Solar Energetic Particle Forecasting with Multi-Task Deep Learning: SEPNet

TL;DR

This work tackles the challenge of forecasting solar energetic particle (SEP) events by introducing SEPNet, a multi-task deep learning framework that jointly predicts SEP occurrence and related solar eruptive activities (flares and CMEs) using SHARP magnetic-field parameters and GOES flare data. SEPNet and its SEPNET2 variant leverage shared representations and sequential modeling (LSTM and transformer) to exploit temporal dependencies, achieving superior performance to classical ML methods and state-of-the-art pre-eruption models, particularly when SHARP features are included. The approach demonstrates robustness across SEPVAL benchmarks and stratified splits, with real-time forecasts showing practical potential for operational space weather alerts, albeit with persistent false-alarm challenges due to data imbalance. The authors provide open-source access to data and code, outline future directions including flux and duration predictions, and emphasize the framework’s potential to improve timely SEP warnings for spacecraft, astronauts, and aviation safety.

Abstract

Solar energetic particle (SEP) events pose severe threats to spacecraft, astronaut safety, and aviation operations, accurate SEP forecasting remains a critical challenge in space weather research due to their complex origins and highly variable propagation. In this work, we built SEPNet, an innovative multi-task neural network that jointly predicts future solar eruptive events, including solar flares and coronal mass ejections (CMEs) and SEPs, incorporating long short-term memory and transformer architectures that capture contextual dependencies. SEPNet is a machine learning framework for SEP prediction that utilizes an extensive set of predictors, including solar flares, CMEs, and space-weather HMI active region patches (SHARP) magnetic field parameters. SEPNet is rigorously evaluated on the SEPVAL SEP dataset (whitman, 2025b), which is used to evaluate the performance of the current SEP prediction models. The performance of SEPNet is compared with classical machine learning methods and current state-of-the-art pre-eruptive SEP prediction models. The results show that SEPNet, particularly with SHARP parameters, achieves higher detection rates and skill scores while maintaining suitable for real-time space weather alert operations. Although class imbalance in the data leads to relatively high false alarm rates, SEPNet consistently outperforms reference methods and provides timely SEP forecasts, highlighting the capability of deep multi-task learning for next-generation space weather prediction. All data and code are available on GitHub at https://github.com/yuyian/SEP-Prediction.git.

Figures (6)

• Figure 1: Timeline visualization of operational SEP ($>$ 10 MeV 10 pfu), flare, CME, and SHARP records used in this study. For each data source, only records occurring between 24 hours before the first SEP event search time and the minimum of the latest recorded times across all sources are included. Each colored band marks the temporal occurrence of a record by type: operational SEP (red), flare (orange), CME (blue), and SHARP (green). Vertical dashed lines indicate the selected time period.
• Figure 2: Diagram illustrating the architectures of the proposed multi-task learning models. Top: SEPNet, composed of shared feed-forward layers with layer normalization, ReLU activations, and dropout, followed by regression and classification heads for predicting flare/CME counts and SEP event probability. Bottom: SEPNet2, an enhanced variant introducing sequential processing via a unidirectional LSTM and transformer encoder before multi-task prediction.
• Figure 3: Performance metrics for SEPVAL prediction models, showing the median and target quantile values across different feature sets and model architectures. The shaded light blue region represents the median and target quantile achieved by state-of-the-art pre-eruption models. Feature set abbreviations: F = flare-related features; S = SHARP parameters; C = CME-related features. Performance metric abbreviations: ACC = accuracy; AUC = area under the curve; FPR = false positive rate; F1 = F1 score; POD = probability of detection; FAR = false alarm rate; TSS = true skill score; HSS = Heidke skill score. Model abbreviations: LR = logistic regression with elastic net regularization; SVM = support vector machines; RF = random forests; XGB = extreme gradient boosting.
• Figure 4: Performance metrics on the 20% testing set for different feature sets and models, targeting classification of general SEP events. Results for each criterion are the median values across five independent random stratified data splits. Feature set abbreviations: F = flare-related features; S = SHARP parameters; C = CME-related features. Performance metric abbreviations: ACC = accuracy; AUC = area under the curve; FPR = false positive rate; F1 = F1 score; POD = probability of detection; FAR = false alarm rate; TSS = true skill score; HSS = Heidke skill score. Model abbreviations: LR = logistic regression with elastic net regularization; SVM = support vector machines; RF = random forests; XGB = extreme gradient boosting.
• Figure 5: Performance of re-validated models (optimize the decision threshold for operational SEP event prediction) compared to original models, targeting classification of operational SEP events. Metrics are derived on the 20% testing set using SHARP parameters with flare features, with results for each criterion being the median values across five independent random stratified data splitting. Performance metric abbreviations: F1 = F1 score; POD = probability of detection; TSS = true skill score; HSS = Heidke skill score. Model abbreviations: LR = logistic regression with elastic net regularization; SVM = support vector machines; RF = random forests; XGB = extreme gradient boosting.