Abstract
Characteristic patterns of the high-latitude, ionospheric electric potential (plasma convection) have been derived for various orientations of the interplanetary magnetic field (IMF), for epochs with and without the occurrence of magnetospheric substorms. These patterns were derived from all available, simultaneous DE-2 and solar wind data, using a least-error-fit technique. The results indicate that for each orientation of the IMF in the GSM Y-Z plane there are distinctive patterns, both with and without substorms in progress. In comparison to the non-substorm electric potential patterns, the substorm-epoch patterns tend to have a greater polar cap potential drop, and the distortion near midnight, the so-called Harang discontinuity, is much more pronounced during substorms. The changes in the electric field s magnitude and orientation near midnight appear to play an important role in the increase of the westward electrojet current during substorms.
Introduction
A critical key to understanding what drives magnetospheric substorms is to know how the global ionospheric electric potential patterns change during these substorms. In the past, the evolution of the ionospheric electric fields has been studied primarily by techniques involving the inversion of ground-based magnetometer data from multiple stations. While this method may yield reasonable results for the global, ionospheric electrojet current patterns, the calculations must rely on rough estimates of the ionospheric conductivity in order to derive the electric fields and potentials. The resulting electric field patterns are only as good as the conductivity estimate, and therefore are of limited utility. In this study the ionospheric electric fields and potentials that have been directly measured by satellite probes are used to derive the ionospheric potential patterns for time periods with and without substorms in progress.
Technique
The electric field data for this study are from the DE-2 satellite [Maynard et al. 1981], using all polar cap passes from the satellite s lifetime during which there were IMF measurements available from the ISEE-3 or IMP-8 satellites (2879 passes). These satellite passes were then divided into two groups, for measurements that were obtained when there was not a substorm in progress, and cases in which the measurements were obtained near the peak of substorms.
This substorm-dependent sorting was accomplished in two stages. In the first step all DE- 2 pass times were input into a computer program which automatically searched for substorms in the Auroral Electrojet digital data (as a manual search through all 2879 cases could take several weeks). More precisely, this program used a convolution pattern-matching to find cases where the AL index had the characteristic shape of a substorm [Weimer, 1994]. In order to be counted in the substorm group, the start of a satellite pass over the polar cap (lasting about 25 minutes) had to occur after expansion onset and no later than 50 minutes after onset, so that the measurements were obtained primarily around the maximum phase of the substorms, from late expansion phase into mid-recovery phase. The number of candidate substorm cases that were found by this method was 356.
As the automatic substorm search is not entirely foolproof, in a second stage this much smaller list was edited manually. Both the AU and AL indices vs. time were examined for each case for obvious substorm signatures. The end result of this two-stage sorting was only 239 satellite passes selected as corresponding to substorm times and 2640 passes in the other group without substorms. The editing process had guaranteed that only true substorms were included in the small, substorm group.
The satellite passes were then sorted into groups according to the angle of the IMF vector in the Y-Z plane, using overlapping bins spaced 22.5 degrees apart (16 steps) and 45 degrees wide. The average IMF for the 40 minute period preceding the mid-point of each pass was used. The data were also purged of cases having abnormally low or high IMF magnitudes, keeping cases having a tangential component (in the GSM Y-Z plane) with a magnitude in the range of 3 to 14 nT. All measurements of the electric potential along the multiple paths in each sorted group were then used to derive a map of the average electric potential over the polar cap for the given conditions. Using the same technique as in Weimer [1995], a least- error fit of the random measurements on the spherical surface was used to find the electric potential in terms of a spherical harmonic expansion:
where theta is a function of the geomagnetic colatitude, phi represents the magnetic local time (MLT), and Plm is the associated Legendre function.
The resulting average ionospheric electric potentials without and with substorms in progress are shown in Figure 1 and Figure 2. As substorms tend to not occur following a northward IMF, there were insufficient passes (less than 26) to get a surface fit for the substorm groups for the +Z IMF orientations. Therefore, these figures show only 9 patterns for the IMF angles from 90 degrees (+Y direction) to 180 degrees (-Z direction) to 270 degrees (-Y direction).
Discussion
Figures 1 and 2 show that there are clear differences in the electric potential patterns between the substorm and non-substorm cases. The substorm patterns tend to be rotated more clockwise (as viewed from over the north pole), with a much more pronounced Harang disconti nuity near midnight. The average polar cap potential drop (difference between peak maximum and minimum voltages) is also higher in the substorm patterns, but most of this difference appears to be due to the fact that the average IMF magnitude for the substorm cases is 7.0 nT, compared to 6.2 nT for the non-substorm cases. Therefore, the higher potentials appear to be more of a cause of more frequent substorms rather than effect produced by substorms.
However, there are subtle differences in how the potentials are distributed, as there appear to be substorm enhancements in the pre- and post-midnight regions. When the IMF is in the +Y direction, the natural orientation of the pattern results the substorm enhancement increasing the positive, dawn potential more than the dusk cell (compare the patterns at 157.5 degrees, for example). The situation is reversed for -Y orientations, but the effect is smaller since the peak of the dusk cell never moves beyond 19 MLT. Also, in the case where the IMF is in the -Y direction with almost zero Z component (270 degrees), the midnight-centered substorm enhancement causes a second peak in the dawn cell potential near 3 MLT, while the normal peak for this orientation remains near 7 MLT.
