Equatorial trench at the magnetopause under saturation

A. Dmitriev1,2 and A. Suvorova3,2

1
Institute of Space Science, National Central University, Chung-Li, Taiwan
2
Institute of Nuclear Physics, Moscow State University, Moscow, Russia
3Center for Space and Remote Sensing Research, National Central University, Chung-Li, Taiwan

Short title: MAGNETOPAUSE TRENCH

Abstract Magnetic data from GOES geosynchronous satellites were applied for statistical study of
the low-latitude dayside magnetopause under a strong interplanetary magnetic field of southward
orientation when the reconnection at the magnetopause was saturated. From minimum variance
analysis, we determined the magnetopause orientation and compared it with predictions of a
reference model. The magnetopause shape was found to be substantially distorted by a duskward
shifting such that the nose region appeared in the postnoon sector. At equatorial latitudes, the shape
of magnetopause was characterized by a prominent bluntness and by a trench formed in the postnoon
sector. The origin of distortions was regarded in the context of the storm-time magnetospheric
currents and the large-scale quasi-state reconnection at the dayside magnetopause.

Keywords: magnetopause, magnetospheric currents, geomagnetic storm

1
1. Introduction saturated by a ‘‘plasmaspheric effect’’: high-density
It is widely accepted that the subsolar magnetopause magnetospheric plasma flows from plasmasphere into
moves earthward due to the reconnection of the magnetopause reconnection site and mass loads the
geomagnetic field with a southward component reconnection such the reconnection rate is locally
(negative Bz) of interplanetary magnetic field (IMF). reduced in the subsolar region. Based on statistical
Stronger southward IMF results in more intense analysis of the magnetopause crossings by
reconnection and, thus, smaller distance to the geosynchronous satellites, Suvorova et al. [2003]
magnetopause. However, when the magnitude of proposed that the reconnection saturation is caused by
negative IMF Bz exceeds a certain threshold of about 20 an enhanced thermal pressure of the magnetospheric
nT, the magnetopause distance does not decrease any plasma and ring current particles during strong
more, i.e. the southward IMF reconnection is saturated magnetic storms. This idea is supported by a recent
[e.g. Yang et al., 2003; Suvorova et al., 2005; Dmitriev model of geosynchronous magnetopause crossings
et al., 2011]. Statistical studies indicate that the IMF Bz (GMCs) [Dmitriev et al., 2011]. Namely, the minimal
threshold for saturation changes from about -20 nT to solar wind pressure required for GMCs under saturation
about -10 nT with increasing geomagnetic activity decreases with the storm-time Dst-variation, which is
[Dmitriev et al., 2011]. The effect of reconnection produced by the intense ring current.
saturation is still poorly investigated because of limited From hybrid MHD and kinetic multi-fluid simulations,
amount of experimental data. Winglee et al. [2008] found that for southward IMF, the
The saturation effect is studied using global MHD subsolar magnetopause is not smooth, as predicted by
simulations. Siscoe et al. [2004] show that at the Dungey [1963] or global MHD, but instead it is the
stagnation point (subsolar magnetopause), the region 1 rippled blunted surface suggested by the multi-X line
field-aligned currents (FAC) cause depletion of the models, and this rippling has implications for spacecraft
geomagnetic field as well as substantial changing of the observations. It means that a slow moving spacecraft
shape of magnetopause such that a dimple develops at will have multiple magnetopause crossings even under
the stagnation point and the magnetopause becomes steady solar wind conditions, with the depth of
blunt and the bow shock recedes. These effects lead to penetration dependent on the size of the flux rope and
decrease and limitation of the reconnection at the its displacement relative to the spacecraft.
dayside magnetopause. Analyzing geomagnetic field Therefore, numerical and empirical models predict that
observations during the strong southward IMF, Ober et the reconnection saturation under a large southward
al. [2006] have found an indirect support of that IMF should be accompanied by prominent distortions of
mechanism. They observe a strong dominance of the the subsolar magnetopause such as a dimple or rippling.
region 1 FAC generated magnetic fields both at high In the present paper, we study the shape of the dayside
