Unveiling the Potential of Deep Learning Models for Solar Flare Prediction in Near-Limb Regions

TL;DR

This study evaluates deep learning models for predicting solar flares within 24 hours using full-disk magnetograms, emphasizing the challenge of near-limb regions, and finds that certain architectures perform better in these areas.

Contribution

It compares the performance of AlexNet, VGG16, and ResNet34 in flare prediction, highlighting the effectiveness of ResNet34 for near-limb flare sensitivity.

Findings

AlexNet achieved the highest overall performance with TSS~0.53.

ResNet34 showed superior prediction sensitivity in near-limb regions.

Models can discern complex spatial patterns for flare prediction near the solar limb.

Abstract

This study aims to evaluate the performance of deep learning models in predicting $\geq$M-class solar flares with a prediction window of 24 hours, using hourly sampled full-disk line-of-sight (LoS) magnetogram images, particularly focusing on the often overlooked flare events corresponding to the near-limb regions (beyond $\pm$70$^{\circ}$ of the solar disk). We trained three well-known deep learning architectures--AlexNet, VGG16, and ResNet34 using transfer learning and compared and evaluated the overall performance of our models using true skill statistics (TSS) and Heidke skill score (HSS) and computed recall scores to understand the prediction sensitivity in central and near-limb regions for both X- and M-class flares. The following points summarize the key findings of our study: (1) The highest overall performance was observed with the AlexNet-based model, which achieved an average TSS$\sim$0.53 and HSS$\sim$0.37; (2) Further, a spatial analysis of recall scores disclosed that for the near-limb events, the VGG16- and ResNet34-based models exhibited superior prediction sensitivity. The best results, however, were seen with the ResNet34-based model for the near-limb flares, where the average recall was approximately 0.59 (the recall for X- and M-class was 0.81 and 0.56 respectively) and (3) Our research findings demonstrate that our models are capable of discerning complex spatial patterns from full-disk magnetograms and exhibit skill in predicting solar flares, even in the vicinity of near-limb regions. This ability holds substantial importance for operational flare forecasting systems.