Pages 999-1015

INTERBALL MAGNETOTAIL BOUNDARY CASE STUDIES

S. P. Savin1, O. Balan6, N. Borodkova1, E. Budnik1, N. Nikolaeva1, V. Prokhorenko1, T. Pulkkinen7, N. Rybjeva1, J. Safrankova8, I. Sandahl11, E. Amata2, U. Auster4, G. Bellucci,2 A. Blagau6, J. Blecki3, J. Buechner4, M. Ciobanu6, E. Dubinin1, Yu. Yermolaev1, M. Echim6, A. Fedorov1, V. Formisano2, R. Grard9, V. Ivchenko16, F. Jiricek5, J. Juchniewicz3, S. Klimov1, V. Korepanov10, H. Koskinen7, K. Kudela13, R. Lundin11, V. Lutsenko1, O. Marghitu6, Z. Nemecek8, B. Nikutowski4, M. Nozdrachev1, S. Orsini2, M. Parrot12, A. Petrukovich1, N. Pissarenko1, S. Romanov1, J. Rauch12, J. Rustenbach4, J. A. Sauvaud14, E. T. Sarris15, A. Skalsky1, J. Smilauer5, P. Triska5, J. G. Trotignon12, J. Vojta5, G. Zastenker1, L. Zelenyi1, Yu. Agafonov1, V. Grushin1, V. Khrapchenkov1, L. Prech8, O. Santolik8

1Space Research Institute, Russian Academy of Sciences, Profsoyuznaya 84/32, GSP-7, Moscow,   117810, Russia, E-mail: SSAVIN@iki.rssi.ru
2Interplanetary Space Physics Institute, CNR, Frascati, Italy

3Space Research Center, Polish Academy of Sciences, Warsaw, Poland
4Max-Planck Institut fur Extraterrestrische Physik, Aussenstelle, Berlin, Germany
5Institute of Atmospheric Physics, Academy of Sciences of the Czech Republic, Praha, Czechia
6Institute of Gravitation and Space Sciences, Bucharest, Romania
7Finnish Meteorological Institute, SF 00 101 Helsinki, Finland
8Faculty of Mathematics and Physics of Charles University, Praha, Czechia
9Space Science Department, ESA, Noordwijk, The Netherlands
10Special Design Division, Ukranian Academy of Sciences, Lviv, Ukraine
11Swedish Institute of Space Physics, Kiruna, Sweden
12Laboratory of Physics and Chemistry of the Environment, CNRS, Orleans, France
13Institute of Experimental Physics, Slovak Academy of Science, Kosice, Slovakia
14Centre d'Etude Spatiale des Rayonnements, Toulouse, France
15Demokritos University of Thrace, Xanthi, Greece
16Kiev University, Kiev, Ukraine

ABSTRACT

We present two examples of INTERBALL-1 data near both the high and low-latitude tail magnetopause (MP) under disturbed conditions. For the high-latitude case, MAGION-4 data determine the scales of the MP current sheets which are in the order of 100-500 km for the main ones, 50-200 km for Flux Transfer Events (FTEs) and a few km for the fine structures and ULF turbulence. The MP speed was 15-30 km/s. The energetic protons in the magnetosheath (MSH) provide evidence of reconnection upstream of the spacecraft (S/C). The tailward flows grow for the northward MSH magnetic field when the reconnection site is believed to be shifted tailward of the cusp. The inner boundary layer (BL) after the disturbance consists of tailward and earthward flowing plasma of MSH origin and cold mantle plasma flowing tailward The earthward flow is evidence of reconnection tailward of the S/C, which is regarded as a specific feature of the disturbed conditions. Local production of a plasma-sheet-like plasma at high latitudes is argued based on the inner BL plasma characteristics. The following features are observed in both cases: (a) FTEs for both northward and southward MSH fields; (b) waves in the current sheet vicinities over ten mV/m and 15 nT peak-to-peak; (c) electron fluxes with scales down to a few km with extra heating especially parallel to the magnetic field; (d) outer turbulent boundary layers with a deflected magnetic field; (e) ions with time-energy dispersion-like features and deflected ion fluxes. In the downstream dawn region at the transition between the low-latitude boundary layer and the plasma sheet (LLBL/PS), multiple MP encounters are observed. In the LLBL parallel electron intensifications correlate with ULF magnetic fluctuations.

INTRODUCTION

The second Inter-Agency Consultative Group (IACG) campaign is directed to the study of "Boundaries in Collisionless Plasmas" with an emphasis on the energy, mass, and momentum entering the magnetosphere. We present here some preliminary INTERBALL-1 results which could be used for the first stage of the campaign.

INTERBALL-1 satellite and MAGION-4 subsatellite (Galeev et al., 1995) have been operating in orbit since August 3, 1995. This article deals with two selected MP crossings from the first year of operation. The observations to be reported here were made by experiments on board the mother/daughter satellites (see INTERBALL: MISSION AND PAYLOAD, 1995 for details):

ASPI: Measurements of the fields and waves aboard the mother satellite,

CORALL: Measurements of the 3D ion distribution function in the energy range 50 eV - 25 keV (INTERBALL-1),

- VDP, VDP-S: Measurements by Faraday cups of ion/electron fluxes and integral energy spectra (only VDP, INTERBALL-1 and MAGION-4),

- ELECTRON: Measurements of 3D electron distribution 10eV-26keV (INTERBALL-1),

- SPS/MPS: Measurements of 2D electron and ion distribution 40eV-25keV (MAGION-4),

- PROMICS-3: Measurements of the 3D ion distribution function and masses in the energy range 4 eV - 70 keV (INTERBALL-1),

- DOK-2, DOK-S: Energetic particle spectrometers (INTERBALL-1 and MAGION-4).

Orbital characteristics of the INTERBALL-1 spacecraft permit studies of the magnetospheric boundary layers both at low latitudes (inbound crossings) and at high latitudes (outbound crossings). We will present two examples that are the most interesting, from our point of view, of magnetospheric boundary crossings.

