GeophysicalM onographS eries Including Maurice Ewing Volumes Mineral Physics Geophysical Monograph Series 1 Antarcticai n the International GeophysicalY ear, A. P. Crary, L. M. Gould, E. O. Hulburt, Hugh Odishaw, and Waldo E. Smith (Eds.) 2 Geophysicsa nd the IGY, Hugh Odishaw and StanleyR uttenberg( Eds.) 3 AtmosphericC hemistryo f Chlorine and Sulfur Compounds,J amesP . Lodge, Jr. (Ed.) 4 ContemporaryG eodesy, Charles A. Whitten and Kenneth H. Drummond (Eds.) 5 Physicso f Precipitation, Helmut Weickmann (Ed.) 6 The Crust of the Pacific Basin, Gordon A. Macdonald and Hisashi Kuno (Eds.) 7 Antarctic Research:T he Matthew Fontaine Maury Memorial Symposium, H. Wexler, M. J. Rubin, and J. E. Caskey, Jr. (Eds.) 8 Terrestrial Heat Flow, William H. K. Lee (Ed.) 9 Gravity Anomalies: Unsurveyed Areas, Hyman Orlin (Ed.) 10 The Earth Beneath the Continents: A Volume of Geophysical Studies in Honor of Merle A. Tuve, John S. Steinhart and T. Jefferson Smith (Eds.) 11 IsotopeT echniquesi n the Hydrologic Cycle, Glenn E. Stout (Ed.) 12 The Crust and Upper Mantle of the Pacific Area, Leon Knopoff, Charles L. Drake, and Pembroke J. Hart (Eds.) 13 The Earth's Crust and Upper Mantle, Pembroke J. Hart (Ed.) 14 The Structure and Physical Propertieso f the Earth's Crust, John G. Heacock( Ed.) 15 The Use of Artificial Satellites for Geodesy, Soren W. Henriksen, Armando Mancini, and Bernard H. Chovitz (Eds.) 16 Flow and Fractureo f Rocks,H . C. Heard, I. Y. Borg,N . L. Carter, and C. B. Raleigh( Eds.) 17 Man-Made Lakes: Their Problems and Environmental Effects, William C. Ackermann, Gilbert F. White, and E. B. Worthington (Eds.) 18 The Upper Atmosphere in Motion: A Selection of PapersW ith Annotation, C. O. Hines and Colleagues 19 The Geophysicso f the Pacific Ocean Basin and Its Margin: A Volume in Honor of GeorgeP . Woollard, George H. Sutton, Murli H. Manghnani, and Ralph Moberly (Eds.) 20 The Earth's Crust: Its Nature and Physical Properties, John G. Heacock (Ed.) 21 Quantitative Modeling of MagnetosphericP rocessesW, . P. Olson (Ed.) 22 Derivation, Meaning, and Use of GeomagneticI ndices, P. N. Mayaud 23 The Tectonic and Geologic Evolution of Southeast Asian Seas and Islands, Dennis E. Hayes, (Ed.) 24 Mechanical Behavior of Crustal Rocks: The Handin Volume, N. L. Carter, M. Friedman, J. M. Logan, and D. W. Stearns( Eds.) 25 Physics of Auroral Arc Formation, S.-I. Akasofu and J. R. Kan (Eds.) 26 HeterogeneousA tmospheric Chemistry, David R. Schryer (Ed.) 27 The Tectonic and Geologic Evolution of SoutheastA sian Seas and Islands: Part 2, Dennis E. Hayes, (Ed.) 28 MagnetosphericC urrents, Thomas A. Potemra (Ed.) 29 Climate Processesa nd Climate Sensitivity (Maurice Ewing Volume 5), JamesE . Hansen and Taro Takahashi (Eds.) 30 MagneticR econnectioinn S pacea ndL aboratorPyl asmasE,d wardW . HonesJ, r. (Ed.) 31 PointD efectsi n Minerals( MineralP hysicsV olume1 ), RobertN . Schock(E d.) 32 The CarbonC yclea nd AtmospheriCc O2:N aturalV ariationsA rcheant o PresentE, . T. Sundquisat nd W. S. Broecke(rE ds.) 33 GreenlandIc e Core:G eophysicsG, eochemistrya,n dt he EnvironmentC, .C. Langway,J r., H. Oeschger,a nd W. Dansgaard( Eds.) 34 CollisionlessS hocksi n the Heliosphere: A Tutorial Review, Robert G. Stone and Bruce T. Tsurutani (Eds.) Maurice Ewing Volumes 1 Island Arcs, Deep Sea Trenches, and Back-Arc Basins, Manik Talwani and Walter C. Pitman III (Eds.) 2 DeepD rilling Resultsin the AtlanticO cean:O ceanC rust,M anikT alwani,C hristopheGr . Harrison, and Dennis E. Hayes (Eds.) 3 Deep Drilling Resultsi n the Atlantic Ocean:C ontinentalM argins and Paleoenvironment, Manik Talwani, William Hay, and William B. F. Ryan (Eds.) 4 EarthquakeP rediction--AnI nternationalR eview, David W. Simpsona nd Paul G. Richards GeophysicalM onograph3 5 Collisionless Shocks in the Heliosphere: Reviews of Current Research Bruce T. Tsurutani and Robert G. Stone Editors American Geophysical Union Washington,D .C. Published under the aegis of the AGU Geophysical Monograph Board: Donald Eckhardt, Chairman; Elaine Oran, James Papike, John Schaake, and Sean Solomon, members. Collisionless Shocks in the Hellosphere: Reviews of Current Research Library of Congress Cataloging in Publication Data Main entry under title: Collisionless shocks in the heliosphere(cid:127) (Geophysical monograph, ISSN 0065-8448; 35) Includes bibliographies. 1. Heliospherc Addresses, essays, lectures. 