to summer pole temperature differences and winter to summer pole helium ratios at 450 km (F10 = 150, Ap = 4) have been calculated for the three atmospheric models. The OGO model, based on mass spectrometer measurements near 450 km, indicates a helium ratio of 12. The CIRA 72 model, based on drag data, gives an inferred helium ratio at 450 km of only 4 due to an underestimate of the temperature difference between the winter and summer poles. The dragimproved OGO 6 model gives a value of 21 for this ratio. The OGO model assumes a factor of 4 larger temperature difference between the poles than the CIRA model (439 °K compared with 106°K); such a large temperature difference will sharply increase the drag-inferred helium ratio at 450 km. The larger amplitude of the drag-deduced winter helium bulge could result from the OGO model underestimating the helium ratio or overestimating the pole to pole temperature difference. Vertical probe data also suggest that the seasonal variations of helium may be in excess of the values given in the OGO model [8]. Results from the neutral particle monopole mass spectrometer aboard the ESRO 4 satellite indicate at least a factor of 20 helium ratio at 270 km [9]. I n addition OGO 6 obtained very limited observations of helium in the summer hemisphere [1]. On the other hand, a seasonal temperature variation in excess of the variation given in the CIRA 72 model but less than that given in the OGO model could result in consistency between drag data and the OGO model helium variation. The large amplitude winter helium bulge apparently results from a major meridional circulation pattern in the lower thermosphere from the summer to winter hemispheres [10, 11]. Variations in the eddy diffusion coefficient may also play a role [12]. 4. Diurnal Variation A diurnal variation in composition near 120 km has not previously been deduced from atmospheric drag data and no drag models include such a variation. Such a variation has been inferred here; the diurnal variation at 120 km is shown as the dashed line in Pig. 3. As may be seen, the drag data indicate that helium concentrations at this altitude maximize in the morning hours. In order to compare these results with the OGO model itself, the effect of drag smoothing on the OGO model's diurnal variation was determined by calculating the theoretical drag effects on the Explorer 19, 24 and 39 satellites orbiting in an OGO 6 model atmosphere. Next densities were determined from the theoretical orbit decay and finally, as in the case of the actual drag measurements, best values of the two primary diurnal coefficients were determined from the densities after setting the other diurnal, semidiurnal and terdiurnal helium coefficients equal to zero. The full curve in Fig. 3 represents the corresponding equatorial diurnal variation at equinox. The diurnal variations of helium deduced for 120 km from the drag and mass spectrometer measurements both show a maximum in the morning hours between

A Critical Evaluation of the OGO 6 Helium Model

277

06 and 09 h. The variation based on actual drag measurements has an amplitude of 1.4. In comparison, the corresponding OGO model variation has a higher amplitude of 2.1 suggesting that the OGO model may overestimate the helium diurnal variation at 120 km. Using the same drag data, a diurnal variation at 120 km maximizing in the morning hours (amplitude 1.5) is also indicated when

Fig. 3. Drag-inferred diurnal variation of helium at 120 km.

the other variations are removed by using an improved version [13] of the US Standard Atmosphere Supplements, 1966. The diurnal maximum near 08 h for helium inferred near 120 km is probably the result of a diurnal redistribution by wind-induced diffusion [14]. 5. Conclusions Exospheric densities measured by means of the orbital decay of the Explorer 19, 24 and 39 satellites have been compared with the OGO 6 atmospheric model to determine the validity of the model and to improve some of the primary model coefficients. The results of the analysis indicate the following: (1) The OGO 6 model underestimates the variation of exospheric temperature and helium concentrations over the solar cycle. This results in helium concentrations and exospheric temperature being seriously overestimated at solar minimum. (2) Assuming the OGO temperature model, drag measurements indicate that the helium concentration ratio between the winter and summer poles is 21 at 450 km, which is significantly greater than the value of 12 indicated by the OGO helium model. This suggests that either the OGO model somewhat underestimates the amplitude of the winter helium bulge or overestimates the latitudinal-seasonal variations in temperature. In contrast the CIRA 72 model apparently underestimates the latitudinal-seasonal temperature variations. A substantially larger difference between summer and winter pole temperatures than that given by CIRA 72 would result in greater consistency between drag and mass spectrometer measurements of the winter helium bulge.

278

G . M . KEATING, E . J . PRIOR e t al.

(3) The first evidence from drag measurements of a diurnal variation in composition near 120 k m is given in this study. T h e inferred morning m a x i m u m of helium is in fair agreement in phase but has a lower amplitude than that given in the OGO model. This morning m a x i m u m m a y result from wind-induced diffusion. Table 1 Drag-Improved OGO Model Coefficients compared with OGO Model OGO 6 Model Spherical Harmonic Coefficients [1] flj,T ¡¡¡¿, He cj 0 , He a11(He 6 n , He

OGO Model

2.67 x l O " 3 - 4 . 4 7 X 10- 3 1.44 -0.37 +0.36

Drag-Improved OGO Model 3.9 XlO- 3 - 2 . 1 0 x 10~3 1.70 -0.02* +0.16*

* other diurnal, semidiurnal and terdiurnal helium coefficients set equal to zero.

References [1] A. E. HEDIN et al., J . Geophys. Res. 79, 215 (1974). [2] CIRA 1972 (COSPAR International Reference Atmosphere 1972), Akademie-Verlag, Berlin 1972. [3] G. M . KEATING a n d E . J . PRIOR, S p a c e R e s e a r c h V I I I , 9 8 2 (1968).

[4] J. A. MTJLLINS et al., NASA TN D-3432 (1966). [5] G. M . KEATING e t al., N A S A T N D - 2 8 9 5 (1965). [6] A . E . H E D I N e t al., N A S A T N D - 7 2 3 9 (1973).

[7] G. M. KEATING et a l , Space Research XII, 765 (1972). [8] G. M. KEATING, paper presented at IAGA meeting, Kyoto, Japan (1973). [9] U . VON ZAHN e t al., J . G e o p h y s . R e s . 78, 7 5 6 0 (1973).

[10] [11] [12] [13] [14]

F. S. JOHNSON and B. GOTTLIEB, Planet. Space Sci. 18, 1707 (1970). C. A. REBER and P. B. HAYS, J . Geophys. Res. 78, 2977 (1973). G. KOCKARTS, Space Sci. Revs., 14, 723 (1973). G. M. KEATING et al., Space Research XIII, 327 (1973). H. G. MAYR et al., J. Geophys. Res. 79, 619 (1974).

Space Research X V — Akademie-Verlag, Beilin 1975

A N N U A L A N D S E M I A N N U A L D E N S I T Y VARIATIONS IN T H E EARTH'S E X O S P H E R E C. WULF-MATHIES1, E . J . PRIOR" a n d G . M . KEATING" a

Institut für Astrophysik und extraterrestrische Forschung der Universität Bonn, Bonn, FRG b NASA Langley Research Center, Hampton, Va, USA

Drag data of the three balloon satellites Explorer 19, 24 and 39 were analyzed for the period 1967—1970 in order to obtain information on the semiannual density variation in the lower exosphere. The observed semiannual effect was decomposed into Fourier terms. The annual and semiannual component are sufficient to represent the observations. The phase of the semiannual component is relatively constant, peaking on the average on 11 April and 10 October. The phase of the annual component varies strongly. Its maxima occur between October and March. The semiannual amplitudes change by a factor of more than two over the period of analysis. The annual amplitudes are much less variable. Hence the variation of the amplitude of the semiannual effect is mainly due to changes of the amplitude of the semiannual component. Comparisons between the observations and predictions of the OGO 6 and CIRA 72 models showed large discrepancies in phases and amplitudes. The semiannual effect as predicted b y the OGO 6 model peaks generally 14 days earlier than the observations.

Two years ago Volland, Wulf-Mathies and Priester [1] tried a new theoretical approach toward the so-called semiannual density variation in the thermosphere based on the analysis of drag data by several authors [2—5]. Their main results were: The semiannual effect is composed of a predominant variation with a period of half a year and a much smaller annual variation. Because the maximum of the annual variation occurred in January it was concluded that its energy source is the annually varying solar heat input into the ozone layer as the earth-sun distance changes. I t was suggested that the heat source of the semiannual component, which generally peaks at the middle of April, could be a combined effect of Joule heating and dissipation of waves at the lower boundary of the thermosphere. In the following we shall apply the term "semiannual effect" for the superimposed effect; otherwise we shall speak of the annual and semiannual component. I t is well established that the semiannual effect can be traced through the whole thermosphere [4], but there is only limited information [3, 6] with regard to the lower exosphere, the helium and hydrogen regime. In order to obtain additional information the orbital decay data of the three NASA Langley Air Density Explorers (Explorers 19, 24 and 39) were analyzed for the period 1967 — 1970. All three are large balloons with diameters of 3.66 m, and their orbits have an orbital inclination of about 80°. Previous studies [7] of the semiannual effect with these satellites for the period 1964—1967 revealed a semiannual variation essentially

280

C. Wulf-Mathies, E. J. Prior and G. M. Keating

in phase with the variations at lower altitudes and a September anomaly with anomalously low densities in September 1964, 1965 and 1966, a phenomenon which disappeared in 1967. In order to calculate air densities at half a scale height above perigee, the changes in the orbital energy due to solar radiation pressure had to be subtracted from the total rate of change of the orbital energy. The solar radiation pressure raised the perigee height of Explorer 19 from around 700 km at the beginnung of 1967 to around 1000 km by the end of 1970, while Explorer 39 kept its perigee altitude from its launch in August 1968 to the end of 1970 at about 750 km. Explorer 24 showed only a small variation around its perigee height of 600 km until it decayed in 1968. The orbital elements for these studies were supplied by the Smithsonian Astrophysical Observatory. The method of drag analysis which was applied in the investigation is described elsewhere [8]. The observed densities which were obtained from the three Explorers were compared with the CIRA 72 model [9] predictions after all variations except the semiannual and the geomagnetic activity effects had been taken into account. In addition the asymmetric seasonal variation of helium was subtracted [10]. In Fig. 1 the ratios of the observed densities to the CIRA 72 model densities are plotted for the time period 1967 — 1970. The abscissa is the time in modified Julian days ( = Julian days — 2400000.5). The CIRA 72 model values were calculated for the appropriate geophysical and solar coordinates and also the appropriate solar 10.7 cm flux. The residual density variations of the three Explorers should exhibit the semiannual effect and to a lesser degree the heating associated with geomagnetic disturbances. If the semiannual effect were absent, the ratios would equal unity. Despite the scatter of the data the semiannual effect can be seen over the whole period, even in 1970 when the curve is strongly affected by some spurious variations. The lower curve outlines the Explorer 19 measurements and the second curve above the respective measurements of Explorers 24 and 39. Explorer 24 decayed in the second half of 1968, two months after the launch of Explorer 39. The appearance of data sets is very similar. In 1967 there is a relatively large semiannual effect; the amplitudes decrease towards 1968, but the effect is still clearly visible. At the end of 1969 and the first half of 1970 the curve is severely disturbed, until the second maximum of 1970 has built up. There exists not only a strong similarity between the semiannual density variation for the three Explorers with their respective perigees varying between 600 km and 1000 km, but also with the appearance of this effect at thermospheric heights. Particularly remarkable are the large amplitudes in 1967 and the relatively small scatter in the observations. This strong variation of the amplitudes of the semiannual effect through the time period of observation is not understandable in the light of the fact that the average solar flux remained relatively constant over the whole period. The semiannual effect seems to be independent of solar cycle variations. The observations of the Explorer satellites were developed into a series of Fourier terms, for each year separately. The two major components of this harmonic analysis are an annual and a semiannual component, whereas the amplitude of the third and higher order harmonics are of the order of the scatter of the data. Hence only the first and second harmonics were taken into account to represent the semiannual effect. The resulting curve is also shown in Fig. 1. Since Explorer 24 reentered in 1968 near the time Explorer 39 was launched, the harmonic analysis during that year is based on data from both satellites. The

Annual and Semiannual Density Variations in Earth's Exosphere

• e1 o

( O'O 0 i

o

f J J

oj£o

0L6L

I

•T $ " Jt

B°

Je©

•tfo wo

5

V A 110 R z > 80

W7X 100 60

90 40

80 20

7

on top.

347

Absolute EUV Photon Fluxes of Aeronomic Interest

minimum in Fig. 2 must be attributed to atmospheric absorption. The rest of the data is under the combined influence of continuous degradation and irregular solar intensity variations. 3. Interpretation of the Data Since a spin stabilized small satellite was used power limitation allowed only a modest accuracy (0.3—4.5 degrees) of solar pointing of the spin axis to be achieved. The impact of these particular pointing conditions on spectral measurements with planar grating type spectrometers was studied theoretically and experimentally [3] and a modified measuring mode for solar mispointing exceeding 1.5° is therefore applied aboard AEROS. For greater mispointing it was found that an accurate "defolding" of the data by means of an appropriate model computation is rather cumbersome because the instrumental parameters such as the optical angles and the actual position of the spin axis with respect to the on-board reference system must be known with very high accuracy. For occasions of rather good pointing however, this difficult procedure can be replaced by a simple selection type operation. For an orbit section of high altitude and small mispointing a series of spectra of good quality has been selected from 1430—1450 UT of 2 March 1973. For this period the weighted fluxes are given in Table 1; spectral intervals have been selected after aeronomic considerations [4]. With the present uncertainty Table 1 Solar EUV Fluxes Wavelength or range (nm)

Identification

103.76 + 103.19

O VI

7.6

0.145

102.57 99.10 97.70 97.25 94.97 94.45 102.7 — 91.1

HLy-2 N III HLy-3 HLy-4 S VI unresolved

6.5 1.2 7.7 1.3 0.6 0.2 3.4

0.125 0.023 0.157 0.026 0.013 0.004 0.070

102.7 — 91.1

integral

20.9

0.418

5.9 5.2 0.9 2.9 1.4 0.7

0.131 0.118 0.021 0.068 0.035 0.017

17.0

0.390

cm

91.1 - 89.0 89.0 — 86.0 83.42 86.0 - 83.0 83.0 - 80.0 91.1 - 80.0

H cont H cont O II, III H cont H cont unresolved

91.1 - 80.0

integral

Photon flux (1013 m- 2 s- 1 )

Energy flux 1 (mW m- 2 S" )

348

G . SCHMIDTKE, K . R A W E R e t a l .

Table 1 (continued) Wavelength or Range nm

Identification

Photon flux (10 13 m" 2 s- 1 )

Energy flux (mW m- 2 s" 1 )

79.02 + 78.77 77.04 80.0 — 77.0 76.04 77.0 - 74.0 70.34 80.0 - 63.0

OIV Ne V I I I H cont OV H cont OIII unresolved

0.5 0.4 0.9 0.4 0.5 0.4 1.1

0.014 0.009 0.023 0.009 0.014 0.013 0.030

80.0 - 63.0

integral

4.2

0.112

62.97 62.53 60.98 58.43 55.44 52.11 50.4 - 47.0 49.93 46.52 63.0 - 46.0

OV MgX MgX He I OIV Si X I I He Icent Si X I I NeVII unresolved

2.3 0.5 1.1 1.9 0.7 0.3 1.2 0.3 0.2 1.1

0.073 0.017 0.035 0.063 0.026 0.013 0.050 0.013 0.008 0.039

63.0 — 46.0

integral

9.6

0.337

46.0 - 37.0

unresolved

1.0

0.050

46.0 -

integral

1.0

0.050

36.81 36.07 33.54 30.38 28.41 37.0 — 28.0

Mg I X Fe X V I Fe X V I Hell Fe X V unresolved

1.0 0.4 0.8 9.0 0.9 4.1

0.053 0.021 0.049 0.587 0.061 0.250

37.0 - 28.0

integral

16.2

1.021

28.0 — 23.1 23.1 - 20.5

unresolved unresolved

4.5 2.0

0.348 0.178

28.0 - 20.5

integral

6.5

0.526

20.5 17.6 -

17.6 15.5

unresolved unresolved

6.6 0.8

0.711 0.092

20.5 -

15.5

integral

7.4

0.803

90.4

3.8

37.0

104 — 15

integral

(1013 m" 2 S"1 = 109 cm" 2 s"1; 1 mW m~2 = 1 erg cm" 2 s"1)

Absolute EUV Photon Fluxes of Aeronomic Interest

349

of absolute flux values more detailed tabulations could not give better information. The absolute fluxes given here could be in error by about 30—35% for the shorter wavelength region and about 40% for the longer wavelengths. This uncertainty is mainly due to the spectral underground determination; calibration data are certainly better. For the day of these observations the solar activity figures were: Relative sunspot number (Zurich) Rz = 38, Covington index (indicating solar flux at 2800 MHz, Ottawa) F 1 M = 101.6 X 10"22 W m" 2 Hz- 1 (adjusted for a burst). Simultaneous X-ray fluxes measured by the Solrad 10 experiment aboard Explorer 44 were for 0.1 to 0.8 nm: for 0.8 to 2 nm:

120 nW m~2 s- 1 , 7 700 nW m" 2 s"1.

Geomagnetically 2 March 1973 was disturbed (daily sum of Kp = 37, Ap = 40). Note that for medium solar activity conditions the intensities of Fe XV (28.4 nm) and Fe XVI are extremely low. Acknowledgments We express our warmest thanks to Drs Hinteregger, Bedo, Hall and Manson of AFCRL, Bedford, USA for helpful discussions concerning the instrumental design. We also thank Drs. Thimm, Drerup and Husmann of Physikalisches Institut of the University of Bonn for their help with instrument calibration. This work was sponsored by Bundesministerium für Forschung und Technologie (grant W R K 99, 192, 226, and 254). References [1] G. SCHMIDTKE, W . SCHWEIZER a n d M. RNOTHE, J . G e o p h y s i c s 4 0 , 5 7 7 (1974).

G. SCHMIDTKE, K . R A W E B , T H . FISCHER and W . LOTZE, Space Research X I V , 1 6 9 ( 1 9 7 4 ) . [3] G. SCHMIDTKE, K. SCHMIDT and C. REBSTOCK, Rep. BMFT-FB W 73-20 (1973). [4] H. E. HINTEREGGER, L. A. H A L L and G. SCHMIDTKE, Space Research V, 1175 (1965). [2]

Space Research XV — Akademie-Verlag, Berlin 1975

MASS S P E C T R O M E T E R M E A S U R E M E N T S OF T H E F 2 REGION ION COMPOSITION FROM T H E S A T E L L I T E COSMOS 274 Y u . A . R o M A i r o v s K Y , V . Y . K A T Y U S H T N A a n d V . G . ISTOMIN

Hydrometeorological Service of the USSR, Moscow, USSR Institute for Space Research, Moscow, USSR

Data from mass spectrometer measurements of the ion composition of the F2 region aboard the satellite Cosmos 274 are reported. The ions H+, N+, 0+, N 2 + , NO+ and 0 2 + were regularly recorded. In the quiet ionosphere the troughs of atomic ions are found in the dawn subauroral region (e.g. the "main trough") and in mid- and low latitudes in the northern hemisphere at post-sunset hours. In troughs the relative concentration of molecular ions is increased. The molecular ion distribution near the F layer peak is presumably controlled by photochemical and 0+ dynamic processes. In storm periods the troughs of atomic ions deepened and new troughs appeared; both absolute and relative concentrations of molecular ions in the F2 region increased. A dawn-dusk asymmetry of the storm-time ionosphere in response to the dawn-dusk asymmetry of the neutral atmosphere is reported.