INTERBALL-1 MAGNETOSHEATH CROSSING DURING THE MAGNETIC CLOUD AND MAGNETOSPHERE ENCOUNTER ON OCTOBER 18, 1995.

The first case we would like to present here is the October 18, 1995 crossing. INTERBALL-1 was pushed out of the magnetosphere by an unusually dense plasma after an interplanetary shock. A so-called "magnetic cloud" with very low density plasma resulted when INTERBALL-1 encountered the magnetosphere [Burlaga et al., 1996]. Details of the magnetic cloud encounter can be found elsewhere in this issue. We will concentrate on the boundary layer effects as a rather full set of satellite and subsatellite data are available and very strong disturbances are seen. INTERBALL-1 encountered the magnetosheath at about 18.19.30 UT, about 45 minutes after the solar wind (SW) dynamic pressure exceeded 15 nP at the WIND position [Burlaga et al., 1996]. The 45-minute delay corresponded to the SW velocity and the WIND position. The INTERBALL-1 GSE coordinates were (-17, -11, 13 Re), i.e. 3-4 Re deeper into the magnetosphere as compared to the average magnetopause position during quiet magnetic time. The MSH features are seen in both the particle data and in the magnetic field (Plate 1 and Figure 1). In Plate 1 dynamic spectrograms of CORALL ions (three top panels), SPS electrons (middle panel) and SPS/MPS ions (three bottom panels) are shown. At 18.10 UT both S/C particle experiments have no signals which correspond to the S/C location in the empty lobe. The MSH encounter is characterized by the appearance of ions and electrons at about 18.19.30 UT on all of the panels of Plate 1 excluding the bottom one. Ion flow is mostly tailward. (Maximum ion flow is seen on the top ion channels for both S/C.) Spin modulation in the mother S/C ion data is due to the sensor field of view (f.o.v.) centering in the direction of 30 degrees from the sun. Modulation in the MAGION-4 ion data is due to deflection of the spin axis (about 30 degrees) from the nominal direction; the I0 sensor (third from the bottom in Plate 1) is looking along the spin axis. Dense low energy MSH electrons appear in the middle panel of Plate 1 as well. The magnetic field in the MP boundary frame (see e.g. Russell, 1995) is presented in Figure 1. The total magnetic field Bt of about 40 nT inside the MP shows that the magnetosphere before the MSH encounter is compressed (as compared to the Bt typical for an empty magnetospheric lobe value of 25-30 nT).

Fig. 1. The FM-3 magnetic field on October 18, 1995. The MSH encounter by INTERBALL-1. From the top to bottom: Bn, Bm, and Bl in the magnetopause boundary coordinates frame, Bt - total magnetic field.

Plate 1. From the top to bottom:
- dynamic spectrogram of tailward flowing ions (INTERBALL-1, CORALL), f.o.v.- 30 (±10) degrees from the sun; 18.41-18.47 and 20.15-20.22 UT - data gaps;
- the same for the perpendicular to the sun direction ion channel (rotating),
- the same for the sunward flowing ions, f.o.v.- 150 (±10) degrees from the sun;
- MAGION-4 MPS/SPS device dynamic spectrogram of electrons, f.o.v.- 135 degrees from the sun;
- the same for the tailward flowing ions, f.o.v.- 0 degrees;
- the same for the perpendicular to the sun direction channel (rotating),
- the same for the sunward flowing ions, f.o.v.- 180 degrees.

One of the striking features of the MP crossing is the layer just outside the MP (18.20-18.40 UT) where Bz in the GSM/GSE frames is positive (not shown) but negative both inside the MP and in the rest of the MSH. The dominant component is Bx (B={-12, 0, 7}nT) which has an opposite sign to that of the internal field (B={26, 6, -3} nT in GSE). Magnetic field variations in the layer were so strong that they approached 80% in the total magnetic field Bt. Many magnetic field spikes are seen there in the normal component Bn (MP normal N is {0.2, -0.58, 0.8}). N was calculated as the vector product of the average magnetic field inside and outside the MP (i.e. before 18.19 UT and at 18.45-19.25 UT respectively). The Bn bi-polar spikes are rather similar to the classic FTEs being described by Russell and Elphic [1978]. We call the turbulent layers just outside the MP hereafter as `Outer Boundary Layers' (OBL). It can be argued that the OBL is not a temporary feature from the fact that a similar feature is observed in both the INTERBALL-1 satellite and the MAGION-4 subsatellite also just before the inbound MP at 19.35-19.51 UT (Figure 1 and Plate 1). Nevertheless, the preliminary comparison with the WIND IMF in Burlaga et al. [1996] shows that there was change of IMF Bz from north to south (or nearly zero) and back to north about 57 minutes before that of INTERBALL-1 (i.e. the magnetic time lag is 12 minutes longer than the lag of the dynamic pressure mentioned above). In Figure 1 these reorientations correspond to the Bm and Bl reorientations at about 18.42 and 19.35 UT respectively. So, it is difficult to judge which observed features of the OBL are temporary and which spatial. Further studies on this MP encounter require detailed comparison with the WIND and/or GEOTAIL magnetic field data.

INTERBALL-1 and MAGION-4 particle data inspections show that heated plasma inside and especially near the edges of the OBLs can be recognized. It is seen both from the intensification of the 0.3-1.5 keV electron spectra (fourth from the bottom panel of Plate 1 {18.20-18.43 and 19.30-19.52 UT}) and from that of the ion spectra in the tailward looking CORALL device channel (third from the top panel of Plate 1). Intensification of the 2-30 keV ion spectra can be also seen on two top panels. In those regions the average magnetic field Bt diminishes down to 5-10 nT (Figure 1) due to the diamagnetic effect of the heated plasma.

