The Outer Heliosphere: The Next Frontiers, Volume 11

COSPAR COLLOQUIA SERIES VOLUME 11

THE O U T E R HELIOSPHERE: THE NEXT F R O N T I E R S

PERGAMON

This Page Intentionally Left Blank

THE OUTER HELIOSPHERE: THE N E X T FRONTIERS Proceedings of COSPAR Colloquium held in Potsdam, Germany 24-28 July 2000

Edited by

Klaus Scherer dat-hex, Katlenburg-Lindau, Germany

Horst Fichtner Institutfiir Theoretische Physik IV, Ruhr-Universiti~tBochum, Bochum, Germany

Hans J0rg Fahr Institut Ff~rAstrophysik und extraterrstrische Forschung, Universittit Bonn, Bonn, Germany

Eckart Marsch Max-Planck-Institutfiir A eronomie, Katlenburg-Lindau, Germany

2001

) PERGAMON An Imprint of Elsevier Science Amsterdam

- London

- New

York

- Oxford

- Paris

- Shannon-

Tokyo

ELSEVIER SCIENCE B.V. Sara Burgerhartstraat 25 P.O. Box 211, 1000 AE Amsterdam, The Netherlands

Copyright 9 2001 COSPAR. Published by Elsevier Science B.V.

This work is protected under copyright by Elsevier Science, and the following terms and conditions apply to its use: Photocopying Single photocopies of single chapters may be made for personal use as allowed by national copyright laws. Permission of the Publisher and payment of a fee is required for all other photocopying, including multiple or systematic copying, copying for advertising or promotional purposes, resale, and all forms of document delivery. Special rates are available for educational institutions that wish to make photocopies for non-profit educational classroom use. Permissions may be sought directly from Elsevier Science Global Rights Department, PO Box 800, Oxford OX5 1DX, UK; phone: (+44) 1865 843830, fax: (+44) 1865 853333, e-mail: [email protected]. You may also contact Global Rights directly through Elsevier's home page (http://www.elsevier.nl), by selecting 'Obtaining Permissions'. In the USA, users may clear permissions and make payments through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; phone: (+1) (978) 7508400, fax: (+1) (978) 7504744, and in the UK through the Copyright Licensing Agency Rapid Clearance Service (CLARCS), 90 Tottenham Court Road, London W1P 0LP, UK; phone: (+44) 207 631 5555; fax: (+44) 207 631 5500. Other countries may have a local reprographic rights agency for payments. Derivative Works Tables of contents may be reproduced for internal circulation, but permission of Elsevier Science is required for external resale or distribution of such material. Permission of the Publisher is required for all other derivative works, including compilations and translations. Electronic Storage or Usage Permission of the Publisher is required to store or use electronically any material contained in this work, including any chapter or part of a chapter. Except as outlined above, no part of this work may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission of the Publisher. Address permissions requests to: Elsevier Science Global Rights Department, at the mail, fax and e-mail addresses noted above. Notice No responsibility is assumed by the Publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made.

First edition 2001 Library of Congress Cataloging in Publication Data A catalog record from the Library of Congress has been applied for.

ISBN:

0 4 4 4 50909 7

@ The paper used in this publication meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). Printed in The Netherlands.

Preface The 11th COSPAR Colloquium "The Outer Heliosphere: The Next Frontiers" was held in Potsdam, Germany, from July 24 to 28, 2000, and is the second dedicated to this subject after the first one held in Warsaw, Poland in 1989. Roughly a century has passed after the first ideas by Oliver Lodge, George Francis FitzGerald and Kristian Birkeland about particle clouds emanating from the Sun and interacting with the Earth environment. Only a few decades after the formulation of the concepts of a continuous solar corpuscular radiation by Ludwig Biermann and a solar wind by Eugene Parker, heliospheric physics has evolved into an important branch of astrophysical research. Numerous spacecraft missions have increased the knowledge about the heliosphere tremendously. Now, at the beginning of a new millennium it seems possible, by newly developed propulsion technologies to send a spacecraft beyond the boundaries of the heliosphere. Such an Interstellar Probe will start the in-situ exploration of interstellar space and, thus, can be considered as the first true astrophysical spacecraft. The year 2000 appeared to be a highly welcome occasion to review the achievements since the last COSPAR Colloquia 11 years ago, to summarize the present developments and to give new impulses for future activities in heliospheric research.

The conference drew not only scientific but also public interest, which became evident from a variety of related TV and radio presentations as well as a couple of articles in the local press. To cover all aspects of outer heliospheric research and future exploration the conference programme was structured into eleven major sessions, namely: the Large-scale Structure of the Heliosphere, Heliospheric and Interstellar Connections, Messengers from the Outside, Echoes from the Heliopause, the Heliosphere and the Galaxy, Views of Space Agencies and Companies, At the Edge of the Solar System, New Kinetic Aspects of Heliospheric Physics, Modern Heliospheric Spacecraft and Missions, Modern Heliospheric Technologies, and Connections to Earth.

my---

Preface We thank all the participants who have submitted a written version of their presentations and also all referees who helped to improve these manuscripts, and, thus, enabled us to compile this volume of high-quality articles. The remaining presentations are included with their abstracts, and, if published elsewhere, a reference is given. A special thanks go to the conveners for putting together the invited talks. Finally, we are grateful that the Colloquium was held under the auspices of COSPAR.

Klaus Scherer Horst Fichtner Eckart Marsch Hans J. Fahr

--

vi'-

Acknowledgements

-- vii-

Prologue Following the example of the first COSPAR Colloquium of the Outer Heliosphere, also during the second COSPAR Colloquium on the same topic a poll about the distance of the heliospheric termination shock has been taken from the participants. The results of both polls are shown in the figure. On the basis of these "data" we are now able to perform a "non-standard" analysis of the shock dynamics.

Mean S h o c k D i s t a n c e 96.96 [AU]

20.0.

COSPAR

g

~

i0(~

90.0 H e l i o s p h e r i c t e r m i n a t i o n s h o c k d i s t a n c e [AU]

Figure 1. The polls taken at the two COSPAR Colloquia on the Outer Heliosphere in New Hampshire (left panel) and in Potsdam (right panel).

Vshock ~

~ 3.27AU yr -1 11 yrs which is, interestingly enough, about the speeds of the Voyager 1 and 2 spacecraft, of 3.5 and 3.1 AU yr -1. We included an additional question in the Potsdam poll: Will the Voyagers reach the shock at all ? The result was, that 83% of the participants of the poll expected this to happen, 11% did not, and 6% made no guess. From our high-quality data set and the fact that Voyager 1 was at 76 AU at the time of the Colloquium, we can estimate the shock encounter to happen in the year 2091 at a distance of about 395 AU under the (probably wrong) assumption that the shock is continuously moving outward. To check these and other (more serious) predictions made during this second Colloquium on the Outer Heliosphere, we suggest to organize a third one in the year 2011, i.e. again another solar activity cycle later.

--

viii-

Opening Remarks by the COSPAR President

Dear Organizers, Dear Participants, Ladies and Gentlemen,

It is a pleasure for me to convey the best wishes of COSPAR Bureau and Secretariat for a successful Colloquium and at the same time our sincere thanks to the organizers and editors for having undertaken the task to organize this event and publish its proceedings. The great number of attendees tells me that it was timely and worth while to do so. Let me formulate a few thoughts which were stimulated by the theme of your meeting, the outer heliosphere. The scales of the universe are incomprehensible. We, who deal every day with them, have of course developed a way to work with these scales by introducing the parsec as unit and thus shrink the universe into something that can be handled and visualized. We can easily feel at home in this vast space and forget how small the part is that we are occupying. The heliosphere, although small by comparison, is well suited to teach us modesty in two respects. We realize the true size of cosmic distances much better when we think that it takes one day for light to cross the heliosphere, with the nearest star being at a distance of four light years. But it makes us even more humble when we reflect about the completely accidental and by no means distinguished location of the heliospere inside some odd filament of a supernova remnant. If we truely tried to absorb these facts, should we not be equallly shaken as the Portuguese were in 1755 after the earthquake of Lisbon? Under the impression that it struck good and evil people, pious ones and heretics without differentiation, they began to ask questions like: is god not good? does god really exist? Should our questions not be: is our life, are we really the culmination of the universe? are we unique? For you who have assembled here during this week the heliosphere is a place of action, of the solar wind colliding with the interstellar medium, of the interplanetary magnetic field winding up and forming shocks, of interstellar ions being picked up to be accelerated to very high energies, of radio waves bringing back messages from the boundaries, of dust particles telling stories about the chemistry in the expanding clouds of exploding stars, and of many other cosmic processes which you can study here in situ in an exemplary fashion. The communication of your latest insights will keep you away from unhealthy thoughts like the above. You might also consider more practical matters like advancements of the propulsion techniques or the technology of large extended surfaces enabling farther and more efficient space travel. You may wonder whether Europe is going to confine itself for ever to "- iX--

Opening remarks by the COSPAR President

the innermost part of the solar system because of its reluctance to fly RTG's. I think that a meeting like this could also serve as a platform from which messages can go to the agencies and the public about the long-range needs of your science and its benefits in understanding our position in the universe. There is also a task in education. I do not think that scientists have to convey to the public all the deep lying discoveries of their respective disciplines. But teaching them respect for the basic laws of nature and making them aware of the reality of our place in the universe is a duty that we should accept again and again. The imagination of our children is so full of aliens, of UFO's and supermen that they may think that everything is possible and never realize how confined we are to our vulnerable planet. Telling stories about the solar system, the true extent and emptiness of the heliosphere, the efforts and times needed to conquer its distances will not only correct some of the light-heartedly accepted illusions, but also raise curiosity about the wider environment and origin of our Earth. Heliospheric physics is eminently suited to fulfill such an educational role. Finally, I have a few words about the publication of your contributions within the COSPAR Colloquium Series. You and also COSPAR want the book to be purchased and read by a community much wider than what is assembled in this room. Therefore, quality of the papers and reasonably wide interest of the topics are of great importance. Not all of the preceding volumes fulfill these conditions. To achieve the goal implies careful refereeing and editorial judgement. One may also want to fill some gaps by accepting additional papers. You who are involved in this process do a great service to your and the wider space science community. Let me express once more my sincere thanks and wish you an exciting week in Potsdam.

Gerhard Haerendel

--"

X

"-"

Contents S e s s i o n 1: L a r g e S c a l e S t r u c t u r e

Physics of the solar wind interaction with the local interstellar medium G.P. Zank and H.-R. Mfiller Relating models of the heliosphere to Lyman-a absorption observed in Hubble spectra H.-R. Miiller and B.E. Wood

13

Interstellar atoms in the heliospheric interface V. Izmodenov

23

MHD modeling of the outer heliosphere: Numerical aspects N.V. Pogorelov

33

Interaction of the local interstellar medium with the heliosphere: Role of the interior and exterior magnetic field A. Barnes

43

Modeling stellar wind interaction with the ISM: Exploring astrospheres and their Lyman-a absorption H.-R. Mfiller, G.P. Zank, and B.E. Wood

53

Stationary MHD-equilibria of the heliotail flow D. Nickeler and H.J. Fahr

57

Non-stationary magnetic field geometry in the heliosphere I.S. Veselowsky

61

Solar cycle heliospheric interface variations: Influence of neutralized solar wind N.A. Zaitsev and V.V. Izmodenov

65

Time dependent radiation pressure and time dependent, 2D ionisation rate for heliospheric modeling M. Bzowski

69

*Magnetic fields in the heliosphere: Introductory remarks W.I. Axford

73

*Temporal variations of Lyman-alpha-intensities, backscattered from the interstellar hydrogen atoms upstream of the Sun" The ULYSSES-Gas observations M. Witte

73

General Discussion

75

S e s s i o n 2: H e l i o s p h e r i c

and Interstellar Connections

Radiative transfer at Lyman-a in the outer heliosphere E. Quemerais

79

A fluid approach to the heliosphere/VLISM problem R.L. McNutt Jr, M. Wiltberger, J. Lyon, and C.C. Goodrich

89

Possible effects of the interstellar magnetic field on the heliospheric structure and H-atom penetration to the solar system V.B. Baranov

99

*Oral papers and posters which were given at the conference, but for which no manuscripts were submitted - - xi - -

Contents

Interstellar gas flow into the heliosphere E. M5bius, Y. Litvinenko, L. Saul, M. Bzowski, and D. Rucinski

109

Non-stationary transport of neutral atoms in the heliosphere A.I. Khisamutdinov, M.A. Phedorin, and S.A. Ukhinov

121

Pickup ion turbulence: A stochastic growth model G.P. Zank and I.H. Cairns

125

A time-dependent, 3D model of interstellar hydrogen distribution in the inner heliosphere M. Bzowski, T. Summanen, D. Rucinski, and E. Kyr51~i

129

Charge exchange ionization rate of interplanetary hydrogen atoms and Lyman-a intensity pattern in the inflow direction of the interstellar gas T. Summanen, T. M~ikinen, E. Kyr51~i, and W. Schmidt

133

New results derived from Pioneer 10/11 UV data H. Scherer and K. Scherer

137

*Composition and characteristics of the local interstellar cloud and the inner source obtained from pickup ions G. Gloeckler

141

*H atom velocity distribution in the heliospheric interface V.V. Izmodenov, Y.G. Malama, M. Gruntman, and R. Lallement

141

*Estimation of interplanetary hydrogen flow parameters from SWAN Lymanalpha observations E. KyrSl~i, J. Jaatinen, T. Summanen, W. Schmidt, T. M~ikinen, R. Lallement, E. Quemerais, J. Costa, and J.L. Bertaux

142

*The influence of a time variable solar H-Lyman-alpha line profile on LISMparameter deductions from interplanetary glow spectra H. Scherer, H.J. Fahr, M. Bzowski, and D. Rucinski

142

General Discussion

143

Session 3: M e s s e n g e r s from Outside The discovery and early development of the field of anomalous cosmic rays H. Moraal

147

Anomalous cosmic rays: Current and future theoretical developments J.A. le Roux

163

Anomalous cosmic ray observations in the inner and outer heliosphere B. Heber and A. Cummings

173

Cosmic rays as messengers from outside the inner heliosphere M.A. Lee and H. Fichtner

183

Modulation region of galactic cosmic rays in the heliosphere: Estimation of dimension, radial diffusion coefficient, intensity out of region L.I. Dorman

187

Latitudinal gradients and charge sign dependent modulation of galactic cosmic

191

-- xii-

Contents

Antiprotons below 200 MeV in the interstellar medium: Perspectives for observing exotic matter signatures I.V. Moskalenko, E.R. Christian, A.A. Moiseev, J.F. Ormes, A.W. Strong

195

Anomalous cosmic rays outside of the termination shock A. Czechowski, S. Grzedzielski, H. Fichtner, M. Hilchenbach, and K.C. Hsieh

199

Ionizing media and the observed charge states of anomalous cosmic rays A.F. Barghouty, J.R. Jokipii, and R.A. Mewaldt

203

ACR modulation inside a non-spherical modulation boundary S.R. Sreenivasan and H. Fichtner

207

The injection problem for anomalous cosmic rays G.P. Zank, W.K.M. Rice, J.A. le Roux, and W.H. Matthaeus

211

Self-consistent acceleration of pickup ions at the termination shock J.A. le Roux, H. Fichtner, G.P. Zank, and V.S. Ptuskin

215

Energetic neutral helium of heliospheric origin at 1 AU A. Shaw, K.C. Hsieh, M. Hilchenbach, A. Czechowski, D. Hovestadt, B. Klecker, R. Kallenbach, E. M5bius, and P. Bochsler

219

Effects of the heliospheric termination shock on possible local interstellar spectra for cosmic ray electrons and the associated heliospheric modulation S.E.S. Ferreira, M.S. Potgieter, and U.W. Langner

223

*Galactic cosmic rays: Overview M. Forman

227

*Galactic cosmic rays: the outer heliosphere (late paper: see page 513) J.R. Jokipii

227

*Modulation of galactic and anomalous cosmic rays E. Christian, W. Binn, C. Cohen, A. Cummings, J. George, P. Hink, R. Leske, R. Mewaldt, E. Stone, T. von Rosenvinge, M. Wiedenbeck, and N. Yanaska

228

*The spectrum of ACR oxygen and its variations in the outer heliosphere from 1992 to 2000 D.C. Hamilton, M.E. Hill, N.P. Cramer, R.B. Decker, and S.M. Krimigis

228

*On the variability of suprathermal pickup He + at 1 AU B. Klecker, A.T. Bogdanov, M. Hilchenbach, A.B. Galvin, E. M5bius, F.M. Ipavich, and P. Bochsler

229

*Observations of pickup ions in the outer heliosphere by Voyager 1 and 2 and implications on pressure balance S.M. Krimigis and T.B. Decker

229

*Consequences of recently observed galactic synchrotron radio emissions on the local interstellar spectrum for cosmic ray electrons U.W. Langner, O.C. De Jager, and M.S. Potgieter

230

*Variation of the fluxes of energetic He + and He 2+ during the passage of corotating interaction regions D. Morris, E. M5bius, M.A. Popecki, L.M. Kistler, A.B. Galvin, B. Klecker, and A. Bogdanov

230

*The cosmic ray electron to positron ratio in the heliosphere M.S. Potgieter and U.W. Langner

231

General Discussion

233

o . .

--

XIII

---

Contents

Session 4: Echoes from t h e H e l i o p a u s e Energetic neutral atom imaging of the outer heliosphere- LISM interaction H.O. Funsten, D.J. McComas, and M. Gruntman

237

Dual spacecraft measurements as a tool for determining the source of low frequency heliospheric radio emissions W. S. Kurth and D. A. Gurnett

245

Theories for radio emissions from the outer heliosphere I.H. Cairns and G.P. Zank

253

Mapping the heliopause in EUV M. Gruntman

263

Energetic neutral hydrogen of heliospheric origin observed with SOHO/CELIAS at 1 AU M. Hilchenbach, K.C. Hsieh, D. Hovestadt, R. Kallenbach, A. Czechowski, E. MSbius, and P. Bochsler

273

Acceleration of pickup ions at the termination shock in the limit of weak scattering S.V. Chalov and H.J. Fahr

277

Doppler shifted photon emission expected due to reactions of energetic protons with the LISM atoms in the heliosphere M. Hilchenbach, K. C. Hsieh, and A. Czechowski

281

A new diagnosis tool to map the outer heliosphere regions R. Ratkiewicz and L. Ben-Jaffel

285

The Lyman-a echo from the heliospheric bow shock region and its observability from earth H.-J. Fahr, H. Scherer, G. Lay, and M. Bzowski

289

*Energetic neutral atoms in the heliosphere M.A. Gruntman

295

*Radio emission from the outer heliosphere D. Gurnett and W.S. Kurth

295

*Energetic neutral atoms as tracers of the ionization state of the local interstellar medium V.V. Izmodenov and M. Gruntman

296

*Simulation of ENA images of the heliospheric termination shock and interface region E.C. Roelof

296

General Discussion

297

Session 5: T h e H e l i o s p h e r e a n d G a l a x y The solar wind: Probing the heliosphere with multiple spacecraft J.D. Richardson

301

Propagation of the solar wind from the inner to outer heliosphere: Three-fluid model C. Wang and J.D. Richardson

311

Relationships of corotating rarefaction regions outside 40 AU with solar observations: Heliospheric mass loading A. Posner, N.A. Schwadron, and T.H. Zurbuchen

315

-- xiv --

Contents

Recurrent ion events and plasma disturbances at Voyager 2" 5 to 50 AU R.B. Decker, C. Paranicas, S.M. Krimigis, K.I. Paularena, and J.D. Richardson

321

Mapping the detailed structure of the local interstellar medium S. Redfield and J.L. Linsky

325

Effect of different possible interstellar environments on the heliosphere: A numerical study H.-R. Mfiller, G.P. Zank, and P.C. Frisch

329

General Discussion

333

Session 6: View of Space Agencies and Companies Solar sail technology development and application to solar system exploration M. Leipold

337

*Contributions to the Outer Heliospheric Missions in frame of the German Space Program O. RShrig

345

*Deep space missions and technological challenges C. Schalinski and K. Eckardt (Astrium Consortium)

345

General Discussion

347

Session 7: At the E d g e of the Solar System The probable chemical nature of interstellar dust particles detected by CIDA on Stardust J. Kissel, F.R. Kriiger, J. Sil@n, and G. Haerendel

351

In-situ studies of interstellar dust from spacecraft I. Mann and H. Kimura

361

Dynamics of interstellar dust at the heliopause A. Czechowski and I. Mann

365

*Kuiper Belt Objects S.F. Green

369

*Dust in the Outer Heliosphere and Interstellar Dust E. Griin

369

*Comets as a source of heliospheric ions J. Geiss, G. Gloeckler, and K. Altwegg

370

*A Cometary and Interstellar Dust Analyzer for the STARDUST Mission J. Kissel, A. Glasmachers, H. von Hoerner, and H. Henkel

370

*Orbital Evolution of Dust in the Outer Heliosphere under the dust-gas drag force K. Scherer

371

General Discussion

373

Session 8: New Kinetic Aspects of Heliospheric Physics The injection problem J. Giacalone

377

A hydrokinetic description of solar wind electrons using hemispheric distribution functions I.V. Chashei, H.J. Fahr, and G. Lay

387

"--)iV --

Contents

*The solar wind electron velocity distribution O. Lie-Svendson

393

*General aspects of Boltzmann H-theorem for generalized collisions R.A. Treumann

393

*Heating and acceleration of ions by cyclotron- and Landau-resonances E. Marsch and C.-Y. Tu

394

General Discussion

395

S e s s i o n 9: M o d e r n H e l i o s p h e r i c Spacecraft and M i s s i o n s

In-space nuclear power as an enabling technology for exploration of the outer heliopause R. Sackheim, M. Van Dyke, M. Houts, D. Poston, R. Lipinski, J. Polk, and R. Frisbee

399

Interstellar probe using a solar sail: Conceptual design and technological challenges P.C. Liewer, R.A. Mewaldt, J.A. Ayon, C. Garner, S. Gavit, and R.A. Wallace

411

Propulsion for interstellar space exploration G. Genta

421

A realistic interstellar probe R.L. McNutt Jr, G.B. Andrews, J.V. McAdams, R.E. Gold, A.G. Santo, D.A. Ousler, K.J. Heeres, M.E. Fraeman, and B.D. Williams

431

Artificial intelligence techniques for the onboard analysis of space science data P.R. Gazis

435

Sunlensing the cosmic microwave background from 763 AU C. Maccone

439

*Propulsion options for the interstellar probe mission L. Johnson

445

*Solar Orbiter- A high resolution mission to the Sun and the inner heliosphere E. Marsch, B. Fleck, and R. Schwenn

445

General Discussion

447

S e s s i o n 10: M o d e r n H e l i o s p h e r i c T e c h n o l o g i e s

Scientific Payload for an interstellar probe mission R.A. Mewaldt and P.C. Liewer

451

Using multilayer mirrors to detect photons from the heliopause B.R. Sandel

465

A cosmic ray detector for an interstellar probe W. DrSge, B. Heber, M.S. Potgieter, G.P. Zank, and R.A. Mewaldt

471

*Future observations of the outer heliosphere M. Hilchenbach and H. Rosenbauer

475

*State-of-the-art solid state arrays and advanced analog microelectronics for heliospheric physics H.D. Voss

475

General Discussion

477

-- xvi --

Contents

Session 11: C o n n e c t i o n s to E a r t h

The heliosphere as viewed from Earth E.N. Parker

481

The heliosphere, cosmic rays, climate K. Scherer, H. Fichtner, and O. Stawicki

493

*Heliospheric changes in the past: Evidence from cosmogenic isotopes in the polar ice J. Beer

497

*The UV radiation climate on Earth and its impact on the biosphere: past, present and future trends G. Horneck

498

General Discussion

499

S e s s i o n 12: M i s c e l l a n e o u s

Aurora vortex structures as a result of disturbed geomagnetic conditions M.A. Danielides and A. Kozlovsky

503

A non-solar origin of the "SEP" component in lunar soils R.F. Wimmer-Schweingruber and P. Bochsler

507

*Disconnection events in P/Halley's plasma tail M.R. Voelzke and H.J. Fahr

511

Late paper:

Cosmic rays in the outer heliosphere and nearby interstellar medium J.R. Jokipii

513

List of participants

521

~176

--"

XVII

---

This Page Intentionally Left Blank

Session 1: Large Scale S t r u c t u r e

This Page Intentionally Left Blank

Physics of the Solar Wind Interaction with the Local Interstellar Medium G.P. Zank and H.-R. Mtfller Bartol Research Institute, University of Delaware, Newark, DE 19716, USA Plasmas in interstellar and interplanetary space are frequently partially ionized. Thus, the solar wind and stellar winds often interact with an interplanetary medium that is an admixture of protons, electrons, other charged ions, and neutral atoms. For example, the very local interstellar medium surrounding our heliosphere may be less than 50% ionized, with the dominant constituent being neutral hydrogen (H). As a result, the composition of the solar wind in the outer heliosphere beyond some 10 - 15 AU is dominated by neutral interstellar H. Our understanding of the complex physics describing the interaction of the solar wind with the partially ionized local interstellar medium has advanced significantly in the last 5 years with the development of very sophisticated models which treat the coupling of neutral atoms and plasma self-consistently. A number of major predictions have emerged from these models, such as the existence of a large wall of heated neutral hydrogen upstream of the heliosphere. Remarkably, in the ensuing years, this prediction has been confirmed by high resolution Hubble Space Telescope Lyman-~ spectroscopic data. An introductory review of the physics, and associated observations, of the interaction of the solar wind with the interstellar medium is presented for this exciting, rapidly developing field. 1. INTRODUCTION The solar wind forms a bubble, called the heliosphere, in the local interstellar medium (LISM), within which the solar system resides and whose size and properties are determined by the manner in which the solar wind plasma and the partially ionized LISM are coupled. In the last decade, great progress has been made in our understanding of the physical processes thought to describe the outer heliosphere. Fundamental to these advances has been the recognition that the interstellar medium is coupled intimately to the heliosphere itself and that much of outer heliospherie physics cannot be understood independently of the local interstellar medium. With the possibility that the aging spaeeerat~ Voyager 1,2 and Pioneer 10,11 might encounter the heliospherie boundaries in the not too distant future, interest in the far outer reaches of our solar system and the LISM has been rekindled. A convenient, if slightly vague, definition of the outer heliosphere, and one that is adopted here, is that it is that region of the solar wind influenced dynamically by physical processes associated with the LISM. Thus, for example, neutral interstellar hydrogen is the dominant (by mass) constituent of the solar wind beyond an ionization cavity of-~6 - 10 astronomical units (AU) in the upstream direction (the direction anti-parallel to the incident interstellar wind). The neutral hydrogen is coupled weakly to the solar wind plasma via resonant charge exchange - a coupling which leads to the production of pickup ions that come to dominate the internal energy of the solar wind. The solar wind changes then from a small plasma beta ]3sw (the ratio of plasma pressure to magnetic field pressure) environment to one in which [3sw > 3. Figure 1 shows the complex interaction between solar wind plasma and the LISM. While clearly simplistic, it serves to illustrate the primary physical processes. The interstellar plasma may or may not impinge supersonically on the heliosphere. The LISM flow is diverted about the heliospherie obstacle either adiabatically or by a bow shock (BS). The LISM plasma is coupled to the neutral interstellar hydrogen (H) primarily through charge exchange. Charge exchange corresponds to a neutral atom surrendering an electron to a passing ion (typically a proton), so creating a new charged particle and (a hydrogen) atom. Thus, when neutral H charge exchanges with LISM plasma that has been decelerated,

-3-

G.P. Zank and H.-R. Miiller heated and diverted about the heliosphere, the neutral H acquires a subpopulation that too is heated, decelerated and diverted. This neutral hydrogen, together with the LISM H, is called component 1 and is created in region 1 of Figure 1. If the charge exchange mean free path is sufficiently small in the region of decelerated LISM flow, a "wall" of neutral hydrogen will form in the upstream direction (the hydrogen wall). Resonant charge exchange is essentially a scattering rather than an extinction process. A boundary, called the heliopause (HP), separates the heliosphere and the LISM plasma, and is either a contact or a tangential discontinuity. The supersonic solar wind is decelerated, diverted and heated by a reverse shock called the termination shock (TS). The location of the TS is determined by the steady and temporal solar wind pressure exerted on the LISM.

H3-Xk'",, /

HAcR I

+ \

\ d ; . X ~ 2,. I

Figure 1. Schematic of the solar wind - LISM boundary regions which act as neutral H sources whose characteristics are identifiably distinct. The solar wind is enclosed by the termination shock (the curve labeled TS) and is identified as region 3. The shocked solar wind, the heliosheath, is bounded by the termination shock (TS) and the heliopause (HP) and is identified as region 2, while the interstellar medium is found beyond the heliopause. The LISM may or may not form a bow shock (BS) depending on whether the interstellar flow velocity is supersonic. Some sample plasma (H§ and pickup ion (H§ trajectories (dashed) are shown, as well as trajectories of neutral H (solid) which comes into the region from the LISM (Husu) and experiences charge exchange (solid arrow heads) outside the HP (component 1 neutrals, HI), between TS and HP (component 2, H2), and in region 3 (fast component 3, H3).

Neutral interstellar hydrogen that crosses the heliopause into the heliosphere can also experience charge exchange. Neutral H atoms that are created by charge exchange with the very hot subsonic plasma downstream of the heliospheric termination shock possess thermodynamic attributes that correspond to their origin, i.e., very hot (--10 6 K) with a large random velocity and an approximately outwardly directed velocity component. This population of neutral atoms, called component 2, is created in region 2 (Figure 1), and it streams away from the heliosphere and is quite distinguishable from component 1. By means of secondary charge exchange with the LISM plasma, component 2 acts to transport heat from the heliosheath (the region located between the heliospheric termination shock and the heliopause) into the LISM and this leads to an increase in the LISM temperature in the immediate vicinity of the heliopause. The final neutral component, component 3, possesses characteristics quite distinct from the other two components and has its origin in the supersonic solar wind (region 3, Figure 1). Component 3 has large outward radial velocities and relatively low temperatures but is dynamically unimportant owing to its very low energy density. The ions born of charge exchange in the supersonic solar wind, the pickup ions, are, however, of considerable dynamical importance since they remove both momentum and energy from the bulk solar wind flow. The solar wind is decelerated, so reducing the ram pressure and hence the expected size of the heliospheric cavity, and the solar wind acquires a tenuous non-thermalized population of suprathermal ions (typical energies-1 keV). The initial pickup ion population is unstable, generating low frequency magnetic turbulence which can scatter both pickup ions and cosmic rays. Some fraction of the pickup ions is further energized, possibly by shock acceleration, to form the anomalous cosmic ray component, although the precise injection mechanism awaits clarification.

-4-

Physics of the solar wind interaction with the local interstellar medium It should be noted that, historically, LISM conditions were very different in the past than they are presently (and indeed will be in the future) and thus the structure of the heliosphere may have been very different from the models discussed here. 1 Furthermore, since the interstellar medium is frequently partially ionized, we can expect that the interaction of winds of stars whose luminosity is relatively low (comparable to the sun, for example) with their LISM will be similar to that of the heliosphere. Thus, we may expect hydrogen walls to form ahead of the astrospheres of, for example, G-type stars. 2 The material presented here has been reviewed extensively by Holzer and Zank 3 and the reader is referred to these works for more detail and an extended bibliography. 2. THE LOCAL I N T E R S T E L L A R MEDIUM Many of the pertinent physical parameters of the LISM are poorly constrained and so, by implication, the detailed solar w i n d - LISM interaction is not yet well understood. Our best inferences about the LISM flow velocity and temperature have been made from observations of neutral interstellar hydrogen (H) and helium (He) within the heliosphere, using either resonantly scattered solar UV light techniques 4 (since H atoms in the heliosphere are illuminated by the Sun in the HI resonance line i.e., the Lyman-~ line) or in situ spacecraft measurements of the He distribution: The bulk velocity of the LISM flow is-~26 km/s and the plasma temperature is-~8000 K. The H and H + number densities are not well determined, although approximate limits can be derived. For HI (i.e., neutral H only), number densities lie in the range of --4).15 - 0.34 cm -3, while for HII (ionized H), the range is -4).03 - 0.14 cm3, depending on whether the EUV radiation field is included or not. 6 This leads to a limit on the total number density of both neutral and ionized interstellar hydrogen in the range --0.15 - 0 . 3 4 cm 3. Although most models assume that the relative motion of the heliosphere through the interstellar medium is supersonic, it is by no means clear that such an assumption is completely warranted. Our knowledge of both the local interstellar magnetic field strength and orientation and the energy density in interstellar cosmic rays is somewhat rudimentary. Cosmic rays with energies of 300 MeV/nucleon and less may contribute -~ (3 + 2) x 10 -13 dynes cm -2, which can yield an interstellar flow that is either sub-magnetosonic or subsonic. 7 The magnetic field orientation appears to be perpendicular to the interstellar flow velocity vector and estimates for the magnetic field strength range from---1.3 to 2 ktG.S Evidently, our knowledge about important LISM plasma and neutral H parameters is rather poor and this must therefore be taken into account when evaluating models of the global heliosphere. Nonetheless, the basic underlying physical processes appear to be reasonably well understood. 3. GLOBAL H E L I O S P H E R I C STRUCTURE The dynamical or ram pressure (given by p U 2 , 13 the fluid mass density and u the flow speed) and thermal pressure p of the solar wind decrease with increasing heliocentric distance and must reach a value which eventually balances the pressure exerted by the LISM. The relaxation towards pressure equilibrium between the solar and interstellar plasmas is characterized by (i) a transition of the supersonic solar wind flow to a subsonic state, and (ii) a divergence of the interstellar flow about the heliospheric obstacle. The transition of the supersonic solar wind is most likely accomplished by means of a termination shock (TS), and it is anticipated that at least the Voyager spacecraft will encounter this boundary in the early 21 st century. The divergence of the LISM flow about the heliosphere may be accomplished either adiabatically (if the relative motion between the sun and the LISM is subsonic) or by means of a bow shock in the case of supersonic relative motion. If we neglect both the deceleration of the solar wind by resonant charge exchange and temporal variations in the

-5-

G.P. Zank and H..R. Mfiller

solar wind ram pressure, the minimum radius Rrs of the solar wind shock transition can be estimated from 9 Rrs=

~/+3

P0u02

(1)

R0 2(~/+ 1) Po0 ' 2 where PoUo is the solar wind ram pressure at R o = 1AU, y the solar wind adiabatic index, and p~ the total LISM pressure. The LISM pressure term can include the thermal gas, cosmic rays, the interstellar ram pressure, the magnetic field pressure, MHD turbulence, and may be expressed as Poo = pu 2 + Pth + 1"1 + PcR + Paust + P~B2 , where the terms are all evaluated in the LISM and the factor r I is an attempt to include the effects of interstellar magnetic field obliquity. 3' 10.18 Although one can estimate the location of the TS and the heliopause (HP) using simple pressure balance arguments, the interaction of the solar wind with the LISM is fundamentally multidimensional. Thus, the main advances in our understanding of global heliospheric structure since the original analytic models have been more recent and based largely on computer simulations. The initial simulations were of one-fluid gas dynamic models and only now has the inclusion of neutral interstellar hydrogen been considered self-consistently, with the prediction of a hydrogen wall. 11 To model the interaction of the solar wind with a partially ionized LISM, the usual set of 3D MHD equations must be solved. The collisional processes coupling plasma and neutral atoms depend on their relative energies and their relative importance can vary from region to region of a complex plasma system such as the solar wind/LISM interface. The Knudsen number Kn = ~,/L (~, the mean free path of neutral atoms and L a characteristic macroscopic length scale, such as the size of the solar heliosphere,-100 AU), which is a measure of the neutral distribution relaxation distance, is >> 1 inside the heliosphere and-1 in the very local ISM. The introduction of neutral atoms into the magnetized heliospheric plasma with a large Knudsen number implies that the neutral and plasma distributions cannot equilibrate and may possess quite distinct bulk flow speeds and temperatures. Charge exchange between the coupled, non-equilibrated neutral and charged particle populations can therefore introduce distinct new populations of neutral atoms and plasma whose characteristics reflect their parent population. The subsequent interaction and assimilation of the newly created plasma and neutral populations into the existing plasma and neutral distributions may then lead to the substantial modification of the overall partially ionized plasma system. Thus, the total neutral distribution cannot relax to a single Maxwellian distribution, and one must therefore use a multi-component transport or kinetic description for the neutral populations. To work directly with the kinetic description of interstellar and heliospheric neutral hydrogen, one must solve the Boltzmann equation 12 -~t + v . V f +

.V v f=

P-L,

(2)

for neutral H directly. Here, jr(x, v, t) is the neutral hydrogen distribution function expressed in terms of position x, v, and time t. F is the force acting on a particle of mass m, typically gravity and radiation pressure. The terms P and L describe the production and loss of neutral particles at (x, v, t) and both terms are functions of the plasma and neutral distributions. An alternative approach, and one that is much less prohibitive computationally, is to recognize that the heliosphere-LISM environment comprises three thermodynamically distinct regions; the supersonic solar wind (region 3), the very hot subsonic solar wind (region 2), and the LISM itself (region 1 ) Figure 1. Each region acts as a source of neutral H atoms whose distribution reflects that of the plasma

-6-

Physics of the solar wind interaction with the local interstellar medium

distribution in the region. One may identify neutral components 1, 2, and 3 with neutral atoms originating from regions 1, 2, and 3. Each of these three neutral components is approximated by a distinct Maxwellian distribution function appropriate to the characteristics of the source distribution. This observation allows the use of simpler production and loss terms for each neutral component. The complete highly non-Maxwellian H distribution function is then the sum over the three components, i.e., 3

f ( x , v , t ) = ~--'f ( x , v , t ) . i=1

Under the assumption that each of the neutral component distributions is approximated adequately by a Maxwellian, one obtains an isotropic hydrodynamic description for each neutral component, and thus a closed multi-fluid transport model describing selfconsistently the interaction of the solar wind with the partially ionized LISM. ~3

Figure 2. (a) A 2D plot of the global plasma structure for the two-shock model when neutral H is included self-consistently. The contouring refers to Logl0[temperature]. (b) A 2D plot of the corresponding component 1 neutral distribution, where the shading refers to the number density. The hydrogen wall is clearly visible between the bow shock and the heliopause.

Models of the heliosphere that do not include neither interstellar neutrals, interplanetary nor interstellar magnetic fields TM show the basic structure of terminations shock and heliopause. The TS possesses a characteristic bullet-shape, quite unlike the approximately spherically symmetric heliosphere that is assumed typically. The supersonic solar wind velocity is constant until the very strong TS (Mach number-~170). The solar wind temperature decays adiabatically with increasing heliocentric distance. An extended and very hot tail of subsonic solar wind is formed (called the heliotail) in the downstream direction. When neutral H is included, a major effect of charge exchange on the heliospheric interfaces is to decrease the distances to the TS, HP and BS. 15The decrease in distance results primarily from the reduction in solar wind ram pressure, this due to the mediation of the wind by charge exchange. Besides the distance to the various heliospheric boundaries, illustrated in Figure 2a, charge exchange affects the shape of the termination shock, making it more spherical compared to the purely ionized gas dynamic description. This is a consequence of charge exchange in the heliosheath ensuring that the sheath plasma remains subsonic in this region.

Owing to the deposit of interstellar protons in the solar wind when charge exchange is included, the solar wind flow now departs slightly from simple spherical symmetry, and the contribution to the internal energy of the supersonic solar wind by pickup ions is substantial. The tail region is also cooler when the proton fluid and neutral fluid are coupled compared to no-charge exchange models. The very hot heliotail is cooled by charge exchange with cooler component 1 neutrals. To enter the heliosphere, a hydrogen atom has to cross three thermodynamically distinct regions. Through resonant charge exchange, the diverging LISM flow and the shocked solar wind flow act to

-7-

G.P. Zank and H.-R. Mfdler

divert some fraction of the incident interstellar neutral H flux away from the heliosphere. Figure 2b, a 2D plot of the component 1 neutral H density distribution, shows that inflowing component 1 neutrals are decelerated substantially and filtered by charge exchange with the interstellar plasma between the BS and HP in the upstream direction. This leads to the formation of a hydrogen wall with maximum densities-~0.2-0.3 cm -3, column densities .--10TM cm 2, and temperatures ranging from 18 000K to 28 000K (depending on the assumed LISM parameters). The pile-up in the neutral gas results from the deceleration and deflection of the neutral flow by charge exchange with the interstellar plasma, which is itself decelerated and diverted due to the presence of the heliosphere. Note that the charge exchange mean-flee-path is typically less than the separation distance between the HP and BS and so a large part of the incident interstellar neutrals experience charge exchange. Component 2, produced via charge exchange between component 1 and hot shocked solar wind plasma between the TS and HP, streams across the HP into the cooler shocked interstellar gas and heats the plasma through a second charge exchange. This leads to an extended thermal foot abutting the outside edge of the HP. This heating of the plasma by component 2 serves to broaden the region between the BS and HP, as well as to (indirectly) further heat the component 1 interstellar neutrals after subsequent charge exchanges. Some heating of the unshocked LISM also occurs upstream of the BS, thereby marginally reducing the Mach number of the incident interstellar wind. The temperature of component 1 neutrals once inside the heliosphere remains fairly constant in the upstream region, and represents a substantial increase over the assumed LISM temperature. A further increase in the component 1 temperature occurs in the downstream region.

Figure 3a. An example of a kinetic simulation described by Mailer et al. 16for a 2-shock model. One-dimensional profiles of plasma density np and plasma temperature Tp are shown as dashed lines over radial distance R from the sun, and neutral density nH and (averaged) temperature THas solid lines. The profiles are obtained in the upstream direction, antiparallel to the LISM flow. The middle row depicts two-dimensional neutral velocity distribution functions (logarithmic density scale) at various locations on that axis, with prominent HLISM and H~ 26km/s neutrals, and evidence of 100-300 km/s H2 and 400 km/s H3 neutrals.

Figure 3b: Normalized H I radial velocity distribution at a point in the hydrogen wall located 160 AU from the Sun in the upwind direction of the interstellar flow into the heliosphere. The jagged line is the distribution given by the Boltzmann code, and the smooth curve is the distribution resolved by the multifluid code through three-component MaxwellianslS. The total distribution is clearly non-Maxwellian.

The number density of component 1 crossing the TS is approximately half the assumed incident LISM number density ("filtration"). The density varies only weakly between the TS a n d - 1 0 AU from the Sun in the upstream region. In the downstream direction, component 1 densities are lower within the heliosphere. The 1D component 1 density and temperature profiles are plotted in Figure 3a, and the heated hydrogen wall is exhibited clearly. A further effect of filtration is to decelerate the upstream neutral gas from 26 krn/s in the LISM to -~19 km/s at the TS in the region of the nose. Deflection of the

-8-

Physics of the solar wind interaction with the local interstellar medium

flow also reduces the radial velocity component at angles away from the nose (a deceleration which is in accord with observations of the Lycx resonance line scattered by neutral H in the heliosphere). Also plotted in Figure 3 (taken from a kinetic simulation ~6) is the evolution of the neutral H distribution from the interstellar medium into the heliosphere. Components 1 and 2 are not always easily distinguished in 1D omnidirectional plots (Figure 3b) but the departure from a Maxwellian distribution is pronounced. The smoothed curve shows the distribution predicted by a multi-fluid model, which assumes a superposition of Maxwellian distributions corresponding to components 1, 2, and 3. While clearly distinct in detail, the basic aspects of the neutral distribution predicted by kinetic and multifluid models are surprisingly similar. Models which assume a subsonic LISM flow do not have a bow shock. Instead, upstream of the heliopause, the LISM flow is compressed adiabatically. This more gradual compression leads to the formation of a lower amplitude hydrogen wall that is more extended in the radial direction. It is also less extended in the transverse direction because of the localized nature of the adiabatic compression. The maximum density of the wall in the upstream direction is smaller than that of the two-shock counterpart (though still larger than the incident LISM number density). However, because it is wider, the column density is comparable with the two-shock case. The heliosphere is smaller due to the higher assumed LISM pressure. Three dimensional models of the solar wind during solar minimum introduce some interesting differences from the 2D models discussed above. The presence of an anisotropic solar wind with a high velocity, low density, high temperature polar steam acts to divert both the subsonic solar wind and the LISM plasma flow into the ecliptic region of the heliosphere. ~7The presence of a high density band of plasma about the ecliptic plane increases the filtration of neutrals compared to the higher polar regions. With an increased neutral flux of H over the poles, the polar solar wind experiences comparatively greater deceleration than the ecliptic wind, so reducing the degree of anisotropy seen in 3D gas dynamic simulations. In addition, the global distribution of neutral H is anisotropic in heliolatitude. Structurally, the heliosphere at low latitudes can revert to a "bullet" shape, typically seen only in models which neglect to include interstellar neutral hydrogen. It should be borne in mind that these remarks pertain primarily to the period of solar minimum observed by the Ulysses spacecraft. Very few simulations include either the interplanetary or interstellar magnetic fields dynamically and those that do neglect neutral H completely or use an extremely simplified description. TM Nonetheless, it appears that the magnetic field in the heliosheath can acquire a very interesting structure. Along the stagnation line in the heliosheath, the flow velocity decreases approximately as r -2 (r the heliocentric radial distance), leading to an amplification in the azimuthal magnetic field in the region. This, together with the J x B force causes the equatorial current sheet to bend either upward or downward (depending on solar cycle) away from the equatorial plane. This leads to the formation of an asynmaetric 3D magnetic shell or wall in the outer region of the upstream heliosheath with a gap in the neighborhood of the displaced current sheet. The region between the termination shock and the magnetic wall is occupied by solar wind originating from the middle and high latitudes whereas, thanks to the gap in the magnetic wall, the solar wind in the region between the magnetic wall and heliopause originates from the low latitude equatorial neutral sheet region. The ram pressure pu2of the solar wind varies by a factor o f - 2 with solar cycle. Since the location of the termination shock and heliopause is determined by a balance of the ram pressure and the interstellar pressure, temporal variation in the solar wind ram pressure must therefore play an important role in determining the global structure of the heliosphere. The time dependent multi-fluid model 15has been used to extend earlier studies of the effect of solar cycle variability on the global structure of the heliosphere. 19The varying ram pressure of the solar wind leads to large arrhythmic

-9-

G.P. Zank and H.-R. Miiller

excursions of the TS and smaller excursions of the HP. The motion of the TS drives pressure waves into the inner heliosheath which steepen into shocks which propagate into the LISM. For a two-shock model, the dynamical train of shocks propagating into the LISM acts to increase the separation distance between the HP and BS, whereas in the one-shock case, the emitted shocks act as a surrogate bow shock(s) for the heliosphere. In both cases, the emitted shocks in the LISM introduce some variability into the neutral H entering the heliosphere on long time scales. Finally, it has been found that ion-neutral friction can destabilize the HP. 19 4. HELIOSPHERIC STRUCTURE INFERRED FROM L Y - ~ MEASUREMENTS

Until Voyager 2 encounters the termination shock and heliopause, we are forced to rely on remote observations to infer the structure of the heliosphere. The most promising approach presently is to use the Lytz absorption line in the direction of nearby stars) ~ A neutral hydrogen pileup or wall may be detectable in directions where the decelerated H is red-shifted out of the shadow of the interstellar absorption if the interstellar column density is sufficiently low. Red-shifted excess absorption in Hubble Space Telescope GHRS Lyo~ observations towards o~-Cen (seen previously by Copernicus and IUE) have been interpreted as evidence for the existence of the solar hydrogen wall) 1 Shown in Figure 4 are observations towards o~ Cen A (the jagged solid curves), and the wavelength scale is relative to the Lyman-o~ line center in the heliocentric rest frame. The upper solid curve is the assumed intrinsic stellar Lyo~ emission profile. The accurate representation of the intrinsic stellar profile is unimportant since the absorption features of interest vary sharply. The remaining smooth solid curve (the saturated absorption curve) of Figure 4 gives the attenuation of the stellar Lyo: emission by interstellar H. The column density for H was fixed by scaling to the Deuterium column density, assuming the commonly accepted value of D/H = 1.6 x 10 -5 for the local interstellar medium) 2 Figure 5 shows very clearly that additional absorption is required both redward and blueward of the interstellar feature if the fit is to be completed. Furthermore, the fit must be applied preferentially to the redward side, so arbitrarily changing the D/H ratio is unacceptable. The additional redward absorption has been interpreted as evidence for the detection of the hydrogen wall. The blueward absorption suggests the possibility of a hydrogen wall about o~ Cen A and B. Figure 4. The solid jagged curves are GHRS Lyman-O~absorption profiles Since the column depth of the hydrogen wall is three orders of towards O~ Cen A. The uppermost magnitude smaller than the column depth in the LISM toward t~ solid curve is the assumed intrinsic Cen, it may be surprising at first glance that the heliospheric optical stellar Lyman-~ emission profile. depth at +0.1 A is of the same order as the LISM optical depth at The smoothed solid line, which exthat wavelength. The key difference is that the hydrogen wall is hibits saturated absorption, corresponds to the intrinsic stellar emisheated and decelerated, which both broadens and redshifts the sion line after absorption from the heliospheric component away from the -0.07 A centroid of the LISM only. Corresponding theoreLISM absorption and toward the +0.1 A wavelength of interest. tical absorption profiles are plotted22 Figure 4 also shows the Lyo~ absorption at the red edge of the LISM feature computed from models for three assumed LISM Mach numbers (supersonic, subsonic and ~sonic) 22. These synthetic data are compared with the observed heliospheric absorption. The primary and very important conclusion to emerge from such modeling is that the synthetic Lyo~ profiles support the interpretation of the observed additional redward Lyo~ absorption towards o~-Cen as evidence for the remote detection of the hydrogen wall. Thus, it appears that the hydrogen wall has indeed been observed! The unprecedented accuracy of the Hubble GHRS observations suggests further that, with the refinement of solar wind - LISM interaction models, the question of whether the heliosphere has a one- or a two-

-10-

Physics of the solar wind interaction with the local interstellar medium

shock structure may be answered in the foreseeable future by a combination of modeling and observations as discussed further elsewhere 23' 24. 5. CONCLUDING R E M A R K S Evidently, many more physical processes remain to be incorporated in models describing the interaction of the solar wind with the LISM. This will undoubtedly be accomplished in the next decade as computational power increases and the underlying physics is better understood. However, the most important limitation that the field currently faces concerns the paucity of information about the state of the local interstellar environment. Until LISM parameters are better constrained, all modeling efforts describing the solar w i n d - LISM interaction must be viewed as somewhat hypothetical. Nonetheless, with new observational techniques emerging, new missions planned, and the increasing sophistication of theoretical modeling efforts, the outlook for understanding the physics of the outer heliosphere is bright indeed. Furthermore, with the detection and modeling of hydrogen walls about nearby low luminosity stars 24' 25, n e w opportunities for investigating the physics of hitherto undetectable stellar winds and their LISM now exist. Acknowledgements. This work is supported in part by an NSF-DOE grant ATM-0078650, a NASA grant NAGS-6469, a Jet Propulsion Laboratory contract 959167, and a NASA Space College Grant award. REFERENCES

1 G.P. Zank and P.C. Frisch, Astrophys. J. 518, 965 (1999); H.-R. Mtiller, G.P. Zank, and P.C. Frisch, "Effect of different possible interstellar environments on the heliosphere: A numerical study", submitted to the proceeding of the Cospar colloquium "Outer Heliosphere", Potsdam, 2000, edited by K. Scherer (Elsevier, Amsterdam, 2001). 2 P.C. Frisch, Astrophys. J. 407, 198 (1993). 3 T.E. Holzer, in Solar System Plasma Physics, edited by E.N. Parker, C.F. Kennel, and L.J. Lanzerotti (NorthHolland, New York, 1979), p. 101; G.P. Zank, Space Sci. Rev. 89 3-4, 413 (1999). 4 J.-L. Bertaux, R. Lallement, V.G. Kurt, and E. Mironova, Astron. Astrophys. 150, 1 (1985); E. Qu6merais, B.R. Sandel, R. Lallement, and J.-L. Bertaux, Astron. Astrophys. 299, 249 (1995); R. Lallement, Space Sci. Rev. 78, 301 (1996). 5 M. Witte, M. Banaszkiewicz, and H. Rosenbauer, Space Sci. Rev. 78, 289 (1996). 6 j. Vallerga, Space Sci. Rev. 78, 277 (1996); G. Gloeckler, ibid, 335. 7 W._H" Ip and W.I. Axford, Astron. Astrophys. 149, 7 (1985). s j. Tinbergen, Astron. Astrophys. 105, 53 (1982); P.C. Frisch, Science 265, 1423 (1994); Space Sci. Rev. 72, 499 (1995). 9 L.E. Davis, Phys. Rev. 100, 1440 (1955); E.N. Parker, Interplanetary Dynamical Processes (Interscience, New York, 1963). ~0R. Ratldewicz, A. Barnes, and J.R. Spreiter, S.S. Stahara, M. Vinokur, and S. Venkateswaran, Astron. Astrophys. 335, 363 (1998); R. Ratldewicz, A. Barnes, and J.R. Spreiter, "Local interstellar medium and modeling the heliosphere", submitted to J. Geophys. Res. ~ V.B. Baranov, M.G. Lebedev, and Y.G. Malama, Astrophys. J., 375, 347 (1991). 12V.B. Baranov and Y.G. Malama, J. Geophys. Res. 98, 15157 (1993); V.B. Baranov and Y.G. Malama, J. Geophys. Res. 100, 14755 (1995); H.-R. Mtiller, G.P. Zank, and A.S. Lipatov, Proceedings, Solar Wind 9, Nantucket, 1998, edited by S.R. Habal et al. (AIP Conf. Proceed., 1999), Vol. 471, p. 819; J. Geophys. Res. 105, 27,419 (2000); V.V. Izmodenov, J. Geiss, R. Lallement, G. Gloeckler, V.B. Baranov, and Y.G. Malama, J. Geophys. Res. 104, 4731 (1999). 13H.L. Pauls, G.P. Zank, L.L. Williams, and D.T. Hall, Proceedings, Solar Wind 8, edited by D. Winterhalter et al. (AIP Conf. Proceed., New York, 1996), Vol. 382, p. 605; Zank et al.15; L.L. Williams, D.T. Hall, H.L. Pauls, and G.P.Zank, Astrophys. J. 476, 366 (1997); Liewer (1997). 14T. Matsuda, Y. Fujimoto, E. Shima, and T. Inaguchi, Prog. Theor. Phys. 81, 810 (1989); V.B. Baranov and Y.G. Malama, J. Geophys. Res. 98, 15157 (1993); R.S. Steinolfson, J. Geophys. Res. 99, 13307 (1994); R.S.

-11

-

G.P. Zank and H.-R. Miiller

Steinolfson, V.J. Pizzo, and T. Holzer, Geophys. Res. Lett. 21,245 (1994); S.R. Karmesin. P.C. Liewer, and J.U. Brackbill, Geophys. Res. Lett 22, 1153 (1995); H.L. Pauls, G.P. Zank, and L.L. Williams, J. Geophys. Res. 100, 21595 (1995); C. Wang and J.W. Belcher, J. Geophys. Res. 103, 247 (1998). 15G.P. Zank, H.L. Pauls, L.L. Williams, and D.T. Hall, J. Geophys. Res. 101, 21639 (1996). 16H.-R. MUller, G.P. Zank, and A.S. Lipatov, J. Geophys. Res. 105, 27,419 (2000). 17H.L. Pauls and G.P. Zank, J. Geophys. Res. 101, 17081 (1996); H.L. Pauls and G.P. Zank, J. Geophys. Res. 102, 19779 (1997);A. Barnes, J. Geophys. Res. 103, 2015 (1998); T. Tanaka and H. Washimi, J. Geophys. Res. 104, 12,605 (1999). 18H. Washimi and T. Tanaka, Space Sci. Rev. 78, 85 (1996); Ratkiewicz et al.l~ T. Linde, T.I. Gombosi, P.L. Roe, K.G. Powell, D.L. DeZeeuw, J. Geophys. Res. 103, 1889 (1998); N.V. Pogorelov and T. Matsuda, ibid, 237. 19G.P. Zank, Proceedings, Solar Wind 9, Nantucket, 1998, edited by S.R. Habal et al. (AIP Conf. Proceed., 1999), Vol. 471, p. 783. 20B.E. Wood, H.-R. Mtiller, and G.P. Zank, Astrophys. J. 542, 493 (2000). 21 J.L. Linsky and B.E. Wood, Astrophys. J. 463, 254 (1996); Gayley et al.22 22K. Gayley, G.P. Zank, H.L. Pauls, P.C. Frisch, and D.E.Welty, Ap.J. 487, 259 (1997). 23 Gayley et a1.22;B.E. Wood and J.L. Linsky, Astrophys. J, 492, 788 (1998); B.E. Wood, J.L. Linsky, and G.P. Zank, Astrophys. J. 537, 304 (2000). 24B.E. Wood, J.L. Linsky, H.-R. Miiller, and G.P. Zank, "Observational estimates for the mass-loss rates of ct Cen and Proxima Cen using HST Lyman-ct spectra", Ap J. Lett., 547, 49 (2001). 25H.-R. Miiller, G.P. Zank, and B.E. Wood, Ap.J. 551,495 (2001).

-12-

Relating Models of the Heliosphere to Lyman-c~ Absorption Observed in Hubble Spectra H.-R. Mfiller ~ and B. E. W o o d b* ~Bartol Research Institute, University of Delaware, Newark, DE 19716 bjILA, University of Colorado and NIST, Boulder, CO 80309-0440 We model the interaction of the solar wind with the partially ionized local ISM using a self-consistent hybrid model in which the Boltzmann equation for neutral hydrogen is solved with a kinetic particle code. The degree of external contribution to the ISM plasma pressure (e.g. due to cosmic rays or to magnetic pressure) is varied as a model parameter, resulting in a family of heliospheric models ranging from two-shock models to one-shock heliospheres (subsonic ISM). We give an overview of the neutral hydrogen distributions in these models. The column density and temperature of the heliospheric neutral hydrogen have been observed to be large enough to produce detectable absorption signatures in Lyc~ spectra of nearby stars. The heliospheric models can be used to predict the amount of absorption for various lines of sight, and we compare these predictions with Lyc~ observations of six nearby stars obtained by the Hubble Space Telescope, sampling lines of sight ranging from nearly upwind to nearly downwind. We find that the Boltzmann models tend to predict too much absorption in sidewind and downwind directions, especially when we assume high Mach numbers for the interstellar wind. 1. I N T R O D U C T I O N The basic heliospheric structure created by the interaction between the fully ionized solar wind and the partially ionized local interstellar medium (LISM) can to first order be modeled by considering plasma interactions alone [1-3]. However, charge exchange, whereby an interstellar neutral loses its electron to a proton, allows the neutrals to take part in the solar-wind/LISM interaction in important ways (see review by Zank [4]). Heliospheric models that treat the neutral gas and plasma in a self-consistent manner predict somewhat different properties than plasma-only models, such as shorter distances to the termination shock, heliopause, and bow shock [5-9]. These models also suggest that neutral hydrogen in the heliosphere should be very hot, with temperatures of order 20,00040,000 K. The importance of this prediction is that this high temperature gas should *This research was supported by NASA grant NAG5-9041 to the University of Colorado, by an NSF travel grant administered by the American Astronomical Society, and by NASA grant NAG5-6469, NSF-DOE award ATM-0078650,JPL contract 959167, and NASA Delaware Space Grant College award NGT5-40024 to the Bartol Research Institute.

-13-

H.-R. Miiller and B.E. Wood

produce H I Lya absorption broad enough to be separable from the interstellar absorption observed toward nearby stars, meaning hydrogen in the outer heliosphere can potentially be directly detectable. This absorption has in fact been observed using observations by the Hubble Space Telescope (HST). Excess Lya absorption is evident in high-resolution HST spectra of the nearby stars a Cen A and B, and this absorption has been demonstrated to be consistent with a heliospheric origin [10,11]. Models that self-consistently treat the neutrals and plasma are essential to help interpret these observations, but such models are a complex theoretical and computational problem. The fundamental difficulty is that neutrals in the heliosphere are far from equilibrium, and their velocity distribution at a given location can be highly non-Maxwellian. Lipatov et al. [12] adopted a particle-mesh method for solving the neutral Boltzmann equation, which should yield accurate particle distribution functions (see also [13,14]). This creates the opportunity to use heliospheric models created by this Boltzmann code to predict heliospheric Lya absorption profiles that can be compared with the HST observations. In this paper, we create models assuming different Mach numbers for the inflowing LISM to see which best matches the observed amount of absorption. For the a Cen line of sight, a similar analysis has previously been carried out using four-fluid rather than Boltzmann models [11], and assuming different LISM parameters. We compare the absorption predicted by the models with lines of sight observed by HST, including a Cen, 36 Oph, and Sirius lines of sight, which all show evidence for heliospheric absorption [15,16]. 2. M O D E L

DESCRIPTION

AND

RESULTS

We compare the HST observations of absorption of stellar Lya to absorption profiles synthesized from models of neutral hydrogen (H) in the global heliosphere. We use the fully time-dependent hybrid code described by [14] that includes neutral H and its charge exchange interaction with plasma in a self-consistent way. It consists of a 2D gas-dynamical description of the plasma of both solar and interstellar origin coupled to a 2.5D particle description of neutral hydrogen. The plasma code uses the Euler equations of fluid dynamics to solve for the bulk plasma variables of number density np, temperature Tp, and velocity u. This treatment is justified because the mean free path of the protons of the plasma is small in comparison to heliospheric length scales, and the proton distribution function is Maxwellian. The Boltzmann code for neutral H does not restrict the neutral distribution function to a Maxwellian, but admits arbitrary distributions. Transporting and influencing the particles in phase space, the particle code indirectly solves the Boltzmann equation for the time evolution of the neutral hydrogen distribution function. The Boltzmann code thus treats even those neutral components correctly whose mean free path is no longer smaller than typical heliospheric scales (component 2 neutrals and splash component, see below). For the sake of comparing the spatial distributions of the neutral and the ionized species, it is customary to take moments of the neutral distribution functions to arrive at the neutral H number density nil, effective temperature TH, and bulk neutral velocity u s . The effective neutral temperature is misleading in the sense that one often associates it with a single Maxwellian, which is an oversimplification of the intricate shapes of the neutral distribution functions that are encountered in the heliosphere.

