1. Space Research Institute, Russian Academy of Sciences, Moscow, Russia
2. Institute of Geophysics and Planetary Physics, University of California, Los Angeles, USA
3. Rice University, Houston, Texas, USA
We present the results of the first two-point study of the cusp/magnetosheath interface, using Interball-1 and Polar magnetic field measurements on May 29, 1996. The strong northward interplanetary magnetic field and the near constancy of the solar wind dynamic pressure provide a unique opportunity for a near-simultaneous detection of the high-altitude magnetopause and the polar cusp. Both Polar and Interball cross field lines that pass through the northern cusp and apparently close to the post-cusp reconnection site. The magnetopause current observed by Interball consists of two quite distinct layers, an inner broad current that is quite turbulent and another current that is quite abrupt and quiet. Polar also crosses current layers, similar to the Interball inner one. The turbulent current sheets are encountered over a wide range of MLT (0830-1500) and geomagnetic latitudes (57-67 degrees). These observations support a model in which cusp field lines experience essentially stochastic behavior but on average provide topological connection between the cusp and magnetosheath. This picture differs from one of laminar reconnection by the presence of the substantial ion heating. For the northward IMF the turbulence is seen inside the main magnetopause current sheets. During other IMF orientations turbulent mixing driven by the magnetosheath flow interaction with the indented near-cusp magnetopause might also be present.
On May 29, 1996 Interball 1 and Polar nearly simultaneously, encountered magnetosheath-like plasma under rather steady solar wind conditions. Because this interval allows us to study the vicinity of cusp-magnetosheath interface simultaneously at two different locations, these data are of great interest. Moreover to our knowledge such a dual encounter is thus far unique in the existing literature. Earlier single spacecraft observations with Heos-2 and later Prognoz-7, 8, 10 have shown that the magnetopause position and magnetosheath plasma flow structures are quite variable near the cusp, a magnetospheric region that is crucial for magnetosheath plasma entry [Paschmann et al., 1976; Haerendel and Paschmann, 1975; Lundin et al., 1991; Savin, 1994]. Recently a turbulent boundary layer has been identified just outside the near-cusp magnetopause [Klimov et al., 1986]. Interball-1 statistics of the magnetopause location for autumn 1995 - spring 1996, and under normal solar wind conditions, indicate that there is also an indentation of about 2 RE depth, which has been called the outer cusp throat [Savin et al., 1997]. This outer cusp region is rather narrow with a variable location. Observations of the highest latitude ISEE magnetopause crossings also suggest a depression in this region [Petrinec and Russell, 1995]. Hawkeye-1 did not map this region well because its orbit took Hawkeye-1 generally far from the cusp compared to those of Heos-2 and Prognoz. Nevertheless statistically Zhou and Russell  did find a magnetopause in this region that was significantly closer to the Earth than at lower latitudes.
The definition used to identify the magnetopause in this region may affect the results [Fung et al., 1997]. In this paper the magnetopause is understood in the sense introduced by Haerendel and Paschmann , i.e. the current sheet where the magnetic field turns from Earth-controlled to magnetosheath-controlled.
Similarly, the term, outer cusp throat, is used here to classify the highly disturbed and/or stagnant magnetosheath plasma, which we believe is located outside the magnetopause. We note that Chen et al.  recently presented a Hawkeye-1 example that looks quite consistent both with the one discussed here and with that presented by Savin et al. . We concentrate here on the magnetic fields using mainly Interball-1 and Polar magnetic field data together with IMP-8 magnetic field and Wind dynamic pressure. For the corresponding instrument descriptions see Klimov et al., ; Russell et al., ; King  and Ogilvie et al., . One can find a comparative study of the Polar data and a global MHD model in Russell et al, . Since the solar wind conditions and the location of Interball were similar during the Interball magnetopause encounter to those of the Polar magnetopause encounter, the same overall sketch of the interaction applies herein. (See their Figure 5).
On May 29, 1996 IMP-8 and Wind were well positioned for monitoring the solar wind parameters. IMP-8 was just upstream of the bow shock (22, 28, -1 RE in GSM), so that the time shift needed to adjust the magnetic data to represent the conditions close to the cusp should be less than four minutes. The corresponding magnetic field measurements in the solar magnetospheric (GSM) frame are shown in Figure 1. The IMP-8 data are not shifted for this small delay. At the bottom of Figure 1 solar wind dynamic pressure, shifted by 39 minutes, is shown. The Wind measurements were obtained at (156, -7, -10) RE in GSM. The solar wind dynamic pressure is high on this day, about 8 nPa, and the magnetosphere correspondingly smaller than usual. The first vertical line from the left in Figure 1 (labeled I) gives the time of the Interball-1 outbound magnetopause crossing; the second vertical line (labeled Po) shows the time that Polar crossed a broad turbulent current sheet and the third line (labeled Pi) shows the time of the Polar entry into the polar cap field lines from a region of cusp-like plasma. The IMP-8 magnetic field data are identical to the Wind data (after the appropriate time shift) except for some minor details.