The energetic particles have a several times higher level in the OBLs as well (Figure 2). The electron intensity in the 21-95 keV range rise in both inbound and outbound OBL, having a several times higher maximum in the outbound OBL (upper panel in Figure 2). The same is true for the 45-133 keV protons measured by the rotating telescope (axis 60 degrees from the sun direction). See the bottom panel in Figure 2. The 21-27 keV protons in the OBLs have a higher minima in the spin modulated signal as compared to the rest of the MSH. The sunward flowing energetic protons (middle panel in Figure 2) have several times lower levels than the tailward protons. Their intensity rises only in the outbound OBL. The dominance of tailward energetic protons in the MSH implies that energetic particle leakage from the magnetosphere occurs upstream of the S/C. The leakage is more intense in the OBL and especially during northward Bz (18.28-18.42 and 19.36-19.53 UT). DOK-S aboard the subsatellite shows that energetic particles behave similarly.

Fig. 2. DOK-2 energetic particles data (INTERBALL-1 S/C) for 18.10-20.40 UT interval on October 18, 1995. The top panel shows sunward flowing electrons of 21-26 keV (upper curve), 39-48 keV and 76-95 keV (bottom curve). On the bottom: tailward flowing ions of 21-27 keV (upper curve), 45-59 keV and 101-133 keV (bottom curve), the vertical spikes once per spin (about 2 minutes) are due to the sunlight reflected from the Earth; and on the second from the bottom panel: the same as previous for sunward ions.

The energetic particle behavior has an explanation in terms of reconnection: for northward IMF the reconnection between magnetospheric and solar wind magnetic field lines should occur downstream of the cusp; for the southward Interplanetary Magnetic Field (IMF) it should be upstream of the cusp (see e.g. Russell, 1995) . In the latter case energetic particles leave the magnetosphere along reconnected lines at longer distances from the S/C. Their flux falls with distance due to particle diffusion and to field line divergence. (The latter is a consequence of the magnetosphere spread downstream of the dayside and of the plasma flow around the magnetosphere in the Y direction at some distance from noon.) For the same reason one can see less frequently the FTE-like Bn spikes in Figure 1 for the negative GSM Bz (18.42-19.32 UT). The difference between the inbound and outbound OBL is in a higher Bm (i.e. positive GSM By) in the former case. This orientation of the field implies that the maximum angle between the internal and the external field (i.e. favorable orientation for reconnection) was in the dusk mantle while the S/C was on the dawn side and that the maximum angle had a lower value. (Bm inside the magnetosphere was negligible. See Figure 1.) So, in this case reconnection should be less probable and reconnected lines should be driven preferably toward the dusk. As a consequence the energetic particles' flux should be lower and less regular. It rather agrees with Figure 2. We think that the distinct FTE-like features in the inbound OBL versus the turbulent Bn disturbances in the outbound OBL can be explained in the same manner as the energetic particle intensities relation.

In the MSH with southward Bz there is an intense spike of energetic electrons at 18.52 UT that corresponds to the most prominent FTE-like Bn spike during negative Bz. There the Bm component changes sign for a short period and the average Bt rises by 5 nT (Figure 1). The smaller electron spike at 19.01 UT correlates with the less intense double FTE-like features, while the following one at 19.08 UT has no FTE counterpart and two FTE-like signals at 19.12 and 19.14 UT have no related energetic electron spikes. In the MSH with a positive Bz OBL, the energetic electrons have poor correlation with the individual spikes in Bn. The only well-defined correlation between the FTE and energetic protons in the MSH is seen at 18.22 UT. Poor correlation of the FTE-like spikes with energetic electrons can be regarded as a lack of connection to the magnetosphere of the majority of the FTE in the MSH at the moments of the encounters. A high-turbulence level, accelerated particles and a deflected magnetic field (i.e. the `OBL' features) are often observed at high latitudes in INTERBALL-1 data. Similar phenomena were seen in PROGNOZ-8,10 data [Klimov et al., 1986; Savin, 1994], but the IMF dependence of these features hadn't been studied yet. In this case it is difficult to say if the high turbulence level (5-10 nT for most MSH crossings, see Figure 1) is a consequence of a very intense solar wind disturbance or that the S/C stayed inside the OBL during the entire MSH encounter and that the OBL properties depended on the IMF.

Plate 2 on the left side displays the electromagnetic turbulence spectra in the range of 0.1-32 Hz. Most intense electromagnetic wideband spikes correlate with the FTE-like Bn features in Figure 1 (i. e., at about 18.33, 18.40, 18.53, 19.13, 19,41, 19.43, 19.45, 19.48 UT etc.). There are also a number of electric field spikes without a pronounced magnetic counterpart (at 18.23.20, 18.25.15 UT and in the 19.51-20.07 UT interval) which are just inside the inbound MP. The electric field fluctuations displayed at the top left panel of Plate 2 have a maximum peak-to-peak amplitude of 10 mV/m, the magnetic ones have up to a 15 nT peak-to-peak. These values are high enough both for local particle acceleration/heating and for diffusion through the magnetopause. In the inner BL at 19.51-20.08 UT the electric field spectral amplitude is nearly the same as in the MSH. The most prominent bursts correspond to the outer boundaries of the magnetic depletion regions. (See Figures 1 and 4 and corresponding descriptions.) There are corresponding magnetic wave bursts with amplitude less than in the MSH. In previous papers intense waves just inside the MP were assumed to be electrostatic [Blecki et al. 1988, Belova et al., 1991, Cattell et al., 1995].

Plate 2. Left side: The MIF-M AC Bz (S/C frame) magnetic field FFT spectra of 0.1-32 Hz on October 18, 1995 (bottom panel), and that of OPERA Ex electric field (top panel).
Right side: The same for September 21, 1995.