-14-

Relating models o f the heliosphere to Lyman-a absorption...

To first order, the dominant interaction between neutral H and plasma is resonant charge exchange. It is incorporated in each piece of the hybrid code through source terms. The source terms of the Euler gasdynamic equations make use of n i l , TH, and UH. Within the particle code, charge exchange is accounted for by loss and production terms in the neutral Boltzmann equation. These terms invoke the bulk plasma parameters as well as an approximation for the Maxwellian plasma distribution function. For the charge exchange cross section, we use an empirical fit [17]. Charge exchange deletes one neutral particle from the neutral distribution, and adds a new neutral that bears the characteristics of the plasma at the site where the exchange took place. The deleted neutral is most likely from the dominant neutral category, i.e., a "component 1 neutral" originally from the LISM neutral population, and its deletion therefore does not significantly change the neutral distribution. The newly born neutral can have very different characteristics, in particular when exchange takes place in the heliosheath or in the heliotail ("region 2") where the plasma is hot and has a low bulk speed, due to the shock heating and deceleration of solar wind plasma at the termination shock. These so-called component 2 neutrals have thermal speeds on the order of 100 km/s, corresponding to 106 K, and travel in essentially random directions. Some of them have trajectories directed towards the inner solar system where they can be detected as energetic neutral (hydrogen) atoms (ENA) in the low energy range below 1 keV [14]. So-called splash neutrals form another distinct component of neutrals, born inside the termination shock ("region 3") through charge exchange with the solar wind plasma that is still supersonic and cold, except close to the Sun. Charge exchange inside the termination shock heats the plasma through the generation of pickup ions, whereas the plasma in the heliosheath between ter~nination shock and heliopause (also in the heliotail) is cooled by charge exchange. In the heliotail, this mechanism works to equilibrate neutral H with the plasma, thereby heating and accelerating neutral H. Some of the hot component 2 neutrals created in the heliosheath deposit their energy upstream of the heliopause through secondary charge exchange, creating an anomalous heat transfer into the oncoming LISM. The resulting plasma temperature gradient in the LISM upstream of the heliopause is communicated to neutral H through charge exchange, creating a neutral temperature gradient as well. This gradient is even more pronounced when the incoming LISM plasma is supersonic. In this case, a bow shock exists where the plasma becomes subsonic, and the associated shock heating of the plasma steepens the plasma temperature gradient and consequently further heats the neutral H between the heliopause and bow shock. The deceleration of neutral H through charge exchange with the decelerated (stagnating) plasma creates what has been termed a hydrogen wall, a density enhanced region of neutral H upstream and sidestream of the heliopause (and bounded on the other side by the bow shock in two-shock models). The filtration of neutral H by the hydrogen wall creates a depletion of neutral H in the heliosphere. For the LISM boundary conditions we choose TH -- Tp - 8000 K, nH -- 0.14 cm -3, np -- 0.10 cm -3, and v - 26 km/s, which are consistent with observations [18-22]. The inner boundary is an inflow of solar wind plasma at 1 AU with T - 105 K, np - 5.0 cm -3, and v - 26 km/s. 7 is set to 5/3 throughout. Figure 1 serves as an illustration of the spatial distribution of neutral density and temperature of a typical model. The Sun is at the origin and the LISM enters from the right. The stagnation axis, which

-15-

H.-R. Miiller and B.E. Wood

cb)t..

+~ -

-

-

b.16,o ~ii.3~I:~ ,~,.c~-o~

9

-,S.; Oo I))} --

~

~..

~0

-,so-j,.,,_.. -soo-

J

."

9 9 0

+

"o

9

Figure 1. Contour plots of (a) neutral H density and (b) effective neutral temperature (for model 5 of Table 1). The Sun is at the origin; the neutral flow field is illustrated by the streamlines in (b). The five density contour levels (in cm -3) and the temperature levels 20,000, 50,000, and 115,000 K are marked in the plots. The small lower panels show 1D profiles along the stagnation axis. Both value axes have a logarithmic scale.

runs parallel to the LISM flow and through the Sun, is a symmetry axis. One can clearly see the density enhancement above the interstellar value of 0.14 cm -3. The 0.16 cm -3 contour approximately follows the hydrogen wall, whose sunward side coincides with the heliopause. The region of maximum density is enclosed by the 0.20 cm -3 contour. The tailward termination shock intersects the stagnation axis at ~100 AU. Downwind of that we see a density depletion and a temperature increase. The neutral temperature also experiences an increase at the wall and in sidestream regions where the plasma has also an increased temperature. Whether or not a bow shock exists upstream of the heliosphere is currently an open question. The LISM plasma temperature and velocity alone would point to the existence of a bow shock (a "two-shock heliosphere") because the Mach number is M - 1.7. However, there is the possibility that mechanisms such as the pressure due to galactic cosmic rays or the pressure of an interstellar magnetic field contribute to the plasma energy, lowering the Mach number of LISM protons to subsonic values and resulting in a one-shock heliosphere. We model such a pressure contribution using a generic factor a to indicate that only a fraction of the total pressure outside of the heliopause is due to LISM protons. The equation of state then reads P - a n v k B T p , where P is the total pressure. The minimal a is 2, expressing the usual assumption that the LISM electrons have the same density and temperature (and therefore the same contribution to the total pressure) as the proton plasma.

-16-

Relating models of the heliosphere to Lyman-a absorption... Table 1 Model parameters. Model # 1 c~ 2.0 Mach# 1.75 Tpl[K] 8,000

2 3.5 1.32 14,000

3 5.0 1.11 20,000

4 7.5 0.90 30,000

5 9.6 0.80 38,000

6 12.5 0.70 50,000

7 18.4 0.58 74,000

We have run seven simulations with differing values for the parameter c~. The values and the corresponding LISM plasma Mach numbers are listed in Table 1. The models range from supersonic without any additional pressure contribution (model 1) to subsonic oneshock models (4-7). The effective plasma temperature Tpl in Table 1 gives an indication what the plasma temperature would be if only LISM protons and electrons accounted for the total pressure. The real temperature for protons, electrons, and neutral H in the LISM is assumed to be 8000 K at the outer boundary in our models.

Table 2 Hydrogen wall results, and c~ M TH,wall [K] 1 2.0 1.75 89,000 2 3.5 1.32 58,000 3 5.0 1 . 1 1 73,000 4 7.5 0.90 78,000 5 9.6 0.80 70,000 6 12.5 0.70 100,000 7 18.4 0.58 80,000

comparison to Lya observations. rtpeak HP Extent Consistency with [cm -3] [ A U ] [AU] 12~ 52 ~ 73 ~ 112 ~ 0.219 104 210 Y N N N 0.229 99 210 Y N N N 0.228 95 210 N Y Y N 0.228 92 250 N Y Y N 0.221 91 260 N Y N N 0.196 88 150 N Y Y Y 0.181 87 180 N Y Y Y

data 139 ~ N N N N N Y Y

148 ~ N N N N N N Y

In Figure 2, we plot one-dimensional profiles of the neutral H density along the stagnation axis for all seven models. The prominent feature in all models is the hydrogen wall, where the densities reach above the 0.14 cm -3 LISM value. The peak density in all the hydrogen walls is the same, rtpeak = 0.22 cm -3 (Table 2), with the exception of models 6 and 7 where it is ,-15% lower. Table 2 contains the locations where the heliopause (HP) intersects the upstream stagnation axis, showing the effect of the increasing interstellar ram pressure for increasing c~. The table also contains estimates for the extent of the hydrogen walls, defined here as the length of the region of the stagnation axis where the neutral density is distinctly larger than the LISM value. This wall extent, too, is constant (~200 AU) for models 1-5 and somewhat thinner for models 6 and 7. The typical effective temperatures of neutral H inside the wall listed in Table 2 don't seem to follow a particular pattern and range from 60,000 K to 100,000 K. Downwind of the hydrogen wall, neutral hydrogen is depleted but hot (~ 105 K, see Fig. l b). In the tail region, the subsonic models have generally lower densities than the two-shock models (see left panel of Figure 2), with the difference between model 1 and model 7 being 20% at 100 AU and

-17-

H.-R. Miiller and B.E. Wood 0.10

I, t~,

Figure 2. 1D density profiles of neutral H upstream (right) and downstream (left) of the Sun, for all models. Note the different scales of the left and right hand panels.

growing to 40% at 1000 AU. In general, higher heliospheric hydrogen temperatures and densities are observed for models with lower values of c~, which is expected since a lower c~ corresponds with a higher LISM Mach number and therefore more heating. Because higher temperatures and column densities will produce broader Lyc~ absorption profiles, models with higher Mach numbers will also generally produce more absorption. The particle Boltzmann code returns the complete phase space information of neutral H. In particular, we obtain the velocity distribution function throughout the heliosphere that we can use to compute the absorption profiles. In Figure 3, radial velocity distributions are displayed for neutral hydrogen particles in the upwind and downwind directions for Model 1. In order to create these distributions, particles are summed within a certain area centered on the locations indicated in the figure. The area measures 104 AU 2 for Figure 3 and 3000 AU 2 for the Lyc~ absorption calculations in Section 3. Poissonian error bars (N ~ are displayed for each velocity bin. The bin size of the histogram is 6 km s -1. Gaussians have been fitted to the distributions (dashed lines), and the poor fits illustrate the non-Maxwellian character of the distributions. Looking upstream, the distribution functions have a broad core at -17 km s -1 which corresponds to moderately heated, decelerated neutral H of LISM origin coming toward the Sun. In the downstream direction, neutral H is moving away from the Sun, with the peak in the distribution at 28 km s -1. In both directions, there is a peak at 400 km s -1 which is the signature of splash component neutrals born through charge exchange in the solar wind. Neutrals with velocities between 100 and 300 km s -1 are component 2 neutrals produced in both the hot heliosheath and the heliotail. 3. C O M P A R I S O N

TO HST DATA

In upwind directions, the heliospheric H I column density is dominated by compressed, heated, and decelerated material just outside the heliopause (the hydrogen wall). For an observer at Earth, the absorption due to LISM H I in the upwind direction is blueshifted relative to the Ly~ rest wavelength. The decelerated heliospheric H I absorption is also blueshifted, but less than that of the LISM. Therefore, the hydrogen wall material accounts

-18-

Relating models of the heliosphere to Lyman-a absorption...

1000.0

-

100.0

10.0

Q) 1.0

o

. ,...~

0.1 150 0 -150 1000.0

100.0

o

10.0

1.0

.........

0.1 100 0

-

)!l JI, ...... h........ 0

-

-100 -200

- 100

0

100

200

300

400

500

( k m s -1) Figure 3. Neutral H radial velocity distributions for two different heliospheric locations based on Model 1, where 0 is the angle relative to the upwind direction of the interstellar flow, and R is the distance from the Sun. The dashed lines are Gaussian fits to the distributions, and residuals of the fits are shown below each panel. From [24].

for most of the non-LISM absorption observed on the red side of the s~turated core of the Lya absorption profile. Component 1 neutral hydrogen develops only small perpendicular velocity components in the hydrogen wall, such that the above holds for all sightlines through the hydrogen wall less than 90 ~ from the upwind direction. At hydrogen walls around other stars, where an observer looks from the outside rather than the inside, this scenario is reversed, meaning that the decelerated material of the stellar hydrogen wall is in fact more blueshifted than the LISM, and the additional absorption shows up at the blue wing of the main (interstellar) absorption feature. Figure 4 shows the observed Lya profile of a Cen B [10], which is 52 ~ from the upwind direction. Excess H I absorption is present on both the blue and red sides of the LISM absorption. The red side excess is due to the heliospheric absorption, while the blue side excess is due to absorption from analogous "astrospheric" material [11]. An additional detection of heliospheric H I absorption only 12~ from the upwind direction was provided by HST observations of 36 Oph [15]. For downwind lines of sight the H I density is much lower than in the hydrogen wall, but the sightline through the heated heliospheric H I is longer, potentially allowing heliospheric Lya absorption to be observed downwind as well [23], again on the red side because neutrals in the tail are accelerated by charge exchange to speeds higher than the LISM speed. A detection of heliospheric absorption along a downwind line of sight toward Sirius has been reported [16].

-19-

H.-R. Miiller and B.E. Wood

2.5 T 2.0

7 O

1.5

~0

o

1.0

" ~ 0.5 N 0.0 1215.0

1215.2

1215.4 1215.6 1215.8 Wavelength

1216.0

1216.2

Figure 4. HST/GHRS Lye spectrum of c~ Cen B, showing broad H I absorption at 1215.6 ~ and D I absorption at 1215.25 ~. The upper solid line is the assumed stellar emission profile and the dashed line is the ISM absorption alone. The excess absorption is due to heliospheric H I (vertical lines) and astrospheric H I (horizontal lines).

We can use the models listed in Table 1 to predict the amount of Lyc~ absorption we expect to see for various lines of sight through the heliosphere [24]. We compare these predictions with the observations of c~ Cen, 36 Oph, and Sirius, which we have already mentioned above as having excess Lyc~ absorption that is presumably heliospheric. In addition to those lines of sight, we also consider three additional lines of sight toward 31 Com, /3 Cas, and c Eri. The HST Lyc~ spectra of these stars show no evidence for excess absorption on the red side of the line that could be heliospheric [25], but these data still provide useful upper limits for the amount of absorption that could be present and they sample different directions through the heliosphere. In Figure 5, the absorption predicted by the seven models in Table 1 is compared with the Lyc~ absorption profiles observed toward the six stars. The 0 values shown in the figure are the angles from the upwind direction, which range from the nearly upwind line of sight toward 36 Oph (0 = 12~ to the nearly downwind line of sight toward c Eri (0 = 148 ~ The predicted Lyc~ absorption is shown after combination with the LISM absorption toward these stars, as determined from previous empirical analyses [10,15,16,25]. An attempt has been made to maximize the agreement between the data and the models by tweaking the assumed stellar emission profiles, as described by [24]. Significant disagreement remains in most cases despite these efforts. Based on Figure 5, we provide in the last six columns of Table 2 our evaluation of which observed stellar lines of sight (labeled by their 0 angles) are inconsistent with which models. None of the models are consistent with every line of sight. The low c~ models do better upwind and the high c~ models do better downwind. In general, the models predict too much absorption, the exception being the 36 Oph line of sight for which most of the models predict too little absorption. Model 7 is the only model that does not predict too much absorption along the downwind line of sight to e Eri.

-20-

Relating models of the heliosphere to Lyman-a absorption...

Cen

1.5 3'6' b ; 'h ' ('0'='1'2'~ . . . . . .

,J o5o 1 0.40

;'1' c;m

i;?;

~ ......

lO 1.o

0.30

8

6

L

0.5

0.20

4 0.10

?

2 ~O

0.00

o

0.0 0

20

40

60

80

o

100

20

40

60

80

20

100

40

60

80

100

~0 Q) e Eri (0=148 ~

'Si; ( 'u; ' 0 ' =' ' l ' 3 ; ~

7

a=2.0 .......... a = 3 . 5

0 v-t v

i/

v.6

N ...... y(/'

--

--

a=9.6 -

/~'1

a = l

0.4

YJ~,, 20

40

60

80

100

0

20

40

60

80

O0

0

20

40

60

80

100

Velocity (km s -1)

Figure 5. The red side of the Lya absorption profiles observed toward six stars, sampling different angles 0 relative to the upwind direction of the interstellar flow into the heliosphere. These data are compared with the heliospheric absorption predicted by the seven models listed in Table 1, which assume different values for a. From [24].

4. C O N C L U S I O N S We have presented models of the heliosphere constructed using a hybrid code combining a fluid treatment for the plasma and a kinetic treatment for the neutral particles. These models are potentially very useful for analyzing the heliospheric H I Lya absorption that has been detected toward a number of stars using HST observations. The models provide detailed information on the spatially-dependent, highly non-Maxwellian neutral H velocity distributions within the heliosphere. In principle, these distributions should allow accurate Lya absorption profiles to be computed for comparison with the HST data. However, we were unable to find a model which successfully reproduced the observed absorption. There are many possible reasons for this lack of success. One is that we simply have not experimented with a large enough range of input parameters. It is only the a parameter that is varied in the set of models in Table 1. It is possible that assuming somewhat different LISM or solar wind parameters, while still forcing them to be within available observational constraints, would improve agreement with the data. Another possibility is that we have neglected one or more physical processes in the models which could be

-21 -

H.-R. Miiller and B.E. Wood

important and affect our results, such as processes involving magnetic fields. One particular physical process that might be a factor is non-charge exchange interactions between the neutrals and the plasma, which are not considered in the current models. Perhaps there are enough neutral-proton collisions to remove a significant number of particles from the far wings of the velocity distributions (see Figure 3), thereby reducing the amount of predicted Lya absorption. This could be especially important for downwind directions where neutrals have traveled longer pathlengths through the heliosphere and have therefore had more opportunity for collisions with protons. And it is particularly in downwind directions where the current models predict way too much absorption. In future analyses, we hope to explore this possible explanation, and others. REFERENCES o

2. 3. 4. 5. 6. 7. o

9. 10. 11. 12. 13. 14. 15. 16. 17.

20. 21. 22. 23. 24. 25.

T. E. Holzer, Ann. Rev. Astr. & Astrophys., 27 (1989) 199. R. S. Steinolfson, V. J. Pizzo, and T. E. Holzer, Geophys. Res. Left., 21 (1994) 245. C. Wang and J. W. Belcher, J. Geophys. Res., 103 (1998) 247. G. P. Zank, Space Sci. Rev., 89 (1999) 413. V. B. Baranov and Y. G. Malama, J. Geophys. Res., 98 (1993) 15157. V. B. Baranov and Y. G. Malama, J. Geophys. Res., 100 (1995) 14755. P. C. Liewer, S. R. Karmesin, and J. U. Brackbill, in Solar Wind 8, ed. D. Winterhalter et al. (New York: AIP, 1996) 613. H. L. Pauls, G. P. Zank, and L. L. Williams, J. Geophys. Res., 100 (1995) 21595. G.P. Zank, H.L. Pauls, L.L. Williams, & D. Hall, J. Geophys. Res., 101 (1996) 21639. J. L. Linsky and B. E. Wood, Astrophys. J., 463 (1996) 254. K. G. Gayley, G. P. Zank, H. L. Pauls, P. C. Frisch, and D. E. Welty, Astrophys. J., 487 (1997)259. A. S. Lipatov, G. P. Zank, and H. L. Pauls, J. Geophys. Res., 103 (1998) 20631. H.-R. Mfiller and G. P. Zank, in Solar Wind 9, ed. S. R. Habbal et al. (New York: AIP, 1999) 819. H.-R. Mfiller, G. P. Zank, and A.S. Lipatov, J. Geophys. Res., 105 (2000) 27419. B. E. Wood, J. L. Linsky, and G. P. Zank, Astrophys. J., 537 (2000) 304. V. V. Izmodenov, R. Lallement, and Y. Malama, Astron. Astrophys., 342 (1999) L13. W. L. Fite, A.C.H. Smith, and R.F. Stebbings, Proc. R. Soc. London Ser. A, 268 (1962) 527. J. L. Linsky, S. Redfield, B. E. Wood, & N. Piskunov, Astrophys. J., 528 (2000) 756. R. Lallement, R. Ferlet, A. M. Lagrange, M. Lemoine, and A. Vidal-Madjar, Astron. Astrophys., 304 (1995) 461. E. Qu~merais, J.-L. Bertaux, B. R. Sandel, and R. Lallement, Astron. Astrophys., 290 (1994) 941. G. Gloeckler, et al., Science, 261 (1993) 70. B. E. Wood and J. L. Linsky, Astrophys. J., 474 (1997) L39. L.L. Williams, D. T. Hall, H. L. Pauls, and G. P. Zank, Astrophys. J., 476 (1997) 366. B. E. Wood, H.-R. Mfiller, and G. P. Zank, Astrophys. J., 542 (2000) 493. A. R. Dring, J. L. Linsky, J. Murthy, R. C. Henry, W. Moos, A. Vidal-Madjar, J. Audouze, and W. Landsman, Astrophys. J., 488 (1997) 760.

-22-

Interstellar atoms in the heliospheric interface V. V. Izmodenov a, aDepartment of Aeromechanics and Gas Dynamics, Faculty of Mechanics and Mathematics, Lomonosov Moscow State University, Vorob'evy Gory, Glavnoe zdanie MGU, Moscow 119899, Russia ([email protected]) In order to enter the heliosphere the interstellar atoms pass through the heliospheric interface - the region of the solar wind and interstellar plasma interaction. In the interface, the interstellar hydrogen atoms strongly interact with the LISM plasma component by charge exchange. This interaction results in the modification of both plasma and interstellar atom flows. Thus, the atoms penetrate into the heliosphere disturbed. This opens a possibility to use interstellar atoms and their derivatives - pickup ions and anomalous cosmic rays - for remote diagnostics of the heliospheric interface plasma structure. However, the interpretations of remote experiments are critical to accurate theoretical models. In this paper advanced self-consistent models of the heliospheric interface are reviewed. Evolution of the atom velocity distribution in the heliospheric interface is discussed with the emphasis on interpretation of present and future space experiments. 1. I n t r o d u c t i o n At the present time there is no doubt that local interstellar medium (LISM) is partly ionized plasma. Interstellar plasma component interacts with the solar wind (SW) plasma and forms the heliospheric interface (Figure 1). The heliospheric interface is a complex structure, where the solar wind and interstellar plasma, interplanetary and interstellar magnetic fields, interstellar atoms, galactic and anomalous cosmic rays (GCRs and ACRs) and pickup ions play prominent roles. In this paper I will review physical processes connected with interstellar atoms. Interstellar atoms of hydrogen are the most abundant component in the circumsolar local interstellar medium. These atoms penetrate deep into the heliosphere and interact with interstellar and solar wind plasma protons by charge exchange. This interaction influences the structure of the heliospheric plasma interface significantly. Being disturbed in the interface, interstellar H atoms and their derivatives such as pickups and ACRs can serve as remote diagnostics of the heliospheric interface. To make constructive conclusions from space experiments one must use an adequate theoretical model. The main difficulty in the modelling of the heliospheric interface is a kinetic character of the interstellar H *The research described in this publication was made possible in part by Award No.RP1-2248 of the U.S. Civilian Research and Development Foundation for the Independent States of the Former Soviet Union (CRDF), INTAS projects # 97-0512 and #YSF 00-163, RFBR grants # 01-02-17551, 99-02-04025, and International Space Science Institute in Bern.

-23 -

K Izmodenov

Figure 1. Structure of the heliospheric interface. The termination shock (TS), the heliopause (HP), the bow shock (BS) separates the heliospheric interface in four regions. Region 1 is the supersonic solar wind. Region 2 is the heliosheath with compressed solar wind plasma. Region 3 is the region of disturbed interstellar medium. The region is extended upstream BS (see, section 3). Region 4 is undisturbed interstellar medium.

atom flow in the heliospheric interface. Since the effects of elastic H-H, H-p collisions are negligibly small as compared with charge exchange, the latter process determines the character of the H atom flow in the interface (see, e.g., [10] and section 3 of this paper). Atoms newly created by charge exchange have local plasma properties. Since plasma properties are different in four regions shown in figure 1, there are four populations of interstellar atoms in the heliospheric interface. Population 1 is the atoms created in the supersonic solar wind. Population 2 is the atoms created in the heliosheath. Population 3 is the atoms created in the region of disturbed interstellar plasma (region 3 in Figure 1). Population 4 is the original interstellar atoms. The strength of H atom-proton coupling can be estimated through the calculation of mean free paths of H atoms in plasma. Generally, the mean free path of s-particle in t-gas can be calculated by the formulae: L - 5 M s~t / S t " Here w~ is the individual velocity of s-particle, S M u t ~ S t is individual Sparticle momentum transfer rate in t-gas, which is a function of local plasma parameters. Table 1 shows the mean free paths of H atoms with respect of charge exchange with protons. The mean free paths are calculated for typical atoms of different populations at different regions of the interface in the upwind direction. For every population of H atoms there is at least one region in the interface where Knudsen number K n N~ 0 . 5 - 1"0. Therefore kinetic Boltzmann approach must be used to MeanFreePath CharacteristicLength

-24-

Interstellar atoms in the heliospheric interface describe interstellar atoms in the heliospheric interface.

Table 1 Meanfree paths of H-atoms in the heliospheric interface with respect to charge exchange with protons, in AU Population At TS At HP Between HP and BS LISM 110 870 4 (primary interstellar) 150 100 58 190 3 (secondary interstellar) 66 40 110 200 2 (atoms originated in the heliosheath 830 200 240 490 1 (neutralized solar wind) 16000 5!0 , ,

.....

....

2. Heliospheric Interface M o d e l Baranov et al. [1] and Baranov and Malama [2] considered a two-dimensional (2D) axisymmetric model of the SW/LISM interaction. The interstellar and solar wind plasma components were treated as fluids. To describe the flow of interstellar atoms the Boltzmann equation was solved: HPf.(r. wp)dwp 0/H(r , WH) ~ F O/H (r, WH) = --fH(r. WH) If IWH -- Wpla.z WH" 0r mH 0WH -+-fp(r, W H ) [fW ~ i

WH [ HP fH(r, WH)dWH 9 9

(~ §

(1)

r, WH).

Here fH(r, WH) is the distribution function of H atoms; fp(r, wp) is the local distribution function of protons, assumed to be Maxwellian; wp and wH are the individual proton and H atom velocities, respectively; a HP is the charge exchange cross section of a H atom with a proton; fli is the photoionization rate; mH is the atomic mass; ~imp~t is the electron impact ionization rate; and F is the sum of the solar gravitational force and the solar radiation pressure force. The plasma and neutral components interact mainly by charge exchange. However, photoionization, electron impact ionization, solar gravitation, and radiation pressure are also taken into account in equation (1). The interaction of the plasma and neutral components leads to the mutual exchange of mass, momentum and energy. The effect of this interaction can be represented by adding source terms Q -- {ql, q2,z, q2,r, q3, O)T to the right-hand side of the plasma equations. The terms ql, q2 = {q2,~,q2,~}, q3 describe the mass, momentum, and energy sources in the thermal plasma component due to interaction with neutrals. These source terms can be expressed through the integrals of the atom distribution function fH:

q~ - n i l " (~i + ~imp~t), nH = f fH(WH)dWH,

wp) f~ (WH)fp (wp)dw~dwp,

-25-

K Izmodenov 1

Here f l r f uaHP(u)fv(wp)dw v is the charge exchange rate, and u is the relative atomproton velocity. Supersonic boundary conditions are used for the unperturbed interstellar plasma and for the solar wind plasma at the Earth's orbit. The velocity distribution of interstellar atoms is assumed to be Maxwellian in unperturbed interstellar medium. 3. On t h e effect of H-H, H-p elastic collisions Recently Williams et al. [3] have suggested that a population of hot hydrogen atoms is created in the heliosphere through elastic H-H collisions between atoms of population 1 and interstellar atoms of populations 3 and 4. Izmodenov et al. [5] examined the approach used by Williams and argued that two assumptions used by Williams et al. result in significant overestimation of the H-H collision effect. 1. Williams et al. [3] applied the momentum transfer cross-section calculated by Dalgarno [4] for quantum mechanically indistinguishable particles in H-H collisions. Treating the colliding particles as quantum mechanically indistinguishable or classically distinguishable could lead to significant differences in the meaning and values of the cross-sections. Dalgarno's approximation is appropriate for a quantum particle ensemble when velocities of distinct particles are correlated with each other. However Williams et al. applied this like-particles H-H collision momentum transfer cross-section to the description of the collisions between different populations of H atoms that are classically distinguishable. Classical statistics should be used under such conditions. Izmodenov et al. [5] calculated the momentum transfer cross-section for H-H collisions in the 0.01-1000 eV energy range treating colliding particles as classically distinguishable. Figure 2A compares the calculated momentum transfer, Dalgarno, and charge exchange cross sections as functions of relative velocity of the colliding particles. It is seen that the cross-section calculated in [5] is significantly smaller than the Dalgarno cross-section when V > 10 km/s. 2. Williams et al. [3] used a Bhatnagar-Gross-Krook (BGK)-like approximation for the Boltzman's equation collision term. This approximation is based upon on an assumption of a complete randomization of particle velocities in the collisions, without making a distinction between 'strong' and 'weak' collisions. If one calculates momentum transfer term for different H-atom populations using the BGK-like approximation, one would obtain the values that 1-2 order of magnitude larger than for our approximation. For calculations with the Dalgarno's cross section, the overestimation introduced by the BGKlike approximation is much smaller, but still significant. This effect can be explained by different functional dependencies of the cross sections and by the BGK-like collision term. Izmodenov et al. [5] concluded that the influence of elastic H-H, H-p collisions is negligible in the heliospheric interface since the dynamic influence of charge exchange is stronger by several orders of magnitude. 4. Basic results of t h e model: H a t o m s Since the elastic H-H, H-p collisions are not important dynamically, Baranov-Malama model [2] describes the main physical processes correctly. In this section main results of

-26-

Interstellar atoms in the heliospheric interface

~ 10- 6

f

l.qarno [1960i -

~

: E

Figure 2. A. Momentum transfer cross section (in cm 2) as functions of relative velocity (in cm sec-~; B. Comparison of momentum transfer of population 2 due to elastic collisions with other H atom populations (in (cm sec) -2) in the upwind direction.

the model are briefly reviewed. Detailed results of the model reported in Baranov and Malama ([2], [6], [7]), Baranov et al. [8], Izmodenov et al. [9], Izmodenov [10]. One of the main and the most spectacular results of the model is prediction of the hydrogen wall, a density increase at the heliopause (figure 3A). The hydrogen wall is made up by secondary interstellar atoms of population 3. The hydrogen wall was predicted by Baranov et al. [1]. Linsky and Wood [11] were first to demonstrate that observed by HST Ly-a spectra toward Sirius can not be understood without taking into account the absorption produced by heliospheric hydrogen wall. Figure 3C shows the heliospheric absorption toward a-Cen calculated on the basis of Baranov-Malama model. There is still a missing absorption on the blue side. This is probably absorption produced by heated neutral gas around the target star (" astrosphere"), as suggested first by Wood and Linsky. Interpretation of stellar spectra can provide us constraints on both the heliospheric interface and astrospheres. The first attempt to constrain the interface with to help of stellar spectra was done by Gayley et al. [12]. Later, Izmodenov et al. [13] interpreted Ly-c~ spectra toward star Sirius-A (Figure 3D) on the basis of the heliospheric interface model. It has been shown that in this direction (41 ~ from downwind) two populations (populations 3 and 2) of neutral H atoms produce a non-negligible absorption. Figure 3B shows number density of population 2 toward Sirius direction. Despite small number density this population produces non-negligible absorptions due to high temperature. During this COSPAR Colloqium Wood, Muller and Zank reported the study of Ly-a spectra toward six nearby stars observed by HST. Comparing their model calculations with data the authors concluded that their model predict too much absorption in sidewind and downwind directions. More theoretical work has to be done to understand these discrepancies between theory and observations.

-27-

K Izmodenov

Figure 3. A. The hydrogen wall is an increase of number density of population 3 at the heliopause. The distribution is shown for direction toward c~Cen. B.The distribution of population 2 toward Sirius. C. HST spectrum of a Cen-A near Ly-a line center. Also shown is simulated stellar profile prior to interstellar absorption, the same profile after interstellar absorption and the expected profile after the cloud plus the calculated heliospheric H absorption. A missing absorption on the blue side is probably "astrospheric". D. HST spectrum of Sirius-A near Ly-a line center. Simulated profiles are similar of profiles on plot C.

5. Basic results of t h e model: influence on p l a s m a flows Interstellar atoms strongly influence the heliospheric interface structure. In the presence of interstellar neutrals, the heliospheric interface is much closer to the Sun than it would be in a pure gas dynamical case (see, e.g., figure 2 in Izmodenov, 2000 [10]). The termination shock becomes more spherical. The Mach disk and the complicated shock structure in the tail disappear. Interstellar neutral atoms also affect the flow of supersonic interstellar plasma upstream of the bow shock, the flow of the supersonic solar wind upstream of the termination shock, and plasma structure in the heliosheath. The flow of the supersonic solar wind upstream of the termination shock is disturbed by charge exchange with the interstellar neutrals. The new created by charge exchange ions are picked up by the solar wind magnetic field. The Baranov-Malama model assumes immediate assimilation of pickup ions into the solar wind plasma. The effect of charge exchange on the solar wind is significant. By the time the solar wind flow reaches the termination shock it is decelerated (15-30 %), strongly heated (in 5-8 times) and mass

-28-

Interstellar atoms in the heliospheric interface

loaded (20-50 %) by the pickup ion component. The interstellar plasma flow is disturbed upstream of the bow shock by charge exchange with the secondary atoms originated in the solar wind and compressed interstellar plasma. Charge exchange results in the heating (40-70 %) and deceleration (15-30 %) of the interstellar plasma before it reaches the bow shock. The Mach number decreases and for a certain set of interstellar parameters (nH,LISM > > ?~p,LISM)the bow shock may disappear. The plasma structure in the heliosheath is also modified by the interstellar neutrals. In a pure gasdynamic case (without neutrals) the density and temperature of the postshock plasma is nearly constant. However, the charge exchange process leads to a strong increase of the plasma number density and decrease of its temperature. Baranov and Malama (1996) [7] pointed out that the electron impact ionization process may influence the heliosheath plasma flow by increasing the gradient of the plasma density from the termination shock to the heliopause. The effects of interstellar atom influence on the heliosheath plasma flow may be important, in particular, for the interpretations of (1) kHz radio emission detected by Voyager (see, [15], [16]) and (2) possible future heliospheric imaging in energetic neutral atom (ENA) fluxes [17].

6. Interpretations of spacecraft experiments on the basis of the BaranovMalama model The Sun/LISM relative velocity and the LISM temperature are now well constrained. Using the new SWICS pickup ion results and an interstellar HI/HeI ratio of 13 i 1 (the average value of the ratio toward the nearby white dwarfs), Gloeckler et al. (1997) [18] concluded that n L r s M ( H I ) -- 0.2 • 0.03 am -3. However there are no direct ways to measure the circumsolar interstellar electron (or proton) density. Therefore, there is a need for indirect observations (inside the heliosphere) which can bring stringent constraints on the interstellar plasma density and on the shape and size of the interface. Such constraints can been done on the basis of theoretical models of the interface. Izmodenov et al. [9] used the Baranov-Malama model to study the sensitivity of the various types of indirect diagnostics of local interstellar plasma density. The diagnostics are the degree of filtration, the temperature and the velocity of the interstellar H atoms in the outer heliosphere (at the termination shock), the distances to the termination shock, the heliopause, and the bow shock, and the plasma frequencies in the LISM, at the bow shock and in the maximum compression region around the heliopause which constitutes the "barrier" for radio waves formed in the interstellar medium. Izmodenov at al. [9] searched also for a number density of interstellar protons compatible with SWICS/Ulysses pickup ion and Voyager, HST, SOHO observations of backscattered solar Ly c~ and kHz radiations observed on Voyager. Table 2 presents estimates of np,niSM obtained from different types of observations using the Baranov-Malama model. Izmodenov et al. [9] concluded that it is difficult to reconcile these estimates without some modification to the model. Two mutually exclusive solutions have been suggested: (1) It is possible to reconcile the pick-up ions and Ly a measurements with the radio emission time delays if a small additional interstellar (magnetic or low energy cosmic ray) pressure is added to the main plasma pressure. In this case, n p , L I S M - 0.07 cm -3 and n H , L I S M - - 0.23 cm -3 is the favored pair of interstellar densities. However, in

-29-

V. Izmodenov

this case, the low frequency cutoff at 1.8 kHz doesn't correspond to the interstellar plasma density, and one has to search for another explanation. (2) The low frequency cutoff at 1.8 kHz constrains to the interstellar plasma density, i.e., rtp,LISM - - 0 . 0 4 a m - 3 . In this case, the bulk velocity deduced from the Ly a spectral measurement is underestimated by about 30-50% (the deceleration is by 3 km s -1 instead of 5-6 km s-~). Model limitations (e.g. stationary hot model to derive the bulk velocity) or the influence of a strong solar Ly a radiation pressure may play a role. In this case, there would be a need for a significant additional interstellar (magnetic or cosmic ray) pressure as compared with case (1).

Table 2 Intervals of Possible Interstellar Proton Number Densities Type of Heliospheric Interface Diagnostics Range of Interstellar Proton Number Density SWICS/Ulysses pick-up ion 0.02 c m - 3 < rtp,LISM < 0.1 c m - 3 0.09 cm -3 < nH,TS < 0 . 1 4 cm -3 Gloeckler et al., [1997] Ly- a, intensity 0.11 cm -3 < nH,TS < 0.17 cm -3 Qugmerais et al., [1994]

np,LISM '( 0.04 cm -3 or

Ly-a, Doppler shift 18 km s -1 < VH,TS < 21 km s -1 Bertaux et al. [1985], Lallement et al. ,

0 . 0 7 c m - 3 < np,LISM
0. The minimum eigenvalue in this case is ,~ = u - af, where af is the largest of the two magnetosonic speeds af and a~ (af > Ca _> as): af,~- ~-

~

+ x/~

+

c~q 4 ~ r p

-37-

x/~

'

(4)

N. K Pogorelov

IBI 2 - n2x + By2 "-t-"n z2 . Let us fill the artificial cell outside the boundary by the parameters at infinity. If the outflow velocity in the center of the cell adjacent to the boundary is superfast magnetosonic, u0 > af0, we need no boundary conditions. If u0 < af0 but u~ > a f t , we introduce a rarefaction wave accelerating the flow to fast magnetosonic speed at the boundary. If u0 < af0 and uo~ < afo~, we can simply solve the Riemann problem between U0 and U~. The eigenvector corresponding to the chosen eigenvalue is (7 is a specific heat ratio)

r-

[1

1)Ca 0, t~By, ~ B z

af aBy, ceBz, ( 7 '

p'

2p

Caas

vf~

c~ - 2 P v ' T - ~ ( d a - a~) -

'

lT ,

a~(a~ - C2a)sgnBx --ca(By2 ~ nz )-BSS ,

(5)

c 2a 47r(a 2 -- c 2) fl - p(C2a - a~) = By2 + B2z "

If we seek the solution to (2) in the form of a simple wave U(x, t) = U(() with ~ = x / t , we obtain ( A - AI)Ur = 0,

A=(,

where I is the identity matrix. Thus, Ur is proportional to the right eigenvector of A corresponding to A = ~. If we denote the proportionality factor as d, we can rewrite system (2) in the form

afd ur = - - ~ , v( -- ceByd, - c~Bzd, p (Bx)r - O , (By)r - /3Byd, (Bz)r - /3Bzd, u-af- d,

(Ca) r =

('7 - 1)cad , 2p

(6)

~.

In system (6) we now substitute the equation for Ca by the equation for af, which can be easily obtained from Eq. (4) by direct differentiation with respect to ~ and by using Eqs. (6), yielding

tgd (af)( = 7 '

tg-

Oaf

r.

One can easily notice that the eigenvector (5) degenerates for C a - - a~. This happens if 0 and Ca < aA. It is apparent from Eqs. (6) that the derivatives ( B y ) ( and (Bz)r become infinite at degeneration points. One can easily obtain, however, that

By2 + B z2 -

-- ca)by,zpr

,

where by,z - By,z/(B2y + B~) 1/2. The right-hand side of this relation never degenerates if we assume by - bz - 1/x/2 for By2 + B z2 - 0 (see, e. g., [22]). If we take into account the relation

18xl a~af-

Ca2v/_~ ,

the equations for the derivatives of the transverse components of the velocity vector can also be transformed to the form suitable for further application,

vr

Bx ~47rpa~(BY)~'

we-

Bx ~(Bz)r

-38-

M H D modeling of the outer heliosphere: Numerical aspects

Thus, by passing from d to PC, we obtain the reduced system of equations that can be approximated, for example, by finite differencing 0er = (fr -f0)/A~). The system becomes:

(af)r

-

(

u0 + a~

O+af

(Ca)F--(Ca)0+(q/--1)

)

,

0

Pr

-

af)

P0 1 -~ 0 + af 0

vr - vo +

()

Bx 47rpaf

o (Byr - Byo),

(8)

i xl -i xlo,

\~p/

0 (B2)r _ B~~ + 87r[(a 2 _ Ca)by]O(DF2 2 Do),

,

(B2)F --

Wr - Wo +

2 + Srr[(a2 nzo 47rpaf

-

2 ca)b2]o(fiF -- rio),

o

3. NONEVOLUTIONARY MHD SHOCKS 3.1. Concept of evolutionarity The important subject of discussion is related to the fact that certain initial- and boundary-value problems can be solved nonuniquely using different shocks or different combinations of shocks, whereas physically one would expect only unique solutions. The situation in MHD differs from that in pure gas dynamics of perfect gas, where all entropy-increasing solutions are evolutionary and physically admissible. The term "evolutionary" means that the necessary conditions of well-posedness for the linearized problem of the shock interaction with small disturbances are satisfied. In MHD, the condition of entropy increase is necessary, but not sufficient for the shock to be admissible. Only slow and fast MHD shocks were found to be evolutionary, while intermediate (or improper slow) shocks were not and therefore must be excluded in ideal MHD. All MHD shocks are plane-polarized, that is, the magnetic field and the shock normal vectors lie in the same plane both ahead and behind the shock. In contrast to evolutionary shocks, the tangential component of the magnetic field behind a nonevolutionary shock acquires the direction opposite to what it had ahead of the shock. Thus, rotations of the magnetic field vector are possible only on two types of discontinuities" Alfv6n ones and nonevolutionary shocks. If the statement of the problem rules out Alfv6n discontinuities, nonevolutionary shocks can fairly easily occur in the numerical solution [5]. There is a simple rule to determine whether the shock is evolutionary. Fast MHD shocks are always super-Alfv6nic, that is, the magnitude of the velocity component normal to the shock is larger than the Alfv6n velocity a A - - [B~l/2x/-~ both ahead of the shock and behind it. Slow shocks are sub-Alfv6nic. Nonevolutionary shocks are trans-Alfv6nic. 3.2. Nonevolutionary shocks in the SW-LISM interaction If the magnetic field vector is parallel to the shock normal, we have a parallel shock. The behaviour of parallel shocks depends on whether the Alfv6n velocity ahead of them is smaller or larger than the acoustic speed of sound. In the former case, slow MHD shocks do not exist, while fast shocks are evolutionary and admissible for all their intensities. In the latter case, on the contrary, slow shocks are admissible for all their intensities, while fast shocks are admissible only in a certain range of parameters ahead of the shock, even if they correspond to a superAlfv6nic flow. Singular shocks are those for which the tangential component of magnetic field is equal to zero ahead of (behind) the shock and not zero behind (ahead of) it. Such shocks are called switch-on (switch-off), as the tangential component of the magnetic field vector is

-39-

N. V. Pogorelov

switched on (off) at them. Switch-on shocks are always fast while switch-off shocks are always slow, since the tangential component of the magnetic field always increases across fast and decreases across slow MHD shocks. Let us consider what happens if we increase the value of the LISM magnetic field Boo with the rest of the LISM quantities being fixed. We have two dimensionless parameters relating the quantities ahead of the shock, namely, the Mach number M ~ = V~/Caoo and the Alfv6n number A ~ = Voo/aaoo. If C~o~ > aao~ for Mo~ > 1, the forward point of the bow shock corresponds to a fast parallel shock which is always realizable. If we further increase Bo~, sooner or later a a ~ will become larger than Caoo with A~ > 1. In this case the parallel shock, though remaining fast, will be still evolutionary until Boo acquires the value corresponding to the interval

( 7 + 1)M 2

1 1, guaranteeing that the flow is subalfv~nic). 3. D I S C U S S I O N

AND CONCLUSIONS

In the following figures there are shown some representative MHD-values. In MHS, isocontours of the thermal pressure and of the current density are field lines. In Figs. 1 and 2 we can see that this is violated in the case of MA ~ O. If we calculate symmetric potential fields it is also possible to choose transformations to get equilibria with asymmetric field strengths, although the trajectories are symmetric. It is also possible to model a global heliosplaere with asymmetric trajectories, so the stagnation point can lie off the x-axis. The coefficients in equation (6) could be chosen in this way that different current directions are possible for the different current sheets. Therefore this technique permits a great flexibility to construct MHD-flows for stellar magnetospheres far away from the central star, moving through the VLISM.

-59-

D. Nickeler and H.J. Fahr

Figure 1. Symmetrical field/stream lines. The length scale is 100 AU, so the standoff distance of the heliopause is -1.5 (= 150AU).

Figure 2. Isocontours of the current density for a certain transformation, which corresponds to the symmetric field lines of the mapped MHS-equilibrium. The tail field lines are open, while the isocontours are closed. The sun is located in (x, y) = (0, 0).

A c k n o w l e d g e m e n t : We are grateful for financial support granted by the Deutsche Forschungsgemeinschaft within the project Fa 97/23-2. REFERENCES

1. H. Grad, H. Rubin, Proceedings of the Second United Nations International Conference on the Peaceful Uses of Atomic Energy, Geneva 1958, Vol. 31, p.190 ft. 2. K. Scherer, H. Fahr, R. Ratkiewicz, A&A, 287 (1994), p. 219 3. J. Birn, JGR Vol. 92 (1987), p. 11101 4. S.T. Suess, S. Nerney, JGR, Vol. 100 (1995), p. 3463 5. E. Hameiri, Physics of Plasmas, Vol. 5 (1998), p. 3270 6. W. Zwingmann, PhD-thesis, Ruhr-Universit/it-Bochum (1984), Germany, p. 53 ft. 7. U. Gebhardt, M. Kiessling, Physics of Fluids, B 4(7) (1992), p. 1689 8. G. Petrie, T. Neukirch, Geophys. Astrophys. Fluid Dynamics, Vol. 91 (1999), p. 269 9. S. J/iger, H. Fahr, Solar Physics, Vol. 178 (1998), p. 631

-60-

Non-stationary magnetic field geometry in the heliosphere I.S.Veselovsky a ~Institute of Nuclear Physics, Moscow State University, Moscow 119899, Russia Model examples of the 3D time-dependent magnetic fields with the open, semi-open and closed configurations in the heliosphere are presented. 1.

INTRODUCTION

The magnetic field in the heliosphere is dependent on space and time because of the imposed boundary conditions and internal dynamical processes (see, e.g., reviews by Mariani and Neubauer, 1990; Burlaga, 1991). Quasistationary and transient structures coexist in proportions varying with the solar cycle. The purpose of this paper is to present model examples showing magnetic field lines with different geometric properties in the heliosphere. 2.

MODEL ASSUMPTIONS

Let us consider magnetic fields under the ideal conductivity approximation neglecting displacement currents. In this case, the governing Maxwell equations read as follows

-

o.

(2)

If the velocity field g(~', t) is given (kinematic approximation), we have four scalar equations (1,2) for three magnetic field vector components. This means that the system (1,2) is overdetermined as a rule, and nontrivial solutions for B are possible only under specific additional restrictions. Boundary conditions for the magnetic fields can be not prescribed arbitrary, but they are adjusted to the given velocity field. The remaining Maxwell equations can be used for explicit calculations of the electric charges and currents via (V. E ) a n d (V x B ) . Hence, the set (1,2) with some boundary conditions totally determines the magnetic field structure in space and time when g(~', t) is given. This approach is well known in applications to the heliosphere when using the method of characteristics. -+

-+

-61

-

LS. Veselowsky

3.

EXAMPLES

1) In the case when g(~', t) - const and the velocity has only the radial component v, the solution of equation (1) can be written as follows -

--

Bo~

T

Bo -

--

Boo

r

(?)

B~ -

Boo

( (

0,7),t

,

(3)

V

O, ~, t

,

(4)

.

(5)

v

O, ~ , t v

The magnetic field vector at the initial sphere is not an arbitrary function of its arguments, but obeys the additional condition J~0r

0 (sin OBoo) + r0 sin 0 gO

--

OBoe]

o~

(6)

- O.

This restriction is very important. In particular, for the stationary axially symmetric boundary conditions this means Boo - 0 and Bo - O. The field line equation dr rdT)

Br

:

(7)

B~

can be integrated in a straightforward manner with the result showing a family of integral curves starting at the sphere with radius r0. Each field line belongs to the cone defined by 0 = 00 and goes from r0 to infinity along the conical spiral when Bo~ -r 0 or along the radial direction when B0~ =0. The overall geometry of the field lines could be termed as semi-open. 2) For the boundary ..+ conditions which are stationary in the reference frame rotating with the Sun, one has B0 (0, 7)- f~t) at the initial sphere r0. Here ft is the angular velocity of the rotation. In the case when Boo = 0 one obtains from (6) the relation v

f~Bo~ +

r sin 0

Bo~ - 0

(8)

and the solution (3-5) looks as follows

Br

-

(~0) -- ~ Bo~(O,~-f~t'),

(9)

7"

(10)

Bo - 0 , B~-(t--~

B o ~ ( 0 , 7 ) - F i t ' ) sin--------~0v f~r~ '

(11)

where t' - t - ~-r0. Field lines are represented in this case by Archimedian spirals v r sin 0 + a7) - const on the cone, 0 - const, where a - v/f~(O) (Parker, 1958). The spirals

-62

-

Non-stationary magneticfieM geometry in the heliosphere are tighter near the equatorial plane because of the differential rotation f~(0). The real heliospheric magnetic fields follow the model (9-11) rather well in a broad range of heliospheric distances only under the appropriate time averaging because of the non-stationary boundary conditions on the Sun (Mariani and Neubauer, 1990). The strong violation of the assumption Bo = 0 takes place for instantaneous magnetic fields. The measurements show that Bo :/= 0, and that Bo is comparable to Be as a rule for instantaneous fields, and the more general relations (3-5) are better applicable. 3) Let us consider solution (3-6) near the stationary "neutral points", where B~ = 0 in the rotating reference frame. This situation corresponds to the long-living "neutral point" B0~ = 0 in the solar wind source region at r = r0. In this case, the "neutral point" is "projected" from the sphere r0 along the Archimedian spiral into the heliosphere and forms the "neutral line" there. Hence, isolated stationary neutral points cannot exist in the heliosphere and the manifold of neutral points is organized in neutral lines of a spiral shape. The field B(0, Bo, B~) is tangential to the heliocentric spheres at the mentioned neutral lines. In the simplest case when B0~ = B~ - 0 one obtains a'family of similar open or closed field lines on concentric spherical surfaces disconnected from the Sun and from the infinity with two components Bo ~ r -~, B~ ~ r -~, obeying the condition of a vanishing divergence. 4) The potential part of the solar magnetic field can be represented by the multipols varying with the Hale cycle (Hoeksema et al., 1983). The dominat poloidal (r, 0) contributions on the source surface are produced by the lowest harmonics. In addition to this, non-potential fields are present and the toroidal (~) component is not negligible. The magnetic field on the Sun and in the heliosphere is more structured during years of higher solar activity. Higher poloidal and toroidal harmonics are stronger at this time. The number of non-stationary zero points and lines increases at the source surface and in the heliosphere. Accordingly, closed and loop-like structures are more common at the high solar activity years on the Sun and in the heliosphere. 5) A full set of the nonlinear MHD equations should be used instead of Eqs. (1,2) if the velocity field g(~', t) is not given. The structures with laminar and turbulent states of the magnetic field are known when the overall geometry is open, closed or intermittent one. More observations, especially multi-point and tomographic ones, are needeed for a better understanding of the heliospheric structure. Sometimes, the linear analysis is applicable. In this case, "perturbations" are small and a background state is well defined. It could be given by the stationary or evolving solution. The "eigen mode problem" arises for small perturbations. This can be solved only in the simplest cases exemplified by discrete and/or continuous spectra of "waves" and "convective branches". Practically, explicit solutions for the background state and perturbations are not easy to find. Partially because of these difficulties, we do not have as yet even a "common language" to describe observations. A dimensionless scaling approach is pomising in this respect. 6) The degree of openness of the heliospheric field increases with the distance from the Sun in the simplest model, when the solar formation region geometry is represented by a superposition of a dipole and a current sheet. The ratio of the closed field line areas at a given distance r from the Sun decreases with this distance approximately as

-63-

I.S. Veselowsky \

-

1 + ar ) 1 where a is proportional to the length scale #/(~. Here # is the magnetic dipole of the "su n, (I) is the magnetic flux of the polar coronal holes, which is proportional to the heliospheric current sheet strength. This model is semi-quantitatively correct for the solar minimum years. The solar magnetic dipole strength decreases with the activity cycle, when the current sheet strength increases and the number of open field lines should increase because of this. As a consequence, the outer corona looks more radially stretched during the maxima. An opposite tendency is imposed by other, time dependent factors. The corona is more structured and dynamical at maxima. Coronal mass ejections are more numerous. Radially expanding non-stationary elements bring their local fields which could be both open and closed, as well as intermittent. As a result, the overall heliospheric field topology needs additional investigations especially during the high solar activity. 4. C O N C L U S I O N S Non-stationary magnetic fields in the heliosphere are partially open, semi-open or closed in different proportions dependent on time and location. The instantaneous magnetic field lines at distances about 1 AU on average strongly deviate from conical Archimedian spirals and provide the non-stationary magnetic coupling between different latitudes, longitudes and distances inside the corresponding correlation lengths, which are determined mainly by variable boundary conditions near the Sun for B and g as well as by non-local dynamical processes in the heliosphere. Because of the heliospheric curent sheets, the degree of openness of the field in general increases with the distance from the Sun. The heliospheric magnetic field is more fragmented, dynamical and has a larger number of neutral lines, separators and closed elements connected and disconnected from the Sun during the years of high activity, but the overall topology remains to be investigated. 5. A c k n o w l e d g m e n t s This work was supportedd by the RFBR grant 98-02-17660, the Federal Program "Astronomy" project 1.5.6.2, the Federal Program "Universities of Russia" project 990600 and the INTAS/ESA grant 99-00727. The author is grateful to the Organizing Committee of the COSPAR Colloquium in Potsdam for the financial support facilitating the attendance at the meeting and for the help in preparation of the final manuscript.

1. Mariani, F., and F.M. Neubauer. The Interplanetary Magnetic Field, in Physics of the Inner Heliosphere I, eds. R.Schwenn and E.Marsch, pp.183-206, Springer Verlag, Berlin, 1990. 2. Burlaga, L.F.E., Magnetic Clouds, in Physics of the Inner Heliosphere II, eds. R.Schwenn and E.Marsch, pp.1-22, Springer Verlag, Berlin, 1991. 3. Parker, E.N. Dynamics of the Interplanetary Gas and Magnetic Fields, Astrophys. J., 328, 848-855, 1958. 4. Hoeksema, J.T., J.M.Wilcox, and P.H.Scherrer, The Structure of the Heliospheric Current Sheet: 1978-1982, J. Geophys. Res., 88, 9910-9918, 1983.

-

64

-

Solar cycle heliospheric interface variations" influence of neutralized solar wind N. A. Zaitsev a and V. V. Izmodenov b* aKeldysh Institute of Applied Mathematics, Russian Academy of Sciences, [email protected] bDepartment of Aeromechanics and Gas Dynamics, Faculty of Mechanics and Mathematics, Lomonosov Moscow State University, [email protected] New non-stationary self-consistent model of the solar wind interaction with a twocomponent (atoms and plasma) local interstellar cloud is proposed. In this model the primary and secondary interstellar atoms are treated as quasi-stationary kinetic gas. Population of H atoms originated in the supersonic solar wind is considered as zero-pressure fluid. Specific non-stationary effects introduced by the solar cycle fluctuations of the neutralized solar wind are explored. 1. I n t r o d u c t i o n The solar wind (SW) interacts with the local interstellar cloud (LIC). The heliospheric interface is formed in this interaction. The structure of the interface and the interface plasma flow depend on the parameters of the SW and LIC and their variations with time. The LIC velocity with respect to the Sun ( VLIC ~ 26 km s -1) and the LIC temperature (TLIC ~ 7000 K) are reliably established. The sonic velocity, aLIC, corresponding to TLIC, is smaller than VLIC. Therefore, the interstellar flow is supersonic and the LIC Mach number, MLIC = VLIC//aLIC, is larger than unity. Therefore, the interstellar plasma flow is supersonic and two-shock plasma structure is formed in the LIC/SW interaction. Interstellar atoms, galactic and anomalous cosmic rays, interstellar and interplanetary magnetic fields may affect the interface. However, the basic features of the heliospheric plasma interface are the same: 1) the heliopause (HP) separates the SW plasma from interstellar plasma, 2) the termination shock (TS) decelerates, heats and compresses the solar wind plasma and 3) there is the pile-up plasma region - plasma wall- between the heliopause and the bow shock. Interstellar H atom charge exchange with protons and significantly influence the heliospheric interface structure. The mean free path of H-atoms is compatible with the characteristic length of the problem considered. Therefore it is not correct to describe the * T h e research described in this publication was made possible in part by Award No.RP1-2248 of the U.S. Civilian Research and Development Foundation for the Independent States of the Former Soviet Union (CRDF), INTAS projects # 97-0512 and #YSF 00-163, RFBR grants # 01-02-17551, 99-02-04025, and International Space Science Institute in Bern.

-65

-

N.A. Zaitsev and K V. Izmodenov

H atom motion hydrodynamically ([1], [3]). The self-consistent axisymmetric model of the SW interaction with two-component (plasma and H atoms) LIC has been developed by Baranov and Malama [2]. A kinetic approach was used to describe H atoms in the model. However, the model does not take into account non-stationary processes in the SW. Steinolfson [4] has numerically investigated the problem of the heliospheric interface response to 180 day period fluctuations in the solar wind ram pressure in the case of subsonic fully ionized interstellar gas. He found that the variation in the distance to the termination shock is only about 1 AU. Karmesin et al. [5] and Baranov and Zaitsev [6] studied the influence of the 11 year variation of the solar wind pressure onto the heliospheric interface with 2D hydrodynamic models. These authors concluded that the response of the position of the termination shock to the changes of the solar wind parameters is within 8 - 12% (or about 10 AU). The response of the heliopause is smaller and the response of the bow shock is negligible. Baranov and Zaitsev also pointed out that in the supersonic solar wind as well as in the heliosheath the plasma flow has a quasi-stationary behaviour, while in the pile-up region between the bow shock and the heliopause there is a sequence of shocks and rarefaction waves. However, the mean plasma distribution is close to the stationary plasma distribution. In this work we investigate the effect of the solar cycle fluctuations of the fast solar neutrals flux onto the heliospheric interface. These energetic neutrals are created by charge exchange inside the heliopause and have significant influence on the compressed interstellar plasma (see, e.g., [3]). 2. F o r m u l a t i o n of t h e P r o b l e m The model [2] clearly shows, that there are four different types of H atoms in the heliospheric interface: i) the unperturbed interstellar neutral H atoms called Primary Interstellar Atoms or PIAs; ii) the compressed, decelerated, and Heated Interstellar Atoms (HIAs) formed by charge exchange with heated interstellar protons outside the heliopause; iii) the neutralized, decelerated, and Heated Solar Wind Atoms (HSWAs) formed in the heliosheath by charge exchange between the neutral interstellar gas and the hot protons of the decelerated and compressed solar wind, and iv) the neutralized Supersonic Solar Wind Atoms (SSWAs). These types of neutrals have different energy and spatial distributions in the interface. In Baranov-Malama model the influence of neutrals on plasma flow were taken into account as source terms Q in the right parts of ithe Euler equations for plasma component. To study properly the solar cycle influence on the heliospheric interface one should solve simultaneously the time-dependent Euler and Boltzmann equations. While the development of the corresponding Monte-Carlo algorithm is still in progress we would like to understand the basic physical effects of the solar cycle on the time-dependent interface from a simplified model. To study the influence of interstellar atoms we made the following assumptions: 1. Since the positions of the BS and the HP do not change significantly and fluctuations of plasma in this region are around stationary distributions, we assume that there is no influence of the solar cycle on the PIA's and HIA's. Hence the source terms QPIA and

-

66-

Solar cycle heliospheric interface variations: Influence of neutralized solar wind

QHIA into the plasma equations do not depend on the solar cycle and can be taken from the stationary solution. 2. We neglect the changes of HSWA's over the solar cycle. 3. For the number densities npIA and nHIA we assume that nH -- exp

7~H,TS~

T VH,TS

where a~z is the charge exchange cross section; rtE, VE a r e the solar wind proton number density and temperature, respectively; nH,TS VH,TS the number densitrnd velocity of interstellar atoms at the termination shock, r is the heliocentric distance, rE--1 AU is the distance to the Earth. This formula is the zero order approximation of interstellar neutral distributions in the interface, but for our objectives it is sufficient. 4. We assume that SSWA's can be treated as a zero-pressure fluid. The governing equations for this fluid are mass and momentum conservation lawis under the condition that the pressure PSSWA = O. The mutual influence of H-atoms and the plasma flow is modelled by the corresponding source terms in the mass conservation law and the momentum conservation law. In the stationary case this model gives the distribution of the SSWA's number density close to that obtained with the Monte-Carlo simulation. In order to simulate the ll-year solar cycle we changed the plasma velocity at the Earth orbit according to a sinusoidal law so that the ram pressure was varied by a factor of two. The sinusoidal variation of VE is the first harmonic of the real time dependence of the SW parameters. 3. N u m e r i c a l r e s u l t s Numerical results discussed below were obtained on the basis of a statinary solution computed for the following parameters [7]" np~ - 0.07 cm -3, n H ~ -- 0.2 cm -3, V~ = 25 kin~s, T~ = 5672K, nEo = 7 cm -3, VEO = 450 kin~s, TEO = 73507K. Our calculations show that the qualitative features of the non-stationary LIC - SW interaction established in [6] take place in the presence of neutral H-atoms as well. But the effect of the solar activity cycle is quantitatively stronger because the interface is closer to the Sun. For example, the TS excursion during the solar cycle on the axis of symmetry is about 30 AU, i.e. about 30% of its mean solar distance. Figure 1 shows the farthest and the nearest discontinuities positions in the upwind hemisphere parameters (solid lines). In the same figure the steady state positions of the BS, HP and TS are shown (by circles). One can see that the region between the BS and HP becomes wider: the mean location of the BS is farther from the Sun and the mean location of the HP is closer to the Sun than the corresponding steady state positions. The mean distributions of the plasma number density on the axis of symmetry in the region between the BS and HP is shown in Figure 2. The stationary distribution of the plasma number density in the same region is also shown. One can see that the mean density in the non-stationary cases is much less than in the stationary solution. This phenomenon can significantly influence the penetration of neutral H-atoms into the solar system. Figure 2 also shows the presense of a sequence of shocks and rarefaction waves moving from the HP to the BS similar to the one obtained in [6].