|Fig. 1. The IMP-8 magnetic field (unshifted) and WIND dynamic pressure (shifted by 39 minutes to adjust to the Interball-1 and Polar positions) on May 29, 1996. I, Po and Pi refer to the times of the Interball and Polar crossings of current sheets.|
In Figure 1 one can see that the Bz component is 10-15 nT northward throughout the Interball and Polar passes. Under strongly northward IMF conditions we would expect the magnetopause position to be principally controlled by the solar wind dynamic pressure [Petrinec and Russell, 1993]. While the Interball-1 magnetopause crossing coincides with the IMF By and Bx sign changes as shown in Figure 1, it appears that the immediate cause of the crossing was the increase of the dynamic pressure from 6.4 to 9.8 nPa that would be expected to decrease the radius of the magnetopause by about 7% [Petrinec and Russell, 1993]. The entry of Polar into the polar cap in Figure 1 occurs at close to 0700 UT during rather stable solar wind and IMF conditions that should be unable to account for the multiple apparent crossings observed.
|Fig. 2a. The Interball and POLAR orbits on May 29, 1996, projected into the dawn dusk plane, Letters I and P denote the crossing of the magnetopause by Interball and entry into the polar cap by POLAR.|
|Fig. 2b. The noon-midnight projection.|
|Fig. 3. The Interball MIF-M magnetic field in GSM frame on May 29, 1996. The main magnetopause current layer observed by INTERBALL-1 is at 03:17 UT, as indicated by the vertical line marked MP. Here the Bx and Bz sign changes from the magnetospheric toward the magnetosheath direction. The rotation is completed by a second discontinuity (RD) at 0323 UT. The T96 model is shown by a dashed line. The model of Urquhart et al.  is shown by a solid green line. The arrows mark a series of bow shock (BS) crossings as Interball enters the solar wind. In the insert are shown Corall ion energy spectra for 0300-0330 UT from the rotating channel (f.o.v. perpendicular to the spin axis pointed to the sun, bottom panel) and ion pitch angle in degrees (upper panel).|
We start with the magnetopause encounter by Interball-1, which was the first to occur. Figure 2 shows projections of the Interball and Polar orbits in the dawn-dusk and noon-midnight planes. We see that Interball leaves the magnetosphere on the dawnside. Polar is initially on the dawnside but crosses over to the dusk side after about 0500 UT. The noon-midnight projection (X-Z plane) shows that Interball is moving sunward throughout the interval with Y about -3.5 RE and crossing the dawn-dusk meridian at about 0115 UT. Polar is roughly about 3.5 RE in front of the Earth and moving upward. Figure 3 shows the magnetic field at Interball-1 in the GSM frame. The observed field is shown by a solid line. At the bottom the spacecraft location in the GSM frame is given in RE. The comparison of the measurements with the Tsyganenko 96 (T96) model indicates that at 03 UT Interball-1 was apparently on tail field lines on the dawn side, the total field being about 10 nT more than the predicted value. The solid green line shows a model of the magnetosphere developed by Urquhart et al.  to explain the observations at Polar. This model predicts the magnetosheath field orientation quite well between 0323 and 0415 UT but predicts the magnetopause and bow shock to be closer to the Earth than observed. The bow shock is identified by the sudden (multiple) drops in field strength to the value observed at WIND and IMP.
|Fig. 4. Magnetic field at 1-second resolution across outbound crossing of magnetopause as observed by Interball. Magnetic field data are given in principal axis system determined over the interval between the 2 vertical lines. The minimum variance direction is (0.83, -0.04, 0.56) in GSM and the maximum variance direction is (-0.56, -0.16, 0.82). The eigenvalues corresponding to the 3 eigenvectors are 2737, 173 and 46 nT2.|
The Interball-1 outbound magnetopause consists of two current layers, one at 0317 UT and one at 0323 UT. The first current sheet is extremely noisy with much high frequency structure along the magnetopause normal direction. The complete crossing of this current layer lasts about 3 minutes. The 1-second resolution magnetic field measurements in the principal axis system are shown in Figure 4 for the period from 0312 to 0326 UT. The second current layer lasts about 10 s and is quiet. Each current sheet has a well-defined maximum variance direction but neither has a well-defined minimum variance direction. Analyzed as one complete crossing of the magnetopause the minimum variance direction for the two sheets is (0.82, 0.00, 0.57) in GSM. This direction is independent of the frequency bandwidth of the analysis (i.e. whether filtered or unfiltered data are used). The normal component across the two current layers is small, about 5 nT, directed inward as would be expected if reconnection were occurring between the northward IMF and the tail field lines. Because the field magnitude is fairly constant across the current sheet and there is a modest normal component across it, we believe the second sheet is a rotational discontinuity.