and low latitudes and diminishing of the Chapman- low-latitude magnetopause with using GMCs occurred
Ferraro current system. Two sets of spacecraft during strong magnetospheric disturbances. The method
observations at high and low latitudes suggest that the of GMC analysis is described in Section 2. The
dayside magnetopause assumes a very blunt shape magnetopause shape during strong southward IMF is
consistent with the predictions of the MHD simulations. analyzed in Section 3. The results are discussed in
However, Nakano and Iyemori [2003] show that during Section 4. Section 5 gives conclusions.
magnetic storms (when all the magnetospheric currents
have been developed), the magnitude of region 2 2. Geosynchronous magnetopause crossings
Birkeland currents is comparable with the region 1 GMCs were collected by using magnetic field data
currents in postnoon and premidnight sectors and, thus, acquired from GOES 8, 9, and 10 geosynchronous
the contribution of the region 1 FAC to the dayside satellites in the time interval from 1994 to 2001. The
geomagnetic field is reduced by the region 2 currents. method of GMC identification is described in detail by
This finding supports the experimental result by Nagai Suvorova et al. [2005]. Briefly, we used high-resolution
[1982], who found that no appreciable variations in the (~1-min) ISTP data
dayside geosynchronous magnetic field are associated (http://cdaweb.gsfc.nasa.gov/cdaweb/istp_public/) from
with the Birkeland currents during geomagnetic storms. GOES satellites and solar wind data from upstream
Modeling the dayside magnetopause by a method of monitors Geotail, Wind, and ACE. A magnetopause
Artificial Neural Network, Dmitriev and Suvorova crossing was identified, when one of two requirements
[2000] found a "dimple" arising in the subsolar region were satisfied: (1) the GOES magnetic field deviated
when IMF Bz < -10 nT. Following to Rufenach et al. significantly from the geomagnetic field and (2) the By
[1989] they interpreted the feature as a result of the and Bz components measured by the GOES correlated
magnetopause erosion. The dimple is a large-scale local with the corresponding IMF components measured by
structure. It is unlikely that such structure is formed by an upstream monitor.
the cross-tail and/or field-aligned currents, which Figure 1 shows an example of GMC identification by
produce a global magnetic effect in the dayside GOES-10 during time interval from 20 to 21 UT on 8
hemisphere. Hence, it should be another phenomenon April 2001. In the dayside magnetosphere, the
responsible for that distortion. geomagnetic field is characterized by a dominant
Using high-resolution 3-D MHD simulations Borovsky northward component Hp > 100 nT. When GOES-10 is
et al. [2008] found that the reconnection can be located in the magnetosheath from 2012 to 2029 UT,
2
from 2043 to 2052 UT, and from 2055 to 2058 UT, the and z = 0 Re) can be different from the nose point,
component Hp turns to southward and the total where the normal nr = (1., 0, 0).
magnetic field H decreases. The By and Bz components, We can see that the measured normal to the
measured by GOES-10 in the magnetosheath, magnetopause is substantially different from the
demonstrate good correlation with the IMF By and Bz reference one. The unusual dawnward deviation of the
components measured by ACE upstream monitor with a normal in the XY plane cannot be explained by the
time delay of 40 min. effect of geomagnetic tilt appearing in the XZ plane.
The time delay for solar wind propagation was The deviation can also result from tilted interplanetary
determined independently from cross-correlation fronts or from time-dependent reconnection and
between the solar wind total pressure Psw and a multiple flux transfer events (FTEs) formed during
pressure PDst, deduced from 1-min SYM-H index strong southward IMF [Omidi et al., 2009]. For the
(equivalent to Dst variation) with expression GMC at 2029 UT, the upstream solar wind did not
∆Dst = b ∆PDst [e.g. Burton et al., 1975]. The exhibit any front of interplanetary discontinuity. Hence,
the unusual magnetopause orientation might result from
coefficient b was calculated from a linear regression of FTE occurred during prolonged interval of strong
the observed Dst and Psw. For the present time interval, negative Bz of ~ -50 to -100 nT in the magnetosheath as
the cross-correlation between Psw and PDst was ~0.56 measured by GOES-10. Here we are unable to verify
that proved our choice of the time delay. The Psw was and analyze the FTE effect, because GOES satellites do