-67-

N.A. Zaitsev and V. V. Izmodenov

Figure 1. The geometrical pattern of the discontinuities: solid lines minimum and maximum heliocentric distances, circles the stationary position of the TS.

Figure 2. The interstellar plasma density: stationary solution (solid line), nonstationary solution (diamonds) and the mean density distribution (dashed line).

4. Conclusions The conclusions of presented study can be summarized as following: 1. The solar cycle influence on the variation of the termination shock is stronger in the case with H atoms than it would be in the case without H atom component. 2 Due to the solar cycle variations of the neutralized solar wind, i.e. atoms created in the supersonic solar wind by charge exchange with solar wind protons, the region between the heliopause and the bow shock become wider and mean plasma density in the region become smaller than for the stationary problem. This can be important for interpretations of heliospheric absorption in Ly-c~. REFERENCES

1. Baranov, V. B., V. V. Izmodenov, and Y. G. Malama, J. Geophys. Res., IDS, 95759585, 1998. 2. Baranov, V. B., and Y. G. Malama, J. Geophys. Res., 98, 15157-15163, 1993. 3. Izmodenov, V. V., this issue. 4. Steinolfson, R. S., J. Geophys. Res., 99, no. 7, pp. 13307-13314, 1994. 5. Karmesin, S. R., Liewer, P. C. and Brackbill, J. U. Geophys. Res. Lett., vol. 22, no. 9, pp. 1153- 1156, 1995. 6. Baranov, V. B. and Zaitsev, N. A. Geophys. Res. Lett., Vol. 25, No. 21, pp. 40514054, 1998. 7. Izmodenov V. V., J. Geiss, R. Lallement, et al., J. Geophys. Res. 104, 4731,1999a.

-68-

Time dependent radiation pressure and time dependent, 2D ionisation rate for heliospheric modelling M. Bzowski ~ * ~Space Research Centre PAS, Bartycka 18A, 00-716 Warsaw, Poland

1. M O T I V A T I O N The objective of this research was to develop time dependent, 3D models of the 3 solar parameters affecting the distribution of neutral interstellar hydrogen in the inner heliosphere: Lyman-c~ radiation pressure, charge exchange rate of H-atoms on solar wind protons, and photoionisation rate. The models are intended to be used in a program calculating the hydrogen density (see Bzowski et al., this volume) and while they are based on actual measurements, some simplifying assumptions were a priori made: the modelled quantities depend as 1/r 2 on the heliocentric distance, the periodicities allowed are longer than ~ 1 year, temporal variations of the solar wind occur immediately throughout the heliosphere; furthermore the functional form of the models was chosen to facilitate their use in numerical calculations. 2. R A D I A T I O N

PRESSURE

In lack of reliable data available it was assumed that the solar Lyman-a radiation pressure is spherically symmetric. The time series of the net solar Lyman-a flux from several spacecraft, recalibrated by Tobiska et al. (1997) and covering a time interval from 1977 till 1997.5, was expressed in the units of solar gravity assuming that the flux at the line centre is by number equal to the net flux. Next a periodogram analysis was performed (Press & Rybicki 1986) and based on the periodicities Pj found, a formula J, # (t) - #o + E (PJ cos cejt + qj sin wit), where wj - 2~r/Pj,

(1)

j=l

was fitted. The time series and the fitted model are presented in Fig.1 (left-hand panel) and the numerical values of the parameters in Table 1. 3. I O N I S A T I O N R A T E The net ionisation rate is an arithmetic sum of photoionisation and charge exchange rates, calculated as follows. *M.B. was supported by Polish SCSR grants 2P03C 004 14 and 2P03C 005 19.

-69-

M. Bzowski Table 1 Fitted parameters of radiation pressure and ionisation rates j Radiation pressure Photoionisation Charge exchange #0 = 1.27864 ~ph0 = 11.1862. 10-8s -1 ~chx0 = 55.5692- 10-Ss -1 qj Pj [y] pj 10 .8 qj 10 .8 Pj [y] pj 10 -8 qj 10 -8 Pj[Y] Pj 1 10.653 0.26043 -3.31286 30.007 1.3110 0.4609 27.261 5.1296 -7.5059 26.1185 -0.45974 0.46164 15.125 -0.0292 0.0046 7.5450 1.8444 2.2619 34.8206 -0.04823 0.12659 10.761 5.1934 0.9897 5.2490 -2.5727 -4.6350 43.1192 -0.17331 -0.26181 8.2470 -0.0822 1.3457 B0 = 25.3629 9 10 -s s-1 , b = 3 . 3 5 9 1 9 f o r n = 2 , B 0 = 2 6 . 0 1 9 3 - 1 0 - a s -1, b = 1 7 4 1 . 7 9 f o r n = 8 .....

Lyman-a radiation pressure 2.5 . . . . . . . . . . . . . . . . . . . . . . . .

1.5 1 0.5~

Solar photoionisation rate

'~J : .

1975

.

.

.

.

.

......... .

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

~

0.5

1980 1985 1990 1995 2000 time [y]

.

1975

.

.

.

.

.

.

.

.

.

.

1980 1985 1990 1995 2000 time [y]

Figure 1. Solar radiation pressure (left-hand panel) and photoionisation rate (right-hand panel): models defined in Eq.1 vs data.

3.1. Photoionisation In absence of sufficiently long time series of actual data, the rate of photoionisation of hydrogen at 1 AU was calculated from a proxy based on the linear correlation between the solar 10.7 cm radio flux and the solar EUV flux, indicated by Hodges & Tinsley (1981). The formula to convert the 10.7 cm radio flux to the hydrogen photoionisation rate at 1 AU from the Sun was developed following a suggestion by Cummings (private communication), who derived this quantity from the solar EUV flux reported by Torr et al. (1979) for the day 113 of 1974 (solar minimum) and day 50 of 1979 (solar maximum). The photoionisation rates for these dates were equal, correspondingly, to 0 . 5 3 2 . 1 0 -7 s -1 and 1.87.10 -7 s -1 and the 10.7 cm adjusted flux to 66.5 and 214.10-22 W cm -2 s -1. Based on these data, a system of two equations with two variables was solved, yielding the following conversion formula: 9ph --

8.7661.1012 F10.7[W m -2 s -1] - 5.84576.10 -1~

(2)

With the use of this formula, daily averages of the 10.7 cm flux, covering the interval from 1948 till 1998 and publicly available in the Internet ( h t t p : / / w w w . d r a o . n r c . c a / i c a r u s / sol_home.shtml), were converted to a photoionisation rate time series and subjected to a similar analysis as in the case of radiation pressure. The periodicities and model parameters found are presented in Table 1 and a plot of the data and the model curve can be seen in Fig.1 (right-hand panel). Good correlation between the solar Lyman-c~ radiation

-70-

Time dependent radiation pressure and time dependent, 2D ionisation rate for heliospheric modeling pressure and the hydrogen photoionisation rate is striking but the model does not fit perfectly to the data; in particular, a 30% discrepancy about 1996 is visible. While it was possible to reduce this discrepancy by clipping the dataset to the recent solar cycle, this solution was rejected because it was giving unrealistic forward extrapolation and wrong timing of minima and maxima of the earlier solar cycles. Since photoionisation is not the dominant sink of hydrogen in the heliosphere, the 1996 underestimate should not be important for hydrogen distribution modelling.

3.2. Charge exchange Details of derivation of the charge exchange model will be published elsewhere because of the lack of space. In this paper results only will be presented.

Figure 2. Charge exchange rate adjusted to 1 AU in ecliptic (left-hand panel) and from the Ulysses Fast Latitude Scan (right-hand panel).

The rate of charge exchange between hydrogen atoms of the neutral interstellar wind and protons of the supersonic solar wind is calculated from the formula ~ch-x (t, (~) -- O'ch_ x (Vrel) ~tSW (t, r

Vrel

(t, r

/r 2, where Vre 1 -- VSW

(3)

and r is the heliocentric distance in AU, r the heliographic latitude, t time, V~el the relative velocity between the colliding particles, Gch-x the reaction cross section, nsw the solar wind density at 1 AU, and Vsw the solar wind speed, assumed independent on r. The charge exchange cross section used to be calculated from a formula by Fite et al. (1962) or by Maher & Tinsley (1977). But in comparison with experimental data compiled by Barnett et al. (1990), Fite's formula is 30 40% too high and Maher & Tinsley's is good from 120 to 450 km/s. Since the velocity range relevant for heliospheric modelling is from just a few to about 800 km/s, a new formula, spanning the whole range and accurate to at least 2.5%, was fitted to Barnett's data (velocity is expressed in cm/s)" a~h-x (v) - -2.01848.10 -17 In 3 v+1.00136.10 -15 In 2 v-1.71725.10 -~4 In v+1.03807.10-~3(4) Based on the data from Ulysses (Marsden 1996) it was assumed that the solar wind is steady in the polar regions and its variability is limited to an equatorial band. Hence a

-71 -

M. Bzowski functional form of the charge exchange rate model was adopted as r 2/Sch_ x ( r

/~o - Bo + ~ (pj cos wjt + qj sin wit)

t) -- B 0 -t-

exp [-b r

(5)

j=l

To find parameters of this formula, time series of daily averages of charge exchange rate were calculated using Equ.(3) from multispacecraft in-ecliptic solar wind data (time span from 1972 till 1998) and from Ulysses Fast Latitude Scan. The first time series was then converted to monthly averages and subjected to an analysis as for the photioinisation rate and yielded parameters COj - - 27c/Pj, pj, qj, and/50 (Table 1, Fig.2, left-hand panel). The latitudinal profiles were fitted to the second time series and yielded B0 and b for the two models assumed (n = 2, referred to as the gaussian bulge, and n = 8, referred to as the rectangular bulge, see Fig.2, right-hand panel). ,--, 10

1

H+ e

(4)

H

sw

(population 1 )

Figure 1. Qualitative picture of the solar wind interaction with the LISM without magnetic field. Regions 1 - 4 are the supersonic solar wind (preshock of the TS), the thermalized (in the TS) solar wind, the thermalized (in the BS) LISM plasma and the LISM plasma respectively. Dotted lines are trajectories of different H-atom populations.

The mutual effect of the plasma and hydrogen components was self-consistently taken into account in [4]. These authors used the simplest possible hydrodynamic scenario, where it was assumed that the velocity VH and the temperature TH of neutral hydrogen remain constant throughout the whole computational domain and that the LISM H-atoms are lost only in the regions 1, 2 and 3 (Figure 1) due to resonance charge exchange. This means that they solved only the continuity equation for neutrals. It was impossible to interpret experiments connected with the Lyman-alpha glow on the basis of this simplest scenario. The model suggested in [4] was criticized in [8] where the correct kineticgasdynamic model was suggested. H-atoms are kinetically described by the Monte Carlo method in this model, because the mean free path of the H-atoms is comparable with the characteristic length of the problem (for example, with the size of the heliopause) and the hydrodynamic approximation is not correct all over for the H-atom component. Recently, authors of [9,10] and [11] revisited the hydrodynamic description of H-atom motion in the problem of the solar wind interaction with the LISM although this description is not correct as we have seen above. It was shown in [12] and [13] that the distribution function of the all H-atom populations is not Maxwellian and, therefore, the multi-fluid model in [10] cannot be used to interpret experimental data. The same result is obtained in [29] for subsonic plasma-plasma interface. In Figure 2 (Figures 5c and 7b from [12]) some results of the kinetic-gasdynamic model in [8] and the multi-fluid model in [10] are compared. From Figure 2a we see that the bulk velocity (along the axis of symmetry) of the LISM hydrogen atoms (primary and secondary), which are responsible

-

100-

Possible effects of the interstellar magnetic field...

Figure 2. Comparison of some results obtained on the basis of the kinetic-gasdynamic model by Baranov and Malama (1993) and the multi-fluid model by Zank et al. (1996): (a) the component (along the axis of symmetry) of the H-atom (born in region 1) velocity and (b) the H-atom (born in region 1) temperature as functions on the heliocentric distance.

for the interpretation of the scattered solar Lyman-alpha radiation, changes its sign in the downwind direction for the model in [10]. In this case the continuity equation is not satisfied in the stationary approximation [10] and, therefore, the results of these authors are not physically real. Figure 2b demonstrates another physically unrealistic result obtained in [10]. Namely, the temperature of the energetic solar wind H-atoms, which are born due to the charge exchange between the LISM H-atoms and the solar wind protons (in the region 1), is two orders of magnitude larger (10 ~ K) than the solar wind proton temperature although there are no reasons for their heating. It follows that the model in [10] can also not be used to interpret results of planned experiments connected with the detection of energetic neutral atom (ENA) fluxes with 1 AU [14]. In the last few years several numerical magnetohydrodynamic (MHD) models of the solar wind interaction with the magnetized LISM were published although at present neither direction nor magnitude of the interstellar magnetic field in the vicinity of the solar system are experimentally known. The example of multi-fluid model results mentioned above (Figure 2) shows that caution must be exercised by observers to interpret experimental results on the basis of theoretical models. That is why the part of this presentation will be devoted to the critical review of MHD models of the solar wind interaction with the magnetized LISM (Section 2). In the Section 3 some problems of the interstellar magnetic field effect on H-atom penetration from the LISM to the solar system will be considered.

-101 -

V.B. Baranov

(a)

Hyperbolicregion

(b) Ellipticregion

(c)

Quasihyperbolic region

BS HI'

I-IP

Figure 3. Possible geometrical patterns of the MHD flow in the case when the interstellar magnetic field Boo and the bulk velocity V ~ are parallel. Hyperbolic, elliptic and quasihyperbolic regions are (a),(b) and (c) respectively.

2. A B O U T T H E F U N D A M E N T A L M H D P R O B L E M C O N N E C T E D W I T H THE INTERACTION BETWEEN THE SOLAR WIND AND MAGNETIZED LISM'S PLASMA It is known that a supersonic flow near hard bodies is characterized by a bow shock formation. The wings of the bow shock are weak discontinuities where parameters are continuous but their gradients are not. In doing so the weak discontinuities coincide with characteristics which are lines of small perturbation propagation. Their direction always coincides with a downwind direction in the classic gas dynamics. We have a more complicated situation in the magnetohydrodynamics (MHD) because there are four velocities of small perturbations (entropic, slow and fast magnetosonic and alfvenic) in the presence of the magnetic field. In particular, characteristics can be directed in the upwind direction as well as in the downwind one (see, for example, in [15]). The heliopause HP plays the role of the hard body in the problem of the solar wind interaction with the magnetized interstellar plasma. Possible geometrical patterns of the flow considered are qualitatively shown in Figure 3 in the most simple case when the magnetic field B~ and the bulk velocity V~ of the interstellar gas are parallel (see [16]). Figure 3a demonstrates a possible shape of the bow shock at V~ > max(a0, aA), where a0 and a A - are the sonic and alfvenic velocities of the interstellar plasma respectively, i.e. at the weak interstellar magnetic field. It is the hyperbolic region and analytical results give rise to real characteristics directed to the downwind direction (see wings of the bow shock in Figure 3a). To investigate the shape of the bow shock in the vicinity of the nose region numerical calculations must be used. All published numerical results (see,

-

102-

Possible effects of the interstellar magnetic field...

for example, [16-21]) are obtained in the case presented in Figure 3a (in the hyperbolic region). The geometrical pattern of the flow considered can be significantly changed at an increase of the magnetic field intensity. For example, the bow shock must be absent in the elliptical region which is characterized by inequalities Vo~ < aoaA/(a 2 + a~) -1/2

and

min(a0, aA) < Vo~ < max(a0, aA)

Real characteristics are absent in this region (see Figure 3b). Characteristics are directed in the upwind direction (see Figure 3c) if inequalities

a0a /(a +


0 solar cycle (as in the 1990's). Ulysses, exploring the "inner" heliosphere, show a remarkable time profile compared to the measurements at Earth and in the outer heliosphere. The heliographic latitude and the spacecraft distance to the Sun are given at the top of Fig. 7. In February 1995 the spacecraft crossed the heliographic equator at a distance of 1.3 AU. At that time Ulysses observations are in good agreement with measurements at Earth. The intensities are then increasing with Ulysses latitude and show a maximum at polar latitudes. Such a behavior is expected from our current understanding of spatial modulation in the heliosphere, and has been intensively investigated

-

179-

B. Heber and A. Cummings

by the Potchefstroom - Ulysses COSPIN/KET collaboration for GCR protons (e.g., Heber and Potgieter, 2000; and Burger et al., 2000) and electrons (e.g., Potgieter et al., 1999). While for GCR the latitudinal gradient was much smaller than anticipated, Fig. 7 clearly demonstrates that these gradients are not negligible for ACR oxygen, leading to the fact that the intensity measured at 2 AU and 80 ~ heliographic latitude corresponds to intensities beyond 40 AU. 1. S u m m a r y

During the minimum phase of the last solar cycle 22 tremendous progress has been made in both measuring and understanding ACRs. The elemental spectra have been determined more precisely than ever before thanks to the advanced instrumentation on Wind and Advanced Composition Explorer at 1 AU. As a result, ACR-like enhancements have been discovered for new species. Measurements of the isotopic composition reflect the corresponding composition of pickup ions and interstellar neutrals, and will also provide insights into the generation as well as acceleration processes of galactic cosmic rays. The current paradigma that ACRs are mainly singly charged was confirmed by measurements of the SAMPEX satellite which uses the geomagnetic field as a filter. Using the geomagnetic filter approach, higher charged ACRs have also been discovered. The charge composition of ACRs can be used to put constraints on the acceleration time scales at the heliospheric termination shock. Another advantage of this approach might be the possibility to track the ACR intensities at 1 AU through the current solar maximum into the next solar minimum around 2003. Voyager 1, at ~78 AU, has not yet reached the heliospheric termination shock. The heliospheric network of space probes available at the 1990's solar minimum showed that the latitudinal gradients in the inner heliosphere are in good agreement with the one in the outer heliosphere. Since these gradients are large compared to the radial gradients, Ulysses at 2 AU 80 ~ observed approximately the same intensity as Voyager 2 beyond 40 AU. ACKNOWLEDGMENT The authors thank E McDonald, R. A. Leske, and R. Marsden for providing IMP helium, 1 AU oxygen and Ulysses oxygen data as well as for helpful discussions. We also thank E. C. Stone for helpful suggestions. This work was partially supported by NASA under Contract NAS7-918. REFERENCES D.N. Baker, G. M. Mason, O. Figueroa, G. Colon, J. G. Watzin, and R. M. Aleman. An overview of the solar, anomalous, and magnetospheric particle explorer (SAMPEX) mission. IEEE transactions on Geos. and remote Sensing, 31:531-541, 1993. 2. R . A . Burger, M. S. Potgieter, and B. Heber. Rigidity dependent of cosmic-ray proton latitudinal gradients measured by the Ulysses spacecraft: Implications for the diffusion tensor. J. Geophys. Res., 2000. in press. 3. A.C. Cummings and E. C. Stone. Composition of anomalous cosmic rays and implications for the heliosphere. Space Science Reviews, 78:117-128, October 1996. 4. A.C. Cummings, E. C. Stone, and C. D. Steenberg. Composition of anomalous cosmic rays 1.

- 180-

Anomalous cosmic ray observations in the inner and outer heliosphere and other ions from voyager observations. In 26th International Cosmic Ray Conference, volume 7, page 531, 1999. 5. H. Fichtner. Anomalous cosmic rays" Messengers from the outer heliosphere. Space Sci. Rev., 2001. 6. L.A. Fisk, B. Koslovsky, and R. Ramaty. An interpretation of the observed oxygen and nitrogen enhancements in low-energy cosmic rays. Astrophys. J. Lett., 190:L35, 1974. 7. M. Garcia-Munoz, G. M. Mason, and J. A. Simpson. A new test for solar modulation theory: the 1972 May-July low-energy galactic cosmic ray proton and helium spectra. Astrophys. J. Lett., 182:L81, 1973. 8. G. Gloeckler, L. A. Fisk, T. H. Zurbuchen, and N. A. Schwadron. Sources, injection, and acceleration of heliospheric ion populations. In AlP Conf. Proc. 528." ACE 2000 Symposium, volume 528, pages 221 +, 2000. 9. G. Gloeckler, J. Geiss, H. Balsiger, L. A. Fisk, A. B. Galvin, F. M. Ipavich, K. W. Ogilvie, R. von Steiger, and B. Wilken. Detection of interstellar pick-up hydrogen in the solar system. Science, 261:70-73, July 1993. 10. G. Gloeckler, J. Geiss, and L. A. Fisk. Heliospheric and Interstellar Phenomena Revealed from Observations of Pickup Ions. Springer-Praxis, 2000. in press. 11. B. Heber and R. A. Burger. Modulation of galactic cosmic rays at solar minimum. Space Sci. Rev., 89" 125-138, 1999. 12. B. Heber and M. S. Potgieter. Galactic cosmic ray observations at different heliospheric latitudes. Adv. Space Res., 26(5):839-852, 2000. 13. J. R. Jokipii. Theory of multiply charged anomalous cosmic rays. Astrophys. J. Lett., 466:L47-+, July 1996. 14. B. Klecker, M. C. McNab, J. B. Blake, D. C. Hamilton, D. Hovestadt, H. Kaestle, M. D. Looper, G. M. Mason, J. E. Mazur, and M. Scholer. Charge state of anomalous cosmic-ray nitrogen, oxygen, and neon: Sampex observations. Astrophys. J. Lett., 442:L69-L72, April 1995. 15. B. Klecker, R.A. Mewaldt, J.~TV. Bieber, A.(~. Cummings, L. Drury, J. Giacalone, J.l~. Jokipii, F.(~. Jones, M.t~. Krainev, M.A. Lee, J.A. Le Roux, R.G. Marsden, F.13. McDonald, R.13. McKibben, C.I). Steenberg, M.G. Bating, D.(~. Ellison, L.J. Lanzerotti, R.A. Leske, J.l~. Mazur, H. Moraal, M. Oetliker, V.S. Ptuskin, R.~,. Selesnick, and K.J. Trattner. Anomalous cosmic rays. Space Science Reviews, 83:259-308, 1998. 16. B. Klecker, M. Oetliker, J.B. Blake, D. Hovestadt, G.M. Mason, J.E. Mazur, and M.C. M c N a b . . In Proc. 25. Inter. Cosmic Ray Conf., volume 7, pages 333-336, 1997. 17. M. A. Lee and L. A. Fisk. Shock acceleration of energetic particles in the heliosphere. Space Sci. Rev., 32:205-228, 1982. 18. R. A. Leske. Anomalous cosmic ray composition from ace. In AlP Conf. Proc. 516." 26th International Cosmic Ray Conference, ICRC XXVI, volume 26, pages 274+, 2000. 19. R. A. Leske, R. A. Mewaldt, E.R. Christian, C/,M.S. Cohen, A. C. Cummings, EL. Slocum, E.C. Stone, T.T. von Rosenvinge, and M.E. Wiedenbeck. Observations of anomalous cosmic ray at 1 au. In AlP Conf. Proc. 528." ACE 2000 Symposium, volume 528, pages 293+, 2000. 20. F.B. McDonald, E Ferrando, B. Heber, H. Kunow, R. McGuire, R. Mtiller-Mellin, C. Paizis, A. Raviart, and G. Wibberenz. A comparative study of cosmic ray radial and latitudinal gradients in the inner and outer heliosphere. J. Geophys. Res., 102:4643-4652, March

-181

-

B. Heber and A. Cummings 1997. 21. E B. McDonald, B. Heikkila, N. Lal, and E. C. Stone. The relative recovery of galactic and anomalous cosmic rays in the distant heliosphere: Evidence for modulation in the heliosheath. J. Geophys. Res., 105:1-8, January 2000. 22. R.A. Mewaldt, J.R. Cummings, R.A. Leske, R.S. Selesnick, E.C. Stone, and T.T. von Rosenvinge. A study of the composition and energy spectra of anomalous cosmic rays using the geomagnetic field. Geophys. Res. Let., 23:617-620, 1996. 23. E. Moebius, D. Hovestadt, B. Klecker, M. Scholer, and G. Gloeckler. Direct observation of he(+) pick-up ions of interstellar origin in the solar wind. Nature, 318:426-429, 1985. 24. H. Moraal. The discovery and early development of the field of anomalous cosmic rays. Space Sci. Rev., 2001. 25. M. E. Pesses, J. R. Jokipii, and D. Eichler. Cosmic ray drift shock-wave acceleration, and the anomalous component of cosmic rays. Astrophys. J. Lett., 246:L85-L88, 1981. 26. M. S. Potgieter. The modulation of galactic cosmic rays in the heliosphere: Theory and models. Space Science Reviews, 83:147-158, January 1998. 27. M. S. Potgieter, S. E. S. Ferreira, B. Heber, E Ferrando, and A. Raviart. Implications of the heliospheric modulation of cosmic ray electrons observed by ulysses. Advances in Space Research, 23:467-470, 1999. 28. M. S. Potgieter and H. Moraal. Acceleration of cosmic rays in the solar wind termination shock, i - a steady state technique in a spherically symmetric model. Ap. J., 330:445-455, July 1988. 29. D. V. Reames. Quiet-time spectra and abundances of energetic particles during the 1996 solar minimum. Ap. J., 518:473+, 1999. 30. J. A. Le Roux. Anomalous cosmic rays: current and future theoretical developments. Space Sci. Rev., 2001. 31. E. C. Stone, A. C. Cummings, D. C. Hamilton, M. E. Hill, and S.M. Krimigis. Voyager Observations of Anomalous and Galactic Cosmic Rays During 1998. In Proc. 26th International Cosmic Ray Conference, Salt Lake City, Utah, USA, August 17-25, 1999, volume 7, page 551, 1999. 32. E.C. Stone, L.F. Burlaga, A.C. Cummings, W.C. Feldman, W.E. Frain, J. Geiss, G. Gloeckler, R. Gold, D. Hovestadt, S.M. Krimigis, G.M. Mason, D. McComas, R.A. Mewaldt, J.A. Simpson, T.T. Rosenvinge, and M. Wiedenbeck. the Advanced Composition Explorer, volume 203, page 48. AIP Conference Proceedings, 1989. 33. V. M. Vasyliunas and G. L. Siscoe. On the flux and the energy spectrum of interstellar ions in the solar system. J. Geophys. Res., 81:1247-1252, March 1976. 34. M. Witte, H. Rosenbauer, M. Banaszkiewicz, and H. Fahr. The ulysses neutral gas experiment - determination of the velocity and temperature of the interstellar neutral helium. Advances in Space Research, 13:121-130, June 1993. 35. G. E Zank, W. K. M. Rice, J. A. le Roux, J. Y. Lu, and H. R. Mtiller. The Injection Problem for Anomalous Cosmic Rays. In R. A. Mewaldt, J. R. Jokipii, M. A. Lee, E. M6bius, and T. H. Zurbuchen, editors, AlP ConfProc.: Acceleration and Transport of Energetic Particles Observed in the Heliosphere: ACE 2000 Symposium, volume 528, pages 317324, 2000.

-

182-

Cosmic Rays as Messengers from Outside the Inner Heliosphere M.A. Lee a and H. Fichtner b aSpace Science Center, EOS, Morse Hall, University of New Hampshire, Durham, NH 03824, USA blnstitut fur Theoretische Physik, Ruhr-Universit~it, 44780 Bochum, Germany The two populations of cosmic rays in the outer heliosphere, galactic and anomalous, are introduced briefly. Their coupling to the solar wind in the process known as solar modulation is described and the transport equation used to describe that process is introduced. Finally new results from the Voyager spacecraft and Interstellar Probe are anticipated, and open questions and challenges are identified. 1. I N T R O D U C T I O N There are two populations of cosmic rays in the outer heliosphere: galactic cosmic rays and anomalous cosmic rays. Both interact strongly with the solar wind. The two populations are shown in Figure 1 which is a schematic diagram of the heliosphere including the Sun, a radially outflowing solar wind with embedded magnetic irregularities, the solar wind termination shock where the solar wind is deflected toward the heliospheric tail (solid circle), the heliopause separating solar and interstellar plasma (inner dashed line), and a possible bow shock where the interstellar plasma flow is decelerated to flow around the heliopause (outer dashed line). The galactic cosmic rays (GCR) are denoted by small dots mostly outside the heliopause (low-energy GCRs with energies _ 300 MeV/nucleon). Outside the heliopause the intensity of lowenergy GCRs should dominate that of higher-energy GCRs since the GCR differential intensity in the several-GeV energy range is proportional to E -2.7. The GCRs originate from outside the heliosphere at supernova shock waves throughout the Galaxy, and also possibly at stellar wind termination shocks, pulsars, or other more exotic objects. They form a sea of energetic particles throughout the Galaxy with a pressure comparable to that of the interstellar plasma and magnetic field. They are partially expelled from the heliosphere by the "sweeping" action of the solar wind in the process known as solar modulation. As shown in Figure 1, the low energy GCRs barely penetrate the heliosheath, that region of space between the termination shock and the heliopause, due to modulation effects. The high energy GCRs penetrate all the way to Earth's orbit and bombard the upper atmosphere, creating the "showers" of nuclear interactions and secondary particles which led to the discovery of cosmic rays and the existence of a modulating environment about the Sun now known as the heliosphere. The so-called anomalous cosmic rays (ACR) are denoted by small dots mostly within the heliopause in Figure 1. The ACRs should perhaps better be called heliospheric cosmic rays since they owe their existence to the heliosphere. The ACRs originate at the solar wind

-

183-

M.A. Lee and H. Fichmer

Figure 1. A schematic diagram of the heliosphere showing the solar wind termination shock (solid line), heliopause (inner dashed line), a possible interstellar bow shock (outer dashed line), anomalous cosmic rays (small dots extending out to the heliopause), high-energy galactic cosmic rays (large dots), and low-energy galactic cosmic rays (small dots outside the heliopause). termination shock in the process of diffusive shock acceleration. They are swept into the heliosheath by the subsonic solar wind flow and may leak across the heliospheric magnetic field at the heliopause into interstellar space. Due to their lower energies of > IVI and that the cosmic ray anisotropy be small, conditions which are readily satisfied with the exception of very low energy ACRs at the termination shock. According to equation (1), f (p, r, t) is insensitive to structures smaller than the scattering mean free path, on the order of several particle gyroradii. The most controversial quantity in equation (1), because it is so difficult to specify, is K, which depends crucially on the behavior of the magnetic field and its fluctuations. Magnetic field-line "braiding" and "random walk" affect the components of K perpendicular to the average field but are only understood qualitatively. The drift velocity VD depends on cosmic ray charge and produces a circulation of cosmic rays in latitude through the heliosphere and along the termination shock. Generally the modulation of electrons and ions show the expected symmetry due to drift, but puzzles remain. The most helpful measurement in this regard would be a comparison of the modulation of electrons and positrons which differ only in the sign of their charge. Terms in equation (1) describing stochastic acceleration by solar wind turbulence and cosmic ray viscosity are small and have been neglected. Although equation (1) is linear in its simplest form in which the solar wind is not modified by cosmic rays, realistic geometries require numerical solutions, while analytical approximations such as the "force-field" solution provide guides for intuition. 3. C O S M I C R A Y S A N D T H E H E L I O S P H E R E : F U T U R E R E S E A R C H

In spite of having a trustworthy transport equation at our disposal, many puzzles remain concerning the transport of cosmic rays, particularly in the outer heliosphere. Voyager and Interstellar Probe should resolve many of these as they traverse previously unexplored regions of space. When the Voyager spacecraft encounters the termination shock in the next few years

- 185-

M.A. Lee and H. Fichmer

it may reveal how pickup ions are injected into the process of shock acceleration, and whether any solar wind ions are injected. Voyager should observe whether the pressure of ACRs modifies the plasma structure of the termination shock. It should also determine whether the nature of modulation is different in the heliosheath where V 9V -- 0 and VD may be reduced. The Interstellar Probe will reveal the nature of the heliopause: What is the nature of the magnetic field adjacent to the heliopause? Do the ACRs escape across the heliopause into the interstellar medium? How effectively do the low-energy GCRs penetrate the heliosheath? Does the pressure gradient of either species modify the flows adjacent to the heliopause? Can ACRs and GCRs be distinguished in this region? Beyond the heliopause is a new frontier: What is the energy spectrum and composition of low-energy GCRs? What is the role of these low-energy cosmic rays in the state of the local interstellar medium? Do escaping ACRs "pollute" this spectrum? What is the interstellar magnetic field strength and does its direction agree with that inferred from the cosmic ray anisotropy observed at -105 GeV? Can low-energy GCR nuclides be detected which reveal new information on the cosmic-ray life-cycle? A half century ago cosmic rays first revealed the existence of a huge volume of plasma surrounding, and controlled by, the Sun, later called the heliosphere. Now we are on a journey to the edges of the heliosphere to reveal for the first time the low-energy cosmic rays in interstellar space. The authors wish to thank the speakers and poster presenters in the cosmic ray session, and the Chair, I. Lerche, for an exciting session which stimulated much discussion. In particular we are grateful to R. Leske for agreeing to present an invited talk on short notice. M.A.L. is very grateful to the organizers for excellent organization of the Colloquium and an outstanding choice of venue in Potsdam. M.A.L.'s portion of this work was supported, in part, by NSF Grant ATM-0091527.

Acknowledgments.

REFERENCES 1. L.J. Gleeson and W.I. Axford, Cosmic rays in the interplanetary medium, Astrophys. J.,

149 (1967) Ll15. 2. J.R. Jokipii, E.H. Levy, and W.B. Hubbard, Effects of particle drift on cosmic ray transport, 1, General properties, applications to solar modulation, Astrophys. J., 213 (1977) 861. 3. E.N. Parker, The passage of energetic charged particles through interplanetary space, Planet. Space Sci., 13 (1965) 9.

- 186-

Modulation region of galactic cosmic rays in the heliosphere" Estimation of dimension, radial diffusion coefficient, intensity out of region L.I. Dorman 1,2 1israel Cosmic Ray Center and Emilio Segre' Observatory, affiliated to Tel Aviv University, Technion and Israel Space Agency; P.O. Box 2217, Qazrin 12900, ISRAEL 2 Cosmic Ray Department of IZMIRAN, Troitsk 142092, Moscow region, RUSSIA ABSTRACT On the basis of neutron monitor data as well as solar activity data for four solar cycles 19-22 the hysteresis phenomenon is investigated by taking into account convection-diffusion and drifts, and estimated the dimension of modulation region, radial diffusion coefficient, expected cosmic ray intensity in the interstellar medium and absolute level of galactic cosmic ray modulation observed on the Earth's orbit. It was estimated that the average time of particle propagation from the boundary of modulation region to the Earth's orbit is much smaller than characteristic time-lag in hysteresis phenomenon, what supports the using in this research of quasi-stationary model of high-energy particle modulation in the Heliosphere. INTRODUCTION A short historical introduction to the research of the hysteresis phenomenon between longterm variations of cosmic ray (CR) intensity observed at Earth with solar activity (SA) is given in Dorman et al. (2001). Analysis made by Dorman (2001) leads to the conclusion that observed long-term CR modulation is caused by two processes: a convection-diffusion mechanism (see discussion and references in Dorman et al., 2001) that does not depend on the sign of the solar magnetic field (SMF), and a drift mechanism (e.g. Burger and Potgieter, 1999) what gives opposite effects with the changing sign of the SMF. In the present paper we try to determine the relative role of convection-diffusion and drifts for neutron monitor (NM) data in solar cycles 19-22. We will determine radial diffusion coefficient and transport path in radial direction, expected CR intensity in the interstellar medium and absolute level of galactic CR modulation. COSMIC RAY L O N G - T E R M VARIATION CAUSED BY C O N V E C T I O N - D I F F U S I O N Because the basic quasi-stationary model of CR-SA hysteresis phenomenon was described in details in Dorman et al. (2001), we give here only the final equations what will be used in our research. According to this model the expected CR long-term modulation at the Earth's orbit is:

ln(n(R,Xo,rE,t)exp)= A-Bx F(t,Xo,W(t-XIXXE), where

-

187-

(1)

L.I. Dorman 1 2(1 W(t_X)/Wmax) X -+( X~ )= ?(W(t-X)/Wmax) 3 3 dX, (2) F t, Xo,W(t- X I X E XE X = r / U , X E = I A U / u , X o=r o/u (X o is in units of av. month =(365.25/12) days

=2.628x106 s),

n(R, Xo,rE,t)exp is expected galactic CR density on the Earth's orbit in

dependence of the value of parameter

X o . Here r is the distance from the Sun, ro is the

expected radius of modulation region, R is the effective rigidity of detected CR particles; W is the monthly sunspot number (or some other parameter of SA), and Wmax is sunspot number in the maximum of SA. Coefficient A determines the CR intensity out of the modulation region, and coefficient B determines radial diffusion coefficient in period of maximum SA. COSMIC RAY L O N G - T E R M VARIATION CAUSED BY DRIFTS According to the main results of the drift mechanism approach to the CR long-term variation (e.g. Burger and Potgieter, 1999), we assume that the drifts are determined mainly by tilt angle T as parabola with 0 points at 15 ~ and 90 ~ and changed sign during periods of the SMF polarity reversal. We used data on tilt-angles from Internet for the period from May 1976 up to June 2000. On the basis of these data we determined the correlation between T and W for 11 month-smoothed data as T : 0.354W + 14.94 ~ (3) with correlation coefficient 0.938. For the period January 1953-April 1976 we determined tilt angle T approximately by Eq. (3) on the basis of 11 month-smoothed W data. Information on SMF polarity reversal periods we used from ftp://ftp.ngdc.noaa.gov/STP/SOLAR_DATA) ANALYSIS ACCORDING TO CLIMAX AND H U A N C A Y O / H A L E A K A L A NM DATA According to the procedure described above we correct the l 1-month-smoothed data of Climax NM on the drift effect for different values of Adr from 0% (no drift effect) up to 4%. The dependence of the correlation coefficient on the value of each value of

Adr

in Figure 1,

Xomax(Adr) can

X o is shown in Figure 1. For

be easy determined when the correlation

coefficient reaches a maximum value Rmax . The functions Rmax (Adr) and

X o max (Adr) are

shown in Figure 2. The function Rmax (Adr) can be approximated with correlation coefficient 0.9994__.0.0002 by parabola

Rmax (Adr )= aA2r + bAdr + c ,

(4)

where a = 0.00472__.0.00006, b = -0.00241+_0.0003, and c = -0.919__.0.011. From Eq. (4) we can determine Adr max when Rmax reaches the biggest value:

Adr max = - b / 2 a = (2.55 _ 0.06)%. The function X o max (Adr) can be

(5) approximated

with

correlation

coefficient

0.99982+__0.00007 by Xo max (Adr)- -(0.367 +_O.OO2)Adr + (16.35 __.0.11) av.month,

(6)

what gives X o max (Adr max )= 15.42 __.0.20 av.month.

(7)

-

188-

Modulation region of galactic cosmic rays in the heliosphere... For Huancayo NM (12S, 75W; height 3.4 km, cut-off rigidity 12.92 GV)/Haleakala NM (20N, 156W; 3.03 km, 12.91 GV) data we found by the same procedure that Adr max = (0.41+_ 0.01)%, X o max (Adr max )= 15.39 _ 0.19 av.month (8)

5w

/ Figure 1. Dependences of the correlation coefficient from X o for different Adr from 0% (no drifts) up to 4% according to Climax NM data (N39, W106; height 3.4 km, cut-off rigidity 2.99 GV) in cycles 19-22.

16

'1

~-

I,LI

e'-

-0.94

E

"~ I11 n,' -0.925 O O

Figure 2. Functions Rmax (Adr) and X o max (Adr) for 19-22 cycles according to Climax NM data.

-

189-

L.L Dorman MAIN RESULTS AND DISCUSSION According to Eq. (7) and (8) the time of solar wind moving from the Sun to the boundary of modulation region is 15.4+_0.2 av. month what gives for the dimension of modulation region ro = u x X o max = 119.0 t 1.5 A U (9) (according to direct measurements on space probes the average solar wind speed for the period 1965-1990 was u=4.41x107cm/s, therefore av. month corresponds to 7.73 A U). Determination of parameters A and B in Eq. (1) gives important possibility to estimate CR intensity out of the modulation region (A = ln(n o (R))) and the radial diffusion coefficient in the maximum of SA (DrmaX(R)= 1.5u2/B ). We can determine regression coefficients A and B only for integer values of X o (because we used monthly data). Therefore, for example, for Climax NM we determine A and B at Adr--2.5% according to Eq. (5), for Xo=15 and 16, and then

by

extrapolation

for

Xomax =15.4__.0.2 what

gives

A=8.37084+0.00009,

B=(0.0132+_0.0002) (av. month) -1 . It means that for Climax NM (10-15 GV primary CR)

In(no(R)) = 8.37084 +_0.00009, Drmax = (5.81__. 0.09)x 1023

cm2/s.

(10)

The transport path and time diffusion of galactic CR 10-15 GV from the boundary of modulation region to the Earth's orbit in maximum SA can be estimated approximately as max (R) ~" r 2 / d D max (R) ~ 0.35 av.month 9 -'rAmax(R)= (5.8 +_0.1)x 1013 cm, r die

(11)

Estimated TdmFC(R) is much smaller than time-lag between CR and SA variations (this supports the using for NM energies of quasi-stationary model of CR modulation in the Heliosphere). By the same way for Huancayo/Haleakala NM data (primary CR 30-40 GV) at Adr--0.4% according to Eq. (8), we obtain: ln(n o (R))= 7.46635 -+ 0.00004, Drmax = (2.19 + max = 0.092 av. month (12) - 0.06)1024 cm2/s ' Tdif The change of radial diffusion coefficient with SA can be described according to Eq. (2) by

D r (R,W)= Drmax (R)x (W/Wma x )-(1/3)-(2/3)(1-W/Wmax ) .

(13)

Now we can determine the absolute CR modulation relative to the intensity in the interstellar space. For example, for Climax NM the absolute decreases were on 23% in February 1958 (cycle 19), on 18% in May 1969 (cycle 20), on 25% in August 1979 (cycle 21), and on 34% in June 1991 (cycle 22). The relative role of drifts for cycles 19-22 for 10-15 GV galactic CR was 2x2.5%/25% - 0.2 and role of convection-diffusion about 0.8. For 30-40 GV galactic CR the relative role of drifts was about two times smaller: 2x0.4%/8% ~- 0.1 and role of convectiondiffusion about 0.9. R E F E R E N C E S

1. R. A., Burger, and M. S. Potgieter, The effect of large heliospheric current sheet tilt angles in numerical modulation models: A theoretical assessment, Proc. 26th Inter. Cosmic Ray Conf., 7, 13-16, 1999. 2. L. I. Dorman, Cosmic ray long-term variation: even-odd cycle effect, role of drifts, and the onset of cycle 23,Adv. Space Res., Paper D1.1-0037, 2001 (in press). 3. L. I. Dorman, N. Iucci, and G. Villoresi, Time lag between cosmic rays and solar activity in solar minimum 1994-1996 and the residual modulation, Adv. Space Res., Paper D1.1-0034, 2001 (in press).

-

190-

Latitudinal gradients and charge sign dependent modulation of galactic cosmic rays B. Heber ~, R Ferrando b, C. Paizis c R. Mtiller-Mellin d, H. Kunow d, M.S. Potgieter e, S.E.S. Ferreira ~, and R.A. Burger e ~Max-Planck-Institut ftir Aeronomie, Germany bCEA/Saclay, Service d'Astrophysique, France ~Universita di Milano, Italy aIEAP, University of Kiel, Germany ePotchefstroom University for CHE, South Africa Ulysses, launched in October 1990, began in December 1997 its second out-of-ecliptic orbit. The solar activity is rising to maximum conditions for this second orbit, whereas it was declining to low activity for the first out-of-ecliptic orbit. According to drift-dominated modulation models, the intensity of galactic cosmic rays depends on the latitudinal extension of the heliospheric current sheet (HCS). The latitudinal gradient as well as the charge sign dependent variation of 2.5 GV protons and electrons observed during the previous Ulysses orbit can be described by such models. In this paper we investigate these two parameters along the Ulysses maximum orbit. The electron-to-proton ratio and the latitudinal gradient are qualitatively in agreement with the model predictions up to fall 1999. Although Ulysses was then moving to higher southern latitudes, the latitudinal gradient is significantly reduced, and the electron-toproton ratio does not depend on the latitudinal extension of the HCS until June 2000, when the electron-to-proton ratio increased again. In both cases, we suggest that these changes are correlated with the reconfiguration of the heliospheric magnetic field. 1. INTRODUCTION The intensity of galactic cosmic rays (GCRs), energetic charged particles entering the heliosphere, is modulated as they traverse the turbulent magnetic field embedded in the solar wind. These particles are scattered by irregularities in the interplanetary magnetic field and undergo convection and adiabatic deceleration in the expanding solar wind. The large-scale heliospheric magnetic field (HMF), approximated by an Archimedean spiral (Parker, 1965), [7] leads to gradient and curvature drifts of cosmic rays in the interplanetary medium (Jokipii et al., 1977) [6]. The latter causes the GCR flux to vary with the 22-year solar magnetic cycle: 1. In an A0 solar magnetic cycle. Electrons are expected to behave oppositely. 2. Latitudinal and radial gradients are ex-

-191 -

B. Heber et al.

r [AU]

1.4

. . . .

t

. . . .

F . . . .

i

. . . .

~ . . . .

i

. . . .

f . . . .

i . . . .

i . . . .

o [o]

i -5

I -20

I -35

I i - ~ - 7 0

li 7060

I 30

i 0

I -15

i

I

-30-45

A+ 1.2

e/p at 1 GV

r

0.8

0.6

0.4

. . . .

i

10

. . . .

i

20

. . . .

i

30

. . . .

I

40

. . . .

i

. . . .

50

i . . . .

60

J

70

. . . .

i , , , ,

80

10

90

1990 1992 1994 1996 1998 2000

et [degree]

Year

Figure 1. Calculated electron-to-proton ratio at 1 GV as function of the maximum latitudinal extent (oz) of the heliospheric current sheet.

Figure 2. 52-day averaged quiet time count rates of > 100 MeV protons at Ulysses (a), at Earth (c) and 2.5 GV electrons from KET on Ulysses (b).

pected to be different for oppositely charged particles. Both the charge sign dependent time profiles as well as the positive and negative latitudinal gradients for GCR nuclei have been observed around solar minimum (Heber and Potgieter, 2000 [3], and references therein). Recently Burger and Potgieter (1999) [1] investigated the proton and electron intensities as a function of the maximum latitudinal extent, c~, of the Heliospheric Current Sheet (HCS). The electron-to-proton ratio as a function of o~ for 1 GV particles during a full A>0 solar cycle is shown in Fig. 1. Burger and Potgieter (1999) [1] found three regimes in the electronto-proton ratio: 1. Below 40 ~ the electron-to-proton ratio is decreasing with increasing oz, 2. between 40 ~ and 80 ~ it is nearly constant, and 3. it is increasing towards the non drift value between 80 ~ and 90 ~ Heber et al. (2000) [4] suggest that drifts are globally important as long as the HMF has a basic, ordered configuration during increased solar activity. In this paper we extend their investigation to later time periods, including the period when the solar magnetic field polarity is reversing. 2. INSTRUMENTATION AND OBSERVATIONS Ulysses, launched on October 6, 1990, encountered in February 1992 the planet Jupiter, and, using a Jupiter gravity assist, began its journey out of the ecliptic plane. The observations presented here were made with the Kiel Electron Telescope (KET) on board Ulysses and the University of Chicago (UoC) instrument on board IMP8 at 1 AU. Fig. 2 displays the 52-day averaged quiet time count rate of > 100 MeV protons measured by KET (a) and the UoC IMP (c) particle sensor. Quiet time fluxes have been defined by analyzing the ~50 MeV proton channels, as described in Heber et al. (1995) [5]. These two measurements allow us to separate spatial from temporal variations in the GCR flux. The intensity time profiles at both locations are very similar. Differences in these profiles due to Ulysses' trajectory can be seen in the figure around solar minimum from 1994 to late 1996; for a detailed discussion see Heber and Potgieter (2000) [3]. Curve (b) in Fig. 2 displays

-

192-

Lat#udinal gradients and charge sign dependent modulation of galactic cosmic rays

Figure 3. Heliographic equator equivalent electron-toproton ratio as a function of the time shifted c~.

Figure 4. Source surface fields in April 1999 (a) and June 2000 (c) (for details see text).

the count rate of 2.5 GV electrons from KET on Ulysses. Comparing curves (a) and (b) it was found that GCR protons and electrons have different latitudinal gradients, in agreement with modulation models. Since the gradient of 2.5 GV electrons is consistent with zero (for a detailed discussion see Heber et al., 1999, [2]) and the > 100 MeV protons have nearly the same rigidity (~1.7 GV), it follows by comparing curves (b) and (c) in 1996 and 1997 that electrons respond differently to changes in c~ than protons (Heber, et al. 1999) [2]. In what follows we discuss the temporal evolution of the electron-to-proton ratio from 1998 on. Since no new IMP data are available, we will refer to the results given in Heber et al. (2000) [4]. 3. D I S C U S S I O N AND C O N C L U S I O N S During the present approach to solar maximum and with Ulysses at high southern heliographic latitudes, it is possible to determine simultaneously the electron and proton count rate ratio and the latitudinal gradient for GCR protons. Since cosmic rays do not respond immediately to a change in c~, but with a certain delay, we shift the calculated c~ at the source surface by 5 solar rotations to later times (cts). Fig. 3 displays the (heliographic equator equivalent) electron-to-proton ratio as a function of c~s, as described in Heber et al. (2000) [4]. The (heliographic equator equivalent) electron-to-proton ratio is calculated from the measured electron and proton count rates by correcting the protons only for the observed latitudinal variation. The corresponding correction factor has been obtained from the latitudinal variation during the fast latitude scan in 1994 and 1995. In contrast to protons electrons do not show any significant latitudinal gradient (see Heber and Potgieter, 2000, and references therein) [3], so that no corrections to the electrons have been applied. The filled and open dots are the ratios during the decreasing and rising phase of the solar cycle. (a) and (c) correspond to two time periods where the source surface maps are shown in Fig. 4. These two maps in April 1999 and June 2000 were obtained from http://quake.stanford.edu/~wso/and

-

193-

B. Heber et al. calculated by using the "radial" boundary conditions at 3.25 solar radii. Within these 10 months the source field changed from a two-sector structure with the northern and southern polarity in the northern and southern hemisphere to a structure with a predominantly southern and northern polarity in the northern and southern hemisphere. The structure in summer 1999 during the time period (b) in Fig. 3, not shown here, was characterized by a more complex behaviour. Therefore the HMF may rapidly change from a relatively ordered (dipole-dominated) configuration to one where higher order components become equally important (Wang et al., 2000) [8] during increased solar activity. As discussed in Heber et al. (2000) [4], the latitudinal gradient before fall 1999 (a) is consistent with a mean latitudinal gradient of Go ~ 0.23 %/AU at solar minimum and drops to low values during period (b). The fact that in April 1999 the observed latitudinal gradient is still consistent with the one measured around solar minimum, suggests that drifts are still important. During the second period (b) the HMF pattern becomes much less regular. At the same time the latitudinal gradient drops and the electron-to-proton ratio reaches a constant value for 10 months. Given that drift motions have a net resulting effect on modulation when extending over large (global) scales, one can argue that when the HMF configuration progressively develops from a well ordered (dipole-like) configuration to a "fragmented" pattern, global drift motions are becoming progressively less important (phase out). It is important to note that this increase in the electron-to-proton ratio is not necessarily indicative of large drift effects, but rather an indication of how drifts phase out for electrons and protons. The newest data (c) indicates a further increase in the electron-to-proton ratio. Such an increase is expected in our "model" when a better ordered HMF configuration has established itself during an A < 0 epoch. The lower panel in Fig. 4 indicates that during this period the solar magnetic field is indeed reversing. Simultaneously, a change from a positive to a negative latitudinal gradient for protons should also be measured, i.e., the GCR proton flux should become lower at high latitudes than at Earth. However, it is difficult to predict how long the magnetic field structure will keep this pattern during extreme solar activity. It might have changed again with the extensive activity in July 2000. Within the following year, with Ulysses at high heliographic latitudes and the possibility to determine the intensity time profiles for protons and electrons we will make progress on how the global configuration of the HMF influences the modulation of GCRs. REFERENCES 1.

2.

3. 4. 5. 6. 7. 8.

R.A. Burger and M. S. Potgieter. In Proc. 26th ICRC, 7, 13, 1999. B. Heber, E Ferrando, A. Raviart, G. Wibberenz, R. Mtiller-Mellin, H. Kunow, H. Sierks, V. Bothmer, A. Posner, C. Paizis, and M. S. Potgieter. Geophys. Res. Lett., 26(14):2133, 1999. B. Heber, and M. S. Potgieter. Adv. Space Res., 26(5):839, 2000. B. Heber, M.S. Potgieter, R. A. Burger, G. Wibberenz, R. Mtiller-Mellin, H. Kunow, E Ferrando, A. Raviart, C. Paizis, and C. Lopate. J. Geophys. Res., 2000. submitted. B. Heber, A. Raviart, C. Paizis, W. Dr6ge, R. Ducros, E Ferrando, H. Kunow, R. MtillerMellin, C. Rastoin, K. R6hrs, H. Sierks, and G. Wibberenz. Adv. Space Res., 16, 1995. J.R. Jokipii, E. H. Levy, and W. B. Hubbard. Ap. J., 213:861-868, 1977. E.N. Parker. Planet. Space Sci., 13:9-49, 1965. Y.-M. Wang, N. R. Sheeley, and N. B. Rich. Geophys. Res. Lett., 27(2): 149-152, 2000.

- 194-

Antiprotons below 200 MeV in the interstellar medium" perspectives for observing exotic matter signatures I.V. Moskalenko a-t, E.R. Christian ~, A.A. Moiseev ~, J.F. Ormes a, and A.W. Strong b aNASA/Goddard Space Flight Center, Greenbelt, MD 20771, USA bMax-Planck-Institut ffir extraterrestrische Physik, 85741 Garching, Germany

Most cosmic ray antiprotons observed near the Earth are secondaries produced in collisions of energetic cosmic ray (CR) particles with interstellar gas. The spectrum of secondary antiprotons is expected to peak at ~ 2 GeV and decrease sharply at lower energies. This leaves a low energy window in which to look for signatures of exotic processes such as evaporation of primordial black holes or dark matter annihilation. In the inner heliosphere, however, modulation of CRs by the solar wind makes analysis difficult. Detecting these antiprotons outside the heliosphere on an interstellar probe removes most of the complications of modulation. We present a new calculation of the expected secondary antiproton flux (the background) as well as a preliminary design of a light-weight, low-power instrument for the interstellar probe to make such measurements.

1. I N T R O D U C T I O N The nature and properties of the dark matter that may constitute up to 70% of the mass of the Universe has puzzled scientists for decades, e.g. see [1]. It may be in the form of weakly interacting massive particles (WIMPs), cold baryonic matter, or primordial black holes (PBHs). It is widely believed that the dark matter may manifest itself through annihilation (WIMPs) or evaporation (PBHs) into well-known stable particles. The problem, however, arises from the fact that a weak signal should be discriminated from an enormous cosmic background, including a flux of all known nuclei, electrons, and v-rays. Antiproton measurements in the interstellar space could provide an opportunity to detect a signature of such dark matter (see [2] and references therein). High energy collisions of CR particles with interstellar gas are believed to be the mechanism producing the majority of CR antiprotons. Due to the kinematics of the process they are created with a nonzero momentum providing a low-energy "window" where exotic signals can be found. It is therefore important to know accurately the background flux of interstellar secondary antiprotons and to make such measurements outside the heliosphere to avoid any uncertainties due to solar modulation. * N R C Senior Research Associate talso Institute of Nuclear Physics, M.V.Lomonosov Moscow State University, Moscow 119899, Russia

-

195-

I. V. Moskalenko et al.

Figure 1. L e f t : Calculated proton interstellar spectrum (LIS) and modulated spectrum (~ = 750 MV) with data [8-11]. R i g h t : Calculated i3 interstellar spectrum (LIS) and modulated spectrum (~ = 400 MV) with data [12,13]. The top curves are the total background. The "tertiary" components (LIS and modulated) are shown separately. Also shown is an example exotic signal [14] extended to lower energies.

