Magnetosheath response to the interplanetary magnetic field tangential discontinuity J. afr nkov, L. Pfech, and Z. N6meeek

Transcription

1 JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 105, NO. All, PAGES 25,113-25,121, NOVEMBER 1, 2000 Magnetosheath response to the interplanetary magnetic field tangential discontinuity J. afr nkov, L. Pfech, and Z. N6meeek Charles University, Faculty of Mathematics and Physics, Prague, Czech Republic D. G. Sibeck Applied Physics Laboratory, Johns Hopkins University, Laurel, Maryland T. Mukai Institute of Space and Astronautical Science, Sagamihara, Japan Abstract. We present a multipoint observational study of the magnetosheath response to the interplanetary magnetic field tangential discontinuities which form hot flow anomaly-like (HFAs) structures. We identify these structures in the vicinity of the bow shock as well as deeper in the magnetosheath. Two or more points of simultaneous observations allow us to describe the gradual evolution and propagation of these HFAs through the magnetosheath. From tens of events recorded by INTERBALL-i, we present two cases. In the first, GEOTAIL identified HFAs in the solar wind near the bow shock, and INTERBALL-1 and MAGION-4 observed related events in the magnetosheath. During the second interval, all three spacecraft observed the HFA features in the magnetosheath. Our analysis of reported events suggests the negligible evolution of these structures in the magnetosheath. A survey of the INTERBALL-1 data has shown that magnetosheath HFAs are observed predominantly during periods of fast solar wind. 1. Introduction hancement. (3) HFAs occur in conjunction with signif- The interaction of the solar wind with the Earth's icant changes in the IMF direction. The angle between magnetosphere generates a population of backstreaming ions directed from the bow shock into the solar wind. This high-energy population was invoked to explain hot flow anomalies (HFAs) [e.g., Schwartz et al., preevent and postevent orientations is typically ø. Thomsen et al. [1986, 1988] have shown that many HFAs occur when the geometry of the bow shock is changing and large fraction of specularly reflected ions 1985; Thomsen et al., 1986] identified as heated regions are present. These and other observations led to many of solar wind plasma flowing nearly perpendicular to theoretical/numerical studies. One-dimensional simulathe Earth-Sun line. The main observational features of tions with a finite length backstreaming ion beam sug- HFAs include the following [Schwartz, 1995]: (1) central regions with hot plasma flowing significantly slower than that in the ambient solar wind in a direction highly deflected (nearly 90 ø ) from the Sun-Earth line. The flow velocities are often roughly tangential to the nominal bow shock shape [Schwartz et al., 1988]. (2) HFAs are bounded by regions of enhanced magnetic field strength, density, and temperature. The outer edges of these enhancements are fast shocks generated by pressure engested that the interaction of the beam and background plasma can produce hot, low-density regions from which the solar wind plasma is largely excluded [Onsager et al., 1990b]. However, an examination of the ion temperature inside HFAs indicated that complete thermalization of the ion beam and subsequent adiabatic expansion of the heated plasma leads to final ion temperatures that are generally below those observed [Onsager et al., 1990a]. Thomas and Brecht [1988] presented a hancements within the core region. The inner edges two-dimensional (2-D) hybrid simulation for a beam of of the enhancements are probably tangential disconti- backstreaming ions relative to the ambient solar wind. nuities [Paschmann et al., 1988]. Published examples The authors demonstrated that the thermalized backindicated that many HFAs are bounded by only one en- streaming ions create a diamagnetic cavity of depressed magnetic field strengths and densities. According to Copyright 2000 by the American Geophysical Union. Paper number 1999JA / 00 / 1999J A $ numerical simulations of Thomas et al. [1991] and Lin [1997], kinetic effects at the intersection of the magnetic discontinuities with the bow shock can create very deflected flows of heated plasma surrounded by enhanced 25,113