2. Shock waves--Addresses, essays, lectures. I. Tsurutani, Bruce T. II. Stone, Robert G. (Robert Gilbert), 1928- . III. Series. QC881.2.H43C65 1985 551.5'11 85-13508 ISBN 0-87590-061-5 ISSN: 0065-8448 Copyright 1985 by the American Geophysical Union, 2000 Florida Avenue, NW, Washington, DC 20009 Figures, tables, and short excerpts may be reprinted in scientific books and journals if the source is properly cited. Authorizationt o photocopyi tems for internal or personalu se,o r the internal or personalu se of specific clients, is granted by the American Geophysical Union for libraries and other users registered with the Copyright Clearance Center (CCC) TransactionaRl eportingS ervice,p rovidedt hat the base fee of $1.00p er copy,p lus $0.20p er page is paid directly to CCC, 21 CongressS treet, Salem,M A 01970.0 065-8448/85/$01+.. 20. This consent does not extend to other kinds of copying, such as copying for creatingn ew collective works or for resale. The reproductiono f multiple copiesa nd the use of full articleso r the use of extracts,i ncludingf iguresa nd tables,f or commericalp urposesr equiresp ermissionf rom AGU. Printed in the United States of CONTENTS Introduction vii MACROSTRUCTURE Theories of Shock Formation in the Solar Atmosphere R.S. Steinolfson 1 Observations of Shock Formation and Evolution in the Solar Atmosphere Jean-Louis Bougeret 13 Review of Interplanetary Shock Phenomena Near and Within 1 AU A.K. Richter, K. C. Hsieh, A. H. Luttrell, E. Marsch, and R. Schwenn 33 Interplanetary Shocks on the Large Scale: A Retrospective on the Last Decade's Theoretical Efforts V. J. Pizzo 51 Interplanetary Shock Phenomena Beyond 1 AU Edward J. Smith 69 Magnetohydrodynamic and Gasdynamic Theories for Planetary Bow Waves John R. Spreiter and Stephen S. Stahara 85 Planetary Bow Shocks C.T. Russell 109 MICROSTRUCTURE Subcritical Collisionless Shock Waves M.M. Mellott 131 Ion Reflection, Gyration, and Dissipation at Supercritical Shocks J.T. Gosling and A. E. Robson 141 Numerical Simulations of Quasi-Perpendicular Collisionless Shocks C.C. Goodrich 153 Oblique, Parallel, and Quasi-Parallel Morphology of Collisionless Shocks Eugene W. Greenstadt 169 Simulation of Quasi-Parallel Collisionless Shocks Kevin B. Quest 185 Electron Velocity Distributions Near Collisionless Shocks William C. Feldman 195 Plasma Waves and Instabilities Donald A. Gurnett 207 Microtheory of Collisionless Shock Current Layers D. Winske 225 The Electron Foreshock Alexander J. Klimas 237 Upstream Suprathermal Ions M.F. Thomsen 253 PARTICLE ACCELERATION Shock Drift Acceleration Thomas P. Armstrong, Mark E. Pesses,a nd Robert B. Decker 271 Diffusive Acceleration Manfred Scholer 287 Notation Introduction Violent expansions of the solar corona cause transient shock waves which propagate outward from the sun at hundreds to thousands of kilometers per second; simple solar wind velocity gradients at the surface of the sun lead to high-speed streams overtaking slower streams, forming corotating shocks; and steady state supermagnetosonic solar wind flow past objects such as the planets lead to standing bow shocks. However, the solar wind plasma is so hot and tenuous that charged particle Coulomb collisions produce negligible thermalization or dissipation on scale sizes less than 0.1 AU. The irrevers- ible plasma heating by these shocks is accomplished by wave-particle interactions driven by plasma instabilities. Hence these shocks are described as (cid:127)collisionless." Collisionless shocks are interesting and important for numerous reasons. Collisionless shocks are the simplest configuration in which a macroscopic flow is regulated by microscopic dissipation, a problem common to many different plasma processes.C ollisionless shocks are therefore of basic plasma physical interest. There are also many important ways in which shocks affect the near-earth environment. Coronal shocks are believed to be responsible for the acceleration of solar flare energetic particles, which then propagate outward to fill the heliosphere. Shock propagation into the outer heliosphere may be a principal cause of the solar cycle dependent cosmic ray modulation. Interplanetary shock interactions with the earth's magnetosphere cause magnetic storms, intense low-latitude aurorae, and radio blackouts. Recent observations of fields and particles near interplanetary shocks and upstream of the earth's bow shock allow us to study particle acceleration processesi n situ, giving us first-hand knowledge of processesw hich are occurring not only in our heliosphere but which may give us important insights into plasma processesw hich are occurring near distant interstellar shocks, pro- cessesw hich are believed to create cosmic rays. These two volumes update our current knowledge of collisionless shocks in the heliosphere, an area in which recent major advances have been provided by space probes from NASA, the European Space Agency, and the Federal Republic of Germany. Individual papers in these volumes will take you from a detailed look at processeso ccurring