2. C A L C U L A T I O N S

OF T H E A N T I P R O T O N

BACKGROUND

made a new calculation of the CR/3 flux in our model (GALPROP) which aims to reproduce observational data of many different kinds: direct measurements of nuclei, p's, e+'s, "),-rays and synchrotron radiation [3-5]. The model was significantly improved, and entirely rewritten in C + + . The improvements involve various optimizations relative to our older FORTRAN version, plus treatment of full reaction networks, an extensive cross-section database and associated fitting functions, and the optional extension to propagation on a full 3D grid. For this calculation, we used a cylindrically symmetrical geometry with parameters that have been tuned to reproduce observational data [5]. The propagation parameters including diffusive reacceleration have been fixed using boron/carbon and l~ ratios. The injection spectrum was chosen to reproduce local CR measurements, ~ flp-2.3s where/3 is the particle speed, and p is the rigidity. The parameters used" the diffusion coefficient, D ~ - 4.6 • 1028/3(p/3 GV) ~/3 cm 2 S-1, Alfven speed, VA -- 24 km s -1, and the halo size, Zh -- 4 kpc. We calculate iv production and propagation using the basic formalism described in [3]. To this we have added i0 annihilation and treated inelastically scattered/3's as a separate "tertiary" component (see [6] for the cross sections). The i0 production by nuclei with Z > 2 is calculated in two ways: employing scaling factors [3], and using effective factors given by Simon et al. [7], who make use of the DTUNUC code, and which appears to be more accurate than simple scaling. (The use of Simon et al. factors is consistent since We

- 196-

Antiprotons below 200 Me V in the interstellar medium: ...

their adopted proton spectrum resembles our spectrum above the f production threshold.) The effect on the p flux at low energies is however small, and the two approaches differ by about 15%. We believe our calculation is the most accurate so far since we used a self-consistent propagation model and the most accurate production cross sections [7]. The results are shown in Figure 1. The upper curves are the local interstellar flux (LIS) and the lower are modulated using the force-field approximation. The two lowest curves in Figure 1 (right) show separately the contribution of "tertiary" ffs, which is the dominant component at low energies. The adopted nucleon injection spectrum, after propagation, matches the local one. There remains some excess of ffs. The excess for the lowest energy points is at the 1 a level. Many new and accurate data on CR nuclei, diffuse gamma rays, and Galactic structure have appeared in the last decade; this allows us to constrain propagation parameters so that the limiting factor now becomes the isotopic and particle production cross sections. At this point we cannot see how to increase the predicted intensity unless we adopt a harder nucleon spectrum at the source in contradiction with constraints from high energy p data [3,5]. More details will be given in a subsequent paper. 3. A N T I P R O T O N

DETECTOR

Very limited weight and power will be available for any experiment on board an interstellar probe [15]. We therefore propose a simple instrument which is designed to satisfy these strict constraints [2]. We base our design (Figure 2 left) on a cube of heavy scintillator (bismuth germanium oxide [BGO]), with mass of the order of 1.5 kg. The cube, 42 g cm -2 thick, will stop antiprotons and protons of energy below 250 MeV. A time-of-flight (TOF) system is used to select low-energy particles. The particles with energy less than 50 MeV will not penetrate to the BGO crystal through the T O F counters, setting the low-energy limit. The separation of antiprotons from protons is the most challenging aspect of the design. A low-energy proton (below 250 MeV) that would pass the T O F selections cannot deposit more than its total kinetic energy in the block. Therefore an event will be required to deposit > 300 MeV to be considered an antiproton. A proton can deposit comparable energy in this amount of material only through hadronic interaction, which our Monte Carlo simulations show requires a proton with energy > 500 MeV. The T O F system can effectively separate low-energy particles (< 250 MeV) from such protons and heavier nuclei. As a conservative estimate, we assume that all protons with energy > 500 MeV have the potential to create a background of "p-like" events and their integral flux in interstellar space is somewhat uncertain but would be ~ 1 cm -2 s -~ sr -~. The exotic p signal, to be significantly detected above the background, should be of the order of 10 -6 cm -2 s -~ sr -1 in the energy interval 50-200 MeV, which corresponds to an expected signal/p-background ratio of ~ 10. We thus can allow only one false antiproton in 107 protons. Simulations to date indicate that the current design will have rejection power of 2 • 106. We expect to get the next factor of five by fine-tuning the design and selections. The efficiency of antiproton selection is shown in Figure 2 (right). The antiproton rate will be 0.1-1 particle per day, and the statistical accuracy will be ~ 10% after 3 years of

-

197-

I.V. Moskalenko et aL

~TOF1

Incident

O

.,..~

antiproton

o

0.9

...............................................................................................................................

,.. . . . . . .

0.8

...............................................................................................................................

i ...........

~=

0.7

................................................................................

oo

0.6

,.-w

TOF2 I

'1 . . . . . .

I

~l

+

~

+

~ ............. i .......................... i........... i

TOF3 .~..~

-~

~

0.5

+

~.~

+

.o TOF4

~

. . . . .

0.3

.....................................................................................................................................................

0.2

......................................................................................................... i .............. !-...........................

m

B I N ) crystal 0.l

......................................................................................................

0

. . . . . . . . . . . . . .

0

0.025

0.05

'

. . . . . . . . .

0.075

0.1

i

. . . . . . . . .

0.125

0.15

i+

. .......................................

;

. . . .

0.175

;

0.2

. . . .

i

. . . .

0.225

0.25

Energy of incident antiproton, GeV

5 cm

Figure 2. Left: Preliminary design of the detector. detection after selections applied.

Right: efficiency for antiproton

observation. I.V.M. acknowledges support from NAS/NRC Senior Research Associateship Program. REFERENCES ,

2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

Trimble, Ann. Rev. Astron. Astrophys. 25 (1987) 525 J.D. Wells, A. Moiseev, J.F. Ormes, Astrophys. J. 518 (1999) 570 I.V. Moskalenko, A.W. Strong, O. Reimer, Astron. Astrophys. 338 (1998) L75 A.W. Strong, I.V. Moskalenko, Astrophys. J. 509 (1998) 212 A.W. Strong, I.V. Moskalenko, O. Reimer, Astrophys. J. 537 (2000) 763 L.C. Tan, L.K. Ng, J. Phys. G: Nucl. Part. Phys. 9 (1983) 227 M. Simon, A. Molnar, S. Roesler, Astrophys. J. 499 (1998) 250 W. Menn et al., Astrophys. J. 533 (2000) 281 M. Boezio et al., Astrophys. J. 518 (2000) 457 T. Sanuki et al., Astrophys. J. 545 (2000) 1135 I.P. Ivanenko et al., Proc. 23rd ICRC (Calgary) 2 (1993) 17 S. Orito et al., Phys. Rev. Lett. 84 (2000) 1078 S.J. Stochaj et al., Astrophys. J. (2001) in press L. BergstrSm, J. EdsjS, P. Ullio, Astrophys. J. 526 (1999) 215 R.A. Mewaldt, P.C. Liewer, These Proceedings

V.

- 198-

Anomalous Cosmic Rays outside of the Termination Shock A. Czechowski ~ S. Grzedzielski ~' b H. Fichtner c M. Hilchenbach d and K.C. Hsieh e ~Space Research Centre, Polish Academy of Sciences, Bartycka 18A, PL 00-716 Warsaw, Poland bService d'Aeronomie du CNRS, B.P. No 3, Verrieres le Buisson, F-91371, France CInstitut fiir Theoretische Physik IV, Ruhr-Universit/it Bochum, 44780 Bochum, Germany dMax-Planck-Institut fiir Aeronomie, D-37191 Katlenburg-Lindau, Germany ~Physics Department, University of Arizona, Tucson AZ 85721, U.S.A. Anomalous Cosmic Rays (ACR) are most probably generated at the solar wind termination shock. When observed in the region inside the shock, the ACR energy spectrum is significantly modulated. The region outside the shock is so far inaccessible to observations. However, the low energy (~ 100 keV) part of the spectrum beyond the shock could be sampled by means of energetic neutral atoms (ENA), into which the ACR ions convert by charge-exchange with the neutral atoms. As the distribution of ACR reflects the large-scale structure of the heliosphere, this offers a possibility of imaging the outer heliosphere by ENA from this source. 1. I N T R O D U C T I O N The possibility of observing energetic neutral atoms (ENA) of heliospheric origin was first discussed by Hsieh et al. [1]. Their discussion included also the ACR ENA, the anomalous cosmic ray (ACR) ions which become neutralized by charge-exchange with the atoms of the background. As the low energy ACR ions (with large charge-exchange cross section) cannot penetrate upstream of the solar wind termination shock, the main source of ACR ENA must be the region beyond the shock. Grzedzielski [2] first observed that the ACR distribution in the outer heliosphere should reflect the asymmetry of the heliospheric plasma flow, with the stagnation point and extended heliotail. Assuming simple models of the heliospheric plasma flow (Parker [3]), the ACR spatial distribution beyond the termination shock was calculated ([4], [5], [6]). Because of convection by plasma flow and the diffusion across the heliopause, the ACR were found to concentrate in the region of the heliotail, so that the ACR ENA flux would have a maximum from the heliotail (anti-apex of the LISM, the local interstellar medium) direction. The observations by CELIAS/HSTOF on SOHO during 1996 and 1997 [7] were consistent with this conclusion: the flux of mass=l 55-80 keV neutral particles detected by the instrument peaked during the periods when the instrument field-of-view included the LISM anti-apex.

-

199-

A. C z e c h o w s k i et aL

'

'

'

'

'

'

'

'

'

'

"

' .....

kousch' Porker

'

'

'

....... ""'"'

o"

".

" ~ "

o*

~E l'Oxl 0 - 4 o

~

Figure 1. ACR density for different models of the heliosphere: Kausch's, Parker's and Kausch's with nH set to 0.1 cm -a as used in Parker's model. The density profiles shown are for 0 = 0 ~ (apex)104 ~ 133 ~ 162 ~ 180 ~

2. A C R D I S T R I B U T I O N

BEYOND

Figure 2. ACR ENA hydrogen flux at 63 keV for the models illustrated in Fig. 1: Kausch's and Parker's. The flux is shown as a function of direction (0 = 0 ~ is the apex direction).

THE SHOCK

The next step ([8],[9]) was to use a more realistic model, based on a numerical solution of the gas-dynamical equations obtained by Kausch [10]. In this model the termination shock is nonspherical (with nonuniform parameters), the flow has nonzero divergence, and the neutral hydrogen from the LISM has the density inside the heliopause reduced (by a factor of 3-4) due to interaction with plasma. All these features affect the transport of ACR. As the simple models used before, Kausch's model is effectively two-dimensional (axially symmetric with respect to the LISM apex-antiapex axis). The ACR distribution function f(x, p, t) is calculated by solving numerically the transport equation (Parker [11]) in the region downstream from the termination shock: O t f -- V . ec. V f - V .

V f nt

1

Of ~ V

301np

. V-

flf

(1)

where V is the velocity of the plasma flow, ~ the particle diffusion tensor, p the absolute value of the particle momentum, and # the loss rate (at ~ 100 keV it is predominantly the charge-exchange loss term #~). The ACR particles are treated in a test-particle approximation. The boundary conditions at the shock prescribe the ACR flux as a function of position and energy. Figures 1 to 4 correspond to the ACR spectrum uniform over the shock (flux ~ E -~42, see [12]). The nonuniform spectrum determined by the local shock parameters was also considered [13]. Most calculations were restricted to the timeindependent case. Also, the solution was assumed to be axially symmetric with respect to the LISM apex-antiapex axis. The diffusion tensor was replaced by a scalar coefficient, equal to ec2 outside and to ~c~ inside the heliopause. As the magnetic field downstream of the shock is expected to have a very complicated form, this assumption, corresponding

-

2 0 0

-

Anomalous cosmic rays outside of the termination shock

> "

0_ 4

m

,,

1

I

Figure 3. Evolution of the ACE energy spectrum downstream from the shock. The dotted lines correspond to omitting the V . V term.

Figure 4. The ENA hydrogen energy spectrum in the anti-apex and apex direction. The ACR spectrum assumed at the shock is also shown.

to the disordered field, may be reasonable as a lowest order approximation between the shock and the heliopause. In most calculations ec2 >> eel was assumed. ~1 was estimated by eel - (1/3)ec I with ec I given by the formula of le Roux et al. [14]. In Fig. 1 the ACR distributions following from the simulations based on the Kausch and Parker models are compared. The density profiles start at the termination shock and have a sharp change in slope at the heliopause (ec2 - 102ec~). With the exception of the heliotail region, the density falls rapidly towards the heliopause, which is approximately the free escape surface. The difference between the models is partly due to different assumptions about the neutral gas density (0.1 cm -a in the calculations based on the Parker model, while Kausch's solution corresponds to 0.02-0.03 cm -a inside the heliopause) but also to the difference in the plasma flows. The difference can be seen both in the ACR density distributions and in the corresponding ACR ENA flux (Fig. 2). The ACR ENA flux has a peak from the heliotail direction (0 - 180~ The shape of the peak is clearly model dependent. Similar results were obtained in the case of nonuniform ACR shock spectrum Figure 3 shows the evolution of the ACR energy spectrum downstream of the shock in the heliotail direction [9]. The spectra at the distances of 187 AU (shock), 275 AU, 380 AU, 600 AU and 990 AU from the Sun are shown. The dotted lines are obtained by neglecting the ~ V - V term in the transport equation. The flow divergence apparently does not affect very strongly the shape of the spectra (the effect is even weaker in the apex direction). The main contribution to modulation of the spectrum is due to chargeexchange processes which reduce the density at low energy. Figure 4 presents the ACR ENA energy spectrum. An important conclusion is that the spectrum, particularly in the heliotail direction, is not as steep as expected for a product of the ACR energy spectrum by the charge-exchange cross section: this is because the fall in the charge-exchange rate reduces the loss rate of ACR, so that the ACR density decreases less rapidly with distance.

-201 -

A. Czechowski et al.

3. D I S C U S S I O N The ACR-generated ENA flux should have a maximum from the heliotail direction (which may be shifted from the anti-apex due to interstellar field, see [15]). The detailed shape of the peak in the ENA flux as a function of direction depends on the model of the heliosphere (Fig. 2). The ACR ENA flux intensity scale is determined by the ACR flux intensity at the source, which (in the test particle approximation) is simply proportional to the ACR flux intensity at the termination shock, a parameter of the model. The estimations of its value rely on the observations of a highly modulated (no low energy part) ACR energy spectra in the solar wind [12]. Extrapolation of the result to the energy range of interest to us (~ 100 keV) is a source of large uncertainty, also because the shape of the ACR spectrum at low energy may deviate from simple power law. The results of model simulations were compared with the CELIAS/HSTOF observations in [8]. One of the problems was the difference in ENA flux intensity between the 1996 and 1997 peaks. Re-evaluation of the CELIAS/HSTOF data (Hilchenbach et al. this conference) changed the situation: the peaks are now of similar height. On the other hand, the new calibration suggests that the absolute ENA intensity may be an order of magnitude higher than reported before, suggesting that the extrapolation of the ACR spectrum used in the ACR ENA calculations underestimates the low energy spectrum at the shock. A.C. acknowledges support from the project KBN 2 P03C 004 14. R E F E R E N C E S

1.

2. 3. 4. 5. 6. 7. 8.

9.

10. 11. 12. 13. 14. 15.

K.C. Hsieh, K.L. Shih, J.R. Jokipii and S. Grzedzielski 1992: ApJ 393, 756 S. Grzedzielski 1993: Adv. Space Res. 13, (6)147 E.N. Parker 1963: Interplanetary Dynamical Processes, Interscience, New York S. Grzedzielski, A. Czechowski and I. Mostafa 1993: Adv. Space Res. 13, (6)261-(6)264 A. Czechowski, S. Grzedzielski and I. Mostafa 1995: A&A 297, 892 A. Czechowski and S. Grzedzielski 1997: in: Proceedings of the 25th International Cosmic Ray Conference, Durban, July 30- August 6 1997, Vol. 2, p. 237 M. Hilchenbach, K.C. Hsieh, D. Hovestadt, B. Klecker, H. Griinwaldt et al. 1998: ApJ 503, 916 A. Czechowski, H. Fichtner, S. Grzedzielski, M. Hilchenbach, K.C. Hsieh, J.R. Jokipii, T. Kausch, J. Kota and A. Shaw 1999: in: Proceedings of the 26th International Cosmic Ray Conference, Salt Lake City, Utah, August 17-25 1999, Vol.7, p. 589-592 A. Czechowski, H. Fichtner and T. Kausch 1999: in: Proceedings of the 26th International Cosmic Ray Conference, Salt Lake City, Utah, August 17-25 1999, Vol.7, p. 523-525 H.J. Fahr, T. Kausch and H. Scherer 2000: A&A 357, 268-282 E.N. Parker 1965: Planet. Space Sci. 13, 9-49 E.C. Stone, A.C. Cummings and W.R. Webber 1996: J. Geophys. Res. 101, 11017 A. Czechowski, H. Fichtner, S. Grzedzielski, M. Hilchenbach, K.C. Hsieh, J.R. Jokipii, T. Kausch, J. Kota and A. Shaw 2001, A&A 368, 622 J.A. le Roux, M.S. Potgieter and V.S. Ptuskin 1996: J. Geophys. Res. 101, 4791 A. Czechowski and S. Grzedzielski 1998: Geophys. Res. Lett. 25, 1855-1858

- 202

-

IONIZING MEDIA AND THE OBSERVED CHARGE STATES OF ANOMALOUS COSMIC RAYS A.F. Barghouty 1, J.R. Jokipii 2, and R.A. Mewaldt 1 1California Institute of Technology, Pasadena, CA 91125, USA 2 University of Arizona, Tucson, AZ 85721, USA ABSTRACT Singly-charged anomalous cosmic rays (ACRs) can give rise to multiply-charged ACR ions when they suffer further ionization during acceleration at or near the solar-wind termination shock. Measurements at 1 AU by SAMPEX have shown that above ~ 25 MeV/nucleon ACR nitrogen, oxygen, and neon ions are multiply charged. These observations have also established that the transition from mostly singly-charged to mostly multiply-charged ACRs occurs at a total kinetic energy of ~ 350 MeV. Recent simulations for ACR oxygen using ambient hydrogen as the only ionizing medium at or near the termination shock are able to successfully model this transition. The simulated oxygen intensity, however, appears deficient at high energies, where multiply-charged ACRs dominate. This paper presents further simulations that now include neutral helium as part of the ionizing medium in addition to the ambient, neutral hydrogen. The inclusion of helium helps reduce the deficiency, but appears to fall short of accounting fully for the observed spectrum. To that end, including heavier neutrals, e.g., oxygen, as well as taking multi-electron stripping into account, are suggested for more realistic modeling of the observed charge states of ACRs.

INTRODUCTION According to the accepted theory of ACRs (Fisk et al., 1974) these cosmic-ray ions are believed to originate as interstellar neutrals that penetrate the heliosphere before getting ionized -either by solar radiation or by charge-exchange collisions with solar-wind protons- to become singly-charged pickup ions. The pickup ions are then convected by the solar wind to the solar-wind termination shock (SWTS) where they are accelerated up to tens of MeV/nucleon via the process of diffusive shock-drift acceleration (Pesses et al., 1981). Unlike galactic cosmic rays or solar energetic particles observed at 1 AU, ACRs are expected to be predominantly singly-charged ions. However, recent observational evidence from SAMPEX (Mewaldt et al., 1996a; Klecker et al., 1998) have shown that ACR nitrogen, oxygen, and neon above ~ 25 MeV/nucleon are multiply charged, with ionic charge states of 2, 3, and higher. At energies below .~ 20 MeV/nucleon most of the observed ACRs are singly charged. SAMPEX observations (Klecker et al., 1998) have further established that the transition from mostly singly-charged to mostly multiply-charged ACRs occurs at a total kinetic energy of ~ 350 MeV. The theory of diffusive shock-drift acceleration (Jokipii, 1996) gives the amount of energy A E an ACR ion gains in drifting along the SWTS (from the equator towards the poles during the 1990s) to be directly proportional to its ionic charge q, i.e., A E ~ 240 (MeV)• As a result, multiply-charged ACRs are accelerated to higher energy per nucleon than the more abundant singly-charged ions. Thus, at energies well above 240 MeV, the theory predicts that multiply-charged ACR ions will dominate if there is a source of such ions to accelerate. The predominance of multiply-charged ACRs at high energy has been interpreted as evidence that some singly-charged ACR ions suffer additional ionization during their acceleration at or near the SWTS (Mewaldt et al., 1996b; Jokipii 1996). Recent simulations (Barghouty et al., 2000) of ACR oxygen using ambient hydrogen as the only ionizing medium at or near the termination shock and a refined set of hydrogen-impact ionization cross-sections (Barghouty 2000) are able to successfully model the transition energy from singly to multiply-charged

- 203 -

A.F. Barghouty, J.R. Jokipii and R.A. Mewaldt

1.0 O

0.8

He/H=0 la; 0/H=0.005

0.6

He: Z ' x 0.13

O

~) 0.4

-

> .,.~ .,m

5

-

-~ 0.2 c~

0.0 10 1

10 2 10 3 10 4 K.E. ( k e Y / n u c l e o n )

10 5

oxygen. The simulated total, i.e., sum over charge states, ACR-oxygen spectrum, however, is found to be deficient at high energies (> 50 MeV/nucleon) where most of the oxygen ions are multiply charged. This deficiency suggests that further ionization during acceleration is still needed so as to provide more multiplycharged ions to be accelerated to higher energies. This paper examines the potential role of helium and heavier neutrals as ionizing media, in addition to hydrogen, at or near the SWTS. Multi-electron stripping in these processes, i.e., heavy atom-impact ionization, can also play a role, an effect that will be explored in detail in a forthcoming publication. AND HEAVY ATOM-IMPACT IONIZATION CROSS-SECTIONS At energies E > few hundreds of keV/nucleon and for ionizing media with nuclear charge Z > 1, the needed ionization cross-sections can be estimated using the scaling relation (Gillespie, 1983): HELIUM

3

Jokipii (1996) and more recently in Jokipii (2000). Very briefly, the model solves the time-dependent Parker heliospheric transport equation in 2 spatial dimensions with drift terms:

Ofq 0 ( Ofq ~ Ofq _ Vd,iOfq 10Vw,i Ofq ~-sourcesq Ot = Oxi _NiJ~ j - Vw,iOxi ~ -~ 3 0 x i Oln p

- sinksq

(2)

where fq(~',p, t) is the distribution function of the ACR ion with charge q = 1, 2, 3,..., Z. From left to right, the RHS terms in the equation depict spatial diffusion, convection, drift, energy gains and losses, and particle sources and sinks due to ionization. The salient transport parameters used for this study are tabulated below. [In Table 1 R is the ion's rigidity, ~ its Lorentz factor, and B the strength of the heliospheric magnetic field at the heliospheric point ~.] Figure 2 illustrates a sample simulation for ACR oxygen at 1 AU. Here we show simulations with and without taking the ionizing contribution of helium into account. Figure 2 suggests that taking helium into

-

204 -

Ionizing media and the observed charge states of anomalous cosmic rays

"6 -5 z 10 > 10 - 6 9

(~'..

!

o

r

~....

~, 10-8 h-,

.,.w

10-9 1

10 K.E. ( M e V / n u c l e o n )

100

account as an ionizing medium, in addition to the ambient hydrogen, does help reduce the deficiency in the simulated intensity at high energies. This addition, however, still seems to fall short of accounting fully for the observed spectrum at high energies. Varying one or more of the salient transport parameters in Table 1 can, in principle, yield a somewhat better fit. In this work, however, they are kept at their nominal values so as to focus on the ionization media and processes at or near the SWTS. Below we discuss the potential contribution and role of multi-electron loss channels in both hydrogen and heavy-atom impact ionization. M u l t i - E l e c t r o n Loss P r o c e s s e s While data remain scarce for proton or hydrogen-impact ionization cross-sections, multi-electron loss cross-sections have been shown to be insignificant compared to single-electron loss in ion-electron collisions at energies relevant to ACR studies (e.g., Krishnakumar and Srivastava, 1992; Deutch et al., 1999). For hydrogen-impact collisions at such energies, multi-electron removal is also expected to be insignificant compared to single-electron removal. For example, the cross section for the two-electron removal process p+He-+p + He +2+2e at 1 MeV/nucleon is found both experimentally (Shah and Gilbody, 1985) as well as in Monte Carlo studies (e.g., McKenzie and Olson, 1987) to be about 2.5 orders of magnitude smaller than that for the single-electron removal process p + H e - + p + H e +1 + l e (see Figure 3). However, as can been seen from Figure 3, for heavy atom-impact ionization, double-electron loss can be ,.o few percent of the single-electron loss cross-section at energies ~ MeV/nucleon. This small, but perhaps not insignificant contribution, will be explored in a future work.

1. Solar wind velocity SWTS radius SWTS strength Heliospheric boundary Neutral-H density Injection energy Heliospheric B-field

350 (equator)- 700 km/s (poles) 100 AU 2.5 160 AU 0.1 cm -3 60 keV/nucleon Parker's+polar modification+ fiat current sheet+ qA > 0 conditions

R1/~ ~/B(~

Parallel diffusion coeff. Perp. diffusion coeff.

0.03• Parallel

- 205 -

A.F. Barghouty, J.R. Jokipii and R.A. Mewaldt ,

.

.

.

.

,

~" 1 0 - 1 5 s

o -16 "~ 10

i

n

g

l

~

0

'~ -17 -~ 10 m 10 - 1 8 m 0

~

lO

-19 .

4,bodY 1 MeV/nucleon

...-~

1 10 N u c l e a r C h a r g e of I o n i z i n g M e d i u m

1

CONCLUSIONS While the simulations presented here for the charge states of ACR oxygen at 1 AU, which are based on the theory of diffusive shock-drift acceleration, appear to capture the essential physics of the charge-changing ionization processes taking place at or near the termination shock the simulated spectrum at high energies remains deficient even after taking helium as an additional ionizing medium into account. This suggests that further ionization during acceleration is still needed. To that end, including heavier neutrals, e.g., oxygen, as well as taking multi-electron stripping into account, are suggested for more realistic modeling of the observed charge states of ACRs. A CKN OWLED G EMENT S Work is supported by NSF grant 9810653 and NASA-JOVE NAG8-1208 (A.F.B.) and by NASA NAS530704 and NAG5-6912 at Caltech. A.F.B. thanks Ed Stone, Alan Cummings, Rick Leske, Conrad Steenberg (Caltech), and Mark Weidenbeck (JPL) for stimulating discussions and insightful comments. REFERENCES Barghouty, A.F., J.R. Jokipii, and R.A. Mewaldt, in Acceleration and Transport of Energetic Particles Observed in the Heliosphere, eds. R.A. Mewaldt et al., pp. 337, AIP ~528, Washington, DC, 2000. Barghouty, A.F., Phys. Rev. A, 61, 052702, 2000. Deutch, H., et al., Int'l. J. Mass Spect., 192, 1, 1999. Fisk, L.A., B. Koslovsky, and R. Ramaty, Astrophys. J. Lett., 190, L35, 1974. Gillespie, G.H., Phys. Lett., 93A, 327, 1983. Gloeckler, G., J. Geiss, and L.A. Fisk, in The Heliosphere near Solar Minimum" the Ulysses Perspectives, eds. A. Balogh et al., Springer-Praxis, Berlin, 2000. Jokipii, J.R., in Acceleration and Transport of Energetic Particles..., ibid, pp. 309, 2000. Jokipii, J.R., Astrophys. J. Lett., 466, L47, 1996. Klecker, B., et al., Space Sci. Rev., 83,259, 1998. Krishnakumar, E., and S.K. Srivastava, Int'I. J. Mass Spect. ~ Ion Proc., 113, 1, 1992. Leske, R.A., et al., in Acceleration and Transport of Energetic Particles..., ibid, pp. 293, 2000. McKenzie, M.L., and R.E. Olson, Phys. Rev. A, 35, 2863, 1987. Mewaldt, R.A., et al., Geophys. Res. Lett., 23, 617, 1996a. Mewaldt, R.A., et al., Astrophys. J. Lett., 466, L43, 1996b. Pesses, M.E., J.R. Jokipii, and D. Eichler, Astrophys. J. Lett., 246, L85, 1981. Shah, M.B., and H.B. Gilbody, J. Phys. B, 18, 899, 1985.

- 206 -

To achieve a self-consistent description of the heliosphere, a fusion of the models addressing its large-scale structure with those of the transport of energetic particles is desirable. We report here on a study of the potential for a thorough analysis of 4-D phase space distributions of anomalous cosmic rays, investigating their energy spectra in further detail. While the computation of such 4-D distributions, thus far, was limited to a spherical, heliocentric shock, we attempt here an investigation of the significance of more realistic shock geometries. The analysis is limited to periods around maximum solar activity for drifts are not taken into account. 1. I N T R O D U C T I O N The overwhelming majority of studies of galactic and anomalous cosmic rays (GCRs and ACRs) within the heliosphere (for a recent review see [1]) is based on the assumption that their phase space distributions do not possess any longitudinal structure. While this assumption appears to be well-justified for heliocentric distances smaller than about 50 AU, it is questionable for the outer heliosphere. There, longitudinal gradients should exist at least for ACRs because: 9 the particle population from which the ACRs originate, i.e. the pick-up ions (PUIs), have longitudinal flux variations (e.g., [2,3]); 9 the injection efficiency of PUIs into the process of diffusive acceleration at the heliospheric shock is likely to be a function of ecliptic longitude, because both the local shock structure (precursor, foot, ramp; e.g., [4,5]) as well as the orientation of the heliospheric magnetic field relative to the shock surface, are likely to change from upwind to downwind within the ecliptic; 9 according to recent models of the large-scale structure of the heliosphere (e.g., [6,7]) the distance to the heliospheric shock increases systematically from the upwind to the downwind direction. We present here results of an extension of our earlier modelling of 3-D [8] and 4-D ACR phase space distributions [9] by considering explicitly a non-spherical heliospheric shock.

-207-

S.R. Sreenivasan and H. Fichtner 2. T H E M O D E L The ACR transport is described on the basis of the time-independent Parker equation + 1

g(V'g~~

Of = 0 ; Olnp

~(~',p)=

(~j_ 0 0

0 ~. 0

0 ) 0 e;ll

(1)

with the distribution function f - f(~', p) in the 4-D phase space defined by heliocentric position f" and momentum p. The solar wind velocity is taken to be radial and constant, +-~ i.e. v~w = 400 km/s g~, and the diagonal elements of the spatial diffusion tensor e; are ~11 - 1022cm2/s (w/c)(P/1 GV)(B/B1Au) and e;• - 0.03~11, where w and P are particle speed and rigidity, respectively. Particle drifts in the heliospheric magnetic field are not included, so that the analysis is limited to periods close to maximum solar activity. The boundary conditions are a vanishing radial gradient of f at the inner boundary at r - to, and prescribed shock spectra at the outer boundary, i.e. the heliospheric shock. They are the same as in [9], i.e. are based on the studies [10] and [11], and depicted as the uppermost lines in the two panels of Fig. 2. 3. T H E L A R G E - S C A L E

ACR DISTRIBUTIONS

Fig. 1 shows the spatial distribution of anomalous hydrogen with a kinetic energy of 31 MeV for a non-spherical heliospheric shock (outermost dashed line) in an equatorial plane (left) perpendicular to the symmetry axis of the heliospheric magnetic field but containing the upwind-downwind axis (horizontal solid line) and in a meridional plane (right) containing both the symmetry axis of the heliospheric magnetic field and the upwind-downwind axis. The shock is elongated in the polar and the downwind direction by factors of 1.3 and 1.5, respectively, which are found in (M)HD studies of the large-scale structure of the heliosphere (see, e.g., [7]). 4. T H E A C R S P E C T R A

AND THEIR COMPARISON

Fig. 2 shows the spectral distribution of anomalous hydrogen for a spherical (left) and a non-spherical heliospheric shock (right), respectively. Since the non-spherical heliosphere is elongated downwind by a factor of 1.5, the boundary spectrum is located at 120 AU in that direction. Therefore, in the right panel of Fig. 2 there are two additional dash-dotted lines giving the spectra at 100 and 120 AU. Obviously, there are significant differences between the outer heliospheric downwind spectra for the spherical and the non-spherical heliosphere. In order to visualize these differences more clearly, Fig. 3 shows the ratio of the downwind spectra R - jsphericat/jnon-spherical for the three distances 66, 78 and 80 AU in the energy interval 1 MeV to 1 GeV. The longitudinal structure is restricted to the outer downwind heliosphere, i.e. to the region r > 50 AU, see left panel in Fig. 1. There, however, the differential flux at a given position is evidently influenced by the large-scale heliospheric structure. While at 66 AU the flux ratio is still rather moderate (R < 5) for all energies above 1 MeV, we find for the interval 1 - 10 MeV 60 > R > 3 at 78 AU and even 110 > R > 4 at 80 AU. For the region beyond the ratio is even higher but since this is downstream of the spherical shock,

-208-

A CR modulation inside a non-spherical modulation boundary

6 "\'~...., T ~..

:

.... ..

,, ! l: ii"

-

2 0 9

-

st

~

S.R. Sreenivasan and It. Fichmer

5. S U M M A R Y

AND CONCLUSION

We confirm our earlier findings about the existence and the order of magnitude of (nonlocal) longitudinal gradients in the phase space distributions of ACRs. In particular we find that" if the heliosphere is non-spherical (see, e.g., Fig. 1), there exists a small longio;.o .... , ................. tudinal gradient in the outer heliosphere (r > 50 AU), shown here as an example for ACR H at a kinetic energy of 31 MeV; 9 this longitudinal structure is most clearly visible from a comparison of upwind and downind spectra (Fig. 2). The effect of heliospheric geometry is probably stronger than that of the variation in the source strength across the shock surface;

Figure 3" The ratio of the differential fluxes R - j~ph~i~g/j~o~-~ph~i~l in the outer downwind heliosphere (see Fig. 2) for the energy interval of interest for ACR H.

9 the polar elongation reduces the absolute flux as well as the latitudinal gradient at a given location in the heliosphere (Fig. 1); 9 the in-ecliptic spectra in the upwind direction (Fig. 2, solid lines) are basically unaffected by the geometry of the heliosphere;

9 the in-ecliptic spectra in the downwind direction (Fig. 2, dashed lines, and Fig. 3) show significant differences between a spherical and a non-spherical heliosphere, but the effect is limited to the outer heliosphere and to relatively low energies (< 10 MeV). In conclusion we find that, at least for anomalous hydrogen, there is a clear signature of the large-scale structure of the heliosphere in its spectra but that it is not likely to be observed in-situ with the presently active deep space probes: the effect is too small in the upwind heliosphere which is the region that is accessible to direct observation with the two Voyager spacecraft. REFERENCES ,

2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

H. Fichtner, Space Sci. Rev. 95 (2001) 639. S.V. Chalov, H.J. Fahr and V. Izmodenov, Astron. Astrophys. 320 (1997) 659. U. Mall, H. Fichtner, E. Kirsch, D.C. Hamilton and D. Rucinski, Planet. Space Sci. 46 (1998) 1375. J.A. le Roux, H. Fichtner, G.P. Zank and V.S. Ptuskin, JGR 105 (2000) 12557. J.A. le Roux, G.P. Zank, H. Fichtner and V.S. Ptuskin, GRL 27 (2000) 2873. H.J. Fahr, T. Kausch and H. Scherer, Astron. Astrophys. 357 (2000) 268. G.P. Zank, this issue. H. Fichtner, H. de Bruijn and S.R. Sreenivasan, Geophys. Res. Lett. 23 (1996) 1705. H. Fichtner and S.R. Sreenivasan, Adv. Space Res. 23 (1999) 535. J.A. le Roux, M.S. Potgieter, and V. Ptuskin, JGR 101 (1996) 4791. C.D. Steenberg, Ph.D. Thesis, Univ. of Potchefstroom, South Africa (1998).

-210-

The Injection Problem for Anomalous Cosmic Rays I G.P. Zank, W.K.M. Rice, J.A. le Roux, and W.H. Matthaeus Bartol Research Institute, University of Delaware, Newark, DE 19716, USA 1. I N T R O D U C T I O N That interstellar pickup ions are related to the anomalous cosmic ray (ACR) component is now well established. However, pickup I-I+ has typical energies of 1-3 keV, sufficiently energetic to form a suprathermal, energetically important plasma component in the outer heliosphere, but well below ACR energies which extend to several 100 MeV/nuc. How some fraction of the interstellar pickup ion population becomes preferentially energized up to such high energies remains a central question in the physics of cosmic ray acceleration. It is generally accepted that for sufficiently energetic pickup ions, first-order Fermi acceleration, otherwise known as diffusive shock acceleration, at the heliospheric termination shock can account for the accelerated ACR spectrum. However, the precise mechanism by which some pickup ions are selected and possibly pre-energized up to energies sufficiently large that they can be Fermi accelerated remains a puzzle. This is the so-called "injection problem" for ACRs. Observations [ 1-3] at interplanetary shocks suggest that (i) accelerated ions emerge directly from the thermal pool; (ii) the injection efficiency appears to depend on the nature of the underlying particle distribution function, and (iii) the observed pickup ion injection efficiency appears to correlate inversely with mass. However, Cummings and Stone [4] used anomalous cosmic ray (ACR) energy spectra measured by Voyager 1 and 2 during 1994 to determine an injection efficiency for ACRs. This approach depends on uncertainties associated with the diffusion tensor model (and associated turbulence and drifts), termination shock location, neutral and pickup ion fluxes, shock strength, and the possible role of ions pre-accelerated by interplanetary shocks. Nonetheless, a basic conclusion is that, even if the estimated injection efficiencies are possibly inaccurate, heavier ions are injected preferentially. Furthermore, besides the ACR "injection efficiencies" [4] being in the opposite sense of those presented by Gloeckler et al. [2], the "injection efficiencies" of the latter authors are considerably higher than those found for ACRs. Finally, it appears [ 12] that heavy ion diffusive acceleration times can be longer than proton diffusive acceleration times. Another important result related to the injection and acceleration of ACRs at the termination shock is that the diffusive shock acceleration time scale for ions must be rapid [5]. To ensure that most ACRs remain singly ionized, one is obliged to assume a weak hard-sphere scattering model at a quasi-perpendicular termination shock [5]. The requirement that the scattering be weak (in the hard sphere sense) implies that the pickup ions already be energetic for them to be Fermi accelerated. If r/~ (defined below in 3) is a measure of the scattering strength, so that r/~ >> 1 corresponds to weak scattering, then for a particle to be scattered multiple times as a field line convects through a shock, the particle velocity v must satisfy v >> v,~o~, where V~his the shock speed [5,6]. Using reasonable This work supported in part by NASA grants NAG5-6469, NAG5-7796, an NSF-DOE award ATM-0078650, and an NSF award ATM-0072810.

-211 -

G.P. Zank et al. parameters [5] requires that pickup ions to have energies close to ~ 1 MeV [7]. As a consequence, it appears that some pre-energization of pickup ions is necessary. A successful resolution of the "injection problem" at the quasi-perpendicular termination shock must relate the above sets of observations consistently. 2. MULTIPLY REFLECTED ION ACCELERATION Since part of the pickup ion distribution function in the shock frame has a very small normal velocity component at the shock interface, these particles are reflected at the electrostatic cross shock potential ~ [7,8]. For a pickup ion shell ahead of the shock, the fraction of the distribution Rref that is incapable of surmounting the cross shock potential barrier is found to be [7]

Zm Rr~ --

2

M 2MI, ( r - 1)

,

where m refers to the proton mass, M and Z to the mass and charge of the particle of interest (pick-up I-F, He +, etc.). These reflected ions are capable of being accelerated to large energies by experiencing multiple reflections at the electrostatic barrier. If we interpret reflection efficiency as injection efficiency, then (i) heavier pickup ions, i.e., with M > m, are less efficiently injected, and (ii) injection efficiency increases with increasing charge. The downstream or transmitted pickup ion distribution is a power law in energy that is extremely hard (v-a) [7]. The balancing of the particle Lorentz force against the electrostatic potential gradient shows [7,8] that the maximum energy gain is proportional to the ratio of an ion gyroradius (whose velocity is that of the solar wind) to the smallest characteristic electrostatic shock potential length scale Lramp. If Lro~p is the thermal solar wind ion gyroradius, the initially very low velocity pickup ions will be accelerated up to no more than the ambient solar wind speed. However, our current understanding of the micro-structure of quasi-perpendicular shocks is that fine structure in the shock potential can be on the order of electron inertial scales [9], so yielding pickup ion energies of several 100 keV at even weak interplanetary shocks. 3. A SYNTHESIS

Although pickup ions are energized by the MRI mechanism, not all the MRI accelerated ions are sufficiently energetic to be further accelerated by a second-stage diffusive shock acceleration process. For diffusion theory to be applicable at a perpendicular shock, particles downstream of the shock must be capable of diffusing upstream. This requires the cosmic ray anisotropy be small. For a perpendicular shock, this implies that the particle velocity v 1/2r , where V,~ is the upstream flow speed in satisfy the condition [10] v >> (3V,~/r)(1 + 772) the stationary shock frame and r the shock compression ratio. For the present, we assume hard-sphere scattering to describe the transport of diffusive particles. Thus, if ;t n is the parallel mean free path ~ll/r_L = l+ 0 z, rlc = 3Xll/(vrg)= ~,ll/rg. Here rg = pc/(QB)is the particle gyroradius (p is particle momentum, Q the charge, c the speed of light, and B the magnetic field strength). In the hard sphere scattering model, r/, is a measure of the strength of scattering: rt~ small implies strong scattering whereas r/, large corresponds to weak scattering. For resonant scattering, Au oc R 1/3 where R = p c / Q is the particle rigidity. It

-212-

The injection problem for anomalous cosmic rays follows then that r/c ,,~ R -2/3, or

0 c rlcp(m/M) 2/3, 77cp -(2ll/rg)proton.

important

implication is that even if scattering is weak for protons, i.e., r/cp >> 1, the inverse dependence of 17C on M implies that 0c can be much smaller for heavy ions. Thus, heavy ions are accelerated diffusively at a much lower threshold velocity than lighter ions. Consequently, a larger fraction of heavy MRI accelerated ions will enter a second-stage diffusive shock acceleration process than light MRI accelerated ions. It remains to determine whether the two competing mass dependence effects conspire to satisfy the Cummings and Stone [4] ACR injection results. We consider two cases; (i) a highly peaked distribution such as a shell, and (ii) a power law distribution such as that obtained by Vasyliunas and Siscoe [1 1]. This gives (i) grey =

Rref

=

Rr~y(H +)[m/M] 1/2, and (ii)

R~/(H+)[m/M] 9 Simulations suggest that for (i), we use a (V/Vo) -4 accelerated spectrum, and for (ii) a

(v/v o)-5

accelerated spectrum. We suspect that the correct model lies between these two extremes. The differential intensity of the ith species ji(m-Zs-'sr-lMeV -1) at the lID

termination shock is given by j~ = pZfa(p,M). In order to calculate the acceleration efficiency in the same way as [4], we rewrite the differential intensity in their form,

10"

q e iF~ -~oiP(q-4 )/ZE(-q+ Z)/Z , w h e r e 47r F~(cm -2s -l ) is the flux of the ith pickup ion

i.e.,Ji = Fig. 1. Inverse acceleration efficiencies for I-I+, He +, O +, and Ne +, plotted as a function of ion mass M, and normalized to He +. The acceleration efficiencies for O +, and Ne + are very similar and greater than those of He + and IV. We would predict this to be true of C + and N + too. The injection/acceleration efficiency for I-I+ is anomalously low. The filled triangles correspond to a pickup ion shell distribution with the reflected 9 --4 MRI accelerated spectrum proportional to v , and the open circles to a Vasyliunus and Siscoe pickup ion distribution and a softer reflected MRI accelerated spectrum proportional to v -5. The derived observations are given by the open squares [4].

species at the termination shock, E0~ the

injection energy of 5.2 x 10 -3 MeV/nuc. used by Cummings and Stone, and e~ is the acceleration efficiency that is to be computed. Note our distinct use of the term "acceleration efficiency" for e~, which is referred to as "injection efficiency" by [4]. We use the former term to distinguish our use of injection efficiency in the context of MRI injection. The injection energies needed to render an ion diffusive range f r o m - 3 3 0 keV for pickup H + to about either ~ 185 (shell) or --75 (power law) keV for C +, N +, O +, and Ne +. Our predicted acceleration efficiency and that inferred by Cummings and Stone can now be compared directly since the same parameters -1 are used. In Fig. 1, we plot, following [4], the inverse acceleration efficiency e~ , normalized to He+, as a function of ion mass. The solid triangles correspond to an in initial pickup ion shell distribution, the open circles to the VS power law case, and the open squares to the

-213-

G.P. Zank et al.

inferred observations presented in [4]. Two points stand out from Fig. 1. The first is the obviously close agreement between the model results and the observed results. The second is that the extreme cases considered here, the initial shell and the VS distributions, yield acceleration efficiencies that are almost identical, suggesting that a more realistic pickup ion distribution at the termination shock is unlikely to alter our conclusions significantly. In the limit of strong scattering, the injection ~0"threshold for ions to be viewed as diffusive is pl virtually identical for all ion species. 10 ~ Consequently, no difference exists between the ~ ~0' injection efficiencies of different mass ion 10' species in the case of strong scattering. This is an important point since it relates for the first lOS time the differential injection efficiency of pickup ions of different masses to the particle t- ~0-'scattering strength. ~o~,0~. One can use the diffusive part of the MRI Energy (MeV) spectrum for pickup H + as the explicit source for ACRs at the termination shock, and hence Fig. 2. Fluxes of pickup I-I+, MRI determine the ACR spectrum at the shock and accelerated I-I+, and I-I+ ACRs at the termination shock, together with the the modulated spectrum at any point within the resulting modulated ACR flux at 57 AU heliosphere. The combined pickup ion, MRI (dashed line), accelerated and ACR spectrum at the termination shock, assumed to be located at 80 AU with parameters given by [4], is illustrated in Fig. 2. The flux of ACRs shown in Fig. 2 is in accord with the expected ACR source spectrum used to model the observed modulated ACR flux within the heliosphere. 4. C O N C L U D I N G R E M A R K S MRI acceleration or shock surfing [7,8] can explain naturally the injection characteristics of pickup ions at interplanetary shocks [2,3] as well as the inferred injection efficiency at the termination shock for ACRs [4,5]. Weak scattering yields an anomalously low injection efficiency for I-I+ compared to He +, C +, N +, O +, and Ne +. Computed ACR termination shock and modulated fluxes compare well to those observed. Strong scattering eliminates the mass dependence of the ACR injection efficiency. See [ 10] for further details and discussion. REFERENCES IGosling, J.T., et al., J. Geophys. Res., 86, (1981) 547. 2Gloeclder, G., et al., J. Geophys. Res., 99, (1994) 17,637. 3Fr/tnz et al., Geophys. Res. Lett., 26, (1999) 17. 4Cummings, A.C., and Stone, E.C., Space Sci. Rev., 78, (1996) 117. 5jokipii, J.R., ApJ, 313, (1987) 842. 6Webb, G.M., Zank, G.P., Ko, C.M., & Donohue, D.J., ApJ, 453, (1995) 178. 7 Zank, G.P., Pauls, H.L., Cairns, I.H., & Webb, G., J. Geophys. Res., 101, (1996)457. 8Lee, M.A., Shapiro, V.D., & Sagdeev, R.Z., J. Geophys. Res., 101, (1996) 4777. 9Newbury, J.A., Russell, C.T., & Gedalin, M., J. Geophys. Res., 103, (1998) 29,581. l~ G.P., Rice, W.K.M., le Roux, J.A., Matthaeus, W.H., ApJ, (2000) in press. llVasyliunas, V.M., and Siscoe, G.L., J. Geophys. Res., 81, (1976) 1247. ~2Scherer, H. et al., J. Geophys. Res., 103, (1998) 2105.

-214-

Self-consistent acceleration of pickup ions at the termination s h o c k J. A. le Roux a, H. Fichtner b, G. P. Zank a, and V. S. Ptuskin c aBartol Research Institute, University of Delaware, Newark, DE 19716, USA bTheoretische Physik IV, Weltraum- und Astrophysik, Ruhr-Universit~it Bochum, Bochum, Germany Clnstitute for Terrestrial Magnetism, Ionosphere, and Radiowave Propagation (IZMIRAN), Troitsk Moscow District, Russia

Self-consistent simulations indicate that when low-energy pickup protons are preaccelerated locally at a nearly perpendicular solar wind termination shock (TS) by the multiply reflected ion (MRI) acceleration mechanism, this might lead to the injection of a sufficient number of pickup protons into standard diffusive shock acceleration to enable reproduction of the observed upstream anomalous cosmic ray (ACR) proton spectra. This is possible despite a significant mediation of the TS by the MRI-accelerated PUIs as well as ACRs that weakens MRI acceleration.

1. INTRODUCTION It is generally accepted that the ACR component is formed when interstellar pickup ions (PUIs) undergo standard diffusive shock acceleration at the nearly perpendicular solar wind TS. At a nearly perpendicular TS, however, a significant energy threshold needs to be overcome by the PUIs before the standard theory applies. It is not known what fraction of PUIs can overcome the threshold by virtue of preacceleration in the upstream solar wind [ 1] or what fraction are injected due to local preacceleration at the TS. Here we investigate the self-consistent, local preacceleration of unaccelerated pickup protons, their injection into, and acceleration by standard diffusive shock acceleration at the TS. By basing the preacceleration and injection of PUIs on MRI acceleration theory [2, 3], an improvement on previous ad-hoc approaches to injection, it is investigated how MRI-accelerated PUI and ACR protons mediate the TS structure, how the mediation affects the efficiency of the two acceleration mechanisms, and what consequences this has for upstream modulation of ACR protons.

2. THE MODEL

The model is time dependent and spherically symmetric with its parameters chosen to reflect conditions in the upwind, equatorial regions. To determine the TS modification by

-215-

J.A. le Roux et aL

MRIs and ACRs, a set of three-fluid equations is solved numerically whereby the fluids are the solar wind plus unaccelarated pickup, MRI-accelerated, and ACR protons with the latter two fluids taken to good approximation to have a low mass density. The fluid equations are closed by calculating the MRI-accelerated pickup proton spectrum and pressure for unaccelerated PUIs, and by solving numerically the standard ACR transport equation to determine the ACR spectrum and pressure. To determine the MRI spectrum and pressure, we first calculate the unaccelerated PUI spectrum at the TS from the analytical solution of the standard PUI transport equation (the standard ACR transport equation without diffusion but with an appropriate source term for the production of pickup protons due to charge exchange). MRI theory is applied to the PUI spectrum to determine the fraction of PUIs involved in MRI acceleration, their expected maximum energy, and the form of the MRI spectrum. The idea behind MRI acceleration is that PUIs with small velocity components normal to the TS are unable to penetrate the crossshock electric field. These particles skim along the TS surface while gaining energy from the motional electric field. Their maximum energy is determined when their Lorentz force overcomes the cross-shock electric field force. The MRI spectrum is updated in the model as the TS is mediated and the cross-shock electric field is weakened under the dynamic influence of the MRIs and ACRs.

\\

84,8

84,9

....

R

85,0

Radial distance (AU) Figure 1. Solar wind speed normalized to Uo = 400 km/s as a function of heliocentric distance in astronomical units (AU). The dashed curves denote the case where only MRI-accelerated pickup protons mediate the TS while the solid curves also include the modifying effects of diffusively TS accelerated ACR protons. The left panel provides a close-up view of the TS shown in the right panel. A portion of the MRI spectrum is diffusively shock accelerated at the TS provided particle speeds satisfy v > Vinj = 4au2(~lJ~• ]/2 where a' = 2, u2 is the downstream solar wind speed and ~,,, is the parallel, perpendicular diffusion coefficient [4]. The radial diffusion coefficient 1err is determined on the basis of quasi-linear theory for parallel diffusion, a theory for the random walk of field lines applied to perpendicular diffusion, and a transport theory for MHD turbulence in the solar wind [5]. For more details about the model see [6, 7].

-216-

Self-consistent acceleration of pickup ions at the termination shock 3. RESULTS The results indicate that both MRI-accelerated pickup and A CR protons might significantly mediate the TS. This is because a considerable fraction of the PUIs (--16%) are MRI accelerated of which a significant amount enters diffusive shock acceleration. Whereas the TS compression ratio is initially -3.1 due the presence of PUIs in the solar wind, MRIaccelerated PUI protons alone is found to weaken the TS compression ratio to s = 2 (dashed curve in left panel of Figure 1). Viewed on a larger scale (20 AU) appropriate for detecting TS modification by ACRs, the TS modification by MRI-accelerated PUIs becomes almost invisible (dashed curve in right panel of Figure 1). When including the effect of the ACR proton pressure on the TS, the compression ratio appears to be s --- 2.3 on this large-scale view of the TS as a large-scale TS precursor due the ACR protons is formed (solid curve in right panel of Figure 1). This gives an idea of the effect of the ACRs on the TS structure. A smallscale view of the TS structure (< 1 AU) shows that the combined effect of MRI and diffusive shock acceleration of PUI protons reduce the TS compression ratio to s = 1.6 (solid curve in Figure 1). The TS modification by MRI and ACR protons results in a weaker cross-shock electric field so that the maximum energy that MRI acceleration provides to PUIs is reduced significantly. Initially, the ~- 1o maximum energy that the MRI-accelerated lo~ pickup protons reach is -170 keV for a 18 1! ~ ~o o (sub)shock ramp with width e = 2 (width ,8 normalized to an electron inertial length) i5 o [8], but after modification of the TS by both 10: 10' MRI and ACR protons the maximum 10-a 10-7 10-0 10-s 10 4 10-3 10-2 10-1 10 0 101 10 2 energy is --74 keV. The maximum energy was further enhanced by a factor of--2 due ,~

0

Figure 2. Differential intensity in particles m -2 s -1 s r -1 MeV -] as a function of kinetic energy in GeV. The top solid curve labeled rsh depicts the simulated combined unaccelered PUI, MRI-accelerated PUI, and ACR proton spectrum downstream of the TS, while the modulated ACR proton spectra upstream are shown at heliocentric distances 1, 23, 42 and 65 AU (lower solid curves from bottom to top). The dashed curve above (below) the top solid curve is the PUI (MRI) spectrum upstream. The filled and open circles denote a reproduction of Voyager 1 data at --65 AU and Voyager 2 data observed a t - - 5 2 AU during 1997, respectively [ 13, 14].

to adiabatic compression of the MRIaccelerated PUIs when crossing the TS. The injection energy was calculated as -50 keV. This threshold allows for enough MRIaccelerated PUIs to be injected into standard diffusive shock acceleration so that the modulated ACR proton spectra upstream compare favorably with observational data recorded by the Voyager spacecraft (see Figure 2). Even though the injection occurs most of the time, it is highly time dependent and sometimes is interrupted as the TS becomes too strongly modified and the cutoff energy for MRI acceleration falls to far below the threshold. In addition, injection might also be interrupted when MRI acceleration itself

-217-

J.A. le Roux et al.

fails. This happens when the TS becomes so strongly mediated that reflection by the crossshock electric field ceases. This is expected too happen when the upstream Mach number of the TS drops to between 1 and 2 [9]. Calculations show that if it is assumed that reflection stops when this critical Mach number is less than 2, then injection becomes sporadic as reflection is interrupted most of the time. The results shown here are for an assumed value of 1.5 whereby reflection occurs continuously. Further simulations underscore the robustness of the model in that comparable results are yielded for e < 10. For all TS (sub)ramp widths of less than about an ion inertial length, the MRI acceleration cutoff energy is sufficiently high to allow injection into diffusive shock acceleration. However, intensities of the modulated ACR proton spectra upstream become too small for reproducing observations when the TS (sub)ramp width approaches an ion inertial length (weak MRI acceleration).

4. SUMMARY AND CONCLUSIONS

In summary, these self-consistent simulations indicate that low-energy pickup protons preaccelerated locally at a nearly perpendicular TS by the MRI mechanism might lead to the injection of a sufficient number of MRI-accelerated PUIs into diffusive shock acceleration for an injection threshold o f - 5 0 keV to enable reproduction of the observed upstream ACR proton spectra. This is possible despite significant mediation of the TS by MRI-accelerated PUIs as well ACRs formed by the diffusive shock acceleration of a fraction of the locally preaccelerated PUIs. For the same injection energy recent work shows that not enough PUIs preaccelerated in the upstream solar wind will be injected to explain ACR observations because the flux of preaccelerated PUIs fall strongly off as 1/r 3 at large heliocentric distances due to adiabatic cooling [10, 11]. Given modeling to the contrary [12], and the uncertainty in the injection efficiency, these conclusions still need further study. REFERENCES

Chalov S. V., & Fahr, H., J. 2000, Astron. Astrophys., 360, 381. Zank, G. P., Pauls, H. L., Cairns, I. H., & Webb, G. M. 1996, J. Geophys. Res., 101,457. Lee, M. A., Shapiro, V. D., & Sagdeev, R. Z. 1996 J. Geophys. Res., 101, 4777. Webb, G. M., Zank, G. P., Ko, C. M., & Donohue, D. J. 1995, Astrophys. J., 453, 178. Zank, G. P., et al. 1998, H., J. Geophys. Res., 103, 2085. le Roux, J. A., Fichtner, H., Zank, G. P. & Ptuskin, V. S. 2000, J. Geophys., Res., 105, 12557. le Roux, J. A., Zank, G. P. Fichtner, H., & Ptuskin, V. S. 2000, Geophys. Res. Lett., 27, 2873. Newbury, J. A., Russell, C. T., & Gedalin, M. 1998, J. Geophys. Res., 103, 29581. 9. Edmiston, J. P., & Kennel, C. F. 1984, J. Plasma Phys., 32, 429. 10. Decker, R. B., et al. 2001 this volume. 11. Rice, W. K. M., Zank, G. P., le Roux, J. A., & Matthaeus, W. H., 2001 Adv. Space Res., in press. 12. Giacalone, J., et al. 1997, Astrophys. J., 486, 471. 13. Cummings, A. C., & Stone, E. C., 1998, Space Sci. Rev., 83, 51. 14. McDonald, F. B., Space Sci. Rev., 83, 33, 1998. ~

2. 3. 4. 5. 6. .

~

-218-

Energetic Neutral Helium of Heliospheric Origin at 1 AU A. Shaw ~, K. C. Hsieh ~ *, M. Hilchenbach b, A. CzechowskF, D. Hovestadt a, B. Klecker a, R. Kallenbach ", E. M6bius f P. Bochslerg ~Department of Physics, University of Arizona, Tucson, AZ 85721, USA bMax-Planck-Institut fiir Aeronomie, D-37189 Katlenburg-Lindau, Germany r

for Space Research, Polish Academy of Sciences, PL-00716 Warsaw, Poland

dMax-Planck-Institut fiir Extraterrestrische Physik, D-85740 Garching, Germany eInternational Space Science Institute, CH-3012 Bern, Switzerland fEOS, University of New Hampshire, Durham, NH 03824, USA gPhysikalisches Institut der Universits Bern, CH-3012 Bern, Switzerland Using measurements from the HSTOF (High-Energy Suprathermal Time-of-Flight sensor) instrument on SOHO (Solar and Heliospheric Observatory) at 1AU, we report on a possible detection of Energetic Neutral Helium Atoms of probable heliospheric origin. Observations made under quiet interplanetary conditions showed an extra component in the helium spectrum between 85 and 141 keV, here we examine the nature of this component and the basis upon which we tentatively identify it as neutral. 1. I N T R O D U C T I O N A singly-charged energetic ion in a space plasma, such as a proton or He + of the anomalous cosmic ray (ACR) population, can become an energetic neutral atom (ENA) by charge-exchanging on an atom of the ambient neutral gas, e.g. H or He of the local interstellar medium (LISM). The resulting ENAs, unmodulated by magnetic fields, can reach regions of space normally inaccessible to the original charged population. This enables us to sample space plasmas in currently inaccessible regions; e.g., ACR in the outer heliosphere. The use of ENAs to probe the ACR population in the outer heliosphere was first discussed by Hsieh et al.[1]. The instrument HSTOF of the Charge, Element and Isotope Analysis System (CELIAS) on SOHO is the first instrument in interplanetary space capable of detecting ENAs, Hovestadt et al.[2], enabling Hilchenbach et al.[3] to measure the intensity of energetic hydrogen atoms (EHA) between 55 and 80 keV at 1 AU under quiet time (QT) interplanetary conditions for the first time. *The work at the University of Arizona is supported in part by NASA grant NAG5-7966 and NSF grant ATM9727080. K.C. Hsieh's travel is partly funded by the University of Arizona's Foreign Travel Grant.

-219-

A. S h a w e t al.

2.

SEARCHING

FOR

ENERGETIC

NEUTRAL

HELIUM

Figure 2. He + and He ++ extraction

Figure 1. HSTOF data 1996 t o 1999

The TOF vs. E' analysis of HSTOF separates the elements into tracks according to mass (Fig. 1). It also effectively reduces noise in the data, requiring each genuine particle to trigger all three detectors: Start, Stop, and PSSD (Pixelated Solid State Detector). These signals from a real event fall in sequence within extremely short time intervals. Noise due to accidental coincidence can be estimated and removed by examining events registered in a region of the TOF vs. E ~plane that do not correspond to genuine particles. Helium exists in three charge states: He ~ (neutral), He +, and He ++. The parallelplate E / Q filter of HSTOF has no effect on He ~ but excludes low energy ions. Fig. 2 displays the energy profile of He + and He ++ during a solar-related event. We use eight such events, during which no significant He ~ flux is expected, to show HSTOF's distinct response to He + and He ++ (Fig. 3). This instrument response set 85 < E < 141 keV as the best energy range for the detection of energetic neutral helium (ENHe), also providing a means to estimate and remove the probable background in that energy range, caused by He + and He ++ during both quiet and non-quiet interplanetary conditions. The periods of quiet interplanetary conditions used by Hilchenbach et al.[3] have been extended from 1997 to late 1999. The total helium flux as a function of kinetic energy, E, accumulated over all these periods - 506 days along the SOHO orbit (1 AU) from 1996 to 1999 - is shown in Fig. 4. The qualitative difference between the energy profiles of the two distinct populations of helium shown in Fig. 2 and Fig. 4 seems to indicate the presence of ENHe in the assigned energy range, 85-141 keV, during the QT periods (~ 100 counts). More analysis is however required to completely rule out possible contamination. We shall now try to determine the direction of arrival of these assumed ENHe.