1. Description of the Experiment The mass spectrometer experiment on the satellite Cosmos 274 was conducted from 24 to 31 March 1969. The satellite was launched into an orbit with apogee 320 km, perigee 220 km, inclination 65°. The apogee was at or above the nighttime and pre-dawn F peak but the perigee was below the daytime P peak. The measurements were carried out with a modified Bennett radiofrequency mass spectrometer MX-6407 P [1]. The sensitivity of this device in the ion mode was ~ 30 cm - 3 ; the mass scan (1—48 amu) took 25 seconds. 2. The Undisturbed Ionosphere The ion species H+, N+, 1 6 0 + , 1 8 0 + , N2+, NO+ and 02+ were regularly recorded in this experiment. A typical distribution of ion composition along the satellite orbit is plotted in Fig. 1. The atomic and molecular ions are distributed in different manners. The molecular ion concentrations depend presumably on height but the height dependence breaks down in the subauroral region. Atomic ion troughs are observed at dusk in low- and middle latitudes and at dawn in subauroral regions; the latter is identified as the "main" trough in the distribution of ionization [2]. Between the troughs a night-time maximum of atomic ions at the geomagnetic equator is found.

352

Y t r . A . ROMANOVSKY, V . V . KATYUSHINA a n d V . G . ISTOMIN

An increase of the relative concentration of molecular ions toward the auroral latitudes is observed; in the troughs the enhancement of molecular ions is as large as 50%. In the subauroral troughs both absolute and relative concentrations of molecular ions are increased; however, in the mid- and low latitude troughs only the relative concentration is increased.

m ,2

10'-

_

yt . v

\

^ J

a

!i w

N 30' 160° IV 1-jO" 60° N ^6°

115° 7° GE

105° 7p°f £ 20° S 48°55° S

33° Ail

240

2Ç0

280

3\0

315290

250

3.5

7,5

2.1 3^2,7

IS

Zpo

23p0

115

11 015

W

Hf 8J0

H (km) L LT

Fig. 1. The variations of the F2 region ion composition along the Cosmos 274 satellite orbit. GE denotes the geomagnetic equator.

To understand the physical mechanisms controlling the molecular ions in the day- and night-time F2 region the experimental data on the absolute concentrations of molecular ions 0 2 + , NO+ and N2+ have been compared with theoretical calculations for the distribution of these ions. Figs. 2 a and 3 a show data for the night-time F2 region and Figs. 2 b, 3 b and 4 show results for the daytime. The theoretical plots were calculated for photochemical equilibrium using the experimental data on the neutral composition [3] for pass 16 and Jacchia's model 1971 [4] for pass 112, the data on UV-flux [5] with corrections [6] and data on diurnal variations of TD and Te in the mid- and low-latitude F2 region [7, 8], The scheme

Mass Spectrometer Measurements of F2 Region Ion Composition

353

of photochemical processes used for our calculations follows t h a t given b y Danilov and Vlasov [9]. The agreement between the experimental and theoretical curves in Figs. 2 a and 3 a (night-time bottomside ionospheres) suggests that the molecular ions for

a)

Night

b) Day 4.0

3.5

3.0

2.5 rœ-momo°K Tœ- 970°K T-850°K

2.0F 1.5

h i i i 220

I L 240

260

niih experim. data Jacchia, 1971

2.0 J I L 280 300

H,km

1.5

; L I 220

I I I 240

250

I I

»

280 H,km

Pig. 2. 0 2 + ions by night (a) and day (b) for undisturbed (pass 112) and storm (pass 16) iononospheres. The solid lines show 0 2 + values calculated for pass 16 using experimental data on neutral composition. The dashed lines show 0 2 + values calculated using Jacchia model 1971 [4]. log NO1

aI Night

4.0

cPo NO

3.0

2.5

o
NO+ + N using experimental data on [N2] and the dashed line gives NO+ values from the reaction 0+ + N 2 N0+ + N. 23

Space Eesearch XV

354

Yu. A. Romanovsky, V. V. Katyushina and V. G. Istomin

both undisturbed and storm conditions are controlled by photochemical processes. The daytime 0 2 + ion data are plotted in Fig. 2 b. The dashed line corresponds to 0 2 + concentration values calculated using Jacchia's model [4] and the solid line to 0 2 + with 0 2 values two-thirds of those given by Jacchia's. The discrepancy found may be explained by the decrease of 0 2 in the daytime thermosphere as a result of photodissociation of 0 2 not considered in Jacchia's model [4]. The

r_M.70-io(kfs

J

210

y-1.4 • W~w, Ferguson et.al.

I I I I I I I I 230

250

270

»

290 H,km

Fig. 4. N2+ ions by day in the undisturbed (pass 112) and storm (pass 16) ionospheres. The dashed line shows N 2 + values calculated using y3 from [11] and the solid line using our temperature dependence for y3.

daytime NO+ ion data are given in Fig. 3 b. The contributions to the NO+ content from the reaction 0+ + N 2 -»• N0+ + O are shown by the dashed line, and from the reaction N 2 + + 0 —> N 0 + + N by the solid line. The discrepancy between these curves and the experimental curves shows that N 0 + sources in the daytime F2 region need to be identified. The daytime N2+ ion data are plotted in Fig. 4; N2+ production from the reaction N 2 + 0+( 2 D) -> N2+ + O was considered. I t has been suggested t h a t gO + ( 2 D) is about 0.2gC)+(3P) [10]. A consistent picture of the observations with photochemical theory (solid line in Fig. 4) is obtained if it is assumed t h a t the rate of reaction of N2+ loss depends on neutral temperature as ^ l i x « - ^ ' " . I n Fig. 4 the dashed line shows the N2+ values, calculated with y3 from [11]. The existence of the night-time maximum of atomic oxygen ions at the geomagnetic equator (Fig. 1) suggests the influence of t h e equatorial electrojet in maintaining the night-time equatorial ionosphere. I t is concluded t h a t the 0+ ion distribution at and under the F2 peak is presumably controlled by dynamical processes—diffusion at middle latitudes and [E X B ] drifts the equatorial ionosphere.

3. The Storm Ionosphere A strong geomagnetic storm (Kp = 8) commenced before the experiment on 23 March 1969. Here we consider data on ion composition variations obtained from four satellite passes (16—19) over the period 0720—1505 UT on 25 March

Mass Spectrometer Measurements of F2 Region Ion Composition

355

1969 during a separate magnetic substorm (Kp = 4 + —5) which occurred during the recovery phase of the storm. Changes in the molecular ion distributions 0 2 + , N0+ and N 2 + occurred during the substorm. I n an earlier paper [3] the first observation of the dawn-dusk asymmetry of the thermosphere during a geomagnetic storm was reported. To search for this phenomenon in the ionosphere we determined the difference between the experi-

Pig. 5. The dawn-dusk asymmetry in the disturbed ionosphere. The values of A log [02+] (see text) for pass 19 (circles) lie on the solid line calculated for the photochemical model with experimental data on neutral composition. The phenomenon of dawn-dusk asymmetry is observed in the subauroral region in the undisturbed period (pass 112, squares).

mental (exp) data on 0 2 + ions and the theoretical (th) photochemical values calculated using Jacchia's model: A log [02+] = log [0 2 +] exp - log [0 2 +] th Fig. 5 illustrates the dawn-dusk asymmetry in the ionosphere; the maximum value of A log [0 2 + ] is observed during dawn hours. The solid line shows the difference A log [0 2 +], calculated using experimental d a t a on neutral composition. The fact t h a t the A log [0 2 +] points lie on the solid line suggests t h a t the dramatic increase of 0 2 + in the dawn ionosphere is caused by an analogous effect in the thermosphere. The dashed curve (for pass 112) shows t h a t the phenomenon of dawn-dusk asymmetry occurs in the auroral zone in quiet conditions. The effect observed was thought [3] to take place due to the enhanced precipitation of energetic particles during morning hours. To illustrate t h a t movements of ionization play an important role in F2 formation during ionospheric disturbances the changes of K = yjcn, where y is the rate of ion-molecule reactions forming molecular ions and 100

1

165

2-

[10]

1

121

1+

[5]

II

III

IV

60-70 forenoon 40-50 forenoon 20-30 40-50 afternoon 60-70 afternoon 75-89 afternoon >100 60-70 forenoon 40-50 forenoon 20-30 75-89 afternoon >100 40-50 forenoon 20-30 40-50 afterhoon 75-89 afternoon

5 5

72-161 76-142

2--4+ 1-—2-

[1, 3 - 5 ] [3, 6 - 9 ]

1

88

4+

[2]

1

86

2+

[2]

1

74

0

[7]

1

110

4-

[12]

2

156, 177

2-,3-

[15]

2 2+

[15]

1 1

73 71

[ H ]

3

145-180

0+-3+

[14, 15,17]

3

156-183

1-3-

[13, 15, 16]

4

93-163

0-3-

[14, 17]

2

94-132

0-2-

[19]

160

5-

[19]

3

159 94-125

0-4

[19] [19-21]

3

180-195

1-3+

[20]

2

98, 155

0+, 5+

[24, 25]

1

1

1-

1

105 85

4 2-

[23] [23]

1

72

3+

[26]

1

T e profiles = No. of profiles. ,F10 7 = range of flux (in flux units of 10" 22 W m 2 Hz- 1 ). Kp — range of Kp index variation.

Electron Temperature Variations at 100—200 km from Probe Measurements

371

Second group: Te profiles obtained in Japan (Kogoshima 31°15' N 131 °05' E ; Michikava 39°34'N 140°04' E) [11 — 17] by means of an electron temperature probe [18]. Third group: Te profiles obtained in USSR (Volgograd 48°41' N 44°21' E) [19—21] by cylindrical Langmuir probes [22]. Fourth group: results of Te measurements not included in the three first groups. This group includes Te profiles obtained by different researchers in Japan (by methods other than the electron temperature probe method) [23,24] , in Woomera (31° S 137° E) [25] and in Hammaguir (31° N 03° W) [26]. h.kmC

200 BOO

1400 WOO

200

600 WOO

Fig. 1. Mean profiles of electron temperature of the first, second and third groups (curves 1, 2 and 3 respectively). Mean scatter of individual Te values is shown by a horizontal bar; profiles constructed from one measurement only are shown by asterisks.

Within each group all the results are divided according to solar zenith angles x> forenoon and afternoon data being considered separately. The number of T e profiles in the different zenith angle intervals used, as well as the range of 10.7 cm solar radio flux (F l o : is given in units of 10 -22 W m - 2 Hz - 1 ) and the Kp index are given in Table 1 (the value of solar flux is taken for the day of the experiment and the Kp index for the three hour period corresponding to the experiment. I n order to learn the main features of electron temperature profiles determined by the different measurement techniques, each of the first three groups was considered separately. I n this case all data within each group were classified according to large intervals of zenith angles: 40—70° in the forenoon; 2 0 - 3 0 ° ; 4 0 - 7 0 ° , 9 0 - 1 0 0 ° , > 100 in the afternoon. Mean values of electron temperature T e at fixed altitudes (with an interval of 10 km) were calculated for each interval of Mean profiles obtained in this way for the first, second and third groups are presented in Pig. 1 (curves 1, 2 and 3 respectively). As is seen, the profiles agree satisfactorily except for two zenith angle intervals (•/ = 20—30° 24*

Te,°K

372

Y u . K . CHASOVITIN a n d N . M . K L Y U E V A

and % > 100°). The mean scatter in the data can reach several hundred degrees; however, in the majority of cases its relative value is not more than 3 0 % ; in some cases deviations from the mean can reach 50—55%. For zenith angles of 20—30° and greater t h a n 100° (Fig. l b , f) there are great discrepancies in the profiles, but it should be noted t h a t different Te profiles are conseructed from

Fig. 2. Solar zenith angle depsndence of electron temperature at fixed altitudes. Mean spread of individual Te values is shown by a vertical bar; light points are the Te values obtained from one measurement. single sets of measurements. Therefore the observed discrepancies in these cases cannot be considered to be genuine. I n the 100—120 km height region the mean Te values obtained in the daytime by all methods are more t h a n 400° and u p to 600 °K, i.e. higher t h a n the neutral gas temperature. This is consistent with the results given by Hirao and Oyama [27], who believe t h a t their results are not affected by contamination of the probe electrode. At night, when the electron temperature in the E region is likely to be lower t h a n 400 °K, the errors are significant. Preference should then be given to the results obtained by the electron temperature probe, which gives T e values close to the neutral gas temperature (Fig. I f ) . Since T e profiles for the various groups differ slightly in the majority of cases, they were considered together in order to investigate in detail the dependence of

Electron Temperature Variations at 100—200 km from Probe Measurements

373

electron temperature on solar zenith angle. Data of different groups were combined according to the zenith angle intervals shown in Table 1 and mean values calculated at fixed altitudes (with an interval of 10 km). For % = 20—30° the Te profile from the second group is not taken into consideration and for % > 100° we used only those Tt profiles which refer to this group. Fig. 2 shows the solar zenith angle dependence of electron temperature. Maximum electron temperature is noted for small solar zenith angles (in this case at 20—30°). I t is seen more distinctly as the height of observation increases, and there is some asymmetry between forenoon and afternoon values. Electron temperature variations depending on % (Fig. 2) agree with the mean diurnal variation given by Hirao and Oyama [27]. I t is seen that there is significant scatter in the values (especially at small %). Mean electron temperature profiles for various solar zenith angles together with \ATe\ values characterizing the mean scatter in the individual values are given in Table 2 (the number of Te profiles used for the averaging is given in brackets). For zenith angles of 20—50° \ATt\ can reach several hundred degrees but relative to T e its value is generally not more than 3 0 % ; in some especially unfavourable cases deviations from T t can reach 60%. Table 2 can be considered as being a preliminary and rather approximate empirical model describing electron temperature variations with solar zenith angle in the 100—200 km height region in middle latitudes. The model constructed is referred to a level of mean solar and geomagnetic activity (Fl0 7 130 X 10~22 W n r 2 Hz - 1 , Kp 2). Further accumulation of experimental data will give the opportunity to improve and refine this model to take into account the main features of electron temperature variations over different geographic points.

2. Solar and Geomagnetic Activity Dependence of Electron Temperature in Middle Latitudes The dependence of electron temperature on solar and geomagnetic activity was investigated as follows. In order to exclude the influence of Te variations with solar zenith angle we used calculated deviations of Te values from the mean profile for the appropriate zenith angle interval (AT e — T,. — Te, where Te is the mean value of electron temperature for the given interval '/_). For fixed altitudes the ATe dependence on solar radio flux F x 0 . 7 was estimated; ATt values corresponding to Kp < 2 and Kp Sg 2 were considered separately. The analysis was carried out for each of three first groups. The results obtained at 120, 150 and 180 km (170 km for the third group) show that there is rather weak dependence of ATe on the value of solar radio flux which in some cases is almost completely masked by the large spread of the points. The same effect was noted by Pfister [28] for the height region considered. Results of the first group show that at 120 and 150 km electron temperature decreases with increase of solar activity at a rate of 3—5°K per unit of flux variation. On the contrary, for measurements referring to the second and third groups at 120, 180 and 170 km Te increases as Ft0 7 increases. This increase is 2 — 3 ° K per unit of flux variation, which agrees with the results of Hirao and Oyama [27].

374

Y u . K . CHASOVITIN a n d N . M . K L Y U E V A

J2,

S 2 - J 2 - 2 Í o

o Ci S

S

S

S

+

i

^

T-

o Oí

H

-

-

o

o

CO IC H

-

H

O O O O O O O O O O O O O O T - C I N O

C O Tt* ^ o o>

o os

>a ^

^

^

^

í ]

o o

o «

o ®

o e

-U+L-H-H-H-H-H^^-^ o o o o o o o o o o í C O i C O N C O M c ú C O C D C O • ^ W C Ô t ' O O C O û O O i O î O i

O O I O O O O O O O O W

W

O

C

O

H

^

c

C

M

h

H

-H-H-H-H-H-H-H-H-H-H o i

o i

o

i

o

o i

o

e

o

M

i

o

o

o

œ

o

œ

o

o

'

H

N

( N < N < N < N e O C O ' < J I ' ^ T j ( < N ( N o

o

IO O -H

o

a

o

e UJ

a 3
1500°K) extended neutral exosphere as has been proposed to slow the solar wind down far upstream of the planet (by means of mass loading due to planetary ions borne in the wind) in order to avoid the formation of a bow shock. This conclusion is also consistent with the low exospheric temperatures inferred from both Mariner 10 ( < 4 0 0 °K) [10] and Mariner 5 (600° ± 50 °K) [11] as well as calculated upper bounds on the extent of the neutral atmosphere and ion-exosphere in the solar wind [12]. Additional support for a bow shock is provided by the observation of the highenergy electron enhancements, upstream of the planet, similar to those observed near Earth's bow shock. We suggest that the flux decrease observed at energies above 100 eV, when the spacecraft was between events A and C on its trajectory, was caused by depletion of the electron population on magnetic flux tubes which passed close to the ionopause. Since the cross section for electron impact ionization of helium has a peak near 100 eV and remains high above that energy, penetration into the neutral helium exosphere is a natural way for electrons at these energies to be selectively removed or to have their pitch angle distribution disturbed. This is consistent with the observed flux modulation. Considering the probable dimensions of the interaction region near the ionopause, a helium density of 106 to 107 cm - 3 is sufficient for the high-energy electron depletion observed, and such densities have been predicted at the altitude of the ionopause [7]. The interaction between the solar wind and the Venusian atmosphere appears to resemble in some ways that thought to occur with a comet. This is supported by the possible penetration of Venus and the depletion of high-energy electrons as they pass through the exosphere near the ionopause. In addition, unusual intermittent features unlike those observed in the terrestrial magnetosheath or in the free-streaming solar wind were observed thousands of scale lengths downstream of Venus during the approach of Mariner 10. The following conclusions may be drawn from the data presented here: (i) the interaction of the solar wind with Venus most probably results in a bow shock; the best fit by hydrodynamic models at the time of Mariner 10 is characterized by Hlr 0 = 0.01; (ii) an extended exosphere which slows down in the solar wind * By the therm "stand off" we mean that the dynamic pressure of the wind is balanced by some other force so that a shock is formed and the wind flows around the planet without touching the surface.

Results from Plasma Science Experiment on Mariner 10

509

without a bow shock appears very unlikely; (iii) a direct interaction between the solar wind and the Venusian atmosphere is indicated by the behavior of electrons of energy between 100 and 500 eV; and (iv) Venus probably has a "tail" hundreds of scale lengths long, suggestive of that of a comet.