calculated as the sum of the solar wind dynamic, not provide any plasma data. Instead, we perform a
thermal and magnetic pressures. Note that during strong statistical study of deviations of the measured
geomagnetic disturbances the contribution of the magnetopause normals from the nominal ones.
thermal and magnetic pressures to the Psw can reach up Figure 2 shows a scatter plot of the collected GMCs in
to 30%. aberrated GSM coordinates. The crossings occupy a
The magnetopause has a systematic distortion related to wide longitudinal sector from -30° to 50° and latitudinal
the Earth rotation around the Sun with orbital speed of sector from -25° to 20°. The wide latitudinal spread is
~30 km/s. In order to eliminate the aberration effect, all due to a superposition of 23.4° tilt of the Earth’s
the coordinates and vector parameters were converted in rotation axis to the ecliptic plane and ~11° tilt of the
the aberrated GSM (aGSM) coordinate system. Here we geomagnetic dipole relative to the rotation axis. The
take into account both the orbital velocity of the Earth wide angular spread of GMCs makes possible statistical
and non-radial solar wind propagation. The aberration is studies of the shape of dayside low-latitude
represented by angles aY and aZ of rotation about the Z- magnetopause.
GSE and aberrated Y-GSE axes, respectively [e.g.
Dmitriev et al., 2003]. 3. Magnetopause shape at low latitudes
The orientation of the magnetopause was determined Figure 3 shows projections to the equatorial plane of the
from the rotation of magnetic field across the magnetopause normals observed during GMCs relative
magnetopause using minimal variance analysis (MVA). to aGSM X-axis and relative to the reference
Note that the results of MVA vary by the length of the magnetopause calculated for the solar wind consitions
time interval chosen for the analysis. If the results vary accompaning the GMCs. The orientation is
greatly for different time intervals, they are unreliable. characterized by a wide random scatter. Significant
By varying the boundaries of the time intervals for each inclinations of the magnetopause normal are revealed in
GMC, we found time ranges for which the results of ~15° vicinity of the nose point (mlon = 0 and mlat = 0).
MVA changed slightly and gradually (i.e. the solution As we mentioned above, very strong magnetopause
of MVA was stable). By this way, we calculated a distortions can result from tilted interplanetary fronts
normal to the magnetopause, i.e. local orientation of the and from multiple FTEs formed during strong
magnetopause. In Figure 1, the magnetopause normals southward IMF. These effects are beyond the scope of
have been determined for GMCs at 2013, 2029, and the present study. Hence for convenience, we exclude
2043 UT. The GMCs at 2053, 2056 and 2058 UT were from further consideration the normals deviated
very short and, hence, the determination of normals was strongly (with a tilt >45°) from the reference normals.
difficult. As one can see in Figure 3a, the number of GMCs with
As an example, we consider the normal to the subsolar strongly deviated normals is small and practically all of
magnetopause (lat = -7.1°, lon = 1.6°) determined in the them are indicated by blue bars, i.e. they occur either
time range from 2026 to 2032. The normal was n = under weak southward IMF (Bz > -5 nT) or under very
(0.73, -0.67, 0.03), i.e. it was pointed sunward and strong solar wind pressures (Psw > 21 nPa).
dawnward. Using a model by Lin et al. [2010] Solar wind and geomagnetic conditions observed by
(hereafter reference model) under the current upstream upstream monitors during GMCs are presented in
solar wind conditions, we calculated a reference normal Figure 4. One can see that the Dst variation (see Figure
nr = (0.99, 0.02, -0.09). As expected, the reference 4a) is mostly negative and large (-310 < Dst <30 nT),
normal is mainly directed sunward. Note that the model i.e. vast majority of GMCs occurs during magnetic
predicts GMCs very well [Dmitriev et al., 2011]. In storms. The upstream solar wind conditions (see Figure
addition, the model predicts a north-south asymmetry of 4b) vary in a very wide dynamic range: -40 < Bz < 40
the magnetopause related to the tilt angle of nT and Psw >5 nPa. The conditions are restricted rather
geomagnetic dipole. Hence for the reference model, the sharply by a lower envelope boundary, below which
subsolar point (a point with coordinates aGSM y = 0 GMCs are not observed. The envelope boundary

