Comparing LSTM-Based Sequence-to-Sequence Forecasting Strategies for 24-Hour Solar Proton Flux Profiles Using GOES Data

2025-10-06

AI Summary

This study addresses the challenge of accurately forecasting 24-hour proton flux time profiles following solar proton events (SPEs). We propose an LSTM-based sequence-to-sequence (seq2seq) model that departs from conventional autoregressive approaches by adopting a one-step-ahead forward prediction architecture, thereby mitigating error accumulation. To enhance performance, we introduce trend-smoothing preprocessing—particularly beneficial for multi-channel (proton + X-ray) inputs—and integrate hierarchical cross-validation, optimized embedding dimensionality, and multivariate fusion. Experiments reveal that models trained on smoothed data exhibit superior generalization, whereas those trained on raw data achieve peak accuracy on certain metrics; overall optimal performance stems from smoothed-data modeling. Crucially, this work demonstrates that architectural design—specifically the seq2seq framework—can amplify and even surpass the benefits of preprocessing. The proposed method establishes a high-accuracy, deployable temporal forecasting paradigm for SPE radiation risk early warning.

Technology Category

• Machine Learning: Time-Series/Data Streams
• Planning, Routing, and Scheduling: Planning with Language Models
• Natural Language Processing: (Large) Language Models

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Web Mining and Content Analysis: Large pretrained models with web data
• Search and Retrieval-Augmented AI: Large language models for search

Abstract

Solar Proton Events (SPEs) cause significant radiation hazards to satellites, astronauts, and technological systems. Accurate forecasting of their proton flux time profiles is crucial for early warnings and mitigation. This paper explores deep learning sequence-to-sequence (seq2seq) models based on Long Short-Term Memory networks to predict 24-hour proton flux profiles following SPE onsets. We used a dataset of 40 well-connected SPEs (1997-2017) observed by NOAA GOES, each associated with a >=M-class western-hemisphere solar flare and undisturbed proton flux profiles. Using 4-fold stratified cross-validation, we evaluate seq2seq model configurations (varying hidden units and embedding dimensions) under multiple forecasting scenarios: (i) proton-only input vs. combined proton+X-ray input, (ii) original flux data vs. trend-smoothed data, and (iii) autoregressive vs. one-shot forecasting. Our major results are as follows: First, one-shot forecasting consistently yields lower error than autoregressive prediction, avoiding the error accumulation seen in iterative approaches. Second, on the original data, proton-only models outperform proton+X-ray models. However, with trend-smoothed data, this gap narrows or reverses in proton+X-ray models. Third, trend-smoothing significantly enhances the performance of proton+X-ray models by mitigating fluctuations in the X-ray channel. Fourth, while models trained on trendsmoothed data perform best on average, the best-performing model was trained on original data, suggesting that architectural choices can sometimes outweigh the benefits of data preprocessing.

Problem

• Forecasting 24-hour solar proton flux profiles using LSTM-based seq2seq models
• Evaluating proton-only versus combined proton and X-ray input data
• Comparing autoregressive and one-shot forecasting strategies for accuracy

Innovation

• LSTM-based seq2seq models predict proton flux
• One-shot forecasting reduces error accumulation
• Trend-smoothing enhances proton+X-ray model performance