R B Sheldon (POLAR/CEPPAD Team)

On January 10, 1997 a coronal mass ejection (CME) arrived at the Earth's orbit causing a mild shock and initiating a small (Dst $<$ 85) magnetic storm. Within 18 hours, however, the relativistic, outer radiation belt electrons experienced a 1000-fold enhancement in intensity. This may account for the Telstar401 failure 24 hours later. A second solar wind density enhancement occurred on January 12, during the storm recovery phase. It did not result in any Dst enhancements, but again, caused a second injection of MeV electrons. This example, like the April 15, 1996 event, had an abrupt, ``clean'' onset of MeV electrons with little or no Dst response, and clarifies the analysis of the acceleration mechanism. As in the April event, this enhancement reveals a source region at L$>$10, which we identify with the geomagnetic cusp. Unlike the April event, the POLAR orbit did not fly through the source region in January, since POLAR encounters the southern cusp at too low an altitude.

We do not believe it was a chance coincidence that MeV electrons and Dst were anti-correlated in both of these storms (as well as the October 23, 1996 Dst storm). This is a natural result of the conditions required for the acceleration of MeV electrons or keV ions respectively. We believe northward IMF, high solar wind speed (which produces energetic tails on the magnetosheath population), relatively high solar wind pressure, and a highly tilted cusp are the necessary prerequisites for an MeV electron enhancement, whereas Dst storms require southward IMF and an ``untilted'' cusp, along with enhanced solar wind pressure, to produce their maximum impact. Thus we argue for a second classification of ``hard rad'' storms which are clearly distinct from the traditional Dst criteria, and are more relevant for operators of spacecraft and forecasts of space weather.