IBM and NASA have introduced a new open-source artificial intelligence model designed to analyze high-resolution solar images and predict solar activity that could impact Earth’s technological infrastructure.
The model, named Surya after the Sanskrit word for the Sun, was trained on nine years of data from NASA’s Solar Dynamics Observatory. It is able to classify and predict solar flares, measure solar wind speeds, and forecast the emergence of active regions on the Sun. Early testing has shown a 16% improvement in flare classification accuracy compared to earlier methods.
Solar activity such as flares and coronal mass ejections can disrupt satellites, navigation systems, power grids, telecommunications, and airline travel. With global reliance on space-based technology continuing to grow, accurate space weather prediction is seen as critical for both research and operational preparedness.
Surya addresses limitations of earlier forecasting techniques by using one of the largest curated heliophysics datasets to train its algorithms. The model achieves unprecedented spatial resolution, enabling scientists to capture fine-scale solar features not previously available in large-scale AI training workflows.
The project is part of a broader collaboration between IBM and NASA to apply AI to planetary and space science. Surya has been released on the Hugging Face platform, allowing researchers worldwide to adapt the model for regional or industry-specific applications.
Source: PR Newswire