Major geomagnetic storms (Dst ≤-100 nT) generated by corotating interaction regions

[ 1] Seventy-nine major geomagnetic storms ( minimum Dst <= - 100 nT) observed in 1996 to 2004 were the focus of a "Living with a Star'' Coordinated Data Analysis Workshop ( CDAW) in March 2005. In nine cases, the storm driver appears to have been purely a corotating interaction region (CIR) without any contribution from coronal mass ejection-related material ( interplanetary coronal mass ejections (ICMEs)). These storms were generated by structures within CIRs located both before and/or after the stream interface that included persistently southward magnetic fields for intervals of several hours. We compare their geomagnetic effects with those of 159 CIRs observed during 1996 - 2005. The major storms form the extreme tail of a continuous distribution of CIR geoeffectiveness which peaks at Dst similar to - 40 nT but is subject to a prominent seasonal variation of similar to 40 nT which is ordered by the spring and fall equinoxes and the solar wind magnetic field direction toward or away from the Sun. The O'Brien and McPherron ( 2000) equations, which estimate Dst by integrating the incident solar wind electric field and incorporating a ring current loss term, largely account for the variation in storm size. They tend to underestimate the size of the larger CIR-associated storms by Dst similar to 20 nT. This suggests that injection into the ring current may be more efficient than expected in such storms. Four of the nine major storms in 1996 - 2004 occurred during a period of less than three solar rotations in September to November 2002, also the time of maximum mean IMF and solar magnetic field intensity during the current solar cycle. The maximum CIR-storm strength found in our sample of events, plus additional 23 probable CIR-associated Dst <= - 100 nT storms in 1972 - 1995, is ( Dst = - 161 nT). This is consistent with the maximum storm strength ( Dst similar to - 180 nT) expected from the O'Brien and McPherron equations for the typical range of solar wind electric fields associated with CIRs. This suggests that CIRs alone are unlikely to generate geomagnetic storms that exceed these levels.