Note: The paper appearing in this CD-ROM
has been slightly revised from that appearing in the hard copy of Advances
in Space Research,Volume 20, Issue 4-5, pages 809-812.
THE LOW-LATITUDE FLANK MAGNETOSHEATH, MAGNETOPAUSE, AND BOUNDARY LAYER: WIND OBSERVATIONS
T.D. Phan1, D. Larson1, M. Moyer1,
J.P. McFadden1, R.P. Lin1, C.W. Carlson1,
M. McCarthy2, G.K. Parks2, H. Reme3, T.R.
Sanderson4 R. Lepping5, and K. I. Paularena6
1Space Sciences Laboratory, University of California, Berkeley, E-mail: firstname.lastname@example.org
2Geophysics Program, University of Washington, Seattle
3Centre d'Etude Spatiale des Rayonnements, Toulouse, France
4Space Science Department of ESA, ESTEC, Noordwijk, Netherlands
5Lab for Extraterrestrial Physics, NASA/GSFC, Greenbelt, Maryland
6Center for Space Research, Massachusetts Institute of Technology, Cambridge, Massachusetts
This paper presents highlights of a survey of the WIND spacecraft crossings of the low-latitude dawn and dusk magnetosheath, magnetopause (MP), and the low-latitude boundary layer (LLBL). In the magnetosheath the plasma flow direction throughout most of the region is found to be remarkably steady, irrespective of the local magnetic field direction. The flow tangential to the MP slows down initially as one moves from the bow shock toward the MP. However, this flow is occasionally observed to speed up as the MP is approached, and is detected in the region where the plasma beta is less then unity. Thus the source of flow acceleration is likely to be associated with draping of the field lines around the MP. In the LLBL, a key finding of the WIND observations is that the flow direction in a substantial portion of the LLBL is nearly aligned with that in the magnetosheath, while the flow tangential to the MP slows down gradually as one traverses the boundary layer. This characteristic of the boundary layer flow seems to be independent of the external magnetosheath magnetic field orientation, and may suggest that the LLBL is locally coupled to the adjacent magnetosheath, irrespective of the magnetic shear across the MP.
The processes responsible for transferring mass, momentum, and energy of shocked solar wind into the magnetosphere are of great interest. Previous investigations of these physical processes have concentrated mainly on the study of the subsolar magnetopause (MP) regions. It was found that the properties of the subsolar MP regions depend strongly on the orientation of the interplanetary magnetic field (IMF).
The subsolar MP is a current sheet when the magnetic shear across the MP is high, while discontinuities in the thermal properties of the plasma, and in the bulk flow direction and magnitude characterize the low-shear MP [e.g., Paschmann et al., 1993]. Reconnection flow signatures are often observed at the dayside high-shear MP [e.g., Paschmann et al., 1979], but are rarely detected at the low-shear MP.
The plasma thermal and flow properties in the magnetosheath region in the immediate vicinity of the MP also seem to depend strongly on the magnetic shear across the MP. A plasma depletion layer (PDL) and magnetic flux pile-up against the MP are generally observed next to the low-shear MP [Paschmann et al., 1993; Song et al, 1993], but are either absent or weak when the shear is large [Anderson et al., 1993; Phan et al., 1994].
Observations of the LLBL have led to different interpretations of the topology of the region. Hall et al.  suggested that the boundary layer is on a combination of open and closed field lines. Fuselier et al. , on the other hand, suggest that the LLBL is on open field lines even when the magnetic shear across the local MP is low. In a separate study, Song and Russell  suggested that under northward IMF conditions, plasma enters the LLBL from high latitude. Le et al.  concluded that the low-shear LLBL is on a combination of open and closed field lines, as a consequence of cusp reconnection.
The magnetosheath, MP, and LLBL regions on the dawn and dusk flanks have been less explored. It has not been established whether physical mechanisms operating in the subsolar region also operate on the flanks. The question concerning the presence/absence of a PDL has not been addressed. In a study of the LLBL, Eastman and Hones  concluded that the LLBL is normally on closed field lines. Mitchell et al. , on the other hand, found that the LLBL is on closed field lines for northward IMF but is on a combination of closed and open field lines for southward IMF.
In this study we survey WIND spacecraft crossings of the flank magnetosheath, MP, and LLBL. The main focus of the present study will be on (1) the presence/absence of a plasma depletion layer next to the MP, (2) bulk flows in the magnetosheath, (3) the structure of the flank LLBL, and (4) the LLBL flow and its relationships to the flow in the external magnetosheath. These observations are expected to shed some lights on the magnetopause processes and on the formation of the LLBL.
The present analysis uses data from the WIND satellite. Details of the plasma and magnetic field instruments are described by Lin et al.  and Lepping et al. .