3. Mercury We now consider the results obtained during the Mariner 10 flyby of Mercury on 29 March 1974. During this time, measurements of plasma electrons were made in the same way as during Venus encounter except t h a t the scan platform rate at Venus was four degrees per second and at Mercury one degree per second. The results, obtained taking the potential of the spacecraft to be equal to space potential, show clearly and unambiguously that the solar wind interaction with Mercury gives rise to a well developed bow shock and sheath region similar to the bow shock and magnetosheath observed at Earth. Inside the sheath there is a region similar to the magnetosphere or magnetotail of Earth which contains a population of electrons whose properties differ from those of the surrounding medium. A detailed discussion of the experimental results on which these conclusions are based is given below. Mariner 10 passed on the dark side of Mercury along a trajectory shown in Fig. 5. Previous to the encounter, the instrument measured typical interplanetary solar wind electron spectra, illustrated by spectrum 1 of Fig. 6. The spectra were similar in form to those taken near 1 AU; they showed separate low energy ("core") and high energy ("halo") components which were both approximately Maxwellian in form and were characterized by temperatures of 1.5 X 10 5 °K and 6 X 10 6 °K, respectively [13, 14]. Fig. 7 shows fluxes at 13.4, 71, and 389 eV, and the density and pressure as a function of time for the period between 2000 and 2200 UT Earth-received time (ERT). A scale with the events referred to the time of closest approach is given at the bottom of the figure, and the UT of spacecraft observation across the top. The density n and pressure p are defined by and respectively, where the integrations have been carried out numerically over the whole energy range of the detector, and extrapolation from 13.4 eV to zero has been made, assuming a Maxwellian form for the distribution function. Modulation of the derived density and pressure by the solar wind flow velocity has not been removed. The peak values of the scan modulation of the density and pressure are the most representative of ambient conditions. From the variation with scan angle of electron flux at 13.4 eV at a time some 20 minutes before encountering the bow shock, we have determined a solar wind velocity of 630 ± 40 km s _1 , and a mean flow direction in a frame moving with the planet of 3 ^ 6° from the west of the sun. The results of the M. I. T. plasma experiments on IMPs 7 and 8 at 1 AU indicate bulk speeds (corrected for propagation and co-rotation) which agree well with this value. The pre-encounter conditions are n = 17 ± 2 cm - 3 and a ram pressure QV2 = 1.1 X 10~7 dyne cm - 2 . We now describe the events illustrated in Fig. 7. On the incoming leg of the trajectory, the bow shock was traversed three times within one minute, begin-

510

H . S . B R I D G E , A . J . LAZARUS e t a l .

Fig. 5. The trajectory of Mariner 10 at the time of encounter with the planet Mercury. Distances are in planetary radii, and the Xsw axis points in the anti-solar wind direction, taken to be three degrees to the west of the sun. The Zsw axis is to the north in this right-hand coordinate system.

ev Fig. 6. Electron spectra at various times, given in the text, during the encounter. Spectrum 1 was taken in the interplanetary medium before reaching the bow shock; 2 in the magnetosheath; 3 and 4 in the magnetosphere ; and 5 between energetic particle events B and C, just before re-entering the magnetosheath. A typical background as observed in the magnetosphere is shown; the background in the interplanetary medium is several times lower.

512

H . S . B R I D G E , A . J . LAZABUS e t a l .

ning at —19 min:38 + 6 sec and ending at —18 min:49 i 6 sec in agreement with the corresponding magnetometer observations [15] to within the resolution of the electron spectrometer. The shock crossings are clearly evident from the increases in flux in the 71 and 389 eV channels and from the related increases in density and pressure, see Pig. 7. This thin shock structure closely resembles the Earth's bow shock observed by an electron detector when the interplanetary magnetic field is approximately perpendicular to the direction of the shock normal ("perpendicular" shock) [13, 14]. Across the discontinuity there was a density change of approximately two times, and a temperature change of approximately three times, in satisfactory agreement with observations of density and temperature jumps by electron detectors which have traversed Earth's bow shock at comparable locations [13, 14]. After the shock traversals, we observed a region analogous to Earth's magnetosheath, where the flux of electrons at, for example, 71 eV was greatly increased. Spectrum 2 of Pig. 6, takenin this region, does not show the " f l a t " form at low energies characteristic of electron spectra taken in Earth's magnetosheath [13, 14], but does show the large increase of electron flux at moderate energies resulting from "thermalization" behind the shock. In the inner sheath, traversed between —17 and —13 minutes, there was a reduction in flux above about 100 eV which varied with the direction of pointing of the detector. The observations made during that period will be discussed below. At —10 minutes there was a well defined change in spectral form (from that shown in Pig. 6 as spectrum 2 to that shown as spectrum 3), accompanied by a drop in density to about 1 cm - 3 . There was an abrupt increase in electron flux between 200 and 680 eV at this time. Comparing these plasma data with similar observations made near Earth [16], we conclude that the spacecraft crossed a boundary analogous to a magnetopause. This interpretation is strengthened by changes in the magnetic field observed at the same time [15]. Fluctuations in the data limit the accuracy with which the time of crossing can be determined to about one minute. The spacecraft passed into the optical shadow of the planet at —4min:48sec, after which the electron spectra gradually assumed the form 4 in Fig. 6 as a result of increasing electron fluxes at all energies. Passage out of the shadow of the planet occurred at + 2 min:39 sec. A second traversal of the magnetopause boundary occurred at approximately + 7 minutes when the density and spectral form became similar to those observed immediately after the shock crossing on the inbound leg of the trajectory. Between + 12 minutes and + 1 7 minutes the spacecraft passed through a highly disturbed region, which we interpret as a pulsating ("parallel") shock [17—19]. This interpretation is consistent with the nature of the incoming shock, the geometry of the orbit and shock boundary, and the measured direction of the magnetic field before and after the encounter [15]. I n this region the plasma properties varied rapidly with time, and it is possible that at least some of the electron spectra are inaccurate because of possible large fluctuations during a single measurement sequence. After this shock passage, typical solar wind spectra and fluid parameters were observed; in addition "upstream events" were seen as indicated in Pig. 7. These events are qualitatively similar to those recorded by electron detectors situated upstream of Earth's bow shock on magnetic field lines intersecting the bow shock [20, 21]. We note that increases in the 71 eV flux before encounter (e.g., at —26 minutes) appear to be different in nature from those observed after encounter; the former coincide with transient decreases in the interplanetary magnetic field [15]. The presence of upstream events is consistent

Results from Plasma Science Experiment on Mariner 10

513

with our interpretation of the outgoing shock as having parallel geometry. We show in Fig. 8 a comparison between the present observations and observations of a parallel shock at Earth made by the triaxial electron spectrometer on OGO 5, which had a similar time response to the present instrument. The time scale according to which Mercury data have been plotted has been adjusted by the ratio [c/(wpe Fsc)]E/[C/(cope FSC)]M to take account of plasma conditions and the speed of the observer F s c . The similarity between observations at Mercury and Earth is quite striking. At the times marked on Fig. 7 by the letters A, B , C, and D, the University of Chicago energetic particle experiment [22] recorded high intensity bursts of

DECIMAL HOUR

Fig. 8. Observations of a parallel shock made at Earth by OGO 5, compared with the present observations at Mercury. A scaling factor of [c/(a>p e F sc )] B /[c/(co pe F s(! )] M has been introduced to compensate for the different environments and spacecraft velocities F s e when observing the two shocks.

energetic electrons of short duration. During bursts B and C, energetic protons were also identified. The response of the plasma instrument showed no change during events A and D, as regards both fluxes and spectral shape. Changes in these quantities did take place during events B and C. After event A, the spectrum was similar to spectrum 4 of Fig. 6. Shortly after event B began (at + 1 min:31 sec), the counting rates in our intermediate energy channels fell to the background value, leaving only low rates in the energy channels below 20 eV and in channels above 389 eV. The spectrum then relaxed to the form shown as 5 in Fig. 6 and retained this shape and approximately the same intensity until + 6 min:43 sec in the middle of event C. There were then two spectra in which the counting rates at intermediate energies were again reduced to their background values. Then the spacecraft entered the magnetosheath, and the spectrum returned to the form 2 of Fig. 6. Having described the major features of the observations, we now consider features of particular interest in more detail and attempt to interpret them in terms of the interaction of Mercury with the solar wind. First, let us consider the possibility that the interaction is with a neutral atmosphere or with an ionosphere as discussed above for the case of the interaction at Venus. In contrast to the Venusian atmosphere, the atmosphere of Mercury is thin or "exospheric", i.e. the mean free path for collisions is greater than the atmospheric scale height. The ultraviolet experiment [23] gives an instrumental background upper limit on the column densities of 3 X 10 13 cm - 2 for Ne, 10 13 cm - 2 for Ar, a measured 33

Space Research X V

514

H . S . B R I D G E , A . J . LAZARUS e t a l .

upper limit of 7 X 1011 cm -2 for He and other column densities with lower values. In addition, an upper limit for the total pressure P j at the terminator, of 2 x 10~9 mbar was inferred from the ultraviolet occultation experiment. To determine whether the atmosphere alone can stand off the solar wind, we consider the change in the velocity of the wind from its free streaming value v to its value v' after it has been slowed down by atmospheric mass addition at the rate nfifJjNi (where w,- is the mass, JI is the ionization rate, and JV,- is the column density of the ith atmospheric constituent which behaves as a fluid). For this case, the one-dimensional continuity, momentum, and energy equations give: v v'

1 (y - 1)»

y+ 1 y y— IT

%WT' QV

2

where y is the specific heat ratio and o is the solar wind mass density. This equation is consistent with the high Mach number limit obtained previously [24—26]. The plus sign corresponds to the formation of a shock; the minus sign refers to no shock. The net ionization rates at the orbit of Mercury (due to photoionization, impact, and charge exchange with the solar wind) are J H e = 2.2 x 10-7 s _1 , JAT = 2.6 X 10 -6 s - 1 , and J N e = 5.2 x 10 -7 s -1 . The maximum mass addition rate preceding the formation of a shock occurs when the argument of the square root vanishes. We note that this fluid equation is realistic only when the ion gyroradius is smaller than the scale of the system, planet plus atmosphere. Following the above, we find that He cannot stand off the solar wind. Since the gyro-radii for Ne and Ar ions are much larger than the scale of the system (the mass loading by Ne and Ar ions is only important within 60 and 30 km, respectively, above the surface), the microscopic point of view must be considered. The time it takes the solar wind to pass through this small part of the atmosphere is of the order of tenths of seconds while the time to accelerate Ne or Ar ions to solar wind speeds is of the order of tens of seconds. Therefore the mass addition by Ar and Ne ions is small and the momentum change of the solar wind in traversing these distances is negligible. Consequently the solar wind, under this hypothesis, would be expected to strike the surface and be absorbed without forming a shock. However, the upper limits on the neutral gas would permit a strong limb shock [27]. A limb shock seems to be excluded by the observed locations of the shock crossing since they occur too far upstream. Therefore we conclude that mass loading of the wind by the atmosphere cannot account for the observations and that the solar wind must be deflected by a magnetic field. One possible source of the magnetic field might be an ionospheric current system induced by the solar wind flow past the planet as seems to be the case at Venus. However, if this were the case at Mercury, the pressure of the Mercurian ionosphere would have to be sufficient to stand off the solar wind. Based on the results of the radio occultation experiment [28], the upper limit of ionospheric electron density is 103 cm -3 . Using this value an ionospheric temperature of > 105°K is required to balance the ram pressure of the solar wind. Such a temperature is untenably high. The most likely source of the interaction at Mercury is thus a planetary magnetic field, either intrinsic or induced by the solar wind. It is instructive to consider the simplest possibility and assume that the magnetic obstacle is a dipolar field. Using a theoretical model [29] for the locations of the terrestrial magnetopause and bow shock, we find the corresponding boundary crossings observed at Mercury can be fitted by varying the strength and location

Results from Plasma Science Experiment on Mariner 10

515

of a planetary dipole. I n fitting to the model we assume the dipole axis to be perpendicular to the local orbital plane of the spacecraft, and that the Ysw—Zav/ trace of the trajectory passes through the origin (Fig. 5). To a good approximation, this plane also contains the solar wind velocity vector, which is along —XaK.

6

6

Pig. 9. The upper section shows the trajectory of Mariner 10 in a coordinate frame moving with the planet for the case of solar wind flow from a direction seven degrees west of the sun. This flow direction is that of the X s w axis. A-A and B-B represent scaled magnetopause and bow shock boundaries, A for the incoming bow shock crossing occurring one minute before B. The lower section shows B'-B', the shock and magnetopause corresponding to solar wind flow from a direction three degrees west of the sun, and the same shock crossing time as B.

Fig. 9 shows the result of such a scaling for three cases: A, using the first inbound shock crossing, the two magnetopause crossings, and a solar wind flow 7° west of the sun; B, using the last inbound shock crossing, the two magnetopause crossings, and a solar wind flow 7° west of the sun; and B', the same as B but with the solar wind flow 3° west of the sun. For case A, we find that the distance from the center of the planet to the nose of the magnetopause is 1.6i?M and the dipole is located at Xsw = 0.4RM, F s w = 0, Z sw = 0. Corresponding values for case B are: nose distance = 1.25 RM, dipole at X sw = — 0.15J?M, Ysw = 0, = 0; and for B ' : nose distance = 1.9i?M, dipole at X3W = 0.55ii M , F3W = 0, Zsw = 0. If we consider the zero degree flow, we find that the dipole must be located off the Xsw axis with a positive Y3W coordinate. These fits give a range 33*

516

H . S . B R I D G E , A . J . LAZABUS e t a l .

of dipole strengths from 4 x 10 - 4 to 9 X 10~4 times that of Earth. These dipole model fits clearly demonstrate the impossibility of a unique and accurate determination of the locations of the shock and magnetopause from the observations made during this single flyby, because of the sensitivity of this quantity to small variations in shock crossing times and flow direction. Furthermore, the dipole may be tilted in a direction different from that assumed, or situated off the Xsw axis, or both; or the model may be inadequate. A preliminary estimate is that the present experiment can only determine the distance to the magnetopause nose to be less than 2.5 planetocentric radii. In the above discussion we have not considered the origin of the magnetic field which causes the plasma interaction observed at Mercury. In general, there are three possible sources for the field: (i) an intrinsic planetary field; (ii) a steady state induction process (unipolar generator); and (iii) an induction process arising from changes in the direction and magnitude of the incident magnetic field. Unfortunately, data from the present experiment do not allow an unambiguous choice among these possibilities. Although the induction processes have been extensively discussed in the literature, a definitive three-dimensional model is still lacking (see, for example, [30—34]). For any model which involves an induced magnetic field, some fraction of the incident plasma flow must contact the surface of the planet; hence in this case the nose of the "magnetopause" should coincide with the surface of the planet. While, in principle, the position of the nose can be calculated from the measured boundary positions and from an adequate theory, as noted above, the present measurements lead to large uncertainties in the calculated positions even assuming that the theoretical models were complete. Given the experimental and theoretical uncertainties, no definite conclusions can be drawn concerning the origin of the magnetic field. We consider next some detailed plasma features or events. The fluxes of electrons with energies greater than ~ 100 eV show interesting variations in the inbound magnetosheath, as illustrated by the 389 eV data field in Fig. 7. On the inbound leg of the trajectory the flux increases by approximately a factor of 102 on crossing the shock into the magnetosheath and remains high for about five minutes. I t then decreases to about twice the solar wind value for one minute, increases by a factor of 10 for two minutes, decreases to slightly greater than the solar wind value for three minutes, and remains low until the magnetopause crossing at —10 minutes. These variations might be temporal; however, the two intervals of decreased flux coincide with the times when the scanning angle of the instrument is directed toward the downstream shock (the instrument looks furthest from the planet), and the increased flux intervals coincide with scanning angles directed away from the downstream shock (the instrument looks closest to the planet). The variations, therefore, probably represent a directional asymmetry in the particle flux. This could be either a pressure or a streaming anisotropy directed downstream along draped field lines. The second possibility is more likely since the stronger upstream bow shock is a more intense source of electrons than the downstream bow shock. A different scan modulation of the flux of the > 100 eV electrons was observed by the same instrument in the magnetosheath of Venus, where the highest fluxes were observed at scan directions furthest from the planet. At Mercury the maximum flux corresponds to electrons coming away from the upstream bow shock. A similar interpretation could be made for the anisotropy observed at Venus since the flux is a maximum when the instrument is viewing along a field line

Results from Plasma Science Experiment on Mariner 10

517

away from the planet and towards the point where the line crosses the bow shock. However, in the intervals of decreased flux at Venus, the fluxes were much less than the ambient solar wind fluxes, requiring an explanation in terms of an anisotropic electron removal rather than an anisotropic electron source. The figure indicates that removal by interaction with the dayside atmosphere of Venus gives an explanation with the right asymmetry. In the outbound magnetosheath of Mercury, variations of the fluxes of energetic electrons also occurred, but these do not coincide with particular scan directions nor with the scan period. They are interpreted as due to time-dependent generation of energetic electrons by the pulsating shock encountered during the outbound passage through the sheath. We note the absence of intervals during which the fluxes of energetic plasma electrons decreased to values less than ambient solar wind values. These events occurred at Venus and were interpreted as atmospheric absorption; they were not seen at Mercury. This absence of absorption is consistent with our earlier conclusion that the atmospheric interaction at Mercury is weak. The changes in plasma properties observed in the Mercury magnetosphere could be either spatial features or temporal events. A single flyby does not permit a unique interpretation. The combination of highly structured plasma electron data, the magnetic field variations [3], and the energetic particle events observed in the magnetosphere suggest that a time-dependent interpretation is a reasonable possibility. If the magnetosphere is either induced or intrinsic, changes in the orientation of the external field can cause dramatic changes in structure. In the case of an induced field, this is obvious since the induced field must change as its driving electric field ( = — VxB or dBj8t) changes. For an intrinsic field of which the Earth's magnetosphere is the best studied example, we know that rapid time changes occur during events known as substorms. There are some striking similarities between the Mercury observations interpreted as temporal events and substorm phenomenology in Earth's magnetosphere. To indicate that a substorm interpretation is a reasonable one to consider, we estimate the possibility of a substorm occurring in the sixteen minutes Mariner 10 was in the Mercury magnetosphere, based on a scaling of the Earth's magnetosphere. The relevant time scale is the "convection time" given by the time to cycle all of the magnetic flux FT in the tail under the action of the convection electric potential. The convection electric field varies from essentially zero when the interplanetary field is oriented antiparallel to the planetary dipole moment to a maximum when the interplanetary field is oriented parallel to the planetary dipole. The maximum value is approximately fpt = Vsv/BSWBM where F s w and J5SW are the ambient solar wind speed and field strength and /¿M is the scale size of the magnetosphere. If B T is the field strength in the tail, the maximum convection electric field gives a minimum cycle time of T0 = Fil 80 MeV for the region R < 13 Jovian radii. We discuss particle absorption by the inner Jovian satellites. It is shown that the particle-satellite collision time, the radial diffusion coefficient, and the resulting absorption probability are expected to be functions of particle energy and species, in qualitative agreement with our observations. We also discuss possible causes for the observed decrease of proton flux at R < 3.5 Jovian radii.