- 220-

Energetic neutral helium of heliospheric origin at 1 A U

Figure 4. HSTOF Quiet Time (QT) Flux

Figure 3. HSTOF He + and He ++ response

Figure 5. HSTOF Look Direction

3. A R R I V A L D I R E C T I O N

Figure 6. Helium Cone[6]

OF ENHe FLUX

HSTOF has a bore-sight lying in the ecliptic plane, 37 ~ west of the SOHO-sun line (Fig. 5). The field-of-view of HSTOF covers 2 ~ either side of the bore-sight in the ecliptic plane, and +17 ~ normal to the ecliptic. Since SOHO moves with the Earth-Sun L1 Lagrangian point, it's daily location can be referenced by Earth's Day of Year (DOY). The direction of the Sun's motion with respect to the LISM is 254 ~ this is the Apex (upwind) direction. HSTOF looks towards the apex around DOY 12, and in the opposite direction (Anti-Apex or downwind) towards the heliotail around DOY 195. The QT ENHe flux as a function of DOY is shown in Fig. 7. Each data point represents 10 days of data. The flux level is higher and more scattered in 1996 than in 1997. The 1997 data suggests a peak ENHe flux at around DOY 200, possibly originating from interactions between He + accelerated in co-rotating interaction regions (CIRs) and the

-221 -

A. Shaw et al.

O:

7

ii:, -

t !: t

,

ii

Figure 7. HSTOF Quiet Time Flux vs. Time

LISM helium cone (Fig. 6), as is the case for EHA[4]. 4. C O N C L U S I O N Though the final absolute helium flux calibration is pending, data from HSTOF/CELIAS/SOHO suggests the first detection of energetic neutral helium (ENHe) in interplanetary space. The arrival direction of the peak flux in 1997 suggests an origin in the interaction between CIRs and the LISM helium cone. If the observed flux is confirmed as ENHe then careful comparison of our results with the observed EHA flux[3], models of EHA production from anomalous cosmic rays (ACR)[5] and CIR accelerated ions[4] must be performed before any definitive conclusions as to the origin of this flux can be established. REFERENCES

1. 2. 3. 4. 5. 6.

K.C. Hsieh, K.L. Shih, K.R. Jokipii, and S. Grzedzielski, Astophys. J. 393 (1992) 756. D. Hovestadt et al., Solar Phys. 162 (1995) 441. M. Hilchenbach, K.C. Hsieh, D. Hovestadt et al., Astophys. J. 503 (1998) 916. J. Kota, et al. submitted J. Geophys. Res. (2000) A. Czechowski et al. submitted Astron. & Astrophys. (2000) E. M6bius, SSRv. 78 (1996) 375.

- 222 -

Effects of the heliospheric termination shock on possible local interstellar spectra for cosmic ray electrons and the associated heliospheric m o d u l a t i o n S. E. S. Ferreira, M. S. Potgieter and U. W. Langner School of Physics, Potchefstroom University for CHE, 2520 Potchefstroom. South Africa

The 'local' interstellar spectrum (IS) for cosmic ray electrons at energies of interest to heliospheric modulation studies is still basically unknown. Recently, new computations of the IS based on advanced modeling of cosmic-ray propagation in the Galaxy [1], and observations including diffuse galactic gamma rays, indicated that the electron IS may be considerably lower at energies below-100 MeV than previously assumed. For this work different scenarios for the electron 'local' IS, and their subsequent modulation in the heliosphere, are studied using a shock-drift modulation model. The effects of the heliospheric termination shock (TS) on each of these scenarios are illustrated, together with the subsequent effects on their modulation in the heliosphere. We find that the computed effect of the TS on galactic electron intensities at 16 MeV is relatively small, in general, but more pronounced if the TS is positioned at 80 AU, than at 90 or 100 AU. The larger the 'local' IS value is, the larger the effect of the TS on electron modulation at this energy becomes. 1. I N T R O D U C T I O N The study of the modulation of cosmic ray electrons in the heliosphere is an important and useful tool in understanding various aspects of heliospheric modulation. Modulated electron intensities in the lower-MeV range give a direct indication of the average parallel and perpendicular mean free paths in contrast to protons that experience adiabatic energy changes below-300 MeV (e.g. [2]). Gradient and curvature drifts become less important for electron modulation at lower energies, with almost no effect below 100 MeV (e.g. [3]). The Pioneer 10 radial-intensity-profiles f o r - 1 6 MeV electrons ([4],[6]) indicate almost no radial gradients out to -70 AU, which put serious constraints on the diffusion tensor. New computations of the interstellar spectra (IS) [1], indicate that the electron IS may be considerably lower at energies below -100 MeV than previously assumed (e.g. [11]). For this work, two different (extreme) scenarios for the 'local' IS for cosmic ray electrons ([1],[5]), and their subsequent modulation in the heliosphere are studied using a shock-drift-modulation model. Satisfying the constraints imposed on the diffusion tensor by the Pioneer 10 electron data in the outer heliosphere, the effects of the location of the heliospheric termination shock (TS) on each of the IS scenarios are illustrated, together with the subsequent effects on their modulation. 2. M O D U L A T I O N

MODEL AND PARAMETERS

The model is based on the numerical solution of the transport equation (TPE) [7]:

- 223 -

S.E.S. Ferreira, M.S. Potgieter and U W. Langner

3fot = - ( V

I(V.V) + (v i) ) ) . V f + V . ( K s . V f ) + ~-

3f +J 3 ln-~

,

(1)

source

where f (r~,t) is the CR distribution function; R is rigidity, r is position, and t is time, with V the solar wind velocity. Terms on the right-hand side represent convection, gradient and curvature drifts, diffusion, adiabatic energy changes and a source function, respectively. The symmetric part of the tensor Ks consists of a parallel diffusion coefficient (Kll) and a perpendicular diffusion coefficient (K• The anti-symmetric element KA describes gradient and curvature drifts in the large scale HMF. J~our~e can be any source, e.g. Jovian electrons, pick-up ions, and/or the local interstellar spectra (IS). Here, we concentrate on the IS for electrons, neglecting all other local sources. The TPE was solved using a two dimensional TSdrift model developed by le Roux et al. [8], and expanded by Haasbroek [9]. The outer modulation boundary is assumed at 120 AU. A TS with a compression ratio of 3.2 < s < 4.0, and scale length of L = 1.2 AU was assumed at rs = 80 AU. The solar wind speed V was assumed to change from 400 km.s 1 in the equatorial plane (0 = 90 ~ to a maximum of 800 km.s ~ when 0 < 60 ~ At the shock, V decreases from the upstream value of V1 = 400 km.s -1 in the equatorial plane according to the relationship given by le Roux et al. [8]: V(r)= VI(S+ 1)_ V I ( S - 1 ) t a n h ( r - r~'] 2s 2s [, L )

(2)

For the diffusion coefficients that describe diffusion parallel, Kjl, and perpendicular, K• to the average heliospheric magnetic field (HMF) as well as the asymmetric coefficient KA, which describes gradient and curvature drifts in the background HMF, we assumed: Be", Kii = K ofif ( R ) --ff-

K_l_r = a 1 + (k,KI]ll/rg

" )2 '

K 2.0

Kll )5," = b 1 + (All/rg

/3n KA = (KA)o ~

(3)

Here/3 is the ratio of the speed of the cosmic ray particles to the speed of light; f(R) gives the rigidity dependence (in GV) with f(R)=R when R > 0.4 GV, and f(R)=0.4 when R ~ 0.4 GV; B is the HMF magnitude modified [12] in qualitatively agreement with Ulysses observations [13]; K0 is a constant in units of 6.0 x l 0 2~ c m 2 S"1", a is a constant which determines the value of K• which contributes to perpendicular diffusion in the radial direction, and b is a constant that determines K• which contributes to perpendicular diffusion in the polar direction. Diffusion perpendicular to the HMF is therefore enhanced in the polar direction by assuming b > a. ([3],[10],[11]). The ratio of the scattering mean free path to the particle gyroradius, )~ll /rg, is larger than unity in view of the Bohm limit, ~'11= rg. The TPE was solved in a spherical coordinate system with the current sheet "tilt angle" (x = 10 ~ for so-called A > 0 epochs (-1990 to present) when electrons primarily drift inward through the equatorial regions of the heliosphere. 3. R E S U L T S

AND DISCUSSION

Figure l a shows the radial profiles of computed 16 MeV galactic electron intensities with an outer boundary at rB -- 120 AU, with the IS of Strong et al. [5] assumed to be the local electron spectrum. Three different radial profiles are shown; the solid line corresponding to a TS at rs = 80 AU, the dashed line to a TS at rs = 90 AU, and the dotted line to a TS at rs = 100 AU. The electron data from Pioneer 10 are presented as shaded areas for radial distances up to

- 224 -

Effects o f the heliospheric termination shock on ...

Figure l a" Computed radial profiles of 16 MeV galactic electron intensities with an outer boundary at rB = 120 AU, and with the IS of Strong et al.[5] assumed to be electron local spectra. The solid line corresponds to a TS at rs = 80 AU, the dashed line to a TS at rs = 90 AU, and the dotted line to a TS at rs - 100 AU. The electron data from Pioneer 10 are presented as shaded areas, lb" The same as in la, but with the IS from the recent calculations of Strong et al. [1]. --50 AU [4], and at --70 AU [6]. The height of the shaded areas incorporate the error bars present in the data, as well as a small time dependent effect due to the solar modulation of the electrons over the period --1972 to --1991. The data are assumed to be of galactic origin, with the Jovian component dominating only for r < 25 AU [4]. The computed intensities are compatible with both data sets. We assumed Ko = 0.5, a - 0.25 and b = 0.6. The effect of the shock is visible in all the radial profiles, indicating an increase in the radial gradient upstream of the shock but a decrease beyond the shock. For a TS at rs -- 80 AU, the effect of the shock on the radial gradients upstream and downstream of the shock is significantly larger than for the two other scenarios. Although the shock is more effective for larger radial distances due to the larger shock radius, the radial diffusion coefficient, due to its radial dependence oc r, is also larger, leading to a smaller effect for the more distant shocks. Figure l b shows the case when the more recently calculated IS of Strong et al. [1] is assumed as the local electron spectrum. Evidently, it is considerably lower than the IS used in Figure la. All three scenarios for the shock locations are again compatible with the observed Pioneer 10 data, but in order to obtain that, we had to assume Ko = 58, a=18 and b=24 in Equation 3. These larger diffusion coefficients are needed in order to produce less modulation between the outer boundary and the data a t - 7 0 AU. The effect of the different TS locations on the computed electron intensity profiles is much less pronounced, with almost no difference between the three scenarios, in contrast to Figure la. These larger diffusion coefficients clearly decrease the effectiveness of the shock.

- 225 -

S.E.S. Ferreira, M.S. Potgieter and U.W. Langner

4. SUMMARY AND CONCLUSIONS Two different scenarios for the electron IS ([1],[5]), and their subsequent modulation in the heliosphere have been studied using a shock-drift modulation model. The effects of the heliospheric termination shock (TS) on each of these scenarios were also studied. At 16 MeV the two electron IS differ by a factor of--100. They were assumed as the local IS at 120 AU, as an outer boundary of the heliosphere. Compatibility between the observed Pioneer 10 ([4],[6]) radial profiles and the model simulations were strictly required for the two IS cases, and the three different TS shock positions, as shown in Figure 1. For the highest IS of Strong et al. [5], very large radial gradients (-~10%/AU) were found between 70 AU, the shock position rs, and the outer boundary. The effect of the shock on the radial intensity profile is more pronounced with rs = 80 AU, than for rs = 90 AU or 100 AU. The radial gradients are found to be larger upstream of the shock than downstream. Although the TS should more effective for larger radial distances due to the larger shock radius, the diffusion coefficients are also larger, leading to a smaller modulation effect for the more distant TS positions. For the lowest IS of Strong et al. [1], the radial gradients beyond 70 AU were significantly less, with the effect of the different TS locations on the computed electron intensities far less pronounced than in the first case. The reason is that larger diffusion coefficients are needed to provide compatibility with the data. These larger diffusion coefficients not only decrease the total modulation in the outer heliosphere, but also the effectiveness of the shock. For the interstellar spectra given in Figure 1, the computed radial intensity profiles for--16 MeV electrons indicate that the detection of the crossing of the TS by a spacecraft registering only these low energy electrons would be ambiguous. Some other observations have to be utilized (e.g. magnetic field or solar wind speed changes) in order to provide information on when a spacecraft actually crosses the TS. REFERENCES ~

2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

A.W. Strong et al., Astrophys. J. 537 (2000) 763. L. J. Haasbroek et al., Proc. 24th ICRC 4 (1995) 706. S. E. S. Ferreira et al., J. Geophys. Res. 105 (2000) 18305 C. Lopate, Proc. 22nd ICRC (Dublin) 2 (1991) 149. A. W. Strong et al., A&A 292 (1994) 82. C. Lopate, private communication (2000). E. N. Parker, Planet. & Space Sci. 13 (1965) 9. J. A. le Roux et al., J. Geophys. Res., 101 (1996) 4791. L. J. Haasbroek, Ph.D. thesis, Potchefstroom University, South Africa (1997). J. K6ta and J.R. Jokipii, Proc. 24th ICRC (Rome) 4 (1995) 680. M. S. Potgieter, J. Geophys. Res. 101 (1996) 24411. J. R. Jokipii and J. K6ta, Geophys. Res. Lett. 16 (1989) 1. A. Balogh et al., Science. 268 (1995) 1007.

- 226-

Oral papers and posters

G A L A C T I C C O S M I C RAYS: O V E R V I E W M.A. Forman Department of Physics and Astronomy, State University of New York at Stony Brook, USA. Cosmic rays from the galaxy bring us clues about energetic processes in stars and in the interstellar medium in our galaxy. They are also one of the important probes of the magnetic conditions all over the heliosphere, and its connection to the interstellar medium. They have been there. Over 50 years ago, before the space age, cosmic ray detectors on Earth revealed that the cosmic rays were positively charged and had energies from at least 5 10 8 electron volts to over 10 18 . 11-year solar modulation of cosmic rays (with energies less than about 10 10 eV) was discovered, and more types of cosmic ray decrease also associated with increased solar activity on time scales of the solar cycle, the solar rotation and after transient solar events. Clearly the Sun was doing some powerful electromagnetic work out there, all the time, in a huge volume of space extending at least to the orbit of Earth, and most likely well beyond. Heliospheric cosmic ray research was one of the first scientific fields to benefit from access to space, because of its obvious need to be in space and because its visionary early experimentalists made rugged instruments ready. We have a pretty good idea of what causes the modulation: large scale moving magnetic fields and cosmic-ray coupling to the ambient plasma through magnetic turbulence. We have a good idea of the distribution of cosmic rays inside 100 AU. We should remember that galactic cosmic rays have already visited the furthest parts of the heliosphere, on all possible routes and tell us about the heliosphere and beyond.

G A L A C T I C C O S M I C RAYS: T H E O U T E R H E L I O S P H E R E J. R. Jokipii University of Arizona, Tucson, AZ 85721 USA.

However, we are still not at the point where we can reliably extrapolate to the interstellar spectrum at energies less than several hundred MeV. A significant problem is our lack of understanding of the outer heliosphere and the interface with the interstellar gas. The use of cosmic rays as remote probes of this region is very important. The current status of our understanding and the results of recent modelling efforts will be discussed, with emphasis on processes ocurring in the outer heliosphere.

- 227 -

Oral papers and posters

MODULATION RAYS

OF G A L A C T I C

AND ANOMALOUS

COSMIC

E. R. C h r i s t i a n (1), W. R. Binns (2), C. M. S. Cohen (3), A. C. Cummings (3), . S. George (3), P. L. Hink (4), R. A. Leske (5), R. A. Mewaldt (5), E. C. Stone (5), T. T. von Rosenvinge (6), M. E. Wiedenbeck (7) and N. Yanasak (7) (1) NASA GSFC Code 661, (2) Washington U., St. Louis, (3) Caltech, (4) Washington U., St. Louis, (5) Caltech, (6) NASA GSFC Code 661, (7) PL. The temporal history of cosmic ray intensities at 1 AU is an important component to the understanding of solar modulation. The large collecting power and high resolution of the Cosmic Ray Isotope Spectrometer (CRIS) and the Solar Isotope Spectrometer (SIS) instruments on the Advanced Composition Explorer (ACE) allow us to investigate the changing modulation on short time scales and over a wide range of rigidities. With these data, we will present the di erences between the short term and long term e ects and the correlation of these e ects with magnetic field, current sheet tilt angle, and other phenomena. The data span the period from the launch of ACE in August 1997 to the present.

T H E S P E C T R U M O F A C R O X Y G E N A N D ITS V A R I A T I O N S I N T H E O U T E R H E L I O S P H E R E F R O M 1992 T O 2000 D.C. H a m i l t o n (1), M.E. Hill (1), N.P. Cramer (1), R.B. Decker (2) and S.M. Krimigis (2) (1) Department of Physics, University of Maryland, (2) Johns Hopkins Applied Physics Laboratory. The Voyager 1/2 LECP instruments are measuring the anomalous cosmic ray oxygen spectrum (0.3 - 40 MeV/nuc) in the outer heliosphere and have observed large intensity variations, particularly in the low energy portion of the spectrum. At Voyager 1, the peak ACR oxygen flux was observed at a nearly constant energy of 1.3 MeV/nucleon during the years 1993 to 1999 as V1 traveled from 53 AU to 76 AU. The flux at that energy increased by a factor of about 100 from 1992 to 1999, reflecting a large decline in modulation. In fact the ACR flux at V1 and V2 continued to increase through 1997, 1998, and into 1999, well after the nominal minimum of the 11-year solar activity cycle in 1996. At Voyager 2 (at 60 AU near the beginning of 2000) it appears that the oxygen flux near the 1.3 MeV/nucleon ACR peak has begun to decrease, declining by about a factor of three from its maximum in mid-1999 to February 2000. At Voyager 1 the picture is less clear. Although the flux has decreased since September 1999, this decrease appears to be in phase with a series of quasi-periodic variations observed over the last three years. These variations at V1 have a period of about one third year and an amplitude of about 50%. It will soon become clear whether this latest downturn is part of that sequence or is the beginning of an increase in long-term modulation. Differences in ACR behavior at V1 and V2 are interesting because V1, at 34 degrees north heliolatitude, is beyond the current sheet while V2, at 21 degrees south, is still within but near the limit of current sheet excursions.

-228-

Oral papers and posters

O N T H E V A R I A B I L I T Y OF S U P R A T H E R M A L 1 AU

P I C K U P H E + AT

B. Klecker (1), A.T. Bogdanov (1), M. Hilchenbach (2), A.T. Galvin (3), E. MSbius (4), F.M. Ipavich (4) and P. Bochsler (5) (1) Max-Planck-Institut fr extraterrestrische Physik, D-85740 Garching, Germany, (2) Max-Planck-Institut fr Aeronomie, D-37819 Katlenburg-Lindau, Germany, (3) University of Maryland, College Park, Md, USA, (4) University of New Hampshire, Durham, NH, USA, (5) University of Bern, CH-3012 Bern, Switzerland. Using data from the STOF experiment onboard SOHO we investigate the variation of suprathermal He+/He 2+ abundances in the energy range 85-280 keV during the years 1997 to 1999. We observe a large variability of the He + abundances ranging from He+/He 2+ < 5% to ,.~ 1. The very large abundances are closely related with the passage of interplanetary shocks. Combining the data from STICS/WIND and STOF/SOHO we are able to identify in these events a pickup He + distribution with the typical cutoff energy at twice the solar wind velocity and a suprathermal tail extending to a few 100 keV. We correlate daily averages of the He + abundances of the suprathermal tail for all days with significant He + flux with solar wind parameters and find a general anticorrelation of He + abundances with solar wind velocity and solar wind thermal velocity. We discuss possible causes of this variability, in particular variations of the source strength of pickup ions and solar wind alphas and variations of the acceleration efficiency for He + and He 2+.

OBSERVATIONS OF PICK-UP IONS IN THE OUTER HELIOSPHERE BY VOYAGERS 1 AND 2 AND IMPLICATIONS ON PRESSURE BALANCE S.M. Krimigis and R.B. Decker Applied Physics Laboratory, Johns Hopkins University, Laurel, MD 20723. Observations from the Low Energy Charged Particle (LECP) instrument on the Voyager 1 (V1) spacecraft at NTOAU during the last solar minimum revealed an excess of counts from the sunward direction in the nominal proton energy range ~40 to ~ 140keV. This, above background, response was also seen by Voyager 2 at ~55AU and was present during the previous solar minimum at V1 at ~28AU. The angular distribution is inconsistent with these counts being due to hot protons convected into the sunward-viewing sector only. We have examined possible sources for the observed counts and have considered that they may be due to singly-ionized heavy (A_>16) ions picked up by the solar wind and convected into the sunward sector of the detector. The spectrum is steep (dj/dE c< E-6), and exhibits a cutoff at ,-~25keV/nuc. The intensities are within range of predictions of stochastic pre-acceleration or phase-space diffusion of interstellar pickup ions in the solar wind (e.g. Le Roux and Ptuskin, 1998; Chalov and Fahr, 1998). Preliminary estimates show that pickup oxygen pressure is c( 10-13 dynes cm -2 at ~70AU, comparable to the extrapolated pickup H+ pressure at 80AU (Whang et al, 1999), at the putative location of the termination shock. The observations suggest that pickup oxygen is present beyond 5AU and, along with other interstellar species, could be the pre-accelerated seed population for anomalous cosmic rays (ACR). LeRoux and Ptuskin, JGR, 103, 4799, 1998. Chalov and Fahr, Astron. Astrophys., 335, 746, 1998. Whang et al, JGR, 104, 28255, 1999.

- 229-

Oral papers and posters

CONSEQUENCES OF RECENTLY OBSERVED GALACTIC SYNCHROTRON RADIO EMISSIONS ON THE LOCAL INTERSTELLAR SPECTRUM FOR COSMIC RAY ELECTRONS U.W. Langner, O.C. de Jager and M.S. P o t g i e t e r School of Physics, Potchefstroom University, Potchefstroom. South Africa. Propagation models for galactic electrons give a synchrotron spectral index which is larger than the recently measured radio index between 22 to 408 MHz. Diffuse gamma-ray data appear to be "contaminated" by Crab-like point sources, so that it is difficult to derive a consistent local interstellar spectrum (IS) for electrons in the 1 to 30 MeV range. Using a phenomological approach, we show that the synchrotron spectral indices calculated from the best-fit IS of Strong and Moskalenko (Proc. 5th Comp. Symp., 1999) - for a 3 to 5 mG field agree well with the spectral indices calculated from their full propagation model in the frequency range of interest. This allowed us to introduce an adjusted local IS, such that the model radio spectral index agrees with observations above 20 MHz. By adding the constraints expected from galactic modulation, we find that the local IS at ,.~4 MeV is marginally above the lower limit for a local IS set by the Pioneer 10 electron data observed in the outer heliosphere

V A R I A T I O N O F T H E F L U X E S O F E N E R G E T I C H E + A N D H E 2+ DURING THE PASSAGE OF CO-ROTATING INTERACTION REGIONS D. Morris (1), E. MSbius (1), M.A. Popecki (1), L.M. Kistler (1), A.B. Galvin (1), B. Klecker (2) and A. Bogdanov (2) (1) (Dept. of Physics and Space Science Center, University of New Hampshire, Durham, NH), (2) (Max-Planck-Institut fr extraterrestrische Physik, Postfach 1603, D-85740 Garching, Germany). With the ACE SEPICA instrument detailed observations of the variation of the energetic He + and He 2+ fluxes have been obtained during the passage of several co-rotating interaction regions (CIR) in 1999 and 2000. For all CIRs under investigation the He+/He 2+ ratio increases consistently from the start of the event towards the end. However, the absolute flux of the energetic ions usually reaches a maximum close to the beginning of the event. Due to the solar rotation and related motion of the spacecraft relative to the CIR, the spacecraft is magnetically connected to the compression region and related shocks between the two different solar wind streams at distances from the sun, which increase with time. Therefore, the increasing He +/He 2+ ratio can be interpreted as an increase of the relative importance of the interstellar gas over the solar wind as a source for the observed energetic ions. The fact that the maximum flux of the energetic ions is observed close to the beginning of the event, supports the recent finding by Chotoo et al. (2000) that efficient particle acceleration also occurs in a region, where no shock has formed yet. Chotoo, K., et al., J. Geophys. Res., in press, 2000.

-230-

Oral papers and posters

THE COSMIC RAY ELECTRON TO POSITRON RATIOS IN THE HELIOSPHERE M.S. P o t g i e t e r and U.W. Langner School of Physics, Potchefstroom University, Potchefstroom. South Africa. The heliospheric modulation of cosmic rays disguises the true spectral form of the local interstellar spectra (LIS) of all cosmic ray species below about 10 GeV. The lower the energy, the more uncertain they seem to become which is especially true for cosmic ray electrons. Recent modeling of the propagation of cosmic rays through the Galaxy by Moskalenko and Strong (Ap. 493, 694, 1998) indicates that the LIS for positrons may be known more reliably than before. Using this information, and the di erent scenarios for the electron LIS, the electron to positron ratios, as modulated through the heliosphere, are computed with a comprehensive numerical drift model. These results can be of use for future missions to the outer heliosphere and beyond, and may assist in establishing the true electron LIS.

-231

-

This Page Intentionally Left Blank

General Discussion Forman to Moraal: I would like to hear some more ideas about the low energy cosmic ray spectrum, what you might learn from that and what it might be. And also the anomalous cosmic rays - do they have higher intensity than the low energy cosmic rays or not? Moraal: As far as the low energies are concerned, if you believe anything about the acceleration mechanism, then you cannot make a distinction between anomalous and galactic cosmic r a y s - anything at low energy should sort of go up like p-4, whether you should call it anomalous or galactic cannot be said. Fahr to Jokipii: What could bc thc rcason, in your vicw, for thc filamcntcd and vcry much disordered structure of the magnetic fields downstream of the heliospheric shock? Jokipii: Remember that the whole structure of the inner heliosphere now is quite consistent with the picture of braided magnetic fields caused by the supergranular motion on the surface of the Sun. If Len Fisk's very interesting field is there, it would also do that and, also, reconnection can occur. Those fluctuations grow in relative importance with radius because of the transverse expansion. And you can show from the number that we think of pretty firm that by the time you get to 100 AU the field line has meandered essentially over 7r radians. And this gets carried out into the heliosheath. Lee to Dorman: How is the distance to the modulation boundary determined? You need to know the diffusion coefficients. Dorman: No, we used only the average velocity. Gruntman to Moskalenko: What is the origin of these anti-protons? Moskalenko: We calculated the anti-proton flux from the interaction of galactic cosmic rays with interstellar gas. So, it's galactic origin on a large scale, the scale of the Galaxy. Marsch: Maybe a general question concerning the diffusion picture. The impression I got from Randy Jokipii's talk is that the Parker equation essentially describes it all. But is this diffusion picture really describing everything that one should look at? I think that there is a lot of convected structure in the flow on mesoscales where usually the features are not easily describable in terms of fluctuations and waves. I am not talking about the convection by the wind in general, there is structure in the wind itself, all sorts of transient features. Lerche: Maybe Gene [Parker] wants to talk about his own equation? Parker: I quite agree [to the existence of convected structures]. That's why there is a convective term in the equation. Also the velocities of structures in the solar wind are in the equation. Jokipii: The real question is the scale of the variation. As soon as you are talking about structures like corotating interaction regions which have scales greater than a fraction of an AU, than you can describe them very well with this transport equation. If we get to very small scales and long mean-free-paths near the S u n - to solar flare p a r t i c l e s - then you have to be a little bit more careful. But for the galactic and anomalous cosmic rays there is no real difficulty with using the equation. And the convcctcd structures arc all

- 233 -

General Discussion

in there. Veselowsky: From the discussion we understand that the equations are sound and reliable. Also the inner boundary conditions near the Sun are under control. But we understand that this is different for the external boundary conditions at the so-called modulation boundary, this mysterious surface. And what we need there is a real understanding of the space-time structure of the LISM. Cummings: Yes, and one question that comes to my mind is: how far do we need to go? Based on what I saw today there seems to be different opinions about how far out you need to get to the undisturbed interstellar medium. Forman: A general question regarding the modelling: Why didn't many observers adopt more sophisticated models including drifts but are rather using simple ones like force field or spherically-symmetric models? Doesn't it make any difference? Wiedenbeck: I think the observers are simply not up to speed on any of these sophisticated codes. I am not sure that they have been put into a form where one can readily use them for interpreting data. Lee: It's just puzzling here, that that we don't see any changes in the nature of modulation. I would have thought that as we get to the termination shock it changes dramatically because the flow beyond the shock is divergence-free. And so, a force-field approximation doesn't apply there. Lerche: Well, or the shock is much further out than you think it is. Veselowsky: The region you are talking about is very complicated in structure because it's keeping the memory of previous solar events. Giacalone to le Roux: A shock would have a foot and a ramp, perhaps an overshoot and some steady oscillations downstream. You discussed a fine structure of the termination shock, and I am not sure what you mean by that? le Roux: From bow shock observations, where they have really perpendicular cases [comparable to the termination shock], it looks that the shock ramp has details. Although the ramp thickness was found to be equal to about an ion inertial length, there are fluctuations insidc thc ramp of much narrowcr lcngth providing a finc structurc. Zank to Czechowski: Is there a sensitivity of the flux of anomalous cosmic rays to the actual structure of the assumed heliosphere? Czechowski: Yes, it is quite sensitive to it.

- 234-

Session 4:

Echoes from the Heliopause

This Page Intentionally Left Blank

Energetic Neutral A t o m I m a g i n g o f the Outer H e l i o s p h e r e - L I S M Interaction Region H.O. Funsten a, D.J. McComas b, and M. Gruntman c a Los Alamos National Laboratory, MS D466, Los Alamos, NM 87545, USA* b Southwest Research Institute, P.O. Drawer 28510, San Antonio, TX 78228, USA c University of Southern California, Los Angeles, CA 90089-1191, USA

Energetic neutral atom (ENA) emission from the region where the heliosphere and the local interstellar medium (LISM) interact will enable remote observation of this region's structure and dynamics. Imaging of these ENAs will allow us to understand the complex physics of the interaction and will help distinguish between competing models of the region. In this paper, we describe the imaging technique of ENA ionization by an ultrathin transmission foil followed by electrostatic deflection and coincidence detection. We also show new laboratory results that demonstrate the ability of this technique to detect ENAs over the energy range of approximately 0.2-6 keV that is critical for understanding the physics of the processes that govern the interaction region. This technique has a large intrinsic geometric factor that will enable imaging of the extremely dim ENA emissions and will allow quantification of the important characteristics of the outer heliosphere-LISM interaction region.

1. INTRODUCTION The interaction of the heliosphere with the local interstellar medium (LISM) is believed to generate energetic neutral atoms (ENAs) that result from charge exchange of hot plasma ions, predominantly hydrogen, with cold interstellar neutrals [1-4]. Once created, these neutral atoms follow ballistic trajectories from the source region and can be remotely detected within the inner heliosphere. These ENAs carry important information about processes occurring at their source region. Models of the interaction region (see [4] and references therein) predict predominant ENA emission at several hundred eV, although at very low fluxes. With increasing energy, the ENA flux generally decreases at a rate that is strongly dependent on the LISM parameters and the physics of the interaction. Imaging of ENAs with an energy Eo greater than-200 eV will provide a sensitive measure for determining the physics of the interaction region, including

* This work was performed under the auspices of the United States Department of Energy. The authors gratefully thank Ed Roelof (APL/JHU) and Hans Fahr (U. Bonn) for numerous insightful discussions on ENA emission from the outer heliosphere.

- 237-

H.O. Funsten, D.J. McComas and M. Gruntman

the type of shock and the evolution of pickup ions, and will distinguish between the various competing models. The primary challenge of ENA imaging is accurate measurement of ENA trajectory and energy against large EUV/UV and charged particle backgrounds. Several ENA imaging techniques covering different energy ranges have been developed to overcome this challenge [5,6]. On the recently successful IMAGE mission, three different imaging techniques were used to measure the ENA emission [7] from the terrestrial magneto-sphere at high time resolution across three different energy ranges: the High Energy Neutral Atom (HENA) Imager utilized a thick foil to block UV, allowing ENAs over 15 keV to pass to the detector section [8]; the Medium Energy Neutral Atom (MENA) Imager utilized freestanding transmission gratings to block UV while allowing a fraction of ENAs to pass to the detector section [9,10]; and the Low Energy Neutral Atom (LENA) Imager employed a reflection surface that ionized a fraction of ENAs followed by electrostatic deflection of ionized ENAs into the detector section [ 11]. These instruments have demonstrated the powerful technique of ENA imaging for viewing global morphology and dynamics of the magnetosphere. However, due primarily to the scientific thrust of the IMAGE mission on the terrestrial magnetosphere, these imagers were not designed to focus on the energy range of several hundered eV to several keV that is critical for imaging of the heliosphere-LISM interaction region. The ENA imagers on IMAGE were designed to detect the comparatively copious ENA fluxes from the magnetosphere at high time resolution and would need to be substantially larger to measure the extremely dim ENA emission from the heliosphere-LISM interaction region. In this paper, we investigate the feasibility of detecting ENAs that are emitted from the heliosphere-LISM interaction region over an energy range of 0.2 to 6 keV. The technique described here of ENA ionization via transmission through an ultrathin foil has certain advantages over other techniques at this energy range, including a geometric factor that is large enough to measure the extremely small ENA flux from the heliosphere-LISM interaction region. This technique has been investigated extensively at somewhat higher energies [ 12-15] but not below several keV.

2. NEUTRAL ATOM IMAGER FOR VIEWING THE INTERACTION REGION The measurement objectives for imaging ENAs from the heliosphere-LISM interaction region include: a large enough geometric factor for statistically significant measurements in one day, an energy resolution AE/E < 1, an imaging resolution in the range of 5o to 10~ and extremely high signal-to-noise ratio for clear identification of ENAs. These objectives are met with a single-pixel imager based on the technique of ENA ionization using an ultrathin foil [12,15]. An imager of this type is shown schematically in Fig. 1. An ENA enters the first set of collimators that consists of a set of alternately biased plates to reject solar wind ions up to an energy-per-charge of several 10s of keV/e from the ENA measurement. ENAs then pass through a second set of collimators that collimate in a direction orthogonal to the first set of collimators. The two collimators result in a single pixel field-of-view (e.g., 7~176 A fraction of the ENAs that pass through the collimator and transit the first ultrathin carbon foil, labeled F1, would be ionized. These ionized ENAs then transit a hemispherical electrostatic energy analyzer (ESA) if their exit energy lies within the selected ESA energy passband. In addition to its energy resolving capabilities, the ESA also serves to prevent UV

- 238-

Energetic neutral atom imaging of the outer heliosphere... light from entering the detector section. The ionized ENAs then enter the detector section that consists of a second ultrathin carbon foil (F2) for generation of secondary electrons that are detected separately from the ENAs for a coincidence measurement. For the energy range 0.2-6 keV, a single pixel instrument is superior to a true imaging instrument having multiple pixels due to significant ENA scattering in foil F1 that results in loss of information of the incident ENA trajectory. The scattering half-angle aPl/2 of ionized ENAs emerging from this foil is related to the incident energy according to apl/2 ~- 13 [keV-deg] / E0 Fig. 1. The ENA imager shown schematically in the [14]. For example, at 1 keV this figure is optimized to measure the faint ENA emission corresponds to an angular FWHM of from the heliosphere-LISM interaction region. The 26 ~ which is far larger than the desired ENA ionization foil can span enormous aperture areas, angular imaging resolution o f - 7 o. This enabling a large geometric factor. As an example, this loss of knowledge of the incident ENA imager would have an aperture area of 64 cm2, a singletrajectory precludes use of a true pixel FOV of 7~ ~ and six contiguous energy passbands covering 0.2-6 keV. imaging instrument such as the ENA imagers on IMAGE. The optimum implementation therefore uses a single pixel instrument in which the field of view is collimated and does not depend on scattering in the foil. The ENA imager could be placed so that its single pixel view is orthogonal to the spin axis on a sun-pointed spinning spacecraft, resulting in a 7~ ~ view each spin. The full heliosphere would be completely viewed twice over each orbit of the spacecraft around the Sun. Broader coverage could be obtained using several single pixel instruments placed at different viewing angles relative to the spacecraft spin axis. An important feature of the imager is that the foil F1 used for ENA ionization can be biased to a high positive potential VVl. ENAs that exit the foil as positive ions (e.g., H +) are accelerated perpendicular to the foil, resulting in three important advantages for enhanced ENA imaging. First, the large angular scattering at low energies, which would normally result in the loss of ionized ENAs by scattering out of the angular acceptance passband of the ESA, is mitigated by this "proximity focusing" effect: the final trajectories after acceleration are nearly perpendicular to the foil, and so these scattered ions are pushed back into the angular acceptance of the ESA and are transmitted through the ESA. Second, the accelerated ENAs transit the detector section at a much higher energy so that (a) the secondary electron yield at the second foil is higher and (b) the ENA detection probability is high. These two features combine to greatly increase the probability of coincidence detection (and therefore increase the geometric factor). Third, the ionized ENAs are accelerated to a higher energy at which the ESA energy passband is broader. This results in a larger apparent energy resolution AE/Eo of the measurement and a proportional increase in the geometric factor.

- 239-

H.O. Funsten, D.J. McComas and M. Gruntman

We illustrate this last effect by considering a spherical ESA having an intrinsic energy resolution AE/Ec = k where Ec is the central energy of an energy passband of width AE and k is a constant dependent only on the ESA geometry. ENAs with incident energy E0 are ionized by the foil, accelerated due to the foil bias, and enter the ESA at energy E0+ q VF1 (q is the ion charge). By equating Ec = E0+ q VF~, the apparent energy resolution AE/Eo from the perspective of the ENAs is

AEEo= k ( l + q V .FE1o)

(1)

The apparent energy resolution is clearly dependent on qVF1/Eo. If qVF1/Eo __ f and angular scattering by density irregularities increases as f ~ fp with a rate or ( 1 - f~/f2)-2 [42]. Figure 4 shows the spatial variations in h ( X , 0) from recent 2-D global simulations [39], where X is along the Sun-nose axis (the symmetry axis) and 0 is the associated polar angle. Density increases at the termination shock, heliopause and outer bow shock are clear. Now superpose a strong GMIR shock (density jump of 4) onto the fp(X, 0 - 0) profile. Foreshock fp radiation is then clearly trapped (by reflection and

-259-

1.H. Cairns and G.P. Zank

-" .

:

"--..

| ~

*%.,.

......

f i

i

1

Figure 4. Profiles fp(X, O) from 2-D, cylindrically-symmetric global simulations [16,39].

scattering) between the shock and the slow density increase at larger R, remaining within a few AU of the shock. However, the slow decrease in fp(R, O) as 0 increases means that the radiation diffuses to larger 0 around the side of the shock and to smaller R closer to the heliopause. Eventually the radiation reaches locations where the shock is sunwards of the heliopause, so that fp 10 nT, well above current estimates, these ideas require further consideration but are potentially very important. 7. C O N C L U S I O N S The GMIR model for the Voyager radio emissions [4], based on observational data, is qualitatively compelling. A new theory [16] provides an underlying theoretical basis for the GMIR model: using the lower hybrid drive mechanism, known radiation processes, scattering, and propagation effects, it removes many previously unexplained theoretical issues with the GMIR model. These include why and where the radiation turns on in the outer heliosheath, the frequencies of the observed radiation, and how radiation reaches the inner heliosphere. The theory provides multiple predictions suitable for observational testing but is not complete. In particular, explanations for the two classes of emission and frequency fine structures remain primitive although several promising research directions exist. Certain previous theories involving emission just sunward of the termination shock or in the inner heliosheath are not viable for the observed emissions but may be relevant to presently undetected emissions that exceed detection thresholds closer to their sources. Financial support from the Australian Research Council, NASA grants NAG5-7390 and NAG5-7796, and JPL contract 959167 is gratefully acknowledged. REFERENCES

1. W.S. Kurth et al., Nature 312 (1984) 27. 2. W.S. Kurth et al., Geophys. Res. Lett. 14 (1987) 49. 3. W.S. Kurth, in Proc. Sixth Int. Solar Wind Conference, Vol. II, NCAR/TN-306+Proc (1988) 667. 4. D.A. Gurnett et al., Science 262 (1993) 199.

-261 -

I.H. Cairns and G.P. Zank

D.A. Gurnett and W.S. Kurth, Adv. Space Res., 16(9) (1995) 279. D.A. Gurnett, S.C. Allendorf, and W.S. Kurth, Geophys. Res. Lett., 25 (1998) 4433. D.A. Gurnett, this volume (2001). McNutt, R.L., Jr., Geophys. Res. Lett., 11 (1988) 1307. A. Czechowski and S. Grzedzielski, Nature 344 (1990) 640. W.M. Macek et al., Geophys. Res. Lett. 18 (1991) 357. I.H. Cairns and D.A. Gurnett, J. Geophys. Res. 97 (1992) 6235. I.H. Cairns, W.S. Kurth, and D.A. Gurnett, J. Geophys. Res. 97 (1992) 6245. Y.C. Whang and L.F. Burlaga, J. Geophys. Res. 99 (1994) 21,457. G.P. Zank et al., J. Geophys. Res. 98 (1994) 14,729. I.H. Cairns and G.P. Zank, Geophys. Res. Lett. 26 (1999) 2605. I.H. Cairns and G.P. Zank, Astrophys. J. submitted (2001). N.R. Labrum and D.J. McLean (eds), Solar Radiophysics, Cambridge U. Press, Cambridge, 1985. 18. D.B. Melrose, Plasma Astrophysics, Gordon & Breach, New York, 1980. 19. C.S. Wu and L.C. Lee, Astrophys. J. 230 (1979) 621. 20. P. Zarka, J. Geophys. Res. 103 (1998) 20,159. 21. H.J. Fahr et al., Space Sci. Rev. 43 (1986) 329. 22. K.G. Budden, Radio Waves in the Ionosphere, Cambridge U. Press, Cambridge, 1961. 23. I.H. Cairns and P.A. Robinson, in Radio Astronomy at Long Wavelengths, Geophys. Monograph 119, AGU, Washington (2000) 27. 24. J.P. Wild, Aust. J. Sci. Res., A3 (1950) 541. 25. M.J. Reiner and M.L. Kaiser, J. Geophys. Res. 104 (1999) 16,979. 26. S.D. Bale et al., Geophys. Res. Lett. 26 (1999) 1573. 27. D.A. Gurnett, J. Geophys. Res. 80 (1975) 2751. 28. S. Hoang et al., J. Geophys. Res., 86 (1981) 4531. 29. I.H. Cairns, J. Geophys. Res. 93 (1988) 3958. 30. G.P. Zank Space Sci. Rev. 89(3/4) (1999) 1. 31. P.J. Filbert and P.J. Kellogg, J. Geophys. Res. 84 (1979) 1369. 32. I.H. Cairns, J. Geophys. Res. 92 (1987) 2315. 33. P.A. Robinson and I.H. Cairns, Astrophys. J. 418 (1993) 506. 34. P.A. Robinson and I.H. Cairns, Sol. Phys. 181 (1998) 395. 35. S. Hoang et al., in Solar Wind 7, Pergamon Press, Oxford (1992) 465. 36. J.B. McBride et al., Phys. Fluids 15 (1972) 2367. 37. Y.A. Omelchenko et al., Sov. J. Plasma Phys. 15 (1989) 427. 38. V.D. Shapiro et al., Physica Scr. T75 (1998) 39. 39. G.P. Zank et al., J. Geophys. Res. 101 (1996) 21,639. 40. H.L. Pauls and G.P. Zank, J. Geophys. Res. 101 (1997) 17081. 41. T. Linde et al., J. Geophys. Res. 103 (1998) 1889. 42. J.-L. Steinberg, S. Hoang, and C. Lacombe, Astron. Astrophys. 7 (1985) 151. 43. J.W. Belcher et al., J. Geophys. Res. 98 (1993) 15,177. 44. G.P. Zank, in Solar Wind 9, AIP Conf. Proc. 471 (1999) 783. 45. I.H. Cairns, J. Geophys. Res. 99 (1994) 23,505. o

6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

- 262-

M a p p i n g the H e l i o p a u s e in E U V Mike Gruntman Department of Aerospace and Mechanical Engineering, MC-1191 University of Southern California, Los Angeles, CA 90089-1191

We know very little about the heliopause, a boundary that separates the solar wind and the galactic plasma of the local interstellar medium (LISM), with the direct experimental data next to nonexistent. We propose to explore the heliopause remotely, from 1 AU by an observer outside of the geocorona. Interstellar plasma ions beyond the heliopause would glow under solar extreme-ultraviolet (EUV) radiation in the resonance lines of oxygen (83.4 nm) and helium (30.4 nm). The measurements of this glow would map the heliopause. Heliopause mapping in EUV is a way to remotely explore the heliospheric interface region and the LISM ionization state and to probe the asymmetry of the interstellar magnetic field. 1. H E L I O P A U S E

The interaction of our star, the Sun, with the surrounding local interstellar medium (LISM) leads to the buildup of the heliosphere, the region where the Sun controls the state and behavior of the plasma environment. The heliosphere is a complicated phenomenon where solar wind and interstellar plasmas, neutral interstellar gas, magnetic fields, anomalous and galactic cosmic rays, and energetic neutral atoms play prominent roles (Figure 1). Experimental data on the sun-LISM interaction region are exceptionally scarce, Figure 1. Possible solar wind interaction (twomostly indirect and often ambiguous. We shock model) with the LISM: TS - termination know very little about the heliopause, a shock; HP-heliopause; BS - bow shock; CRboundary that separates the solar wind cosmic rays; ISP(G) -interstellar plasma (gas); plasma and the interstellar plasma. 1 B - magnetic field. Interstellar plasma flows Many important questions remain outside the heliopause (gray area). Angle 0 is unanswered, for example" What is the counted from the upwind direction. distance to and shape of the heliopause? What is the ionization state of interstellar gas in the LISM? What is the direction and magnitude of the interstellar magnetic field? Is the interstellar wind subsonic or supersonic? Is a bow shock formed in front of the heliosphere?

- 263 -

M. Gruntman

Voyager 1, now approaching 80 AU, is expected to cross the termination shock and explore the heliospheric sheath properties in one point-direction. The spacecraft may not however reach the heliopause by the end of the mission in ---2025. The interstellar wind, solar wind latitudinal variations, and interstellar magnetic field make 10E+01 1 TS 'P ' the heliosphere asymmetric. Only remote ~ ]1 H ] l techniques, complemented by the "ground truth" ~ ~ I~ Voyager and future Interstellar Probe in-situ o can provide a global view of the ~ time-varying three-dimensional heliosphere on a ~ ' Iv continuous basis. ~ ~ ,~ The heliopause separates the interstellar and ~ 1.05-03 ~ ., solar wind plasmas with the number densities 2 orders of magnitude different. Figure 2 illustrates 1.05-04 ............................ ~.................. the dependence of the plasma number density 2 0 100 200 300 400 50c on the heliocentric distance in the approximately heliocentric distance, AU upwind direction (with respect to the interstellar Figure 2. Typical plasma number density wind). The plasma density rapidly decreases with in the upwind direction. The arrows the expansion of the solar wind. Interstellar indicate the positions of the termination plasma cannot cross the heliopause and flows shock (TS), heliopause (HP), and bow around it. One can imagine the sun surrounded shock(BS). by an interstellar ion "wall" beyond the empty cavity, the "heliopause moat," limited by the heliopause boundary. Is the heliopause stable under such conditions? This heliopause moat suggests a way of remote, from 1 AU, mapping of the heliopause. 3'4 Singly charged interstellar ions (He +, O +) would scatter the corresponding solar extremeultraviolet (EUV) line emissions. Measurements of this scattered radiation, the LISM plasma glow, would open an access to the heliopause and the region beyond. Heliopause imaging would map the heliopause and provide an important insight into the LISM ionization state and the asymmetry of the interstellar magnetic field. Heliopause EUV mapping, combined with the heliosheath plasma imaging in energetic neutral atoms, 5'6'7'8 will explore in detail the three-dimensional time-varying region of the sun-LISM interaction. r

2. HELIOPAUSE MAPPING

Interstellar neutral atoms are unsuitable for heliopause mapping because they penetrate deep into the heliosphere. Most of the glow of heliospheric neutrals, as seen by an observer near 1 AU, would originate within the 10-AU region. Interstellar protons cannot be imaged optically at all. Interstellar helium (He +) and oxygen (O +) ions are ideally suited for heliopause mapping. Helium is the most abundant interstellar gas constituent (---10%) after hydrogen, with the exceptionally bright corresponding solar line (30.4 nm). Measurement of the ionization state of interstellar helium will provide an important insight into heating and cooling of the LISM. Oxygen is the most abundant interstellar gas minor constituent (~0.1%). The ionization states of hydrogen and oxygen are tightly coupled by efficient charge exchange. Therefore, the experimental determination of the ionization state of oxygen will establish the ionization state of hydrogen.

- 264 -

Mapping the heliopause in EUV

For an observer at 1 AU looking in the antisolar direction, the radiance (photon sr -1) F(O) is an integral along the line of sight

c m -2 s -1

oo

1 R~ ~N, + F ( O) - --4-~

(R,O)gI(R)dR

where the g-factor (scattering rate per ion per second) gl(R) depends on the distance from the sun, NI+(R,0) is the local ion number density, RE = 1 AU, and the scattering phase function is assumed isotropic. For a simplified case of 1) vanishing ion number density inside the heliopause (R < RHp) and 2) interstellar plasma (beyond the heliopause) at rest with a uniform number density and constant temperature, the radiance would be F(O) =

NI+ gI.E R2E 4~r

1 Rzp

where gx.E is the g-factor at 1 AU. The sky brightness is thus inversely proportional to the distance to the heliopause in the direction of observation. Measuring the directional dependence (imaging) of the interstellar plasma glow is a way to establish the size and shape of the heliopause. The detailed temperature, velocity and number density flow fields of the LISM plasma are needed for accurate treatment of the problem, and the g-factors should be calculted for the specific local velocity distribution functions of the ions. The plasma flow field was calculated for this work by Vladimir Baranov and co-workers in the Russian Academy of Sciences, Moscow using their two-shock sun-LISM interaction model 2 with the following LISM parameters (at infinity)" velocity, 25 km s-l; temperature, 5672 K; electron (proton) number density, n~ = 0.07 cm-3; neutral hydrogen number density, nH = 0.14 cm -3. The solar wind was assumed to flow spherically symmetric with a velocity of 450 km s -1 and a number density of 7 cm -3 at 1 AU. Interstellar helium and oxygen abundances were assumed 0.1 and 7x 10-4, respectively, by the number of atoms relative to hydrogen.

3. SOLAR LINES, BACKGROUND, AND FOREGROUND The solar emissions in the He + and O + resonance lines are well k n o w n . 9'10 Figure 3 shows the line profiles used in this work. The total solar flux in the helium line (30.4 nm) was assumed to be 6.0• 109 cm -2 s-1 at 1 AU and the line FWHM 0.01 nm (0.1 A). 9 An oxygen ion O + has a triplet transition at 83.2754, 83.3326, and 83.4462 nm. Both the solar emission O + and the nearby O 2+ multiplet could excite moving interstellar and heliospheric O + ions. The total solar flux l~ in all the lines was assumed to be 5.3• c m "2 s -1. The continuum c m -2 S "1 nlTl "1 for the helium and oxygen contribution is 1.8x108 c m -2 S "1 nlI1-1 and 4.0• lines, respectively. This contribution is important for the glow of oxygen ions, but unimportant for helium. There are two other major sources of radiation at 30.4 and 83.4 nm, the solar wind produced foreground and the galactic background. The feasibility of heliopause mapping critically depends on spectral properties and relative strength (brightness) of this interfering radiation. One requires the LISM plasma glow (the "heliopause glow") to be brighter and/or spectrally separated from the background and foreground radiation.

- 265 -

M. Gruntman

0

2.E+11

0t.--

Q- ].E+I]

Figure 3. Solar emissions in the He+ (30.4 nm) and O+ (83.4 nm) resonance lines used in this work. Arrows mark the OII triplet lines; other nearby lines are of the OIII multiplet. The continuum is important for the oxygen irradiance, but unimportant for helium. The glow of the He + and O + ions in the solar wind would produce the foreground radiation. These ions are produced by ionization of interstellar neutrals penetrating the heliosphere. The newly formed ions are picked up by the solar wind flow and carried to the termination shock as the pickup ions. 11'12 The pickup ions are singly charged and characterized by a spherical shell velocity distribution function. The glow of the pickup ions is the line radiation spectrally similar to the LISM plasma glow. We calculated the number densities of the pickup ions describing the inflowing interstellar neutral helium and oxygen by the hot model. Interstellar helium is only slightly affected by the crossing of the heliospheric intreface region. In contrast, the properties of interstellar oxygen are significantly modified by the crossing, which requires use of modified gas parameters at infinity. The details of the calculations can be found elsewhere. 4'13 The glow of the pickup ions was calculated similarly to that of the LISM plasma beyond the heliopause. We used the spherical shell velocity distribution function in calculations of the gfactors. The observed asymmetry 14'15 of the ion distribution function would only slightly modify the expected pickup ion glow and was disregarded in this work. At 30.4 nm, there is another important source of the foreground, viz. the emissions produced in charge exchange between the solar wind alpha-particles (He 2+) and heliospheric atomic hydrogen. The existence of this emission was known for some time, 16'17 but only recently it was analyzed in detail. 18 This emission is spectrally separated from the glow of the LISM and pickup ions and would not interfere with heliopause mapping (see below). We also note here that the all-sky images in the charge-exchange emissions will remotely reveal the three-dimensional solar wind flow properties everywhere in the heliosphere, including in the regions over the sun's poles and on the other (from the observer) side of the sun. 18 Two main sources of the EUV background are the radiation emitted by hot interstellar plasmas (diffuse galactic background) and by the stars (stellar radiation field). The stellar

- 266-

Mapping the heliopause in EUV radiation field is a continuum with negligible line emissions, while hot plasmas emit a continuum with prominent line emissions. The stellar radiation dominates the EUV continuum background at wavelengths > 20 r i m . 19'20 The total (at least 90% complete) continuum radiation field is ~13 photon c m -2 s -1 nm -1 and ~9.3 photon cm -2 s 1 nrn -1 at 30.4 nm and 83.4 nm, respectively, z~ The stellar EUV Figure 4. Projections on the ecliptic plane of the most important background radiation is stellar sources (Adhara, G191-BZB, and Feige 24) of the highly anisotropic. 2~ A background radiation at 30.4 nm and 83.4 nm; X and [3 are the single bright star, ecliptic longitude and latitude, respectively. Also shown are the Canis Majoris (Adhara interstellar wind, solar apex, and the closest star, a Centauri. or Adara), produces most of the radiation at 83.4 nm. Two white dwarfs, Feige 24 and G 191-B2B dominate the background at 30.4 n m . 20 For isotropic background, the estimated stellar radiation 2~ field translates into the ~l.3x 10 .2 mR/nm and ~9.3x10 -3 mR/nm at 30.4 nm and 83.4 nm, respectively. (1 Rayleigh = 1 R = l0 s mR = 106 ~tR = 106/(4 ~) phot cm 2 s~ sr-1.) For observations in the directions other than toward these bright sources (Figure 4), the stellar continuum background would be significantly smaller. Hot (~106 K) interstellar plamas efficiently emit EUV radiation that includes both the line emissions and continuum. The sun is embedded in a relatively small, a few parsec long, and dense (~0.1 cm 3) local interstellar cloud. This local interstellar cloud (LIC), our LISM, is too cold (~7000 K) to emit EUV radiation. LIC is positioned in the center of a region, the Local Bubble, filled with hot and dilute plasmas. These plasmas would emit the EUV radiation that after partial absorption in the LIC would reach the sun. The hot plasma EUV background (line emissions and continuum) was calculated using a standard plasma emission model 21 with the temperaure 106 K, emission measure 0.0006 cm -6 pc, and 1018 cm -2 column density of the absorbing LIC hydrogen. 18 Charge exchange of the solar wind alpha-particles would form, with some probability, single charged helium ions in the metastable state. 18 The metastable ions would decay by a two-photon emission process contributing to the background for X>30.4 nm. Actually, this two-photon decay would dominate the continuum background in most directions in the 3 5-90

- 267-

M. Gruntman

nm wavelength range. 18 The contributions from various sources to the background EUV continuum are summarized in Table 1. Table 1 Background EUV continuum spectral radiance, mR/nm Wavelength

4.

Stellar field (isotropic)

Local Buble plasma emission

30.4 nm

1.3

10-2

6.7

10 -4

83.4 nm

9.3

10-3

4.3

10 -6

RADIANCE

Solar Wind charge exchange emission

(1.0-2.8)

10-3

AT 30.4 NM AND 83.4 NM

Figure 5 shows the calculated radiance of the glow of the LISM plasma beyond the heliopause and the glow of the pickup ions. The helium glow is much brighter (milliRayleighs) as compared to the oxygen glow (micro-Rayleighs). The initial slight increase of the LISM plasma brightness with the angle 0 is due to the Doppler effect as the velocity radial component diminishes in the plasma turning around the heliopause.The glow falls as the heliopause moves away from the sun and the Doppler effect reduces the g-factor. If the plasma had the constant number density, velocity and temperature beyond the heliopause, then the glow angular dependence would have been inversely proportional to the distance to the heliopause. This latter inverse radial dependence is shown by the solid curves. The difference between the solid curves and the calculated radiance (empty circles) illustrates the sensitivity of heliopause mapping to the plasma flow field beyond the heliopause. 18

16

/

14

E o C-

8

~

L

-ul

6

Figure 5. Glow at 30.4 nm and 83.4 nm of the LISM plasma beyond the heliopause and the solar wind pickup ions. The solid curves are the function 1/R~ normalized at 0 = 0, R ~ is the distance to the heliopause.

- 268 -

Mapping the heliopause in EUV The assumed ion-to-neutral ratios were 0.3125 and 0.5 for helium and oxygen, respectively. The brightness of the LISM plasma glow is roughly proportional to the number density of the interstellar gas ionized component, while the pickup ion glow is roughly proportional to the number density of the neutral component. Figure 6 illustrates this effect for oxygen, 4 showing the angular dependence of the radiance for three different ion-to-neutral ratios, 1"2, 1:1, and 2:1. Helium glow properties exhibit a similar dependence on the ionization state. By measuring the upwind-to-downwind brightness ratio one would establish the ionization state of helium and oxygen. 5. SPECTRAL RADIANCE

10

t

~

t

~

2 "1 ne = 0.07 cm^(-3) -

Figure 6. Sky radiance directional dependence at 83.4 nm for various ionization states of the LISM. Interstellar oxygen is assumed ionized similarly to hydrogen. Total LISM number density is 0.21 cm3, and the oxygen relative abundance is 7• 10-4.

The expected spectral radiance at 30.4 nm is shown in Figure 7. The LISM plasma and pickup ion glows are in practically the same spectral range. Most of the adjacent plasma line emissions (from the Local Bubble) are produced by interstellar OIII, AIIX, and SiXI ions. The continua of the plasma emission and the stellar radiation field are negligible. The

Figure 7. Typical spectral radiance at 30.4 nm. The glow of the LISM plasma beyond the heliopause is shown as white (empty) bars. Also shown are the pickup ion glow (light gray bars), Local Bubble interstellar emissions (black bars), and the solar wind charge-exchange emission (dark gray bars; the two peaks correspond to the fast/polar and slow/ecliptic solar wind).

- 269-

M. Gruntman

Doppler shifted solar wind chargeexchange emissions 18 are clearly spectrally separated from other emissions. Measurements of the heliopause glow with the spectral resolution-~0.025 nm would allow one to eliminate the interfering contributions of the solar wind emissions. In order to efficiently remove the contributions of the interstellar lines, one would require the resolution of 0.005 nm. Figure 8 shows the spectral radiance at 83.4 nm. The glows of the LISM plasma and the solar wind pickup ions are of comparable total radiance of several milliRayleighs. However, the spectral radiance of the LISM plasma glow is about one order of magnitude brighter since it is concentrated in the narrower spectral range (Figure 8a, b). The continuum radiation due to two-photon decay of the metastable ions in the solar wind 18 and stellar radiation field 2~ are not negligible (Table 1). The combined spectral radiance from all sources is shown in Figure 8c for a typical continuum 0.16 ~R/(0.01nm). The glow measurements with a 0.01nm spectral resolution would allow identification of the contributions from the LISM plasma and from the pickup ions. Much modest resolution of~0.1 nm, should provide the combined glow radiance. The recently developed EUV spectrometers EURD advanced the sensitivity of the diffuse radiation detection to 1 mR. 22 The new space mission CHIPS, presently under preparation, will study the diffuse galactic radiation with a sensitivity of ~1 mR/line at X< 26 nm. The proposed heliopause mapping requires, however, significantly higher (a factor of 100) spectral resolution at 30.4 nm and significantly higher sensitivity (a factor of 100) at 83.4 nm. Development of diffuse EUV radiation spectrometers with such advanced performance characteristics is a challenging but not impossible task.

8

E r-

o o --- 0 . 2

F

d O t--

._~ "~

0.1

I-k. O

O~

0.0

---

E t-

.c,-

o

2.0

~

1.5

do 1.0 t~

Figure 8. Spectral radiance at 83.4 nm. a) LISM plasma glow; b) pickup ions glow; c) combined spectral radiance including the continuum.