3
corresponds to minimal solar wind conditions required magnetopause might be rather symmetric around a new
for GMCs and can be presented numerically by the X’-axis, which is shifted toward dusk. Such
following expression [Suvorova et al., 2005]: representation is in a good agreement with properties of
the magnetopause dawn-dusk asymmetry reported by
16.2 Dmitriev et al. [2004; 2005].
Psw = 21 − (1) An obtuse shape of the low-latitude dayside
1 + exp{0.2( Bz − 2.)} magnetopause under saturation is also revealed in the
meridional (X-Z) plane shown in Figure 6. Here we
The right branch (Bz → ∞), asymptotically approaching consider separately the prenoon and postnoon sectors.
Psw = 21 nPa, corresponds to a regime of pressure As one can see in the left panel of Figure 6, in the both
balance, under which the dayside magnetopause is only sectors, the normals in the northern (southern)
driven by solar wind pressure. The left branch (Bz → - hemisphere tend to tilt southward (northward) relative
∞) approaches Psw ~4.8 nPa under strong southward to the reference normals that indicate to a larger
IMF and is attributed to the regime of reconnection curvature of the magnetopause than that predicted by
saturation. the reference model.
In order to study the magnetopause shape under An asymmetry between the prenoon and postnoon
saturation, we select GMCs accompanied by upstream sector is revealed in the orientations of the
solar wind conditions with IMF Bz < -5 nT and Psw < magnetopause normal relative to aGSM X-axis in
21 nPa. For Psw > 21 nPa, both southward IMF and meridional plane, as shown in right panel of Figure 6. In
strong solar wind pressure can cause a GMC and, the prenoon sector, the normal demonstrates a regular
hence, it is hard to distinguish between the increase of the tilt with the magnitude of latitude. In the
magnetopause effects produced by two different postnoon sector, the orientation of normal exhibits more
upstream drivers. The threshold of IMF Bz < -5 nT is complicated behavior. Note that for the present case, the
chosen because of two circumstances. On the one hand, statistics of normals is slightly asymmetric relative to
this allows collecting enough statistics in close vicinity the equator: the Southern hemisphere is more
and inside the saturation regime. On the other hand, in represented by GMCs. It seems in the totally aberrated
Figure 1 we demonstrate that during IMF Bz ~ -5 nT coordinate system, the asymmetry has rather statistical
and strong solar wind pressure Psw ~ 18 nPa, the nature than any physical meaning. Moreover in average,
southward component of magnetosheath magnetic field, the orientation of the normals can be considered to be
which affects directly the magnetopause, can be symmetrical relative to the equator. At latitudes >10°,
extremely large (up to -100 nT) that is considered as a the normal is predominantly tilted outward from the
sufficient condition for the reconnection saturation [e.g. equator, i.e. poleward that is similar to that observed in
Siscoe et al., 2004; Borovsky et al., 2008]. Hence, as a the prenoon sector. In the latitudinal ranges from 5° to
first approach we can attribute the regime of 10°, the average tilt is close to zero that indicates to the
reconnection saturation at geosynchronous orbit to planar magnetopause perpendicular to the X-axis.
upstream solar wind conditions with IMF Bz < -5 nT At equatorial latitudes <5°, we find that 4 out of 5
and Psw < 21 nPa. normals are tilted toward the equator that indicates to a
Under saturation, the orientations of normal to the negative curvature of the magnetopause, or dimple.
magnetopause exhibit a specific pattern. In the Table 1 lists the basic characteristics of the GMCs in
equatorial (X-Y) plane, the normals are mainly tilted the dimple region. The GMCs occurred during 3
dawnward relative to the reference ones in both prenoon different magnetic storms, under very strong southward
and postnoon sectors as one can see in the left panel of IMF Bz < -14 nT and at various aGSM longitudes from
Figure 3. Comparing the orientation of normals with ~6° to 43°. GMCs #1 and 3 were accompanied by
aGSM X-axis (right panel of Figure 3), we find that at extremely strong southward IMF (Bz < -20 nT), which
longitudes from ~5° to 35°, the average orientation is is definitely proper for the reconnection saturation. For
almost parallel to the X-axis. In the prenoon sector, the the “abnormal” GMC #5, whose normal is tilted
normals are mainly tilted dawnward. The duskward tilt outward from the equator, Table 1 shows the weakest
is found in the late postnoon sector (lon > 40°). In magnitude of negative Bz = -14.3 nT and very large
Figure 4a one can see that most of these GMCs occur solar wind pressure Psw = 19.6 nPa. It might be that for
under not very high solar wind pressures Psw < 15 nPa the present case, the effect of very strong pressure
and during strong magnetic storms, namely at the main masked the effect of relatively weak southward IMF.
phase when Dst < -50 nT. Only three GMCs occur Figure 7 shows a sketch of meridional cut of the
during storm commencement (Dst > 0 nT) related to magnetopause in the postnoon sector. In contrast to the
strong Psw. reference model, the shape of magnetopause is not
Statistically, the observed orientation of the normals can smooth. In general, it is characterized by a larger
be represented as a bluntness extended to the postnoon curvature. But at low latitudes, the curvature increases
sector with simultaneous duskward skewing of both infinitely (plane) and then takes a negative value that
prenoon and postnoon wings of the magnetopause. A implies a dimple at equatorial latitudes. The dimple-like
sketch of the magnetopause cross-section in the aGSM orientation of the normals appears in a wide range of
equatorial plane is shown in Figure 5. One can see that longitudes. Geometrically, this feature can be described
under saturation, the equatorial magetopause is as an equatorial trench at the postnoon magnetopause.
characterized by a blunt nose region, which is located in
the postnoon sector. In the first approach, the dayside

