Solar Limb Flare Prediction using SDO/AIA EUV Intensities

We propose an unprecedented solar limb flare prediction model using SDO/AIA data. Projection effects in solar white light and magnetogram data hinder accurate flare forecasts near the solar limb. Through the statistical flare analysis of SDO/AIA EUV data, our model predicts the probability of major solar limb flare occurrence within a day. We use SDO/AIA 94 Å and 131 Å wavelengths that have high correlations with X-ray flare class from 2010 to 2022. For the same period, we select 252 flares stronger than M1 class and located at the limb region(larger than 60 degrees in longitude). For our study we define the limb intensity as the summation of SDO/AIA over the limb region. We compute the flare occurrence probability as a function of total intensity and limb intensity. Our major result can be summarized as follows. First, the major limb flare occurrence probability strongly depends on both intensities. Second, for 94 Å, the largest probability is about 33.3 % and the smallest probability is about 1.1 %. For 131 Å, the largest probability is about 34.6 % and the smallest probability is 0.9. Our study provides positive feasibility for major limb flare prediction by using solar EUV data.