at and near the shock itself (plasma instabilities responsible for dissipation of the streaming energy, and heating and acceleration of the plasma) to the characteristics of the upstream and downstream particles and their role in the shock structure. They discussh ow these processes change as the strength of the shock and the properties of the upstream plasma and the orientation of the magnetic field are varied. The past development and present state of gasdynamic, MHD, multifluid, and hybrid numerical simulation codes are summarized, and numerous applications to collisionless shocks are discussed.S everal papers deal with the development of shocks near the sun and the earth and their outward propagation and eventual interactions with slower solar wind and/or other shocks in the outer heliosphere. Particle acceleration mechanisms for parallel and perpendicular shocks are described. The relationship between theory and observations forms the central dialogue of these volumes. The monographs are intended to serve as summaries of the present status of our knowledge of collisionless shocks in the heliosphere. Volume 34, composedo f four tutorials, serves as a general text for entering graduate students and, for a scientist from a related field, provides the background needed to understand fully the topical reviews of volume 35. The 19 papers of volume 35 focus on specific topics of collisionless shocks and are intended to be reviews (and a source of references) of the present status of research in this field. Acknowledgments. We wish to thank the referees, listed below, for the countless hours they devoted to critical review of the papers appearing in the two volumes: K. A. Anderson, T. P. Armstrong, A. Barnes, L. F. Burlaga, D.C. Ellison, D. H. Fairfield, M. A. Forman, D. W. Forslund, R. W. Fredricks, B. E. Goldstein, M. L. Goldstein, J. T. Gosling, E. W. Greenstadt, C. A. Gurgiolo, T. Hada, J. D. Huba, M. K. Hudson, J. R. Kan, P. J. Kellogg, A. J. Lazarus, M. A. Lee, R. P. Lepping, R. P. Lin, W. H. Matthaeus, K. Papadopoulos, G. K. Parks, G. Paschmann, K. B. Quest, P. Rodriguez, C. T. Russell, F. L. Scarf, J. D.-Scudder, D. D. Sentman, N. R. Sheeley, Jr., G. L. Siscoe, E. J. Smith, S.S. Stahara, S. T. Suess, J. A. Van Allen, K. P. Wenzel, R. T. Woo, and C. S. Wu. The Chapman conference, held in Napa, California, February 20-24, 1984, and these books owe a major portion of their successt o the tireless efforts of Christina Brokl of the Jet Propulsion Laboratory and Barbara Holland of the Goddard Space Flight Center. These books could not have been possible without the support of the National Science Foundation and the National Aeronautics and Space Administration and, in particular, the moral and financia! support of the International-Sun-Earth-Explorer Project. We also wish to express our special appreciation to the American Geophysica! Union's publication staff, who provided exceptiona! manage- ment support throughout this endeavor. BRUCE T. TSURUTANI Jet Propulsion Laboratory California Institute of Technology Pasadena ROBERT G. STONE NASA Goddard Space Flight Center Greenbelt, Maryland Geophysical Monograph Series Collisionless Shocks in the Heliosphere: Reviews of Current Research Vol. 35 Theories of Shock Formation in the Solar Atmosphere R. S. STEINOLFSON 1 High Altitude Observatory, National Center for Atmospheric Research, Boulder, Colorado 80307 Models for the generation of coronal shocks by large-scale, disruptive solar events are reviewed. The theoretical requirements for shock formation in the corona are discussed.V arious approaches used to model the shock-initiating events are analyzed with distinctions drawn between the driving mechanisms in the theories. The impor- tance of the pre-event coronal magnetic field configuration and magnitude in influ- encing the subsequent coronal response and shock structure is also discussed. Partic- ular emphasis is placed on the information furnished by the theories which can be compared with observations and, similarly, the observational results which provide critical tests for the theory. Such comparisons will be shown to suggest a revised representation of the shock-transient association. Recent models of the observed spa- tial relationship of type IIs and transients are examined. 