An empirical model for prediction of geomagnetic storms using initially observed CME parameters at the Sun

In this study, we discuss the general behaviors of geomagnetic storm strength associated with observed parameters of coronal mass ejection (CME) such as speed (V) and earthward direction (D) of CMEs as well as the longitude (L) and magnetic field orientation (M) of overlaying potential fields of the CME source region, and we develop an empirical model to predict geomagnetic storm occurrence with its strength (gauged by the Dst index) in terms of these CME parameters. For this we select 66 halo or partial halo CMEs associated with M-class and X-class solar flares, which have clearly identifiable source regions, from 1997 to 2003. After examining how each of these CME parameters correlates with the geoeffectiveness of the CMEs, we find several properties as follows: (1) Parameter D best correlates with storm strength Dst; (2) the majority of geoeffective CMEs have been originated from solar longitude 15 W, and CMEs originated away from this longitude tend to produce weaker storms; (3) correlations between Dst and the CME parameters improve if CMEs are separated into two groups depending on whether their magnetic fields are oriented southward or northward in their source regions. Based on these observations, we present two empirical expressions for Dst in terms of L, V, and D for two groups of CMEs, respectively. This is a new attempt to predict not only the occurrence of geomagnetic storms, but also the storm strength (Dst) solely based on the CME parameters.