Helio Research, Bolder, Colorado

On the basis of the location, orientation, and surrounding magnetic field configuration of an erupting filament observed on September 27, 1997, an Earth-encounter with a magnetic cloud having specific topological properties and an ensuing geomagnetic storm were predicted for September 30, assuming an average Sun-Earth transit time. A subsequent comparison with solar wind and geomagnetic data surrounding September 30 yielded a mixed level of success. Although a geomagnetic storm occurred on October 1, ironically, the strong southward magnetic field responsible for it was not encountered at the leading edge of the cloud, as predicted, but rather in the preceding sheath. In retrospect, the most likely sources of the southward field in the sheath were either a preceding CME or magnetic flux from high in the corona, not draped or turbulent flux from the ambient solar wind as is often supposed. With the aid of a flux rope model, we show that the elevation and polarity of the cloud axis and chirality of the cloud were predicted correctly but that the trajectory of the spacecraft (and Earth) through what was deduced to be the leg of a flux rope loop precluded encounter with its leading southward field. The storm began with a sudden commencement near 0100 UT on October 1, corresponding to passage of a shock-like discontinuity, and the Dst storm index reached a minimum of -108 nT 14 hours later at 1500 UT. The leading edge of the cloud, however, coincident with the onset of counterstreaming electrons confirming its identification, was not encountered until 1700 UT. The disagreement between the predicted and actual arrival time reflects not only the lack of speed prediction capabilities but also the inability to predict whether encounter will be head-on with the apex of a flux rope loop or, as deduced in this case, passage through one of its legs.

J. Geophys. Res., Vol. 106, No. A12, 29,185-29,194, 2001