During the first 20 months of the WIND mission, 22 complete passes of the dayside and flank magnetosheath, MP, and LLBL have occurred. Three crossings occurred in the subsolar region. The remaining crossings took place on the flanks. Of the 22 crossings, two are crossings of the low-shear MP (defined to be when the rotation of the magnetic field across the MP is less than 45 degrees). One low-shear crossing occurred near the subsolar point, and the other on the dusk flank. The dusk event is the low magnetic shear event studied in this paper. We also study a high-shear event where the LLBL crossing duration is long in order to resolve the structure of the high-shear boundary layer.
Low Magnetic Shear Event
Figure 1 shows an overview of the inbound pass from the dusk bow shock to the magnetosphere on January 12, 1996. At ~15:10 UT, the satellite crossed the bow shock and moved into the magnetosheath. At 19:55 UT, the satellite encountered a sharp increase in the proton temperature, Tp (Fig. 1b), together with abrupt drops in the bulk plasma flow speed and magnetic field magnitude (Fig. 1f), but with almost no changes in the magnetic field direction (Fig. 1g). These features are similar to signatures of subsolar low-shear MP [Paschmann et al., 1993]. From 19:55 UT to 20:50 UT, the spacecraft was in the LLBL. The Plasma sheet was reached at 20:50 UT.
Figure 1. Overview of the inbound pass on Jan 12, 1996, extending from the solar wind across the magnetosheath and the MP into the plasma sheet. From top to bottom, the figure shows the proton number density Np; the perpendicular and parallel proton temperatures; the proton bulk speed Vp; the azimuth angle, jvp, in the LMN boundary normal coordinate system [Russell and Elphic, 1979] with 0o along the L axis and +90o along the M axis; the magnetic field strength, B; the azimuth angle, jB; and the plasma b. The direction of the MP normal is taken from the Fairfield  MP model. IMP-8 (R= 39 RE ,Lat= 35 degrees, Lt= 19 hrs) solar wind measurements are displayed with a 10 min shift backward in time. The IMF magnitude is displayed with an amplification factor of 3.2.
Throughout the magnetosheath the flow angle jvp is remarkably steady at ~-90 degrees, i.e., the flow is nearly confined to the equatorial plane. The flow direction is independent of the direction of the ambient magnetic field.
Some variations in the ion density in the magnetosheath are noted: In the region just downstream of the bow shock (15:10-16:20 UT), where the density is higher than in the remainder of the magnetosheath, and in the 19:00-19:50 UT interval where the density decreases on approach to the low-shear MP. Neither of these density variations are present in the solar wind as detected by the IMP-8 spacecraft (Fig. 1a). Thus the magnetosheath density variations seem to be spatial rather than temporal features.
The source of the density drop at 16:20 UT is presently not understood. The gradual density decrease in the 40 minute interval preceding the MP crossing, on the other hand, indicates the presence of a plasma depletion layer.
The gradual increase of the magnetic field magnitude from the bow shock toward the MP (Figure 1e) is related to an increase of the solar wind magnetic field strength observed by IMP-8, except in the region closest to the MP (19:00-19:55 UT) where the increase of the field is significantly larger than the already amplified IMF strength. This enhancement of the field close to the MP indicates pile-up of the magnetic field pile up against the low-shear MP, thus consistent with the plasma measurements which indicate the presence of a PDL as a result of field pile-up.
Figure 1c displays the bulk flow speed. The Figure shows that the flow slows down initially as one moves from the bow shock toward the MP. However, close to the MP where the plasma beta is less than unity (Fig. 1g), this trend reverses and the flow speeds up as the MP is approached. Since the solar wind flow speed, measured by IMP-8, was rather steady during this interval, this flow enhancement represents an acceleration of the tangential magnetosheath flow near the MP.
In the LLBL interval (19:55-20:50 UT), the plasma parameters such as the proton density, temperature, and flow velocity are highly fluctuating, and the plasma flow direction vp in a major portion of the LLBL is nearly aligned with the magnetosheath flow.
High Magnetic Shear Event
Figure 2 shows an overview of the inbound pass from the dusk bow shock to the plasma sheet on Aug 19, 1995. At 20:40 UT, the satellite crossed the bow shock and moved into the magnetosheath. The MP was encountered at ~22:40 UT where a magnetic shear of ~120 degrees was observed. From 22:40 UT to 23:05 UT, the spacecraft was in the LLBL. A feature similar to the previous low-shear example is the steadiness of the flow angle vp throughout the entire magnetosheath.
Figure 2. Overview of the inbound pass on Aug 19, 1995. The format is the same as in Figure 1. IMP-8 (R= 41 RE, Lat= 35 degrees, LT= 4.2 hrs) solar wind measurements are displayed with a 20 min shift backward in time. The IMF magnitude is displayed with an amplification factor of 3.2.
As in the previous example, the density just downstream of the bow shock (20:40-21:10 UT) is higher than in the remainder of the magnetosheath. However, unlike the low-shear case, the density variations near the bow shock in this case seem to be associated with solar wind density variations observed by IMP-8. Also, the evidence for the presence of a PDL next to this high-shear MP is not so clear since the ion temperature does not decrease toward the MP, and the pile-up of the magnetic field against the MP is less evident.
