Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://solarwind.cosmos.ru/pdf/sg.pdf Äàòà èçìåíåíèÿ: Wed Nov 23 15:46:55 2005 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 19:57:22 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: êàñêàä êåìáëà

Surveys in Geophysics (2005) 26: 95­133 DOI 10.1007/s10712-005-1874-4

ñ Springer 2005

MAGNETOSHEATH INTERACTION WITH THE HIGH LATITUDE MAGNETOPAUSE
S. SAVIN1, A. SKALSKY1, L. ZELENYI1, L. AVANOV1, N. BORODKOVA1, S. KLIMOV1, V. LUTSENKO1, E. PANOV1, S. ROMANOV1, V. SMIRNOV1, YU. YERMOLAEV1, P. SONG2, E. AMATA3, G. CONSOLINI3, T. A. FRITZ4, J. BUECHNER5, B. NIKUTOWSKI5, J. BLECKI6, C. FARRUGIA7, N. MAYNARD8, J. PICKETT9, J. A. SAUVAUD10, J. L. RAUCH11, J. G. TROTIGNON11, Y. KHOTYAINTSEV12 and K. STASIEWICZ12
1

Space Research Institute, Profsoyuznaya 84/32, Moscow, 117810, Russia E-mail: ssavin@iki.rssi.ru; 2 University of Massachusetts, Lowell, MA, USA; 3 Interplanetary Space Physics Institute, CNR, Roma, Italy; 4 Boston University, Boston, MA, USA; 5 Max-Planck Inst, Aeronomie, Katlenburg-Lindau, Germany; 6 Space Research Center, Warsaw, Poland; 7 University of New Hampshire, Durham, NH, USA; 8 Mission Research Corporation, Nashua, NH, USA; 9 University of Iowa, Iowa City, IA, USA; 10 CESR, Toulouse, France; 11 LPCE, Orleans, France; 12 IRF-U, Uppsala, Sweden

(Received 9 February 2003; Accepted 30 April 2004)

Abstract. We present both statistical and case studies of magnetosheath interaction with the high-latitude magnetopause on the basis of Interball-1 and other ISTP spacecraft data. We discuss those data along with recently published results on the topology of cusp-magnetosheath transition and the roles of nonlinear disturbances in mass and energy transfer across the high-latitude magnetopause. For sunward dipole tilts, a cusp throat is magnetically open for direct interaction with the incident flow that results in the creation of a turbulent boundary layer (TBL) over an indented magnetopause and downstream of the cusp. For antisunward tilts, the cusp throat is closed by a smooth magnetopause; demagnetized `plasma balls' (with scale ~ few RE, an occurrence rate of ~25% and trapped energetic particles) present a major magnetosheath plasma channel just inside the cusp. The flow interacts with the `plasma balls' via reflected waves, which trigger a chaotization of up to 40% of the upstream kinetic energy. These waves propagate upstream of the TBL and initiate amplification of the existing magnetosheath waves and their cascade-like decays during downstream passage throughout the TBL. The most striking feature of the nonlinear interaction is the appearance of magnetosonic jets, accelerated up to an Alfvenic Mach number of 3. The characteristic impulsive local momentum loss is followed by decelerated Alfvenic flows and modulated by the TBL waves; momentum balance is conserved only on time scales of the Alfvenic flows (1/fA~12 min). Wave trains at fA~1.3 mHz are capable of synchronizing interactions throughout the outer and inner boundary layers. The sonic/Alfvenic flows, bounded by current sheets, control the TBL