2 , 25,114 AFR,/ NKOV,/ ET AL.' MAGNETOSHEATH RESPONSE densities and magnetic field strengths. Although such features should be swept downstream, only a few HFAs have been observed in the magne- tosheath [Paschmann et al., 1988; Thomsen et al., 1988; Schwartz et al., 1988]. Almost all reported magnetosheath events have been identified near the bow shock, and thus the question of their evolution and further propagation through the magnetosheath remains open, although their impact on the magnetosphere can be remarkable as demonstrated by a recent case study of $ibeck et al. [1998, 1999]. The authors described the magnetospheric response to an interplanetary magnetic field (IMF) tangential discontinuity which first struck the postnoon bow shock and magnetopause and then swept past the prenoon bow shock and magnetopause. Although unaccompanied by any notable plasma variation, the IMF discontinuity interacted with the bow shock to form an HFA. The depressed pressure in the core of the HFA then resulted in a sunward magnetopause displacement of more than 5 Re. This case is probably extreme, but it shows that HFAs can drive significant magnetopause boundary motion which then complicates the data interpretation. We have taken advantage of International Solar Ter- restrial Physics (ISTP) era observations by many spacecraft simultaneously monitoring the solar wind, foreshock, bow shock region, and magnetosheath to carry out a multipoint study of the HFA-like structure evolution. We used Wind and IMP 8 observations to detect the motion of the discontinuity through the solar wind into the foreshock (or into the vicinity of the bow shock) and INTERBALL-I, MAGION-4, and GEOTAIL observations to determine the effect of the event on the magnetosheath. We are presenting two case studies and a survey of three months (January- March 1997) of the INTER- BALL observations. During this interval, INTERBALL- Tail probes spent long time intervals in the magnetosheath and observed 26 events which could be at- tributed to the arrival of IMF discontinuities and the formation of HFA-like structures at the bow shock. 2. Observations To demonstrate HFA properties, we have chosen two representative examples with favorable positions for all mentioned spacecraft. The first case study shows an interval when two of three IMF tangential discontinuities identified in Wind solar wind data resulted in HFAs. These HFAs are observed in the bow shock region by GEOTAIL and simultaneously by INTERBALL-1 in the magnetosheath. The second case is related to the propagation of HFAs through the magnetosheath. We have found an interval when the same HFA is observed by GEOTAIL in the dayside magnetosheath and by INTERBALL-1 and MAGION-4 in the dusk magnetosheath flank, 20 Re downstream. In order to show that the solar wind velocity can be an important factor for the HFA creation, we have reanalyzed data published by nsager et al. [1990a] and complemented their survey of International Sun Earth Explorers (ISEE) and Active Magnetospheric Particle Tracer Explorers (AMPTE) observations by our analysis of INTERBALL- 1 measurements Case 1 On August 31, 1996, the INTERBALL-1 satellite registered a series of HFA-like events in the dawn magnetosheath at (-0.2,-17.2,-5.1)csE Re. Figure I shows observations during two of these events. Figure I pre- sents 15 s time resolution Omnidirectional Plasma Sen- sor (VDP)ion flux [$afrankova et al., 1997] and I s time resolution three-axial fluxgate DC/AC magnetometer (MIF-M) magnetic field [Klimov et al., 1997] observa- tions for the interval from 0740 to 0820 UT. The first HFA at UT as well as the second one at UT can be identified by rapid decreases of the ion flux bounded by transient enhancements (Figure la). The time resolution of the measurements does not allow us to see the full amplitude of the enhancements in this case, but our analysis of similar events has shown that they can exceed the mean magnetosheath flux by a factor of 3. The cone angle (Figure lb) of INTERBALL-1 31-Aug '... '... ' ool lo... i... i... ' 15 lo... '...' m i... i oooo t'i,.-...,'i, [, , \- -, ; ,.[.-.,,, -.-:,;, ). ':": :.:'.i:i.'.:; :::; ::=.::':.!.::':. :' iili:?- :,, -- - ' ' '-;!;.. [i :. :..'- : ::::?:.-": :-..-'...' ::: '..'..... loo i :.... -, --,, -, :; :,,-'-:-i', :,,, 'i :- -? i',,, "...?' ;"':??':i? i '? ",," :, '/ :20 UT Figue 1. An observation Of HFAs in the magnetosheath: INTERBALL-1 data. (a) The fvi P (ion flux), (b) alpha (cone angle between ion bulk velocity and the XcsE coordinate), (c) fi O C (flux of electron in the range kev), (d) B (magnetic field magnitude), Bx, (e) By, and Bz (components of the magnetic field), (f) Ee0 (electron energy spectrogram). 3.0 '"'