The Polar magnetic field data for the period 0300 to 0800 UT on May 29, 1996 have been discussed by Russell et al. . Polar appears to traverse an extensive region of magnetosheath plasma on magnetospheric field lines and then to enter into the polar cap at 0710 UT. In some respects it is similar to the Interball-1 outbound magnetopause, occurring at nearly the same XZ coordinates and under similar dynamic pressure and Bz conditions. Here the magnetic field increases from low levels and becomes quiet and nearly completely reverses its direction. This boundary is not the crossing from the magnetosheath to the magnetosphere because the plasma data shows that the spacecraft was on field lines connected to the ionosphere prior to this current sheet crossing. Moreover, Polar entered this region slowly via a gradual rotation of the average field direction immersed in a high level of turbulence, from approximately 0400 to 0445 UT [Russell et al., 1998]. Similar turbulence could be seen during the Polar inbound crossing into the polar cap at 0640-0710 UT (see e.g. Figure 5), where the peak-to-peak disturbance amplitudes reach 90% of the average total field. Similar to the Interball outbound magnetopause at 0315-0319 UT in Figure 4, the total magnetic field drops indicate a diamagnetic effect caused by heated plasma.
|Fig. 5. The magnetic field at 0.125 second resolution observed by POLAR as it crossed from the region of the cusp to the polar cap field lines. Measurements are given in principal axis coordinates determined over the interval between the two vertical lines. The minimum variance direction is (0.73, -0.13, 0.67) in GSM coordinates, and the maximum variance direction is (-0.37, 0.75, 0.55). The eigenvalues corresponding to the 3 eigenvectors are 1060, 179 and 33 nT2.|
The minimum variance analysis for the final Polar crossing interval (0705-0706 UT) gives a normal direction of (0.73, -0.13, 0.68) with an inward directed normal magnetic field of 8 nT. This direction is similar to that observed by Interball four hours earlier, nearby on the tail magnetopause. Figure 5 shows the Polar magnetic field observations rotated into the principal axis system. We note that the magnetic field is quite turbulent through the entire interval prior to the magnetopause crossing but somewhat less so as the final crossing at 0706 UT is approached. In the case of both the Interball and Polar current sheet crossings, strong short-lived magnetic components are observed along the current sheet normal. These spiky transients are symptomatic of the low-latitude magnetopause as well [Russell, 1995].
The 3D particle measurements on Interball can assist in interpreting these results. We present here the Corall experiment ion energy spectra [Yermolaev et al., 1997]. In the insert in Figure 3, the upper panel shows the ion pitch angle and the lower panel shows the ion energy spectra versus time. The ion distribution displays a strong mantle-like, field-aligned, tailward flow at 0300-0307 UT (high pitch angles in the insert in Figure 3). It has been proposed that for the northward IMF the mantle may be due to the direct magnetosheath plasma injections at locations upstream of the spacecraft (Lundin et al., 1991). Thus it is important to attempt to distinguish the topology of the magnetic field throughout this interval.
During the measurements obtained from 0316 to 0318 UT straddling the time of the changing of the sign of the Bz and Bx components, ions, accelerated up to 20 keV, were observed. The ion temperature there is 2.5 times higher than in the undisturbed magnetosheath. These fluxes are a sign that reconnection occurred poleward of the point of observation. The respective particle data will be discussed in detail in the separate paper. Here we would like to point out only that the region of the ion acceleration/heating coincides with the high turbulence level zone inside the first current sheet (see Figure 4). After crossing the second discontinuity at 0323 UT the unaltered magnetosheath plasma is observed.
The Interball-1 data presented here agrees with the Dungey  model for northward IMF. Figure 5 of Russell et al.  illustrates the draping field pattern when Polar was near the magnetopause. While we have not shown the equivalent figure in this paper, such figures have been drawn and show that Interball also crossed field lines that passed near the northern cusp reconnection point. At both spacecraft the magnetic field lines have the overdraped geometry suggested by Russell . This interpretation is confirmed with the Interball-1 ion data. Beyond the outermost discontinuity is the undisturbed magnetosheath flow. Inside this region (from 0318-0323) the flow has been altered but the inner current layer at 0318 UT has a much greater effect on the plasma. In fact the flow reverses in space, across this layer. This is again clear evidence for high-latitude reconnection. This interpretation, however, has no counterpart in the model of Urquhart et al.  whose predictions reproduce the magnetic field behavior prior to the magnetopause crossing in Figure 3. Both the particle data and the MHD model indicate that the near-magnetopause field lines (0316- 0318 UT) at Interball are connected to the northern ionosphere.