96

S. SAVIN ET AL.

spectral shape and result in non-Gaussian statistical characteristics of the disturbances, indicating the fluctuation intermittency. We suggest that the multi-scale TBL processes play at least a comparable role to that of macro-reconnection (remote from or in the cusp) in solar wind energy transformation and population of the magnetosphere by the magnetosheath plasma. Secondary micro-reconnection constitutes a necessary chain at the small-scale (~ion gyroradius) edge of the TBL cascades. The thick TBL transforms the flow energy, including deceleration and heating of the flow in the open throat, `plasma ball' and the region downstream of the cusp. Keywords: boundary layer, cusp, magnetopause, magnetosheath Abbreviations: BL ­ Boundary Layer; FOV ­ Field Of View; MP ­ Magnetopause; GSM ­ Geocentric Solar Magnetic; MSH ­ Magnetosheath; SW ­ Solar Wind; ULF ­ Ultra-Low Frequency; UT ­ Universal Time; MLT ­ Magnetic Local Time; HEOS ­ Highly Eccentric Orbiting Satellite; ISEE ­ International Sun-Earth Explorers; GDCF ­ Gas Dynamic Convected Field model; FOV ­ Field Of View; GSE ­ Geocentric Solar Ecliptic

1. Introduction Early single spacecraft observations with Heos-2 and later Prognoz-7, 8, 10 have shown that the magnetopause (MP) position and magnetosheath plasma flow structures are quite variable near the cusp, a magnetospheric region that is crucial for magnetosheath plasma entry (Haerendel and Paschmann, 1975; Paschmann et al., 1976; Klimov et al., 1986; Savin, 1994). Haerendel (1978) first introduced the turbulent boundary layer (TBL) to cusp physics in a discussion on the interaction of the magnetosheath flow with the magnetopause at the flank of the tail lobe. We reproduce his TBL sketch in Figure 1a: a laminar hydrodynamic flow interacts with an obstacle by generation of a TBL both in front of the obstacle (marked by ``1'') and behind it (marked by ``2''). The zone ``1'' corresponds to the funnel-shaped cusp throat in Figure 1b; the obstacle is presented by uprising magnetic field tubes at the tailward cusp wall. The downstream zone ``2'' has been poorly studied (see Savin et al., 2004, and references therein). Because of differences in characteristics, researchers have divided the high altitude cusp into a number of layers and regions. Since full agreement in terminology is not yet achieved, we provide our definitions of the regions discussed in this paper. These regions, as shown in Figure 1b, are the outer and inner cusps, the open throat (OT) of the outer cusp, and the turbulent boundary layer. We will demonstrate, however, that the interaction pattern in Figure 1 should be further modified for winter cusp crossings. In Figure 1b, the OT (slant-line shaded region) is outside the MP, the outer cusp (gray) is just inside the MP, and the inner cusp (black) is deeper in the magnetosphere. We identify the MP (inner white line) as the innermost current sheet where the magnetic field turns from Earth-controlled to magnetosheath-controlled (Haerendel and Paschmann, 1975). The outer cusp (OC) is a region with three different particle populations: newly injected


THE HIGH LATITUDE MAGNETOPAUSE

97

Figure 1. (a) Generation of a turbulent boundary layer in the process of interaction of hydrodynamic flow with an obstacle (from Haerendel, 1978). ``1'' ­ marks open cusp throat, ``2'' ­ stands for high latitude boundary layer downstream of the cusp. (b) Sketch for MSH/ cusp interface in the noon-midnight plane from (Savin et al., 2002a). The boundaries and subregions are described in the text.

MSH ions, MSH ions reflected from the ionosphere, and quasi-perpendicular ions trapped in the local magnetic field minimum near the cusp (Savin et al., 1998b; Sandahl et al., 2002). There are also electrons accelerated along the field lines. The newly injected and quasi-perpendicular ions dominate over those that are reflected. This is one of the characteristics distinguishing the outer cusp from the inner cusp and from the distant mantle. The outer cusp is also characterized by moderate magnetic noise, while in the inner cusp (IC) there is a similar type of noise observed primarily only at the boundaries (Pottelette at al., 1990). The outer cusp consists of the entry layer and the portion of the plasma mantle adjoining the entry layer (Paschmann et al., 1976). According to the work of Yamauchi and Lundin (1997) the entry layer and mantle that are parts of the outer cusp form one continuous region. At the cusp the magnetopause can be indented. This indentation was first predicted by Spreiter and Briggs (1962) and then detected by HEOS-2 (Paschmann et al., 1976), ISEE (Petrinec and Russell, 1995), and Hawkeye-1 (S. Chen et al., 1997). Interball-1 early statistics show that the indentation is on the average about 2 RE deep (Savin et al., 1998b). We call this part of the exterior cusp the OT. The plasma in the OT is highly disturbed and/or stagnant MSH plasma. The turbulent boundary layer (TBL) is a region dominated by irregular magnetic fields and plasma flows. It is located just outside and/or at the near cusp magnetopause and has recently been found to be a permanent feature (Savin et al., 1997, 1998b, 2002a; Sandahl et al., 2002). Here the energy density of the ultra low frequency (ULF) fluctuations is comparable to the ion kinetic, thermal, and DC magnetic field densities. The ULF power is usually several times larger than that in the MSH, and one


