Geoeffectiveness of ICMEs during 1996-2009

Some 340 interplanetary coronal mass ejections (ICMEs) passed the Earth during 1996-2009, a period encompassing solar cycle 23, as cataloged by Richardson and Cane, Solar Physics 264, 189, 2010. We summarize the geomagnetic response to these ICMEs and their associated upstream sheaths and identify the parameters that are most closely related to the strength of this response. The average minimum value of Dst attained is -76~nT, while the most probable value is ∼-40 nT. Both exceptionally quiet and exceptionally strong activity levels are relatively rare. Some 26% of the ICMEs were associated with intense storms (Dst<-100 nT), and 6% with severe storms (Dst < -200~nT). As a complication, some of these important storms involved multiple ICMEs, interactions with corotating high-speed streams, and shocks moving through ICMEs. An excellent correlation (cc=-0.903) is found between the minimum value of Dst and the maximum value of the y component of the electric field (E_y) in the ICME or the upstream sheath based on 1-hour averaged solar wind data, characterized by Dst(nt)= -11.9E_y(mV/m)-13.3. A high degree of correlation (cc=-0.891) is also found with the maximum strength of the southward magnetic field component in the sheath or ICME, with Dst(nT) =-8.23B_s + 3.74. The correlation with ICME speed is much weaker (cc= -0.539 for minimum Dst) such that speed is a poor predictor of geoffectiveness (at least as measured by Dst). We find that there is a >50% probability of an intense storm being generated when E_y>6 mV/m in the ICME or sheath and that all cases with E_y>10~mV/m generate such storms. The probability of generating a severe storm is around 80% for E_{y ∼} 15-20 mV/m. The observations suggest that even if ICME arrival at the Earth can be accurately predicted by imaging and/or modeling, the large majority of the forecast ICMEs will not produce important geomagnetic effects. They emphasize the importance of estimating or measuring directly (e.g., using an upstream monitor) B_s and E_y for forecasting the geomagnetic impact of ICMEs. Successful modeling of magnetic fields will likely require the field strength and direction in the ICME and the sheath to be predicted on time scales of ∼1~hour ( ∼0.01~AU).