In Figure 3 the magnetic field and ion flux (Faraday cup VDP(1) looking to the sun) from the main S/C are shown along with the ion flux (Faraday cup VDP-S(0) looking to the sun) from the subsatellite for the outbound MP crossing. At the first outbound MP at 18.19.40 UT no time lag between the S/C and subsat Faraday cup signals is seen. This is probably due to the substantial average Bn of 5 nT prior to the MP current sheet (such a large Bn implies that the MP surface is deflected from its average position) which seems to be parallel to the satellite-subsatellite separation vector. There are explicit MP crossing delays of 22 s and 15 s at about 18:23:05 and 18:25:40 UT); the MP motion along the average normal yields a MP speed Vmp = 18± 3 km/s for the outbound MP motion at 18.23.55 UT and Vmp = 26± 4 km/s for the MP inbound motion; the distance between the S/C along N being about 400 km. The case at about 18.24.10-18.25.15 UT seems to be similar to the latter one from VDP/VDP-S ion flow data, while the Bl behavior does not show clear MP out/in crossings. Nevertheless the outbound MP speed, deduced from Faraday cup lag in this case, is essentially the same; while the inbound speed is 1.5 times lower than in previous cases. Taking into account the complicated magnetic field pattern at 18.24-18.25 UT, we suppose that 26 km/s is the proper estimate of the inbound MP speed. The estimated speeds give the characteristic MP current layer scales in the order of 400-500 km at 18.19.40 UT, 270 km at 18.23.05 UT and 120 km at 18.25.40 UT. In the disturbances (e.g. at 18.20, 18.22.15 and 18.26.45 UT in Figure 3) and in FTE-like sites, the main current sheets had scales of 50-200 km. The total FTE width is estimated as 500-3000 km. It is comparable or less than the energetic proton gyroradius. The latter explains the lack of correlation of individual FTE-like spikes in the MSH and the energetic protons' intensity (Figures 1-3). The fine structures in Figure 3, have characteristic times of fractions of a second, which give the shortest scales down to a few km.

Fig. 3. The multiple outbound MP crossing on October 18, 1995. From the top to bottom: Bn, Bm, Bl in the magnetopause boundary coordinates frame, Bt - total magnetic field, VDP(1) (crosses, mother S/C) sunward looking Faraday cup and the nearest to sun direction VDP-S(0) (dots, MAGION-4) signals (the latter is saturated in MSH due to unusually high SW dynamic pressure), MSH encounters (shadowed) for both S/C inferred from VDP/VDP-S data.

The inbound MP crossing at about 19.51 UT is followed by the Bt drop and a quasi-periodic disturbance. The latter is seen mostly in Bl and Bt (Figure 1). The VDP/VDP-S data (not shown) yield about an 18 s delay which gives Vmp =18 ±3 km/s (for 330 km distance along N). In Plate 1 at 19.50-20.20 UT one can see heated electrons (fourth panel from the bottom). The most prominent spikes (at 19.52, 19.57, 19.58 UT) correlate with the Bt drops in Figures 1 and 4. Inspection of the ELECTRON device data (not shown) confirms the intensification of 0.1-1 keV electron fluxes at those times and at 20.03.30 UT in conjunction with the Bt drop while the subsatellite data does not show such a feature for the latter case (Plate 1.) In Figure 4 the Bt drops correlate with the appearance or intensification of the ion fluxes as well. So, the diamagnetic effect of the dense plasma is an explanation of these quasiperiodic magnetic structures. In Figure 2 there are four explicit maxima in the intensity of the energetic sunward flowing protons in the range of 22-60 keV and weak ones in the 101-122 keV range. (Tailward protons are measured by the rotating sensor and a short maxima can be affected by the rotation. See the bottom panel of Figure 2). The maxima are in a phase of different energies and directions. The energetic protons' behavior can be interpreted as a wave-like movement of plasma sites. The maximum amplitude of the movement is evaluated in the order of 101-132 keV protons gyroradius, i.e. 2500 km (as these protons are only slightly modulated). The minimum amplitude should be evaluated at 500 km, a reference to the visibility of the electron intensification at 20.03.30 UT only by the main S/C device. The quasi-periodic movement can be either due to a MP surface wave or to the Kelvin-Helmholtz instability of the inner edge of the flowing BL plasma. The surface wave could modulate the adjacent BL directly or via the injection rate modulation at the injection site.

Fig. 4. CORALL and FM-3I (INTERBALL-1) data for inbound MP crossing after the October 18, 1995 MSH encounter. From top to the bottom: - dynamic spectrogram of tailward flowing ions (CORALL), f.o.v.- 30 ±10 degrees from the sun- the same for the perpendicular to the sun direction channel (rotating), - the same for the sunward flowing ions, f.o.v.- 150 ±10 degrees from the sun;-Bx, By, Bz in the GSE frame (FM-3I); -Bt - total magnetic field.

The most striking features of the thermal plasma in the inner boundary layer at 19.51-20.40 UT are the fluctuating sunward ion flows which correlate with the Bt minima, the density just inside the MP being about ten times less than that in the MSH. The sunward component of the flow is better seen from a comparison of the tailward and sunward ions measured aboard the subsatellite (third from the bottom and bottom panels in Plate 1 respectively). CORALL ion channels show the same tendency (third from the top and top panels in Plate 1 and Figure 4). The PROMICS-3 data (not shown) confirms this conclusion. In the boundary layer at 19.51-20.10 UT the sunward ion flow dominates over the tailward flow; while in the MSH and in the rest of the inner BL, the tailward flow does. Noting that Bx is the main magnetic field component in the mantle, one can conclude that ion flows were injected tailward of the S/C and reached it mainly along field lines. The sunward field-aligned 200 eV electron injection is seen as well in the ELECTRON device data (not shown) at 19.56 UT. A 70 eV tailward injection occurs at 19.51 UT. At 19.51-19.57 electrons tend to be field aligned. We will discuss a mechanism of parallel electron acceleration by whistlers in the next section. At 20.20-20.40 UT tailward cold ions of, most probably, ionospheric origin are seen. This corresponds to the rather usual mantle encounter.

