CME Prediction Using SDO, SoHO, and STEREO data with a Machine Learning Algorithm

Abstract:

It is unclear whether a flaring active region will also produce a Coronal Mass Ejection (CME). Usually, active regions that produce large flares will also produce a CME, but this is not always the case. For example, the largest active region from the last 24 years, NOAA Active Region 12192 of October 2014, produced many X-class flares but not a single CME. We attempt to forecast whether an active region that produces an M- or X-class flare will also produce a CME. We do this by analyzing data from three solar observatories -- SDO, STEREO, and SoHO -- using a machine-learning algorithm. We find that the role of horizontal component of the photospheric magnetic field plays a crucial component in driving a CME, a result corroborated by Sun et al. (2015). We present the success rate of our method and the potential applications to space weather forecasts.