Prediction of Coronal Mass Ejections from Vector Magnetograms: Results from More Active Regions

In a previous pilot study of four predominantly bipolar active regions observed by the MSFC vector magnetograph in 1991-92, we found that two quantitative measures of the global nonpotentiality of an active region were promising predictors of whether the active region produced coronal mass ejections (CMES) during its rotation across the Sun. The two quantitative measures are 1) the length (L(sub ss)) of the segments along the active region's main neutral line on which the vector magnetic field is both strong and strongly sheared, and 2) the global net current (I(sub N) flowing up one side of the bipole and down the other. Two of the active regions had large measured global nonpotentiality (L(sub ss) approximately >10(exp 5) km and I(sub N) approximately >4x 10(exp 12) Amp) and produced three CMEs each. The other two active regions had much smaller measured global nonpotentiality and produced no CMES. We have now expanded our sample of active regions by a factor of 3 with active regions from 1991 to the present. We have calibrated the line-of-sight magnetic field by comparing it to the line-of-sight field measured by Kitt Peak and/or SOHO/MDI. We have resolved the 180 deg. ambiguity in the transverse field by using three different methods, and compared the resulting global net currents. We will compare the two measures of global nonpotentiality to each other and to the CME productivity of each active i-eolon in our expanded set. We will examine the extent to which these results from the expanded set confirm or modify the tentative rule for CME prediction that we found in our pilot study.