3 AFR NKOV, ET AL.' MAGNETOSHEATH RESPONSE 25,115 the ion flux shows that the flow is highly deflected. Whereas the undisturbed magnetosheath flow is deflected by - 20 ø from the Sun-Earth's line, consistent with INTERBALL-I's location near the terminator, the ions inside the structures are highly deflected. They flow nearly perpendicularly to this line during our second event. The ELECTRON spectrometer [$auvaud et al., 1997] observed a hotter electron population during both events (Figure lf), and, also, the Energetic Particle Experiment (DOK-2) [Lutsenko et al., 1995; Kudela et al., 1995] count rate for electrons in the energy range from 28 to 31 kev increased during both events (Figure lc). The ion temperature inside both structures (which is not shown in Figure 1) is higher than that in the surrounding magnetosheath. The magnetic field strength fluctuated, but its magnitude was roughly correlated with the ion flux. From the 10-s running averages shown in Figure 1, one can clearly see that the aforementioned enhancements of the ion flux bounding the HFA corresponds to the local maxima in the magnetic field magnitude, whereas clear magnetic field depressions are observed during intervals of low and deflected ion flux. Both observed HFA-like events correspond to changes in the magnetic field direction, and in fact the mean values of all components have different signs before and after events. The events resemble all features reported for HFAs with one exception, both of them exhibit a double structure. Ion flux increases at 0747 and 0809 UT divide the core regions into two parts with different magnetic field orientations. Moreover, the increased fluxes of highenergy electrons only occur in the first part. This behavior is common in the HFA-like structures analyzed in the present study, but it has not been previously reported in either the solar wind or the magnetosheath. During studied time interval, three spacecraft monitored the solar wind and IMF conditions. Wind was far upstream, but near the Sun-Earth line at (X, ¾, Z)GSE = (98.5,-3.5,-5.5)GSE Rs, IMP 8 was on the dawn side at (-7,-34.8, 16)GSE Rs, and GEOTAIL crossed the bow shock at UT and moved outbound. At 0800 UT, it was located - I Rs upstream of the bow shock at (0, -23.5, -2.8)GSE Rs. Figure 2 presents the IMP 8 IMF and Wind plasma and IMF measurements with I min time resolution. The WIND & IMP-8 31-Aug ,5...,...,..., 2.0 1'5 i - I, 0.0 ' :... ::... I... I... ( 6 f... '... i (c) 4 ', _ o -2 Bz. 6 ',......,... 2 = o (d) ]... ' 4 ' (e) , UT (WIND +26 min.) Figure 2. Solar wind dynamic pressure and interplanetary magnetic field as measured by Wind (subscript WI) and IMP 8 (subscript I8) during the INTERBALL- 1 observation of magnetosheath HFAs. tion and tangential discontinuities, but only the last discontinuity in Figure 2 exhibits a notable change of the magnetic field magnitude, which is usually considered as a simple test to distinguish tangential and rotational discontinuities. If the first discontinuity is tangential, we can compute its normal as (B 1 x B2)/(IBll. lb21) where the sub cripts 1 and 2 refer to the region upstream and downstream of a discontinuity, respectively. We tested the normal direction by minimum variance analysis of the Wind high-resolution (3 s) data. Both approaches yield the same direction of the discontinuity normal (n ,0.67, 0.70). The normal defines the discontinuity plane, and we can compute the time Wind data are shifted by 26 min to obtain the best coincidence of the measurements from both spacecraft. The solar wind was quiet (mean speed km.s - lag for observations of the discontinuity at Wind, IMP 8, and GEOTAIL (GEOTAIL observations will be discussed in detail later). Figure 3 shows the positions of and mean density cm -3) as is indicated by the the satellites projected into an ecliptic plane and the inunchanged dynamic pressure in Figure 2a. The profiles and values of the IMF components are tersection of the estimated discontinuity with this plane. The discontinuity was observed by Wind at 0719 UT, similar at both spacecraft, indicating that all observed by IMP 8 at 0745 UT, and by GEOTAIL at 0747 UT. features are of solar wind origin. One can identify The observed lags are different from those obtained by three clear discontinuities in the IMP 8 data: at 0745, simple calculations based on the solar wind velocity and 0758, and 0805 UT. The location of the spacecraft (Fig- spacecraft separation in the solar wind direction. This ure 3) suggests that the first and third can be associ- calculation yields, for example, a lag of min beated with the events observed at 0744 and 0805 UT by tween Wind and IMP 8 instead of - 26 min observed. INTERBALL-1. According to Schwartz [1995], HFAs are associated with large rotations in the IMF orienta- However, we supposed that a planar discontinuity with the normal derived above from the Wind magnetic field

