Towards Interpretable Solar Flare Prediction with Attention-based Deep Neural Networks

TL;DR

This paper presents an attention-based deep learning model for solar flare prediction that improves interpretability and operational readiness, achieving competitive skill scores and focusing on relevant active regions in full-disk magnetograms.

Contribution

The study introduces an attention mechanism into deep learning for solar flare prediction, enhancing interpretability and demonstrating effective prediction of near-limb flares from full-disk magnetograms.

Findings

Achieved an average TSS of 0.54 and HSS of 0.37 for flare prediction.

Model learns features from active regions in full-disk magnetograms.

Predicts near-limb flares with relevant active region focus.

Abstract

Solar flare prediction is a central problem in space weather forecasting and recent developments in machine learning and deep learning accelerated the adoption of complex models for data-driven solar flare forecasting. In this work, we developed an attention-based deep learning model as an improvement over the standard convolutional neural network (CNN) pipeline to perform full-disk binary flare predictions for the occurrence of $\geq$M1.0-class flares within the next 24 hours. For this task, we collected compressed images created from full-disk line-of-sight (LoS) magnetograms. We used data-augmented oversampling to address the class imbalance issue and used true skill statistic (TSS) and Heidke skill score (HSS) as the evaluation metrics. Furthermore, we interpreted our model by overlaying attention maps on input magnetograms and visualized the important regions focused on by the model that led to the eventual decision. The significant findings of this study are: (i) We successfully implemented an attention-based full-disk flare predictor ready for operational forecasting where the candidate model achieves an average TSS=0.54$\pm$0.03 and HSS=0.37$\pm$0.07. (ii) we demonstrated that our full-disk model can learn conspicuous features corresponding to active regions from full-disk magnetogram images, and (iii) our experimental evaluation suggests that our model can predict near-limb flares with adept skill and the predictions are based on relevant active regions (ARs) or AR characteristics from full-disk magnetograms.