-270-

Mapping the heliopause in EUV

Mapping the all-sky radiance at 30.4 nm and 83.4 nm will establish the distance to and shape of the heliopause. The ionization states of interstellar helium and oxygen (and correspondingly hydrogen) will be obtained from such measurements. The asymmetry of the insterstellar magnetic field would also be pronounced and identified in the all-sky maps. The geocorona is bright at 30.4 nm and 83.4 nm, and too many photons would reach the night side through multiple scattering. 23 The heliopause mapping experiment can be performed only from a spacecraft outside the geocorona. REFERENCES:

1. S.T. Suess, Rev. Geophys., 28, 97-115, 1990. 2. V.B. Baranov and Yu. G. Malama, J. Geophys. Res., 98, 15157-15163, 1993. 3. M. Gruntman and H.J. Fahr, 25, 1261-1264, 1998. 4. M. Gruntman and H.J. Fahr, J. Geophys. Res., 105, 5189-5200, 2000. 5. M.A. Gruntman, Planet. Space Sci., 40, 439-445, 1992. 6. M. Gruntman, Rev. Sci. Instrum., 68, 3617-3656, 1997. 7. E.C. Roelof, this volume. 8. H.O. Funsten et al., this volume. 9. G.A. Doschek,, W.E. Behring, and U. Feldman, Astrophys. J., 190, L 141-L 142, 1974. 10. R.R. Meier, Geophys. Res. Lett., 17, 1613-1616, 1990. 11. E. Moebius et al., Nature, 318, 426-429, 1985. 12. G. Gloeckler et al., Science, 261, 70-73, 1993. 13. M.A. Gruntman, J. Geophys. Res., 99, 19213-19227, 1994. 14. L.A. Fisk et al., Geophys. Res. Lett., 24, 93-96, 1997 15. E. M6bius et al., J. Geophys. Res., 103,257-265, 1998. 16. F. Paresce et al., J. Geophys. Res., 86, 10038-10048, 1983. 17. M.A. Gruntman, Geophys. Res. Lett., 19, 1323-1326, 1992. 18. M. Gruntman, J. Geophys. Res., 106, N.A5, in press, 2001. 19. K.-P. Cheng and F.C. Bruhweiler, Astrophys. J., 364, 573-581, 1990. 20. J. Vallerga, Astrophys. J., 497, 921-927, 1998. 21. M. Landini and B.C. Monsignori Fossi, Astron. Astrophys. Suppl. Ser., 82, 229-260, 1990. 22. S. Bowyer, J. Edelstein, and M. Lampton, Astrophys. J., 485,523-532, 1997. 23. R.R. Meier, R.R., Space Sci. Rev., 58, 1-185, 1991.

-271 -

This Page Intentionally Left Blank

E n e r g e t i c N e u t r a l H y d r o g e n of Heliospheric O r i g i n O b s e r v e d w i t h SOHO/CELIAS

at 1 AU

M. Hilchenbach a K.C. Hsieh b D.Hovestadt c R.Kallenbach d A. Czechowskie E.MSbius f and P.Bochsler g aMax-Planck-Institut fiir Aeronomie, D-37191 Katlenburg-Lindau, Germany bphysics Department, University of Arizona, Tucson AZ 85721, U.S.A. CMax-Planck-Institut fiir Extraterrestrische Physik, D-85740, Garching, Germany dInternational Space Science Institute, Hallerstr. 6, CH-3012, Bern, Switzerland eSpace Research Centre, Polish Academy of Sciences, Bartycka 18A, PL 00-716 Warsaw, Poland flnstitute for the Study of Earth, Oceans, and Space, University of New Hampshire, Durham, NH 03824 gPhysikalisches Institut, University of Bern, CH-3012, Bern, Switzerland

The High-Energy Suprathermal Time-of-Flight sensor (HSTOF) of the Charge, Element and Isotope Analysis System (CELIAS) on the Solar and Heliospheric Observatory (SOHO) near the Lagrangian point L1 is capable of observing energetic hydrogen atoms (EHAs). The EHAs are accumulated from 1996 to 2000 under quiet interplanetary conditions and the observed EHA flux level, their energy spectrum and their anisotropy will be discussed. 1. I N T R O D U C T I O N In 1992 Hsieh et al. [1] proposed to study the acceleration and propagation of the anomalous cosmic rays (ACRs) in and out of the heliosphere via energetic neutral atoms (ENA) originating from ACRs charge exchange with atoms of the local instellar medium (LISM). In 1995, Czechowski et al.[2] modelled the expected ENA flux originating from the ACRs in the outer heliosphere and predicted an apex- tail flux anisotropy. In 1998, Hilchenbach et al. [3] observed energetic hydrogen atom (EHA) fluxes during quiet-time interplanetary conditions in 1996 and 1997 with a high flux anisotropy coming from the approximate tail direction of the heliosphere, e.g. as predicted for ACR produced EHAs. In this follow-up study we address the in-flight calibration of the HSTOF instrument, the selection of interplanetary quiet-time periods and the energy spectrum of the EHAs.

-273-

M. Hilchenbach et al.

Figure 1. In-flight cross calibration of SOHO/CELIAS/HSTOF with ACE/ULEIS and ACE/EPAM [4].

Figure 2. Proton and EHA flux for 'less quiet' times (826 days). About a third of the low energetic particle events can be attributed to transmitted protons.

2. M e t h o d s and C a l i b r a t i o n

To determine the flux of EHAs requires accurate knowledge of the HSTOF detection efficiency in order to subtract the low-energy protons from the particle spectrum measured with HSTOF. The remaining events are either due to EHAs or accidental coincidence events. More details about this method and its rationale as well as the instrument are described in Hilchenbach et al. [3]. For energetic-particle events we compared proton fluxes measured with HSTOF with fluxes measured by instruments on the Advanced Composition Explorer (ACE, Level 2 Data, time interval DOY 301/1998 to DOY 99/2000 [4] ). ACE is, as SOHO, on a halo orbit near the Lagrangian point L1. With this comparison we could determine the inflight efficiency of HSTOF for protons. It turned out to be ~ 10 times lower than expected from pre-flight calibrations (Fig. 1). The origin of this large discrepancy is currently under investigation. The HSTOF efficiency did not vary systematically beyond a level of about ~ 15% since launch (data not shown). 3. R e s u l t s For the present study, we analysed HSTOF data in the time interval from DOY 43/1996 to DOY 99/2000. We selected interplanetary quiet-time periods according to the registered events in the mass = 1 and in the 5 0 - 6 6 0 k e V energy range: < 12 cts/day for 'very quiet', < 20 cts/day for 'quiet' and < 100 cts/day for 'less quiet' time periods. In the 'less quiet' times, we found that about a third of the particles in the 5 8 - 8 8 k e V energy range could be attributed to transmitted quiet-time protons (Fig. 2). For the quiet-time protons

-274-

Energetic neutral hydrogen of heliospheric origin... .

--*--

:

.

.

EHA

.

i

.

.

.

.

] r~

:

,4

~5 1 0 .4

(keV)

Figure 3. Energetic particle spectrum at 'quiet' times (555 days). Compared to Fig. 2 the proton flux is reduced due to the lowered quiet-time selection threshold.

Figure 4. Particle spectrum for 'very quiet' times (401 days). Compared to Fig. 2 and Fig. 3, the EHA flux is consistent.

we found a power law with 7 = -2.5 which agrees with previous observations (Richardson et al. [5]). Using only data from quiet and very quiet-times reduces the proton flux below the ~ 10% level. The EHA flux in the energy range 58 - 88keV remains unaffected (Fig. 3 and Fig. 4). Again, the quiet-times reveal an anisotropic flux arising from the tail region of the heliosphere (Fig. 5). For the newer data reported here, the quiet-time EHA observations were hampered by the 3-month loss of SOHO in 1998 as well as the return of enhanced solar activity. The energy spectrum of the EHAs during very quiet-times originating from the apex and tail regions is shown in Fig. 6 (tail region DOY 170-220 and apex region DOY 1-169, 221-365(366)). Calculations based on Kausch's heliospheric model (Fahr et al. [6], multiplied by factor of 10 for the tail and 40 for the apex region) are shown for comparison (dotted lines in Fig. 6). 4. D I S C U S S I O N We cross calibrated the HSTOF instrument in flight with instruments on the ACE satellite. The reduced sensitivity of HSTOF transforms into an EHA flux which is higher than assumed previously [3]. The cause is still under consideration, e.g. part of the timeof-flight foils could have been altered during the SOHO launch. We checked the quiet-time threshold criteria for the EHA analysis. We found that the proton flux is a function of the selection threshold and the EHA flux is independent. The EHA flux observed by HSTOF is about 10 times higher than previously explained by model calculations (e.g. Czechowski et al.[7]). Furthermore, recent modelling of another EHA source, originating from accelerated protons in co-rotating interaction regions (CIRs), show a significant

-275-

M. Hilchenbach et al.

Figure 5. Quiet-time 5 8 - 8 8 k e V EHA flux as observed by HSTOF (left panel), the background level plotted for comparison (right panel) and the heliocentric ecliptic longitude coordinate system (top labels, heliotail at 74 ~

Figure 6. EHA energy spectrum for 'very quiet times' (tail and apex regions).

EHA flux anisotropy from the direction of the LISM helium focusing cone, overlapping the heliospheric tail region (Kdta et al., [8]). We acknowledge the work of the people involved in the ACE ULEIS and EPAM instruments who made the CELIAS/HSTOF in-flight calibration study possible [4]. We thank the referee, R.F. Wimmer-Schweingruber, for his helpful comments. REFERENCES

1. 2. 3. 4. 5. 6. 7. 8.

K.C. Hsieh, K.L. Shih, J.R. Jokipii and S. Grzedzielski 1992: ApJ, 393, 756 A. Czechowski, S. Grzedzielski and I. Mostafa 1995: A&A, 297, 892 M. Hilchenbach, K.C. Hsieh, D. Hovestadt et al. 1998: ApJ, 503, 916 ACE Science Center Level 2 Data, http://www.srl.caltech.edu/ACE/ASC/ I.G. Richardson,D. V. Reames,K.-P.Wenzel,J.Rodriguez-Pacheco 1990: ApJ, 363L, L9 H.J. Fahr, T. Kausch and H. Scherer 2000: A&A, 357, 268-282 A.Czechowski, H.Fichtner, S.Grzedzielski et al. 2001: A&A, 368, 622 J. Kdta et al. 2000" JGR, submitted

-276-

Acceleration of pick-up ions at the termination shock in the limit of weak scattering S.V. Chalov ~* and H.J. Fahr bt ~Institute for Problems in Mechanics of the Russian Academy of Sciences, Prospect Vernadskogo 101-1, 117526 Moscow, Russia UInstitut fiir Astrophysik und Extraterrestrische Forschung der Universitgt Bonn, Auf dem Hiigel 71, D-53121 Bonn, Germany It is generally agreed that interstellar pick-up ions constitute a seed population for anomalous cosmic rays (ACRs) originating at the solar wind termination shock (TS) through the shock-drift acceleration and first-order Fermi mechanism. Acceleration of ACRs at the TS is usually described in the limit of strong scattering when the cosmic-ray velocity distribution function is almost isotropic and continuous through the shock front. We consider here the opposite case of weak scattering taking into account anisotropy of the velocity distribution which can be of great consequence for the acceleration process at least at low energies. In order to describe the motion of pick-up ions in the upstream and downstream parts of the solar wind flow near the TS the Fokker-Planck transport equation for anisotropic velocity distribution function is used, while through the shock front the conservation of the magnetic moment of particles is assumed. It is shown that downstream spectra of accelerated pick-up ions close to the ecliptic plane depend strongly on longitude due to longitudinal dependence of the angle between the shock normal and heliospheric magnetic field connected with departure of the TS from sphericity. 1. I N T R O D U C T I O N

Acceleration of ACRs at the TS is often considered in the diffusive approximation which is valid only in the case of small pitch-angle anisotropy. However, the velocity distribution of particles in the vicinity of the TS can be essentially anisotropic if pitchangle scattering is not too strong, so that the application of the diffusive theory is highly problematic at least for low and moderate energy particles. As an example we refer to the recent paper [1]. In the paper the differential energy spectra of 50-keV to 20 MeV protons accelerated at forward and reverse shocks at corotating interaction regions observed at Ulysses during 1992 and 1993 have been compared with the predictions of two models based on diffusive shock acceleration theory [2,3]. It has been shown in [1] that the *SVC was partially supported by Avard No. RPI-2248 U.S. Civilian Research and Development Foundation, RFBR Grant 99-02-04025, and DFG Grant 436 RUS 113/110/6-2. tHJF was partially supported by DFG Grant Fa-97/24-2 and RFBR Grant 99-02-04025.

-277-

S. V. Chalov and H.J. Fahr

observations are inconsistent with the main predictions of the theory. Specifically, the observed spectra are regularly much harder than predicted. In the present paper we present results concerning acceleration of interstellar pick-up protons at the TS up to energies of ACRs taking into account anisotropy in their velocity distribution in the vicinity of the shock. An essential point of our model is preacceleration of pick-up protons by solar wind turbulence. 2. S T O C H A S T I C

AND REGULAR

ACCELERATION

OF P I C K - U P

IONS

The transport of pick-up ions in the heliosphere up to the TS is described by the well-known transport equation for the isotropic velocity distribution function including continuous production of pick-up ions, their convection with the solar wind velocity U, adiabatic deceleration, and energy diffusion. It should be pointed out that isotropy of the velocity distribution is assumed here only for particles which did not suffer action of the TS. Solar wind turbulence is considered as consisting of small-scale Alfv(~nic turbulence and nonlinear large-scale fluctuations in the solar wind velocity and magnetic field. Thus the energy diffusion coefficient can be written in the form: D - DA + DL, where DA and DL depend on the intensities of corresponding fluctuations ~AE - - ( ( ~ B ~ E ) / S ~ and Cbs - (aU~s)~/2/Us at 1 AU (for more details, see [4]). Figure 1 shows fluxes (in the solar wind rest frame) of accelerated pick-up protons in front of the TS (rsh -- 90 AU in the upwind direction). The solar wind and LISM velocities and number densities are adopted to be Us - 450 km s -1, n~,s - 7 cm -a, cm -a, respectively. Curve 1 corresponds -- //H,LISM -- 0.14 V L I S M - - 26 km S - 1 , • e , L I S M to the case when only Alfv(~nic turbulence is taken into account" ~AE - - 0 . 4 , (~LE - - 0. Curves 2 and 3 show fluxes in the case when the level of Alfv(~nic turbulence is lower (Cas - 0.2) but acceleration of large-scale fluctuations is included (~LS -- 0.3 and 0.5 for curves 2 and a, respectively). One can see that the large-scale fluctuations have a dominant role in formation of high energy tails in spectra of pick-up ions accelerated by solar wind turbulence. The preaccelerated pick-up ions experience further acceleration at the TS. We consider here the TS as a discontinuity neglecting the cross-shock potential. Then in the case of weak scattering considered here one can use the assumption of conservation of the energy of pick-up ions in the de Hoffmann-Teller frame and the assumption that the magnetic moment of a particle is the same before and after the encounter with the shock. From these two assumptions one can obtain conditions under which particles are transmitted through the shock or reflected from the shock due to abrupt change of the magnetic field (see [5]). Due to the reflection particles gain energy and move along magnetic field lines away from the shock front. However, the interaction of the particles with upstream Alfv~nic turbulence leads to their pitch-angle scattering, and the result is multiple reflection from the TS. In order to describe the process of the multiple reflection the following transport equation for anisotropic velocity distribution function of pick-up ions f(t, x, v, #) in the vicinity of the TS is considered here"

Of

at +

(Ux +

Of _

x)b-7

(1)

f + O(x, v),

-278-

Acceleration of p&kup ions at the termination shock...

. . . . . . . .

~"

1E+2

1E+1

1E+t

1E+0

|

. . . . . . . .

!

. . . . . . . .

O

.-

t/J tg)

le-1 E ._o_

E

1E-1

O

1E-2

.,..,

1E-3

~-

1E-4

\

1E-4

III

"-a

x ii

1E-6 |

1E-7 1E-1

. . . . . . .

i

. . . . . . . .

i

. . . . . . . .

1E+1 KINETIC E N E R G Y (keV)

Figure 1. Fluxes of accelerated by solar wind turbulence pick-up protons in front of the TS. 1 - r = 0.4, ~LE = 0; 2 - r = 0.2, (Ls = 0.3; 3 - ~ , s = 0.2, (Ls = 0.5.

Figure 2. Downstream fluxes of accelerated pick-up protons. 1 - ~b - 70 ~ 2 - ~b - 80 ~ 3 - ~b - 85 ~ , 4 - ~b - 89 ~ .

where the z-axis is directed perpendicular to the shock which is considered to be planar, v is the velocity of pick-up ions in the solar wind rest frame, # is the cosine of the pitchangle of ions, X is the cosine of the shock normal angle ~b, the source Q corresponds to pick-up ions arriving at the TS (see Figure 1), and S f is the scattering operator including pitch-angle scattering and energy diffusion (see [6]). In the following, we consider regions of the heliosphere close to the ecliptic plane. Since the shape of the TS is not spherically symmetric due to the interaction of the solar wind with the moving interstellar medium the shock normal angle ~b is a function of longitude measured from the upwind direction. In the upwind and downwind direction ~b = 90 ~ while at the flanks of the TS ~b can reach 65 ~ [7]. Figure 2 shows downstream fluxes of pick-up protons accelerated at the TS for different shock normal angles. The initial upstream flux is given by curve 3 in Figure 1. The level of Alfv~nic turbulence in front of the TS is adopted to be CA = 0.01 (weak scattering). The key distinction of fluxes in Figure 2 from fluxes predicted by the diffusive theory is their nonmonotonic behaviour. In the case of weak scattering the downstream fluxes can be considered as consisting of two parts. The low energy parts are formed by protons which were transmitted through the shock and did not experience multiple reflections. The high energy parts are formed by protons which experienced multiple reflections at the shock. The strong dependence of the fluxes on the shock normal angle and, therefore, on longitude can be seen in Figure 2. Downstream fluxes at the flanks of the TS exceed fluxes at its nose part in the energy range from about 100 keV to several MeV. It is connected with increasing reflection efficiency at the shock with decreasing ~b.

- 279-

S. K Chalov and H.J. Fahr

,

,

|

,

,

|

,

,

|

,

,

,

|

,

,

1E-4
70 MeV/nuc counts measured by the cosmic ray subsystem (CRS) on Voyager 2. Four clear density enhancements are seen, one in early 1997, one in late 1998, and one each in early 1999 and 2000. Burlaga and Ness [13] identified the 1998 event as a MIR based on the enhanced IMF during this time period. The other events have very similar plasma signatures and we suggest these are also MIRs. The frequency is increasing towards solar maximum, as expected. The large IMF during these events acts as a barrier to the inward penetration of cosmic rays. After each MIR identified here a decrease in the CRS counts is observed, although CRS decreases sometimes occur when no density enhancement is present. In past solar cycles, the CRS counts increased towards solar maximum, with minor decreases triggered by MIRs followed by recoveries, until a large event causes a major decrease in flux which persists until the next solar cycle [14]. The Voyager 1 CRS counts (gray) look

-305 -

J.D. Richardson

Figure 4. Diagram of speed profiles versus latitude for solar minimum and maximum.

almost identical to those observed by Voyager 2; thus we know that the effects of these MIRs cover a latitude range of at least 50 ~ and a longitude range of at least 45 ~ longitude.

Figure 5. Solar wind density (black), CRS counts (gray), and CME ejecta (gray line).

The driving mechanism for large MIRs may be large CMEs, or groups of CMEs, occurring on the Sun. Signatures of CMEs near Earth are low temperatures, smooth field rotations (magnetic clouds), counter streaming electrons, and enhanced helium abundances [12]. The signature most likely to survive until the outer heliosphere is the helium abundance. Wang and Richardson [15] recently completed a survey of helium abiundance enhancements (HAEs) observed by Voyager, defined as periods where the He++/H + ratio is greater than 10%. This condition seems sufficient to identify some CMEs, but other CMEs do not have He++/H + ratios of this magnitude. The only multiple day HAE event in the time period covered by Figure 5 is shown by the vertical line just before 1999.5; the He++/H + ratio was enhanced over a 6 day period probably resulting from a series of CMEs. The location of the HAEs on the trailing edge of the MIR is a clear indication

-

306-

The solar wind: Probing the heliosphere with multiple spacecraft that the MIR is driven by the CMEs in this case. More work should be done for the other MIRs, looking for He++/H + ratios which are below the 10% threshold but are still enhanced to see if CME drivers can be identified. 4. S O L A R W I N D C H A N G E S

WITH DISTANCE

4.1. Solar W i n d S l o w d o w n

One difficulty in understanding solar wind observations is decoupling changes due radial distance, temperature, and solar cycle effects. Figure 2 shows that the Voyager 2 speeds in 2000 are well below those observed by IMP 8 and Ulysses. This speed difference is at least partially due to the addition of interstellar neutrals to the solar wind; when these neutrals are ionized they are accelerated to the solar wind speed by the Lorenz force. The energy for this acceleration comes from the kinetic energy of the solar wind. Quantifying the speed decrease provides estimates of the density of H in the local interstellar medium. The first attempt to quantify the speed decrease [16,17] found a decrease of about 30 km/s near 30 AU giving a density for interstellar neutrals of 0.05 cm -3, but this result was controversial [18]. Isenberg [19] argued that the slowdown in this period was greater than 30 km/s and thus not inconsistent with values of the density of the local interstellar medium derived by Gloeckler et al. [20] of 0.115 cm -3 at the termination shock and 0.22 cm-3 in the undisturbed interstellar medium. Wang et al. [21,22] took advantage of two time periods when Ulysses and Voyager 2 were at similar latitudes to look for the speed slowdown when the spacecraft were separated in radius by 30 and 55 AU. They used the Ulysses data as input to a 1-D MHD model and propagated the solar wind out to the radial distance of Voyager 2. Wang and Richardson [23] show histograms of speeds observed at Voyager 2, speeds predicted for no density of the local interstellar medium, and speeds for a density of the local interstellar medium density of 0.05 cm -3 at the termination shock. This value for the density of the local interstellar medium gives good agreement with the observations, but is lower than results from observations of pickup ions and from UV measurements. These observations give densities of about 0.22 in the density of the local interstellar medium of which roughly half is lost before the termination shock [20]. Thus the values derived from the speed decrease are about half of other determinations of this value. Doubling the density of the local interstellar medium in our model would roughly double the observed speed decrease, which would not be consistent with the observations. We note that the Gloeckler results are for a limited set of solar wind conditions, whereas the slowing of the wind is an integral process occurring over 60 AU, which could explain the apparent discrepancy. As Voyager gets further from the Sun the speed decrease will grow even larger and perhaps provide the opportunity for more determinations of the slowdown. 4.2. T E M P E R A T U R E

Many competing effects determine the temperature of the thermal component of the solar wind. The plasma cools as it moves outwards due to adiabatic expansion. But it can also be heated by numerous effects. The velocity difference between different solar wind streams results in the formation of shocks which transfer energy from the solar wind flow into the internal plasma energy. Dissipation of speed shear and magnetic turbulence results in heating of the plasma [24]. Pickup ions are created with energies equal to the

- 307-

J.D. Richardson

flow energy, about 1 keV, and this energy may be transferred to the thermal protons. Figure 6 shows a comparison of proton temperatures observed at Voyager 2 and IMP 8. The temperature decreases from Voyager launch to somewhere between 1986 and 1990, when Voyager was at 20-30 AU. The minimum in 1986 is probably a solar minimum effect due to Voyager 2 being in very slow solar wind flow. The decrease in temperature is much less than the R -4/3 adiabatic expansion would predict, falling as R - 5 - R -7 from 1-30 AU [8,16]. After 1990 the temperature increases to about 1997, then begins to decrease. Comparison with the IMP 8 temperature profile shows a very good correspondence for feature with scales on order of a year out to almost 60 AU. After 1998, the temperatures at IMP 8 increase while those at Voyager 2 decrease.

Figure 6. Temperatures (solid lines) and speeds (dotted lines) at IMP 8 and Voyager 2.

Also plotted on Figure 6 are scaled speed profiles measured by each spacecraft. At 1 AU the speeds and temperatures show very good correspondence. For Voyager 2, speed and temperature features generally match quite well. The surprising feature of this plot is the very good agreement between the magnitudes of the traces after 1983 when Voyager was at about 15 AU. The decrease in temperature relative to the speed inside 15 AU is due to the adiabatic cooling lessened by the effects described above. The similarity of the traces outside 15 AU suggests that the heating which occurs compensates almost exactly for the expected adiabatic cooling over a distance of 45 AU. The remaining fluctuations are then explained quite well by the speed variations. 5. S U M M A R Y Solar wind parameters vary with changes in the solar cycle, with changes in heliolatitude, and with changes in radial distance. Even with multiple spacecraft it can be difficult

-308-

The solar wind: Probing the heliosphere with multiple spacecraft to deconvolve these various effects. We find that the solar wind dynamic pressure varies by a factor of 2 over the solar cycle at all latitudes and distances. The speed and density (which determine the dynamic pressure) have dramatically different latitude profiles at solar maximum and solar minimum. At solar minimum, a narrow, 20-30 ~ latitude wedge of slow, dense solar wind at low latitudes is surrounded by fast tenuous solar wind at high latitudes, whereas at solar maximum the highest speeds and lowest densities are at low latitudes with a slow decrease in speed towards higher latitudes. The speed of the solar wind decreases with radial distance due to the pickup of interstellar neutrals; at 60 AU this slowdown is about 40 km/s. The solar wind is heated as it moves outwards so it does not cool adiabatically. The observed temperatures in the outer heliosphere are are still very strongly correlated with the solar wind speed, a relation initially imposed at the solar source. 6. A c k n o w l e d g m e n t s I thank D. McComas for the Ulysses data used in this paper and E. Stone for the Voyager CRS data. This work was supported under NASA contract 959203 from the Jet Propulsion Laboratory to the Massachusetts Institute of Technology and by Heliospheric GI grant NAGW-6473. R E F E R E N C E S

1. D . J . McComas et al., J. Geophys. Res., 105 (2000) 10419. 2. P.R. Gazis, Rev. Geophys., 34 (1996) 379. 3. A . J . Lazarus and R. L. McNutt, Jr., in Physics of the Outer Heliosphere, edited by S. Grzedzielski and D. E. Page, Pergamon Press, New York (1990) 229. J. D. Richardson, K. I. Paularena, A. J. Lazarus, and J. W. Belcher, Geophys. Res. Lett., 22 (1995) 1469. S. R. Karmesin, P. C. Liewer, and J. U. Brackbill, Geophys. Res. Lett., 22, (1995) 1153. C. Wang, and J. W. Belcher, J. Geophys. Res., 104 (1999) 549. 7. J. L. Phillips et al., Solar Wind 8, edited by D Winterhalter et al., (1996) 416. 8. P. R. Gazis, J. Geophys. Res., 98 (1993) 9391. 9. J. D. Richardson and K. I. Paularena, Geophys. Res. Lett., 24 (1997) 1435. 10. Miyake, W., et al., Plan. Sp. Sci. 36 (1998) 1329. 11. E. J. Smith, A. Balogh, R. J. Forsyth, D.J. McComas, COSPAR 2000, E2.2-0037, Warsaw (2000). 12. J. T. Gosling, Coronal Mass Ejections, edited by N. Crooker, J. A. Joselyn, and J. Feynman, Geophysical Monograph 89, American Geophysical Union (1997) 9. 13. L. F. Burlaga and N. F. Ness, J. Geophys. Res., 105 (2000) 5141. 14. F. B. McDonald, Sp. Sci. Rev., 83 (1998) 33. 15. C. Wang and J. D. Richardson, J. Geophys. Res. (2000), in press. 16. J. D. Richardson, Physics of Space Plasmas (1995), No. 14, edited by T. Chang and J. R. Jasperse (1996) 431. 17. J. D. Richardson, K. I. Paularena, A. J. Lazarus, and J. W. Belcher, Geophys. R es. Lett., 22 (1995) 325. .

.

o

-

309-

J.D. Richardson

18. P. R. Gazis, Geophys. Res. Lett., 22 (1995) 2441. 19. P. A. Isenberg, Solar Wind 9, edited by S. R. Habbal, R. Esser, J. V. Hollweg, and P. A. Isenberg, (1999) 189. 20. G. Gloeckler, L. A. Fisk, and 3. Geiss, Nature, 386 (1997) 374. 21. C. Wang, J. D. Richardson and a. T. Gosling, J. Geophys. Res. 105 (2000) 2337. 22. C. Wang, J. D. Richardson and J. T. Gosling, Geophys. Res. Lett. 26 (2000) 2429. 23. C. Wang and J. D. Richardson, this volume. 24. W. H. Matthaeus, G. P. Zank, C. W. Smith, and S. Oughton, Phys. Rev. Lett., 82 (1999) 3444.

-310-

Propagation of the solar wind from the inner to outer Heliosphere" Three-fluid model C. Wang* and J. D. Richardson Center for Space Research, Massachusette Institute of Technology, Cambridge, MA 02139, USA We take advantage of two time periods (1991 and the beginning of 1999) when Ulysses and Voyager 2 were at about the same latitude to study the propagation of the solar wind from the inner to outer heliosphere (up to 60 AU). A three-fluid model of the solar wind which consists of solar wind protons, pickup ions and electrons is employed to examine the effect of pickup ions. As expected, the pickup ions play an important role in the evolution of the solar wind in the distant heliosphere, especially beyond 40 AU. We find a decrease of about 40 km/s or 10% in the radial velocity near 60 AU. This speed decrease implies an interstellar neutral density at the termination shock of 0.05 cm -3. 1. I N T R O D U C T I O N The propagation of the solar wind in the outer heliosphere, especially shock evolution and interaction, have been studied intensively [1,2]. However these studies could not take advantage of recent spacecraft observations out to 70 AU and did not appreciate the importance of pickup ions in the distant heliosphere. Only recently have investigations of the effect of pickup ions on dynamical processes in the solar wind begun appearing in the literature [3-6]. However, almost all these pickup models assume that wave-particle interactions proceed sufficiently quickly that pickup ions are soon assimilated into the solar wind, becoming indistinguishable from solar wind protons. A substantial increase in the solar wind temperature with increasing heliocentric distance beyond ~ 5 AU is thus predicted. Such a temperature increase is, however, not observed in the outer heliosphere. A model which distinguishes the pickup ions from the solar wind ions is appropriate and a simple version was developed by Isenberg [7]. We will extend the Isenberg three-fluid model in this study. Voyager 2 continues to explore the outer heliosphere as Ulysses studies the latitudinal dependence of the solar wind. The trajectories of Ulysses and Voyager 2 are shown in Figure 1. During the year 1991 these spacecraft were within 2~ latitude and their radial separation was larger than 30 AU. They were at about the same latitude again at the beginning of 1999 and their radial separation was as large as 55 AU. This latitudinal proximity provides a good opportunity to study the evolution of the solar wind from the inner to outer heliosphere and to investigate the effect of pickup ions. *Also at the Lab. for Space Weather, Chinese Academy of Sciences, Bejing, China

-311 -

C. Wang and J.D. Richardson

Figure 1. Ulysses.

Trajectories of Voyager 2 and Figure 2. Steady-state Heliosphere.

2. T H R E E - F L U I D

MODEL

Isenberg [7] developed a three-fluid model which consists of co-moving thermal populations of solar wind protons, pickup ions and electrons. The steady-state solution yields a solar wind with a "core" proton distribution which cools, essentially, adiabatically. This simplified three-fluid model, which does not consider the interaction of the solar wind protons and pickup ions and assumes that all the energy of the pickup process remains with the pickup ions, does not reproduce the observed temperature profile. We extend this three-fluid model by depositing part of the energy from the ionization and pickup process into the solar wind protons. We suggest an adjustable parameter e(t, r, v, n, ...) to represent the ratio of this portion of the energy to the total thermal energy generated from the ionization and pickup process. That is E~ CEtotal. For simplicity, we assume c to be constant. With this constraint, we find that c = 20V0 gives the best fit to the observations, as illustrated in the third panel in Figure 2. The cold neutral density distribution for the interstellar neutrals, nil(r), is taken as [8] =

(1)

e -AIr

where A = 4 AU and UH~ = 20 km/s. The subscript infinity refers to the boundary values (the boundary being the termination shock). The source terms for charge exchange were -312-

Propagation o f the solar wind from the inner to outer heliosphere: ....

summarized in our previous work [9]. The solar wind speed, proton density, proton temperature, pickup ion density, and pickup temperature as functions of heliocentric distance are shown in Figure 2 from the top to bottom panels, respectively.

Figure 3. Evolution of the solar wind (1991 Figure 4. Comparison of the model results data). with Voyager 2 observations.

3. E V O L U T I O N

OF T H E S O L A R W I N D

In order to study solar wind propagation from the inner to outer heliosphere, we feed the five streams from day 140 to 210 in 1991 from the Ulysses data into our three-fluid model as an input. We follow the evolution of the solar wind structure to the location of Voyager 2. Figure 3 shows the theoretical development of these streams as if they were "observed" at successive distances (10, 20 AU and the location of Voyager 2 (~35 AU)). The interaction and evolution of the streams observed by Ulysses produce a quite different solar wind structure at Voyager 2. The structure loses the traces of the original stream structure and form two strong forward shocks at Voyager 2, which agrees roughly with the observations. 4. S L O W D O W N

OF THE SOLAR WIND

During 1998 and 1999, it took about 6 months for the solar wind to travel from the location of Ulysses to Voyager 2. So the solar wind observed by Ulysses during the time period from 1998.5 to 1999.5, when the latitudinal difference between these two spacecraft was less then 10 ~ reached Voyager 2 during 1999. We carry out numerical calculations

-313-

C. Wang and,I.D. Richardson

g

0.2 0.0 300

0.0

350

400

450

500

550

600

300

550

400

450

5.00

550

600

. . . .

300

350

L 1 ] .............

400

450

500

550

600

s

Figure 5. Histograms of the solar wind speed.

using two different models, namely an MHD model and an MHD model which mimics the effect of the interstellar neutral hydrogen (MHD PI model). Figure 4 shows comparisons of the observed flow speed at Voyager 2 (solid lines) with those simulated by the MHD model (without pickup ions, dotted line in the top panel) and the MHD PI model (with interstellar neutral density at the termination shock n/zoo - 0.05 cm -a, dotted line in the bottom panel). Overall, the speed predicted by both MHD and MHD PI models roughly lies in the range of the observations; however, neither of them reproduce the fine structure of the speed profile observed by Voyager 2. As shown in Figure 4, the MHD model in general predicts higher speeds than are observed at Voyager 2. As expected, we need to incorporate pickup ions into our model to decelerate the solar wind speed sufficiently to match the observations. Figure 5 shows histograms of the frequency of occurrence of various speeds for both the Voyager 2 observations in 1999 and the model results using the Ulysses observations as input. From the left to the right panels, the figure shows the Voyager 2 observations, the MHD and the MHD PI model speed predictions (with n~oo - 0.05 cm-a), respectively. Both the MHD and MHD PI models reproduce the observed histogram shape. The average speed of the observations for 1999 is 392 km s -~. However, the MHD model gives a higher average speed of 432 km s -~. By contrast, the MHD PI model with n~oo - 0.05 cm -3 gives an average speed of 389 km s -1, very close to the observed value. These results are similar to our previous finding using a one-fluid MHD model [6]. REFERENCES

1. 2. 3. 4. 5. 6. 7. 8. 9.

A . J . Hundhausen, AGU Monograph, 34 (1985) 37. Y. C. Whang, Space. Sci. Rev. 57 (1991) 339. G.P. Zank, and H. L. Pauls, J. Geophy. Res. 102 (1997) 7037. W . K . M . Rice, and G. P. Zank, J. Geophy. Res. 104 (1999) 12563. C. Wang, J. D. Richardson and J. T. Gosling, J. Geophys. Res. 105 (2000) 2337. C. Wang, J. D. Richardson and J. T. Gosling, Geophys. Res. Lett. 26 (2000) 2429. P.A. Isenberg, J. Geophys. Res. 91 (1986) 9965. V.M. Vasyliunas, and G. L. Siscoe, J. Geophys. Res. 81 (1976) 1247. C. Wang, and J. W. Belcher, J. Geophys. Res., 104 (1999) 549.

-314-

Relationships of corotating rarefaction regions outside 40 AU with solar observations: Heliospheric mass loading A. Posner, ~ N.A. Schwadron, ~ and T.H. Zurbuchen ~ * ~Department of Atmospheric, Oceanic and Space Sciences, University of Michigan, 2455 Hayward, Ann Arbor, MI 48109-2143, USA In mid-1994 the solar wind instrument on the Voyager 2 spacecraft observed a recurrent pattern of compression and rarefaction regions on the timescale of the solar rotation period. The spacecraft location at that time was at 12 ~ southern heliographic latitude. The time period of the release of the large scale structures at the Sun was coincident with observations of recurrent solar wind structures by IMP 8 in Earth orbit. Additionally, the Yohkoh/SXT soft x-ray telescope made synoptic observations of the upper solar corona. With the data available we identified the coronal sources of solar wind structures observed by Voyager 2 at ~43 AU in the Yohkoh/SXT observations of the Sun at the time the solar wind was released. Using the Sun's rotation as a clock a significant slowdown of the solar wind can be derived. A numerical solution for the mass loading of the solar wind with pickup ions is used to infer the interstellar neutral hydrogen density as n~ = 0.127+0.019. 1. I N T R O D U C T I O N Bryant [1] discovered a periodicity in the intensity of heliospheric energetic protons that resembled the synodic solar rotation period. Barnes and Simpson [2] associated these with quasi-stationary solar wind structures in the heliosphere. Krieger et al. [3] discovered coronal holes as the source of high speed streams in the ecliptic plane by ballistic backmapping of the radially propagating solar wind. Due to solar rotation fast streams interact with ambient slow solar wind to form sets of Corotating Interaction Regions (CIRs) and Rarefaction Regions (CRRs) (see sketch in [4]). Solar sources for CIRs and CRRs are the western and eastern boundary regions of coronal holes, respectively. Outside 10 AU CIRs start to merge into MIRs (Merged Interaction Regions) [5]. In the literature, values from 12 AU [6] to 45 AU [7] are found for the maximum distance that a CIR can exist without merging with CIR structures from another rotation. Tilted dipole 3-D MHD simulations [8], however, suggest that forward and reverse shocks pass each other at different latitudes [9], leaving behind undisturbed rarefaction regions in the outer heliosphere. The ballistic backmapping of CIRs is inaccurate due to the interaction of the converging streams, resulting in acceleration and deflection of the bulk solar wind flow. Ballistic backmapping that focuses on corotating rarefaction regions avoids the effects of these shocks. It is therefore the most efficient and easiest method to relate solar *This work was supported, in part, by NASA contracts NAG5-2810 and NAG5-7111.

-315-

A. Posner, N.A. Schwadron and T.H. Zurbuchen

wind observations to solar source regions. 2. O B S E R V A T I O N S

AND DISCUSSION

Figure la shows one hour averages of the solar wind speed measured by Voyager 2/PLS in 1994 at 43 AU from the Sun, at 12~ S heliographic latitude. The long-term solar wind speed increase and the recurrent structure on the basis of solar rotations (shading) are caused by a stable southern polar coronal hole extension observed by Yohkoh/SXT and its associated fast solar wind stream observed by IMP 8 in Earth orbit in early 1994. In Carrington rotation (CR) 1880 and 1881 an equilibrium is reached with the beginning of the decay of the coronal hole extension. Here ballistic backmapping should most accurately map the rarefaction region back to its coronal source longitude. Figure l b shows backmapped solar wind data of Voyager 2 and IMP 8 along with synoptic maps of the solar corona by Yohkoh/SXT for two consecutive CRs. The origin of the rarefaction is found at ~230 ~ Carrington longitude (CL) in the synoptic map. IMP 8 observations reveal a dwell in backmapped solar wind data at these CLs in CR 1880 (data gap in 1881). For Voyager 2 observations this dwell is shifted from 230 ~ CL in CR 1881 to 40 ~ CL in CR 1880. Streamlines, which, according to the Parker model, are associated with the magnetic field Archimedian spirals, would map back to their coronal source. If the solar wind is slowing down, for example by the interaction with pickup ions, the backmapping based on a Parker spiral associated with the observed solar wind speed is more tightly wound than to the actual stream line. Therefore the backmapping gives a deviation to the west of the actual source of the solar wind. We quantify the slowdown of the solar wind to about 10% for the 600 km/s solar wind observed in the rarefaction. The original speed in the inner heliosphere was ~667 km/s. 3. D E R I V I N G

THE INTERSTELLAR

NEUTRAL

DENSITY

Given the slowdown in solar wind speed (~10%), presumably due to mass loading from interstellar pickup ions, we may solve a simple model to infer the interstellar density. The equations below describe the mass, momentum, and energy flow in the solar wind: r 2 d r ( r 2 p u ) - ~3prop

1 r 2 dr

dP r

(2)

+ d-7 = --~Tnnp~t2

5p) r2

(1)

nn

_

1

7

O)

[12]. According to [13] the photoionization rate is given by 3p - 0.9.10-7/s. After charge

-316-

Relationships of corotating rarefaction regions outside 40 A U...

Figure 1. a) Voyager 2/PLS observations at 43 AU, 12~ S heliographic latitude during 1994. From top to bottom: (top) one hour averages of the solar wind speed, shaded according to the consecutive association to Carrington rotation of solar wind release, (middle) 50 day running mean of the speed (light) and proton density (dark), and (bottom) the solar wind proton density shaded as in the top panel. Carrington rotation numbers are given on top. b) Yohkoh/SXT synoptic maps and backmapped solar wind data for Carrington rotations 1880 and 1881. The middle panel shows the solar wind speed (thick black lines) and density (thin gray lines) observations by IMP 8 in Earth orbit. The dominant magnetic field polarity is indicated according to IMP 8 in situ observations. The fast solar wind stream for these two consecutive rotations tracks the longitudes of the southern polar coronal hole extension with a rarefaction region at ~240 ~ CL in CR 1880 (data gap in CR 1881). The lower panel shows backmapped Voyager 2 solar wind speeds (thick black lines) and densities (thin gray lines) from a distance of 43 AU. Two vertical lines are connected with a horizontal. The left vertical line indicates the approximate source longitude of the rarefaction region identified in the Yohkoh/SXT synoptic map. The right vertical line indicates the dwell longitude of the rarefaction region solar wind observed at Voyager 2. The horizontal line itself indicates the difference in the longitudes that is equivalent to the time delay At = Ak/co e.

-317-

A. Posner, N.A. Schwadron and T.H. Zurbuchen

Figure 2. The interstellar neutral density depending on the slowdown of the solar wind from 667 km/s at 43.5 AU distance from the Sun. The error bars indicate the statistical uncertainty of our method to quantify the solar wind slowdown and the tolerance of the parameters used for the numerical integration.

exchange or photoionization the ions of interstellar origin are picked up by the frozen-in solar magnetic field. The momentum of these ions is transferred to the solar wind, which leads to a slowdown of the bulk solar wind. The location of Voyager 2 in 1994 is ~30 ~ off the interstellar upstream direction. We derived the slowdown of the solar wind to ~10% (667 km/s to 600 kin/s) based on the observations of corotating rarefaction regions at 43 AU combined with the ballistic backmapping technique. Numerical integration of the equation system leads to the dependency of the interstellar neutral density after passage through the termination shock on the slowdown of the solar wind at 43 AU shown in Fig. 2. With our method we derive the interstellar neutral hydrogen density, as the major contribution for the momentum transfer to the solar wind, to 0.127 cm-3+ 0.019 cm -3. The given error refers to the statistical error in the observed speed. Our density estimate is consistent with the result of Gloeckler et al. [14] of 0.115 cm-3+ 0.025 cm -3. 4. C O N C L U S I O N S An attempt was made to associate the recurrent structures in the solar wind speed profile at Voyager 2 with coronal observations. The solar wind appeared to be slowed down. We interpret this slowdown to be according to mass loading of the solar wind with interstellar pickup ions. By comparison of the backmapped structures with the source of these structures in the corona we derived a slowdown in the order of 10% of the original solar wind speed. A numerical integration of the solar wind mass, momentum, and energy equations led to an interstellar neutral density of 0.127 cm-3+ 0.019 cm -3, which is close to the result of Gloeckler et al. [14] based on observation of pickup ions with Ulysses/SWICS. Other methods, like observations of resonant back-scattering of solar UV radiation, give results ranging from 0.03 to 0.3 cm -3 [15], [16] and suffer from

-318-

Relationships of corotating rarefaction regions outside 40 A U... the dependence on complex back-scattering models with poorly known parameters as well as on the uncertainty of a galactic UV emission background. In our method to determine the slowdown of the solar wind we cannot entirely exclude the possibility that other causes, for example merging shocks, made a significant contribution. For the following reason it appears unlikely that MHD effects may have affected the results significantly. The contribution of CIR-associated shocks on the high speed solar wind (667 km/s) could only lead to a further slowdown of the solar wind, therefore our method would still give an upper limit to the pickup ion momentum transfer and interstellar neutral density. However, attempts are planned to constrain the slowdown of the solar wind for Voyager 2 observations with improved and more sophisticated MHD codes [17] in the near future. A more detailed version of this paper can be found online at http://www-p ersonal, umich, edu/~ ap o/isdens.html.

~

.

3. 4.

7. 8. 9. 10. 11. 12.

17.

D. A. Bryant and T. L. Cline and U. D. Desai and F. B. McDonald, Phys. Rev. Lett., 14 (1965), 481. C.W. Barnes and J.A. Simpson, Astrophys. J. Lett.,210 (1976), L91. A. S. Krieger and A. F. Timothy und E. C. Roelof, Solar Phys., 29 (1973), 505. I. G. Richardson and L. M. Barbier and D. V. Reames and T. T. van Rosenvinge, J. Geophys. Res., 98 (1993), 13. P. R. Gazis and F. B. McDonald and R. A. Burger and S. Chalov and R. B. Decker and J. Dwyer and D. S. Intrilligator and J. R. Jokipii and A. J. Lazarus and G. M. Mason and V. J. Pizzo and M. S. Potgieter and I. G. Richardson and L. J. Lanzerotti, Space Sci. Rev., 89 (1999), 269. Z. K. Smith and M. Dryer and R. S. Steinolfsen, J. Geophys. Res., 90 (1985), 217. Y. C. Whang, J. Geophys. Res., 103 (1988), 17,419. V.J. Pizzo, J. Geophys. Res., 99 (1994), 4173. V.J. Pizzo, personal communication (2001). L. F. Burlaga and N. F. Ness, J. Geoph. Res. 101 (1996), 13,473. P. Isenberg, J. Geophys. Res., 91 (1986), 9965. C. F. Barnett, Atomic Data for Fusion, Tech. Rep. ORNL-6086, Vol. 1 Oak Ridge National Laboratories, TN, 1990. D. Rucifiski and A. C. Cummings and G. Gloeckler and A. J. Lazarus and E. MSbius and M. Witte, Space Sci. Rev., 78 (1996), 73. G. Gloeckler and L. A. Fisk and J. Geiss, Nature, 386 (1997), 374. D.P. Cox and R.J. Reynolds, Annu. Rev. Astron. Astrophys., 25 (1987), 303. E. Quemerais and J.-L. Bertaux and B.R. Sandel and R. Lallement, Astron. Astrophys., 290 (1996), 941. P. Riley, personal communication (2000).

-319-

This Page Intentionally Left Blank

R. B. Decker a, C. Paranicas a, S. M. Krimigis a, K. I. Paularena b, and J. D. Richardson b ~Applied Physics Laboratory, Laurel, MD 20723, U.S.A. bMassachusetts Institute of Technology, Cambridge, MA 02139, U.S.A. Models of heliospheric shock propagation predict that pickup ions dominate the pressure in the outer heliosphere and play a key role in weakening CIR-associated shocks beyond a few tens of AU. To test this prediction, we used Voyager 2 PLS and LECP data to examine the temporal relationship between recurrent (~ 26-day) intensity increases of energetic ions and shocks or shock remnants. We focused on the periods 1983-1985 (11-19 AU) and 1992-1994 (36-45 AU). We found that ion intensity peaks lag further behind shocks or shock remnants during the later period, consistent with model predictions. 1. I n t r o d u c t i o n Forward and reverse shocks bounding corotating interaction regions (CIRs) accelerate ions to energies of at least 10 MeV/nucleon [1]. Between ~ 3-20 AU, intensity peaks of these recurrent (~ 26-day) ion events correlate with the passage of shocks, as inferred from plasma and magnetic field data. Surveys of data from the Pioneer and Voyager spacecraft show that beyond ~ 15-20 AU, the occurrence rate and strength of CIR shocks decrease [2,3]. In previous work, we used Voyager 2 data beyond 30 AU to show that CIRs or their remnants are still associated with well-defined peaks in quasi-recurrent ion intensity increases. However, the temporal association between the measured abrupt increases in plasma flow speed and ion intensity peaks becomes less clear farther from the Sun [4]. Models of shock propagation in the heliosphere show that beyond a few tens of AU, pickup ions dominate the pressure, thereby weakening CIR-associated shocks [5]. This weakening reduces a shock's efficiency for injecting and accelerating ions. A shock with speed Vs will outrun the intensity peak of accelerated ions (formed when the shock was relatively strong) that convects outward at the post-shock (downstream) solar wind speed Vd < Vs. Thus, peak ion intensities should lag the shock or shock remnant in the outer heliosphere when the contribution to the total pressure from pickup ions dominates [6]. 2. Observations The Voyager 2 LECP data we discuss herein are 6-hour averaged intensities of 0.521.45 MeV protons. These data have been carefully corrected by removing background counts due to penetrating cosmic rays. We have also examined intensities of eight Z>_1 ion channels, 0.040-4.0 MeV. These data corroborate our conclusions based on analysis of the proton channel. To determine if the predicted lag is consistent with observations, we

-321 -

R.B. Decker et al.

Figure 1. Quasi-recurrent (~ 26 day) increases in 0.52-1.45 MeV proton intensities during" 1983-86 (top) and 1992-95 (bottom).

analyzed Voyager 2 data during two time periods, one solar activity cycle apart. These periods are 1983-1985 and 1992-1994, when Voyager 2 was between 11-19 AU and 36-45 AU, and lie within the declining phases of the solar activity cycles during the peak CIR "season." The intensity of 0.52-1.45 MeV protons during the two periods of interest is shown in Figure 1. Vertical lines indicate arbitrary 26-day intervals. If the model is correct, the time lag between observed ion intensity peaks and signatures of shocks or shock remnants should increase with radial distance from the Sun. For each time period, we selected candidate shocks from the Voyager 2 data set and estimated the lag time between that measurement and the measured ion intensity peak. Only single peaks were chosen for this study and the times used were located at the mean along the FWHM of the peak. Representative data are presented in Figure 2. CIR-associated shocks weaken in the outer heliosphere [4], so changes in solar wind parameters become less discrete at larger helioradii. We are thus comparing shocks in the first period with shock remnants in the second period. As Figure 2 reveals, the size of the flow speed changes associated with a shock remnant are often comparable to those associated with a shock. For simplicity, we will refer to both kinds of discontinuities in the solar wind parameters as "shocks" in this paper. Shocks identified for this study originated from a candidate list of discontinuities in solar wind parameters. The P LS team compiled this list by identifying changes in solar wind speed, total density, thermal ion temperature, and, when available, magnetic field strength from the MAG instrument. The specificity of these criteria decline with time due to changes in the nature of the shocks, the solar wind parameters, and the sensitivities of the various instruments. The list of candidate discontinuities was therefore more comprehensive for the earlier study period, 1983-1985. For this reason, we used energetic particle data to expand the list of shocks in the outer heliosphere. The presence of sig-

- 322-

Recurrent ion events and plasma disturbances et Voyager 2: ...

Figure 2. Top to bottom" Solar wind speed, energetic proton rates, and delay time during 1984 (left) and 1994 (right). Vertical lines show shock or shock remnant passage. Dots in lower panels show delay times (ordinate) and times of particle intensity peaks (abscissa).

nificant intensity peaks occurring nearly simultaneously with changes in the solar wind parameters were regarded as shocks. We used the criteria that proton increases must have single peaks and that the peak count rates exceed 0.02 c s -~. This also allowed us to identify shocks where the plasma instrument had data gaps, but where the criteria for a shock was satisfied on both sides of the gap. The modified candidate list produced 17 (13) events in the 1983-1985 (1992-1994) period. We also assessed the variation of peak proton intensity with heliospheric distance. These results are shown in Figure 3(a). Only proton peaks associated with shocks were selected. Figure 3(a) shows that between these two study periods, the peak proton intensity decreased with helioradius as r -2"9. Time delays are shown in Figure 3(b). 3. S u m m a r y and conclusions During 1983-1985, the average delay time between the shock and 0.52-1.45 MeV proton intensity peak was ~ 0.6 days (5= 0.9 days) whereas during 1992-1994, it was ~ 3.4 days (+ 2.8 days). This radial increase in time lag, which is consistent with model predictions [6], and the c< r -3 peak intensity decrease have implications for the processes of shock evolution and particle transport in the outer heliosphere and for the expected intensity of CIR-associated seed ions incident on the heliospheric termination shock. Energetic

- 323 -

R.B. Decker et aL

Figure 3. Peak intensity (a) and delay time (b) of 0.52-1.45 MeV protons versus helioradius for quasi-recurrent events observed during 1983-1985 and 1992-1994.

ions are accelerated closer to the Sun, before the CIR shocks have dissipated, and then convect outward with the solar wind, undergoing spatial diffusion, adiabatic deceleration, and possible reacceleration by solar wind turbulence and other interplanetary shocks. When they reach the termination shock, a small fraction of these superthermal ions will be accelerated to ACR (anomalous cosmic ray) energies (~ 10-100 MeV/nucleon). This work was supported at APL by NASA Grant NAG5-4365 and at MIT by NASA Grant NAG-6473 and NASA Contract 959167. REFERENCES

1. 2. 3. 4. 5. 6.

G.M. Mason, et al., Space Sci. Rev. 89 (1999) 327. P.R. Gazis, et al., Space Sci. Rev. 89 (1999) 269. P.R. Gazis, J. Geophys. Res. 105 (2000) 219. A.J. Lazarus, et al., Space Sci. Rev. 89 (1999) 53. G.P. Zank and H.L. Pauls, J. Geophys. Res. 102 (1997) 7037. W.K. Rice, et al., Geophys. Res. Lett. 27 (2000) 509.

-324-

Mapping the Detailed Structure of the Local Interstellar Medium S. Redfield ~ and J.L. Linsky b ~JILA, University of Colorado Boulder, CO 80309-0440, USA bjILA, University of Colorado and NIST Boulder, CO 80309-0440, USA High resolution Hubble Space Telescope (HST) absorption spectra are used to construct a model of the three-dimensional shape of the Local Interstellar Cloud (LIC). We use spherical harmonics as our basis functions and apply a least-x 2 fitting routine to determine the best fit. This technique provides a powerful means for determining the true structure of the Local Interstellar Medium (LISM) by providing predictions of the amount of LIC material for any arbitrary line of sight. As a first attempt, our model is successful at modeling the LIC with a quasi-spherical shape, having a volume of ~ 93 pc 3 and a mass of ~ 0.32 3//o. An axis of symmetry in the direction of 1 ~ 315 ~ may indicate that the shape of the LIC is influenced by the flow of hot gas from the Scorpius-Centaurus Association. Our model places the Sun near the edge of the LIC at a distance of _< 0.19 pc, and the Sun should encounter the boundary of the LIC in _< 5000 years. Implications for applying this technique to other nearby clouds are also discussed. Physical parameters of the LIC are listed. The homogeneity of the LIC is discussed with reference to a sample of Hyades stars observed with the Space Telescope Imaging Spectrograph (STIS) aboard HST. 1. L I C M o d e l We develop a three-dimensional model of the LIC by calculating distances to the edge of the LIC from high resolution HST spectra, and using spherical harmonics to fit all available observations [1]. We adopt the technique described in [2] in which we determine the distance to the edge of the LIC, dedge(LIC), along a given line of sight from the hydrogen column density (inferred from the deuterium column density in HST spectra). We assume that the interstellar gas moving with the LIC speed has a constant density, rtHI = 0.10 cm -a, and the LIC extends from the heliosphere to an edge determined by the value of Nm(LIC) along each line of sight. We fit spherical harmonics to 32 lines of sight for which we have values of d e d g e ( L I C ) o r upper limits from HST, EUVE, and Ca II data. The LIC model is clearly not a long thin filamentary structure such as those seen in optical images of some interstellar clouds (e.g., reflection nebulae in the Pleiades), but neither is it spherical in shape. As seen from the North Galactic Pole, the LIC has an axis of symmetry that points in the direction 1 ~ 315 ~ (see Figure 2 in [1]). Since the

-

325

-

S. Redfield and J.L. Linsky

Figure 1. The location of the HST, EUVE, and Ca II stars used in this analysis are shown in Galactic coordinates. The symbols identifying the stars are listed in the first columns of Tables 1 and 2 of [1]. The shadings indicate values of NHI in units of 10 is cm -2 from the Sun to the edge of the LIC, based on Data Set B. From darkest to lightest, the shadings designate > 2.0, 1.0-2.0, 0.5-1.0, 0.25-0.50, 0.10-0.25, 0.05-0.10, and < 0.05 in these units. Figure from [1].

direction of the center of the Sco-Cen Association is 1 - 320, the shape of the LIC could be determined by the flow of hot gas from Sco-Cen. The LIC model shows that the Sun is located just inside its edge in the direction of the Galactic Center and toward the North Galactic Pole. The absence of Mg II absorption at the LIC velocity toward a Cen indicates that the distance to the edge of the LIC in this direction is < 0.19 pc and the Sun should cross the boundary between the LIC and the G cloud in less than 5,000 yr [3,4]. The LIC model can be seen at the Colorado Model of the Local Interstellar Medium web site http://casa.colorado.edu/~sredfiel/ColoradoLIC.html. The input data, prescription for computing the model, and a tool for calculating the hydrogen column density in any direction are also at this web site. As new data appear and we can model other warm clouds, updated versions of the LISM model will be placed in this web site.

-

326-

Mapping the detailed structure of the local interstellar medium

Table 1 Physical and Morphological Properties of the LIC Value Property 7000 + 1000 K Temperature 0.10 cm -3 Neutral hydrogen density (nHi) 0 911+0-12 Electron density (he) ~-0.06 cm -3 0.52-+-0.18 Hydrogen ionization (X(H) = np/(nHi + rip)) +1280 Gas Pressure (P/k) 1620_630 cm - 3 K -1.1 + 0.2 Log (depletion of Mg), D(Mg) -1.27 Log (depletion of Fe), D(Fe) -o.25 Log (depletion of O), D(O) 6.2pc Maximum dimension 4.7 pc Minimum dimension 1.9 x 1018 cm -2 Maximum NHI 1.5 x cm -2 Minimum NHI

Reference [5] [2] [6]

[1] [1] [5] [7]

[7] [1] [1] [1] [1]

1.1. P h y s i c a l P r o p e r t i e s

Table 1 lists our adopted empirical properties of the LIC and the references from which these data were taken. The physical parameters and hydrogen column density of the LIC are roughly consistent with the warm ISM models that assume pressure and ionization equilibrium [8], but the empirical hydrogen ionization is much higher and the gas temperature lower than the theoretical models predict. The high ionization is naturally explained by the LIC gas being in a recombining phase following shock ionization from a nearby supernova as proposed by [9]. The higher ionization increases the gas cooling, which can explain why the gas is 2400 K cooler than the ionization equilibrium models predict. Computed and observed temperatures are remarkably in agreement for a theoretical model with the observed LIC electron density.

1.2. H o m o g e n e i t y "

the Hyades Sample

The Space Telescope Imaging Spectrograph (STIS) instrument aboard HST observed 18 members of the Hyades star cluster. This dataset was acquired under observing program 7389 with principal investigator E. Bohm-Vitense. Only the Mg II h and k lines were observed. Because all 18 stars are members of the Hyades cluster, their lines of sight are very closely spaced on the sky. Thus, they sample a very small region of the LIC, and therefore offer a unique opportunity to study the homogeneity of the LIC. Figure 1.2 shows the difference in Mg II column density as a function of angular distance. The Mg II column density difference does not consistently exceed a factor of two for stars less than 8 ~ away. The hydrogen column density calculated by our LIC model [1] successfully predicts the gradient of the Mg II column density in the Hyades sample [10].

-327-

X RedfieM and J.L. Linsky

Figure 2. The difference in Mg II column density as a function of angular distance. With 18 target stars, there are 153 baselines, and therefore many angular distances to compare. The Mg II column density difference does not consistently exceed a factor of two for stars less than 8~ away. Figure from [10].

REFERENCES 1.

2. 3. 4. 5. 6. 7. 8. 9. 10.

