IBM and NASA Develop a Digital Twin of the Sun to Predict Future Solar Storms (www.wired.com)

IBM and NASA have unveiled Surya, a pioneering AI foundation model designed as a digital twin of the Sun to better understand and predict solar storms. Trained on nine years of high-resolution data from NASA’s Solar Dynamics Observatory, Surya integrates diverse solar observations—including magnetic field measurements and temperature estimates across different electromagnetic wavelengths—to simulate the complex dynamics of solar activity in real time. This digital twin approach allows scientists to manipulate and study the Sun’s behavior more efficiently than ever before. Surya leverages a long-range vision transformer architecture enhanced by spectral gating, which reduces noise and memory usage for sharper data analysis. Unlike traditional models that depend on extensive labeled datasets, Surya learns directly from raw data, enabling faster adaptation and improved predictive accuracy. During early testing, it successfully integrated data from multiple solar missions and achieved a 16% boost in solar flare classification accuracy, offering a two-hour advance warning for solar flares—doubling the prediction lead time of existing systems. This advance is critical for safeguarding astronauts, satellites, and terrestrial infrastructure from disruptive solar events. Beyond heliophysics, NASA highlights Surya’s adaptable architecture, which could revolutionize planetary science and Earth observation fields. By enhancing the speed and precision of analyzing solar behavior, Surya empowers the AI/ML community with a powerful tool to mitigate the risks of solar storms that threaten global telecommunications, power grids, and navigation systems—paving the way for more resilient technology infrastructure on Earth.