1. Introduction The existence and some characteristics of trapped relativistic electrons and the Jovian magnetic field were deduced from synchrotron models for the decimetric radiation observed at the earth [e.g. 1,2]. However, information on the trapped protons and non-relativistic electrons has only recently become available as a result of direct measurements by experiments on Pioneer 10 (preliminary reports of these results were published in the 25 January 1974 issue of Science). In December 1973 Pioneer 10 became the first spacecraft to traverse the magnetosphere of Jupiter. The University of California at San Diego (UCSD) experiment on Pioneer 10 measured protons with energies Ev 80 MeV and electrons with energy thresholds in the range 0.1 Ee sS 35 MeV. Some aspects of the results of this experiment, including the description and characteristics of the instrument have been published [3, 4]. The analysis in this paper was performed by using the magnetic field model derived from direct measurements on Pioneer 10 [5]. In particular, the values of the magnetic shell parameter L (= R cos -2 A), where X is the magnetic latitude and R is the jovicentric distance in units of the planetary radius Rj, are a function of the magnetic field model. In the present paper we restrict our analysis to the region R < 157?J; where a dipole representation of the magnetic field is usable [5]. Based on particle measurements, the Jovian magnetosphere is also naturally divided into two regions with the boundary at R ~ 20RJ [e.g. 3, 4]. In §2. we present radial profiles of electrons with Et > 0.16 MeV, Ee > 9 MeV, and Ee > 35 MeV and of protons with Ep > 80 MeV. In § 3. we develop the concepts needed for an analysis of the interaction of the Jovian satellites with the

522

A. Mogro-Campero, R. W. F u l i u s and C. E. Me I I wain

trapped particles, and we show why the published treatments of particle absorption by Jupiter's satellites are inadequate for some of the energies and values of L for which we have measurements. We find that there is a considerable variation in the drift frequencies for the particles we measure, and we discuss particle absorption by the innermost Jovian satellites with an emphasis on the effects due to the differences which can be expected as a function of particle energy. We show that the qualitative characteristics of the particle-satellite interactions which we observe can be understood in terms of these ideas. I n § 4 we discuss possible causes for the observed decrease of proton flux at L < 3.5. We conclude that loss due to pitch angle scattering and absorption by Amalthea are possible explanations, but that there is insufficient information in our data to decide between these.

2. The Data The data reduction procedures, including calculations of the geometrical factors and background corrections when necessary have been described in [4]. To illustrate the structure in the radial profiles of electrons we have chosen three integral flux measurements {Ee > 0.16 MeV, Ee > 9 MeV, and Ee > 35 MeV), as shown in Fig. 1. Although the satellites revolve around Jupiter at a fixed jovicentric distance, the non-centered and tilted position of the magnetic dipole with respect to jovigraphic coordinates results in a spread in the L values traversed by each

13.03

13

II

9

7

5

L IN JUPITER 3 3

RADII 5

7

II

13

TRAPPED ELECTRONS IN JUPITER'S RADIATION BELTS MEASURED BY THE UCSD TRAPPED RADIATION DETECTOR ON PIONEER 10 r„n„n, EUROPA EUROPA l_

14

E >0.16 MeV,

16 18 20 22 HOURS IN DEC. 3, 1973

4

6

8

10

12

14

HOURS IN DEC. 4, 1973

Fig. 1. Radial profiles of electron fluxes in the inner Jovian magnetosphere. The hours are in spacecraft time. The shaded bands indicate times when Pioneer 10 was on L shells traversed by Jupiter's satellites.

Electrons and Protons in Jupiter's Radiation Belts

523

satellite as shown in the figure. As can be easily seen, some prominent features in the radial profile seem to be due to absorption of particles by satellites, but it is immediately evident that all energies do not behave in the same manner (e.g., at Europa electrons with Ee > 0.16 MeV are affected strongly, whereas those with Ee > 9 MeV seem to pass by with no effect). We discuss why an energy dependence would be expected in the next section. 74AM-4-45

2

3

4

5

6

LIRj)

Fig. 2. The integral flux of protons > 80 MeV as a function of L.

We also see from Fig. 1 that the energy spectrum flattens toward lower L values, since with decreasing L the fluxes of higher energy electrons are rising at a faster rate than those of lower energy electrons. The peaks in the lowest energy electrons inside of Io on both the inbound and outbound trajectories are suggestive of acceleration effects at Io as proposed, for example, in [6], In Fig. 2 we show the proton flux (Ep > 80 MeV) as a function of L. The inbound pass corresponds to the higher fluxes, and the data gap on the outbound trajectory is due to occultation of the spacecraft by Jupiter. These data were stored in the spacecraft and transmitted to earth after occultation. Although the reduction procedure has not been completed due mainly to timing problems, it is clear from the occultation data that a good approximation to the actual profile is a straight line connection for the missing portion in the figure. Therefore, it is established that the narrow peak observed inbound at L ~ 3.5 occurs also on the outbound trajectory, where the reduction in flux can be understood in terms of magnetic latitude (i.e. the inbound pass was closer to the magnetic equator). In § 4 we discuss possible reasons for this dramatic decrease in proton flux for L < 3.5.

524

A . MOGRO-CAMPERO, R . W . F I L L I U S a n d C . E . M C I L W A I N

3. Particle Absorption by Jupiter's Satellites I t appears likely that the radial diffusion of trapped particles in the inner region of the Jovian magnetosphere may be driven by electric fields associated with the upper atmosphere dynamo which is driven by neutral winds in the ionosphere [7, 8]. I t seems natural then to consider motion in an inertial reference frame, in which steady state winds in the planet's upper atmosphere produce electric field fluctuations at zero frequency [7].

L(Rj)

Fig. 3. Drift frequencies of protons and electrons of selected energies as a function of L in an inertial frame. The ranges of L values traversed by the satellites are indicated at their appropriate rotation frequencies (JV is Amalthea, JI is Io, JII is Europa, and JIII is Ganymede). The dashed line corresponds to negative frequencies (i.e. opposite to the direction of planetary rotation).

I n this frame, the frequency of rotation around Jupiter of protons and electrons of different energies is shown in Fig. 3. This drift frequency is a combination of the usual drift frequency produced by the curvature and the gradient of the magnetic field and the planetary rotation frequency (assuming co-rotation). We see that for both species drift frequencies at a fixed energy are represented by straight lines which intersect L = 0 at the planetary rotation frequency, and that larger slopes correspond to higher energies. The range of L values traversed by each of the innermost four satellites is shown in this figure at the rotation frequency of the satellite. The range of L values for each satellite is a measure of the radial extent

Electrons and Protons in Jupiter's Radiation Belts

525

of its "sweeping region". For the electrons there is a characteristic energy Et at a fixed L value at which the drift frequency / is zero, and at the average L value traversed by a satellite there is an energy Ea at which the particle drift frequency equals the satellite rotation frequency. At the average L values of the satellites these electron energies in MeV are (Et, Es): Amalthea (90, 14.5), Io (39, 30), Europa (24, 21), and Ganymede (15, 14). Theoretical treatments of Jovian trapped particle diffusion and their interactions with the satellites [8—10] have been limited to particles whose drift frequency equals the planetary rotation frequency, and therefore, as illustrated in Fig. 3, they may be inapplicable at the higher energies for which we have measurements. As noted in [9], because of the tilt of the magnetic dipole axis with respect to the planetary rotation axis, particles with large pitch angles will have less probability of being absorbed by the satellites. An examination of pitch angle distributions both before and after the traversal by Pioneer 10 of the orbit of Io reveals that this effect is occurring, but we postpone a detailed discussion of this effect and others related to pitch angle distributions to a future publication. We also defer treatment of the effects of radial and latitude excursions of the satellites, since the preliminary magnetic field model [5] may soon be improved (E. J . Smith, private communication). These latter effects have not been dealt with in the theoretical treatments, but have recently been considered in [11]. We proceed by exploring the consequences of the frequency distribution illustrated in Fig. 3, realizing that the effects we are omitting and which are mentioned above may modify this simplified treatment. Two characteristic parameters determine particle absorption by satellites: the time Ta taken by the particles to traverse the sweeping region of the satellite, and the time taken by the particles to impact the satellite once they find themselves in its sweeping region. The probability of absorption by a satellite is high for Ts Ti, and low for Ta. I t will be shown that the probability of absorption of electrons has two minima. We assume that Ti is inversely proportional to the relative drift frequencies between the particles and the satellite. This means that for electrons Ti will be very large at energies near Es, the energy at which the particle drift frequency equals the satellite rotation frequency. For radial motion, Ts a -D-1, where D is the radial diffusion coefficient, so that the absorption probability will have an additional energy dependence if the diffusion coefficient is a function of energy, as it is likely to be. As an example of the energy dependence of D, we proceed as follows. The diffusion coefficient for motion produced by fluctuating electric fields is proportional to the power spectrum of these fields at the drift frequency [12]; therefore the diffusion coefficient can be constant only if the fields have a white spectrum. A frequency dependence of the power spectrum will result in an energy dependence for Ts. As an example, we assume a power spectrum which is inversely proportional to the square of the frequency. I n this case the diffusion coefficient for electrons will have a maximum at Et, the energy at which the drift frequency is zero. Other forms of the power spectrum might be preferable, for example a maximum might be expected at the planetary rotation frequency. However, for simplicity, in this paper we consider only the case described above. I n Fig. 4 we show T^ and Ts as a function of energy for electrons at L = 9.5 (the average location of Europa) with the assumptions stated above. Since the time scale is independently arbitrary for Ta and T{, we have fixed their relative positions so that low energy electrons will suffer some absorption (Ts > Tt),

526

A . MOGRO-CAMPEBO, R . W . F I L L I U S a n d C. B . M O I L W A I N

whereas high energy electrons (3 MeV < E e < 42 MeV) will not, as is observed (Fig. 1). ~ Fig. 4 shows t h a t large variations in and Ta can be expected. Our assumption for the frequency dependence of the power spectrum is not critical, since even a constant Ta would lead to a rather broad energy window of easy particle access across the satellites, with absorption at the low energies. We find t h a t at Io, a

Pig. 4. The absorption time and the radial diffusion time Ts as a function of energy for electrons at the average L value traversed by Europa. The highest absorption probability occurs for Ts > at the energies shown at the top of the figure. similar analysis results in the energy window for little or no absorption moving to higher energies (since both Et and Ea are larger t h a n a t Europa). Therefore, at Io our electron observations shown in Fig. 1 can again be understood qualitatively. Our observations near Amalthea can also be made consistent with these ideas by a similar analysis. I t seems reasonable t h a t both T; and Ta should be the same for protons and electrons in the low energy limit, so t h a t once the positions of Ta and T[ are fixed with respect to each other to satisfy the observations at a given electron energy, qualitative predictions can be made for electrons and protons of all other energies.

4. The Maximum in the Radial Profile of Protons I n Fig. 2 we have seen t h a t the protons with Ev > 80 MeV decrease abruptly at L < 3.5. I n this section we consider possible causes for this decrease. These causes have also been recently considered in [11] with respect to their data. If the decrease in protons is produced by a loss mechanism, this mechanism will become effective for TL < Ts, where TL and Ts are the characteristic loss and source times. If the source of particles is radial diffusion, we may use T s ~ Rj2jD.

Electrons and Protons in Jupiter's Radiation Belts

527

For loss due to pitch angle scattering into the loss cone, TL ~> 4L 4 i?j/a [13], where a is the velocity of the particle. If we use for a the velocity of an 80 MeV proton, and L = 3.5, we obtain an upper limit for the diffusion coefficient, D < 2.2 X 10 -3 iJJ2(s~1). This upper limit is ~ 5 orders of magnitude greater than the diffusion coefficient derived in [8] for this L value. However, we point out that the diffusion coefficient derived in [8] to explain the observed Jovian decimetric radio emission is only valid for particles with drift periods near the planetary rotation frequency; the drift frequency of 80 MeV protons at L = 3.5 is a factor of ~ 3 greater than the planetary rotation frequency. For Jovian trapped protons, plasma turbulent precipitation loss by electromagnetic ion cyclotron waves and by quasi-electrostatic loss-cone waves have been considered [13]. The maximum flux of protons we observe (Fig. 2) is well below the limiting flux expected as a result of the instabilities considered in [13]. This is consistent with the notion that most of the particles being limited are below our threshold energy. The electrons are not observed to decrease as the protons do (Fig. 2), but this might be due to their lower drift frequencies and consequently higher diffusion coefficients (§3), and to the limiting effect of synchrotron radiation. For the ion cyclotron wave, significant wave growth results only if E > Ee = B2l8jin [13], where n is the cold plasma density, so that our observations in Fig. 2 imply that Ec < 80 MeV. Therefore, at L = 3.5 this means n > 4.5 cm -3 . The lower limit of n is more than an order of magnitude higher than the plasma density at 1 ^ L ^ 5 derived for Jupiter in [14], but we note that this model does not take into account the plasma contributed by the satellites. I t may be that absorption by Amalthea is responsible for the drastic reduction even though the electrons we observe do not behave similarly. An analysis for Amalthea similar to that shown for Europa in Fig. 4, shows that this is possible. Since Ti is smaller for protons than for electrons at the relevant energies, and Ts is larger, both of these lead to a higher probability of absorption for protons (see § 3). In summary, on the basis of our data, it seems possible in principle that the proton decrease at L < 3.5 may be due to a pitch angle scattering loss mechanism. If the plasma density is n > 4.5 cm -3 , the ion cyclotron instability is a possibility. Proton absorption at Amalthea is also plausible for explaining this phenomenon. However, it is clear that many more data are needed to solve this problem; the behavior of the radial profile at L < 3 (Pioneer 11) and proton measurements at other energies can be expected to throw more light on this puzzle. Acknowledgment This work was supported by NASA contracts NAS 2-5602 and NAS 2-6552, and by NASA Grant NGL 05-005-007.

References [1] J. W. WARWICK, Space Sci. Rev. 6, 841 (1967). [2] J. N. CLARKE, Radio Sci. 5, 529 (1970). [ 3 ] R . W . FILLIUS a n d C. E . MCILWAIN, S c i e n c e 1 8 3 , 3 1 4 (1974). [4] R . W . FILLIUS a n d C. E . MCILWAIN, J . G e o p h y s . R e s . 7 9 , 3 5 8 9 (1974).

[5] E. J. SMITH et al., Science 188, 305 (1974).

528

A . MOGRO-CAMPERO, R . W . FILLIUS a n d C. E . MCILWAIN

[6] [7] [8] [9] [10] [11] [12] [13] [14]

S . D . SHAWHAN e t a l . , S c i e n c e 1 8 2 , 1 3 4 8 (1973). N . BRICE a n d T . R . MCDONOUGH, I c a r u s 1 8 , 2 0 6 (1973). T . BIRMINGHAM e t al., J . G e o p h y s . R e s . 79, 87 (1974). G . D . MEAD a n d W . N . HESS, J . G e o p h y s . R e s . 78, 2 7 9 3 (1973). W . N . HESS, T . J . BIRMINGHAM a n d G . D . MEAD, J . G e o p h y s . R e s . 79, 2 8 7 7 (1974). J . A . SIMPSON e t al., J . G e o p h y s . R e s . 79, 3 5 2 2 (1974). C . - G . FALTHAMMAR, J . G e o p h y s . R e s . 70, 2 5 0 3 (1965). F . V . CORONITI, C. F . KENNEL a n d R . M . THORNE, A s t r o p h y s . J . 1 8 9 , 3 8 3 (1974). G . IOANNIDIS a n d N . BRICE, I c a r u s 1 4 , 3 6 0 (1971).

Space Research XV — Akademie-Verlag, Berlin 1975

COSMIC D U S T I N F L U X TO T H E E A R T H DAVID W . HUGHES

The University of Sheffield, England

The influx of cosmic dust to the earth's surface has been deduced from observations obtained by satellite-borne detectors, from the analysis of backscatter radio echoes off meteor trails, and from visual meteor counts. Observations of zodiacal light have also been used to deduce a value for the density of cosmic dust in space. Recent measurements of the geocentric velocity of particles in the cosmic dust cloud have also been reviewed. These observations are combined to give a value for the total influx of cosmic dust to the earth (in the mass range 10~13 g to 106 g) and the spatial density of this dust at 1 AU from the sun.

1. Introduction The number of particles intercepted by the earth or detected by a satellite at a distance of 1 AU from the sun is usually given in terms of the cumulative flux, 6>2ji, of particles incident on an area of one square metre of the surface per second from above (i.e. per 2n steradians) having a mass greater than a given mass m(g). The flux observations given in this paper are mean values, the diurnal, seasonal and latitudinal variation having been averaged out. The mass density q of the cosmic dust cloud is related to 02n by 0 =

0.25gw(l + V / « 2 )

(1)

where 0.25 is the ratio between the cross-sectional area and the total area of the earth's surface, u is the unperturbed (gravitationally) geocentric velocity of the particles and ve is the escape velocity from the earth (see [1]). The term in brackets in Eq. (1) accounts for the gravitational enhancement and is obviously of great importance for low velocity particles. Ground-based observations, for example of radio and visual meteors (to > 10-6g), give the influx 02 „ and also VG the observed (perturbed) geocentric velocity. F G is related to u by u = (F G 2 — we2)1/2- Satellite-borne detectors usually measure 0 2jt and some are also now measuring the particle impact velocity which can be related to u or F G depending on the position in space of the satellite. Zodiacal light observations have been used to calculate q. More recent observations of the Doppler shift of the spectral lines in the zodiacal light give values for u; thus 02n can be found for the small particles responsible for this light by using Eq. (1). 34*

532

DAVID W . HUGHES

One obvious problem in the interpretation of flux data is caused by the dispersion in particle velocities. The value of u used in Eq. (1) is the median of the observations. Also measurements of F G , by radar and visually, suffer considerably from observational selection, the faster particles being the "brightest" and therefore the most easily detected (the fact that "brightness" is proportional to approximately the fourth power of F G exacerbates the problem, making slow particles virtually undetectable). 2. Cosmic Dust Influx to the Earth Fig. 1 is a compilation of the latest influx data from satellite, radar meteor and visual meteor sources. 2.1. Satellite Observations The satellite observations are for near earth space (i.e. 1 ± 0.2 AU from the sun), the mass sensitivity of the satellite-borne detector having been calculated for an impact velocity of 20 km s - 1 . These have been taken from [2]. The hatched

log10

Partide

Mass

(g)

Pig. 1. The cumulative flux of particles to the earth's surface as a function of mass. Satellite data from [2], radio and visual data from [8, 10]. The dashed line is an interpolation between satellite and visual data, sm is the mass distribution index.

Cosmic Dust Influx to the Earth

533

areas in K g . 1 represent the maximum degree of uncertainty in the data. The resulting satellite curve has the formula logio &2* = - (10.08 ± 0.15) -

(0.55 ± 0.02) log 10 m (10-

13

< m
10~6 g particles. This flux is, however, an important contribution to the mass accreted by the earth and it can be seen from Fig. 1 t h a t 0 2 „(1O- 6 ) is 1.7 X 10-7 particles n r 2 s- 1 2n ster- 1 .

Fig. 2. Zodiacal light observations of the cumulative flux, compared with the mean satellite data — and the mean radar data - • -. The zodiacal light observations have been taken from R. S. Powell et al., NASA SP-150, 233 (1967); J. L. Weinberg, Annls d'Astrophys. 27, 718 (1964); F. G. Gillett, Thesis, Univ. of Minnesota p. 101 (1966); H. Elsasser, Mitt. Astr. Inst. Tubingen No. 35, 73 (1958); M.F.Ingham, Mon. Not. R. Astr. Soc. 122, 171 (1961); D. E. Blackwell et al., Mon. Not. R. Astr. Soc. 136, 325 (1967); H. C. van de Hulst, Astrophys. J. 105, 471 (1947) and E. J. Opik, Irish Astr. J. 4, 84 (1956).