98

S. SAVIN ET AL.

or two orders of magnitude larger than that inside the magnetopause. As recent studies conclude (see, e.g., Belmont and Rezeau, 2001, and references therein), the strong ULF fluctuations that occur just outside of or at the magnetopause can independently result in micro-reconnection and local plasma penetration all along the magnetopause surface, even without the presence of quasi-stationary global reconnection. Examples of highly turbulent magnetic and electric fields in the exterior cusp have been reported by Paschmann et al. (1976) from Heos-2, by Klimov et al. (1986) from Prognoz10, by Savin (1994) and Blecki et al. (1998) from Prognoz-8 and by S. Chen et al. (1997) from Hawkeye-1 data. The main goal of this paper is to survey achievements in this area and explore solutions to the problems associated with the TBL and exterior cusp physics in the Interball era. The recent baseline case studies are described in detail in Savin et al. (2001, 2002a, b, 2004). An Interball-Polar case on 19 June 1998 is utilized to display the characteristic TBL features. It demonstrates the asymmetry of boundary layers for positive (sunward) Earth magnetic dipole tilts in summer and that of the negative (anti-sunward) tilts in winter. We reproduce the most interesting results from the previous studies and analyze detailed dynamics of the ion energy and of Poynting flux to clarify the pattern of nonlinear interactions in the upstream TBL. The wave packets, going from MP upstream in a subsonic MSH flow, occur to stimulate partial randomization of the flow far in front of the MP, while the SW driver plays a minor role. The interaction with the upstream going waves launches downstream-accelerated jets at about sonic speed. The jets break up the homogeneous equilibrium streamlining by carrying down-flow up to half of the flow momentum density. That signifies the cascade-like non-linear energy transformation in the TBL, proposed by Savin et al. (2001). We present, for the first time, a full statistical review of the high level ULF magnetic turbulence (i.e. of the TBL) from the Interball-1 data, concentrating on the MP asymmetry for the summer and winter hemispheres. The permanent plasma heating in the TBL is regarded as a result of transformation of MSH flow energy into the random and thermal energies in the process of the MSH flow interaction with the outer cusp throat. We exhibit de-magnetized large-scale `plasma balls' inside the winter MP and study their statistics versus that of stagnant MSH plasma outside the NP in summer. We also present 3D maps of the TBL dependence on the fluctuation power and on the dipole tilt and study the `plasma ball' occurrence that depends on the tilt, magnetic shear and interplanetary magnetic field. Finally, we discuss the presented and published Interball-1 data in relation to the MSH plasma penetration and acceleration both due to plasma percolation and turbulent heating and due to multi-scale reconnection of anti-parallel magnetic fields. We address the practically unexplored interaction of MSH flow with stagnant high-beta `plasma ball' via the highly super-Alfvenic jets and decelerated


THE HIGH LATITUDE MAGNETOPAUSE

99

Alfven flows embedded into the coherent pattern of cascade-like interactions in the upstream TBL.