4 25,116 AFR. NKOV. ET AL.' MAGNETOSHEATH RESPONSE WIND '1 '1 WIND (182,-12.4,17.2)! /(98,-3.5,-5.5) i x ', B l GE,/ Y /a,,, ', j_b G E B2 1 in the magnetosheath is shown for comparison. The plasma parameters depicted in the figure were computed from two Low Energy Particle instrument (LEP) sensors. Intervals when data from the Solar Wind ion analyzer (LEP-SW) are used are distinguished by dotted lines above the density profile, whereas the Energyper-charge Analyzer (LEP-EA) is used for the times remaining (see Mukai et al. [1994] for details). Unfortunately, both detectors were saturated during the events. This led to an underestimate of the density and overestimate of the temperatures. For a better understanding, Plate 1 presents a set of GEOTAIL electron and ion energy spectrograms for this time interval. One can note the increase of the electron counts in all direc- tions from 0747 to 0750 UT and again from 0810 to 0814 UT. These increases indicate a rise in the electron / Janua 31,1997 August31,1996 x temperature. The ion energy spectra of the SW analyzer (Plate lc) are lost during this time interval due to Figure 3. A projection of the satellite positions into ecliptic plane during discussed events. The right part a large deflection of the ion flow. The ion spectra from of the figure belongs to August 31, 1996, event; the left the EA detector (Plate lb) show that the ion flow is part refers to January 31, 1996, event. The intersection oriented antisunward and dawnward (i.e., roughly tanof discontinuities with the ecliptic plane is shown by gentially to the nominal bow shock in Figure 3) durdouble lines. ing intervals of increased electron temperature. The broadness of the EA spectrograms between 0747 and 0749 UT and between 0810 and 0814 UT, in a cornparis frozen into and propagated with the solar wind, and thus we computed the lags in three dimensions. The lags obtained this way are close to those observed. To test the frozen-in assumption and the stability of the discontinuity, we have run a minimum variance analysis of the IMP 8 magnetic field. Although IMP 8 was separated from Wind by more than 100RE in space and - 26 min in time, the discontinuity conserved its normal direction with an accuracy - 5 ø. The normal directions of the other two discontinuities are very similar, but the uncertainty of their determination is higher because the changes of the magnetic field are less pro- _ INTERBALL-1 & GEOTAIL 31-Aug nounced. The discontinuities propagate with the solar :88 wind velocity, and thus there cannot be a significant o flow across them. The normal component of the mag netic field obtained by the minimum variance analysis -200 was negligible in all cases. Thus we can conclude that both necessary conditions (Bn - 0 and un - 0 where B and u are normal components of the magnetic field 0-50 and velocity, respectively) are fulfilled and that all dis continuities in this time interval are tangential discontinuities. Such discontinuities can create HFAs, when the motional electric field sw x ] is oriented toward the discontinuity on one or both sides of the current sheet 2O [Linet al., 1997; Thomsen et al., 1988]. We tested this 15 criterion and found that it is obeyed for the first and 10 5 third discontinuities but not for the second one. Thus 0 only the first and third discontinuities can create HFAs when interacting with the bow shock. This is consistent O800 UT with magnetosheath observations presented in Figure 1 Figure 4. A comparison of magnetosheath and sowhere only two HFA-like events are seen during the time lar wind observations during the HFA events. (a) fib interval under question. magnetosheath ion flux (INTERBALL-l), (b)nge, (c) Figure 4 shows G EOTAIL observations in the vicin- VXGE, (d) VyGE, (e) VZGE, (f) TiGE, and (g) BGE, the ity of the bow shock at the times of the events discussed above. The ion flux measured by INTERBALLsolar wind density, velocity, ion temperature, and magnetic field magnitude (GEOTAIL).

5 , ; AFR,NKOV, ET AL.' MAGNETOSHEATH RESPONSE 25,117 I M- lo = (a) m m RAM-B _ 10 1 O m m m ' o.1 O.Ol I ' " 0.1 o.1 O.Ol 10 1 I. o.1 _ o.1 o.ol 'm. ' :?'! ' ' ' '" ' ':r" I,,,I 4.. ; 10 "., 0. " ' ' - " 5 _ "'" - m.'- " -';" ', I o.-j..; '-' '"' ' I " ' I - - ' ' '"Z,, " ' ' %'.1 m -- ' ' 20 I ' ' - lo " I -' C o. -.'..-.. i :-'- s m..¾... m m t %1 "!1 I I.. to. ; ".L. '. 2O, I "11 ' ', '1. ' "..' ' '",:, -., ',o on =.1:::' i,.,....' 5 PJ M-,A II, II, III I ml RAM-B =,....:. -..,.,. E ---,. _., O I m 'r m i m,,, i m i,'m m m i m m m i, m! m 'm i m m m 07: 07:45 07:50 07:55 08: :05 08:10 08:15 08:20 u- Universal Count ample,'..., ck B of Da ' Plate 1. The GEOTAIL observation of HFAs at the bow shock region. (a) Electron energy spectrograms in four directions, the counts correspond to the left energy scale. (b) The ion energy spectrograms in the same directions (EA device). During the intervals discussed in the paper, the device was in the wide-energy mode (RAM-A) with the energy scale on the left. (c) The ion spectrogram of the SW device. RAM-B 2o lo

