We have reported new observations of the magnetospheric responses to a typical CME incident upon the Earth.
The first response observed was the propagation of the MHD compression through the polar cap, displacing the geopause inward past POLAR, immersing it in mantle plasma. The mantle flows consisted of a mixture of hot magnetosheath protons (and He) reflected from the cusp region, with relatively cold O outflows of similar parallel velocity. The mantle was bursty, apparently in response to gusty solar wind conditions.
The second response observed was the arrival at POLAR of the leading edge of what we refer to as an ionospheric mass ejection. Relatively cold and slow plasma, principally O, but including light and molecular ions at times, was observed to flow from the ionosphere into the polar lobes, indicating energy inputs that evidently originated with the incidence of the CME, and subsided after it passed. We interpret the change from an H/O mixture to an O-dominated outflow as resulting from the effects of CME incidence upon the magnetosphere, and the dissipation of energy throughout the topside ionosphere in the dayside auroral zone or cleft region, likely by field-aligned current intensifications.
Topside ionospheric heating and outflow was observed in the dayside southern hemisphere auroral zone after several hours of magnetospheric buffeting by strong and variable solar wind dynamic pressure (and also after the development of negative IMF .) Plasma upwelling was much stronger in flux (see below) during this perigee pass than during the preceding pass. The POLAR apogee pass following this one early on 25 Sept. exhibited O-dominated outflow throughout the pass, but at significantly lower flux levels. Conditions similar to those before the event reestablished themselves in the polar cap outflows about two orbits (36hrs) after the leading edge of the CME arrived. Higher time resolution observations of the low altitude heating and field-aligned currents are available from the FAST spacecraft [R. Strangeway, personal communication].
To place these results in broader context, we plot in Figure 2 the relationship observed during the DE-1 epoch, between peak upwelling flux and solar wind dynamic pressure variability [Pollock et al., 1988]. A subset of the O flux events used in the [Pollock et al., 1990] study were selected, for which solar wind and magnetic field data sets from the ISEE-1 spacecraft could be obtained from the NSSDC. The 1 min average peak upwelling flux was correlated with prior hour-averaged interplanetary conditions, delayed to the magnetopause. The Akasofu dynamo power parameter , the Reiff polar cap potential proxy , the plasma momentum flux , and its variability were studied. The peak upwelling O flux was found to be best correlated with the variability of the dynamic pressure at the magnetopause (R=0.76). Correlations with the IMF were poor (R ±0.1-0.2), suggesting that the energy coupling to the dayside ionospheric outflows is controlled by the dynamic pressure and relatively unaffected by dayside reconnection.
The perigee pass outflow fluxes from the present event have been plotted with open symbols in Figure 2. The present observations lie below but otherwise follow the trend established by the DE-1/ISEE-1 study. The 1981-2 DE-1 data were obtained in a period of average 10.7cm flux 208  while the present data were obtained at a time of lower solar activity with 10.7cm flux = 136. The F10.7 effect on O outflow noted by Yau et al.  suggests a factor of 2 difference between 1981 and Sept. 1998. Other factors, such as the recent history of internal geomagnetic activity, may have contributed to the somewhat larger ratio of the fluxes for these two epochs.