Effects of a high-density plasma sheet on ring current development during the November 2-6, 1993, magnetic storm

The growth and recovery of the November 2-6, 1993 magnetic storm was simulated using a drift-loss ring current model that was driven by dynamic fluxes at geosynchronous orbit as an outer boundary condition. During the storm main phase, a high-density plasma sheet was observed by the Los Alamos National Laboratory geosynchronous satellites to move into and flow around the inner magnetosphere over a period of similar to 12 hours [Borovsky et al., 1997; this issue] during the storm main phase. Densities at the leading edge of this structure reached 3 cm(-3) as compared with more typical values <1 cm-3. The factor of 3 change in the plasma sheet density from quiet to active times produced a factor of 3 enhancement in the strength of the simulated ring current. In addition, a short-timescale recovery in the Dst index at 1600 UT on November 4 was driven by changes in the outer boundary condition and appeared even in the absence of collisional losses. An overshoot in the minimum Dst* occurred in the simulated ring current compared with observed values at similar to 0200 UT on November 4 and is taken as evidence of a loss process not included in the ring current-atmosphere interaction model (RAM). The storm onset was associated with a compression of the entire dayside magnetopause to within geostationary orbit starting at 2307 UT and continuing for a half hour. It is suggested that a possible additional loss may have resulted as ions drifted to the compressed dayside magnetopause. In fact such losses were found in another simulation of the inner magnetosphere for the same storm by Freeman et al. [1996]. The energy supplied to the inner magnetosphere, relative to the total energy input during this magnetic storm, was examined by comparing two widely used energy input functions, the epsilon parameter [Akasofu, 1981] and the F parameter [Burton et al., 1975] against energy input to the ring current model based on geosynchronous plasma observations at the outer boundary. It is found that the epsilon parameter [Akasofu,1981] overestimates the ring current energy input compared to the drift-loss model by almost an order of magnitude during the main phase. However, the integrated energy input from epsilon, over the 4 day interval of the storm, is in very good agreement with the total energy input inferred from observations. On the other hand, F more closely approximates the magnitude of the ring current energy input alone as calculated in the drift-loss model. An energy budget is constructed for the storm that shows energy inputs from the solar wind and energy dissipation due to ring current buildup and decay, auroral electron precipitation, Joule heating, ion precipitation, and energy storage in the magnetotail in reasonable balance. The ring current energy input accounts for only 15% of the total dissipated energy in this storm interval. A more complete energy budget that extends to November 11,1993, was compiled by Knipp et al. [this issue].