NASA AI Model Predicts Solar Storm Regions Before They Appear

• NASA researchers developed an AI model that predicts the emergence of active solar regions up to 12 hours before they become visible.
• The model uses a "sliding-window transformer" architecture to analyze acoustic wave fluctuations and magnetic field changes beneath the Sun's surface.
• This early warning system aims to protect astronauts, satellites, and communication infrastructure from solar flares and coronal mass ejections.

Technology and Methodology

• Developed by NASA’s COFFIES (Consequence Of Fields and Flows in the Interior and Exterior of the Sun) center, the team includes researchers from NJIT, Princeton, and NASA Ames.
• The model identifies "precursors" in the Sun's acoustic power—subtle changes described as rhythmic shifts in a noisy environment—as magnetic structures rise toward the surface.
• Unlike previous deep learning approaches that analyze the entire surface at once, the transformer model focuses on recent data sequences while maintaining context from earlier patterns.

Impact on Space Weather Forecasting

• Current operational forecasts by NOAA and the U.S. Air Force rely on monitoring already visible sunspots to calculate flare probability.
• The new predictive capability could allow for more proactive safety measures for Artemis missions and future crewed travel to Mars.
• While not yet operational in real-time, the model will undergo further validation across a wider range of solar events to improve accuracy.

Collaboration

• The COFFIES team is part of a broader NASA effort to integrate research into tools used by the Moon to Mars Space Weather Analysis Office (M2M SWAO) and the Space Weather Prediction Center.