Comparison with the smooth Sibeck et al.  model, derived principally from low-latitude observations, and with the Zhou and Russell  model, derived from high-latitude Hawkeye observations, shows that the Interball outbound magnetopause and Polar main magnetopause current sheet encounters are 1.1-1.7 RE inward of the low-latitude magnetopause, but right on the expected location of the high-latitude magnetopause for northward IMF for the enhanced dynamic pressures observed on this day. This location is also close to the average Interball value in Savin et al. . The two current layers crossed by Polar at about 0430 UT and 0700 UT occur inside the magnetopause as defined in the Hawkeye study. The turbulent current layer (see Figures 4 and 5) contains accelerated/heated magnetosheath particles (see insert in Figure 3), the temperature at 0316 UT is 268 eV (i.e. 2.5 times higher than at 0328 UT in the unperturbed magnetosheath). The diamagnetic effect of the ion temperature excess corresponds to an average drop of the total magnetic field of about 30 nT for the measured ion density of 8 cm-3. The bursty total field drops in Figure 5 are consistent with ion temperature jumps in this turbulent current layer up to 500 eV if the ion density is of the order of 50 cm-3. These values should be clarified later in more detailed data analysis.
The observation of this turbulent current layer at three points (in the range 0830-1500 MLT and 57-67 degrees of geomagnetic latitude) over a period of four hours is a strong argument for the rather permanent existence of this turbulent current layer. Thus in the vicinity of the cusp a substantial part of the reconnected lines should cross the turbulent current layer where the high-level of turbulence makes them stochastic and randomizes the magnetosheath ions. This picture resembles the stochastic percolation model of Kuznetsova and Zelenyi . This turbulence allows the cusp magnetic field lines to be disconnected from the magnetosheath by the turbulence in the current layer with very different particle distributions just inside and outside cusp-magnetosheath interface (see also Savin et al., 1997).
Strong northward IMF differs substantially from other IMF orientations by the confinement of the turbulence to the region mostly inside the main magnetopause current sheet. The outer turbulent boundary layer observed for other IMF directions might result from turbulent mixing, driven by the magnetosheath flow interaction with a deformed near-cusp magnetopause. The reason why this does not happen in the present case may be due to the high latitude extension of the subsolar plasma depletion layer with the high magnetic field magnitude. The MHD simulation has a rather smooth boundary in the vicinity of the polar cusp for the conditions seen on this day. Current sheets with a high level of magnetic fluctuations are not a unique feature of the near-cusp magnetosheath-magnetosphere transitions. They can be observed both on reconnected near cusp lines (e.g. overdraped as in the case under study) and inside sufficiently thin low-latitude current sheets themselves (see the percolation model of Kuznetsova and Zelenyi, 1990).
Finally, this initial look at the cusp/magnetosheath interface as seen by Interball 1 and Polar shows that the multipoint and multi-instrument data are very promising. Future more detailed study of the particle data will enrich substantially our understanding of these complicated and intriguing processes.
We thank K. W. Ogilvie and the SWE team for providing of WIND solar wind dynamic pressure data and J. H. King for providing the IMP-8 magnetic field data. We appreciate the help in the paper preparation by I. Dobrovolsky, V. Prokhorenko and V. S. Romanov. Work was partially supported by INTAS through grant INTAS-93-2031 and through the National Aeronautics and Space Administration under grant NAG5-4066.
Chen, S.-H. et al., Exterior and interior polar cusps: Observations from Hawkeye, J. Geophys. Res., 102, 11335-11347, 1997.
Dungey, J. W., The structure of the exosphere, or adventure in velocity space in Geophysics the Earth's Environment, edited by C. DeWitt, J. Hieblot and A. Lebeau p.505, Gordon and Breach, New York, 1963.
Fung, S. F., T. E. Eastman, S. A. Boardsen and S.-H. Chen, High-altitude cusp positions sampled by the Hawkeye satellite, Physics and Chemistry of the Earth, accepted 1997.
Haerendel, G. and G. Paschmann, Entry of solar wind plasma into the magnetosphere, in Physics of the Hot Plasma in the Magnetosphere, edited by B. Hultqvist and L. Stenflo, p.23, Plenum, NY, 1975.