4
Discussion current and thermal pressure can hardly explain the
From GOES observations of the magnetopause appearance of magnetopause bluntness and trench.
crossings during reconnection saturation, we have found Siscoe et al. [2004] proposed that the region 1 FACs
prominent distortions of the magnetopause in result in distortions at the dayside magnetopause.
comparison with the reference model by Lin et al. However, during magnetic storms the region 2 FACs
[2010]. Figures 5 and 7 show the sketches of can be intense such that the net contribution of the
magnetopause cross-section in the equatorial and Birkeland currents to the dayside magnetic field is small
meridional planes, respectively. We can combine the [Nakano and Iyemori, 2003; Nagai, 1982]. In addition
orthogonal cross-sections into a 3-dimensional sketch of to the Birkeland and ring currents, the cross-tail current
the dayside magnetopause as shown in Figure 8. An contributes to the geomagnetic field. During magnetic
important feature of the resulting 3-D picture is storms, the cross-tail current is very intense and its
equatorial trench at the magnetopause in the postnoon magnetic effect at the dayside magnetopause might
sector. In the prenoon sector, the magnetopause is become comparable with the geodipole field [Maltsev et
blunted and smooth. At low latitudes, the magnetopause al., 1996; Alexeev et al., 2001; Turner et al., 2000].
is shifted duskward such that the nose region occurs in The depletion of the dayside geomagnetic field by the
the postnoon sector. cross-tail current is well established [Maltsev et al.,
Note that the sketches presented in Figures 5 and 7 are 1996; Alexeev et al., 1996]. This depletion results from
based on limited statistics. Here we have to remind that storm-time intensification of the current and from
geosynchronous crossings of the magnetopause at low sunward motion of its inner edge such that the enhanced
latitudes in vicinity of noon are very rare because of current approaches to the dayside magnetopause. The
two circumstances [Suvorova et al., 2005]. In the noon geometry of magnetic field, produced by the storm-time
region, geosynchronous satellites spend most of time at cross-tail current at the magnetopause, is apparently
middle latitudes (>20°) and, in addition, the solar wind different from the dipole. The field is rather close to a
conditions required for GMCs are quite infrequent. linear configuration formed near the edge of a large-
Hence, in order to collect “sufficient” amount of scale flat electric current. The flat current produces
observations, which can definitely prove the results, we almost constant magnetic effect at different latitudes.
need to accumulate several decades of continuous data. The negative magnetic effect of the “linear“ magnetic
Up to date, only ~10 years of continuous observations field should be biggest near the magnetic equator,
of GMCs and solar wind conditions are available. Based where the dipole magnetic field is weakest. Hence, it is
on these limited data, we found a pattern of reasonably to propose that the bluntness of the dayside
magnetopause features, which were observed magnetopause is related to a strong contribution of the
independently during different time intervals and at cross-tail current. But the trench at the equatorial
different longitudes. The wide spatial and temporal magnetopause is hardly related to the global magnetic
ranges support the validity of limited statistics. In effect of the tail current.
addition, the revealed pattern does not contradict to the The equatorial trench in the nose region might be
existing findings and complements the previous produced by the strong region 1 FAC, when the region
expectations. 2 FAC is still undeveloped, i.e. in the beginning of
It is commonly accepted that the magnetopause, a magnetic storm. Such conditions might appear for GMC
current layer separating the magnetospheric magnetic #4 from Table 1. However for most GMCs, the
field from the interplanetary medium, is formed in magnetopause is characterized by a prominent dawn-
interaction of the solar wind with the geomagnetic dusk asymmetry that testifies to the strong partial ring
dipole. The interaction results in generation of a current and a well-developed magnetic storm activity on
complex geomagnetic field and a system of the main phase. Hence, the storm-time formation of the
geomagnetic currents. Hence, the solar wind plasma trench can be explained hardly by the effect of region 1
streams and IMF are primary external drivers of the FAC.
magnetopause and geomagnetic field [e.g. Parker, Under storm conditions, the equatorial trench might
1996; Vasyliunas, 2001]. Under reconnection regime, result from a large-scale quasi-state reconnection at the
the magnetosphere states in a dynamic equilibrium dayside magnetopause. During the reconnection, the
when the effect of strong southward IMF as external magnetic flux tubes are transferred from the equatorial
driver of the magnetopause is counteracted by effects of magnetopause to the tail. This transport produces an
the magnetospheric origin, which we can call internal indentation at the low-latitude magnetopause and an
magnetospheric drivers. expansion of the magnetopause at middle and high
Discussing possible drivers of the magnetopause dawn- latitudes. For very large negative Bz, the reconnection is
dusk asymmetry, Dmitriev et al. [2004] show that mostly effective in the nose region. We have shown that
asymmetries in the upstream solar wind drivers are under saturation, the nose region is shifted in the
unlikely responsible for the duskward skewing of the postnoon sector. Hence, the strong reconnection at the
magnetopause. It was concluded that the storm-time dayside magnetopause might be a possible mechanism
partial ring current is rather a source of the dawn-dusk of the trench formation in the postnoon sector. Under
asymmetry. The thermal pressure of hot plasma saturation, the equatorial magnetopause deflation,
populating the storm-time ring current was also produced by reconnection, can be balanced by a thermal
proposed as a force, which restricts and saturates the pressure of the hot plasma from the strong ring current.
magnetic reconnection at the magnetopause [e.g. Further experimental studies of the plasma and particles
Dmitriev et al., 2011]. However, the effects of ring

