Ring current asymmetry from global simulations using a high-resolution electric field model

[1] We study ring current evolution during the 10 January 1997 geomagnetic storm, comparing results from two inner magnetospheric convection electric field models: (1) the Kp-dependent Volland-Stern (V-S) model and (2) the high spatial and temporal resolution assimilative mapping of ionospheric electrodynamics (AMIE) model, coupled with our global ring current model. We have added a penetration electric field (driven by partial ring current closure in the ionosphere) to the AMIE model which improves the agreement at low L shells with Polar/EFI measurements, and we mapped the modified AMIE convection electric potentials (MACEP) to the equatorial plane. Both V-S and MACEP convection models predict a very asymmetric local time distribution of ring current energy density during the main and early recovery phase of the storm. However, the peak of the medium energy ions is located during the main phase near midnight when the MACEP model is used, while it is located near dusk using V-S. In both models the energy density peak is located near dusk during the early recovery, and the ring current becomes symmetric during the late recovery phase. Ring current injection is larger, penetrating to lower L shells, and the Dst index is significantly better reproduced using MACEP rather than using V-S model. We compare model results with Polar data and find reasonably good agreement with both models at larger L shells and near dawn. The enhanced storm time distributions at low L shells near dusk are better modeled with MACEP. Electromagnetic ion cyclotron (EMIC) waves are predominantly excited near Dst minimum, have larger wave gain, and cause stronger ion precipitation using the MACEP model. In this case the calculated ion precipitation patterns exhibit features consistent with storm time Polar/IPS observations and show enhancements within the plasmaspheric bulge and along the duskside plasmapause. In the dusk to midnight sector at L > 5, however, there seems to be a need to include the effect of an additional scattering process during highly active conditions.