Magnetic Causes of Solar Coronal Mass Ejections: Dominance of the Free Magnetic Energy over Either the Magnetic Twist or Size Alone

We report further results from our ongoing assessment of magnetogram-based measures of active-region nonpotentiality and size as predictors of coronal mass ejections (CMEs). We have devised improved generalized measures of active-region nonpotentiality that apply to active regions of any degree of magnetic complexity, rather than being limited to bipolar active regions as our initial measures were. From a set of approx.50 active-regions, we have found that measures of total nonpotentiality have a 75-80% success rate n predicting whether an active region will produce a CME in 2 days after the magnetogram. This makes measures of total nonpotentiality a better predictor than either active-region size, or active region twist (size-normalized nonpotentiality), which have a approx.65% success rates. We have also found that we can measure from the line-of-sight magnetograms an active region's total nonpotentiality and the size, which allows use to use MDI to evaluate these quantities for 4-5 consecutive days for each active region, and to investigate if there is some combination of size and total nonpotentiality that have a stronger predictive power than does total nonpotentiality. This work was funded by NASA through its LWS TR&T Program and its Solar and Heliospheric Physics SR&T Program, and by NSF through its Solar Terrestrial Research and SHINE programs.