AI Boosts Space Weather Forecasts, Protecting Satellites

The study presents a machine-learning-based pipeline for forecasting megaelectron-volt (MeV) electron flux in the Earth's outer radiation belt, a critical capability for safeguarding satellites. The research leverages precipitating electrons, near 1-MeV electron flux, solar wind measurements, and geomagnetic activity indices to train algorithms, including linear regression, convolutional neural networks, and Transformer-Encoder models. A key innovation is the application of the TimesFM timeseries foundation model, particularly in a hybrid configuration (TimesFM+Cov) that combines ridge regression with TimesFM inference.

The results demonstrate that TimesFM+Cov significantly outperforms other models, achieving an average R2 of 0.9 across L-shells in out-of-sample testing. This represents a substantial improvement in forecasting accuracy, especially at higher L-shells. The ability to accurately predict MeV electron flux variations is essential for mitigating risks to satellites, as these electrons can cause damage to electronic components and disrupt spacecraft operations.

This research highlights the potential of machine learning and foundation models in space weather forecasting. By adapting and refining these techniques, scientists can develop more reliable and comprehensive forecasting systems, enabling proactive measures to protect valuable space assets. However, further work is needed to address limitations in accuracy at higher L-shells and to validate the models under a wider range of space weather conditions. The integration of additional data sources and the development of more sophisticated algorithms could further enhance the performance of these forecasting systems.

*Transparency Disclosure: The AI model used in this analysis is Gemini 2.5 Flash. The analysis is based solely on the provided source content and adheres to EU AI Act Article 50 compliance standards.*

_Context: This intelligence report was compiled by the DailyOrbitalWire Strategy Engine. Verified for Art. 50 Compliance._