2. Turbulent boundary layer and ``plasma ball'' on 19 June 1998 Data from 19 June 1998 (Figure 2) illustrate the recent findings at the MSH/ cusp interface. Geotail provided high-resolution SW magnetic field and synchronization with WIND plasma data, Polar traced the northern (summer) stagnation region, and Interball-1 entered the southern cusp from the MSH. The Gas Dynamic Convection model (GDCF), which is described in detail by Song et al. (1999a, b) and Dubinin et al. (2002), links the multipoint observations. While several recent papers describe some features of this event (Savin et al., 2002a, b, 2004), we report new, valuable findings in the interaction pattern. 2.1. Inbound magnetopause crossing We present a sketch in Figure 3 (turned upside down for easy comparison with Figure 1b) to provide a guide for a different topology on 19 June 1998 compared with that in Figure 1. The southern hemisphere MP here has no indentation. The disturbances of the ion flow in the XZ plane (`outer BL') start in the upstream (relative to MP) flow and result in the appearance of the accelerated jets first at ~09 UT and then in front of the MP in the tailward stream (upstream TBL). The criteria fulfilling the definition of the TBL (see Section 3) are marked by gray bars on the bottom. In the TBL map in Figure 2.A1 this case appears as a 20 min interval, centered at ~09 UT, with the interval at 09:20­10:40 UT colored according to the Df magnitude (black curve in Figure 2.A6i, see discussion below). The MP transition is thick and imbedded into the TBL, which terminates in a `plasma ball' (PB, see Savin et al., 2002a). The PB is a high-beta (in this case up to 15) large-scale (few RE) sub-region of the OC (Figure 3). As Savin et al. (1998b) and Kirpichev et al. (1999) have shown, the general OC (and PB) feature trapped MSH-origin ions ­ are often seen in low-beta and small-scale regions (i.e. the PB occurrence is much smaller than that of the OC, see Section 4). In the PB the magnetic field is reduced, marking the boundary between the dayside and mantle/polar cap magnetic field lines. The PB average position should tend to shift towards the anti-parallel magnetic field through the MP that corresponds to the place of minimal B2/2l0, predicted by the fields' vector sum (cf. Spreiter and Briggs, 1962). We will refer hereafter to the OT only in the case of the configuration of Figure 1b, which seems to be characteristic for positive tilts (see Figure 2.A2­2.A4 and related discussions in Sections 3 and 4).


100

S. SAVIN ET AL.

Figure 2. (A1) TBL level map (nT) in 3 GSM planes (gray color ­ D(Bx)>3 nT). (A2) TBL tilt dependence for D(Bx)>8 nT. (A3) PB tilt (degrees) dependence (gray color ­ TBL). (A4) Tilt dependence of PB (blue, green) and open OT. (A5) SW By spectrogram (Geotail) on 19 June 1998. (A6) TBL and PB on 19 June 1998. (a) energy densities; (b) normal electric field and Poynting vector; (c) magnetic clock angle; (d) |B|; (e) ion Vx velocity; (f) ion temperature; (g) electron spectrogram and intensity (>30 keV); (h) ion spectrogram and intensity (>30 keV); (i) magnetic By ­ spectrogram and Bx -variation; (j) ion kinetic energy spectrogram; (k) ion thermal energy spectrogram; (1­n) energetic ions (FOV 180 and 62 deg. from Sun) and electrons (FOV 180°. from Sun).


THE HIGH LATITUDE MAGNETOPAUSE

101

Figure 3. Sketch for the interaction pattern of MSH plasma flow with outer cusp on 19 June 1998; spacecraft orbit (Interball-1 moves from left to right), with characteristic ion velocity ~ vectors in the XZ GSE plane; N- normal to MP in GSE frame ~(0.7, 0.07, )0.71); Pn-projection ~ of Poynting vector on N; VMS, VA-magnetosonic and Alfven speeds, see also Figure 4; MP is shown by the thick broken curve; OC, IC ­ see Figure 1 and related discussions.