534

DAVID W . HUGHES

2.2. Zodical Light The zodiacal light is a further manifestation of the existence of a dust cloud in the solar system, the particles responsible for this light having masses similar to those observed by satellite detectors. Leinert [3] has taken the zodiacal light data of other authors and converted these into particle fluxes to the earth's surface. The fluxes obtained are shown in Fig. 2 (the results of van de Hulst and Opik have also been added to this figure). Following Verniani [4] and Kornblum [5] low values of density (1.0 g cm -3 ) and velocity u = 10.0 km s _ 1 were assumed. The resulting flux curves would have to be moved to the right for higher values of density and moved upwards for higher values of velocity. Opik [1] states that velocities Vq of the zodiacal dust particles lie in the range 11.1 to 12.2 km s - 1 corresponding to an upper limit to the unperturbed (extra-terrestrial, unaccelerated) geocentric velocity u of 5.1 km s - 1 . This is supported by Bandermann and Singer [6] who find geocentric velocities u of 5—15 km s - 1 . Also measurements of the Doppler shifts of the zodiacal spectral lines [7] confirm the fact that the geocentric velocity of the particles is small ( < 5 km s- 1 ). Returning to Fig. 2 it can be seen that the heavy dashed line which indicates the mean satellite flux curve (Eq. (2)) and the calculated zodiacal flux curves agree within the experimental error; i.e. for m < 10 -6 g both techniques are detecting the same particles and arrive at the same result. However, for m > 10~6 g the problem becomes much more complicated. These particles can only be detected by ground-based meteor radar systems if they have high velocities ( > 20 km s - 1 ). Zodiacal dust with its mean velocity of entry to the earth's atmosphere of only 12.5 km s~ l would have a negligible ionizing efficiency thus rendering it practically undetectable by radar. Thus the radar curve shown in Fig. 1 is the lower limit to the cumulative flux, representing only the high velocity component of the particles in the 10 -6 < m < 10~2 g mass range and not the complete influx. 2.3. Radio Meteor Data The radio meteor line in Fig. 1 has been taken from [8]. In this paper a minimum deviation fit was made to all available radio meteor flux data, the resulting curve being log10 02„(ocz) = (2.42 ± 0.04) - (1.03 ± 0.01) log 1 0 «, (10.7 < log, 0 s between specific mass limits, e.g. IO -13 < m < 10® g, 10-13 g being the mass of the smallest particle that can remain in the solar system against radiation pressure and IO6 g being the mass of the largest meteoroid that can escape from the gravitational attraction of a cometary nucleus. The annual mass influx to the earth is : Satellite Radio Visual Total 13 6 6 2 2 10- < m < 10- g 10- < m < 10" g m > 10~ g lO"13 < m < Full line in 3.3 X 109 g 4.5 x 108 g 4.2 x 108 g 4.2 X 109 g Fig. 4 Dashed line 3.3 X 109 in Fig. 4

1.24 X 1010 g

5.5 X 108 g

10 6 g

1.62 X 1010 g

I t can be seen from Fig. 4 that the integration limits have very little effect on the total mass influx; in the case of the full line 99% is made up of particles in the 10 - 1 0 < m < 10 1 g range; for the dashed line 99% of the influx is in the 10~9 < m < 10° g range. Assuming that this second approach is correct it can be seen from the above table that radio observations will only detect ~ 3.5% of the mass incident in the 10~6 < m < 10~2 g range, visual meteor observations however detecting about 75% of the particles with m > 10~2 g.

5. Discussion and Conclusions The influx to the earth of interplanetary matter is found (from the dashed line) to be 1.62 X 1010 g yr" 1 (equivalent to 9.9 X 10" 17 g c n r 2 s - 1 ). Error calculations indicate that the upper and lower limits are about 3.0 and 0.8 X 1010 g y r - 1 respectively. This result compares with tho value of 1.03 x 10 10 g y r - 1 obtained by Dohnanyi [16] 1.59 x 1010 g yr- 1 obtained by Whipple [17], 2.01 X 1010 obtained by Dohnanyi [18], all for the total surface of the earth, and 12.7 X 10 - 1 7 g cm - 2 s _ 1 obtained by Ganapathy et al. [19] for the lunar surface. An upper limit can be found for the spatial density of cosmic dust at 1 AU by using the dashed curve in Figs. 1 and 4 and assuming that the large majority of particles in the mass range 10~13 < m < 106 g have low geocentric velocities.

538

DAVID W . HUGHES

-12

- 1 0 L o g ^

- 8

- 6

Partie!«

- U M a s s

- 2 ( g )

Fig. 4. The mass influx on to the earth's surface in each logarithmic mass interval. The solid and dashed lines correspond to the solid and dashed lines in Fig. 1.

Following Opik [1,20] u in Eq. (1) is taken to be 6 km s giving a Vq of 12.6 km s- 1 . The upper limit of the space density is then 1.5 X 10~22 g cm - 3 . The important conclusion of this paper is that radio observations of the influx of particles cannot be included in the cumulative flux curve because the particles observed by radar only represent a small fraction (probably around 4%) of the influx in the 10 -6 < m < 10~ 2 g mass range. This small fraction is simply the high velocity component of the total influx in that region. Satellite-borne apparatus detects the particles which scatter zodiacal light and from Doppler shift measurements these particles all seem to have very low values of geocentric velocity whereas radio meteoroids [9] all have high velocities.

Acknowledgments I would like to thank Professor W. M. Alexander for supplying the satellite data prior to publication and Professor E. J . Opik for his helpful suggestions and private communications. i

Cosmic Dust Influx to the Earth

References [1] E. J . ÖPIK, Irish Astr. J. 4, 84 (1956). [2] W . M . ALEXANDER a n d J . L . BOOT, S p a c e R e s e a r c h X I V , 7 4 9 (1974).

[3] G. LEINERT, Space Research X I , 249 (1971).

[4] F. VERNIANI, Space Sei. Rev. 10, 230 (1969). [5] J . J . KORNBLUM, J . G e o p h y s . R e s . 74, 1 9 1 1 (1969).

[6] L. W . BANDERMANN a n d S. F . SINGER, Rev. Geophys. 7, 759 (1969). [7] N . K . REAY a n d J . RING, N a t u r e 219, 710 (1968). [8] DAVID W . HUGHES, i n p r e p a r a t i o n . [9] F . VERNIANI, J . G e o p h y s . R e s . 78, 8 4 2 9 (1973). [10] DAVID W . HUGHES, M o n . N o t . R . A s t r . S o c . 1 6 6 , 3 3 9 (1974).

[11] E. J . ÖPIK, Irish Astr. J . 6, 3 (1963). [12] DAVID W . HUGHES, S p a c e R e s e a r c h X I V , 7 8 9 (1974).

[13] G. S. HAWKINS a n d R . B. SOUTHWORTH, Smithson. Contr. Astrophys. 2, 349 (1958).

[14] J. A. M. MCDONNELL, Space Research XI, 415 (1971). [15] DAVID W . HUGHES, P l a n e t , a n d S p a c e Sei. 2 0 , 1 9 4 9 (1972).

[16] J . S. DOHNANYI, Science 178, 558 (1971). [17] F. L. WHIPPLE, Smithson. Astrophys. Inst. Prepr. No. 239 (1967). [18] J. S. DOHNANYI, Icarus 17, 1 (1972). [19] R . GANAPATHY, R . R . KEAYS a n d E . ANDERS, S c i e n c e 1 7 0 , 5 3 3 (1970).

[20] E . J . ÖPIK, Ann. Rev. Astron. Astrophys. 7, 473 (1969).

539

Space Research X V — Akademie-Verlag, Berlin 1975

N E A R - E A R T H COSMIC DUST F L U X E S OBTAINED FROM SKYLAB EXPERIMENTS C. L . HEMENWAY», D . S . HAXLGBEN" a n d C. D . TACKETT1» "Dudley Observatory and the State University of New York at Albany, Albany, New York, USA ^Dudley Observatory, Albany, New York, USA

Three exposures (34,46 and 33 days) of thin films and polished metal plates with a total area of 0.12 m 2 per exposure were carried out during the Skylab S-149 experiment. First study of the materials recovered indicates that theS-149 experiment contains important information concerning cosmic dust in the near-earth vicinity. Craters and penetration holes have been found ranging from 135 ¡¿m diameter to less than 0.5 ¡j.m. A cosmic dust flux curve in the mass range 10~ 16 —10 -7 grams is presented. Directional characteristics and evidence of particle fragility are seen in penetration holes in the multilayer thin film materials exposed. Preliminary composition data concerning the particle residue within the copper crater walls are given.

The Skylab S-149 experiment consists of an instrument designed to accept a set of four cassettes containing various experimental surfaces and foils to study micrometeorite impacts and penetrations in near-earth space. The instrument allows the unsealing and deployment of these cassettes, the carrying out of long duration exposures, and the closing and resealing of the cassettes after each exposure. Fig. 1 shows the equipment with the sample pans with blank slides (2.9 cm X 6.8 cm) in position as if they were ready for exposure. Polished metal slides of stainless steel, copper and silver were used, as were sets of 1 cm2 polished slides of copper and stainless steel for scanning electron microscope study, stacked aligned layers of thin (400 A) carbon-coated, nitrocellulose films on electron microscope grids, nitrocellulose films over glass, and two layers of 500—800 A gold foils over stainless steel. Fig. 2 shows the location of the S-149 hardware during the antisolar exposure (—Z) and during the solar facing exposure (+Z). The solar facing exposures were manually deployed on the ATM dish by the Skylab crews during their extravehicular activities. The -\-Z axis of the spacecraft was maintained toward the sun approximately 9 9 % of the time during these exposures. Table 1 shows the beginning and end times of the three exposures carried out at an altitude of approximately 430 km, and returned to earth thus far. A fourth set of cassettes is currently being exposed in the hope that an exposure of several years duration can be recovered on a future manned visit to Skylab. Fig. 3 shows a typical impact crater observed in a copper slide with a scanning electron microscope. The craters, with diameters between ~ 1 and ~ 100 ¡xm, all have typical high velocity impact structures such as a raised lip and sidewise compression of the host material. Fourteen craters have thus far been found by

542

C. L. HEMENWAY,

D. S. HALLGREN a n d C. D .

TACKETT

Fig. 1. S-149 micrometeorite collector in open position.

Fig. 2. S-149 exposure locations.

optical scanning at 200 X ; they have been analyzed to detect residual material of the impacting particles. Some foreign material was detected in eight of these craters. Table 2 shows preliminary data for all of the impact craters in copper studied thus far. The stainless steel craters require further study. The distribution of the craters in the various pans and covers is shown in Table 3. The directions indicated in Table 3 are the directions perpendicular to the exposed surfaces in the pans and covers with respect to the spacecraft as shown in Fig. 2. It will be noticed that the pans have significantly more craters than the covers

543

Near-Earth Cosmic Dust Fluxes from Skylab Experiments Table 1 Exposure Times Type

Beginning

End

Duration

Antisolar (SL 2/3) Solar (SL 3) Solar (SL 4)

23 J u n . 73 6 Aug. 73 22 Nov. 73

28 Jul. 73 22 Sept. 73 25 Dec. 73

34 days 46 days 33 days

Fig. 3. I m p a c t crater in copper.

Table 2 Residual Elements in Copper Craters Slide number

A-3-3-5 B-3-15-9 C-2-9-5 C-2-9-9 (1) C-2-9-9 (2)

Outside diam. (fxm)

Inside diam. ((Jtm)

Elements detected Al Si Fe

143.0 14.3 3.0 5.0 2.0

86.0 11.7 2.6 4.5 1.8

X X X X

Cr

X X

and that the antisolar covers appear to have more impact craters than the solar facing covers. We anticipate that further work will enable the directional characteristics of large micrometeorites to be discerned from these data. Fig. 4 shows a scanning electron micrograph of a penetration hole in the 5 0 0 - 8 0 0 A thick gold foil. One can see evidence of the non-perpendicular nature of the impact in the asymmetry of the streaks radiating from some of the holes. Many of the penetration holes show tears.

544

C. L. Hemenway, D. S. H a l l g r e x and C. D. T a c k e t t Table 3 Crater Locations

Antisolar (SL 2/3) Solar (SL 3) Solar (SL 4)

Covers Pans Covers Pans Covers Pans

A

B

C

D

Total

2 (-Z) 6 (-Y) 0 (+Z) 4 (+y) 0 (+Z) i (+y>

1 (-Z) 4(+X) 0 (+Z) 2 (+X) 0 (+Z) 1 (+ *)

i (-Z) 4 (+Y) 0 (+Z) 3 (-7) 2(+Z) 13 ( - 7 )

2 (-Z) 2(-X) 1 (+Z) 3(-X) 1 (+Z) 3(-X)

6 16 1 12 3 18

Fig. 4. Penetration in gold foil.

Fig. 5. Multiple holes in gold foil below penetration hole.

Near-Earth Cosmic Dust Fluxes from Skylab Experiments

545

Fig. 5 shows the foil immediately below a penetration hole of ~ 20 ¡xm diameter. I t will be noticed that the penetrating particle broke up into many smaller fragments. This demonstrates the meteoroid shield principle on a miniaturized scale, and suggest that some particles are rather fragile, since the dimensions of the particles are very large compared with the thickness of the foil. Some impact craters have been found in the stainless steel substrates beneath the two layers of gold foil. In one case, two small craters 1.5 and 0.5 ¡j.m in diameter were found in the substrate after the particle penetrated two layers of gold foil. The hole in the top layer was 25 ¡xm in diameter.

Fig. 6. Silver samples.

Fig. 6 shows an optical photograph of three exposed silver slides compared with an unexposed control silver slide. A thick black corrosion layer developed on the silver slides during the solar exposures and a thinner layer resembling tarnish developed during the antisolar exposure. X-ray diffraction analysis has shown the black material to be silver oxide (Ag 2 0). The copper slides showed a smaller but noticeable amount of oxidation. We suspect that the oxidation is the result of a chemical reaction with atomic oxygen in the ionosphere. Such an oxidation mechanism may explain at least a part of the small particle break-up and subsequent clustering of events observed during these S-149 exposures as well as previous space exposures and collection experiments [1, 2]. Fig. 7 shows the flux curve drawn from the data analyzed thus far. I t will be noticed that in the large particle size range (10~12—10~7 grams) the data are consistent with the S-10 data from Gemini 8—10 [1], and somewhat higher than the Explorer data [3]. The fluxes obtained with the crater results appear to be significantly higher in the antisolar exposure than in the solar exposure suggesting that these large particles are spiraling inward toward the sun. These "evil eye" fluxes observed during the solar exposures appear to be within the range of fluxes of submicrometer particles observed in the rocket collection experiment [2] and are consistent with the fluxes implied by the "evil eyes" of the S-12 experiment of Gemini 9 [1]. The crater sizes were converted to particle masses by assuming that the particle densities are 3 g cm - 3 and that the particle diameters 35

Space Research XV

546

C. L . H e m e n w a y , D . S. H a l l g r e n a n d C. D .

8=

Taokett

Evil Eye fluxes

\ ™

\

\ \ \

\

V \

\

Has

Antisolar Impa :t Crat tr Fit1X«S "

^sol or irr pact

—ta

Cra tar Fl uxes

-17

-16

-15 -14 -13 -12 LOG

-II

-10

-9

-8

-7

-6

PARTICLE MASS IN GRAMS

Fig. 7. S-149 particle fluxes.

are one third the crater inside diameters. The "evil eye" fluxes were converted by assuming a density of 3 g cm - 3 and a ratio of hole diameter to particle diameter of two. The fluxes have been corrected for solid angle of view and earth shielding. The data support the concept of the solar system containing two populations of particles, a submicrometer component flowing outward from the sun [2, 4, 5] and a component spiraling inward toward the sun.

Acknowledgments We wish to express our appreciation to the flight crews and flight controllers of Skylab for their competent and enthusiastic assistance, to Mr. W. Schneider, Mr. K. Kleinknecht and Mr. M. Willcox and their associates for their extraordinary support and encouragement in carrying out this experiment. We also thank L. Bourdillon, C. Jones, A. Laudate, B. Reynolds, J . Tarnawski, L. Brennan, P. Hutchison, W. Radigan and R. Schwarz of the Dudley staff for the design, qualification of the instrument, the sample preparation and the many hours of painstaking scanning. This work was carried out under NASA Contract NAS 9-10380.

Near-Earth Cosmic Dust Fluxes from Skylab Experiments

547

References [1] C. L . HEMENWAY, D . S. HAILGREN a n d J . P . KERRIDGE, S p a c e R e s e a r c h V I I I , 5 2 1 (1968).

[2] C. L. HEMENWAY, invited paper presented at the Whipple Retirement Symp., Cambridge, Mass. 1973, in press. [3] C. T. D'AIUTOLO, W. H. KINABD and R. J . NAUMANN, in: Proc. Symp. on Meteor Orbits a n d D u s t , Cambridge, Mass. 1965, NASA SP-135, 239 (1967). [4] C. L . HEMENWAY, D . S. HALLGKEN a n d D . C. SCHMALBERGER, N a t u r e 2 3 8 , 2 5 6 (1972). [5] C. L . HEMENWAY, J . W . ERKES, J . M . GREENBERG, D . S. HALLOREN a n d D . C. SCHMAL-

BERGER, Space Research XIII, 1121 (1973).

35*

Space Research X V — Akademie-Verlag, Berlin 1975

T H E I N T E R P L A N E T A R Y A N D N E A R JOVIAN D U S T E N V I R O N M E N T : SOME E X P E R I M E N T A L R E S U L T S J . M . ALVAREZ, D . H . HUMES, W . H . KXNAED a n d R . L . O'NEAL NASA Langley Research Center, H a m p t o n , Va, USA

The cosmic dust environment in interplanteary a n d near J o v i a n space has been measured b y Langley Research Center experiments on board Pioneers 10 a n d 11. The Pioneer 10 experiment detects microparticles capable of penetrating 25 (im steel sheet, while t h e Pioneer 11 experiment detects microparticles penetrating 50 ¡ m steel sheet. Analysis of t h e Pioneer 10 a n d 11 penetration experiment observations shows t h a t there is a gap in t h e cosmic dust environment. The gap s t a r t s a t 1.16 AU, as observed in b o t h experiments, a n d extends t o 1.35 AU on Pioneer 10 a n d 1.70 AU on Pioneer 11. The mechanism responsible for it is n o t well understood. Our experiments indicate t h a t t h e asteroid belt does n o t play a significant role in t h e production of 10 - 8 —10 - 9 g particles. The penetration rates f r o m b o t h experiments decrease w i t h increasing solar distance with no evidence of a n increased meteoroid population in t h e asteroid belt. Analysis of t h e Pioneer 10 J u p i t e r encounter d a t a indicates t h a t t h e twoorders-of-magnitude rise in penetration r a t e a t J u p i t e r periapsis is apparently due to J u p i t e r ' s strong gravitational field.

1. Introduction Although it is generally believed that micrometeoroids originate from asteroids and comets, the first in situ detection of interplanetary micrometeoroids in the range 10~8—l(h9 g indicates that the asteroid belt apparently plays little part in the production of these particles. This conclusion is the result obtained from the Langley Research Center's meteoroid detection experiments aboard Pioneers 10 and 11. These experiments have evolved from Langley Research Center detectors used from the beginning of the space age in near-earth orbit [1, 2] and more recently near the moon [3]. The meteoroid detection instruments on the Pioneer 10 and 11 spacecraft consists of 234 cells pressurized with a gas mixture 75% argon and 25% nitrogen. The cell walls are 25 ¡xm thick stainless steel on the Pioneer 10 experiment and 50 |im on Pioneer 11. When a meteoroid penetrates the wall, the gas escapes from the cell and the loss of pressure is detected. 2. Instrumentation The pressure switch used to indicate the loss of pressure is a cold cathode device. The switch consists of two electrodes (the potential difference between them is 500 V) in a pressure cavity connected by a copper tube to the pressure cell.

550

J . M. ALVAREZ, D. H . HUMES et al.

