5. Concluding remarks
Perhaps the most interesting outcome of the simulations is the appearance of the filamentary field-aligned currents connecting the region just behind the dipolarization front with the high-latitude ionosphere. In the previous section we have made the association between these currents and the discrete aurora. At this point, however, the association is far from definitive. The instabilities that generate the waves shown in Figure 11 need to be investigated further to determine the precise excitation mechanism and to characterize the wave lengths. It does seem as if the code is fully resolving the waves. If that is the case, then an additional mechanism, perhaps involving electron inertia [Hui and Seyler, 1992] is needed to further steepen the waves to produce electron inertial lengths thicknesses typical of auroral arcs.
Assuming the association between the filamentary currents and the discrete aurora, the code has reproduced many of the morphological features associated with substorms. Among them are the poleward expansion of aurora, the apparent dipolarization in the near-Earth plasma sheet and plasma sheet thinning during the growth phase and in the region tailward of the dipolarization. At no point, however, did there seem to be a defining moment, like an instability onset or triggering event to mark the beginning of a substorm. Rather, convection drove plasma earthward, the plasma splashed off the dipole field and the manifestations of the substorm expansion phase began to appear. Moreover, at no time during the simulations did magnetic reconnection seem to play a role. This is not to say that reconnection does not occur. Reconnection processes may well play a necessary role in maintaining the convection electric field that drives the system.
There has long been considerable debate over whether the magnetospheric substorm is a driven or an unloading process. Based on these simulations, we tend to agree with Fairfield et al. [1981] that it is both. Convection results in an accumulation of energy in the magnetotail, as well as causing substorm effects. As seen in the example shown in Figure 13, when convection is turned off, the tail field will collapse and substorm manifestations will still occur. In fact, it seems as if substorm manifestations will occur almost no matter what happens. What then is needed to produce a quiet magnetosphere and an absence of aurora? It seems that an absence of earthward tail flow is needed. A condition for this is that both an absence of driving convection and an adequate earthward pressure gradient are needed to balance the tension of the tail field A possible outcome of the dipolarization process would be to leave the magnetotail filled with sufficiently high-pressure plasma to resist earthward flow.
Another subject of debate relates to the possible role of the ionosphere in the substorm. In this simulation, the ionosphere is a passive conductor, and the substorm features originate from conditions completely external to the ionosphere. There are substorm phenomena, such as the current wedge [Akasofu, 1972], which of necessity must involve the ionosphere, because the wedge is formed by field aligned currents which complete their path through the ionosphere. The intensity of the westward electrojet and field-aligned currents undoubtedly depends on the ionospheric conductivity, which may be in turn influenced by discrete auroral ionization. But the simulations suggest that processes taking place in the magnetosphere initiate events.