In Figure 3 the cusp throat is closed by the smooth MP at a larger distance, when compared with Figure 1b. A principal problem is distinguishing dynamic interactions of the SW with the MP from local disturbances. For this purpose in Figure 2.A5 a wavelet spectrogram (see details in Savin et al., 2002b) of the IMF GSE By-component from Geotail is shown for 1.6­100 mHz range (octave-based frequency scale in Hz and color scale for logarithrm of the wave power in [nT2/Hz] are shown on the left side). We compare the Geotail data with those of Interball-1, given in panel i in Figure 2.A6, where a black line represents the Bx-variation for 2-min intervals from 4 Hz-sampling of the magnetic field (Df, scale on right side in [nT/Hz1/ 2 ]). The MP crossing (from MSH to OC, see Figure 3) is marked at the bottom of Figure 2.A6, along with the Interball-1 position in the GSM frame at two points. In panel a of Figure 2.A6 we display energy densities in eV/cc; ion thermal nTi (n and Ti-ion density and temperature) ­ blue line; DC magnetic B2/2l0-violet line; kinetic energy Wkin-black line; the red curve presents GDCF predictions for Wkin, multiplied by a factor of 0.8 to adjust the measured value in the middle MSH (following Savin et al., 2004). The time lag is chosen for best fit at 08:40­09:50 UT, while it is certainly different for 07:30­ 08:30 UT (see Dubinin et al. (2002) and Figure 4 below). Panel b of Figure 2.A5 shows the electric field (En, black line), calculated from the vector product of ion velocity and magnetic field, along the MP normal (N ~ (0.7, 0.07, )0.71) in the GSE frame (see Savin et al. 2002a, b), and that of the Poynting vector Pn at 5­50 mHz (blue line), also in the N


102

S. SAVIN ET AL.

Figure 4. Tracing of disturbances in TBL on 19 June 1998; black vertical line ­ approximate MP crossing. (a) and (b) wavelet correlation time for ion thermal and kinetic energy densities, left ­ octave frequency scale in mHz, right ­ gray scale in periods of coherent signal at every frequency, black horizontal lines ­ 1st, 2nd and 4th spacecraft spin harmonics; (c) Poynting flux along MP normal (Pn, see Figure 2.A6 and related discussions) for 2­50 mHz; (d) electric field along MP normal and its GDCF prediction (thick gray line), the time lag at 08:30­ 10:30 UT is the same as in Figure 2.A6, the lag at 07:30­08:30 UT is 7.5 min less (numbers mark events discussed in the text); (e­g) Poynting flux along GSE X, Y and Z for 2­50 mHz; (h) ion density and its GDCF prediction (thick gray line). Top ­ black bars mark flows with ~ magnetosonic speed VMS, bottom ­ gray bars mark flows with ~ Alfven speed VA .

direction. The GDCF proxy for En is represented by the red color, and shifted by an extra 1 mV/m for better adjustment with the experimental data. In panel c the magnetic field clock angle is presented in degrees: the black line is for the Interball-1 data, and the violet line shows SW monitoring by Geotail shifted in time to adjust to the average Interball-1 data at 09:30­ 10:00 UT. Predictions of GDCF at the Interball location are represented by the red line. All three curves correlate at ~08:40­09:50 UT, proving an MSH encounter. Panel d of Figure 2.A6 displays |B| in the same format (scale for SW in nT on the right side). Systematic discrepancies between the data and SW proxy, which we call `plasma balls', are shaded blue (cf. Figure 3). Note that at ~09:33 UT a similar field depression at Interball is predicted by GDCF. Only the appli-