When a cell is punctured by a meteoroid the device acts as a glow tube due to ionization of the gas and it conducts current in a limited pressure range from 130 torr to 2 torr as the cell leaks down [4]. The detectors were fabricated in 13 panels (20 cm X 30 cm), each having 18 cells, in a configuration resembling an air mattress, and mounted on the back of each spacecraft's high gain antenna dish. The experiment carried on each spacecraft was split into two independent instruments to increase reliability. When any of the cells on that channel are punctured, the pulse resulting from the conduction of the switch is shaped and fed to the event counter which advances one count. The counter is then electronically locked out for a period of approximately 80 minutes so that the alternating current produced by the pressure switch circuit will not be interpreted by the system as several meteoroid penetrations. More detailed information on the experiment may be obtained from [5].

3. Observations Fig. 1 presents the Pioneer 10 penetration data, which show a clustering of events and various long time intervals during which no punctures were recorded. The data were checked statistically for very long and very short time between punctures. One very long interval, between 1.16 and 1.35 AU from the sun, was found which had a probability of occurrence of 0.0064 based on a penetration rate calculated from the preceding 13 punctures. The existence of this " g a p " is verified by Pioneer 11 data, show in Fig. 2; these data also contain a gap starting at the same location (1.16 AU from the sun) but extending to 1.70 AU. Thus it is much more improbable than the Pioneer 10 gap which extended only to 1.35 AU. The Pioneer 10 penetration experiment is sensitive to meteoroids about 10~9 g and larger while the Pioneer 11 penetration experiment is sensitive to meteoroids 1

2

3

-i

70

i

Y Y

distance from sun, AU 4 r

i

T-

Y

60 g 50

f

eQ- «

Ic:

t

I 30

.1

20 10

A

1

s

"

astenia

J xgap

I

Y Kirkwood's

gaps

oert

probabi/ity-0.0064

,

200

,

300

J

L

400 500 BOO exposure time, days

700

Fig. 1. Pioneer 10 penetration data: 25 ¡zm.

Interplanetary and Near Jovian Dust Environment

551

about eight times more massive. Thus, the fact t h a t the gaps are of different lengths implies that a size-dependent force may be involved in the mechanism responsible for producing the gaps. If gravitation is to play a role in the production distance from sun, AU 2.0 2.5

1.5

1.0

3.0

3.5

—i—

30 mm

%20

m t$ m M ¡pt Ili Ü®

-PIONEER 10 data gap

Y Kirkwood's gaps

i-« gap probability-5 x 10

x

asteroid betf

J

100

L

200 exposure time, days

300

Fig. 2. Pioneer 11 penetration data: 50 |j.m.

of the gaps, then the earth must be involved since it is a major perturbing influence from 1.16 AU to about 1.3 AU. Perhaps the Poynting-Robertson effect and earth resonance effects are responsible for the different lengths of the gaps. Figs. 1 and 2 also clearly show the lack of an increased microparticle concentration in the asteroid belt. Kirkwood's gaps [6] are indicated in an attempt to see if any of the known characteristics of the asteroid belt were observed in our data. (Kirkwood's gaps are gaps in the orbital period distribution of the asteroids; these temporal gaps have been transformed into spatial gaps by assuming asteroidal dust to be in near-circular orbits.) There is no obvious correlation here with Kirkwood's gaps. In fact, one would not suspect the existence of an asteroid belt from our data. Therefore, it is difficult to see how the asteroid belt can play a significant role in the production of meteoroids in the 10~8— 10 _9 g range. To analyze the interplanetary portion of the mission, we used a computer simulation technique. The simulation program calculates particle concentration (number of particles per cubic meter) and the penetration rate through our detectors as a function of distance from the sun, considering the meteoroid orbital parameter distribution, the empirical penetration equation, and the velocity and orientation of the Pioneer spacecraft. In effect, the program transform the particle orbital parameter distribution and mass distribution functions into particle concentration as a function of distance from the sun and the penetration rate through our detectors. Fig. 3 presents the penetration rates derived from measurements made by Pioneer 10 and 11 as a function of solar distance. The values of the penetration rates were obtained by a least-squares technique [7] using 6 or 7 punctures for

552

J . M . ALVAREZ, D . H . H U M E S e t a l .

each calculation. As the figure shows, the Pioneer 10 penetration rate can be fitted fairly well by an Rdependence on solar distance R. The Pioneer 11 data have been fitted in a preliminary fashion with an R~2A dependence using the seven punctures which occurred after Pioneer 11 went through the gap.

Fig. 3. Pioneer 10 and 11 penetration rates.

We have tried various orbital parameter distribution functions in attempting to simulate our experiments and we have found t h a t : (a) The gap detected on Pioneer 10 can only be simulated by particles having fairly circular orbits (eccentricity less than about 0.2). (b) All the orbital parameter and mass distribution functions used to fit the Pioneer 10 data gave a value of about 10~9 m~3 for the concentration of 10~9 g meteoroids at the earth (c^ 1 AU). (c) The dependence of concentration on solar distance is less marked than the dependence of penetration rate on solar distance. Thus the concentration decreases from the earth at a rate slower than R 1 for meteoroids in the 10~9 g range. (d) An orbital parameter and mass distribution function which reproduces the penetration rate through both Pioneer 10 and 11 detectors has yet to be found. This is because the penetration rates through the two thicknesses appear to differ in their dependence on solar distance. If this is a real effect, then either the mass distribution changes as a function of solar distance, or the orbital parameter distribution depends on mass, or both. I t is still premature to say definitely that the penetration rates through the two thicknesses differ in their dependence on solar distance because of the few punctures experienced by Pioneer 11 after the gap. By the time Pioneer 11 flies by Jupiter, however, there should be sufficient information to decide whether the dependences on solar distance are different for the two experiments. Jupiter encounter results are shown in Fig. 4. The ten punctures which occurred

Interplanetary and Near Jovian Dust Environment

553

near Jupiter periapsis are seen to be essentially symmetrical about periapsis. There are no indications that the spacecraft went through a Jupiter belt since such an occurrence would result in punctures being very close together in one or two segments of the spacecraft trajectory. There appears to be no evidence to indicate that Jupiter has a ring structure similar to Saturn. The top half of Fig. 4 Igr © PIONEER 10 encounter data S

Fig. 4. Pioneer Jupiter encounter data.

presents the number of punctures expected during periapsis assuming time is measured from the first puncture which occurred at about —50 hours from periapsis. Opik's flux enhancement expression [8] was used to calculate the increase in penetration rate due to micrometeroids affected by Jupiter's gravitational field. The penetration rate for interplanetary space was used as the flux at infinity and various values for the meteoroid velocity at infinity were used in order to determine whether gravitation alone could explain the data. As the figure shows, Jupiter gravity is enough to explain the increase in penetration rate during encounter. 4. Conclusions I n conclusion, Pioneer 10 and 11 penetration data indicate t h a t : (i) There is a gap in the interplanetary meteoroid environment which starts at 1.16 AU. I t extends to 1.35 AU for approximately 10~9 g particles. For about 10 -8 g particles the gap extends to 1.70 AU. The reason for the gap is not well understood. (ii) There was no indication that the asteroid belt plays a significant role in the production of small (10~8—10~9g) particles. (iii) The penetration rates decrease with solar distance. The Pioneer 10 penetration rate for 10~9g meteoroids decreases as R-1 with solar distance R. The concentration of such meteoroids decreases less rapidly than R'1 from a near-earth concentration of about 10~9 rrr 3 . (iv) The rapid rise in the Pioneer 10 penetration rate during Jupiter encounter appears to be caused solely by Jupiter's gravitational field.

554

J . M . ALVAREZ, D . H . HUMES e t a l .

References [1] C. T . D'AIUTOLO e t a l . , N A S A T N D - 2 4 6 8 (1964). [2] R . L . O'NEAL e t al., N A S A T N D - 4 2 8 4 .

[3] C. A. GUBTLER a n d G. N. GREW, Science 161, 462 (1968). [4] W . K . MESHEUIAN e t al., J . S p a c e c r . a n d R o c k e t s 7, 1 2 2 8 (1970). [5] W . H . KINARD a n d R . L . O'NEAL, N A S A S P - 2 6 7 , 1971 (p. 607).

[6] V. M. BLANCO and S. W. MCCTTSKEY, Basic Physics of the Solar System, Addison-Wesley, Reading, Mass. (p. 264). [7] J . M. ALVAREZ, NASA TN D-5668 (1970). [8] E. J . OPIK, Proc. Roy. Irish Acad. 54, 25 (1951).

Space Research XV — Akademie-Verlag, Berlin 1975

B O U N D S FOR T H E I N T E R S T E L L A R TO SOLAR SYSTEM M I C R O P A R T I C L E F L U X RATIO OVER T H E MASS R A N G E IO-"-10

1 3

g

J . A . M . MCDONNELL3 a n d 0 . E . BERG»

"•University of Kent, Canterbury, Kent, England t>NASA-Goddard Space Flight Center, Greenbelt, Md, USA A statistical analysis of the front film and grid coincidence events recorded over 5 years of observation on deep space probes Pioneer 8 and 9 is presented using computer integrated graphs. The data show a predominance of particles towards the probe apex for mass ~ 10 -11 g in accordance with a 'sweeping up' of particles in degraded cometary orbits, but for particles ~ 10 -13 g a predominance of particles incident from an apparent solar direction is observed. Average omnidirectional, probe apex and "solar" oriented fluxes are presented. The data have been analysed to investigate the possibility of a component from an interstellar source. Although the analysis indicates a small excess of particles corresponding to celestial directions 240°—360° longitude, which could indicate a celestial component arising from the solar apex of motion, the magnitude of this effect is entirely commensurate with normal statistical variations of the impact rate. A bound on the maximum magnitude of an interstellar component is placed at < 4% for particles of mass > 10 -13 g.

1. Introduction The Pioneer multi-coincidence microparticle sensing system provides the most comprehensive set of directly measured data on microparticle fluxes in space to date. The experiment has been described [1] and various results presented [2, 3]. The data analysed here have been accumulated over the years 1968 to 1970 for Pioneer 8, and 1969 to 1970 for Pioneer 9. During this time the two Pioneers ranged from 0.75 AU to 1.25 AU heliocentric distance. The experimental data fall into two main classes, termed Time-of-Might (TOF) and Front Film and Grid (FFG) events. For TOF events detection on the rear film and grid matrix is achieved and the radiant, velocity and mass of an incident particle is thereby derived. In this class of events, 19 have been reported by Berg and Grün [3]. We investigate here data where detection in coincidence by the front film and grid assembly has been observed and a pulse height analysis (PHA) obtained. The time-of-flight ¡information is not available for this class of event because of either the reduced geometrical acceptance angle required for the rear assembly detection or particle volatilization or comminution at the front film penetration. I t should be emphasized that penetration is not required for FFG events since the grid ion collector assembly can collect charge from the front and rear sides of the front film. With the absence of specific information on radiant and velocity,

556

J . A. M. MCDONNELL and O. E . BEKG

we must, for FFG events, use a different approach to the analysis. The pulse height information contains a term dependent on the product of mass X velocity2-6 [3] and we cannot, therefore, directly separate the two parameters of mass and velocity for each event. The radiant information of an incident particle can only be obtained from the spacecraft pointing direction at the time of impact; since the front film sensor has a relatively wide field of view, the radiant information for one event will be subject to uncertainties as large as ± 7 5 ° , although for larger numbers of particles the radiant trends will certainly become valid and, indeed, important factors in the interpretation of the data. Despite the shortcomings of these FFG events, the larger number of events available permits better evaluation of many statistical properties of the flux and, of course, the PHA analysis extends to those smaller masses which will not penetrate the film to give TOF data. The data of 5 years' observations on Pioneers 8 and 9 have been reduced to punched cards and from these 138 computer graphs obtained by incorporating an optimal filtering technique. Some earlier statistical properties of these events have been analysed by McDonnell et al. [4]. 2. Analysis Procedure In Fig. 1 we illustrate relevant parameters known at the instant of an FFG impact on a Pioneer's sensor. The spacecraft spins in its own heliocentric orbital plane which corresponds to that of the ecliptic; for each event we know the spacecraft pointing direction 0 measured by a sensor relative to the sun's position) which is in this analysis referred to the probe apex of motion. From the orbital position we obtain the same spacecraft aspect relative to Aries, 20 km s _1 ) particles observed by radio and visual techniques, other origins have been put forward for the smaller (say < ICh6 g) particles observed by satellite-borne detectors and as the scatterers of zodiacal light. For example Singer [1] concludes that collisions between asteroids in the asteroidal belt is a likely source; Roach et al. [2] give them a common physical origin with the outer solar corona; Opik [3] considers them as the remnants of primordial dust which has been spiralling inwards from the outskirts of the solar system since its origin; Greenberg [4] suggests that a considerable fraction of these particles have been directly accreted from interstellar space, whereas Whipple [5] and Harwitt [6] return to comets, considering comet Encke and comet Halley as being prime sources. Needless to say if more than one of these mechanisms are at work the physical characteristics of the produce are not necessarily the same. 2. Observations of the Cometary Contribution As the earth, on its journey around the sun, ploughs through the cosmic dust cloud it picks up about 9.9 X 10 -17 g cm -2 s _ 1 (see [7]), this value being obtained by averaging over the earth's surface and removing diurnal and seasonal variations.

566

DAVID W . H U G H E S

When the earth passes through a meteor stream (the debris of a decaying or decayed comet) a meteor shower occurs and the influx rate increases. The shower particles arrive from one direction only, and also have a lower mass distribution index sm than the continuous sporadic background, indicating that streams contain relatively more large meteoroids than the solar system dust cloud. (The mass distribution index sm can be defined by the formula dN a mrSm dm where dN is the number of particles having masses between m andTO+ dm; if sm > 0 the number of particles decrease as the particle size increases.) Hughes [8] found the maximum influx from the Quadrantid, Perseid and Geminid meteor streams to be 3.3 X 10"17, 0.28 X 10-17 and 2.4 X 10~17 g cm-2 perpendicular to radiant, s- 1 respectively, this being in the mass range 10 -13 < m < 106 g. These values are considerably less than the continuous sporadic influx but showers are easily detected in the radio and visual regions of the mass spectrum because the majority of shower influx is in this region, the majority of sporadic influx being made up of smaller particles. Also most meteor streams have high geocentric velocities making them stand out above the sporadic background. A considerable difference is found between the mass distribution index sm of particles in meteor streams and the sm values of the sporadic background. Much more work requires to be done in this field to obtain more measurements of the meteor stream sm values. Tentatively Hughes [9] proposed that the mean stream values changed from 1.71 for lfh 6 < m < 10 _ l g to 2.27 forTO> 10 - 1 g, these values being obtained by averaging known values of sm for the major meteor streams observed by radar and by visual observers. 3. Meteor Streams Fig. 1 shows how the particles in the proposed mean meteor stream are distributed between the different logarithmic mass intervals (i.e. the interval containing particles with masses between 10™ and 10 n+1 g). This has been calculated using the formulae developed in [10] assuming that the dust] in an average meteor stream has a mass of about 10 ls g [8] and that the upper and lower limits of the particle size in this debris are 10~13 g and 106 g respectively. The dashed line in Fig. 1 shows the distribution of the 1018 g after it has been perturbed from the meteor stream into the solar system dust cloud. This line is obtained by scaling the data given in Fig. 1 of [7] to give a total mass of 1015 g. It can be seen from Fig. 1 that the most probable particle size changes drastically as the matter is transferred from stream to dust cloud. For example in the stream 48% of the mass is in the form of particles in the size range 10 -2 —10° g, 73% in the range 10~3—101 g and 86% in the range 10 -4 —10 2 g. This contrasts with the solar system dust cloud (i.e. sporadic meteors) where 52% is in the range 10~6 to 10-4 g, 82% in the range 10" 7 —10" 3 g and 95% in the range 10~8—10~2 g. As the meteor stream decays the most probable particle size decreases. It is interesting to note that Fig. 1 indicates that very little of the cosmic dust detected by satellites is made up of meteor stream particles. For example if the satellite region is divided into intervals 10~7 < m < 10 _9 g, 10~9 < m < 10~ n g and 10 -11 < TO < 10~13 g the percentages (by mass) of the total influx made up of stream particles are 9, 17 and 30 respectively. These are only very tentative observations because the mass distribution index of 1.71, measured in the radio region has been extrapolated into the satellite region, but if correct they indicate

The Cometary Contribution to Cosmic Dust

567

that showers could become more and more obvious as one looks at smaller and smaller particles. Have meteor streams been detected by satellites? Soberman [11] concluded that satellite fluxes "show no significant variation from that expected statisti-

Log

Particle

Mass

(g)

Fig. 1. Mass of dust in each logarithmic mass interval of the total of 1015 g in an average meteor stream. It has been assumed that the stream particles have a mass distribution index of 1.71 in the region 10~13 < m < 10 _1 g and 2.27 in the region 10 _1 < m < 106 g. Distribution of dust after it has been perturbed into the solar system dust cloud.

cally even during known showers". However, Alexander et al. [12] detected increases of up to 100% in the spatial density of picogram particle in selenocentric space using microphone sensors on board the satellites Lunar Explorer 35 and OGO 3. These increases were found to coincide with the major meteor showers but it was thought that the dust might be ejecta from the lunar surface and not primary stream particles. Observations of the zodical light scattered from dust particles smaller than 10~6 g could also indicate if detectable streams are present in this mass range. Levasseur and Blamont [13] measured the zodical light intensity in the plane perpendicular to the earth—sun line using the low altitude D2A Tournesol satellite and found that on certain days the intensity increased by up to 100% over normal values, this increase lasting for a few days. By making observations in consecutive years the increases were found to occur at the same points around the earth's orbit. Contrary observations have, however, been made by Sparrow and Ney [14] who measured zodiacal light intensity for four years

568

DAVID W . HUGHES

from the satellite OGO 5. They failed to detect any temporal variations in the brightness and concluded that this variation if present at all is less than ± 10% of the mean value.

4. Variations in the Mass Distribution Index during Decay Fig. 2 gives a clearer idea of the effect of changing the mass distribution index. The five regions of this figure each show a fixed mass of material, this mass being made up of a collection of spherical particles of constant density. There are five different sizes of particles (see Table 1). The particles in each region have been laid out in two dimensions and each region has a different mass distribution index. The numbers of particles in each region are given in Table 1 together with the total cross-sectional area. I t can be seen from Fig. 2 that not only does the number of small particles increase enormously as sm increases but also the collision cross-sectional area of a fixed mass of material changes, this change being ~ 350% between 1.5 and 2.0 and ~ 66% between 2.0 and 2.5. The arrows in Fig. 2 indicate the transition direction under the influence of collisional fragmentation. Due to the change in cross-sectional areas the mean free paths of the particles decrease as sm increases and the illustrated transition process will therefore speed up. The decay of the meteor stream into the dust cloud produces considerable changes in the mass distribution indices. Fig. 3 shows sm as a function of mass, the dust cloud values being taken from the mean cumulative flux curve given in [7], the meteor stream values coming from [9]. I t can be seen that decay invariably causcs an increase in the sm value. The direction and magnitude of this change, /ism, should provide vital clues as to the structure of the meteoroids and to the actual decay process responsible for stream dissipation. Table 1 The Number and Sizes of Particles in a Fixed Mass of Material of differing Mass Distribution Index sm These sets of particles are shown in Fig. 3. Sizes of particles (arbitrary units) Radius Mass

16 4096

8 512

4 64

2 8

1 1

Mass distribution index (sm)

Number of particles of each size

Cross-sectional

1.50 1.75 2.00 2.25 2.50

2 1 0 0 0

0.29 0.40 1.00 1.39 1.66

* relative to sm — 2.00

4 4 4 2 1

12 19 34 22 10

34 93 272 293 242

94 438 2178 4000 5564

areas*

The Cometary Contribution to Cosmic Dust

569

s= 2-50 Fig. 2. Each circle shows a fixed mass of material made up of a collection of particles having different mass distribution indices. The numbers and sizes of the particles in each circle are given in Table 1. The arrows show the probable direction of transition under the influence of collisional fragmentation.

