Predictions of active region flaring probability using subsurface helicity measurements

Solar flares are responsible for a number of hazardous effects on the earth such as disabling high-frequency radio communications, interfering with GPS measurements, and disrupting satellites. However, forecasting flare occurrence is currently very difficult. One possible means for predicting flare occurrence lies in helioseismology, i.e. analysis of the region below the active region for signs of an impending flare. Time series helioseismic data collected by the Global Oscillation Network Group (GONG) has been analyzed for a subset of active regions that produce large flares and a subset with very high magnetic field strength that produce no flares. A predictive parameter has been developed and analyzed using discriminant analysis as well as traditional forecasting tools such as the Heidke skill score. Preliminary results show that this parameter predicts the flaring probability of an active region 2-3 days in advance with a relatively high degree of success.