5
near the magnetopause during strong magnetic storms of the hot magnetospheric plasma from the strong
are required in order to verify this scenario. asymmetrical ring current into the pressure balance at
the dayside magnetopause during magnetic storms. The
Conclusions bluntness can result from the magnetic effect of the
intensified cross-tail current. The origin of equatorial
GOES observations of the magnetopause at
trench in the postnoon region might be related to a
geosynchronous orbit allow determining the shape of
large-scale quasi-state reconnection at the dayside
dayside magnetopause under saturation related to the
magnetopause.
effect of strong negative Bz. We found a prominent
dawn-dusk asymmetry consisting in duskward shifting
of the magnetopause such that the nose region is located Acknowledgements
in postnoon sector. At low latitudes, the magnetopause This work was supported by grants NSC-99-2111-M-
is blunt. In the postnoon sector, a trench is formed at the 008-013 and NSC-100-2111-M-008-016 from the
equatorial magnetopause. National Science Council of Taiwan and by Ministry of
The effects of reconnection saturation and duskward Education under the Aim for Top University program at
skewing can be explained by a substantial contribution National Central University of Taiwan.

Table 1. Characteristics of GMCs in the trench region

# Date and UT time mlat, mlon, deg PS, deg Dst, nT Bz, nT Psw, nPa
1 2000/4/06 1835 -4.7, 34.4 17 -94 -23.37 9.13
2 2001/3/31 1822 3.5, 32.8 14 -268 -17.53 15.36
3 2001/3/31 1651 -1.3, 6.6 15 -243 -28.65 5.14
4 2001/8/17 1937 -1.7, 42.8 21 -47 -18.67 20.37
5 2001/8/17 1947 -2.1, 42.0 21 -55 -14.25 19.57

References

Alexeev, I. I., E. S. Belenkaya, V. V. Kalegaev, Y. I. Feldstein, and A. Grafe (1996), Magnetic storms and magnetotail
currents, J. Geophys. Res., 101, 7737.
Alexeev, I. I., V. V. Kalegaev, E. S. Belenkaya, S. Y. Bobrovnikov, Ya. I. Feldstein, and L. I. Gromova (2001),
Dynamic model of the magnetosphere: Case study for January 9-12, 1997, J. Geophys. Res., 106, 25,683.
Borovsky, J. E., M. Hesse, J. Birn, and M. M. Kuznetsova (2008), What determines the reconnection rate at the
dayside magnetosphere? J. Geophys. Res., 113, A07210, doi:10.1029/2007JA012645.
Burton R. K., R.L. McPherron, and C.T. Russel (1975), An empirical relationship between interplanetary conditions
and Dst, J. Geophys. Res., 80, 4204.
Dmitriev A.V., and A. V. Suvorova (2000), Three-dimensional artificial neural network model of the dayside
magnetopause, J. Geophys. Res., 105(A8), 18,909-18,918.
Dmitriev A., J.-K. Chao, D.-J. Wu (2003), Comparative study of bow shock models using Wind and Geotail
observations, J. Geophys. Res., 108(A12), 1464, doi:10.1029/2003JA010027.
Dmitriev A.V., A.V. Suvorova, J.-K. Chao, Y.-H. Yang (2004), Dawn-dusk asymmetry of geosynchronous
magnetopause crossings, J. Geophys. Res., 109, A05203 doi: 10.1029/2003JA010171.
Dmitriev A., J.-K. Chao, M. Thomsen, A. Suvorova (2005), Geosynchronous magnetopause crossings on October
29-31, 2003, J. Geophys. Res., 110(A8), A08209, doi:10.1029/2004JA010582.
Dmitriev, A. V., Suvorova, A. V., J.-K. Chao (2011), A predictive model of geosynchronous magnetopause
crossings, J. Geophys. Res., 116, A05208, doi:10.1029/2010JA016208.
Dungey, J. W. (1963), The structure of the exosphere or adventures in velocity space, in Geophysics of the Earth’s
Environment, edited by C. De Witt et al., pp. 505–550, Gordon Breach, New York.
Lin, R. L., X. X. Zhang, S. Q. Liu, Y. L. Wang, and J. C. Gong (2010), A three-dimensional asymmetric
magnetopause model, J. Geophys. Res., 115, A04207, doi:10.1029/2009JA014235.
Maltsev, Y. P., A. A. Arykov, E. G. Belova, B. B. Gvozdevsky, and V. V. Safargaleev (1996), Magnetic flux
redistribution in the storm time magnetosphere, J. Geophys. Res., 101, 7697.
Nagai, T. (1982), Local time dependence of electron flux changes during substorms derived from multi-satellite
observations at synchronous orbit, J. Geophys. Res., 87, 3456.
Nakano, S., and T. Iyemori (2003), Local time distribution of net field-aligned currents derived from high-altitude
satellite data, J. Geophys. Res., 108(A8), 1314, doi:10.1029/2002JA009519.
Ober, D. M., N. C. Maynard, W. J. Burke, G. R. Wilson, and K. D. Siebert (2006), ‘‘Shoulders’’ on the high-
latitude magnetopause: Polar/GOES observations, J. Geophys. Res., 111, A10213, doi:10.1029/2006JA011799.
Omidi, N., T. Phan, and D. Sibeck (2009), Hybrid simulations of magnetic reconnection initiated in the
magnetosheath, J. Geophys. Res., 114, A02222, doi:1029/2008JA013647.
6
Parker, E. N. (1996), The alternative paradigm for magnetospheric physics, J. Geophys. Res., 101(A5), 10,587.
Rufenach, C. L., Jr., R. F. Martin, and H. H. Sauer (1989), A study of geosynchronous magnetopause crossings, J.
Geophys. Res., 94, 15,125.
Siscoe, G., J. Raeder, and A. J. Ridley (2004), Transpolar potential saturation models compared, J. Geophys. Res.,
109, A09203, doi:10.1029/2003JA010318.
Suvorova A., A. Dmitriev, J.-K. Chao, M. Thomsen, Y.-H. Yang (2005), Necessary conditions for the
geosynchronous magnetopause crossings, J. Geophys. Res., 110, A01206, doi:10.1029/2003JA010079.
Suvorova, A., A. Dmitriev, J.-K. Chao, Y.-H. Yang, M. Thomsen (2003), Necessary conditions for the
geosynchronous magnetopause crossings, Abstract on EGS-AGU-EUG Joint Assembly, Nice, France, 7-11 April
2003.
Turner, N.E., D.N. Baker, T.I. Pulkkinen, and R.L. McPherron (2000), Evaluation of the tail current contribution to
Dst, J. Geophys. Res., 105, 5431.
Vasyliunas, V. M. (2001), Electric field and plasma flow, what drives what? Geophys. Res. Lett., 28(11), 2,177.
Winglee, R. M., E. Harnett, A. Stickle, and J. Porter (2008), Multiscale/multifluid simulations of flux ropes at the
magnetopause within a global magnetospheric model, J. Geophys. Res., 113, A02209,
doi:10.1029/2007JA012653.
Yang, Y.-H., J. K. Chao, A. V. Dmitriev, C.-H. Lin, and D. M. Ober (2003), Saturation of IMF Bz influence on the
position of dayside magnetopause, J. Geophys. Res., 108(A3), 1104, doi:10.1029/2002JA009621.

