Real-Time 24-hour Ahead Forecasting of Solar Energetic Particle Events with SEPNET-PRISM-R

Abstract Solar energetic particle (SEP) events, operationally defined as proton flux exceeding10 pfu in the >10 MeV channel, threaten spacecraft systems, astronautsafety, and high-latitude aviation, yet continuous real-time forecasting with adequatelead time remains difficult. We present SEPNET-PRISM-R, a near-real-timesystem that issues hourly, 24-hour-ahead probabilistic SEP forecasts. The modelintegrates multimodal solar data—including active-region patches, flare features,coronal mass ejection (CME) characteristics, and historical proton flux—to concurrentlypredict SEP occurrences, peak proton flux, soft X-ray flux, and CMEvelocity, with the primary objective being SEP forecasting. Building on existingliterature, SEPNET-PRISM-R enhances operational performance by incorporatinga post-processing module that enables cross-validated decision-making and interpretablequantification of feature importance. The training dataset spans fromFebruary 1986 to September 2025, with subsequent forecasts generated sequentiallythrough the present via a real-time space weather prediction platformthat provides predictive probabilities, feature attributions, and historical analogs,enabling more reliable and informed operational decision-making. The proposedframework substantially suppresses false-alarm rates while maintaining robustevent-detection capabilities, as evidenced by case studies and aggregate statistics.