In the detailed data of the CORALL ion energetic spectra and the total magnetic field in Figure 4 just inside the MP at 19.52 UT, there is a burst of accelerated up to 10 keV ions with the time-of-flight energy/time signature. (The time delay of 10-20 s is much less than the spin period.) A substantial count rate is observed in the direction perpendicular (second from the top) to the sun channel. If the delay between the high and low-energy ion arrivals of 10-20 s is mostly due to the time of the ions' flight, the injection was in about 1-2 Re from the S/C, so that it should be the local tail MP event and the plasma of the MSH origin should be accelerated/heated several times in energy space during the injection. As the GSM S/C coordinates were (-18, -8, 15 Re), the injection in 2 Re could be explained neither by the near-cusp injection (which is more than 18 Re sunward from the S/C while the ions flow from the tail) nor by LLBL or PS injections. The latter are supposed to be more than 10 Re from the S/C and the plasma should be transported perpendicular to the magnetic field. For the rest the BL (and if the injection distance estimation is not valid at 19.51-19.57 UT), the PS or LLBL could be a source of the registered plasma. As at 19.35 UT magnetic field in the MSH turned northward and the IMF stayed northward for about 20 minutes. (See discussion on OBL above.) One could estimate the upward drift velocity for the PS plasma reaching the S/C during this time period - for a 10 Re distance to the PS the drift velocity should be in the order of 50 km/s with a respective cross-tail dusk-to-dawn electric field in the order of 2 mV/m. The northward magnetic field duration is nearly the same as the lag between the northward field turn and the BL plasma encounter (Plate 1). So, one should suppose that the PS origin plasma can be seen at the S/C position for a much shorter time period than 20 minutes as, after the southward IMF turning, the plasma should drift back toward the PS. This means that the PS plasma upward drift cannot account for the total BL encounter during almost 30 minutes.

To account for the sunward flowing BL by a PS origin, one should let some earthward injection occur in the PS (LLBL) in about a 40 Re distance tailward from the S/C (the distance to the PS multiplied by the ratio of sunward plasma speed to the vertical upward plasma drift). The northward field would reach 40 Re downstream in more than 15 minutes moving with MSH plasma speed. This is inconsistent with the estimated 50 km/s upward drift for the about 2 mV/m cross-tail electric field as then the time lag between the MSH Bz turn to the north and the PS plasma registration should be more than 35 minutes instead of 20. So, if the PS or LLBL could provide plasma into the observed BL during the northward IMF, at least the sunward flowing plasma at 19.51-20.10 UT should be of MSH origin being injected tailward of the S/C. As there is no substantial difference between the sunward and tailward flowing BL plasmas, we intend to regard the high-latitude MSH as the main source of the BL plasma in the case under study. The burst reconnection is one of the main candidates for causing the ion injection earthward at the high-latitude tail magnetopause. A brief (about five to ten minutes) excursion of the IMF to the positive Y values just prior to the southward turn [Burlaga et al., 1996] could be the particular reason for the reconnection bursts for this period (approximately 19.50-20.00 UT at the S/C position if the 57 minutes magnetic field lag with WIND holds. See text above.) All of the IMF components had an opposite sign to that of the mantle field. We suppose that the reconnection tailward of the S/C (X = -18 Re) is a feature of the strong disturbance interaction with the Earth's magnetosphere.

So, our data analysis in the inner boundary layer (at 19.51-20.30 UT) shows that the BL plasma is (at least partially) the depleted and locally accelerated/heated MSH plasma (with characteristic energies being several times higher than into the MSH). This plasma resembles that of the cold plasma sheet. It means that in the tail at large distances from the magnetic equator such a PS-like plasma can be locally produced during disturbed periods.

PLASMASHEET, LOW-LATITUDE BOUNDARY LAYER AND MAGNETOPAUSE CROSSINGS ON SEPTEMBER 21, 1995.

In August/September 1995 at the inbound leg of its orbits, INTERBALL-1 crossed the low-latitude magnetopause, the low-latitude boundary layer and the plasmasheet. The crossing on September 21, 1995 was chosen for this presentation because the predicted plasmasheet was observed along with unexpected magnetopause crossings. Unfortunately, the MAGION-4 data were not available at this time period.

The GSE S/C coordinates for the middle of the interval are (3.4, -12, -1.8) Re. Only the Z GSM coordinate differs substantially from its GSE counterpart, being equal to 2 Re. In Figure 5 the ASPI DC magnetic and electric field data together with the RMS plot of the plasma current oscillations in the range of 1-8 Hz for the interval 02:00-03:00 UT on September 21, 1995 are shown. The Fast Fourier Transform (FFT) colour electric and magnetic spectrograms for the same period are shown at the right side of Plate 2. The magnetosphere was compressed (the total magnetic field value being about 1.5 times higher than that predicted by the magnetospheric model) and the observed By component is also smaller then that expected from the model. In the middle of Figure 5 at 02:28 and 02:33 UT Bt and Bz drops down to 3-10 nT which implies MSH encounters. At 02.38 UT we believe that only the near-magnetopause turbulent layer was encountered for a time period of about two minutes. (See also Plate 3.) Only the event at 02:33 UT is long enough to determine the value of the magnetosheath DC electric field from the signal measured with the long dipole antenna which is modulated by the spacecraft spin. The Ey GSE (Figure 5) corresponds to an antisolar plasma velocity of 350 km/s which is a rather typical value for the MSH. In the PS, Ey GSE is negative which implies that the sunward plasma speed is about 100 km/s. This observation is supported by the northward direction of the MSH field. The GSE By is practically aligned with the average MP normal, so the bi-polar By event which is observed at 02:32 is very similar to the FTE-like events in Figures 1 and 3.

Fig. 5. From the top to bottom: Bx, By, Bz in the GSE frame, Bt - total magnetic field for 02-03 UT on September 21, 1995, DC electric field Ey in GSE frame, 1-8 Hz RMS split probe signal.

Plate 3. PROMICS-3 and ELECTRON data for the 02-03 UT interval on September 21, 1995, from the top to bottom: - tailward flowing ions, f.o.v. - 10 degrees from the sun
- the same for the perpendicular to the sun direction ion channel (rotating),
- the same for the sunward flowing ions, f.o.v.- 170 degrees.
- tailward flowing electrons; f.o.v. 0-22 degrees from the sun;
- the same for the perpendicular to the sun direction electron channel (rotating),
- the same for the sunward flowing electrons, f.o.v. 158-180 degrees.