THE HIGH LATITUDE MAGNETOPAUSE

103

cation of the model can provide a reliable tool to determine this crossing of the low-shear MP (at ~ 09:53 UT), which is imbedded in the TBL. The change of the sign of the ion velocity (± signs are marked by green/blue) in panel e (Vix, GDCF ­ red line, scale in km/s) confirms the identification of the MP encounter. In panel f, the ion temperature Ti slightly reduces prior to the MP and then rises by a factor of 1.5­2. This reduction reflects a diminishing effective temperature of the core MSH ions as the Ti is fitted to the 3D Maxwellian ion distribution. The Ti does not account correctly for the input from higher-energy protons, clearly seen in panel h from 08:53 UT on. A similar remark is also applicable for the absolute value of nTi in panel a. In panels g and h of Figure 2.A6, electron and ion color-coded (scales for counts per second on the left side) energy spectrograms are presented, with the energy-per-charge scales in eV depicted on the left. Black lines give count rates of the ions and electrons with energies >30 keV, which flow generally towards the Sun (count rate scales on the right vertical axes). Panel j of Figure 2.A6 depicts a wavelet spectrogram of the ion kinetic energy density, and panel k that of the thermal ion energy (vertical frequency scales are octave-based, i.e. logarithmic, cf. Figure 2.A5 and Figure 2.A6i). In panels 1­n we present color-coded (see the logarithmic scale on the left side in cm)2 keV)1 ster)1 s)1) spectrograms of energetic ions flowing towards the Sun (FOV 180° from the Sun, cf. black line in panel h), from the Sun (rotating FOV at 62° from the spacecraft spin axis, pointed to the Sun) and sunward flowing electrons (panel n, FOV 180°, cf. black line in panel g). Returning to a comparison of the By-spectra from Geotail and Interball-1 (Figure 2.A5, 2.A6i), the time lag between Geotail and Interball)1 should be 5­15 min (Savin et al., 2002a). Within those lags a SW disturbance at ~07:52 UT on Interball-1 is quite similar to that at Geotail. In the middle of the MSH at ~ 08:30 UT another disturbance practically coincides with that of the SW, with the low-frequency part being strengthened in the MSH. At 09:00­10:50 UT in the MP vicinity, wide-band Interball-1 fluctuations are seen; most of them have no counterparts in the SW, and vice-versa. This implies that driving by the SW is not dominant for the near-MP period analyzed; note multiple spectral maxima in this region (TBL, see Figure 3), which are related in a complicated manner. At frequencies >0.7 mHz the disturbances in the TBL have higher intensity levels and different frequency dependencies, as compared with the MSH; therefore we think that the MSH is also not the major source for the fluctuations in the TBL. This is in agreement with the cross-correlation of By at 07:30­10:10 UT being <0.5; at 09:30­10 UT the cross-correlation is 0.23, with a time lag for Geotail of 12 min, with Geotail Vx=)478 km/s. Considering possible different tilts for SW disturbances (Maynard, 2003) would hardly improve the correlation substantially. Greater resolution of these variations is presented in Figure 4, which will be discussed in detail in Section 2. After implementing


104

S. SAVIN ET AL.

different lags in Figure 4d, we can see that in events 2, 3 and 10 the GDCF (i.e. SW) disturbances produce clear responses in the Interball data, while the TBL perturbations of a comparable magnitude in events 5­7, 9 and 11 have no counterparts in the SW (cf. also Figure 2.A6, panel c). Strong differences in the magnetic spectral shapes in the Interball-1 and Geotail data are also in agreement with the local nature of the TBL turbulence (Savin et al., 2002a). Just inside the MP, energetic electrons have a high count rate (Figure 2.A6, panel g) that marks the boundary of closed magnetospheric field lines. This provides extra support for locating the PB inside the MP. Waves in the 2­50 mHz range (panels i­k) correlate with the intensity of energetic protons (black line in panel h and panels l, m) upstream of the MP, starting from ~08:53 UT. The main wave bursts have counterparts in the energetic electrons (panel g). The low-shear MP (~80°) and PB encounters take place at a tilt ~)21° (i.e. the Southern dipole axis being turned from the Sun towards the tail). The IMF Bz turned to positive values about 10 min prior to the MP. The ion plasma beta at 09:56­10:03 UT reached 15; in the rest of the blue-shaded sites it exceeds 2. Similar PB encounters occur on one previous and two following Interball-1 orbits, on 15­27 June 1998 (Savin et al., 2005, submitted).