Figure Captions

Figure 1. An example of magnetopause crossing identification by GOES-10 magnetic data from 20 to 21 UT on 8 April
2001 (from top to bottom): GOES-10 horizontal Hp (black solid curve) and total H (blue dotted curve) magnetic field;
solar wind pressure Psw (black solid curve) and Dst-deduced pressure PDst (blue dotted curve); aberrated GSM (aGSM)
Bz, By, and Bx components of magnetic filed measured by ACE (black solid curves) and GOES-10 (blue dotted curves);
solar wind aberration angles aY (black solid curve) and aZ (blue dotted curve), aGSM latitude mLat and local time MLT.
The Bz and By components measured by GOES are divided by 10 and Bx component is divided by 5. The vertical red
dotted lines indicate magnetopause crossing. The vertical blue dashed-dotted lines restrict the time interval for MVA
analysis.

Figure 2. Location of geosynchronous magnetopause crossings (GMCs) observed by GOES in aGSM coordinates. The
red dots correspond to IMF Bz < -5 nT and Psw <21 nPa.

Figure 3. Projection of the magnetopause normal, observed during GMCs, to the equatorial plane for all events (blue and
red segments) and for those during the reconnection saturation for IMF Bz < -5 nT and Psw < 21 nPa (red segments).
Left and right panels show the orientation, respectively, relative to the reference magnetopause normal nr and relative to
aGSM X-axis. The magnetopause is characterized by a strong dawn-dusk asymmetry and a prominent bluntness in the
postnoon sector.

Figure 4. Scatter plots of the solar wind pressure Psw during GMCs versus a) Dst variation, and b). IMF Bz in aGSM
coordinates The horizontal dashed line indicates the Psw = 21 nPa. The thick solid curve indicates an envelope boundary
of necessary conditions for GMCs [Suvorova et al., 2005]. The red circles correspond to the conditions proper for
reconnection saturation.

Figure 5. A sketch of the magnetopause cross-section in the aGSM equatorial plane: for the reference magnetopause
(dashed black curve) and the magnetopause under saturation (solid red curve). For the latter case, the magnetopause
shape is characterized by a prominent bluntness in the nose region shifted to the postnoon sector. The new axis of
symmetry X’ of the magnetopause is denoted by the red solid arrow, which is shifted duskward from the nominal X-axis
denoted by the black dashed arrow.

