JW-Flare: Accurate Solar Flare Forecasting Method Based on Multimodal Large Language Models

TL;DR

JW-Flare leverages multimodal large language models fine-tuned on solar physics data to achieve state-of-the-art accuracy in predicting powerful solar flares, surpassing traditional models and revealing insights into the model's knowledge sources.

Contribution

This paper introduces JW-Flare, the first multimodal large language model tailored for solar flare forecasting, demonstrating superior performance and explainability in a domain-specific application.

Findings

JW-Flare accurately predicts all 79 X-class flares in the test set.

It achieves a TSS of 0.95 and TPR of 1.00, outperforming traditional models.

Model scaling laws are confirmed in the context of solar physics applications.

Abstract

Solar flares, the most powerful explosive phenomena in the solar system, may pose significant hazards to spaceborne satellites and ground-based infrastructure. Despite decades of intensive research, reliable flare prediction remains a challenging task. Large Language Models, as a milestone in artificial intelligence, exhibit exceptional general knowledge and next-token prediction capabilities. Here we introduce JW-Flare, the first Multimodal Large Language Models (MLLMs) explicitly trained for solar flare forecasting through fine-tuning on textual physic parameters of solar active regions and magnetic field images. This method demonstrates state-of-the-art (SOTA) performance for large flares prediction on the test dataset. It effectively identifies all 79 X-class flares from 18,949 test samples, yielding a True Skill Statistic (TSS) of 0.95 and a True Positive Rate (TPR) of 1.00, outperforming traditional predictive models. We further investigate the capability origins of JW-Flare through explainability experiments, revealing that solar physics knowledge acquired during pre-training contributes to flare forecasting performance. Additionally, we evaluate models of different parameter scales, confirming the Scaling_Law of Large Language Models in domain-specific applications, such as solar physics. This study marks a substantial advance in both the scale and accuracy of solar flare forecasting and opens a promising avenue for AI-driven methodologies in broader scientific domains.