Similar to the high-latitude magnetopause observations in the previous case, the magnetosheath and magnetopause turbulent layer encounters are accompanied by intense electromagnetic noise (Plate 2). The most intense wave bursts correlate with the sharp changes of the magnetic field. The highest amplitude magnetic field burst in Plate 2 (up to 10 nT peak-to-peak) is seen exactly at the time interval where the DC field sharply rotates (02.32.08 UT). The frequency of the dominant signal is about 0.8-1.5 Hz. The ion cyclotron frequency is about 0.3 Hz. The most intense electric field bursts (up to 10 mV/m peak-to-peak) are registered just inside the magnetopause. One component, electrostatic waves with a frequency of a few Hz and amplitude of about 10 mV/m just inside the high-latitude dayside magnetopause, has been reported earlier by Vaisberg et al. [1983], Klimov et al. [1986], Blecki et al. [1988] and Belova et al. [1991]. Similar observations at the low-latitude magnetopause have been described by Cattell et al. [1995]. The Ex spectra at the MP and FTE have an intense 1-10 Hz component, while in the rest of the magnetosphere the Ex waveform sensitivity is too low. The magnetic field channel, in contrast, records the 1-10 Hz waves outside the DC magnetic field disturbances as well. The comparison with the plasma current RMS fluctuations (measured by the split current probe, bottom panel of Figure 5) shows that at 04:20-04:45 UT the spikes in the current signal are similar to that of Bz (Plate 2); but at 04.06, 04.09, 04.14, 04.49, 04.56-05.00 UT, there are intense events which have no counterparts in the magnetic dynamic spectra in Plate 2. We suppose that at these moments electrostatic waves are recorded. (The split probe have better sensitivity than the electric field waveform.) Supposing the speed of the MP about 20 km (cf. with the previous case), one could get the scales for the currents in the range of 1-8 Hz of about 3-20 km. It is compatible with the shortest scales near the high-latitude MP on October 18, 1995.

Let's now compare the field and wave data with particle measurements. In Plate 3 the ELECTRON experiment data are shown. At 02.32- 02.33 UT the typical isotropic MSH electrons are seen. The PS ions practically disappear in the PROMICS-3 data (Plate 3) with an exception of the tailward low-energy ions and the beam of about 10 keV. CORALL ion data (not shown) confirms the MSH ions encounter. Both magnetospheric energetic ions and electrons are not seen in the MSH in the DOK-2 data (not shown). At about 02.27 UT the lower energies of the MSH-like electrons are depleted as compared to the 02.32-02.33 UT case (Plate 3) which indicates that, presumably, the MSH particles' injection inside the magnetosphere occurs. Indeed, the bi-polar signature is seen in the By component which is close to the average MP normal direction which is the FTE feature (Figure 5). The PROMICS-3 ion data shows only PS ion depletion and anisotropy in contrast with the observation at 02.33, where they totally disappeared and the high energy part of the MSH ion distribution is seen. In the next strong DC magnetic and electric field disturbance at about 02.37 UT, very few MSH electrons could be found; the PS ions are depleted for a shorter time as compared with the latter case. The electrostatic waves inside the event have maximum level (Figure 5, Plate 2), which is on the same order as that of the MP in/out crossings at 02:32/02:34 UT. The DC magnetic field reveals a complex structure (Figure 5) with possibly multiple FTE-like signatures.

In Plate 3, just outside the MSH-like electron regions (at about 02.27, 02.33 and 02.37 UT), one could see intense accelerated MSH origin electrons (AME, cf. electrons in the BL at 19.50-20.10 UT in Plate 1 on October 18, 1995, in the previous section). The modulation of the measured electron flux with a rate of twice the spin period indicates that the electron flux is anisotropic. Its maximum values are observed being close to the magnetic field. The interval prior to 02.26 UT we identify as the LLBL from the ELECTRON data in Plate 3. In this interval the dominant electrons are field-aligned AME (30-500 eV) with inserts of PS-like electrons, which are mostly perpendicular to the magnetic field with the average energy higher than in the normal PS (at 02:38-03:00 UT). The intensity of the AME has a maximum in the BLs and FTEs (as it was pointed out earlier). Hence it is argued that these regions are the source of the LLBL field-aligned electrons. The electrons from the source regions can easily go away only along field lines. If this is so, we do need more detailed study of the pitch-angle distribution in the AME source regions. The boundaries in the case under study are too sharp compared with 3D electron distribution function time of measurements by the ELECTRON device (spin period), so statistical study and/or combination with the VDP data in electron mode will be needed for further studies.

At 02:12-02:20 UT the electron spectra intensities along and perpendicular to the magnetic field are comparable; the energy of the perpendicular electrons is much higher, approaching the PS energy. (See the middle ELECTRON panel in Plate 3.) We call this kind of distribution the "hybrid electron distribution" (HED). The AME distribution function seems to transfer rather continuously into the HED distribution. The HED could be a source of the PS electrons, which became more isotropic and energetic while travelling tailward and back along field lines in the PS. It could be the case that the AME from the LLBL (i.e. at 02:00-02:26 UT) have no permanent access to the tail PS due to the peculiarity in the magnetic field topology (i.e.. current layer(s), potential barriers, etc.) at the transition of a dipole-like field into the tail field. The disturbances at 02:26-02:39 UT described here are close to this transition. To clarify the reason for the MP crossings, comparison with the WIND data is needed.

Comparison of the ion PROMICS-3 data in the LLBL and the PS (02.00-02.26 and 02.38-3.00 UT respectively, see Plate 3) does not show dramatic differences as in the case of the electrons. The smaller intensity and higher variability of the ions in the LLBL could only be outlined here. In the vicinity of the MP in/outbound at 02.32/02.34 UT and at about 02.38 UT (cf. magnetic field and wave disturbances in Figure 5 and Plate 2) the accelerated ions were observed along with the time-of-flight-like ion spectra features. The latter are the MSH plasma injection signatures, which can be explained by either burst reconnections or impulsive penetration of the MSH plasma filaments with excess momentum [Lemaire, 1977].