570

D A V I D W . HUGHES

Meteoroid density q varies from stream to stream, Verniani [15] giving a table for photographic (m ~ 10-1 g) meteoroids showing the Geminids with a density of 1.06 g em -3 at one extreme and the Draconids at the other with a density of

E

3

,

-

:

SOLAR SYSTEM DUST CL0UD\» , '

ic

!•

/

o

/

3 a 2 ui

/ /

•6

C

i METEOR STREAM

o

S

I

i

.

i

-12

-10

-8

Log

i -6

Particle

i -

4

i -

i

2

Mass

0

i

2

(g)

Fig. 3. The mass distribution index as a function of mass: cloud into which it decays.}

meteor stream;

dust

< 0.01 g cm -3 . There is also a density variation with mass, Verniani dividing the mass spectrum of the dust cloud into three regions q ~ 3.5 g cm -3 for m < 10~5 g, q a 0.8 g cm-3 for 10-5 < m < 10~2 g and o ~ 0.3 for m > 10~2 g. This indicates that the most probable stream particles (mass ~ 10_1 g) are porous fragile crumbly objects made up of loosely conglomerate sponge-like material [16]. There are four probable stream decay processes: (i) Stream-stream particle collisions As all stream particles return in their orbits to the place where they were emitted from the cometary nucleus, most of these collisions between stream particles will occur near perihelion. A t this point the relative velocities between the particles will be a few tens of metres per second. These low velocity collisions should lead to the partial fragmentation of the meteoroids and to slow stream broadening. (ii) Stream-dust cloud particle collisions Here the interparticle velocity is high, ~ 40 km s -1 , i.e. in the region of the mean stream particle heliocentric velocity. Collisions become more prevalent closer to the sun where the density qc of the cloud is higher (i.e. assuming particles have highly elliptical orbits, nc a r-1-5, r being the distance from the sun). The particles after the collision join the dust cloud. (iii) Planetary perturbation This is only important for close encounters of Jupiter and Saturn. For example the gravitational attraction of Jupiter is equal to that of the sun when the particles are 0.16 A U away from the planet and is 1% of the sun's at 1.23 AU. The Quadran-

The Cometary Contribution to Cosmic Dust

571

tid meteor stream which intercepts Jupiter's orbit will find about 10% of its meteoroids affected every time Jupiter passes through (i.e. every 12 years). (iv) Solar

perturbations

The Poynting-Robertson effect will cause the particles to spiral into the sun, the small ones moving faster. This increases sm on the sunward side of the stream and decreases it on the antisolar side. I t is thought that mechanism (ii) is the most important of these decay processes giving a stream life of between 5000 and 10000 years and producing a Asm of ~ 0.5. 5. Discussion One approach to the problems of meteoroid structure and stream decay is to start with the observations given in Figs. 1 and 2 and work backward from these. This contrasts with the normal procedure of assuming certain particle parameters and collision conditions and working forwards. For example the change in the most probable particle size from 10 _1 g in the stream to 10~6 g in the dust cloud could indicate that stream meteoroids are made up of a conglomerate of smaller elemental particles having masses of around 10~5—10-7 g and that stream meteoroids tend to break up into many of these smaller particles when they collide. This thesis is supported by observations of meteor flares which are thought to be produced by these small grains breaking off the main meteoroid as it ablates in the atmosphere [17—20]. Secondly the low value of sm (1.71) for stream meteors (10 -6 < to < 10_1 g) could support the accretion hypothesis of cometary dust particles. Napier and Dodd [21] performed a theoretical collision analysis where bodies coalesced on collision. Starting with a collection of similar sized bodies, sm quickly reached a value of 1.5. With a second model using collisional fragmentation and erosion hypotheses sm was found to increase from the starting point of 1.8, the size of the increase being a function of collision energy. A Asm of 0.5 obviously favours highly energetic collisions and thus process (ii) for stream decay. However the Asm produced by hypervelocity impacts of dissimilar massed fragile particles is thought to be considerably less than that which would be obtained if the particles were solid and brittle. The sm value could therefore be a good indicator of particle density and structure. The values obtained for the mean sm values (obtained by taking the gradients of specific portions of the mean flux curve) shown in Fig. 3 are supported by a review of individual sm values made recently by Millman [22]. Millman does not go into the problems of transferring from the observed quantity (i.e. crater diameter, puncture hole size, electron line density, meteor visual magnitude, lunar crater size and asteroidal brightness) to the particle mass; however, his analysis does indicate mass distribution indices changing from 1.47 (to < 10~8 g) to 1.97 (to ~ 10"6 g) to 2.30 (10-4 < to < 10° g). There are no observations in the (10° < to < 104 g) region; above this the sm value is dramatically low again: e.g. ~ 1.8 for meteorites (to ~ 10s g), 2.02 for the bodies responsible for lunar craters and 1.77 for asteroids. This variation of sm with mass leads to the conclusion that there are at least two sources of cosmic dust. Since increases in sm are produced by collisional fragmentation whereas low values of sm can be pro-

572

DAVID W . HUGHES

duced by accretion, the origin of the majority of the particles with m > 10 -7 g (sm > 1.6) is likely to be comets, the particles having high velocities (a few tens of km s _1 ), the differing sm values possibly being due to the differing collision lifetimes and mean free paths. An asteroidal origin is unlikely because of the dynamical difficulties of bringing a particle from the collision position in the asteroidal belt into the inner solar system. For example the mean relative collision velocity of the asteroids is around 5 km s - 1 and an impulse of this order in the retrograde direction is required to change the orbit into one that intersects the earth. The majority of the particles with m < 10 -7 g (sm < 1.6) probably have a different origin, being formed by accretion from much smaller particles in the solar corona or the interstellar dust. They also seem to have geocentric velocities considerably lower than those of the cometary debris [7]. The rather sharp change in the (sm, log m) curve in the region of 10~7 g might be an indication of the change of source. A cautionary note must be introduced, however, because most theories of cometary origin have accretion as one of the main building processes, and this possibly accounts for the low sm values found for most meteor streams. References [1] S. P. SINGER, in: Meteorite Research, D. Reidel Publ. Co., Dordrecht, Holland 1969 (p. 590). [ 2 ] F . E . ROACH, H . B . PBTIT, E . TANDBERG-HASSEN a n d D . N . DAVIES, A s t r o p h y s . J . 1 1 9 ,

[3] [4] [5] [6] [7] [8] [9] [10]

253 (1954). E. J. ÖPIK, Irish Astr. J. 4, 84 (1956). J. M. GESENBERG, Space Research IX, 111 (1969). P. L. WHIPPLE, NASA SP-150, 409 (1967). M. HARWITT, J. Geophys. Res. 68, 2171 (1963). DAVID DAVID DAVID DAVID

W. W. W. W.

HUGHES, HUGHES, HUGHES, HUGHES,

Space R e s e a r c h X V , 531 (1975). Space Research X I V , 709 (1974). Space Research X I V , 789 (1974). P l a n e t , a n d S p a c e Sei. 20, 1949 (1972).

[11] R. K. SOBEEMAN, Rev. Geophys. and Space Phys. 9, 239 (1971). [12] [13] [14] [15] [16]

W . M. ALEXANDER, C. W . ARTHUR a n d J . L . BOHN, S p a c e R e s e a r c h X I I I , 1 0 3 7 ( 1 9 7 3 ) . A . C. LEVASSEUR a n d J . E . BLAMONT, N a t u r e 2 4 6 , 2 6 ( 1 9 7 3 ) . J . G. SPARROW a n d E . P . N E Y , S c i e n c e 1 8 1 , 4 3 8 ( 1 9 7 3 ) . P . VERNIANI, S p a c e S e i . R e v . 1 0 , 2 3 0 ( 1 9 6 9 ) . DAVID W . HUGHES, J . B r i t . A s t r . A s s o c . 8 4 , 2 7 2 ( 1 9 7 4 ) .

[17] R. E. MCCROSKY, Ph. D. thesis, Harvard University (1955). [18] H. J. SMITH, Astrophys. J. 119, 438 (1954). [ 1 9 ] A . N . SIMONENKO, K o m e t y i M e t e o r y 1 5 , 3 4 ( 1 9 6 7 ) . [ 2 0 ] V . I . MUSIJ a n d I . S . SHESTAKA, K o m e t y i M e t e o r y 1 6 , 2 0 ( 1 9 6 8 ) . [ 2 1 ] W . M c D . NAPIER a n d R . J . DODD, M o n . N o t . R . A s t r . S o c . 1 6 6 , 4 6 9 ( 1 9 7 4 ) .

[22] P. M. MILLMAN, The Moon 8, 228 (1973).

Space Research XV — Akademie-Verlag, Berlin 1975

T H E B I S T R I B U T I O N OF D U S T ALONG T H E E A R T H ' S O R B I T D E D U C E D FROM S A T E L L I T E M E A S U R E M E N T S OF Z O D I A C A L L I G H T A . - C . LEVASSEUR a n d J . E . BLAMONT

Service d'Aeronomie du CNRS, Verrieres-le-Buisson, France

The zodiacal light intensity has been measured for 26 months from the D2A satellite. The elongation was 90° degrees. Under normal conditions, the intensity remained between 57 S10 (vis)* at the ecliptic poles and 150 S10 (vis) at the equator. It was observed to vary slowly at high ecliptic latitudes with a one-year period. It could also increase rapidly from one day to the next; such increases occurred sometimes over a limited part of the sky. The parallel trends of the evolution with time for selected directions in two consecutive years show that the variations are due to the position of the earth in its orbit. Long-period variations are explained by a zodiacal cloud symmetrical with respect to the solar system invariant plane. Small-period increases may be due to local meteor streams.

1. Experimental Conditions and Data Reduction The experiment has been described in [1]. I t is a photometer with one channel at Balmer alpha and a second one at 6530 A. Only 6530 A data will be discussed here. The bandwidth was 20 A and the integration time ~ 0.9 s. Because of this integration time, the field of view was extended from 1°20' X 2°40' either to 2°40' X 2°40' or to 6°35' X 2°40' depending on the spin period of the satellite. A 100 S 10 (vis)* signal provided a count rate of about 670 counts s - 1 . The on-board calibration source and the calibration on stars crossing the field of view or on H I I regions (a complete mapping of galactic nebulae has been obtained) enabled us to demonstrate that the instrument response was constant within 5% and that no short term fluctuations occurred. The D2A Tournesol satellite was launched in April 1971 and turned off in June 1973. Its spin axis was oriented towards the sun. The line of sight of the photometer was perpendicular to the spin axis, and therefore scanned all the ecliptic latitudes every minute, or every 4 minutes (Fig. la). About two orbits were completed every 24 h, that is to say, during the night phase about 1200 measurements were made. Only 60440 measurements which satisfy the conditions given in Table 1 will be analysed here. Very strict conditions on the moonlight have been used, to make comparison possible with the results of Sparrow and Ney [2). After this editing of the data, the recorded signal N is due to the zodiacal light I , * S10 (vis) = Brightness unit: number of solar type 10th magnitude stars per square degree.

574

A . - C . LEVASSEUR a n d J . E . BLAMONT

Table 1 Reduced Measurements for Zodiacal Light Observations Solar depression angle Impact parameter Galactic latitude |&nl No star brighter than magnitude New moon; (line of sight, moon) angle No Lagrangian point of earth-moon system Spacecraft outside South Atlantic Anomaly Balmer a emission nominal

> 150 km

| no nightlow

>30°

\

gg "> ^

i small starlight contribution | no moonlight contamination J - dark current nominal

the dark current n, and the integrated starlight contribution, deduced from the tables of Roach and Megill [3]: I — {N — n) K — h(lu,

bn).

Fig. l b shows a typical example of zodiacal light measurements obtained over two nocturnal parts of the orbit. The results are in good agreement with previous measurements [4—7]. H*rlh «cliptic pal«

Fig. 1. (a), geometry of the observations of zodiacal light from D2A Tournesol; (b) typical example of the intensity, in S w (vis) units, as a function of the ecliptic latitude /S in degrees, measured on 4 June 1973. The spin period was 4 min.

The instrumental arrangement allows a measurement of zodiacal light variations from day to day. Fig. 2 presents the results obtained at the north ecliptic pole, where the stellar background was constant all over the year. There were almost 15 different measurements on each day, and the rms remained about 7 Si0 (vis). Two types of features may be noticed: a weak annual variation, and short-period increases.

2. One Tear Period Variations: Symmetry of the Zodiacal Cloud 2.1. Observations at the North Ecliptic Pole Fig. 3 a presents the data obtained at the north ecliptic pole, after removing all the points that may be suspected of representing short-period increases. The average intensity is 57 S10 (vis) in July—September, and 69 S10 (vis) in Feb-

Mean

value, and

r.m.s.

value

for

one

day

«i

10 z UJ so tz

AP.| M A Y

iW

L à S ^ ^è-V-A-Aj!

-

jjUNE

r

r ^ 5T.

jJULY

- —

| AUG. |SEPT. | OCT.

*

| NOV. | DEC. | J A N .

| F E B . | MAR.

|APR.

CO

N O

¡MAT

Fig. 2. Mean value of the intensity measured at the north ecliptic pole from April 1971 to June 1973. invariant

Ecliptic

plane

plane

CN N o>

U

tv O

m io o 100

CO

1/1 2

liJ

V

il 50

o

K*

CN

O JUNE

JULY

AUG.

SEPT.

OCT.

NOV.

DEC.

JAN.

FEB.

MAR.

APR.

MAY

Fig. 3. (a), evidence for variation with one-year period at the north ecliptic pole; (b), explanatory diagram.

576

A.-C. L e v a s s e u r and J. E. B l a m o n t

120 -Oo


10 s cm"

Zone I

Zone I I

Zone I I I

10*

2

4 6 10s

TRACK

2

* e

DENSITY

(Cm*)

107

2

Zone I V 6

t O 8 f>

Fig. 3. Histograms showing t r a c k density distribution in feldspar crystals and pyroxenes

olivines

H

f r o m different depths of t h e L u n a 20 soil sample. The shaded p a r t | | j refers to track density g > 108 c m - 2 .

Track Investigations of Lunar Soil from Luna 20

605

To compare the degree of irradiation by heavy nuclei in the lunar regolith we used, as described in [11], two parameters: NH/N, the fraction of crystals irradiated at depth H (for Qes > 108 cm -2 ) on the surface of regolith and oq, the quartile track density such that 25% of all examined crystals have track densities lower than Qq. The results obtained for NH/N and oq are given in Table 1. About 30% of all crystals studied were irradiated by the low-energy heavy nuclei of solar cosmic rays. This indicates that the Luna 20 matter is a "feebly" irradiated sample of the lunar soil [12, 13]. These data (see the histogram in Fig. 3) clearly show the existence of several groups of crystals being distinguished by their track density with Q < 10® cm -2 , 10® 2 cm, we have obtained a minimum exposure age of this surface regolith of Luna 20 soil. Using the rate of track formation in regolith given in [14], we have found the values of effective exposure ages corresponding to the observed crystal groups to be (0.2—1) X 10® yr, (2—20) X 106 yr and (20—200) X 106 yr. As is seen from Table 1, the relative abundance of the feldspar crystals entering these three groups varies significantly with depth (by a factor of ~ 2 ) only for those samples with o 1.5 kg cm - 2

Bearing strength 0.15-1.5 kg cm- 2 Bedding depth of hard base < 1 0 0 mm Bearing strength < 0.15 kg cm- 2

explained by the presence of dense soil layers at the top and of more loose material beneath them. This case was classified as "soil with enhanced penetrability". Table 2 lists the percentages of the above typical cases during the operation of Lunokhod 2. 2.3.3. Mechanical

'properties

of stones and some lunar

formations

The strength of stones and outcroppings was estimated from the results of the experiment with the conic-vane penetrometer and when Lunokhod's wheels rode into them [4]. Stone-like formations with loose structure easily crumble. During their destruction smooth vertical surfaces are formed and the fragments are pounded by the wheels. When Lunokhod rode into such obstacles its orientation remained practically unchanged; this is also indicative of the low strength of these formations. Most stones investigated had high strength. When a load over 20 kg was applied, the penetration of only the pointed section of the penetrometer to a depth of 0.5—2.0 cm was observed. The pressure was 5—10 kg cm - 2 . Occasionally (0.5% of the overall number of measurements) splitting of the stone took place. The bearing strength of lunar soil on the rims of some craters is noticeably lower than at the horizontal sections and on the slopes of craters. This is also confirmed by the analysis of the character of soil strain under Lunokhod's wheels from photographs. On the rim of a crater the track is much deeper. The bearing strength on the rim of relatively fresh craters 1 —1.5 m in diameter is on average 0.1—0.15 kg cm - 2 ; on the rim of a crater 13 m in diameter 0.25—0.3 kg cm - 2 . Between craters and on strongly eroded "old" craters the bearing strength of the soil was somewhat higher than average. The uppermost lunar layer is dust-like. Under the effect of a wheel a clear-cut trail 0.5—1.0 cm deep is formed. Soil is, in the main, packed under the ninth wheel without noticeable bulges sideways. The bearing strength of this soil layer determined by calculations on the basis of the depth of the track of the ninth wheel was much lower than for the main mass of soil.

Luna 16 and 20 Investigations of Physical and Mechanical Properties

615

2.3.4. Interaction between Lunokhod and soil Processes of interaction between the self-propelled vehicle and the soil were studied along the entire paths of Lunokhods 1 and 2. With this aim in view the torques on the wheels, the angles of inclination of the surface and skidding were measured [1, 2]. The averaged results of measurements are shown in Fig. 7. The values of the skidding coefficient a are given along the horizontal axis, the traction coefficient T and the coefficient of momentum yi are given along the vertical axis. The relationships between these terms have the following form: (7)

V = fmzxO' T = (Tm,x + T0) a> -

T0

where y m a x is the coefficient of momentum in the case of full skidding (a = 1), T 0 the traction coefficient when skidding is absent (a = 0), T m a x the traction coefficient when a = 1, and j is a constant coefficient. The averaged values of the coefficients in Eq. (7) are: V m a x = 0.72, T m a x = 0.48, T0 = 0.15, j = 0.4. Lunokhod's motion at different speeds allows the evaluation of the effect of speed and the time of application of load on soil strain processes. I t has been found t h a t in the speed range of 1—2 km h r - 1 the strength and strain properties of soil do not vary. The value of the vertical speed of soil strain was 0.12 to 0.3 m s - 1 while the time of the effect of external load was 0.2—0.4 sec. The ground trials of the self-propelled vehicle on soils of different types show t h a t the closest analogue to lunar soil, with regard to its traction and cohesion characteristics, is fine-grained quartz sand (Fig. 7). Fig. 8 shows the average maximum value of the traction coefficient Tmax as a function of the bearing strength q of the soil. This dependence was obtained during the trial of a single wheel on soils of a wide granulometric spectrum and under conditions simulating lunar

Fig. 7. The traction and-cohesion characteristics of Lunokhod's propulsive device: T, traction coefficient; yi, coefficient of momentum; a, skidding coefficient. 1, obtained on the moon; 2, obtained during ground trials.

