Part 11/12:
Using nine years of data from NASA’s Solar Dynamics Observatory, Surria creates a digital twin of the sun, predicting realizations an hour ahead with 16% improved accuracy over existing tools. Its architecture—based on a long short vision transformer—handles enormous images comfortably, and learns complex patterns, including the sun’s irregular rotation.
Beyond flares, Surria predicts solar wind speeds, ultraviolet radiation, and magnetic activity—key factors in space weather risks. Its accessibility democratizes space weather modeling, offering governments, space agencies, and researchers a powerful tool for early warning and disaster mitigation.