1. Introduction The ultimate test for any theory of coronal shock for- Large-scale disruptions of the solar corona have been mation is a direct comparison with available observa- studied extensively since the start of data acquisition tions. Correlative studies, demonstrating that interplan- from orbiting white-light coronagraphs. These disrup- etary shocks can often be associated with coronal mass tions are generally referred to as coronal transients (the ejections [Gosling et al., 1975; Sheeley et al., 1983] and term used herein) or, for the subclass which involves ex- solar activity [Hundhausen, 1972; Gosling, 1975], indi- pulsion of coronal material, coronal mass ejections. Com- cate that at least some white-light observations may in- pressive waves form as a natural consequence of such clude shocks. Data from three orbiting coronagraphs energetic transient phenomena and, under suitable con- (Skylab [MacQueen et al., 1974], Solwind [Sheeley et al., ditions, steepen into shocks in either the corona or inter- 1980], and SMM [House et al., 1981]) and the ground- planetary space. The term corona will be used here to based K-coronameter [Fisher and Poland, 1981] are cur- refer to that portion of the solar atmosphere within ap- rently available. These instruments record electron scat- proximately 5 solar radii of the surface. tered light and thereby provide information on the spa- Potential sources of transients include flares, eruptive tial and temporal evolution of the line-of-sight coronal prominences, and unstable or rapidly evolving coronal mass. A density increase due to a sufficiently strong magnetic field configurations. Transitory coronal shock would be observed by coronagraphs, although the changes also occur on a much slower time scale than in shock usually would not appear as a sharp discontinuity the above impulsive events, although this review will be in the integrated measurement. Weak shocks and shocks limited, for the most part, to the more energetic, rapid with a very thin compressed region could be lost in in- events that are likely to result in shocks. We will not strument noise. White-light images, then, furnish infor- consider other potential sources of coronal shocks, such mation on the transient producing the shock and, possi- as standing shocks that occur in some solar wind solu- bly, on the shock wave itself. tions [Hasan and Venkatakrishnan, 1982; Whang, 1982; Another source of observations is metric type II radio Habbal and Tsinganos1,9 83]a nd shocksp roduced bursts (such as those recorded by the Clark Lake Radio subphotospheric oscillations [Hollweg et al., 1982]. Observatory and the Nancay and Culgoora radiohelio- graphs), which are generally interpreted as being associ- 'Permanent affiliation: Department of Physics,U niversity of ated with shocks. Satellite observations of kilometer California, Irvine, California 92717. wavelength solar radio bursts also provide information Copyright American Geophysical Union Geophysical Monograph Series Collisionless Shocks in the Heliosphere: Reviews of Current Research Vol. 35 2 SHOCK FORMATION IN THE SOLAR ATMOSPHERE i i i 1 tions 4.2 and 4.3. The final section contains a brief sum- t = :30 rntn 45 min 1.0 - (a) ---• . • mary. 2. Shock Formation in the Corona 0.5 Ai- =r 15• min •, The essential requirement for finite-amplitude distur- , \ bances to steepen into shocks (at least in a one- dimensional model) is that the local characteristic (fast 0 mode) speed in the disturbance exceed the characteristic 0.5 - (b) speed in the medium through which it is traveling. Consequently, compressive MHD waves, formed during •,T i = 15m ln '•.• eruptive solar events, should steepen as they propagate i away from their point of origin. Whether or not the 2 waves evolve into fully developed shocks in the corona, SOLARR ADIUS(R /R s) or even in the interplanetary medium, is another matter since the steepening process requires a certain time Fig. 1. Shock formation in the lower corona for (a) rela- tively large-amplitude and (b) small-amplituded isturbances.D is- period. We are neglecting here the possibility of the de- turbances were initiated at t- 0 at the solar surface, and the velopment of solitons, in which dissipative processesb al- curves represent the steepening profiles at two successive times ance the nonlinear steepening. for several cases. For instance, in Figure lb the wave does not From a theoretical standpoint, a shock forms when steepen into a shock for a 5 min rise time A•i, while a shock does characteristics of the same family cross so the solution at form when A•i is reducedt o 1 min. The initial atmospherew as the point of intersection is no longer unique [Zel'dovich assumed to be in hydrostatic equilibrium with an embedded and Razier, 1966]. This principle has been used to derive magnetic field parallel to the surface. Average values of sound analytic expressions for shock formation time for MHD speeda nd beta are 150 km s-' and 0.04, and 7 = 1.05. waves by Montgomery [1959] and Cohen and Kulsrud [1974]. Numerical simulations by Steinolfson [1981] pre- on electron acceleration by coronal shocks [Cane et al., dict