From conventional magnetograms and chromospheric and coronal images, it is known qualitatively that the fastest coronal mass ejections (CMEs) are magnetic explosions from sunspot active regions in which the magnetic field is globally strongly sheared and twisted from its minimum-energy potential configuration. In this paper, we present measurements from active region vector magnetograms that begin to quantify the dependence of the CME productivity of an active region on the global nonpotentiality of its magnetic field. From each of 17 magnetograms of 12 bipolar active regions, we obtain a measure of the size of the active region (the magnetic flux content, Phi) and three different measures of the global nonpotentiality (L-SS, the length of strong-shear, strong-field main neutral line; I-N, the net electric current arching from one polarity to the other; and alpha = muI(N)/Phi, a flux-normalized measure of the field twist). From these measurements and the observed CME productivity of the active regions, we find that: ( 1) All three measures of global nonpotentiality are statistically significantly correlated with each other and with the active region flux content. ( 2) All three measures of global nonpotentiality are significantly correlated with CME productivity. The flux content has some correlation with CME productivity, but at a less than statistically significant confidence level (less than 95%). (3) The net current is less strongly correlated with CME productivity than is, and the correlation of flux content with CME productivity is weaker still. If these differences in correlation strength, and a significant correlation of alpha with flux content, persist to larger samples of active regions, this would suggest that active region size does not affect CME productivity except through global nonpotentiality. ( 4) For each of the four global magnetic quantities, the correlation with CME productivity is stronger for a +/-2 day time window for the CME production than for windows half as wide or twice as wide. This plausibly results from most CME-productive active regions producing less than one CME per day, and from active region evolution often significantly changing the global nonpotentiality over the course of several days. These results establish that measures of active region global nonpotentiality from vector magnetograms ( such as L-SS, I-N, and alpha) should be useful for prediction of active region CMEs.