Figure 6. Projections of the magnetopause normal to the meridional plane observed under reconnection saturation (Bz < -
5 nT and Psw < 21 nPa): (left) in comparing with the normals to the reference magnetopause and (right) relative to the
aGSM X-axis in the prenoon (blue segments) and postnoon sectors (red segments). In the left panel, the orientation of
normals exhibits a strong bluntness of the magnetopause. In the right panel, the pattern of deviations in the postnoon
sector indicates a dimple at low-latitudes.

Figure 7. A sketch of the magnetopause cross-section in the aGSM meridional plane in the postnoon sector: for the
reference magnetopause (dashed black curve) and the magnetopause under saturation (solid red curve). For the latter
case, a dimple is formed at the low-latitude magnetopause.

Figure 8. A sketch of the dayside magnetopause under saturation. The magnetopause is skewed duskward. In the
afternoon sector, a trench is formed at low latitudes.

7
Figure 1. An example of magnetopause crossing identification by GOES-10 magnetic data from 20
to 21 UT on 8 April 2001 (from top to bottom): GOES-10 horizontal Hp (black solid curve) and total
H (blue dotted curve) magnetic field; solar wind pressure Psw (black solid curve) and Dst-deduced
pressure PDst (blue dotted curve); aberrated GSM (aGSM) Bz, By, and Bx components of magnetic
filed measured by ACE (black solid curves) and GOES-10 (blue dotted curves); solar wind
aberration angles aY (black solid curve) and aZ (blue dotted curve), aGSM latitude mLat and local
time MLT. The Bz and By components measured by GOES are divided by 10 and Bx component is
divided by 5. The vertical red dotted lines indicate magnetopause crossing. The vertical blue dashed-
dotted lines restrict the time interval for MVA analysis.

8
 20

10
Latitude, deg

-10

-20

-40 -30 -20 -10 0 10 20 30 40

Longitude, deg
Figure 2. Location of geosynchronous magnetopause crossings (GMCs) observed by GOES in
aGSM coordinates. The red dots correspond to IMF Bz < -5 nT and Psw <21 nPa.

9
 40 40

30 30

20 20

Longitude, deg
Longitude, deg

10 10

0 0

-10 -10

-20 -20

-30 -30

-40 -40
> >
nr X
Figure 3. Projection of the magnetopause normal, observed during GMCs, to the equatorial plane for
all events (blue and red segments) and for those during the reconnection saturation for IMF Bz < -5
nT and Psw < 21 nPa (red segments). Left and right panels show the orientation, respectively,
relative to the reference magnetopause normal nr and relative to aGSM X-axis. The magnetopause is
characterized by a strong dawn-dusk asymmetry and a prominent bluntness in the postnoon sector.

10
 Psw (nPa) 10

-350 -300 -250 -200 -150 -100 -50 0 50 100

Dst (nT)

a
Psw (nPa)

10

-40 -30 -20 -10 0 10 20 30

Bz-aGSM (nT)
b

Figure 4. Scatter plots of the solar wind pressure Psw during GMCs versus a) Dst variation, and b).
IMF Bz in aGSM coordinates The horizontal dashed line indicates the Psw = 21 nPa. The thick solid
curve indicates an envelope boundary of necessary conditions for GMCs [Suvorova et al., 2005]. The
red circles correspond to the conditions proper for reconnection saturation.

11
 Y

X’

Figure 5. A sketch of the magnetopause cross-section in the aGSM equatorial plane: for the reference
magnetopause (dashed black curve) and the magnetopause under saturation (solid red curve). For the
latter case, the magnetopause shape is characterized by a prominent bluntness in the nose region
shifted to the postnoon sector. The new axis of symmetry X’ of the magnetopause is denoted by the
red solid arrow, which is shifted duskward from the nominal X-axis denoted by the black dashed
arrow.

12
 20 20

10 10

Latitude, deg
Latitude, deg

0 0

-10 -10

-20 > -20 >

nr X
Figure 6. Projections of the magnetopause normal to the meridional plane observed under
reconnection saturation (Bz < -5 nT and Psw < 21 nPa): (left) in comparing with the normals to the
reference magnetopause and (right) relative to the aGSM X-axis in the prenoon (blue segments) and
postnoon sectors (red segments). In the left panel, the orientation of normals exhibits a strong
bluntness of the magnetopause. In the right panel, the pattern of deviations in the postnoon sector
indicates a dimple at low-latitudes.

13
 Z

X’

Figure 7. A sketch of the magnetopause cross-section in the aGSM meridional plane in the postnoon
sector: for the reference magnetopause (dashed black curve) and the magnetopause under saturation
(solid red curve). For the latter case, a dimple is formed at the low-latitude magnetopause.

14
 Z

trench
X

Figure 8. A sketch of the dayside magnetopause under saturation. The magnetopause is skewed
duskward. In the afternoon sector, a trench is formed at low latitudes.

15