2.2. Direct interaction of the magnetosheath flow with a `plasma ball' After this discussion of the general TBL and MP features on 19 June 1998, we embark on a detailed investigation of the TBL properties. The main physical problem to address is how the practically demagnetized PB is interacting with the incident MSH flow in the collisionless plasma. Due to the high beta both in the MSH and the PB, the magnetic forces are small, and only local electric fields and waves can provide the MSH flow deflection and/ or dissipation. The electric field En near the MP can be supported by a surface charge at the MP-related current sheet (s), and it can deflect the incident MSH plasma to flow along the MP, while it cannot stop the normal flow in the absence of wave-induced effective collisions. The `local' En, should be seen as a regular difference between the measured and SW-induced field (i.e. GDCF one). The MP transition at )09:53 UT is manifested in the different sign as compared with the GDCF one. Figure 2.A6b shows that the difference is mostly wave-like; the only systematic difference upstream of MP is visible at 09:48­09:53 UT. Such a negative En (relative to the GDCF one) might contribute to the flow turning, but it should accelerate the incident particles towards the MP (instead of stopping them) if the measured highamplitude waves provide an effective perpendicular conductivity. Thus, the waves constitute the major means for the boundary and the MSH plasma interaction in the case under study.


THE HIGH LATITUDE MAGNETOPAUSE

105

In Figure 4a­b we display wavelet correlation times (see Savin, 2002b, 2003a, for details) for the thermal and kinetic ion energy densities, which indicate for how many periods (at each frequency) the signal is coherent. Usually a signal conserving coherence for more than 2 periods at several consecutive analysis intervals can be regarded as a regular or coherent one. Panels c, e­g display the Poynting vector normal to the MP and its GSE components in the frequency band 2­50 mHz. In Figure 2.A6, the latter serves to outline the weak sunward moving disturbances (i.e. with positive Pn). In panel d we present En with different time lags at 07:30­08:30 and 08:30­10:30 UT. In Figure 4 h and i, the measured and GDCF ion densities and ion kinetic energy density are depictured with the same time lags. Thick gray vertical lines with interruptions mark characteristic disturbances to be discussed; they are numbered in panels d and e. At 07:30­08:20 UT strong post-shock activity is well seen in Figure 2.A6, (cf. Savin et al., 2002b). Both the spectral and correlation time maximums at 1­2 mHz are recognizable throughout MSH (Figure 2.A6j and Figure 4a­b). Savin et al. (2002a, b) outlined similar maxima in magnetic spectra on Interball and Polar at these times, but they couldn't detect continuous or coherent signals. We draw attention to events 2 and 3 (Figure 4), where the measured En on average reproduces that of the GDCE. Thus, these events represent SW disturbances moving in the downstream MSH far from the disturbed TBL. Accordingly, all Poynting flux components are negative in these events (cf. Figure 3 and Interball-1 GSM position in Figure 2.A6 at 09 UT). The same is, most probably, valid for events 1, 4 and for the low-frequency En trends between events 9­10 and 11­12. So, moderate SW disturbances provide a validation for our Poynting flux measurements in the MSH. At 08:35­ 08:53 UT a weak activity in panels i­k (Figure 2.A6) resembles that of Figure 2.A5 and thus is driven by the SW. The respective maximum at ~4­5 mHz in correlation time (Figure 4b) could be traced from the post-shock region at 07:50­08:50 UT. The region at 08:50­09:50 UT is characterized by strong disturbances, which are not SW-driven ones (cf. Figure 2.A5 and panels a­e and i­k in Figure 2.A6). Soft energetic ions are registered there (panels h, 1, m, Figure 2.A6) that correlate with the strong energy fluctuations and with the drop in the MSH kinetic energy. The latter drop is well seen after 09 UT in Figure 2.A6a and Figure 4i as a systematic difference between the black and thick red traces. Figure 2.A6e also demonstrates the larger departure of Vix from the model after that time. Figure 2.A6e also demonstrates the larger departure of Vix from the model after that time. We check the density correspondence to the model in Figure 4h; the average measured density follows the GDCF proxy rather well until the diffuse MP encounter, with two exceptions at ~07:50 and 08:45 UT. The first density departure can be affected by a partial shock crossing, while in the second one the ion momentum (~nVix) is close to the GDCF prediction, but