616

A . K . LEONOVICH, V . V . GROMOV e t a l .

gravity in a flying laboratory. The average value of bearing strength of 0.38 kg c m - 2 corresponds to T m a x = 0.48, which is in good agreement with the results of direct measurements. Tmax' 0.8

OJB OA 0.2 u

0.2

OA

0,6

OJB

10

I2q-

Fig. 8. Traction coefficient Tmas as a function of bearing strength of soil q (kg cm -2 ).

References [1] A. P. VINOGRADOV, Preliminary Data on Lunar Soil Samples Brought by the Luna 16 Unmanned Spacecraft, Geokhimiya (Geochemistry), No. 3, 1971. [2] Yu. I . STAKHEYEV, Y E . K . VTTLFSON, A. V . IVANOV et al., Izv. Akad. Nauk SSSR, Geological Series, No. 1, 1972. [ 3 ] A . P. VINOGRADOV, Preliminary Data on Lunar Soil Sample Brought by the Luna 2 0 Unmanned Spacecraft, Priroda (Nature), No. 8, 1972. [4] A. K. LEONOVICH, V. V. GROMOV etal., in: The Movable Laboratory on the Moon—Lunokhod 1, Izd. Nauka, Moscow 1971. [5] A. K. LEONOVICH, V. V. GROMOV, V. A. LOZHKIN et al.,paper presented at 24th IAF Congr. Baku, USSR, October 1973. [ 6 ] V . V . GROMOV, A . K . LEONOVICH, A . D . D M I T R I Y E V et al., Kosm. Issled. 9, No. 5 , 1 9 7 1 . [7] Surveyor Project Final Report, Part 2, Science Result, NASA, 1968. [8] 1 . 1 . CHERKASOV and V. V. SHVARYOV, The Principles of Lunar Soil Science, Izd. Nauka, Moscow 1971.

Space Research XV — Akademie-Verlag, Berlin 1975

RESULTS OF RADAR E X P E R I M E N T S PERFORMED ABOARD THE LUNA 19 AND 20 AUTOMATIC STATIONS N. N. K b o u p e n i o , A. G. B a l o , E . G. R u z s k h , V . A. L a d y g h e n , V . V . C h e r k a s o v and V . S. Fomin Space Research Institute, Academy of Sciences, Moscow, USSR

The Luna 19 automatic station, the lunar artificial satellite, and the Luna 20 automatic station, which landed on the lunar surface, performed measurements on the reflection characteristics of radiowaves in the 3 cm range. Luna 19 showed that in two regions near the Riimker crater the dielectric constants £ are 2.35 ± 0.65 and 3.2 ± 0.2. The rms angles of surface inclination over a base of about 30 cm in these regions a a are 10° i 1° and 8.5° ± 1° respectively. At the Luna 20 landing site (Mare Foecunditatis) e = 1.7 ± 0.2 and

m

X CO m

111

¥ 180 c

270°

_L_

0° In

L 90"

180"

Pig. 12. High-energy gamma ray emission from the galactic plane as observed by SAS 2. Some results of OSO 3 have been included where SAS 2 had no data.

Composite data SAS-I | b111«ICK0 OSO-II lb 1 ! 15°

I il 180°

270°

if mit} IM J

L

90°

180°

Fig. 13. Composite data of SAS 2 and OSO 3 for the entire galactic plane. Results have been weighted according to accuracy of measurement. 44*

692

K . PlNKAU

however, will produce gamma rays by Compton collisions on starlight. Since starlight is strongest in the galactic centre and shows no toroidal structure, the ensuing gamma rays will in turn exhibit a distribution with galactic longitude that is not flat but rather shows a maximum at I11 = 0° (Fig. 16; but see also

Pig. 14. Energy spectrum of high-energy gamma rays recorded from the direction of the galactic centre.

15

Wj for 0 - 9 0 ° Data (from Fig. 3 ) Surface Density of Giant HH regions from Mezger (1970) c o en

10

Hi CC

E

1=1 X

LU

IOOM»V

Fig. 16. Attempt to explain the observed gamma ray distribution in terms of jt°-production of cosmic rays on matter in the spiral arms and interarm regions, and of inverse Compton collisions of cosmic ray electrons on starlight.

Cowsik and Voges [36]). It is in this context that the mixed distribution of the OSO 3 and SAS 2 results shown in Fig. 13 is interesting. Furthermore, if the spectrum from the centre region is as hard as indicated by some of the results in Fig. 14, an inverse Compton process would give a harder spectrum than a 7i°-spectrum at these larger energies. 2.4. Localized Sources Recently, SAS 2 has also added to the wealth of data existing on the Crab, and the new compilation of results is shown in Fig. 17. I t appears that the entire flux is pulsed. Another feature has emerged from the SAS-data that had been anticipated by the present author several years ago [38]. This is the observation of the Vela supernova remnant [39] (Fig. 18). The centre of the gamma ray enhancement is within 1° of Vela-X and P S R 0833-45. The observed excess is 5 X 10" 6 photon cm- 2 s- 1 > 100 MeV. Pinkau [38] had predicted a flux of 4.5 X 10~8 photon c n r 2 s - 1 if the product of total released cosmic ray energy times local matter density had a value of 7 x 1050 erg cm - 3 . While these cosmic rays slowly diffuse away from the source, they produce gamma rays by interactions with the interstellar gas. The observed gamma rays are consistent with a jr°-spectrum. From considerations of the distance and age of supernovae, flux values for other sources are given in Table 1 [38]. Observations like these of the seats of cosmic ray nucleon production are likely to help settle whether cosmic ray

K . PINKAU

io

11 UNII—1111iim—111 hihi—I 11iinq—111 HIHI

io

\

--0.7E"09 PULSED FLUX (LAROS it ol, 1973)

\

\ 21.8 E- TOTAL X-RAY FLUX \ V / (LAROS el Ol., 1973) tN ^

'+>A

10 10 10 ' 10 10 10 10 " i i IO*

I SAS-2 KURFESS (1971) i McBREEN «I ol.,(1973) HELMKEN AND HOFFMAN (1973) > ALBATS Ol Ol ,(1972) l BROWNING «I Ol.,(1971) > KINZER et Ol.,(1973) > PARLIER •> 01.(1973) • ORWIG «t Ol.,(1971) ' FISHMAN il Ol.,(1969) ' KETTENRINS «I ol.,( 1971)

io"

+

' i "•"•' i i • I II Hill 1 ' " 10* I0T 10" 10* PHOTON ENERGY (*/)

10

Fig. 17. Energy spectrum of hard (pulsed) photons from the Crab nebula.

Fig. 18. Result of t h e observation of energetic gamma rays from the Vela region.

Gamma Ray Astronomy (0.1 — 1000 MeV)

s n •«'fl Ë7Â P 3 Pk « 'i X 3, CS V . ö5 P N tj tí

t

2-

-30*-20* -10° 0

10° 20* 30*

_1

I

-30* -20* -10° O

b,

I I L_ 10" 20* 30*

b,

Fig. 3 (a). Distribution of high-energy (E y > 100 MeV) gamma rays summed from In = 335° to l u = 25° as a function of 6 t I . The solid line represents the sum of two distributions with equal areas, one representing only the detector resolution and the other a Gaussian with 100 MeV) gamma rays summed from 90° < l n < 170° and 200° < 111 < 270°, where data exist.

other a Gaussian with a standard deviation of 6°. As will be discussed later, this result implies that the origin of the radiation is about equally divided between close ( < 2 or 3 kpc) and more distant regions. In Fig. 3, right, [11], the distribution in b11 for that portion of the region 90° < ln < 270° for which data exist is plotted, except that the Crab region has been excluded. Here there is no narrow peak—suggesting that most of the radiation is coming from relatively close regions as expected since the sun is fairly far from the galactic center and there are no strong sources thought to be near the rim of the galaxy. The other striking feature in Fig. 3 is the very much greater intensity of the galactic radiation in the 335° < l n < 25° region, which was also seen in Fig. 1. There are two peaks, at 260° < I11 < 270° and 70° < Z" < 80°, in addition to that due to the Crab nebula (180° < lu < 190°). The spectra from these regions are also quite hard and indistinguishable from the other regions just mentioned. The region of greater intensity of the two, that between 260° and 270°, has been discussed by Thompson et al. [4]. This enhancement is centered around b11 = —3 (±1)° rather than b11 = 0°, and is significant at the 8a level.

High Energy Gamma Ray Astronomy

703

As noted by Thompson et al. [4], this enhancement could be the result either of a galactic arm feature or a compact source, but the limited extent, the fact t h a t it is centered at 611 = —3 (¿1)°, and that the Vela X supernova remnant is consistent with its location strongly suggest the latter explanation. This possibility will be pursued further in § 3.4.2. 3.1.2. Discussion

of Galactic

Radiation

The high energy galactic gamma rays are generally thought to result from the interaction of cosmic rays and interstellar matter. This concept is supported by the relatively hard energy spectra observed here for most of the galactic plane. This hypothesis will be examined in terms of some of the models that have been proposed, after a brief review of the points in the basic calculation of particular importance here. The number and energy spectrum of the gamma rays produced by cosmic rays interacting with interstellar matter has been calculated in detail for the case of the cosmic radiation in intergalactic space by several authors (e.g. [12, 13]). The flux of gamma rays with energies greater than E at a distance r is given by the expression

where S is the number of gamma rays produced on the average for one interstellar nucleus/sec and a cosmic ray energy density and spectrum equal to that near the earth, n is the intergalactic nucleon density, and g has been introduced here to represent the ratio of the cosmic ray density to that in the vicinity of the solar system. Following Stecker [14], S is taken to be 1.5 X 10~25 s - 1 . The principal contribution to the high energy gamma radiation from the cosmic ray interactions with interstellar matter comes in the cosmic ray energy range from a few tenths of 1 GeV to a few tens of GeV. Below that energy range the parent mesons are not produced, and at higher energies the contribution is very small because the cosmic ray energy spectrum is decreasing much faster with energy ( ~ ¿J1'4). Hence, when cosmic rays are mentioned here, the energy range mentioned above is implied. In the first attempts to compare the observed high energy gamma ray intensity with calculated values, it was assumed (e.g. [1]) that the cosmic ray density was uniform throughout the galaxy so that g could be taken outside the integral in Eq. (1), and was usually set equal to one. Using the 21-cm data to estimate columnar hydrogen density Kraushaar et al. showed that whereas the calculated intensity was fairly close to that expected in the anticenter direction when the expected intensity was integrated over the solid angle of the detector (which had a Gaussian angular sensitivity with standard deviation about 15°), the observed intensity in the galactic center region was about four times the calculated value. Thus, the galactic longitudinal dependence was clearly inconsistent with this model, and it could, therefore, not be brought into agreement by assuming a uniformly higher value of the cosmic ray density or by assuming that the total matter density was uniformly much higher because a significant portion of the interstellar hydrogen was in molecular form, for example. More recently, Strong et al. [15] have assumed that the cosmic-ray density has a smooth distribution, but one which increases towards the galactic center

704

C. E . FICHTEL

according to the equation :

Z2 exp 20 (

g oc i Z e x p

exp

100 )

( - x )

t1 +

4

C/3 1.2

i —j

0.8 0.4

-

i

i

I

—80 - 6 0

i

-40

i -20

i

I

0

20

1

40

I

60

1

80

GALACTIC LONGITUDE (£] Fig. 4. Comparison of the longitudinal distribution of galactic -/-radiation observed on SAS 2 with the distribution given by the theoretical model of Stecker et al. [17].

agree in detail with more recent SAS 2 results. This work, however, is important as one of the papers departing from the traditional constant density cosmicray concept. Stecker et al. [17] have proposed that the galactic cosmic ray flux varies with the radial distance from the galactic center and is about an order of magnitude higher than the local value in a toroidal region between 4 and 5 kpc. They further suggest that this enhancement can be plausibly accounted for by Fermi acceleration caused by a hydrodynamic shock driven by the expanding gas in the " 3 kpc" arm and invoked in some versions of galactic structure theory. This theory does provide a possible explanation of the general enhancement in the central region as shown in Fig. 4, but possibly not some of the fine details now beginning to appear. There is, of course, also the question of whether or not the Fermi acceleration exists. If it does, then, clearly, the accelerated cosmic rays could play a role. In pursuing the problem of galactic gamma radiation, it is important to realize that the one-dimensional full-width angular resolution of the high energy gamma ray detectors flown thus far has been either several degrees, for SAS 2, or about

High Energy Gamma R a y Astronomy

705

25° in the case of OSO 3. Thus, the observed intensity of a feature with a thickness comparable with that of the disc of the galaxy will decrease approximately inversely with the distance once it is more than 2 kpc away from SAS 2 (and closer for OSO 3), and faster if it is also small in extent within the plane. Hence, more distant regions of the galaxy have to be substantially more intense than local ones to explain an observed intensity of gamma rays in any given direction with the present instruments. This consideration, together with the geometrical distribution of the intense high energy gamma radiation, particularly the broad, relatively flat distribution of the gamma radiation in galactic longitude over 70° to 90° in the central region of the galaxy, suggested to Kniffen et al. [3] and Bignami and Fichtel [18], that the source of the enhancement is possibly predominantly diffuse radiation from the spiral arm segments closest to the sun in the direction of the galactic center. Bignami and Fichtel [18] have proceeded further and proposed that in general the cosmic rays are enhanced where the matter is greatest, namely in the arm segments and clouds. This hypothesis is supported by the following considerations: First, it is assumed that the cosmic rays and magnetic fields are galactic and not universal. Then, as shown by Bierman and Davis [19] and Parker [20] in more detail, the magnetic fields and cosmic rays can only be contained by the weight of the gas through which the magnetic fields penetrate; and, hence, they are tied to the matter. The galactic cosmic ray energy density cannot substantially exceed that of the magnetic fields, or the cosmic pressure will push a bulge into the fields, ultimately allowing the cosmic rays to escape. The local energy density of the cosmic rays, however, is about the same as the estimated energy densities of the average magnetic fields and of the kinetic motion of matter. Together the total pressure of these three effects is estimated to be equal to the maximum that the gravitational attraction can hold. This feature suggests that the cosmic ray density may generally approach the limit the magnetic fields can contain. This concept is also given some theoretical support by the expected slow diffusion rate of cosmic rays in the magnetic fields of the galaxy and the very possible high production rate of cosmic rays, which together also suggest that in general the cosmic rays should be plentiful in a given region and should not move quickly to less dense regions. Therefore, it was assumed that the energy density of the cosmic rays is at or near its saturation value. As a trial assumption, Bignami and Fichtel [18] assumed the cosmic ray density to be proportional to the matter density. The fluctuations in matter density are then quite important in determining the expected gamma ray intensity calculated by Eq. (1), since the gamma radiation becomes proportional to n 2. The density distribution of interstellar matter has generally been estimated from 21-cm radio data which, however, indicates only the atomic hydrogen and not the ionized and molecular hydrogen. There are in addition some problems associated with the direct interpretation of the 21-cm data as discussed, for example, by Simonson [21], First, there is clearly significant absorption of the 21-cm line over a band in galactic longitude about the galactic center, and also there are indications of high optical depth along spiral arm segments. Second, the interpretation of the observed intensity in the 21-cm line in terms of density depends on the assumed galactic velocity field, and there is increasing reason to believe the velocity pattern is not as simple as assumed in the earliest models. I t is actually this latter problem which is of greater concern here, because it affects the peak to valley ratio of the matter density distribution. 45

Space Uescarcli XV

706

C. E.

Fichtel

14 M IO r n — i — i — i — i — i — r

~l

1

1

1

1—;

t-IO* 30 MeV. The initial phase separation of matter and anti-matter leads ultimately to regions of pure matter and pure anti-matter of the size of galactic clusters. Stecker et al. [35] have predicted the gamma ray spectrum which would be expected from annihilation at the boundaries of such clusters from the beginning of their existence to the present. This spectrum is very similar (essentially indistinguishable) to the one in Fig. 7 in the energy range for which data exist, and is not included in the figure for that reason. The final model involves cosmic ray interactions with the blackbody radiation at an early point in cosmological time. Wolfendale [36] has shown that this theory is also a possibility. 3.3. Low Energy Gamma Ray Bursts In 1973, Klebesadel et al. [37] reported the detection of low energy gamma ray bursts. These have subsequently been confirmed by other groups. The most recent catalog of these hard X-ray—soft gamma ray bursts is that of Strong et al. [38]. The SAS 2 anticoincidence dome provides a very large detector for the high energy portion ( > 0.3 MeV) of these events and for completeness the observations of SAS 2 will be reported here, even though they fall in the low energy gamma ray category. From any side or the top of the SAS 2 experiment, there is an effective area of about 2.5 X 103 cm2. The energy threshold varies with position over the 2-cm thick dome and with incident angle at any one point. However, except in the vicinity of the bottom rim its response is fairly uniform. On averaging over the dome it is found that the effective threshold for detection is about 0.15 MeV, 46*

710

C . E . FICHTEL

the efficiency rises to about 4% at 0.2 MeV, 15% at 0.4 MeV, and 20% at 0.6 MeV. The counting rate of the anticoincidence dome when there is no increase due to the trapped radiation in the South Atlantic magnetic anomaly is about 4.2 x 103 counts s - 1 and it remains quite steady. During the period of its operation, 20 November 1972 through 8 June 1973, SAS 2 [11] detected two events observed by other satellites at 2327.53 on 2 March

VELA (5A)

VELA (6A)

J

k IT 30 32 34 36 38 : (SECONDS AFTER 7 HRS 7 MINS UTI

9 10 II 12

1

1

1

!

1

1

1 -

)

-

—

1

1 region

-20

/

y*

-40 -

^

\ maximuny

opticalaxis

-60

-SO

1

I

ISO 140

1 120

1

[

1

60

1

40

1

1

1

!

1

1

1

1

1

20 0 340 320 300 260 260 240 220 200 equatorial right ascension

Fig. 3. Region of the sky seen by the experiment of 21 November 1973. Regions of minimum and maximum according to OGO 5 results are indicated. The broken line shows the trace of the zero Doppler shift circle. t h e sun, a t an ecliptic longitude of 47°14'. I t s velocity was 26 k m s _ 1 a n d it was pointing in t h e direction « = 123°, ô = 0.6° in t h e heliocentric ecliptic system of coordinates. T h e intensity recorded shows a modulation with a period equal to t h e spin period of t h e spacecraft d u e to t h e well-known intensity distribution of t h e L y m a n 47

Space Research XV

724

J . L . BERTATTX, J . E . B L A M O N T e t a l .

alpha emission of the interstellar wind. A strong signal due to the Orion region is superimposed; its relative position to the interplanetary features provides the direction and phase of the rotation of the probe. The reduction factor indicates the absorption by hydrogen of Lyman alpha light emitted by atoms with the same velocity as the cell. If the velocity of the source is different from the velocity of the cell, the absorption is decreased because of the Doppler shift (Fig. 1) and the reduction factor is increased. The reduction factor is minimum for directions of sight which are perpendicular to the relative velocity of the probe and the in-

J

Rayleighs

1

1

t

1

rf T

71

mai

__

1

V -

20'mm

- w

1 1 1

101 WO

1 270

!

o

,

se

i

,

m