that the steepening process may occur somewhat 1981]. Recent efforts to analyze simultaneousr adio emis- faster than these theories estimate. sion and white-light data for the same transient [e.g., The time of shock formation in the above analyses de- Stewart et al., 1982; Gergely et al., 1984; Gary et al., 1984] pendso n the ambientm ediuma nd the spatial gradienti n are extremely useful in establishing the shock-transient the initial disturbance. A complication occurs in the relationship. solar atmosphere because the energy source producing Some additional data sources( such as X ray and EUV), the propagating disturbance may increase in amplitude while not furnishing any information on the shock or with time. If the increase is too slow, the disturbance transient directly, are, nonetheless,v aluable in the study may propagate out of the corona before a shock forms. A of shock formation. They place constraints on the simple estimate for how rapid the energy increase must amount and time-dependenceo f the energy released in be, in order that a shock form before the disturbance has the solar event and the magnetic field before and follow- traveled a specified distance, can be derived using the shock intersection criteria for fast mode MHD character- ing the eruption. Ha observationsm ay be useful for un- derstanding transients in those casesw here the material istics. If a solar event increases the velocity in the re- emitting in white light can be identified as originating in sulting disturbance by an amount Av above the ambient a prominence. velocity v, then a shock forms within l providing the time The steepening of compressivem agnetohydrodynamic required to reach Av is less than Azi given by (MHD) waves into shocksd ependso n the magnitudeo f lAv the wave disturbance, gradients in the wave, and the am- A•i = (1) (C• + v)(C• + v + Av) bient medium. This general problem is addressedi n the following section,w ith specifica pplication to shock for- whereC s = Cs(1+ 2/7•)' /2. The quantitiesC sa nd fl are mation in the solar atmosphere.S ome of the models that the ambient sound speed and plasma beta (ratio of ther- have been suggesteda s representative of coronal tran- mal to magnetic pressures),r espectively. sients are discussed in section 3. Most of these models do The above expression was evaluated in the corona not include the transient-atmosphere interaction (and, using a one-dimensionalM HD code, which solves the therefore, shockst hat may form in that interaction) as an complete nonlinear, time-dependent equations. The re- integral part of the analysis.T hus, in relating theory and sults, in a static atmosphere,f or two values of Av, repre- observations in section 4, the quantitative comparisons sentative of the energy input in the solar event, are are limited to models which include shock formation. shown in Figure 1. The evolving velocity profiles near These data and a critical evaluation of the theory sug- the leading edge of the disturbances are shown at two gest revised pictures of the shock-transienta nd the different times following their initiation at the solar sur- shock-transient-typeI I associationsa s discussedi n sec- face. For the lower energy event in Figure lb, the profile Copyright American Geophysical Union Geophysical Monograph Series Collisionless Shocks in the Heliosphere: Reviews of Current Research Vol. 35 STEINOLFSON 3 shows no indication of steepening for the slowest rise It must be kept in mind that the analysis of this section time (Azi - 15 min). The profile is clearly steepening for is based on a one-dimensional model in which the ambi- the 5-min rise time, although it has not yet shocked. Re- ent characteristic speeds are relatively well-behaved. ducing A•i to I min causes the shock to be fully formed, That is, they do not undergo large variations on small as indicated by the steepness of the profile and by its (solar radius) spatial scales. Multi-dimensional effects faster propagation speed. These results agree with (1), would tend to retard shock formation, and significant which predicts that the shock should form before reach- changes in the ambient atmosphere would certainly alter ing 5 solar radii if A• _<2 min for this case. The larger the results given here. amplitude disturbance in Figure la shocks for a 5-min 3. Models of Coronal Transients rise time, which also agrees with the (1) estimate of A• i _< 6 min. There are two main areas where the model predictions Having established that (1) may provide a reasonable and the observations overlap. First, the mass excess com- approximation of theoretical shock formation times, it is puted in a particular model can be converted to bright- of interest to determine if the