6 25,118 AFR NKOVA ET AL.' MAGNETOSHEATH RESPONSE ison with the spectrograms immediately following, suggests a large ion temperature. To avoid a confusion, we would like to point out that the EA detector was operating in two energy scanning modes which was switched automatically in accordance with the SW detector data. The modes differ by the energy range, and they are denoted as RAM-A (30 ev- 39 kev) and RAM-B ( kev) at the top of these particular spectrograms. We have discussed only the data received in the wide-energy (RAM-A) mode. The corresponding energy scale is on the left-hand side of the spectrograms. The events observed by GEOTAIL exhibit some HFA features such as a flux depression bounded by enhancements in the magnetic field and an increase of the temperature but no significant plasma depletion. A possible explanation of the density profile is that the arrival of the IMF discontinuity formed the HFA which than caused outward motion of the bow shock [Schwartz et al., 2000; $ibeck et al., 1999], and GEOTAIL under- went short excursions into the magnetosheath where the plasma density is higher. The ion velocity was highly deflected from the Sun-Earth line as usually observed inside HFAs. Outside of the events, GEOTAIL observes solar wind features similar to Wind and IMP 8. Precise timing is impossible because the HFAs significantly disturb all the parameters. Nevertheless, the beginnings of the disturbances coincide with the estimated arrival times of the IMF discontinuities at GEOTAIL's position. The fact that INTERBALL-1 in the magnetosheath observes the events earlier than GEOTAIL in the solar wind can be explained taking into account the spacecraft positions and the geometry of the disturbance (see Figure 3). If the disturbance is created at the bow shock, it will propagate to INTERBALL-I's position from the subsolar region at the magnetosheath velocity. An estimated path of the disturbance through the mag- netosheath is shown with a dashed line in Figure 3. On the other hand, the normal to the IMF discontinuity has a significant component perpendicular to the solar wind flow and thus the discontinuity sweeps along the bow shock rather slowly. This means that the disturbances observed by GEOTAIL may not be the cause of those observed by INTERBALL-1. For this reason, we cannot say if the differing durations of the events observed by these two satellites is due to different dimensions of HFAs in the subsolaregion (where the INTERBALL-1 events originated) and on the flank or if it is caused by a further evolution of HFA in the magnetosheath Case 2 A similar situation occurred on January 31, The IMF tangential discontinuity observed by Wind at 1953 UT resulted in the creation of an HFA observed in the dusk magnetosheath by three spacecraft: GEO- TAIL, INTERBALL-I, and MAGION-4. For the sake of simplicity, Figure 5 shows only the ion flux and magnetic field magnitudes observed by these spacecraft, but we have carefully examined all parameters to ensure INTERBALL-I, GEOTAIL & MAGION-4 31-Jan-1997,.-., 8... '..., , 30 = 20,,.- lo o 4.o ,,--., 15 c lo m UT Figure 5. Propagation and evolution of HFA through the magnetosheath. B and f stand for the magnetic field magnitude and ion flux, respectively, and subscripts are abbreviations of the spacecraft names. that the magnetosheath event has HFA characteristics. It should be noted that the characteristics of the event differ a little at different spacecraft positions. The locations of the spacecraft projected onto the ecliptic plane are schematically shown in the left part of Figure 3. GEOTAIL was located in the magnetosheath near the magnetopause and registered the leading enhancement of HFA at 2034 UT. Owing to the decreased density in the core region of HFA, GEOTAIL crossed magnetopause and entered the plasma sheet. It exited into HFA at 2041 UT to observe the trailing enhancement of the ion flux. INTERBALL-1 and MAGION-4 observed the same event at higher latitudes from 2044 to 2047 UT. The ion flux inside the core region of the HFA fell to zero, and the temporal increase of the magnetic field magnitude observed by both spacecraft in the core of the event suggests that they both entered the lobes. Nevertheless, during intervals of low-magnetic field strength, they observed the deflected flow which characterizes HFAs. Since all three spacecraft crossed the magnetopause, we cannot analyze the evolution of the plasma parameters during the event and instead concentrate on the duration of the event as seen at different locations. The observations indicate that the HFA propagates downstream because it is seen first by GEOTAIL in the subsolaregion at (7.2, 10.1, 0)½SE RE and later by MAGION-4 and INTERBALL-I, which were located