King, J. H., Availability of IMP-7 and IMP-8 data for the IMS period, in The IMS Source Book, edited by C. T. Russell and D. J. Southwood, 10-20, American Geophysical Union, 1982.
Klimov, S. I. et al., Investigation of plasma waves by combined wave diagnostic device BUDWAR PROGNOZ-10-INTERCOSMOS, Cosmic Research (Transl. from Russian), 24, 177, 1986.
Klimov, S. et al., ASPI Experiment: Measurements of Fields and Waves Onboard the INTERBALL-1 Spacecraft, Ann. Geophys., 15, p.514-527, 1997.
Kuznetsova, M. M., and L. M. Zelenyi, The theory of FTE: Stochastic percolation model, in Physics of Magnetic Flux Ropes, edited by C. T. Russell, E. R. Priest, L.C. Lee, pp.473-488, American Geophysical Union (1990).
Lundin, R., J. Woch and M. Yamauchi, The present understanding of the cusp, in Proceedings of the Cusp Workshop, European Space Agency, Spec. Publ., ESA SP-330, p.83-95, 1991.
Ogilvie, K. W. et al., A comprehensive plasma instrument for the WIND spacecraft, in The Global Geospace Mission, Space Science Rev., 71, 55-77, 1995.
Paschmann, G., G. Haerendel, N. Sckopke, H. Rosenbauer and P. C. Hedgecock, Plasma and magnetic field characteristics of the distant polar cusp near local noon: The entry layer, J. Geophys. Res., 81, 2883, 1976.
Petrinec, S. M. and C. T. Russell, Factors which control the size of the magnetosphere, in Solar-Terrestrial Predictions IV, Vol. 2 edited by J. Hraska, M. A. Shea, D. F. Smart and G. Heckman, 627-635, NOAA, ERL, Boulder, Colorado, 1993.
Petrinec, S. M. and C. T. Russell, An examination of the effect of dipole tilt angle and cusp regions on the shape of the dayside magnetopause, J. Geophys. Res., 100, 9559-9566, 1995.
Russell, C. T., The configuration of the magnetosphere, in Critical Problems of Magnetospheric Physics, edited by E. R. Dyer, 1-16, IUCSTP Secretariat, Washington, D.C., 1972.
Russell, C. T., The structure of the magnetopause, in Physics of the Magnetopause, edited by P. Song, B. U. O. Sonnerup and M. F. Thomsen, p.81-98, American Geophysical Union 1995.
Russell, C. T. et al., The GGS Polar magnetic fields investigation, Space Sci. Rev., 21, 563-582, 1995.
Russell, C. T et al., Entry of the POLAR spacecraft into the polar cusp under northward IMF conditions, Geophys. Res. Lett., in press, 1998.
Savin, S. P., ELF waves near the high latitude magnetopause, in Abstracts of AGU Chapman Conference on Physics of the Magnetopause, March 14-18, p.41, 1994.
Savin, S. P. et al., Interball tail probe measurements in outer cusp and boundary layers, in Encounter Between Global 5Observations and Models in the ISTP Era, edited by D. Gallagher, J. Horwitz, T. Moore, American Geophysical Union, submitted 1997.
Sibeck, D.G., R.E. Lopez, E.C. Roelof. Solar wind control of the magnetopause shape, location and motion, J. Geophys. Res., 96, 5489-5495, 1991.
Urquhart, A. L. et al., Magnetic field models for Polar magnetopause crossings of May 29, 1996, J. Geophys. Res., submitted, 1998.
Yermolaev, Yu. I. et al., Ion distribution dynamics near the Earth's bow shock: first measurements with the 2D ion energy spectrometer CORALL on the INTERBALL/Tail-probe satellite, Ann. Geophys., 15, p.533-541, 1997.
Zhou, X.-W. and C. T. Russell, The location of the high latitude polar cusp and the shape of the surrounding magnetopause, J. Geophys. Res., 102, 105-110, 1997.
S. P. Savin, S. A. Romanov, A. O. Fedorov, L. Zelenyi, S. I. Klimov, Yu. I. Yermolaev, E. Yu. Budnik and N. S. Nikolaeva, Space Research Institute, Russian Academy of Sciences, Profsoyuznaya 84/32, GSP-7, Moscow, 117810, Russia, (e-mail: firstname.lastname@example.org)
C. T. Russell and X-W. Zhou, Institute of Geophysics and Planetary Physics, University of California, Los Angeles, 3845 Slichter Hall, Los Angeles, CA 90095-1567, (e-mail: email@example.com)
L. Urquhart and P. H. Reiff, Rice University, 6100 Main Street, Houston, TX 77005-1892.