The averages of the simultaneous Auroral Electrojet indices (AU, AL, and AE) have also been evaluated for each group of satellite passes. The results of graphing these indices vs. the polar cap potential drop are shown in Figure 3. The solid lines show the non- substorm cases, drawn as a continuous loop as the IMF orientation moves from 0 degrees to 180 degrees and back to 0. The dashed lines show the substorm cases, for IMF orientations from 90 to 270 degrees. It is interesting that the AU index (middle graph) shows almost no break in continuity between the two sets, and is very linear. On the other hand, the AL index for the substorm line appears to be on a level above the non-substorm line, while the slope is roughly the same. As the slope of these lines should be proportional to ionospheric conductivity, this could be construed as evidence that the conductivity enhancement is not the primary cause of the increased electrojet current during substorms.
In order to get a picture of what the electrojet and field-aligned currents look like with and without substorms, these results have been combined with a numerical model of the ionospheric conductivity, from Hardy et al. [1987]. Maps of the Pedersen and Hall conductivities are shown in Figure 4 for Kp=3. There is one major problem in that this conductivity model was derived from satellite particle data that was sorted by 3-hr Kp indices; as it contains a mixture of all IMF orientations and substorm phases, this model is not self-consistent with the electric potential patterns that are derived from this study.
Nevertheless, the results shown in Figure 5 and Figure 6, for one IMF orientation, are illuminat ing. It is particularly gratifying that in the non-substorm case ( Figure 5) a continuous ring of upward (negative) current is obtained; this oval resembles the auroras that were imaged with the DE-1 satellite [Frank and Craven, 1988] much more than the upward current patterns that are usually obtained by the magnetometer inversion technique. The substorm case (Figure 6) also shows evidence of enhanced currents before and after midnight, and a hooked feature resembling the head of a westward surge. Also note that while the same conductivity model was used, the peak in the westward current density increased by 65%, while the polar cap potential drop had increased by only 38%. Therefore, a good portion of the westward current s increase appears to be due to a rotation of the electric field to a south-west orientation, so that there is a positive reinforcement to the westward electrojet from both the Hall and Pedersen components.
Summary
The ionospheric electric potential (plasma convection) patterns for substorm and non- substorm cases have been derived for various orientations of the IMF, using a least-error-fit technique. The results indicate that for each orientation of the IMF there are distinctive patterns, both with and without substorms in progress. In comparison to the non-substorm electric potential patterns, the substorm patterns tend to have a greater polar cap potential drop, and the Harang discontinuity distortion near midnight is much more pronounced during substorms. Most of the increased potential can be accounted for by an increase in the average IMF magnitude. However, the average AL index during substorms increases by an even greater amount in comparison to the non-substorm cases. Yet the slope of the AL vs. potential graph is not greater for substorms, indicating that an increased conductivity does not seem to be the major contributor to the substorm s enhanced electrojet. Instead, the changes in the electric field s magnitude and orientation near midnight appear to be important.
Acknowledgments. This research was supported by NSF grant ATM-9506169. The author thanks Nelson Maynard for helpful discussions and for use of the electric field data from the DE 2 satellite. The data from IMP 8 and ISEE 3 were provided by the National Space Science Data Center. Ron Lepping and Alan Lazarus are the Principal Investigators for the magnetometer and plasma instruments on IMP 8. Edward Smith and John Gosling are the Principal Investigators for the magnetometer and plasma instruments on ISEE 3.
References
Frank, L. A., and J. D. Craven, Imaging results from Dynamics Explorer 1, Rev. Geophys., 26, 249-283, 1988.
Hardy, D. A., M. S. Gussenhoven, R. Raistrick, and W. J. McNeil, Statistical and functional representations of the patterns of auroral energy flux, number flux, and conductivity, J. Geophys. Res., 92, 12,275-12,294, 1987.
Maynard, N. C., E. A. Bielecki, and H. F. Burdick, Instrumentation for vector electric field measurements from DE-B, Space Sci. Instrum., 5, 523-534, 1981.
Weimer, D. R., Substorm time constants, J. Geophys. Res., 99, 11,005-11,015, 1994.
Weimer, D. R., Models of high-latitude electric potentials derived with a least error fit of spherical harmonic coefficients, J. Geophys. Res., 100, 19,595, 1995.
Figure Captions
Figure 1. Electric potentials derived at nine IMF angles for non-substorm conditions. The angle is stepped from -90 ( -Y) through 0 ( +Z) to +90 (+Y) in 22.5 increments, as noted in the upper left corner of each graph. The numbers at the lower left and right corners show the minimum and maximum potentials in unit of kV.
Figure 2. Electric potentials derived at nine IMF angles for substorm conditions, in the same format as Figure 1.
Figure 3. Average Auroral Electrojet indices vs. polar cap potential. The solid lines show the curves for all 16 non-substorm patterns as the IMF angle rotates from 0 (+Z) to 180 degrees (-Z) and back to 0. The dashed lines show the curves for the 9 substorm patterns for IMF angles from 90 degrees (+Y), through 180 degrees, to 270 degrees (- Y).
Figure 4. Maps of Pedersen and Hall conductivities. Derived from the Hardy et al. [1987] model for Kp=3, these conductivities are used to calculate the currents shown in Figures 5 and 6.
Figure 5. (Top) Map of field-aligned current calculated from the average electric field for the non-substorm case at an IMF angle of 247.5 degrees. Upward currents are negative, shaded in blue. The maximum current densities in micro-A/m2 are indicated in the lower left and right corners. (Bottom) Map of horizontal current vectors, superimposed on the electric potential contour map. Peak currents in A/m are indicated below the graph, and the locations of the peaks are marked with the black squares.
Figure 6. Maps of field-aligned (top) and horizontal (bottom) currents for the substorm case, in the same format as Figure 5, for the same IMF orientation.