7 AFR NKOV/[ ET AL.' MAGNETOSHEATH RESPONSE 25,119 GEOsTAI.L,. interba,ll-,1,& MAGION-4 31-Jan o 3.'-.2 -=1 o 3 -' i UT (GEOTAIL sec.) Figure 6. A detail of the HFA ion flux profile as measured by three satellites. From top to bottom: (a) GEO- TAIL, (b) INTERBALL-l, and (c) MAGION-4. on 0 20-RE tailward at (-5.4, 20.3, 11.8)GSE-RE and (-6.0, 21.1, ll.4)gse RE, respectively. Figure 3 depicts the positions of the satellites and the orientation of the IMF discontinuity. Figure 6 shows a detail of the ion flux profile as measured by all three spacecraft. The 420 s shift of the GEOTAIL data corresponds to the separation of the spacecraft in the direction of the magnetosheath flow and measured magnetosheath speed ( km.s -1 from the GEOTAIL data). After this shift the events coincide, and we can conclude that all spacecraft observed an HFA carried by magnetosheath flow. However, the duration of the event is much shorter at INTERBALL-1 than at GEOTAIL. Differing durations can be attributed either to temporal evolution or a variation in the width of the HFA along the ZGSE axis. The later possibility is more probable because the duration of the event is 0 12 s shorter in the MAGION-4 we have taken into account only those events which had all basic HFA characteristics, i.e., a region of tenuous highly deflected flow and low magnetic field bounded by short regions of significantly enhanced ion density and magnetic field magnitude. We exclude events con- nected with bow shock or magnetopause crossings because the plasma and magnetic field behavior are more complicated in such cases and a complex analysis of all parameters is needed to decide whether the particular event is connected with the HFA. We identified 26 clear HFAs in the magnetosheath like those in Figure 1. Of these, 23 occurred during intervals when the solar wind speed exceeded 400 km.s- (usually km.s - ) despite the fact that during the 3 month period the solar wind speed was less than this value for 1168 hours and greater than it for only 883 hours. The high-speed solar wind stream has, as a rule, a low density. In our survey, the upstream density during HFA events never exceeded 8 cm -3 (it was usually 2-3 cm-a). We have analyzed the influence of other upstream parameters (as Mach number or plasma/ ), but we did not find any clear correlation with the occurrence of HFAs. 3. Discussion We have presented two multipoint case studies of the magnetosheath response to the arrival of IMF tangential discontinuity at the bow shock. In both of them, an HFA is created at the bow shock and then observed in the magnetosheath. The INTERBALL-1 observation of HFA in the magnetosheath is supported by simultaneous GEOTAIL observations at the bow shock region in our first case. The analysis of this event shows that the characteristics of HFA in the solar wind and in the magnetosheath are basically the same. HFAs can be distinguished as regions of the hot tenuous plasma bounded by density enhancements in both regions. data than in the INTERBALL-1 data and MAGION-4 In the second case, a single HFA was observed in the is located on 0.4_RE further from the ecliptic plane than INTERBALL-1. dusk magnetosheath near the magnetopause by three spacecraft. We suggested that the differing profiles The exact determination of the event duration is of the events at different points of the magnetosheath rather questionable because the shape of the event differs in the measurements by different spacecraft due to low-frequency waves which are always present in the could be attributed to the shape of a HFA cavity rather than to a temporal evolution of its dimensions on a timescale of several minutes. We suggesthat the duramagnetosheath. A further observational fact support- tions of HFAs depend on the locations where IMF dising our conclusion that the different duration of HFA continuities intersect the bow shock. Thus the different is a spatial effect is the shape and amplitude of the durations of the events observed by INTERBALL-1 and bounding flux enhancements which do not exhibit any GEOTAIL in our first case may result from GEOTAIL's systematic evolution A Survey of INTERBALL HFA separation from the subsolaregion where the plasma observed by INTERBALL-1 enter the magnetosheath. Simulations indicate that HFAs need 10 ion gyroperi- Observations ods to develop into notable disturbances [Thomas et al., Both cases analyzed above occurre during a period 1991]. However, if the angle between the normal to the of relatively high solar wind speed (505 km.s -1 for the original IMF discontinuity and the solar wind velocity first case and 495 km.s -1 for the second case). In or- is larger than 45 ø (as in our case 1), the discontinuity der to determine whether or not this is a random oc- sweeps along the shock surface with a speed increasing currence, we surveyed 3 months of the INTERBALL-1 from the subsolar point toward the flank. In this case, observations (January- March 1997). For our survey one expects HFAs to be more developed in the subsolar