In the PS ion regions (Plates 2 and 3, excluding 02:26-02:39 UT), we would like to point out the spectra of the electromagnetic waves. Similar waves were found as being characteristic for the plasma sheet boundaries using the data of previous missions (Blecki et al., 1994 and references therein). However, most of the previous results were based only on the electric field data. At 02:05-02:12 UT the modulation of the Ex and Bz spectra could be seen in Plate 2. The DC magnetic field components at this time are also fluctuating within the period 1.5-2.5 minutes (Figure 5). The absence of such oscillation in the plot of the total magnetic field value would imply either an Alfven wave or a surface wave on the magnetopause.

In the LLBL the presence of the 0.1-2 Hz electromagnetic waves correlates with the AME and the amplitude correlates with the AME intensity. The most intense wave bursts up to 30 Hz correspond to the transition from the AME to HED (at 02:23, 02:28, 02:26, 02:20, 02:13, 02:08 and 02:02 UT); as a rule, there are inhomogeneities and accelerated PS ions at these moments (Plate 3). At 02:15 UT the burst of sunward accelerated ions (up to 20 keV) is accompanied by weak waves of 0.1-1.5 Hz without a substantial disturbance in the electron distribution function. In the latter case the split probe signal rises. (See the lower panel in Figure 5.) In the other cases the split probe signal has spikes as well. The intensity correlation with the electron spectra seems to be poorer.

In the PS at 02:38:30-02:42:30 UT intense waves similar to the LLBL are in the region of the electron energy and anisotropy changes. Anisotropically accelerated ions with energies higher than 15 keV are also present there. At 02:42:30 UT the wave burst up to 20 Hz corresponds to the sharp gradient in the electron energy (to the lower energy electrons insert). The less intense electron energy gradient at 02:43 UT correlates with the weaker waves and, possibly, with the tailward ion bulk flow and/or accelerated ions. Rather strong waves are seen at 02:49 UT in the region where the electron energy drops down and anisotropy rises, ion gradients and accelerated ions are also seen. Very similar ion disturbances at 02:38-03.00 UT are accompanied by wave emissions at frequencies below 2 Hz. However, there is no evidence of any electron gradients observed simultaneously with these electromagnetic waves. The split probe has the signal risings in conjunction with the electron and ion distribution changes in the PS region as well (Figure 5 and Plate 3).

So, the main-wave activity in the LLBL and the PS in the 1-20 Hz frequency band correlates with the electron distribution function changes. The waves observed at frequencies below 2 Hz well correlate with the peculiarities of the ion distribution. However, further study is required to find out the origin of these lower frequency waves. Both ion and electron distributions reveal certain peculiarities which could lead to the wave generation, and vice versa, high amplitude waves are able to accelerate/heat the MSH origin particles.

We think that the measured high-level ELF turbulence near the MP heats the MSH electrons to the AME-like distribution as it was proposed by Blecki et al. [1987]. The theory of acceleration along the magnetic field by the waves in the low-hybrid range (up to 25 Hz for our cases) was developed in Vaisberg et al. [1983a]. They predict that the accelerated electron energies are proportional to the electric field amplitude to the power 4/5 and to the magnetic field to the power minus 2/5. For a magnetic field of 5 nT and the electric field of 1 mV/m, they had the characteristic electron energy of 100 eV. In our case, for 20-30 nT and 5 mV/m we would estimate an energy of 200 eV, which is in satisfactory agreement with the AME data in Plate 3.

The counterpart of the OBL could be found in Figure 5 and Plate 2 at about 02:32-02:34 UT; while due to very fast low-latitude MP crossings, the only Bz deflection in the OBL prior to the trailing MP could be unambiguously distinguished (Figure 5) and wave turbulence in Plate 2 is certainly higher close to the MP. The latter holds just inside the MP as well. But in this case at low latitudes we have rather dense plasma from both sides of the MP.

An example of the low-latitude magnetospheric boundary crossing described here is rather unusual for such a spacecraft position. It demonstrates a substantial ELF wave activity both in the relatively quiet and in the highly disturbed regions. The high-amplitude electric and magnetic oscillations at a frequency in the order of 1 Hz observed near the magnetopause transition are of a particular interest. Their amplitudes can reach 10 nT and 5 mV/m respectively. We have confirmed the existence of the ELF electric oscillations in the plasma sheet and measured their magnetic component. The described crossing can be a subject for further in-depth studies on the multi-instrument basis when all INTERBALL-1 experiments were operating in fast modes. The WIND magnetic and plasma data are needed for an understanding of the SW disturbances nature and their interaction with the magnetosphere close to the dipole to tail-like magnetic-field transition.

CONCLUSIONS

We presented here two crossings of high and low-latitude GEOTAIL boundaries under disturbed conditions. The main comparative features could be summarized as follows.

The "outer boundary layer" just outside the MP in high latitudes contains a deflected magnetic field, high ELF turbulence and accelerated/heated MSH particles. The energetic particles flow downstream in the MSH and grow inside warm plasma sites. The magnetic component of ELF noise near the current sheets is substantially suppressed inside the MP in the low density BL, which is compatible with the drift current waves predictions for a low-beta situation into the high-latitude magnetosphere (i.e. Vaisberg et al. [1983] or Cattell et al. [1995] and references therein). As for the low-latitude BL the magnetic field deflection and turbulence is seen from both MP sides; energetic particles have maxima just inside the MP, dropping outside it. So, in contrast to the disturbed high-latitude case, no evidence of magnetospheric energetic particle leakage is seen. The amplitude of an electric component of the ULF waves tends to have a higher level inside the MP as well (i.e. for lower plasma beta).