Redfield, S., & Linsky, J.L. 2000, ApJ, 534, 825 Linsky, J.L., Redfield, S., Wood, B.E., & Piskunov, N. 2000, ApJ, 528, 756 Linsky, J.L., Wood, B.E. 1996, ApJ, 463, 254 Wood, B.E., Linsky, J.L., & Zank, G.P. 2000, ApJ, 537, 304 Piskunov, N., Wood, B.E., Linsky, J.L., Dempsey, R.C., & Ayres, T.R. 1997, ApJ, 474, 315 Wood, B.E., & Linsky, J.L. 1997, ApJ, 474, L39 Linsky, J.L., Diplas, A., Wood, B.E., Brown, A., Ayres, T.R., & Savage, B.D. 1995, ApJ, 451,335 Wolfire, M.G., Hollenbach, D., McKee, C.F., Tielens, A.G.G.M, & Bakes, E.L.O. 1995, ApJ, 443, 152 Lyu, C.-H., & Bruhweiler, F.C. 1996, ApJ, 459, 216 Redfield, S., et al., in progress

- 328-

Effect of Different Possible Interstellar Environments on the Heliosphere: A Numerical Study H.-R. Miiller ~, G. P. Zank ~, and P. C. Frisch b* ~Bartol Research Institute, University of Delawar(i, Newark, DE 19716 bDepartment of Astronomy and Astrophysics, University of Chicago, Chicago, IL 60637 At present, the heliosphere is embedded in the local interstellar cloud (LIC) which is warm (8000 K) and moderately dense (~0.2 cm-3). The sun will leave the LIC in perhaps about 10,000 years. This suggests that during its lifetime, the heliosphere has been, and will be, exposed to different interstellar environments. By means of a multi-fluid model, the interaction of the solar wind with a variety of partially ionized interstellar media is investigated. We assume that the basic solar wind parameters remain/were as they are today but we consider a range of ISM parameters (hot and cold, low density and ultrahigh density). In response to different interstellar boundary conditions, the heliospheric structure changes, as does the abundance of neutrals in the inner heliosphere, with possible implications for pickup ions and cosmic rays in the vicinity of Earth. 1. I N T R O D U C T I O N The Sun and its heliosphere is currently embedded in a warm interstellar cloud that has a relative Sun-cloud velocity of 26 km s -1. The main components of this cloud are neutral hydrogen (H) atoms and a plasma consisting of protons and electrons. The interaction of the local interstellar medium (LISM) with the fully ionized solar wind gives rise to the heliospheric morphology which includes the heliopause HP (a contact discontinuity separating solar wind and LISM), the termination shock TS (where the solar wind becomes subsonic and is diverted downstream to form a heliotail), and, when the LISM plasma is supersonic, a bow shock upstream of the heliopause. These general boundaries are created by the plasma interaction, yet the presence of neutral H and its coupling to the plasma protons via charge exchange greatly influences the details of the heliospheric morphology and the location of its boundaries (see [1] for a review). By studying the absorption of stellar profiles of nearby stars (within ..o100 pc) at various wavelengths along many lines of sight, it is possible to explore the fine structure of the interstellar medium surrounding the solar system [2]. There is evidence for different clouds and cloudlets, ranging in scale from fewer than 1 pc to tens of pc [3]. Some of these clouds possess characteristics that are quite different from those found in the local interstellar cloud (LIC). The clouds' inferred velocities can be different, their temperatures can be *This research was supported by NASA grant NAG5-6469, NSF-DOE award ATM-0078650,JPL contract 959167, and NASA Delaware Space Grant College award NGT5-40024.

- 329-

H.-R. Miiller, G.P. Zank and P.C. Frisch

higher or lower, and their densities and ionization states can span a wide range of values. Moreover, it is conceivable that the interstellar magnetic field, which is difficult to infer, differs from region to region. The Sun entered the cloud system of the so-called Scorpius-Centaurus association some 105 years ago [3]. There are indications that the Sun will leave our current immediate LIC in under 104 years from now [2]. This suggests that throughout Earth's history the solar system has moved between different clouds (or, depending on the viewpoint, different clouds have flowed past the solar system). Different interstellar environments should be capable of producing noticeable changes in the environment at 1 AU, as indicated by the amount of neutral H, anomalous (ACR), and galactic cosmic rays (GCR) at 1 AU. It has been suggested [4] that lunar soils contain an archive of elemental abundances that are different from the particle environment of the present era. Antarctic ice cores show signatures that may be interpreted as cosmic ray background variations at Earth [5]. These possibilities have motivated us to study how the global heliosphere would behave under LISM conditions that are different from today's. We undertake a numerical study, probing the parameter space of LISM density, temperature, and velocity. The parameter space is huge, of course, and we present the results of ten choices that are somewhat arbitrary, but are in a range suggested by interstellar clouds elsewhere or by other star's different velocity with respect to their LISM.

2. M O D E L S

We use the multifluid model developed by [6] to generate ten heliospheric models. The multifluid code tracks the plasma component and three thermodynamically distinct populations of neutral H. The models are essentially three-dimensional with spherical geometry; however, we assume azimuthal symmetry about the stagnation axis (the axis parallel to the LISM flow that contains the Sun) which reduces the dimensionality to two. The inner radial boundary condition is a prescribed solar wind plasma that corresponds at 1 AU to an averaged plasma density of 5.0 cm -3, a temperature of 10~ K, and a radial velocity of 400 km s -1. Table 1 lists the varied LISM boundary conditions. The first three models represent the effect of increasing relative LISM velocities, ranging from 26 to 100 km s -1. The remaining seven models are ordered in decreasing heliospheric size, where the upstream termination shock location ranges from 180 to 9 AU. The LISM parameters have been inspired by real astrophysical systems. Model 1 corresponds to the e Ind system (see Mfiller, Zank, and Wood contribution in this volume), model 6 to c~ Cen. Models 0, 4, and 5 represent variants of the contemporary heliosphere. Model 2 is an example of a superbubble shell, a possible LISM environment that is the result of a nearby supernova explosion sweeping up the outflow of a starforming region. Models 3 and 7-9 correspond to several different types of clouds in the Orion association that have been observed in the ISM in the direction of 23 Orionis [7]. All ten models have a supersonic LISM plasma, which leads to the formation of a bow shock where the plasma is heated and decelerated. All models have a termination shock, a heliopause, and a hydrogen wall. As an example, Figure 1 shows data from model 9, with boundaries marked in the plot. In upstream directions, the hydrogen wall is typically confined between the HP and the bow shock (BS). In downstream directions,

-330-

Effect of differentpossible interstellar environments on the heliosphere:...

Table 1 Model parameters and results. LISM parameters X nH np T v cm -3 K km/s

6 0.4 0.14 0.10 7 0.5 0.10 0.10 8 ~0 11.00 0.15 9 ~0 15.00 0.20 Ionization fraction ) / =

Morphology TS HP BS AU AU AU

5650 25 98 8000 26 86 100 26 18 3000 26 9.5 np/(nH +np); nil, np,

Neutrals/Pickups nH(TS) npI(TS) nH(1AU) cm -3 cm -3 cm -3

140 280 0.04 120 245 0.04 32 130 1.10 15 35 1.70 npi: number density H,

0.00007 0.0009 0.00007 0.001 0.008 0.08 0.02 0.27 plasma, pickup ions.

the enhancement in neutral H typically does not follow the HP. The TS either has a bullet shape (models 1, 2, 4, and 7) or is spherical (models 0, 3, 5, 6, 8, and 9). The spherical case occurs when the LISM neutral H density is high (with the exception of model 1 where the high LISM speed creates a bullet shape), and is characterized by a heliosheath and heliotail plasma that is subsonic throughout. In contrast, in the bullet shape cases the initially subsonic plasma of the nose of the heliosheath accelerates to supersonic speeds like in a nozzle, shocks to subsonic speeds further downstream, and meets heliotail plasma at a contact discontinuity. There is a characteristic triple point where heliosheath shock, termination shock, and the contact discontinuity meet. 3. R E S U L T S

An obvious result of the comparison between the ten models is the variation in sizes of the heliosphere, as expressed in the location of upstream TS, HP, and BS in Table 1. A shrinking heliosphere is caused by the strength of the LISM ram pressure (dominated by the speed in cases 1 and 2) or similarly through the weakening of the solar wind ram pressure by an increased rate of charge exchange inside the heliosphere owing to an increased neutral number density (models 5, 8, and 9). Models 3 and 4 are especially large due to a lower LISM plasma ram pressure, and both a lower LISM temperature or a lower neutral density lets the heliosphere expand more. The responses of the heliosphere to these parameters are non-linear, and the inclusion of ACR acceleration and its feedback on the TS would further increase this non-linearity. There are several models listed in Table 1 whose heliosphere is so small that some of the outer planets find themselves beyond the TS in the hot heliosheath, at least for parts of their orbits (e.g. Fig. 1). The upstream distances from the Sun to the TS range from 9 to 180 AU, those to the HP range from 12 to 240 AU. The bow shock of a cool, tenuous LISM (model 3) is as far as 700 AU away from the Sun.

-331 -

H.-R. Miiller, G.P. Zank and P.C. Frisch

. . . . . . . .

i

. . . . . . . .

i

. . . . . . . .

:

"

'

. . . . . . .

i

. . . . . . . .

i

. . . . . . . .

T

10 s ++

101

nH

P + "

TH ,

"'.:'l'IP

TS

p

,

el__

101

H

m

if) i

i

i

I

i

i iii

i

i

i

i

i

i iii

i

i

I

i

. . . . .

i

i

i

i

i llI

i

i

i

i

....

|

i

i

i

i

i

i

i

i

Figure 1. Profiles of density (cm -3) and temperature (K) of model 9 along the upstream stagnation axis.

The density of neutral hydrogen that enters through the termination shock at the upstream stagnation axis varies over a wide range in the ten models; however the value of today's heliosphere (model 0, 0.04 cm -3) is reached or surpassed in all but the lowest density model 4. This finding even occurs at the fixed distance of 1 AU (last column of Table 1), suggesting that if the LISM hydrogen has been only slightly denser or faster in the past, there was a larger background of neutral H arriving at Earth with potential consequences for the terrestrial atmosphere [8,9]. We try to assess the consequences of different ISM environments for the ACR density at Earth by using the density of pickup ions at the TS as a proxy, calculated with a Vasyliunas-Siscoe type model [10]. The ACR and the GCR environment will influence planetary magnetospheres, terrestrial climate, atmosphere, and biology [11]. Again, most models have a higher value than that of the model of the current heliosphere. These values, however, are only a rough indicator of the terrestrial cosmic ray background. The shrinking distance to the heliospheric boundaries alone (models 7-9) will increase the GCR background [11], and the high neutral and pickup density in models 8 and 9 leads to a strong modification of the TS through the ACR acceleration processes. REFERENCES .

2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

G. P. Zank, Space Sci. Rev. 89 (1999) 413. J. L. Linsky, S. Redfield, B. E. Wood, & N. Piskunov, ApJ 528 (2000) 756. P. C. Frisch, Space Sci. Rev. 72 (1995) 499. R. F. Wimmer-Schweingruber & P. Bochsler, AIP Conference Proceedings 528 (2000). G. M. Raisbeck et al., Nature 326 (1987) 273. G. P. Zank, H. L. Pauls, L. L. Williams, & D. T. Hall, J. Geophys. Res. 101 (1996) 21639. D. E. Welty, L. M. Hobbs, J. T. Lauroesch, et al., ApJ Supp. 124 (1999) 465. M. Bzowski, H. J. Fahr, & D. Rucinski, Icarus 124 (1996) 209. G. P. Zank & P. C. Frisch, ApJ 518 (1999) 965. V. M. Vasyliunas & G. L. Siscoe, J. Geophys. Res. 81 (1976) 1247. K. Scherer, in: The Outer Heliosphere: Beyond the Planets Copernicus Gesellschaft (2000).

-332-

General Discussion Baranov to Richardson: Are pick-up ions included in your observations? Richardson: No, their fluxes are too low and they are too hot for us to see. Jokipii to Richardson: Given that the model does not even come close to represent the data, would it be fair to say that there is no real evidence for the wind slowing down? Richardson: The wind is definitely slowing down. Wc arc definitely seeing the 40 km/s decrease. We are just not seeing the 80 km/s decrease one would expect with George Gloeckler's numbers. Barnes to Richardson: With respect to the temperature plateau you see: isn't it true that there is a variation in energy in various forms between what's left of the stream structure, pressure imbalances and so on - isn't there sufficient cnerg3z to account for the heating? Richardson: That's certainly possible. But I would expect more heating near solar maximum rather than less. Marsch to Richardson: Did you look at the coronal conditions during the time period when the temperature drops? Because it's generally known that the slow solar wind has a lower tempertaure to start with, so there is a natural variation of the temperature in the source region of the solar wind. Richardson: Yes, it is possible that it is the difference between slow and fast wind temperatures. Fahr to Lallement: You gave us the view that we are sitting in one of these cloudlets and we are just moving along the periphery. But these cloudlets are co-existing with the hot medium around them, and you may suspect that there is a transition region. Is it possible that we move through such a transition region? Lallement: Yes, but we don't know. It is also possible that the cloudlets are touching each other, so that no hot medium is in between. One of the biggest problems in the physics of the Local Bubble is that nobody knows about the interfaces between the cloudlets and the hot gas. Cummings to Lallement: Do you assume a 40% ionization of helium in the local interstellar medium, or where do you get it from? Lallement: That is a consistent result from line-of-sight measurements to white dwarfs. But we cannot exclude a local variation. Dorman to Lallement: Do you have some information on the magnetic field in the cloudlets? Lallement: If there is a pressure equilibrium between the cool and the hot gas (and given the age of the local bubble there should be) the only plausible way to produce the equilibrium is to have a strong magnetic field at the boundaries of the cloud.

-333-

This Page Intentionally Left Blank

Session 6: View of Space Agencies and Companies

This Page Intentionally Left Blank

-337-

M. LeipoM

-338-

Solar sail technology development and application...

-339-

M. Leipold

-

3 4 0

-

Solar sail technology development and application...

-341

-

M. LeipoM

-342-

Solar sail technology development and application...

-

343

-

M.. Leipold

-

3 4 4

-

Oral papers and posters

CONTRIBUTIONS TO OUTER HELIOSPHERE MISSIONS IN FRAME OF THE GERMAN SPACE SCIENCE PROGRAMME O. R S h r i g German Aerospace Center, Directorate Space Programmes, D-53 183 Bonn. A short overview of the German Space Science Programme will be presented with emphysis on the past, present and future German engagements in heliospheric missions. Special attention will be given to the highlights, starting with the German/US project HELIOS, and the strong participation in the ESA projects ULYSSES and SOHO. German experiments have flown on the US Pioneer and Voyager Probes. But also ion release experiments and ground based experiments on German initiative have contributed to the picture of the sun which we have today. An outlook into the future will deal with the German intention to participate, e.g. in the missions STEREO, Solar Probe, Solar Orbiter. At last, solar sail technology may become a promising tool for future outer heliospheric missions.

DEEP SPACE LENGES

MISSIONS

AND

TECHNOLOGICAL

CHAL-

C. Schalinski and K. Eckardt Astrium GmbH, Friedrichshafen. Current and future missions require advanced technologies for instruments and spacecraft, e.g. formation flying of separated satellites with a high degree of onboard autonomy for interferometric missions like LISA or DARWIN. Astrium GmbH (former Dornier Satellite Systems) has acquired leadership among European Space companies as main industrial contractor for cornerstone missions in ESA's science program since the successful launch of ULYSSES. We will present progress on current projects and study results.

- 345 -

This Page Intentionally Left Blank

General Discussion Mewaldt to Leipold: What are the goals of the deployment [of a solar sail] in space? Are you also going to use the photon pressure to accelerate or navigate? Leipold: The objective is to keep it low cost, so we have to limit it to a pure deployment mission. Gruntman to Leipold: What are the physical limits on the minimum thickness of a solar sail? Leipold: The thinnest film samples that have been coated and that I am aware of are of the order of 1 micron. Genta to Leipold: What is the largest sail you think you can built with your boom design technique? Leipold: Right now it's 40 by 40 m. The problem with larger sails will be the load on the boom root. Marsch to Leipold: Since you expose the sail to a plasma: did you look at the problems of electrostatic charging? Leipold: Not yet. We have actually designed some experiments that will be performed this year. Sackheim to Leipold: Has anybody looked at a hybrid approach- i.e. an ion propulsion followed by a sail mode? Leipold: We have not looked at that at this point.

-347-

This Page Intentionally Left Blank

(1)

t~ C/3 O

or) Imio

~b

0 --,1

0

O0 ('b

This Page Intentionally Left Blank

The Probable Chemical Nature of Interstellar Dust Particles Detected by "CIDA" Onboard "STARDUST" J. Kissel a, F.R. Kruegerb, J. Silen c, and G. Haerendel a a Max-Planck-Institut fuer Extraterrestrische Physik, Giessenbachstrasse, Postfach 1312, D-85741 Garching, Germany b lngenieurbureau Dr. Krueger, Messelerstr.24, D-64291 Darmstadt, Germany c Finnish Meteorological Institute, Vuorikatu 24, SF-00101 Helsinki, Finland

The dust impact mass spectrometer CIDA onboard the NASA spacecraft STARDUST has detected five interstellar particles and recorded mass spectra during its first measuring period. Due to their unexpected complexity the analysis is an arduous task and requires new methods. However, the dominating substance class - namely, polymeric heterocyclic aromates and aliphates - and the particle masses could be determined. Both together are consistent with optical properties of those interstellar particles being able to reach the inner solar system.

1. INTRODUCTION On the 7th of February 1999 the spacecraft STARDUST had been launched from Cape Canaveral/F1. to its long journey to comet p/Wild-2, and further back to earth. Its main task in January 2004 is sampling cometary dust by aerogel plates made from silicon dioxide of very low density (approx. 0.03 g/ml), bringing them back to earth in January 2006. However, it is expected that the organic component of the dust may be altered or even be destroyed during the collection by the aerogel. As a result the information gathered will be limited to the mineral component, and to the isotopic distribution of all elements contained in the dust, except for silicon and oxygen, of course. In order to perform an in-situ chemical analysis of the cometary dust during encounter we implemented a dust particle impact mass spectrometer onboard, named CIDA (Cometary Impact Dust Analyzer). Not only will CIDA analyze the molecular composition of the dust of comet p/Wild-2, but also that of interstellar dust intercepted during its flight from earth to the comet and back to earth again. Due to the fact that the trajectories of the interstellar dust streams are often perpendicular or at least inclined relative to those of the interplanetary dust, there are periods of some months each during which the instrument is sensitive to interstellar

-351 -

Jr. Kissel et aL

dust exclusively. One of those periods is already completed, and the results are reported here. Before encounter with p/Wild-2 there are two more such periods, and, if all goes well, also during the spacecrafts's journey back home.

2. THE MEASURING PRINCIPLE OF "CIDA" Already as early as 1986 three dust impact mass spectrometers (PIA on board GIOTTO, and PUMA-1 and -2 onboard VEGA-1 and -2) encountered comet p/Halley. With those we were able to determine the elementary composition of the cometary dust. Molecular information, however, was only marginal. This was due to the fact that the relative velocities between each spacecraft and the coma of this retrograde comet, were extremely high (GIOTTO: 69 km/s; VEGA: 78 km/s). Because p/Wild-2 is a prograde comet, the velocity relative to STARDUST is only about 6 km/s. The velocities relative to the interstellar dust are estimated to be in the order of 25 km/s, but may be even lower due to the effect of the radiation pressure on these particles. In any case, with these velocity regimes CIDA is more sensitive to molecular rather than atomic information. Depending on the position and pointing of the CIDA instrument, interstellar, interplanetary, or cometary particles may impact on the about 150 cm 2 circular silver target (Fig.l). During impact they instantaneously evaporate forming molecular and atomic fragments (their yields depending on the impact speed, as already mentioned), which are in part electrically charged, either positively or negatively. By a voltage of U=+/-1 kV applied to the target either positive or negative ions are accelerated through the grounded acceleration grid. Due to the fact that the main chemical information is "contained" in the positive ions, only those are measured until encounter with the comet. Namely, the detection of negative ions is considered more risky due to the higher voltage required for the instrument's ion detector (which in any case measures secondary electrons after cathodic conversion). After encounter also negative ions can be measured during some flight periods. As a result special information about the oxygenand sulfur-chemistry of some organic functional groups (or mineralic parts) may be gained.

-352-

The probable chemical nature of interstellar dust particles

Fig. 1" CIDA-sensor schematics, for explanation see text. After acceleration, all ions of different molecular (or atomic) masses m (in dalton) possess the same kinetic energy E = 1/2 m v 2 = e U . Consequently, their velocity v in the subsequent drift space is inversely proportional to the square-root of their mass m, respectively (empirical evidence suggests ions with one single elementary charge e exclusively). However, a certain velocity distribution width is due to different initial energies of the ions produced. This uncertainty is compensated in first order by a reflector. Namely, those ions of the same mass m, travelling "too fast" due to their higher initial energy, will fly deeper into the electrostatic reflector than the others. Due to their thus longer flight path they arrive at the same time at the detector as the slower ones of the same mass m. Thus the total travel time (from the target all the way to the detector) t of each ion of mass m is directly proportional to the square-root of m. Thus yields t = a*sqrt(m) + b ; with "a" being a function of the flight path effective length and the voltages applied; in principle, "b" is given by the electronic signal travel times relative to the time t=0 for (an imaginary) mass m=0. The time spectrum of the ion current at the detector can thus be converted into a mass spectrum, exact knowledge of a and b provided. (If the impact time is u n k n o w n - which indeed is the case with three of the five impact events there are great mathematical difficulties to overcome in order to determine b.) The mass resolution m/delta-m is about 200. Thus the mass line to m=200 is still fairly resolved from that to m=201. The quality of the above transformation "time into mass" is best if there are

-353-

J. Kissel et al.

only spectral peaks due to INTEGER mass numbers AND a and b are in agreement with the actual parameters of the instrument. This transformation has been performed semiautomatically as follows. In the mathematical space ( a , b ) - due to the above transformation equation t=a*sqr(m)+b the fraction of those ion signals which corresponds to an interval of any integer mass number +- 0.1 daltons was calculated. If the time signals of the ions would have been randomly distributed over the entire time/mass regime, trivially about 20% of the ion flight times would be after transformation be found within the m+-0.1 windows. A good mass spectrum, however, would typically contain about 90% of all flight times in these mass windows. Indeed, near "canonical" (calculated by flight paths, voltages values, and pulse travel times) of a and b such transformations had been found with the impact mass spectra as discussed here. As a consequence, the quality of the transformation is good. The chemical information about the dust particle impacted is then "hidden" in the intensity distribution of the detector signals over the entire mass regime. In order to gain this information some other mathematical routines had been used, as described in the next chapter.

3. THE ANALYSIS OF THE MASS SPECTRA When analyzing substance mixtures in the laboratory, generally those mixtures are first (e.g., chromatographically) separated, and subsequently each single substance is analyzed mass spectrometrically. Naturally, any pre-separation is impossible with such fast dust particles. The mass spectrum generated by a particle impacting on the CIDA target thus represents the mixture within the particle as a whole for large impact velocities. With low impact velocities the mass spectra are more sensitive to the surface rather than to the interior of the particle. Consequently, one can only hope that at least an analysis of substance CLASSES majorizing the particle's matter is possible. Then it comes as a necessary condition for class analysis that the mixture is not too heterogeneous chemically. However, even in a favorable case this class type analysis challenges a completely new problem. ( - In this very context it appeared as a great help to us, that we, together with a group of chemists in Vienna, are just developing a chemometric method into the direction of analyzing functional groups in substance mixtures (1). As early as 1987 we made use of a principally similar, but very simple, method in analyzing the molecular ion contribution in p/Halley's dust impact mass spectra (2). However, a further methodological development is needed for our experiment COSIMA (COmetary Secondary Ion Mass Analyzer) onboard the ESA-spacecraft ROSETTA to comet Wirtanen, which will be launched in 2003. In rendezvous with that comet we will softly collect cometary dust on targets from metal black, and subsequently analyze it by Secondary Ion Mass Spectrometry (SIMS). As we know from laboratory experiments the chemical process of ion formation in SIMS is very much comparable to that in dust impact in the lower velocity (10 kN), moderate specific impulse (>800 s) bums, specific masses below 0.01 kg/kW should be attainable. Unlike proposed high energy density systems that have not yet proven feasible (e.g., fusion), generating a self-sustaining fission reaction is quite straightforward. All that is required is for the fight materials to be placed in the fight geometry--no extreme operating conditions or reaction drivers (e.g., lasers, magnetic fields) are needed. As an example of enabling performance, consider the velocities required by interstellar precursor missions such as a Kuiper object rendezvous (Figure 2). If these types of missions are to be completed in an acceptable time (e.g., _900 s. Extensive Russian experience also indicates that solid core nuclear thermal propulsion systems with specific impulses >900 s may be feasible. Liquid fissile fuels may enable Isp's >1200 s, and gas/plasma fissile fuels may enable Isp's >2500 s. ~3 3.3 Phase 1 system capable of providing 300 kWt currently under investigation One candidate for a phase 1 space fission propulsion system is the Safe Affordable Fission Engine (SAFE)-300. The SAFE-300 is a potential near-term, low-cost space fission propulsion system capable of producing 300 kWt even following multiple failures. The system consists of 284 fuel pins and 103 heat pipes. The outer diameter of both the fuel pins and the heat pipes is 1.27 cm. The fuel pin inner diameter is 1.1 cm. The dimensions of the hexagonal core are 25.4 cm across the fiats by 40 cm long. The system is fueled with uranium dioxide with an average smear density of 92% in the ground-fueled zone and 85% in the space-fueled zone. Uranium enrichment in both zones is 97%. A cross section of the SAFE-300 is shown in Figure 7. A picture of a SAFE-30 full-core primary heat transport test is shown in Figure 8. The SAFE-300 builds on experience gained from the heat pipe bimodal system effort that has been ongoing at Los Alamos National Laboratory (LANL) since 1995.

- 406 -

In-space nuclear power as an enabling technology for exploration...

Figure 7. Cross section of the SAFE-300.

Figure 8. Full core primary heat transport test of SAFE-30.

-407-

R. Sackheim et al.

The SAFE design is optimized to ensure that safety, cost, and schedule goals are met. The nuclear portion of the SAFE is virtually non-radioactive at launch ( 140 microns (Hauser et al. 1998), but shorter wavelengths could not observed because of the very bright foreground emission from zodiacal light. The zodiacal dust decreases with heliocentric radius and beyond 10 AU Interstellar Probe may be able to detect the CIRB at wavelengths 100 m tip-to-tip, a preamplifier, and a single 5-kHz wideband receiver. The payload processor would perform Fourier transforms to provide spectral information and average spectra, find peaks (as a means of identifying bursty plasma waves) and perform data compression. Solar Wind, Pickup Ion and Interstellar Plasma: The plasma instrumentation on the Interstellar Probe will map the solar wind up to the termination shock, follow its thermalization in the heliosheath, and detect the transition from solar to interstellar plasma at the heliopause. It should also determine the composition and velocity distribution of interstellar pickup ions and the interstellar plasma itself. Electron and Ion Sensors: The plasma instrument should include an ion sensor to measure beam-like plasmas in the sunward and anti-sunward directions and an electron sensor to measure core and halo electrons over-~4n. These sensors need to measure plasma temperature, density, speed, and pressure at intensities considerably lower than typical solar wind conditions. The energy range should exceed 10 keV/Q and the geometric factor should be > 10.2 cm 2 to allow good sensitivity for heavy ions. Because interstellar plasma is relatively cold (-~104 K), an energy per charge (E/Q) measurement will result in good mass per charge (M/Q) resolution. Suprathermal electron measurements (and the implied magnetic topology) will provide information on the nature of the termination shock and heliosheath. Due to telemetry limitations, pitch-angle distributions may be computed on board. Pickup-Ion and Interstellar-Plasma Distribution and Composition Sensor: This sensor will determine the elemental and isotopic composition of the LISM by observing interstellar pickup ions inside the heliosphere and interstellar plasma beyond the heliopause, including the isotopes of key refractory elements such as C, Mg, Si, Ca, and Fe. Science objectives include the ionization-state of the LISM and the nuclear and chemical history of interstellar

-459-

R.A. Mewaldt and P.C. Liewer

material. In addition, this instrument will provide the source composition of ions accelerated at the termination shock. While only moderate mass resolution is needed for elemental composition and ionization-state studies, a mass resolution of M/AM > 40 is required for isotope measurements. The opposing flow directions of pickup ions and interstellar plasma can be covered with separate sensors or with a top-hat design. Conventional time-of-flight sensors are adequate for elemental composition studies. Isotope resolution can be achieved with an isochronous time-of-flight instrument with 360 ~ field-of-view. To provide these capabilities within the resources indicated in Table 1 will require developments in electronics, high-voltage supplies and sensor materials. Interstellar Neutral Instrument: The objectives of the interstellar neutral sensor are to provide the density, flow direction, and temperature of interstellar H, O and C, and possibly He. In addition, resolution of deuterium would provide the important interstellar D/H ratio. These parameters will vary along the flight path from the heliosphere through the heliosheath, heliopause, and hydrogen wall, and on into the undisturbed ISM. While He enters unimpeded, H and O are affected by resonant charge exchange in the heliospheric interface region, and a key objective is to determine directly the H and O depletion along the trajectory through the hydrogen wall. The ionization-state of the VLISM will be determined by combining neutral and ionized abundances of key species such as H, O, C and He. These capabilities can be achieved by a sensor based on neutral to negative-ion conversion during reflection off a suitable conversion surface, combined with electrostatic deflection and acceleration, and subsequent time-of-flight analysis. Suitable surfaces are under study but require further development. Suprathermal Ions and Electrons: The suprathermal ion and electron sensors cover the energy range above the plasma regime where particles are accelerated out of the bulk distribution. Important objectives include measuring the injection and acceleration of pickup ions at the termination shock, the suprathermal extension of heated solar-wind and pickup-ion distributions in the heliosheath, and searching for new particle components beyond the heliosphere. The electron sensor will survey electron acceleration in the heliospheric boundary region. To fulfill these objectives requires overlap with the plasma instrtmaent at lower energies and the cosmic ray sensors at higher energies. A 4n angular acceptance is important to measure the expected distributions. To differentiate source populations elemental discrimination is necessary (at least H, He, C, O, Ne, Ar, Fe). Of these species H, He, O, Ne and Ar have a high ionization potential, while Fe has a low ionization potential and might originate from other sources. Cosmic Ray It, He, and Electron Sensor: This instrument will measure the energy spectra of cosmic ray protons, 3He, and 4He with energies from ~1 to 300 MeV/nucleon, where they are excluded from the heliosphere by the solar wind. It should also measure cosmic ray electrons from ~1 to 100 MeV and identify positrons over a more limited range for comparison with radio and gamma ray measurements of interstellar electrons. Because of dynamic range considerations it is best to measure these light species in a separate instrument from heavier elements (see below). Note that this energy range includes anomalous H and He that are accelerated at the termination shock, as well as cosmic ray ~H, 3He, and 4He in an energy range that is not accessible at 1 AU. One concept is described in DrSge et al (2001). Anomalous and Galactic Cosmic Ray Composition: Measurements of the composition of anomalous and galactic cosmic rays from-~1 to--300 MeV/nucleon can address key questions about the acceleration of particles at the termination shock and at supernova shock waves, the transport of particles in the ISM, and the origin of the accelerated material. To discriminate

- 460 -

Scientific Payloadfor an interstellarprobe mbssion between suggested sources of ACR ions and acceleration models requires measurements of abundant ACR species (He, C, N, O, Ne, and Ar) as well as rare species (C, Mg, Si, S, and possibly Fe) whose origin is controversial. Isotopic studies of C, O, Ne, and Ar will provide key information for understanding the origin and evolution of neutral interstellar material, complementing studies at plasma energies. Other objectives are to measure the composition and energy spectra of interstellar cosmic rays with energies that are excluded from the heliosphere by the solar wind ( 30 MeV, H and He isotopes from 4 to 130 MeV/nucleon, and positrons from ~ 0.1 to ~ 3 MeV. 1. Scientific O b j e c t i v e s

1.1. T h e Local Interstellar S p e c t r u m Cosmic rays diffuse through the galaxy before arriving at the heliosphere and finally at Earth. During propagation, nuclear interactions modify the composition of the primary cosmic rays. The diffusive radio synchrotron spectrum, as well as the gamma-ray spectrum produced by the interaction with ambient matter, reflect the energy density and the shape of the total electron spectrum in the galaxy. Fig. 1 displays the result of such an analysis

-471-

W. DrOge et al.

Figure 1. Calculated electron and positron spectra, together with modulated electrons observed close to Earth [5].1999).

Figure 2. Galactic cosmic ray protons measured by the Kiel Electron Telescope in 1995 close to Earth [2], and anomalous protons at 63 AU [4].

by Strong et al. [6] where the shaded part shows the uncertainty in the GCR electron spectrum. The uncertainty in these calculations increases with decreasing electron energy. Since low energy nuclei undergo significant modulation (see next paragraph), including large energy loss, while traveling from the local interstellar medium to an observer at Earth, the local interstellar spectrum is not measurable by space probes in the inner heliosphere, as displayed in Fig. 2. In the outer heliosphere measurements of the low energy nuclei spectrum, e.g. the hydrogen spectrum at 63 AU in Fig. 2, is dominated by anomalous cosmic rays (ACRs) so no inference can be made about the local interstellar spectrum within the heliosphere. Only a space probe penetrating into the local interstellar medium will allow us to determine the different nuclei LIS. 1.2. T h e t e r m i n a t i o n shock, a s o u r c e for e n e r g e t i c p a r t i c l e s Since the 1970's it has been established, that the neutral interstellar gas is ionized in the heliosphere and picked up by the solar wind. These pickup ions are getting accelerated to cosmic ray energies by the termination shock. Therefore measurements close to the termination shock are of special interest to investigate the properties of ACRs. These measurements are of astrophysical importance since the termination shock will be the only accessible shock to provide a model for similar, more energetic processes in supernova shocks. 1.3.

Cosmic

ray modulation

To reach the Earth, cosmic rays enter the heliosphere experiencing modulation by their interaction with the solar wind and its frozen-in magnetic field. While the transport of cosmic rays within the termination shock has been studied over the last decades, only little is known about the modulation beyond the termination shock. Recently McDonald et al. [3] found evidence for cosmic ray modulation in the region beyond the termination shock, by analyzing Voyager GCR and ACR time profiles. Studying the propagation of GCR electrons, positrons and nuclei through the heliosphere and the local interstellar medium would lead to important insights in transport processes in astrophysical magnetized plasma. Simultaneous measurements of the particles, as well as the local plasma

-472 -

A cosmic ray detector for an interstellar probe

conditions would be an important diagnostic for the microphysics of particle and magnetized low density plasma interactions.

1.4. Low energy (~ MeV) primary positrons Positrons are prevalent in the galaxy, as the recently observed galactic positron fountain illustrates. They also have been observed via diffusive 0.511 MeV gamma-ray annihilation radiation in the interstellar medium and in several discrete sources. Positrons in the MeV range from different astrophysical sources, as well as electrons, could be present in the local interstellar medium. The effects of the transport from their sources to the heliosphere can be studied in the particle spectra and, comparatively, in the derived photon radiation. These electromagnetic interactions limit the lifetime of the e + and ein the galaxy, a limitation not imposed on the nucleonic component. Combined with the nuclear secondary particles (e.g. antiproton, 2H, 3He, X~ the e+e - data can contribute to a new picture of particle confinement and transport in the Galaxy.

2. Experiment Description The objectives of our proposed cosmic ray detector for the Interstellar Probe are to provide the differential energy spectra of the interstellar cosmic ray hydrogen, deuterium, tritium, and both helium isotopes as well as electrons and low energy positrons. In addition, the instrument should be capable of resolving the flow directions of low energy electrons and protons to investigate remotely acceleration processes close to the termination shock and a possible bow shock. The instrument is a multi-element array of solid state detectors with anticoincidence to measure the energy spectra of electrons from ~ 0.1 to > 30 MeV, H and He from 4 to 130 MeV/nucleon, and positrons from ~ 0.1 to ~ 3 MeV. The total mass of the instrument is 2.3 kg, the total power consumption is 2 W, and the telemetry rate after onboard data compression is 3 bits per second. The sensor aperture points at an angle of 90 ~ with respect to the s/c spin axis. Figure 3 shows a schematic view of the telescope, which comprises two different functional units: a stack of four silicon detectors constitutes an entrance telescope, and a Cesium Iodide scintillator is used as a calorimeter. The four passivated ion-implanted detectors (D1-D4) define, together with the top part of the plastic anticoincidence detector (D6), the 46 ~ full width conical field of view with a geometric factor of 2.5 cm 2 sr. Detectors D1 and D2 are divided in multiple segments, permitting for sufficient corrections for path length variation to resolve isotopes of hydrogen and helium at energies below ~ 30 MeV/amu. Another important advantage of segmentation is the capability it provides to implement a commandable or self-adaptive geometric factor. On detection of high count rates, as can be expected as the ISP approaches the termination shock, the logic will disable all but the inner segments of detectors D1 and D2, reducing the effective geometric factor by a factor of 10 or more. Measurements of fluxes as high as 106 particles/(cm 2 s sr) will therefore be possible without significant dead time losses. The 45 mm thick CsI(T1) scintillation detector DS, read out by two groups of photodiodes, stops electrons up to ~ 30 MeV and hydrogen and helium nuclei up to 130 MeV/amu. The bottom part of the anticoincidence (D7) will allow particles stopping in D5 to be distinguished from penetrating particles. The whole stack of detectors is mounted in an aluminum housing.

-473 -

W. Dr6ge et aL

positrons electrons gamma rays hydrogen helium D1-D4 D5 D6-D7 Measurement techniques Geometry factor View cone Power TLM bitrate Weight Dimensions Figure 3. Cross section of the cosmic ray telescope with a possible track of a 1 MeV positron.

0.1 3 MeV 0.1 30 MeV 0.1 5 MeV 4 130 MeV/n 4 130 MeV/n Silicon Detectors CsI(T1) Scintillator Plastic Scintillators AE x E (e-, H, He), and annihilation (e+) 2.5 cm 2 sr 46 ~ full cone 2W 3 bps 2.3 kg 250 x 200 x 250 mm 3

Figure 1. Instrument Characteristics and Resource Requirements

Detector specifications are summarized in Table 1. The response of the telescope to electrons, positrons, protons, and helium nuclei was studied by means of a Monte-Carlo method using the CERN Library program GEANT 3 [1]. It is found that the energy range for measurements of electrons will be 100 keV through 50 MeV, with a probability of 95 % that all secondary electrons, created by 30 MeV primaries, are contained within the calorimeter. Low energy ( ~ MeV) positrons will be identified by the following method: when a positron stops and annihilates in D2 or D3, one of the two 0.511 MeV annihilation gamma rays may be absorbed in the Cesium Iodide calorimeter (D5), and the other gamma-ray might escape without interacting with the instrument. An example of a positron-like event signature would be a particle stopping in D3, no signal in D4, and the deposition of ~ 0.511 MeV in D5, as displayed in Fig. 3 for a 1 MeV positron. Preliminary results show that a 5% fraction of positrons in the MeV energy range could be detected out of the electron background.

REFERENCES R. Brun, F. Bruyant, M. Maire, A.C. McPherson, and P. Zanarini. GEANT3. CERN DATA HANDLING DIVISON, 1987. (DD/EE/84-1). 2. B. Heber and M.S. Potgieter. Adv. Space Res., 26(5):839-852, 2000. 3. F.B.McDonald, B. Heikkila, N. Lal, and E.C. Stone. J. Geophys. Res., 105:1-8, 2000. 4. F.B. McDonald. Space Sci. Rev., 1998. 5. M.S. Potgieter, S.E.S. Ferreira, B. Heber, P. Ferrando, and A. Raviart. Adv. Space Res., 23:467-470, 1999. 6. A.W. Strong, I.V. Moskalenko, and O. Reimer. Astrophys. J., 537:763-784, 2000. 7. W.R. Webber. Astrophys. J., 506:329-334, 1998. 1.

-474

Oral papers and posters

FUTURE

OBSERVATIONS OF THE OUTER HELIOSPHERE

M . H i l c h e n b a c h and H. Rosenbauer Max-Planck-Institut fr Aeronomie, D-37189 Katlenburg-Lindau, Germany. The next generation of space-borne particle instruments should enable us to deepen our understanding of the physics of the outer heliosphere. The new instruments should be capable to observe, for example, by remote and in situ observations the local interstellar medium and its interaction with the solar wind and energetic particles. We will discuss the challenges for instrument designs and possible detector concepts for the exploration of the outer heliosphere.

STATE-OF-THE-ART VANCED ANALOG SPHERIC PHYSICS

SOLID STATE ARRAYS AND ADMICROELECTRONICS FOR HELIO-

H . D . VOSS Taylor University, Upland, IN 46989.

Fundamental advances in the understanding of the heliosphere and magnetosphere are possible using pixel arrays of cooled solid state detectors (SSD) and analog microcircuits. The SEEP experiment on the $81-1 satellite achieved 1.5 keV SSD energy resolution with high-sensitivity thereby giving new insights into the microstructure of radiation belt particles (E 4 keV). The CEPPAD/SEPS spectrometer on the POLAR satellite has over 500 SSD pixels that map continuously, for the rst time, the source and loss cone with unprecedented high angular resolution. New PVDF particulate sensors and microcircults capable of burst and time of flight impact analysis have increased our understanding of orbital debris with the SPADUS experiment recently launched on the ARGOS earth satellite. The ADS energetic particle instrument included along with PVDF dust sensors in the SPADUS instrument pushes the temporal resolution down to fast 8 ms accumulation intervals. The HENA instrument (ENA) on the IMAGE satellite implements a new type of low noise, thin window, and 240 pixel SSD sensor with associated microchips that is decoupled from the fast (noisy) time-of-flight front end analyzer. The new generation of detectors and analog microelectronics will produce comprehensive images with simultaneous mass, energy, and charge information for remotely unraveling the dynamics of ENA in the heliosphere and magnetosphere.

-475

-

This Page Intentionally Left Blank

General Discussion Roelof to Hilchenbach: I am very optimistic that we will be able to interprete the images we will get from the outer heliosphere. But I would like to suggest that the direction of tomography as a concept for the extraction of information from these images is not very promising for two reasons. First, most inversion techniques assume an ideal and exactly known instrument response function. And the set of instruments you showed us is not going to have such very well defined responses. Second, there is another set of dimensions in the inversion problem you discussed. Most tomographic techniques assume that you are inverting a density which is a scalar. When we are talking about energetic neutral atoms or moving ions with doppler shifts in the emission, we are talking about a phase space distribution which we are trying to discover. Hilchenbach: I think we totally agree that we need a very good model to do this deconvolution, because even for the Sun is doesn't work yet. But my talk was really about future observations. It's not within the next ten years, but it's something we can start now. Kissel to Mewaldt- It just occured to me that if you have a 40 by 40 m solar sail you have a very efficient dust detector. So far we have about 1 particle per month - you will have 5000 a day. Mewaldt" That was definitely discussed. Our present plan is to drop the sail because we worry that it would interfere with the plasma and magnetic field measurements. Obviously, if you could figure out a way how to make use of the sail, there might be a portion of the mission where one would be able to do that. MSbius to Sandel: What are the accuracy requirements for the thickness of the multilayers? Sandel: They need to be controlled to a few Angstroem. Marsch to Sandel" W h a t you showed us was all imager with a limited field of view that looks at an object like the Earth' magnetosphere. Wouldn't you rather want to design a camera that has a large field of view? Sandel: This camera head has a ficld of view of 30 ~ That's about as far as you can push this. The IMAGE instrument uses three of these heads in order to cover the plasmasphere of Earth in a single exposure. Klecker to Voss" Your chips are radiation-hardened. How hard are they? Voss: Depending on tile money you want to pay you get different types of hardness. On IMAGE you were at about 8- 10 4 radhard, which is medium. If we go through the full radiation-hardening process we have had them as hard as 10 a radhard. Jokipii to DrSge: Just a comment: the distiguishing between positrons and electrons is a major problem throughout the heliosphere. DrSge: I completely agree with that. Moskalenko to Dr6ge: What is the limiting factor which restricts the energy range of positron detection?

-477-

General Discussion

Dr5ge: What limits the sensitivity of this kind of detector is, of course, the background of electrons. Our estimates show that you can detected positrons if they are at the few percent level. If it's below that it's very difficult. Krimigis to DrSge: Such spectrometers of the current generation weigh 10 kg or more. Where are the principal savings that makes your weight only 2.5 kg? DrSge: The most part of the weight comes from the Cs-Iodide, because you need mass to stop 20 or 30 MeV electrons. The electronics for a single instruments would be about 2 kg but it might be achieved that the electronics can be integrated for all the cosmic ray instruments.

-478-

S e s s i o n 11: C o n n e c t i o n s to E a r t h

This Page Intentionally Left Blank

The heliosphere as viewed from Earth E.N. Parker ~ a Enrico Fermi Institute and Depts. of Physics and of Astronomy, University of Chicago, Chicago, IL, 60637, USA The first sketches of the heliosphere were made nearly half a century ago in response to the recognition of universal solar corpuscular radiation. The hydrodynamical origin and structure of the heliosphere soon followed from the extended million degree solar corona. The heliosphere is understood today in terms of diverse interpenetrating particle populations. Almost all stars evidently create their own "astrospheres" more or less along the lines of our local heliosphere. 1. INTRODUCTION Perhaps the place to begin is the realization that almost all stars evidently have "astrospheres", all too transparent to be seen but undoubtedly as complex as our own heliosphere. The stellar wind is the creator of the astrosphere, just as the solar wind sweeps out the cavity in interstellar space that we call the heliosphere. Thus the origin of the heliosphere and the astrosphere traces back to the hydrodynamics of the million degree solar and stellar coronas. The solar corona appears to be created by the dissipation of mechanical and magnetic energy in the tenuous gas above the dense photosphere. It is that dissipation, evidently in the form of the microflaring in the magnetically "quiet" regions of the Sun, that creates the heliosphere. The staggering complexity of the convective and magnetic machinations on all scales down into the unresolved microstructure of the solar activity gives some idea of the mystery of the stellar corona and astrosphere. Indeed, the mystery does not stop with the microflaring, for we are in the dark as to the origin of the fibril magnetic fields that seem to drive the system from below the visible surface. With the variety of stellar types and circumstances that may be presumed to create stellar winds and astrospheres, the inquiry into the heliosphere and the extrapolation to other stars is bewildering. The first primitive model of the heliosphere was sketched some 45 years ago, and the subject has come a long way since that time with the advent of the space age. We begin, then, by noting that the heliosphere evidently has been in place since the formation of the Sun and Earth some 4.6 x 109 years ago. Unknown to classical astronomy, the heliosphere remained "silent" until the advance of technology and science first began to uncover its effects. Only in the last half century have we appreciated its existence. Then once we ventured into space the "silent" heliosphere became noisy indeed. There are a number of terrestrial effects, but in the early years they were more puzzling than informative. Some effects are obvious, e.g. the aurora, while others e.g. geomagnetic fluctuations, cosmic ray variations, etc. are detected only by scientific instruments. It was the geomagnetic storm that a century ago first suggested bursts of "solar corpuscular radiation" from the Sun, consisting mainly of protons and an equal number

-481-

E.N. Parker

of electrons to provide electrical neutrality. Otherwise space was regarded as a hard vacuum capable of supporting unlimited electric potential differences, at the same time that the zodiacal light was interpreted as sunlight scattered from about 500 free electrons/cm 3 at the distance of Earth (1AU). Then about half a century ago B iermann's ([ 1], [2]) studies of the anti-solar acceleration of comet tails led to his fundamental pronouncement of the perpetual universal emission of solar corpuscular radiation. The velocity of the solar corpuscular radiation had long been estimated at 10 a km/sec, from the time delay of a couple of days between the flaring on the Sun and the impact of the corpuscular radiation against the outer boundary of the geomagnetic field. B iermann inferred from the measured anti-solar acceleration of gaseous comet tails that the number density of the solar corpuscular radiation at the orbit of Earth is in excess of 103 electrons and ions per cm 3, later revised downward to perhaps as little as 500/cm a based on resonant charge exchange with the cometary atoms. This density seemed to be confirmed by the comparable interplanetary electron density inferred from the intensity of the zodiacal light, considered at that time to be Thomson scattering of sunlight by free electrons. So the solar corpuscular radiation was powerful stuff. Its impact against the geomagnetic dipole field was calculated to confine the field to a distance of about five Earth's radii on the sunward side. Leverett Davis ([6]) conceived the first sketches of the heliosphere, reproduced in Fig. 1, based on Biermann's declaration of universal solar corpuscular radiation. Davis referred to it as the "cavity in the galactic magnetic field", the term heliosphere originating only thirteen years later in an article by A. J. Dessler. From the existing estimates of the density and velocity of the solar corpuscular radiation Davis suggested that the corpuscular radiation pushed back the interstellar gas and field to a radius of the order of 200 AU. He recognized that the radius of the heliosphere would vary with the i 1-year magnetic cycle of the Sun, and he suggested that the varying size of the heliosphere was responsible for the observed variation of the cosmic ray intensity within the heliosphere.

Figure 1. Two sketches of the cavity in the galactic magnetic field (from Davis [6]) with different suggested solar magnetic field forms.

It should be noted here that the origin of the solar corpuscular radiation at the Sun was a mys-

-482 -

The heliosphere as viewed from Earth

tery at that time, with vague ideas about acceleration in or around the magnetic fields of active regions, sunspots, and flares. Thus the origin was made even more mysterious by B iermann's basic point that the Sun emitted corpuscular radiation in all directions at all times, regardless of the presence or absence of magnetic active regions. Now by 1956 John Simpson ([25], [13]) had succeeded in determining the energy spectrum of the variation of the cosmic ray intensity with the varying level of activity of the Sun. The variations were first detected by Scott Forbush, using ion chambers, which are sensitive to the muons produced in the atmosphere by cosmic ray protons with energies of 10-20 Gev and up. Simpson invented the cosmic ray neutron monitor which responds to the nucleonic component in the atmosphere, thereby registering the effect of cosmic ray protons down to about 1 Gev, where the time variations are much larger. Using five neutron monitors distributed from the geomagnetic equator to Chicago (at 55 ~ geomagnetic latitude) he exploited the geomagnetic field of Earth as a magnetic spectrometer. He showed that the variations had an energy spectrum that could not be a consequence of an electrostatic potential difference in space, which would be presumed to decrease the energy of each particle by the same amount. Instead, the variations, apart from the bursts of solar cosmic rays from the occasional large flare, showed simply a removal of particles that increased with declining cosmic ray particle energy. He noted that the variations suggested time varying magnetic fields in space. The great cosmic ray flare of 23 February 1956 showed direct passage of the solar cosmic rays from their origin on the Sun to Earth, arriving promptly at Earth from the direction of the Sun ([12]). Thereafter the solar cosmic ray intensity was observed to decline slowly as if escaping by diffusing through a magnetic barrier beginning at about the orbit of Mars and extending outward to the orbit of Jupiter. The simplest model suggested by the observations was a radial magnetic field extending from the Sun out to the orbit of Mars, with a disordered nonradial magnetic field beyond. Collectively this indicated a dynamical state of the solar corpuscular radiation and magnetic field in interplanetary space. The challenge, then, was to understand how the corpuscular radiation and interplanetary magnetic field were created by the Sun.

2. The solar wind concept The development of the ideas that eventually led to understanding the origin of the solar corpuscular radiation, or solar wind, got under way in the thirties, with the astonishing million degree temperature of the outer atmosphere of the Sun, i.e. the solar corona, firmly established by about 1942, thanks to the combined efforts of Grotrian, Edlen, and Lyot (cf. Billings [3]). Then in 1957 came Chapman's ([5]) recognition of the extension (through thermal conductivity) of the million degree coronal temperatures far out into space. He showed from the equation for barometric equilibrium that the corona extends past the orbit of Earth, with a density of perhaps 10 3 protons and electrons per cm a at the orbit of Earth. That is to say, Earth orbits within the corona of the Sun. In fact this created a dilemma, because both the corona and the solar corpuscular radiation, each composed of equal numbers of electrons and ions, represent plasmas. The two-stream plasma instability was known by that time, and it was clear that two collisionless plasmas can not stream through each other. The rapid growth of the instability would lock them together. Yet the existence of both plasmas and the large relative velocity were established beyond reasonable doubt.

-483 -

E.N. Parker One could understand the situation if, and only if, the strongly bound static corona near the Sun somehow became the solar corpuscular radiation at large distance from the Sun, so that there is only a single plasma rather than two interpenetrating plasmas. The first step was to show ([ 18]) that, with the extended temperature (T ~ 1/r 2/7) there can be no static equilibrium if the inward pressure of the interstellar magnetic field, wind and cosmic rays is no more than the estimated 0.5 • 10 -12 dynes/cm 3. Then it had to be realized that hydrodynamics was the appropriate treatment for the dynamical state of the nonstatic corona ([17]). Integration of the hydrodynamic equation for the simple case of radial outflow showed that the hydrodynamic expansion of the extended hot corona into the vacuum of interstellar space automatically creates the supersonic solar wind, thereby forming the heliosphere ([18]). In the final analysis the heliosphere is to be understood as the direct product of coronal heating ([20]). Once it was realized that the solar corpuscular radiation is to be identified with the supersonic hydrodynamic expansion of the outer corona of the Sun, the general structure of the heliosphere was apparent with, or without, an interstellar wind beyond 100 AU. Interplanetary space is filled with the magnetic fields stretched out by the expanding corona. In the ideal case of a uniform solar wind velocity v the magnetic field lies along the Archimedean spiral r - a = (v/aQ)(qb - ~b0) for the equatorial field line originating at the Sun (r = a) at azimuth ~b = ~b0 where Q is the angular velocity of the Sun ([ 18]). Thus the field is nearly radial inside the orbit of Earth, declining as 1/r 2 and reaching a 45 ~ inclination to the radial direction at the orbit of Mars. Beyond Mars the field becomes principally azimuthal and declines as 1/r. The outward sweep of the magnetic field tends to convect the galactic cosmic rays out of the inner solar system with varying degrees of vigor, depending particularly on the small-scale irregularities in the magnetic field ([ 19]). Thus the reduction of the cosmic ray intensity at the orbit of Earth is variable and strongest when the Sun is most active. Blast waves from explosions at the Sun (coronal mass ejections and flares) represent outward sweeping belts of concentrated magnetic field, providing transient reductions (Forbush decreases) in the cosmic ray intensity ([21 ]). The nonuniformity of the corona automatically produces the slow and fast solar wind streams, with the same Archimedean spiral form r = (v/a~)c~ as the field lines. The collision of fast streams with the rear (concave) side of the slow streams provides the co-rotating interacting regions in the solar wind, with their forward and backward propagating shocks, particle acceleration, etc ([23]; [4]). The hydrodynamics went on to predict the general form of the heliosphere, illustrated in Figure 2 for a static exterior interstellar gas and magnetic field, and in Figure 3 in the presence of an interstellar wind [22], [23]). The termination shock in the supersonic wind is shown in Figure 2 for various stagnation pressures in the solar wind. This was all theoretical before the advent of the space age, of course, and there was a widespread disbelief of the idea that the hydrodynamic expansion of the corona would reach supersonic velocities to become the solar corpuscular radiation. Fortunately the space age soon arrived, and the solar wind was first detected in situ by Gringauz in 1960 ([8]). The speed and density were first measured unambiguously by Snyder and Neugebauer with an instrument on the Mariner II spacecraft outward bound to Venus in 1961 ([26], [15], [16]). The density proved to be of the general order of 5 protons and electrons/cm 3, which was about a factor of 102 less than the indirect inference of 500/cm 3 already mentioned. It followed that the zodiacal light represents sunlight scattered from interplanetary dust grains rather than free electrons. It was also realized that the interaction of the solar corpuscular radiation or wind with the ions in the comet tail is primarily through the magnetic fields carried in the wind rather than by particle

-484-

The heliosphere as viewed from Earth

interactions. The first measurements of the spiral interplanetary magnetic field were accomplished by Heppner et al. ([9], [14]), the difficulty with earlier space measurements being the magnetic contamination by the

Figure 2. Terminal shock location and outer boundary of the heliosphere confined by a uniform interstellar magnetic field, for the indicated values of the stagnation pressure YI (from Parker [22], [23]).

spacecraft. They found an average strength of about 60 microgauss at 1 AU inclined to the radial direction by about 40 ~, as expected. It should also be remarked that they found the field to fluctuate wildly about the mean direction, and the origin of the fluctuations is not properly understood to the present day.

3. Structure of the heliosphere The expanding corona, or solar wind, extends far beyond the planets, pushing back the surrounding tenuous interstellar gas and magnetic field to distances in excess of 100 AU. That radius is determined simply by equating the ram pressure pv 2 of the solar wind to the estimated impact pressure of the interstellar wind against the heliosphere. It is the large dimensions of the heliosphere that make its exploration so difficult. It has taken the two Voyager spacecraft two decades (at about 4 AU/year) to reach their present positions, with Voyagers I and II in the

-485 -

E.N. Parker

Figure 3. The streamlines in a heliosphere confined by a subsonic solar wind (from Parker [22], [23]). The termination shock was omitted from this drawing.

vicinity of 80 AU and 60 AU, respectively. It is reasonable to expect that Voyager I will reach the termination shock sometime in the next decade, thereby showing the termination shock at the outer bound of the supersonic solar wind where the wind drops to subsonic velocity and the temperature of the shocked gas is of the general order of 106 - 107 K. It will be a landmark in space science, for up to the present time we have only theoretical estimates to guide our thinking. The question is how long the spacecraft will remain in good health and to what distance communication can be maintained. It must be appreciated that at the present position of Voyager I, at a distance of 80 AU, the Sun is a very bright but distant star, with sunlight only 1/6400 as bright as we see it here at Earth. That is about 70 times brighter than moonlight, and enough to read a newspaper, but only dimly. With an intensity of about a sixth of a watt/m 2, compared to a kilowatt/m 2 at Earth, there is not enough sunlight to power solar cells. So the Voyager spacecraft carry their own operating fuel in the form of radioactive thermal generators, using the heat from the decay of radioactive nuclei to operate a thermopile. I never cease to marvel that the collecting power of a large radio dish and modem electronic amplification here at Earth are able to pick up the signal of only a few watts from the relatively small antennas pointed our way from the spacecraft. A narrow bandwidth and the associated low bit rate are part of the game, of course, along with phase locking to pull the signal out of the noise. Needless to say, the radioactive heat sources are gradually cooling and the available power slowly declining.

-486-

The heliosphere as viewed from Earth

One might even go so far as to hope that the spacecraft will still be alive and talking to us when crossing the heliopause and passing out of the shocked solar wind into the interstellar plasma beyond (see discussion in Wang and Belcher [29] and Zank [30]). Now from our end of things at Earth, the heliosphere seems like a vast region of space. The characteristic diameter of 200 AU is one light day. The signals received from the spacecraft have been travelling for about twelve hours when they are picked up at Earth. On the other hand, if we think in terms of the stars and the Galaxy, the heliosphere is insignificantly small. The distances to the nearest stars are 4 and 5 light years, or about 1600 times the diameter of the heliosphere. If we were to make a large drawing of the relative positions of the nearby stars and the Sun, with lm between stars, the heliosphere would have a diameter of the order of 0.6 mm - a small dot visible to the naked eye. When we recall the nearly three decades of travelling for the Voyager spacecraft to reach 100 AU and note that it will then be only 1/3200 of the distance to the nearest stars, the immensity of interstellar space is readily apparent. At that rate it is nearly 105 years to the nearest star. There is no physical process known to contemporary science that can project a communicating device to arrive at a nearby star in a human lifetime, or even within the lifetime of a coordinated national political entity, generally not in excess of a few centuries. Interstellar travel times, as presently available, are more nearly comparable to the age of homo sapiens sapiens. The supersonic coronal expansion and solar wind stretches out some of the weaker magnetic fields of the Sun, filling the heliosphere to its farthest reaches with solar magnetic field. The rotation of the Sun continually winds the magnetic field as the field is carried outward in the solar wind. The solar wind travels the Sun-Earth distance of 1 AU (1.5 x 10 ~a cm) in about 4 days, so the journey to 100 AU requires 400 days, during which time the low latitude regions of the Sun rotate through 16 revolutions. Thus the magnetic field spirals 16 times around the heliosphere between the Sun and 100 AU. Beyond the orbit of Mars the field becomes nearly azimuthal, declining more or less in proportion to 1/r to a value of the order 0.4 microgauss at 100 AU. Note, then, that with a mean density of about 5 ions/cm a at the orbit of Earth, the wind density at 100 AU is 0.5 x 10-a/cm a. This is to be compared with an estimated ambient interstellar magnetic field of 3 x 10 -6 Gauss, and a gas density of the order of 0.1 ions/cm a, and 0.2 neutral atoms/cm a.