specified relationships in ness and compared with the brightness of a transient the equation coincide with observations. First of all, observed by a white-light coronagraph. Thus, the simu- since A• is an upper limit, one would expect that a very lated and observed transient shapes and their motion are rapid energy deposition, which occurs on, for instance, directly comparable. The actual brightness levels provide the time scale of the impulsive phase of the flare, would a quantitative comparison of somewhat less value due to produce shocks in the corona. This is consistent with assumptions of transient line-of-sight depth and back- correlative studies which show that type II bursts (in- ground density. Second, if shocks are predicted in the dicative of coronal shocks) tend to be associated with the model, they can be compared with radio observations. faster ( > 400 km s-1) coronal mass ejections [Sheeley et A common denominator in all models is the magnetic al., 1984], which, in turn, are preferentially associated field. The mechanism driving the transient generally in- with flares [Gosling et al., 1976]. On the other hand, a volves this quantity, either directly or indirectly. The slow energy input, representative of that which may atmospheric response is also substantially influenced by occur during an eruptive prominence, may not produce a the ambient magnetic field magnitude and configuration. shock in the corona, although it may eventually steepen The shock formation process is generally not included in interplanetary space. Once again, this is in agreement as a constituent part of the transient formation and prop- with the above correlations. agation in the models discussed here. Two exceptions are Another significant quantity affecting coronal shock the stationary-driver models and the self-similar solu- formation is the value of fi in the ambient atmosphere. tions of Low [1982]. This serious shortcoming must be Smaller values of fi, larger magnetic fields, require a addressed at some point in the development of each indi- more impulsive event, in order to obtain the necessary vidual model before it can be used in quantitative com- steepening. Although the sound speed does not vary ap- parison with observations. In the self-similar solutions, preciably in the lower corona, (1) predicts that shocks which cannot model the early stages of transient devel- form later in hotter atmospheres. opment, the problem of shock formation is avoided by The above formulation applies to solar events in which assuming that the shock is sufficiently ahead of the tran- the major driving mechanism remains essentially station- sient. ary near the solar surface. Some models for coronal tran- All models contain enough free parameters that they sients, as discussed in the following section, require that can reasonably reproduce at least one observed the driver be an integral part of the transient. The driver quantity--the leading-edge velocity along a single radial motion must be accounted for, in this case, in deriving a line. This necessary requirement, however, is virtually relation similar to that in (1). useless in determining whether a model can satisfactorily There are a couple of points worth emphasizing with describe the physics of the driver-transient-shock interac- respect to Figure 1. Until the shock forms, the leading tion throughout the entire disturbance. edge of the transient always travels at the local fastest A major simplifying assumption in the models, at least characteristic speed, independent of the magnitude of the in their mathematical description, is that a complete velocity increase in the disturbance (providing the dis- three-dimensional treatment is not necessary. There is, of turbed velocity is less than the fastest characteristic course, no observational evidence to support a two- speed). Note that it is not necessary for the fluid velocity dimensional representation, although a strictly one- in the disturbance to exceed any characteristic speed in dimensional treatment is clearly inappropriate. One ad- the ambient medium in order that shocks form. For exam- ditional mutual simplification is that dissipative effects ple, the velocity increase in the case in Figure lb is less can be neglected, except at shocks. than half the ambient sound speed. Finally, for a rapidly Some of the models are now described with a critical increasing driving energy so large that the disturbed ve- examination of their advantages and disadvantages in locity exceeds the ambient fast speed, shocks form almost their current state of development. Recent reviews that immediately. discuss theories for transient formation have been given Copyright American Geophysical Union