In the LLBL interval, the flow in most of the region is aligned with the magnetosheath flow.
We have investigated two WIND passes across the dusk magnetosheath, MP, and LLBL. In the first case, the magnetic shear across the MP was less than 30 degrees. In the other case, the magnetic shear across the MP was ~120 degrees. Below we discuss the implications for the physical processes in the flank regions.
Structure of the magnetosheath
Previous reports have indicated that a plasma depletion layer is generally present in the magnetosheath next to the subsolar low-shear MP, but is either absent or weak next to the high-shear MP. This finding was interpreted as evidence for a rather closed (open) MP when the magnetic shear across the local MP is low (high). The present observation indicates that a PDL is also present on the flank MP when the local magnetic shear is low. The high-shear case presented in this paper does not give conclusive evidence for the presence or absence of a PDL. Examinations of the remaining 17 high-shear flank MP crossings by WIND have not produced any clear evidence for a PDL either.
Plasma properties in the LLBL, and their relationships to the external magnetosheath plasma
It was pointed out in section 3 that the plasma parameters and the magnetic field in the LLBL interval (19:55-20:50 UT) are highly fluctuating. The fluctuations in the density, however, is correlated with those of the plasma flow speed, while they are anti-correlated with temperature fluctuations. These correlations suggest that the LLBL profiles may be smooth and gradual, and the fluctuations are due to the oscillatory motion of the MP instead. To investigate such a possibility, we examine the relationships between the various parameters measured in the LLBL. Figure 3a shows the proton temperature as a function of the proton density for the time interval 19:40-21:00 UT for the low-shear case, and for the interval of 22:30-23:30 UT for the high-shear case. These intervals contain the entire LLBL, in addition to short intervals of the magnetosheath and the plasma sheet. We find that the relationship between the temperature and the density in the LLBL is almost one to one.
Figure 3. Scatter plots of proton temperature (panel a), flow angle (panel b), and flow component VM (panel c) vs. the proton density. The left and right panels show the low- and high-shear cases, respectively.
Fig. 3b shows the flow angle vp as a function of the density. It is noted that for the low shear case, the flow direction jv in the outer part of the LLBL (Np ~ 4-17 cm-3) is nearly aligned with the flow direction in the magnetosheath (Np> 17 cm-3). In the high-shear case, the portion of the LLBL where the flow direction is similar to that of the magnetosheath is even wider. Figure 3c shows the east-west component of the velocity tangential to the MP, VM. A clear trend for the density and the magnitude of VM to be correlated is seen, i.e., the LLBL flow speed decreases gradually as the plasma sheet is approached. In the low-shear case, the decrease of the flow speed across the LLBL starts at the MP whereas in the high-shear case, the it begins somewhat Earthward of the MP.
Theoretically, the existing models of the LLBL involve either viscous or magnetic reconnection processes. Eastman et al.  invoked viscous interactions as the formation mechanism. A quantitative model was later developed by Sonnerup  which includes couplings to the ionosphere via field aligned current. In the viscous model the plasma parameters are expected to vary smoothly and gradually across the LLBL and the flow in the LLBL is expected to be aligned with the magnetosheath flow.
Existing models of the LLBL involving reconnection are less quantitative, with the exception perhaps of the model by Song et al.  for the low-shear LLBL. In that model, the transfer takes place at high latitude via cusp reconnection. The LLBL profiles are expected to be step-like.
Several features of our LLBL observations seem to be consistent with viscous models: (1) the flow direction is the same as in the magnetosheath and the flow speed decreases gradually across the LLBL; (2) smooth and continuous variations in plasma parameters across the LLBL. Step-like profiles as predicted by the Song et al. model would have resulted in clusters of points in the Tp-Np plot.
It should be pointed out that the gradual variation of the flow speed across the LLBL has also been reported for the tail flank LLBL [Fujimoto et al., 1996]. They suggest, however, that reconnection may still play an important role in the formation of the LLBL.
One difficulty associated with the viscous model of the LLBL is that the flow characteristics observed in the LLBL seem to be independent of the external magnetosheath orientation. The difficulty lies in the fact that high-shear LLBL has been shown in previous studies to be at least partially on open field lines [e.g., Mitchell et al., 1987; Hall et al., 1991]; a feature which is more naturally explained by reconnection.
Thus while the LLBL profiles favor diffusive processes, the topology of the region suggests a significant role for reconnection in the formation of the boundary layer on open field lines when the magnetic shear is high.
We thank Ron Lepping, the principal investigator for the Magnetic Field Investigation (MFI) instrument, and the MFI team for providing the magnetic field data used in the present analysis. This research was funded in part by NASA Contract NAS5-30366 and NASA grant NAG5-2815 at U. C. Berkeley.