4. The heliosphere over time

It is amusing to contemplate that in its youth (t < 108 years) the Sun evidently rotated more rapidly, with a period of only a few days. Thus, for instance, when the period of rotation was 4 days, the same wind velocity as today would cause the magnetic field to circle about 100 times around the heliosphere out to 100 AU, becoming strongly azimuthal even before reaching the planet Mercury. On the other hand, a higher wind velocity would reduce the degree of spiraling to some degree, and we have no idea what those early wind velocities might have been. The loss of angular momentum from the Sun indicates only a massive wind blowing out through a stiff magnetic field, so that the departing gas carries away a lot of angular momentum per unit mass. Now Earth resides very close to the center of the heliosphere where the solar wind is complicated by the collisions of fast and slow streams of wind, coronal mass ejections, interplanetary particle acceleration (in shock waves) and the sometimes lethal bursts of fast particles from

-487-

E.N. Parker

flares. The magnetospheric "roof" protecting Earth from the wind is continually rattled by the turbulence, blast waves, and rapid magnetic reconnection with the field in the passing solar wind. What is more, cosmic ray conditions here at Earth depend to large degree on activities in the outer heliosphere (~ 100 AU), where the state of the plasma and magnetic field is influenced by conditions in the interstellar wind and galactic magnetic field impinging on the heliosphere from outside. So we still do not have a clear picture of the whole cosmic ray modulation process. In this connection it is interesting to consider past variations of the heliosphere. The essential point is that the heliosphere extends out to where the solar wind becomes so tenuous that the interstellar gas and magnetic field can block it somewhere in the vicinity of 100 AU in the upstream direction. In fact, the variation in the ram pressure of the solar wind at low heliographic latitudes varies by about a factor of two through the 11-year magnetic activity cycle, with the minimum pressure at solar maximum and maximum pressure a year or two later ([ 10]). The result is some complicated and unstable breathing in and out at the termination shock and heliopause ([30]). The radial displacement of the termination shock is estimated to be 10-20 AU. So the outer heliosphere is an active place, with dynamical complications and a disordered magnetic field so far known only through theoretical considerations (cf. [31 ]). The passage of a single spacecraft, e.g. Voyager I, through the region will be a good start, but there is clearly a lot of activity out there waiting to be discovered, and the first passage will produce more mysteries than answers, we can be sure. The projected Interstellar Probe Mission will have its work cut out for it, and hopefully Voyager I will be able to define the problems to aid in planning and designing the scientific instruments for the Interstellar Probe. If this is the present state of affairs, consider, then, the past variations in the interstellar environment of the heliosphere over the approximately 20 orbits around the Galaxy since the Sun and Earth were formed. First of all, there is reason to believe that the solar wind from the young Sun was much denser than at present, as already noted, with the strong solar magnetic field and dense wind carrying away most of the initial angular momentum. We would expect that the dense wind of that early time pushed the outer boundary of the heliosphere much farther away into interstellar space than the present wind for any given interstellar condition. Further, we may reasonably expect that dense interstellar gas clouds (say 100 atoms/cma), have been occasionally encountered by the Sun over the last 4.6 x 109 years, and for those brief moments the interstellar wind might have pushed the boundary of the heliosphere in as far as the orbit of Jupiter (5 AU), or closer (cf. [32]). In such case the dynamical coupling of the interstellar gas to the solar wind would be through resonant charge exchange. There is no way of knowing how the cosmic ray intensity was affected at Earth, since we do not have a proper quantitative understanding of the total cosmic ray reduction today. At the most primitive level we would expect that the proximity of the interstellar magnetic field and cosmic rays means an increase in cosmic rays at the orbit of Earth (cf. [32]). Presumably the anomalous cosmic ray intensity would be enhanced approximately in proportion to the increase of the neutral density in interstellar space above the present value of 0.2/cm 3, multiplied by the geometrical factor of the reduced cross section of the compressed heliosphere, because the anomalous cosmic rays are the end product of infalling neutral interstellar atoms. The anomalous cosmic rays would represent an energy density and pressure substantially in excess of the galactic cosmic rays at the termination shock. In that connection, it is interesting to note that recent studies by Linsky et al. ([11], [24]; see paper by Redfield and Linsky, these Proceedings) of the nearby interstellar gas indicate that the present warm (7000-8000 K) partially ionized (about 0.1 hydrogen ions/cm 3 compared

-488

-

The heliosphere as viewedfrom Earth

to 0.2 neutral atoms/cm 3) local interstellar gas is limited to a cloud of some 7 parsec extent. The cloud is drifting past the heliosphere at 26 km/s, suggesting that the heliosphere has been engulfed in this cloud for the last 12 x 105 years. They find that the heliosphere is now very close to the rearward end of the cloud, and in another 3000 years or so it will leave us, presumably exposing the heliosphere to the very hot (106K) tenuous interstellar gas component of very low density. One expects that the hydrostatic pressure of the hot component is about the same as the pressure in the present cloud, suggesting a density of the general order of 2 x 10 -3 ions/cm 3 and essentially no neutral atoms whatever. What the relative motion of the hot component will be is not known, but the impact against the heliosphere will be greatly reduced because of the very low density. This suggests that the heliosphere will increase its size, particularly in the direction toward the impacting interstellar wind. The principal confining force will be the galactic magnetic field, of perhaps 3 microgauss, with some help from the pressure of the hot ionized component, for a total of the order of 0.5 x 10 -~2 dynes/cm 2, compared to the present impact pressure of about 2 x 10 -~2 dynes/cm 2. With these estimates, the termination shock would move out to somewhere in the vicinity of 200 AU. The effect on the cosmic ray reduction at Earth is not known because of present ignorance of the effects in the distant heliosphere now and in 3000 years. The anomalous cosmic rays can be expected to fall to negligible levels for lack of infalling neutral interstellar atoms. The few anomalous cosmic rays to be found would arise from the small numbers of neutral atoms from the exposed surfaces of the moon, asteroids, interplanetary dust, the atmosphere of Venus, comets, etc. It is not incorrect to say that the heliosphere is notable for the interpenetrating and interacting particle populations, with resonant charge exchange the principal interaction between the neutral atoms and the solar wind plasma. The result is a variety of pick up ions in the solar wind and the conversion of those ions into anomalous cosmic rays at the termination shock. In fact it has been shown that there is substantial charge exchange in the shocked interstellar wind upstream from the heliopause, so that the infalling interstellar neutrals are "filtered" before entering, with their velocity and density distributions significantly modified ([7], [31 ], [32]). A final remark concerns the point made some years ago that the heliosphere is an excellent garbage disposal machine. If one were to vaporize garbage and pitch it out into the solar wind (at a prohibitive cost per gm), the garbage molecules would soon be ionized through charge exchange with the solar wind. Picked up by the solar wind, the garbage ions would be transported out to the termination shock in a little more than a year. From there into interstellar space and gone forever. However, upon reflection, it is evident that the story is not so simple, because a large fraction of the garbage ions would be turned into anomalous cosmic rays at the termination shock, and we would find our garbage back at 1 AU at 10 Mev per nucleon. We would leave a tainted wake in the interstellar wind, I suppose, although any amount of garbage that we can hurl out into the solar wind from Earth would represent quite a negligible effect when spread out over 100 AU. Now the solar wind plasma direct from coronal expansion exhibits ion temperatures in approximate proportion to the ionic mass, with slightly higher wind velocities for the heavier ions compared to the wind velocity of the protons. We might think of it as several superimposed interpenetrating solar winds, each wind for a different ionic species. The maintenance of the transverse ion temperatures in the transversely expanding solar wind indicates some strong form of ion heating, probably by resonant scattering from waves in the plasma to account for the higher heavy ion temperatures.

-489-

E.N. Parker

The galactic cosmic rays represent the most strongly interpenetrating component, arising from the relativistic temperature of the cosmic ray gas, i.e. the large cyclotron radius and relativistic speed of the individual particles. As already noted, the variable reduction of the intensity of the galactic cosmic rays penetrating into the inner heliosphere can be fully understood only when Voyager I, or other spacecraft, penetrates out of the heliosphere.

5. The distant heliosphere Recent studies by Svensmark and Friis-Christensen ([28], [27]), suggesting that terrestrial cloud cover correlates more closely with the local cosmic ray intensity than with any other index of solar activity, have led them to speculate that there may be a cosmic ray effect on terrestrial climate. If indeed there is a physical connection, it would be through the nucleation of aerosols and ice crystals by the ions created by the passage of cosmic rays through the terrestrial atmosphere. The essential point is simply that, if it turns out that there is something to this speculation, then the dynamical state of the outer heliosphere, with detectable cosmic ray effects at Earth, has direct effect on the climate of Earth. Finally, we should not fail to mention that the subsonic flow of shocked solar wind gas beyond the termination shock in the solar wind trails off downstream in the interstellar wind. The interstellar wind impacting the heliosphere is also presumed to be shocked, and the net result is a complex downstream interstellar wake left by the passage of the heliosphere through the interstellar medium. The wake contains the double magnetic spirals composed of the northern and southern hemispheres of the heliospheric magnetic field. The extremely hot shocked solar wind ions are quenched by resonant charge exchange with the interstellar neutral atoms, so that the thermal structure of the wake soon degenerates to ions with thermal velocities of the order of the 26 km/s interstellar wind velocity, having lost the original fast solar wind ions as fast neutral atoms (~300 km/sec or more) following the charge exchange. Then the escaping high speed neutral atoms charge exchange with interstellar ions, and the process eventually degenerates. To get some idea of the scale of the wake, note that the heliosphere represents a mass loss of about 106 tons/s, or some 0.6 x 10 a6 protons/s. The interstellar wind impacting the 100 AU radius of the heliosphere involves approximately 3 x 10 36 neutral hydrogen atoms/s. So there are approximately five neutral atoms for each solar wind ion. The interstellar magnetic field is evidently of the order of 3 x 10 -6 Gauss, while the magnetic field in the solar wind is perhaps 4 x 10 - 7 Gauss, a tenth as strong, compressed then to as much as 1 - 2 x 10 -6 Gauss in passing through the termination shock. So it appears that the interstellar wind and magnetic field dominate the scene. The characteristic charge exchange time for a solar wind ion in the interstellar neutral atom number density N is of the order of 107/N s. Thus for N = 0.2/cm a the time is 5 x 107s or a little less than 2 years. During this time the 26 km/s interstellar wind velocity carries the gas only about 9 AU downstream, which is small compared to the 100AU characteristic transverse scale of the wake. It would appear, then, that the downstream wake immediately relaxes and broadens into the surrounding interstellar gas, rapidly approaching a bland asymptotic state in which the original solar wind particles are spread out and soon lost in the vast reaches of interstellar gas. It is difficult to say how far downstream in the wake the magnetic field of the heliosphere might be discernible. However it would appear that the passage of the heliosphere through interstellar space probably leaves no significant spoor.

-490-

The heliosphere as viewed from Earth

REFERENCES

1. 2. 3. 4.

L. B iermann, Zeit. f. Astrophys. 29 (1951) 274. L. Biermann, Observatory 77 (1957) 103. D.E. Billings, A Guide to the Solar Corona, Academic Press, New York. L.E Burlaga, Interplanetary Magnetohydrodynamics, Oxford University Press, New York, 1997. 5. S. Chapman, Proc. Roy. Soc. London A 253 (1959) 462. 6. L. Davis, Phys. Rev. 100 (1955) 1440. 7. H.J. Fahr, Space Sci. Rev. 78 (1996) 199. 8. K.I. Gringauz, V.V. Bezmkikh, V.D. Ozerov, and R.E. Rybchinsky, Space Res. 2 (1961) 539. 9. J.E Heppner, N.E Ness, C.S. Scearce, and T.L. Skilman, J. Geophys. Res. 68 (1963) 1. 10. A.J. Lazarus and R.L. McNutt, Physics of the Outer Heliosphere, S. Grzedzielski and D.E. Page (eds.), Pergamon Press, New York, (1990) 229. 11. J. Linsky, S. Redfield, B. Wood, and N. Piskunov, Astrophys. J. 528 (2000) 756. 12. E Meyer, E.N. Parker, and J.A. Simpson, Phys. Rev. 104 (1956) 768. 13. E Meyer and J.A. Simpson, Phys. Rev. 99 (1955) 1517. 14. N.E Ness, C.S. Scearce, and J.B. Seek, J. Geophys. Res. 69 (1964) 3531. 15. M. Neugebauer and C.W. Snyder, J. Geophys. Res. 71 (1966) 4469. 16. M. Neugebauer and C.W. Snyder, J. Geophys. Res. 72 (1967) 1823. 17. E.N. Parker, Phys. Rev. 107 (1957) 924. 18. E.N. Parker, Astrophys. J. 128 (1958a) 644. 19. E.N. Parker, Phys. Rev. 110 (1958b) 1445. 20. E.N. Parker, Astrophys. J. 132 (1960) 821. 21. E.N. Parker, Astrophys. J. 133 (1961a) 1014. 22. E.N. Parker, Astrophys. J. 134 (1961b) 20. 23. E.N. Parker, Interplanetary Dynamical Processes, Interscience Div. John Wiley and Sons, New York, ( 1963). 24. B. Schwarzschild, Physics Today (January) (2000) 17. 25. J.A. Simpson, Phys. Rev. 94 (1954) 426. 26. C.W. Snyder and M. Neugebauer, Space Res. 4 (1964) 89. 27. H. Svensmark, Phys. Rev. Lett. 81 (1998) 5027. 28. H. Svensmark and E. Friis-Christensen, J. Atmos. Solar Terrest. Phys. 59 (1997) 1225. 29. C. Wang and J. W. Belcher, Ninth Intern. Solar Wind Conf., AIP Conf. Proc. 471, American Institute of Physics, Woodbury, New York (1999). 30. G.E Zank, Ninth Intern. Solar Wind Conf., AIP Conf. Proc. 471, American Institute of Physics, Woodbury, New York, (1999a) 783. 31. G.E Zank, Space Sci. Rev. 89 (1999b) 413. 32. G.E Zank and EC. Frisch, Ninth Intern. Solar Wind Conf., AIP Conf. Proc. 471, American Institute of Physics, Woodbury, New York, (1999) 831. 33. G.E Zank, A.S. Lipatov, and H. Mtiller, Ninth Intern. Solar Wind Conf., AIP Conf. Proc. 471, American Institute of Physics, Woodbury, New York, (1999) 811.

-491 -

This Page Intentionally Left Blank

The Heliosphere, Cosmic Rays, Climate Klaus Scherer ~ Horst Fichtner, Olaf Stawicki b ~dat-hex, 37191 Katlenburg-Lindau bInstitut fiir Theoretische Physik IV: Weltraum- und Astrophysik, Ruhr-Universits Bochum, 44780 Bochum The heliospheric shield protects the inner heliosphere from the direct contact with the interstellar medium. Changes in the interstellar medium during the Keplerian evolution of the Sun around the galactic center cause variations in the heliospheric distance of the shield. This effects the environment of the planets, especially that of the Earth. A moderate increase in the interstellar density causes a strong rise in the flux of the cosmic rays at Earth orbit. Possible effects on the evolution of life on Earth are discussed. 1. I n t r o d u c t i o n

2. T h e variable h e l i o s p h e r e In the upper panel of Fig.1 the heliosphere is displayed as seen in the rest frame of the Sun for the present-day ISM with a temperature of T ~ 8000 K, an ISM inflow velocity of v - 2 5 k m / s , a proton number density of np = 0.1cm - 3 and neutral gas density

-493 -

K. Scherer, H. Fichtner and O. Stawicki

of n H = 0.1 cm -a , and a fictitious ISM with ten times higher number densities (lower panel). The difference between the two cases is significant: for the higher ISM density the heliosphere shrinks by more than a factor of two, with a HTS location at about only 35 A U in the upwind direction. It is a complicated task to predict the CR fluxes in a

Figure 1. Comparison of the heliospheric shield for two different interstellar media (see text). In the upper panel the present-day situation is modelled, while the lower panel shows a hypothetical shield for a factor 10 interstellar gas density. Both cases are shown on the same scale. such 'modified' heliosphere. While the shock is further in and, thus, the modulation region is smaller as compared to the current situation, the HTS, representing the source location for ACRs and the main modulation barrier for GCRs, might be strongly modified. Furthermore, the solar wind flow and heliospheric magnetic field structure inside the shock surface might be significantly changed and, thus, the modulation of the ACR and GCR fluxes might be very different from that observed nowadays [3,4]. Nonetheless, in order to get a basic idea of how the CR fluxes at 1 A U would change due to a smaller modulation region, we neglect principal changes in the particle transport and concentrate on the effects of (i) an HTS closer to the Sun and (ii) a lower solar wind speed.

The heliosphere, cosmic rays, climate

and 2 - 3 times in the interval 0.1-1 GeV. For the lower solar wind speed of 200 kms -1, the flux levels are further increased by factor of two above 0.1 GeV and up to 1000 below 0.1 GeV. Therefore, combining both effects the CR flux in the energy range 0.1-1 GeV increases at Earth by a factor of about six. Primarily, this energy interval affects the Earth environment, because these CRs penetrate through the magnetosphere and reach the atmosphere or even ground level. It was recently demonstrated [1] that the cloud coverage of Earth is correlated with the high energy CR flux variations due to solar activity and that mainly the CR with energies between 0.1 to 1 GeV influence the lower atmosphere. Consequently, one also should expect the much stronger CR flux variations due to a changing interstellar environment of the heliosphere to affect the climate on Earth. In view of the high flux increases and the much longer time-scale on which they appear (from tens to millions of years) the corresponding cloud coyerage of Earth could have dramatic conse- Figure 2: The productivity indices of quences, mankind through the Maunder minimum, Even, when the Sun is quiet for a long from [6] time, as it happened during the Maunder Minimum, one observes a higher production of cosmogenic elements, i.e. elements produced in the Earth atmosphere by cosmic ray bombardment [7]. The human behavior is also affected through that times as can be seen in Fig. 2, where the productivity index for painting, poetry, and science in the western and chinese culture is plotted during the Maunder Minimum. The depression in all three aspects during the quiet time of the Sun is well observed. The chain of reactions may not be well understood, but other adaptions of mankind to the variation of the solar cycle may be found. Since at present conditions measurements of the radiation budget of Earth revealed that clouds reflect more energy into outer space than they absorb [1], an increased cloud coverage results in a net cooling effect. While these processes, in particular the exact chain of physical processes providing the correlation between CR flux and cloud coverage are not well understood, further evidence for the cooling effect comes from cosmogenic nuclei, which are produced by the bombardement of CR onto the atmosphere. Thus a higher CR flux produces more cosmogenic nuclei, than a smaller flux, which can be observed, for example, in the A C 14 rates during the solar cycle (A C 14 is produced by thermal neutron capture).

-495 -

K. Scherer, H. Fichmer and O. Stawicki

3. S u m m a r y a n d discussion We have presented evidence of the influence of the galatic environment on the climate on Earth. A rise in the galactic cosmic ray flux at Earth can trigger cold ages, like that of the Maunder minimum, or ice ages. Previous attemps (see discussion in [3]) connecting ice ages or other climatic catastrophes discussed a very high number density of the interstellar medium (hundred to thousands of particles per cubic-centimeter), so that either the termination shock was pushed inside the Earth orbit or that enough dust was available falling on Earth blocking the Sun's light. In contrast, in our scenario only small variations of the interstellar number density lead to a remarkable increase in the CR fluxes influencing Figure 3: The dots mark cometary imthe climate on Earth. pacts correlated with mass extinctions. The most important changes of the local inThe circle is a good approximation to terstellar medium are caused by crossings of the heliosphere through galactic spiral arms (see the solar path. While the arcs mark Fig. 3). Thus, not only hazardous cometary im- 'quiet' periods, in which due to modpacts on Earth causing mass extinctions, but erate changes in the ISM, as shown in also CRs via their imprints on cosmogenic nu- Fig. 1, the planetary environment was clei in ice cores or sediments are wittnesses of the affected, and, in turn, the life on Earth. solar path during the last few millions of years. Moreover, a comet, which originates in the Oort Cloud at about 20000 AU and is perturbed by a changing gravitational potential e.g. due to a spiral arm crossing, needs about 1 to 1.5 million years to reach the Earth. For that time-scale changes in the interstellar medium are nearly instantaneously communicated to Earth e.g. via heliospheric transport of CRs. Thus, a cometary impact marks only a late singular catastrophic event in a long-term planetary environmental change induced by a weakened heliospheric shield. REFERENCES 1. Svensmark, H., Friis-Christensen, E., J. Atmos. Terr. Phys., 59, (1997), 1225-1232 2. Pudovkin, M.I., Verentenenko, S.V., J. Atmos. Terr. Phys., 57, (1995) 1349-1355 3. $cherer, K., in: The Outer Heliosphere: Beyond the Planets, Eds: K. Scherer, H. Fichtner, E. Marsch, Copernicus Gesellschaft, (2000) 327-356 4. Scherer, K., Fichtner, H., Stawicki, O., J. Atmos. Sol. Terr. Phys., accepted 5. O. Stawicki, O., Fichtner, H., Schlickeiser, R., Astron. Astrophys. 358, (2000) 347 6. Ertel, S., Bursts of creativity and aberrant sunspot cycles, in: The scientific study of human nature: Tribute to Hans J. Eysenck at Eighty, Ed.: H. Nyborg, (1997) 491-510 7. Beer, J., Long-term indirect indices of solar variability, Space Sci. Rev. 94, (2000) 53-66

-496-

Oral papers and posters

HELIOSPHERIC CHANGES IN THE PAST EVIDENCE FROM COSMOGENIC ISOTOPES IN POLAR ICE J. Beer EAWAG, CH-8600 Duebendorf. Cosmogenic isotopes such as 10Be are produced continuously in the atmosphere by the interaction of cosmic ray particles with nitrogen and oxygen. The energy spectrum of the cosmic ray particles is modified when travelling through the heliosphere. The magnetic field carried by the solar wind influences mainly the low energy particles. As a consequence the cosmic ray flux and therefore also the production rate of cosmogenic isotopes depends on the solar activity. The higher the activity the lower is the production rate. 1~ is removed from the atmosphere after a mean residence of 1-2 years mainly by precipitation. Polar ice sheets that preserve the precipitation in annual layers provide a natural archive containing information about heliospheric changes over many millennia. We will present results obtained so far from ice cores in Greenland and discuss the potential and the limitations of this approach.

-497-

Oral papers and posters

T H E U V R A D I A T I O N C L I M A T E O N E A R T H A N D ITS I M PACT ON THE BIOSPHERE: PAST, PRESENT AND FUTURE TRENDS Gerda Horneck German Aerospace Center DLR, Institute of Aerospace Medicine, Linder HShe, 51147 KSln, Germany.

Solar UV radiation, on the one hand, is a dynamic driving force of organic chemical evolution, on the other hand it may set severe constraints in biological evolution, especially if the biologically harmful UV-C radiation reaches the surface of a planet. During the history of life on Earth, the UV-radiation climate has dramatically changed: The atmosphere of the early Earth was virtually devoid of oxygen and hence ozone, which means that solar UV-C (200 -280 nm) and UV-B (280 - 315 nm) radiation could reach its surface and hence the early biosphere virtually unattenuated during the first 1.5 - 2 Ga of life's existence. This short wavelength range of solar UV radiation is a potent mutagen and a selective agent since it is effectively absorbed by the genetic material, the DNA. Radiative transfer models and biological experiments in space have demonstrated that the biologically effective irradiance at the surface of the late Archean Earth was about 3 orders of magnitude higher than it is experienced today. Action spectra for inactivation, DNA damage and mutation induction, using polychromatic UV radiation from solar simulators confirm this value. Hence, during its early evolution life on Earth had to cope with an intense UV radiation climate of high mutagenic potential. In response to this genetic and physiological stress, the primitive life forms were forced either to withdraw into UV refuges or to develop strategies to tolerate the UV radiation influx, such as the development of internal and external UV screens or of cellular mechanisms to repair the UV-induced damage. The situation changed with the advent of oxygenic photosynthesis whereby oxygen was enriched in the atmosphere. At present, the stratospheric ozone layer serves as a cut-off filter protecting the surface of the Earth from the detrimental short-wave solar UV of A < 290 nm. However, even today, life has invented several protection mechanisms to cope with the UV radiation climate. In view of the current seasonal ozone depletion, the consequences for the biosphere of a further depletion have been estimated: an extreme erosion of the ozone column, (e.g., by about 80%) would result in a 100 fold higher damage to the DNA. Likewise, the photosynthetic productivity would be reduced which could finally reduce the oxygen concentration in the atmosphere. This might result in a negative feed back mechanism with less ozone produced in the atmosphere. However, this scenario is rather unlikely because it would require much higher amounts of industrial outgassing than is currently envisaged for the near future.

-498-

General

Discussion

Bertaux to Parker: You mentioned the problem of finite pressure a.t infinity. Chamberlain showed that if you have an exospheric model that problem disappears. I remember that you had a conflict with Chamberlain about the solar wind vs. solar breeze. Could you comment on that? Parker: When you allow for outward expansion the only solutions wtlich fit the boundary conditions are the supersonic solutions. The exospheric treatment of the corona is very interesting and you can duplicate hydrodynamic results. But the observations of protons and electron temperatures at the orbit of Earth and their near-isotropy inspite of the anisotropic expansion indicates that the particles are somehow continuously scattered, probably by small-scale waves. That makes the exospheric approach very difficult. Fahr to Parker: How would you imagine the ignition of the solar wind? How do you see the system of differential equations to find the right solution? Parker: I've never looked at the time-dependent equations. However, if you allow the tcmpcraturc to incrcasc sufficicntly slowly, thcn I would assumc that wc stay fairly closc to the steady states. And, I guess, you wouldn't get any outflow until the temperature is up to about half a million degrees. Vondrak to Horneck: You were saying that eucaryotes were developing after an ozone layer. Does that mean that if Mars, for example, lacks an ozone layer, you would not expect to find eucaryotes? Horneck: The first known eucaryotes on Earth are from about 1.5 billion years a g o - it's not exactly but more or less the time when we had the protection by the ozone layer. Mewaldt to Beer: Is there any evidence at all for a nearby supernova or any increase [in cosmogenic isotopes] that you can't explain using geomagnetic or solar activity? Beer: I would say for the last 100000 years there is no indication. Fahr to Beer: Is there any idea around what could be the driver of this Gleisberg cycle of about 90 years? Beer: I don't think there is any clear idea about this. Lallement to Beer: I was told the Vostok data should allow to go back 300000 years. Could you comment on that? Beer: Yes, even more. Right now, it's 440000 years. The problem there is that the accumulation rate is very s m a l l - about 2 cm of ice per year. And, therefore, you don't have a very high time resolution. Vondrak to Scherer: I guess the real difficulty is finding the linkages between some of the events that you showed and the heliosphere. Scherer: Indeed. I think, we have to have interdisciplinary research to do to link all this together. Gazis to Scherer: There might be a connection between climate and cosmic rays, as was suggested by Svensmark and Friis-Christensen. Could you comment on that?

- 499 -

General Discussion

Scherer: Yes, this relation is being discussed. While there is no detailed modelling going on yet, this connection seems rather plausible. So, a fascinating thing seems quite possible now, namely that mankind can be influenced by the interstellar environment.

-

5 0 0

-

Session 12: Miscellaneous

This Page Intentionally Left Blank

Auroral Vortex Structures as a Result of Disturbed Geomagnetic Conditions M.A. Danielides a and A. Kozlovsky b Department of Physical Sciences, P.O. Box 3000, FIN-90014 University of Oulu. b Sodankyl/i Geophysical Observatory, FIN-99600 Sodankyl/i, Finland. ABSTRACT The aim of this study is to understand electrodynamics of the arc-like auroral structures observed in Alaska close to the local midnight during a substorm on February 11, 1997. The highly variable arc segments were moving and rotating. Rotation of the auroral structures monitored by all-sky camera has been compared with rotation of the ionospheric equivalent currents derived from ground magnetic observations. The comparison indicates that the arc segments were associated with the Hall current and corresponding plasma flow directed across the arc-like structures. The obtained features are discussed in terms of possible electrodynamical model for the auroras. 1. INTRODUCTION Since very beginning of the magnetosphere-ionosphere research, auroral displays and their relation to ionospheric electrodynamics have been investigated. As a basic kind of auroras, quiet auroral arcs were mostly discussed. These studies allow one to obtain the electrodynamical scheme of an auroral arc [2], and the relation has been investigated between motion of the arc and the large-scale ionospheric plasma convection [8]. However, disturbances in the magnetosphere-ionosphere system can result in a very complex auroral structure. Vortex structures of various spatial scales can be associated with quiet arcs [4]. Arc fragments during recovery of substorms also demonstrate complex behaviour and vortex features [7]. Rotation of auroral structures during disturbed geomagnetic conditions is under investigation in this study. 2. PHYSICAL BACKGROUND In a quiet auroral arc a disturbance may occur e.g. due to Kelvin-Helmholz instability associated with a convection shear along the arc. Also, a vortex of plasma flow can arise due to some kind of the Rayleigh-Taylor instability [6]. Figure 1 presents a sketch of wave-like distortions in the field-aligned currents and plasma flows connected to an auroral arc. Associated with the plasma flows ionospheric Hall current can be detected by ground-based magnetometers [5]. One may assume three different elementary electrodynamical mechanisms [1], which can be associated with an active arc-like structure. They are presented in Figure2" a) Upward field aligned currents (FAC) are usually observed inside a quiet auroral arc. Outside the arc one can expect return currents (downward FAC). Electric field and Pedersen currents are located between the up- and downward FAC. Parallel to the arc Hall current arises. This Hall current results in a ground magnetic effect, which can be expressed in terms of the total equivalent current (TEC) (Figure 2 a).

-503-

M.A. Danielides and A. Kozlovsky

b) The Cowling channel is associated with enhanced ionisation within an arc. Arc-aligned component of the large-scale background electric field causes a strong Hall current across the arc where Hall conductivity is enhanced. This leads to a polarization electric field oriented across the arc, and corresponding arc-associated Hall current along the arc. Figure 1. Wave-like distortion in the field- Ground magnetic effect in this case is similar aligned currents and plasma flows connected to to one shown in Figure 2 a, but this pattern is an auroralarc, not associated with field-aligned currents (Figure 2 b). e) Wave-like disturbances in arcs in Figure 1 are associated with alternating up- and downward orientated FAC. Corresponding electric field and plasma flow vortices result to Hall currents across the arc (Figure 2 c). Thus, the electrodynamical patterns in Figure 2a and 2b result in the ground- measured TEC directed along arc-like structures, whereas the pattern in Figure 3c is associated with TEC perpendicular to the arc. When the TEC is rotating, the differential equivalent current (DEC) is perpendicular to it. 3. DATA ANALYSIS TECHNIQUES Aiming to learn electrodynamics of auroras, we will compare rotation of the arc fragments observed all-sky camera with rotation of the ionospheric currents derived from ground Figure 2. Electrodynamical models for dis- magnetic observations. In our study we use turbed auroral forms. For more details see data of all-sky TV observations during Auroral text. Turbulence 2 rocket experiment, which occurred on February 11, 1997 in a vicinity of Fort Yukon (FYU, 66.6 N, 214.8 E), Alaska. The observation took place after a substorm onset. More detailed information on the experiment and geophysical background has been presented in [3]. During recovery of the substorm, we observed three cases when rotating arc segments were passing over Fort Yukon and Poker Flat (PKF). Figure 3 show such kinds of an arc segment in all-sky TV image (white). Such frames have been digitised at every 5-second. In every frame, the sharply defined lower edge of the arc was identified and the azimuth and elevation were measured for several points uniformly distributed along the edge. Assuming altitude of the lower luminosity edge at 110 km, we calculated geographic coordinates of the measured points. Using linear fit, we determined orientation of the arc segment with respect to latitude (positive angles relate to a clockwise arc rotation, i.e. it is located in the north-west to south-east sectors).

- 504-

Aurora vortex structures as a result o f disturbed geomagnetic conditions

Figure 4. Location of auroral vortex from Figure 3. Auroral vortex forms seen by all-sky Figure 3 and TEC sketched on a geographic camera (ASC) from Fort Yukon, Alaska on 1lth February 1997 at 08:40:35 UT. map of Alaska. The circle represents the field of view of the ASC.

Figure 5. Angular changes in optical auroral structures and DEC vectors (PKF,FYU) versus UT. Orientations of three observed arc segments versus time are presented in Figure 5. Below these three auroral structures are referred as I, II, and III. In addition to the optical observations ground-based magnetic observations are used in this analysis. By composing all optical and magnetic observations together on a geographic map (Figure 4), one receives an approximate overview of the momentary situation. The TEC vector lines shown in Figure 4 represent the ionospheric Hall current, and ground magnetic Z-component is marked as boxes scaled by the intensity. Temporal variations of the orientation of the TEC vectors, which is represented as differential equivalent currents (DEC), are shown in Figures 5 together with orientations of the optical structures.

-505-

M.A. Danielides and A. Kozlovsky 4. DISCUSSION

From Figure 5 one obtains that the angular changes in the optical auroral structures (arc segments) and the DEC vectors are related and orientated in the same sense of rotation, if the arcs were observed close to the points of the ground magnetic measurements. That indicates in favour of the electrodynamical pattern, which is associated with alternating up- and downward orientated FAC (Figure 2c). As it was mentioned in Introduction, such a structure may result from Kelvin-Helmholz instability or some kind of the RayleighTaylor instability in the magnetosphere-ionosphere system. In Figure 5, a counter-clockwise sense of rotation for the vortex structure of event I and II is seen. The rotation of the equivalent current is in agreement with this observation. Later, for event III the sense of rotation switches to clock-wise. The orientation of rotation of the ionospheric equivalent current switches as well. This behaviour can mean that the different auroral structures are related to regions of field-aligned currents of different directions. The region of upward field-aligned current may be associated with decrease of electron precipitation, which can leads to decrease of auroral luminosity and a break of the arc. The spatial scale of the auroral vortex is ~ 100 km and could therefore be classified as spiral [7]. The direction of rotation (clockwise) is usual for the spiral also. However, sometimes we observe contra-clockwise rotation, which can be explained by the spatial alternating between up- and downward FAC. 5. SUMMARY AND CONCLUSION In this study, optical auroral arc-like rotating structures have been compared with ground-based magnetic signatures. Both rotation of the optical structures and ionospheric equivalent currents were in the same direction and the TEC was perpendicular to the arcs. An electrodynamical model for the auroras has been chosen and proven to be reliable. The model includes alternating of up- and downward orientated FAC, which probably result from instability in the magnetosphere-ionosphere system.

Acknowledgements We would like to thank J. Olson for magnetometer data of the Geophysical Institute, University of Alaska. Without the all-sky images provided by Prof. Dr. T. Hallinan this work would have not been possible. REFERENCES

1. Amm et al., Ann. Geophysicae 16, 413, 1998. 2. Blixt and Brekke, Geophys. Res. Let., 23, 2553, 1996. 3. Danielides et al., Geophysica, 35 (1-2), 33, 1999. 4. Davis and Hallinan, J. Geophys. Res., 81(22), 3953, 1976. 5. Kamide et al., J. Geophys. Res., 86 (A2), 801,1981. 6. Kozlovsky and Lyatsky, Ann. Geophysicae, 12, 636, 1994. 7. Royrvik and Davis, J. Geophys. Res., 82, 4720, 1977. 8. Williams et al., Ann. Geophysicae, 16, 1322, 1998.

- 506 -

A Non-Solar Origin of the "SEP" Component in Lunar Soils? R. F. Wimmer-Schweingruber and P. Bochsler ~ aPhysikalisches Institut, Universit/it Bern, Sidlerstrasse 5, CH-3012 Bern, Switzerland Solar wind implanted in surface layers (~< 0.03#m) of lunar soil grains has often been analyzed to infer the history of the solar wind. In somewhat deeper layers, and thus presumably at higher implantation energies, a mysterious population, dubbed "SEP" for "solar energetic particle", accounts for 10% - 40% of the implanted g a s - ~ 3 - 4 orders of magnitude more than expected from the present-day flux of solar energetic particles [9]. In addition, its elemental and isotopic composition is distinct from that of the solar system. While the heavy Ne isotopes are enriched relative to 2~ 15N is depleted relative to 14N - a behavior that is hard to explain with acceleration of solar material. Here we show that variations in solar activity are not responsible for this component. 1. S O L A R A C T I V I T Y

IN THE PAST

An enhanced solar activity in the past has often been called upon to explain the "SEP" component. However, such an interpretation has severe difficulties explaining the isotopic composition of the various "SEP" gases. Isotopic fractionation in the solar wind is small and limited to at most few percent per mass unit even in the potentially most fractionated slow wind (see e. g. Bochsler [2] for a review). Moreover, the anticorrelation of the heavy isotopes of nitrogen and of neon can not be explained by acceleration of solar wind material. In addition, the isotopic composition of "SEP" He is not enriched in aHe, as would be expected if it were implanted suprathermal He. Impulsive events are strongly enhanced in 3He, with aHe/4He ratios sometimes exceeding unity. For quiet times, Mason and coworkers [5] have found an enhancement of the aHe/4He ratio by a factor of 10 above solar, inconsistent with the "SEP" component (aHe/4HesEp = 2.17 x 10 -4, i. e. below its solar value.). They interpret this as due to remnant suprathermal particles from impulsive events. A more active Sun in the past would imply a higher flux of impulsive-flare-related and aHe-enriched material. Next we show that a more active Sun in the past would also not explain the amount of the "SEP" component. The flux of solar particles in the past can be estimated quantitatively in the following manner. Solar wind particles leaving the Sun carry with them angular momentum, L. As they cross the Alfv6n radius, r A, they decouple from the Sun and remove this angular momentum at a rate L -

(1)

Here fl is a dimensionless factor describing the geometry of the problem and r A is the Alfv6n radius, i. e. the radius where Vsw - V a - B a / v / - f i - o - f i . Since solar wind outflow time

-507-

R.F. Wimmer-Schweingruber and P. Bochsler is fast compared to the age of the Sun, this reduces the solar angular rotation t -

(2)

where a - 2/5 for a homogeneous sphere. Equating Eqs.1 and 2 we can find the spin-down rate of the Sun: .

~'

.

.

.

s

.

~

vAm~o

,

(3)

where mass loss is given by r h - pVA47rR~ and we have related BA -- B(rA) to B o via BAR2A -- B o R ~ . Assuming that B scales linearly with solar rotation

B,(t)-

(4)

we have vA(t) -- VA(T)[W(t)/w(7)] 1/2, where 7 is today. differential equation for solar rotation with time, w(t), A

5/2

(t),

Inserting in Eq. 3 we have a

(5)

where we have absorbed all constant numerical factors in A. Integrating, and setting the integration constant such as to reproduce the present-day solar rotation frequency, we obtain

go(t)

(1 + a A ( t _ T ) ) 2 / 3 .

(6)

This trivial calculation represents the observed rotation periods of stars of various ages remarkably well as can be seen in Figure 1. There we show the derived rotation curve as a solid line which passes through the Sun (circled), for which we have precise knowledge of its age and rotation period. The rotation periods of a number of stars has been measured within the framework of the HK project (see [1]) and for stars in the Hyades (see e. g. [3]). The age of the Hyades is well known, while those of the other stars have been inferred from eq. 3 of [8]. The resulting uncertainties are large and are indicated for the Sun too, in order to highlight them. The spread of the data around the solid line is also due to the large spread in initial angular momentum of the stars. The only free parameter is the "slope" of the curve (the exponent in the time dependence). Obviously, the agreement with the data is good. As we have already mentioned, the curve was derived under the assumption that the magnetic field strength scales linearly with rotation. Another relation, e. g. the well-known Skumanich law [7] yields another exponent and hence another "slope" and results in less good agreement with observations of other Sun-like stars. The energy for the acceleration of solar particles has to come from the energy available in the magnetic field. This scales as B 2 and a long-term average can now easily be computed. (I)sw(t)

=

cd2(t)

(I)sw(T) w2(T)

=

1

(7)

Jr- 3A ( t -

-508-

A non-solar origin of the "SEP"component in lunar soils

,

i w

|

1

:

i w

I

I

I

i

I

i

]

I

10 9

i

i

I

I

I

I

I

10 ~~

Figure 1. Solar rotation versus time. Rotation periods of solid circles are from [6,1], their ages have been inferred using eq. 3 of [8]. The empty squares represent data for the Hyades (from [3].). The theoretical curve (solid line) is discussed in the text.

This can be integrated from the time of the formation of the first lunar regolith (that have been collected by astronauts) at an age of about 109 years to the present (4.57 x 109 years). This evaluates to about 2.5 times the present value of the solar wind flux. Hence the long-term average flux of solar particles cannot have been larger than about 2.5 times the present flux, insufficient to explain the overabundance of the "SEP" component. The average enhancement by a factor of 2.5 is in excellent agreement with that derived by Geiss [4] from measurements in lunar soil Kr and Xe. 2. C O N C L U S I O N S The "SEP" component observed in lunar soils and other solar-system samples is not necessarily of solar origin. No present-day suprathermal particle population is known that exhibits the compositional characteristics of the "SEP" population. Based on energy considerations we can rule out a strong enhancement in the past. In recent work [11,10], we have shown that interstellar pick-up ions (PUIs) which are ionized and accelerated in the heliosphere and subsequently implanted in lunar regolith grains can account for the properties of the "SEP" population. On average, interstellar pick-up ions were more abundant in the past than they are today, and must have been a substantial if not the dominant part of the suprathermal particle population in the heliosphere. This implies that lunar soils preserve samples of the galactic environment of the solar system and may

- 509 -

R.F. Wimmer-Schweingruber and P. Bochsler eventually be used as an archive for solar system "climate". Therefore, we recommend that the "SEP" be cautiously renamed to "HEP" for heliospheric energetic particles. A CKN OWLED G EMENT S This work was supported by the Schweizerischer Nationalfonds. REFERENCES

S. L. Baliunas, R. A. Donahue, W. H. Soon, J. H. Horne, J. Frazer, L. WoodardEcklund, M. Bradford, L. M. Rao, O. C. Wilson, Q. Zhang, W. Bennett, J. Briggs, S. M. Carroll, D. K. Duncan, F. Fiuearoa, H. H. Lanning, A. Misch, J. Mueller, R. W. Noyes, D. Poppe, A. C. Porter, C. R. Robinson, J. Russell, J. C. Shelton, T. Soyumer, A. H. Vaughan, and J. H. Whitney. Chromospheric variations in main-sequence stars II. Astrophys. J., 438:269- 287, 1995. P. Bochsler. Abundances and charge states of particles in the solar wind. Rev. Geophys., 38:247- 266, 2000. J. Bouvier, R. Wichmann, K. Grankin, S. Allain, E. Covino, M. Ferng~ndez, E. L. Mart~nand L. Terranegra, S. Catalano, and E. Marilli. COYOTES IV: the rotational periods of low-mass Post-T Tauri stars in Tauris. Astron. Astrophys., 318:495 - 505, 1997. J. Geiss. Solar wind composition and implications about the history of the solar system, volume 5, pages 3375 - 3398, 1973. Proceedings of 13th International Cosmic Ray Conference. G. M. Mason, J. E. Mazur, and J. R. Dwyer. 3he enhancements in large solar energetic particle events. Astrophys. J., 525:L133- L136, 1999. R. W. Noyes, L. W. Hartmann, S. L. Baliunas, D. K. Duncan, and A. H. Vaughan. Rotation, convection, and magnetic activity in lower main-sequence stars. Astrophys. J., 279:763- 777, 1984. A. Skumanich. Timescales for Ca, II emission decay, rotational breaking, and Li depletion. Astrophys. J., 171:565- 567, 1972. D. R. Soderblom, D. K. Duncan, and D. R. H. Johnson. The chromospheric emissionage relation for stars of the lower main sequence and its implications for the star forming rate. Astrophys. J., 375:722- 739, 1991. R. Wieler, H. Baur, and P. Signer. Noble gases from solar energetic particles revealed by closed system stepwise etching of lunar soil material. Geochim. et Cosmoschim. Acta, 50:1997- 2017, 1986. 10. R. F. Wimmer-Schweingruber. Lunar soils: A long-tern archive for the galactic environment of the solar system? Habilitation Thesis, 2000. Physikalisches Institut, Universit/it Bern, Switzerland. 11. R. F. Wimmer-Schweingruber and P. Bochsler. Is there a record of interstellar pick-up ions in the lunar regolith? In R. A. Mewaldt, J. R. Jokipii, M. A. Lee, E. Moebius, and T. H. Zurbuchen., editors, Acceleration and Transport of Energetic Particles Observed in the Heliosphere, pages 2 7 0 - 273, Woodbury, New York, 2000. AIP conference proceedings. .

o

.

o

.

.

o

.

-510-

Oral papers and posters

D I S C O N N E C T I O N E V E N T S IN C O M E T P / H A L L E Y ' S P L A S M A TAIL M.R. Voelzke and H.J. Fahr Institut fiir Astrophysik und Extraterrestrische Forschung. Universit~t Bonn. D-53121 Germany. Cometary and solar wind data are compared with the purpose of determining the solar wind conditions with comet plasma tail disconnection events (DEs). The cometary data are from The International Halley Watch Atlas of LargeScale Phenomena. A systematic visual analysis of the atlas images revealed, among other morphological structures, 47 DEs along the plasma tail of comet P/Halley. These 47 DEs documented in 47 different images allowed the derivation on 19 onsets of DEs, i.e., the time when the disconnections begin was calculated. The solar wind data are in situ measurements from IMP-8, which are used to construct the variation of solar wind speed, density and dynamic pressure during the analysed interval. This work compares the current competitive theories, based on triggering mechanisms, in order to explain the cyclic phenomena of DEs, i.e., the ion production effects, the pressure effects and the magnetic reconnection effects are analysed.

-511-

This Page Intentionally Left Blank

Cosmic Rays in the Outer Heliosphere and Nearby Interstellar Medium J. R. JokipiP* ~Department of Planetary Sciences, University of Arizona, Tucson, AZ, 85721 The effects of the outer heliosphere on cosmic rays are discussed. The outer heliosphere is assuming a greater importance in our understanding of both galactic and anomalous cosmic rays. In addition, these effects determine the distance beyond which we may say that we are free of the effects of the heliosphere in observing these particles and we see the unmodified interstellar spectrum. Alternatively, this may be where one should place the "modulation boundary", specifying the interstellar cosmic-ray spectrum, when solving the transport equation for galactic and anomalous cosmic-ray transport inside the heliosphere. We find that the location of this outer boundary for cosmic rays depends on the poorly-known diffusion coefficients of cosmic rays in the local interstellar medium, or alternatively, on the interstellar magnetic field and its fluctuation spectrum. If the scattering by the fluctuations in the local interstellar magnetic field is small, we have a free escape boundary, whereas if the scattering is significant, the boundary is very complicated. 1. I n t r o d u c t i o n . The heliosphere is observed to influence profoundly the fluxes of galactic cosmic rays (hereinafter GCR) observed within it. These effects are collectively called "the solar modulation" of the GCR. It is observed that the heliosphere significantly distorts the flux of galactic cosmic rays below an energy of about 10 ~2 eV, and effectively prevents any galactic cosmic rays with energies below about 250-300 MeV from being seen in the inner heliosphere. In addition, the interaction of the solar wind with the local interstellar medium (hereinafter termed LISM), results in the acceleration of anomalous cosmic rays (hereinafter termed ACR), which produce their own distortion of the spectrum observed within the heliosphere. In all of this, the outer heliosphere is of particular importance because both GCR and ACR must pass through it on their way to the inner heliosphere. The outer heliosphere, particularly beyond the termination shock, remains poorly understood. In this paper the effects of the outer heliosphere on cosmic rays is discussed, with emphasis on effects in the heliosheath and beyond. This region is becoming increasingly important as the Voyager spacecrafte approach the solar wind termination shock. Effects in these outer regions which can be treated relatively crudely for the study of cosmic rays in the inner heliosphere become more important. In addition, it is possible that these heliospheric effects may, distort and modify the intensities of the cosmic rays well outside the heliospere, in the LISM. The nature of the effects depends in large part on the poorly

-513-

J.R. JoMpii

determined transport parameters in the LISM. As far as we know, these effects have not been seriously addressed before in the published literature. We present here an initial analysis of the general nature of the expected effects. It is found that, if the interstellar magnetic-field spectrum is such that the local interstellar cosmic-ray diffusion mean free path is of the order or less than the scale of the heliosphere, or a few hundred AU, the effects of the heliosphere on the cosmic-ray flux extends several hundreds of AU out into the LISM.

Fig. 1. Schematic illustration of the ezpected configuration of the heliosphere and its interaction with the local interstellar medium. The wind flows out, passes through the terminations shock and is deflected to flow downstream in the direction of the interstellar flow. The interstellar plasma is deflected around the solar plasma and the surface between the solar plasma and the interstellar plasma is the heliopause. Upstream of the heliosphere there may be a bow shock if the flow of the interstellar plasma is super sonic or superalfv~nic. The meandering interplanetary magnetic field is represented by the dotted lines. The diffusion coefficients for cosmic rays in the various regions are denoted by I~.

On the other hand, if the local interstellar diffusion mean free path for cosmic rays is much larger than the scale of the heliosphere, the effects are more complex. The cosmic rays will propagate quite rapidly away from the heliosphere and in most cases the

-514-

'

Cosmic rays in the outer heliosphere and nearby interstellar medium guiding of the particles by the magnetic field, there will likely be local regions where the effects of the heliosphere are seen far from the heliosphere. The effects may be similar to those discussed by Mazur, et al [1], and Giacalone, Jokipii and Mazur [2] and seen in solar cosmic rays near the Sun. 2. I n t e r a c t i o n of Cosmic Rays with the Heliosphere The expected configuration of the heliosphere is illustrated schematically in figure 1. A variety of cosmic-ray effects arrise in this system. The galactic cosmic rays (herein GCR) are an essentially constant bath of particles coming in from the interstellar medium. In addition, anomalous cosmic rays are accelerated to energies greater than 1 GeV at the termination shock. The phenomenon of solar modulation of cosmic rays observed inside the heliosphere has been discussed extensively over the past few decades. Confrontation with in situ spacecraft observations both in the inner and outer heliosphere, and in the polar regions, has led to a successful physical picture. The basic phenomena observed by spacecraft within the termination shock are reasonably-well understood, and sophisticated numerical simulations have been carried out which explain the major observed effects. Remaining uncertainties revolve primarily around the values of the transport parameters, the location of the termination shock, etc. Similarly, the acceleration and transport of ACR are welldescribed by the same physical picture, where the ACE are accelerated at the termination shock of the solar wind. For a recent review of the status of this field, see the book edited by Fisk, et al [3]. Because of the complexity of the plasma and magnetic field beyond the termination shock, and the lack of observations, the situation there is much more poorly understood. Here, the flow of the solar wind is distorted by the flow of both the ionized and the neutral components of the interstellar gas, and magnetic-field stresses probably play a significant role. The interstellar and interplanetary plasmas and magnetic fields are separated by a "contact surface" of as yet unexplored shape and location. In the most recent cosmic-ray simulations, the Parker transport equation [4,5] is solved in a simplified heliosphere which captures the essential structure of the inner heliosphere, including effects of shocks, CIR's, etc, but in which the outer boundaries are taken to be spherical. The Parker transport equation for the pitch-angle-averaged distribution function f (r, p, t) of cosmic-ray particles at position r, momentum magnitude p, and time t may be written Ot =Oxi

~s)

-U

.Vf-Vd

"Vf+~V'U

Ognp

where the successive terms on the right-hand side correspond to diffusion, convection, particle drift, adiabatic cooling or heating and any local source Q. Here, for particles of speed w, momentum p and charge q, the drift velocity is Vd--

pcw v x

(2)

3q

where c is the speed of light and ~I.~) is the symmetric part of the diffusion tensor. In

-515-

J.R. Jokipii the diffusion coefficients parallel and perpendicular to it as

"

(3)

This equation is quite general, and applies whenever there is enough scattering by magnetic irregularities that the pitch-angle distribution is kept nearly isotropic. When equation (1) is applied to the problem of ACR or GCR, as described above, the equation is solved subject to a boundary condition where the distribution function f approaches the interstellar value at some radius D, which has typically taken to be some 50% larger than the radius of the termination shock. Clearly, this is a very crude approximation to the actual situation. Just what is the physical meaning of this radius? As far as we have been able to determine, this question has not been seriously discussed in the literature. This will be the topic of the next section.

3. The B o u n d a r y Conditions Models of cosmic-ray transport in the heliosphere require knowledge of the boundary conditions at the interface with the interstellar medium, as well as at the Sun. The specification of the boundary conditions is a complex physical problem that has not been discussed clearly in the literature. This has traditionally been handled by assuming that there is (generally spherical) boundary where the (steady over the relevant time scales) interstellar spectrum is taken to be the boundary value. Although the form of this spectrum is not well-determined, a reasonable guess can be made. The physical nature of this "modulation" boundary was not much discussed. In early models, this boundary was taken to be a sphere which was where the solar wind ended - or the termination shock. Some 15-20 years ago, it was realized that the the termination shock should be part of the heliosphere, since it accelerated the anomalous cosmic rays and affected significantly the galactic cosmic rays. In most applications the termination shock was simply set to be a spherical surface at some suitable radius, where the radial wind speed was reduced by the shock ratio,r~h. Very recent work has included the influence of cosmic rays on the shock and even included a non-spherical, self-consistent shock [6]. In these models, the boundary then moved outward to some sphere beyond the termination shock, where again the physics was not clearly determined. Fortunately, many of the consequences for the inner heliosphere are not very sensitive to the details of the transport beyond the shock. However, as we have seen above, some of the observations in the outer heliosphere are seeing the effects of heliosheath. It is important to treat this boundary more correctly. In order to do so, we must define it more explicitly. The question as to just what this boundary corresponds to, physically, has not been clearly discussed in the literature, but this must be done if it is to be treated more accurately. Figure (1) illustrates, schematically, the four relevant spatial regions, with their associated cosmic-ray diffusion coefficients. Within each of the regions the value of the coefficient may vary with position, and each t~ is really an anisotropic tensor. For the present purposes, I assume that they do not vary by large amounts as a function of position. More importantly, the diffusion coefficients mav varv considerablv as a function

-516-

Cosmic rays in the outer heliosphere and nearby interstellar medium

The most-important parameter is EISM, the diffusion coefficient in the local ISM. There are two important dimensionless parameters which will determine the nature of the boundary condition. They are (4)

,7(T) - R . V,s~, I'~ISM

and

(5)

I'~ISM

where the energy dependence of the numbers is emphasized. It is readily seen that if atSM is large enough that both ~(T) and r,+(T) are much less than unity, the diffusion in the local ISM is very rapid, and the boundary may be taken to be the standard "free escape" boundary at the heliospuse. Beyond this point the particles move so rapidly that the anomalous cosmic-ray intensity may be taken to be zero and the galactic cosmic-ray intensity takes on the full interstellar value. This is essentially what has been done in simulations carried out up to now. Putting numbers into equations (4) and (5), we find that this would require that t~rSM > > 1023cm2/sec. If arSM were to be of the order of 1023 or less, then the boundary condition would be much more complicated, and depend on the local structure of the interstellar medium and its magnetic field. 3.1. T h e Interstellar Diffusion Coefficient The interstellar medium contains a very wide spectrum of turbulence ranging from scales of the order of 10 ~9 cm down to perhaps 10 ~? cm. This spectrum is measured primarily in the electron density, but it is reasonable to expect that a similar spectrum is present in the magnetic field as well. Hence, we may regard energetic particles such as cosmic rays as propagating in a magnetic field which has an average component of some 5# gauss, with fluctuations around this average which are a manifestation of the broadband turbulence. As a consequence, the cosmic rays will gyrate about the magnetic field with a gyro-radius equal which is some few astronomical units. As these particles gyrate, their pitch angle and position relative to the field will be subject to fluctuations because of the turbulence. These effects lead to diffusion and the motion of a large number of particles may be shown to be well-approximated by equation (1). However equation (1) is still quite complex and involves a number of processes for which our understanding is limited. For example, the convection term can be important at low energies, but the structures of the flows are not known. Also, the drift velocity terms are complicated. Fortunately, we can show on quite general grounds that not all of the terms in the transport equation are large enough to be important for the cosmic rays of most interest.. It is easy to demonstrate that for the typical 1 GeV particles in the 5pgauss interstellar magnetic field the drift velocity is small because of the relatively small gyroradius compared to the macroscopic scales. The convection is likewise slow compared to the diffusive motions. Finally, the adiabatic energy change (proportional to V . U) terms may be neglected. In this case, we are left with the diffusion equation Ot =

(6)

+ Q

-517-

J.R. Jokipii

as a reasonable first approximation for the typical 1 GeV cosmic ray. This equation (or an even simpler approximation to it as discussed below) is the one most often used in discussing cosmic rays in the galaxy. The diffusion tensor aij can in principle be determined theoretically from the statistics of particle motion in the magnetic field. At present there are two approaches to doing this - the quasi-linear approximation or by following particle trajectories in synthesized turbulence. The quasi-linear approximation has been used for more than 30 years and is still quite popular [5], [8]. The determination of the motion parallel to the background magnetic field is relatively straightforward. The scattering in pitch angle is found to be determined by the magnetic field spectrum at scales comparable to the gyroradius in the background field. For simple models of the turbulence it is found that the parallel diffusion coefficient for a power law spectrum with index a is proportional to a power law in momentum,

(7) where the constant of proportionality depends on the nature of the turbulence (i.e., whether the turbulence depends on 1, 2 or three spatial dimensions. This, applied to the interstellar turbulence, gives a diffusion coefficient of the order of 1028 cm 2/sec. The perpendicular motion is less well determined. Although the general transport equations discussed above apply to the transport of cosmic rays in the galaxy, its application is difficult because some of the basic, underlying parameters are poorly known. Instead, it is often adequate to work in terms of an even more simplified picture in which the diffusion term in equation (6) is replaced by a loss time TL. If L is the characteristic dimension of the disk, then, in order of magnitude, L2

(8)

It turns out that he principal observed properties of the cosmic-ray intensity in the energy range near 1 GeV can be understood in terms of a simple picture in which the galaxy is treated as a uniform, homogeneous storage vessel, into which the particles are injected and then confined for a mean time for loss from the galaxy ~-L (which is, in general, a function of energy), after which they escape from the galaxy. The value TL turns out to be much smaller than the age of the galaxy, so the system is in a steady state. In addition, 7L turns out to be much greater than the interval between supernova explosions, so it is assumed that the source is also homogeneous and continuous in the storage vessel. The equation expressing particle conservation can then be written as Of ,.~ 0 ,.~ Ot

f

(9)

+Q

TL

or

(10)

f -- 7-LQ

where f is the cosmic-ray distribution function (the observed sDectrum di/dT can be

-518-

Cosmic rays in the outer heliosphere and nearby interstellar medium

of particles. To the accuracy required, we may assume the particle to be relativistic, in which case pc - T, and energy losses due to ionization may be neglected. Considerations such as this lead to an approximate value of the loss time 7- and thence a diffusion coefficient of the order of 102s - 1029cm2/sec, a value remarkably similar to that obtained from quasilinear theory. From this one may conclude that,, if the general conditions in the interstellar medium apply to the local vicinity of the heliosphere, the proper boundary condition is that of a free escape boundary. Note that, if the interstellar irregularity spectrum is indeed as this would imply, the fluctuations in the magnetic field at scales of the order of a few parsecs are of order unity. Since in Kolmogorov turbulence, the mean square fluctuations over a spatial scale ~ scale as t~2/a, we find that the fluctuations at the ~ 1AU scale resonant with a several GeV galactic proton are given by ~B B "~ 1 0 - 3 - 10-2" (11) This is a very smooth magnetic field. A spacecraft would only see a .1% variation in the magnetic field direction or magnitude in a distance of a tenth of an AU.

4. S u m m a r y and Conclusions The above discussion suggests that the commonly-used free escape boundary condition for galactic and anomalous cosmic rays is a good approximation. Only if the fluctuations in the interstellar magnetic field scales of the order of 1 AU are much larger than implied by our current knowledge of interstellar turbulence, would the boundary condition change. Such fluctuations would require a significant input of energy into the local interstellar medium over scales of the order of hundreds of AU to pruduce magnetic fluctuations at scales of an AU. While present knowledge doesn't rule this out, it seems unlikely. Since the magnetic field is likely to be quite smooth on scales of the particle gyroradius, the effects of the heliosphe on cosmic rays will extend out into the interstellar medium on those magnetic field lines whih connect to the heliosphere. This is similar to the effects seen in connection with impulsive solar-flare events [1,2].

REFERENCES 1. ,I.E. Mazur, G.M. Mason, J. R,. Dwyer, J. Giacalone, J. R,. Jokipii, and E. C. Stone, Ap. J., 532 (2000) L79. 2. J. Giacalone, J. R. Jokipii, and J.E. Ap. J., 532 (2000), L75. 3. L. A. Fisk, J. R. Jokipii, G.M. Simnett, R. von Steiger, and K.-P. Wenzel (Eds), Cosmic Rays in the Heliosphere (1998) Kluwer. 4. E.N. Parker, Plan. Space Sci., 13, (1965) 9. 5. J.R. Jokipii, 1986, Rev Geophys. and Sp. Phys, 9, (1971) 27. 6. V. Florinski and J. R. Jokipii, AP.J., 523, (1999) L185. 7. J.R. Jokipii, in Physics of the Outer Heliosphere, Cospar Colloquia Volume 1, (1990)p 169. 8. ,John W. Bieber and William H. Mattheaus, Ap. J., 485 (1993) 655.

-519-

This Page Intentionally Left Blank

COSPAR Colloquium on The Outer Heliosphere Potsdam, Germany, 24-28 July 2000 List of Participants

-521

-

List of part&ipants

-

522-

List of participants

-

523-

List of participants

-524-