106

S. SAVIN ET AL.

it departs from GDCF after 09 UT. The general agreement confirms the reliable ni measurements and, thus, the local dramatic Wkin decrease (Figure 4i). Note that the average MSH flow is subsonic (Wkin B2 =2l0 ). On the top of Figure 4, black bars mark flows with nearly the magnetosonic speed, VMS ; on the bottom gray bars mark flows with nearly the Alfven speed, VMS . This is also shown schematically by thin arrows in Figure 3. Besides the accelerated MSjet at 09 UT, there are a number of smaller MS-jets in the upstream TBL. The decrease in Wkin mentioned above after 09 UT corresponds to a decelerated Alfvenic flow. In the upstream TBL those flows are mixed with the MS-jets. In the MSH upstream of the TBL at 08:40­08:50 UT the standard deviations of ion energy densities are: dWkin~ 22%, dnTi~ 10%. Figure 2.A6k shows that at 08:53­09:35 UT nTi fluctuates, at 09:03­ 09:15 UT the nTi-disturbances dominate over those of Wkin : dWkin ~ 49%, dnTi ~ 20%, dWkin/dnTi ~ 0.48. In the middle of the upstream TBL both the kinetic and thermal ion energies are quite disturbed (09:15­09:35 UT): dWkin ~ 71%, d nTi ~ 25%, d Wkin /dnTi ~ 0.996. Relative to the unperturbed MSH the standard deviations in the TBL center are: d Wkin ~ 45%, dnTi ~ 17%. The lower limit for the energy conversion into the irregular fluctuations in the TBL (i.e. the difference of standard deviations) is 23% of MSH kinetic energy and 7% of its thermal. As the TBL spectra in Figure 2.A6j and k are quite different from the upstream MSH ones, we would like to accept the higher limits for the MSH energy chaotization: dWkin ~ 30­40% and dnTi ~ 10­15%. A strong deficit of the average Wkin at 09:03­09:15 UT (and of the ion momentum) is displayed even in the sum of Wkin+dWkin, which constitutes only 47% of the average upstream kinetic energy. The only candidate to carry off the momentum and kinetic energy excess is the strong impulse in the Wkin and Vix in Figure 2.A6 and Figure 4 at ~09 UT. The respective hodogram of the ion speed vector tip in the plane of GSE (Vx, Vy) is shown in Figure 5a. The average speed in the depicted interval 08:59:11­ 09:00:08 UT is ()252, )45, )88) km/s, and the possible inferred vortex component (i.e. the loop in the hodogram) has a radius of ~50 km/s (~4% of its average Wkin). Those estimates are given in the spacecraft frame (which is close to the MP frame), while in the frame of the unperturbed MSH the vortex kinetic energy is ~0.3 Wkin . Since the magnitude of the Wkin-pulse reaches the local value of nTi (Figure 2.A6a), that means a nearly magnetosonic (MS) speed of the jet (VMS $ Ï2Ti =Mi ÷1=2 in the high-beta plasma (Mi ­ proton mass, Ti ­ in energy units)). A magnetic loop at this time has been found at half this frequency. This plasma jet is in the middle of a current sheet (bi-polar disturbance in the clock angle, Figure 2.A6c), bounded by |B| drops down to a few nT (`diamagnetic bubbles') and a bi-polar En - spike (~5 mV/m), which can be


THE HIGH LATITUDE MAGNETOPAUSE

107

Figure 5. Left: Hodogram of ion Vx, Vy at 08:59:11­09:00:08 UT, 19 June 1998. Right: Hodogram of ion Vx, Vy 09:42:28­09:43:46 UT, 19 June 1998.

accounted by surface charges at the current sheet. In Figure 4 it is event 6 that contains the impulsive density rise. The negative spike of the Poynting flux in all compon