USRA Supports Development of Surya, a New AI Model for Space Weather Research

Artificial intelligence is now being applied to one of the most complex scientific challenges: understanding the Sun’s behavior. Researchers from academia, industry, and the Universities Space Research Association (USRA) have collaborated to develop Surya, an open-source foundation model that uses AI to analyze and forecast solar activity.

The model represents an important advance for space-weather prediction, a field critical to protecting satellites, communications systems, and infrastructure on Earth. By drawing on nine years of data from NASA’s Solar Dynamics Observatory, which continuously records high-resolution images of the Sun, Surya can detect patterns and anticipate solar events with improved accuracy.

Developed through a collaboration that includes NASA centers, universities, and private organizations, Surya adapts the principles of large-scale machine learning familiar from natural-language AI systems—to model the dynamics of solar physics. Its results suggest new potential for studying and predicting space-weather phenomena.

USRA contributed to the project by supporting model implementation, evaluating architecture, and providing technical review. Dr. David Bell, Director of USRA’s Research Institute for Advanced Computer Science, noted that the model “marks a new era in how we study and prepare for solar activity that affects both space- and Earth-based technologies.”

Early applications of Surya include:

• Forecasting the emergence of active solar regions
• Predicting solar flares and solar-wind speed
• Modeling extreme ultraviolet (EUV) spectra

Improving these capabilities could help reduce the economic impact of solar storms, which have previously disrupted satellite launches, GPS signals, and agricultural operations.

Source: Adapted from a press release distributed by PR Newswire on Oct. 25, 2025/ Universities Space Research Association.