8 25,120 ; AFR,&NKOV,& ET AL.' MAGNETOSHEATH RESPONSE region. However, if the normal to the discontinuity is aligned with the solar wind, the apparent velocity of its motion along the shock surface is very high in the subsolar region. This effect may partly explain why HFAs are observed rather rarely under this condition [Schwartz et al., 2000]. Another reason why the observational evidence for magnetosheath HFAs is limited may lie in our finding that they tend to occur during periods of high solar wind speed. A similar tendency can be inferred from the list of HFAs observed by ISEE and AMPTE IOnsager et al., 1990a]. The authors have analyzed a similar number (21) events and summarized the parameters in a table. From this table it follows that for 18 of the 21 events the solar wind speed exceeded 400 km.s -. In both sets of HFA observations the high velocity was frequently connected with a low density (less than 8 cm -3 in our study and less than 11 cm -3 as given by Onsager et al. [1990a]). It is hard to say whether the principal factor favoring the HFA creation is the high upstream speed or low upstream density and further efforts in data analysis and modeling are needed. Our survey indicates that HFA-like structures are rather frequent in the central magnetosheath, but despite this, there is little evidence in the literature for them. The magnetosheath flow is often highly disturbed, and some HFA features may often be attributed to other phenomena. For example, Nemecek et al. [1998] described transient flux enhancements (TFEs)in the magnetosheath, but some of these enhancements can be related to the magnetosheath HFAs. The authors discussed the INTERBALL-1 measurements dur- ing the time interval from 0730 to 1200 UT on August 31, 1996, and related the observed ion flux profile to TFEs, but a few of the enhancements observed during this time are connected with HFAs, as we demonstrate in Figure 1. However, it should be noted that the flux spikes bounding HFA meet the criteria for TFE identification. TFEs were attributed to the interaction of the foreshock magnetic field discontinuities with the bow shock, and thus they can be observed only behind the quasiparallel shock. However, HFAs can originate regardless of the IMF orientation whenever the discontinuity meets the criteria specified above. This led to the clear identification of HFAs in the magnetosheath behind the quasiperpendicular bow shock, whereas behind the quasiparallel shock they can be mixed with other phenomena. The double structure, which some of the magnetosheath HFAs exhibit, is probably connected with the mechanisms of the HFA formation in front of the bow shock. An indication of such structure is seen in GEO- TAIL ion-energy spectrograms at 0812 UT (Plate 1) when GEOTAIL was located at the bow shock. Moreover, careful examination of simulated profiles of the plasma parameters by Lin [1997] shows a slight density enhancement at the middle of the HFA core region. It seems that the "seed" for this structure is created at the bow shock, but the structure evolves in the magnetosheath. 4. Conclusion HFAs are an important magnetosheath phenomenon. They can be encountered throughout the magnetosheath from the bow shock to the magnetopause. Our observations are limited by the INTERBALL-1 satellite's orbit to the range IXGsEI < 10 RE, but the negligible evolution of reported events suggests that HFAs should be observed further tailward, if the duration of the event (.. several min) and the temporal resolution of observations allow their identification. We have pointed out a double structure of magnetosheath HFAs. A complete description of this feature requires a multidevice analysis of high-time resolution data. The magnetosheath HFAs exhibit a clear tendency to occur predominantly during periods of enhanced solar wind speed. This needs more careful examination because high-speed streams have different characteristics than slow-speed streams, perhaps computer simulation can determine which factors are most important in HFA