We have identified the ionospheric response to a sudden southward turning of the IMF and found that the response is nearly instantaneous (<2 min) over a large portion of the dayside ionosphere. Prior to the identified transition, there was no indication in the high-latitude convection pattern that a change in IMF had occurred at the magnetopause. Subsequent to this time, there was a rapid enhancement of convective flow velocities that spread over the high-latitude ionosphere on a time scale of less than 30 s.
Observations from the GEOTAIL spacecraft, located near the subsolar magnetopause, WIND located in the upstream solar wind, and IMP8 located down the flank of the magnetosheath, provide evidence that the field lines of the new IMF state were draped over a large portion of the dayside magnetopause when reconnection began. To illustrate the draping of the new IMF at the time of reconnection at the magnetopause a minimum variance analysis was performed on the WIND magnetic field data and a planar surface containing the field lines determined. The planar structure was propagated at the solar wind speed, measured at WIND, for three time intervals. Where the planar structure passed into the magnetosheath it was propagated in a manner consistent with the flows predicted by the gas dynamic model of Spreiter and Stahara . Figure 4 shows the intersection of this plane with the X-Y plane for four times: 1551 UT, the observation time at WIND; 1652 UT, the observation time at GEOTAIL; 1701 UT, two minutes prior to the time that enhanced ionospheric convection was observed; and 1713 UT, the observation at IMP8.
The lines drawn in Figure 4 are undoubtedly somewhat in error due to the difficulties in determining the orientation of an assumed planar structure in the solar wind, propagating such a feature in the solar wind, and the unknown details of the flows in the magnetosheath near the magnetopause. However, even considering these errors it is apparent that, for this case, the field lines were significantly draped over a large portion of the dayside magnetopause when reconnection began. The draping was most dramatic in the northern hemisphere due to the vertical tilt of the magnetic plane (not shown). Our analysis does not require any delay in the onset of reconnection once the new IMF state has contacted the magnetopause.
In this particular case the draping of magnetosheath field lines over a large portion of the dayside magnetopause and an associated extended line of reconnection, could explain the observed nearly instantaneous, large-scale convection response to a change in the IMF. This picture is inconsistent with earlier models of single-point reconnection causing ionospheric responses that spread anti-sunward at a few km s-1. We do not dispute that the latter situation may, on occasion, occur, but we believe that our observations may help to resolve some of the ongoing controversy regarding rapid, large-scale ionospheric responses following changes in the IMF.