Wave-like disturbances in the inner BL within a one to five minute period are registered in both cases. At the high latitudes two-point particle data are compatible with the plasma sites or flows having quasiperiodic movements with amplitudes of 500-3000 km. The MP crossings seem to have a periodicity of five to ten minutes in both cases (bottom panels of Figure 3 and Plate 3). It is an open question if the latter periodicities are driven by SW disturbances or by MP surface waves (produced by step-like SW disturbances or by Kelvin-Helmholtz instabilities).

The accelerated and deflected ions (which reveal sometimes time-energy dispersion-like features) could be seen in both cases presented in our paper. In the nearly antiparallel magnetic field configuration across the main current layer (first case), the burst reconnection is one of the main candidates to cause the ion deflection. The FTE-like feature observations are present in both the high and low-latitude data sets. For the low-latitude case no evidence of reconnection downstream of the S/C is found, while at high latitudes the sunward ion flows seem to prove the occurrence of such reconnection.

The two S/C measurements in the first case provide the estimate of the MP velocity value of about 20 km/s. It gives the scale of the MP current layers in the order of 100-500 km for the main current sheets, 50-200 km for FTEs currents, and a few km for the fine turbulent structures. The latter corresponds to several electron inertial lengths. Such scales were predicted by many theoretical works and simulations. From the magnetic field and plasma data similarities, one can estimate the characteristic scales at low-latitude BLs being of the same order as at the high-latitudes BLs.

ACKNOWLEDGEMENTS

We are grateful to C. T. Russell for helpful discussions. We thank the referees for many valuable remarks on the paper improvement. Work was supported by INTAS through grant INTAS-93-2031.

REFERENCES

Belova, E.V., J. Blecki, M. Denis, L.M. Zelenyi, and S.P. Savin, Excitation of ion cyclotron waves at the boundary of the magnetosphere, Sov. J. Plasma Phys., v. 17, 555 (1991).

Blecki, J., K. Kossacki, B. Popielawska, S.I. Klimov, S.A. Romanov, S.P. Savin, L.M. Zelenyi. ELF plasma waves associated with plasma jets near the Earth magnetopause as observed by PROGNOZ-8. Physica Scripta, v.37, 623-631 (1988).

Blecki, J., B. Nikutowski, P. Koperski, B. Popielawska, K. Kossacki, S.A. Romanov, S.P. Savin, ELF plasma waves in the hot and cold plasma fluxes observed by PROGNOZ-8 in the magnetospheric tail, Ann. Geophys., v. 12, 1006, (1994).

Blecki, J., K. Kossacki, R. Wronowski, D. Zwolakowska, S.I. Klimov, S.P. Savin, L.J.C. Woolliscroft. The distribution of the small scale structures observed by PROGNOZ-8 and AMPTE-UKS satellites in the vicinity of the Earth's bow shock and magnetopause, in 2nd National Symposium PLASMA'95 "Research and applications of plasmas". Warsaw, June 26-28, Vol.1.Contributed Papers, p.245-248 (1995).

Burlaga, L.F., R.P. Lepping, W. Mish, K.W. Ogivie, A. Szabo, A.J. Lazarus, J.T. Steinberg, A magnetic cloud observed by WIND on October 18-20, 1995. NASA preprint, GSFC, Laboratory for Extraterrestrial Physics, February (1996).

Cattell, C., J. Wygant, F.S. Moser, T. Okada, K. Tsuruda, S. Kokubun, T. Yamamoto, ISEE-1 and GEOTAIL observations of low-frequency waves at the magnetosphere, J. Geophys. Res., v. 100, A7, 11823, (1995).

Galeev, A.A., Yu.I. Galperin, L.M. Zelenyi, The INTERBALL Project to study solar-terrestrial physics, in INTERBALL Mission and Payload, CNES-IKI-RSA, pp.11-42 (1995).

Klimov, S.I., M.N. Nozdrachev, P. Triska, Ya. Vojta, A.A. Galeev, Ya.N. Aleksevich, Yu.V. Afanasiev, V.E. Baskakov, Yu.N. Bobkov, R.B. Dunetz, A.M. Zhdanov, V.E. Korepanov, S.A. Romanov, S.P. Savin, A.Yu. Sokolov, and V.S. Shmelev, Investigation of plasma waves by combined wave diagnostic device BUDWAR (PROGNOZ-10-INTERCOSMOS), Cosmic Research (Transl. from Russian), v. 24, 177, (1986).

Klimov, S.I. et al., ASPI experiment: Measurements of fields and waves onboard of the INTERBALL-TAIL mission, in INTERBALL mission and payload, IKI-CNES, p.128 (1995).

Lemaire, J., Impulsuve penetration of filamentary plasma elements into magnetosphere of the Earth anf Jupiter., Planet. Space Sci., v. 25, p. 887, (1977).

Russell, C.T., and R.C. Elphic, Initial ISEE magnetometer results: magnetopause observations, Space Sci. Rev., 22, 681 (1978).

Russell, C.T., The structure of the magnetopause, in Physics of the Magnetopause, edited by P.Song., B.U.O. Sonnerup, M.F. Thomsen, pp.81-98, American Geophysical Union, Washington, D.C., (1995).

Savin, S.P., ELF waves near the high-latitude magnetopause, in Abstracts of AGU Chapman conference on physics of the magnetopause, March 14-18, 41 (1994).

Vaisberg, O.L., A.A. Galeev, L.M. Zelenyi, G.N. Zastenker, A.N. Omelchenko, S.I. Klimov, S.P. Savin, Yu.I. Yermolaev, V.N. Smirnov, M.N. Nozdrachev., Fine structure of the magnetopause from the measurements of the satellites PROGNOZ-7 and PROGNOZ-8. Cosmic Research (Transl. from Russian), v. 21, 49, (1983).

Vaisberg, O.L., A.A. Galeev, G.N. Zastenker, S.I. Klimov, M.N. Nozdrachev, R.Z. Sagdeev, V.Shapiro, Electron acceleration at the front of strong collisionless shock waves. Sov. Phys. JETP (Engl. Transl.), v. 58, 716, (1983a).