formation. Acknowledgments. The authors thank the Wind, IMP 8, GEOTAIL, and INTERBALL working teams for the magnetic field and plasma data. The present work was supported by the Czech Grant Agency under Contracts 205/99/1712 and 205/00/1686 and by the Charles University Grant Agency under Contract 181. The support of the NATO grant CRG.CRGP which allowed the mutual cooperation of Czech and American authors is greatly acknowledged. Janet G. Luhmann and the authors thank Shigeru Fujita and Zuyin Pu for their assistance in evaluating this paper. References Klimov, S., et al., ASPI experiment: Measurements of fields and waves onboard the Interball-1 spacecraft, Ann. Geophys., 15, , Kudela, K., M. Slivka, J. Rojko, and V. N. Lutsenko, The apparatus DOK-2 (project INTERBALL): Output data structure and modes of operation, Tech. Rep. UEF , 65 pp., Institute of Experimental Physics, Slovak Academy of Science, Kosice, Slovakia, Lin, Y., Generation of anomalous flows near the bow shock by its interaction with interplanetary discontinuities, J. Geophys. Res., 102, 24,265-24,281, Lutsenko, V. N., J. Rojko, K. Kudela, T. V. Gretchko, J. Balaz, J. Matisin, E. T. Sarris, K. Kalaitzides, and N. Paschalidis, Energetic particle experiment DOK-2 (Interball project), in: INTERBALL Mission and Payload, edited by Y. Galperin, T. Muliarchik, and J-P. Thouvenin, pp , CNES-IKI-RSA, Paris-Moscow, Mukai, T., S. Machida, Y. Saito, M. Hirahara, T. Teresawa, N. Kaya, T. Obara, M. Ejiri, and A. Nishida, The low energy particle (LEP) experiment onboard the GEOTAIL satellite, J. Geomag. Geoelectr., J6, , Nemecek, Z., J. Safrankova, L. Prech, S. Kokubun, T. Mukai, and D. G. Sibeck, Transient flux events in the magnetosheath, Geophys. Res. Lett., 25, , Onsager, T. G., M. F. Thomsen, J. G. Gosling, and S. J. Bame, Observational test of a hot flow anomaly formation mechanism, J. Geohys. Res., 95, 11,967-11,974, 1990a. Onsager, T. G., M. F. Thomsen, and D. Winske, Hot flow

9 anomaly formation by magnetic deflection, Geophys. Res. Lett., 17, , 1990b. Paschmann, G., G. Haerendel, N. Sckopke, E. Moebius, H. Luehr, and C. W. Carlson, Three-dimensional plasma structures with anomalous flow directions near the Earth's bow shock, J. Geophys. Res., 95, 11,279-11,294, Safrankova, J., G. Zastenker, Z. Nemecek, A. Fedorov, M. Simersky, and L. Prech, Small scale observation of magnetopause motion: Preliminary results of the Interball project, Ann. Geophys., 15, , Sauvaud, J.-A., P. Koperski, T. Beutier, H. Barthe, C. Aoustin, J. J. Throcaven, J. Rouzaud, E. Penou, O. Vaisberg, and N. Borodkova, The INTERBALL- Tail ELEC- TRON experiment: Initial results on the low-latitude boundary layer of the dawn magnetosphere, Ann. Geophys., 15, , Schwartz, S. J., et al., An active current sheet in the solar wind, Nature, $18, , Schwartz, S., Hot flow anomalies near the Earth's bow shock, Adv. Space Res., 15, , Schwartz, S. J., R. L. Kessel, C. C. Brown, L. J. C. Woolliscroft, M. W. Dunlop, C. J. Farrugia, and D. S. Hall, Active current sheets near the Earth's bow shock, J. Geophys. Res., 93, 11,295-11,310, Schwartz, S. J., G. E. Paschmann, N. Sckopke, T. M. Bauer, M. W. Dunlop, A. N. Fazakerley, and M. F. Thomsen, Hot flow anomalies revisited, Int. J. Georaag. Aeronoray, in press, Sibeck, D. G., N. L. Borodkova, G. N. Zastenker, S.S. Romanov, and J.-A. Sauvaud, Gross deformation of the dayside magnetopause, Geophys. Res. Lett., œ5, , 1998., AFR, NKOV, ET AL MAGNETOSHEATH RESPONSE 25,121 Sibeck, D. G., et al., Comprehensive study of the magnetospheric response to a hot flow anomaly, J. Geophys. Res., 10d, , Thomas, V. A., and S. H. Brecht, Evolution of diamagnetic cavities in the solar wind, J. Geophys. Res., 95, 11,341-11,353, Thomas, V. A., D. Winske, M. F. Thomsen, and T. G. Onsager, Hybrid simulation of the formation of a hot flow anomaly, J. Geophys. Res., 96, 11,625-11,632, Thomsen, M. F., J. T. Gosling, S. A. Fuselier, S. J. Bame, and C. T. Russell, Hot diamagnetic cavities upstream from the Earth's bow shock, J. Geophys. Res., 91, , Thomsen, M. F., J. T. Gosling, S. J. Bame, K. B. Quest, C. T. Russell, and C. T. Fuselier, On the origin of hot diamagnetic cavities near the Earth's bow shock, J. Geophys. Res., 95, 11,311-11,325, T. Mukai, Institute of Space and Astronautical Science, Yoshinodai 3-1-1, Sagamihara, Kanagawa 229, Japan. Z. N mes. ek, L. P ech, and J. afr nkov, Faculty of Mathematics and Physics, Charles University, V Hole5oviSok ch 2, Prague, Czech Republic. (jana.safrankova@ mff. cuni.cz) D. G. Sibeck, Applied Physics Laboratory, Johns Hopkins University, Johns Hopkins Road, Laurel, MD (Received November 29, 1999; revised March 16, 2000; accepted April 3, 2000.)