AI model improves long-range space weather forecasts
Surface observations of solar activity were linked to deep magnetic dynamics using AI, paving the way for earlier warnings to protect satellites and power grids.
Scientists from Southwest Research Institute and the National Center for Atmospheric Research, supported by the National Science Foundation, have created an experimental tool that could extend space weather forecasts from hours to several weeks.
Longer lead times would help operators protect satellites, navigation systems, and power infrastructure from solar disturbances. Research focuses on predicting where flare-producing solar active regions form.
By analysing magnetic data captured by the Solar Dynamics Observatory, scientists reconstructed hidden magnetic conditions beneath the Sun’s surface, showing that these regions follow structured magnetic bands rather than appearing randomly.
PINNBARDS, a physics-informed AI model, connects surface observations with deep tachocline dynamics that drive solar magnetic evolution. Better modelling could provide earlier warnings of solar flares and coronal mass ejections, helping protect communications and astronaut safety.
Funding from NASA and Stanford University supported the work. Researchers describe it as a foundation for next-generation forecasting systems capable of anticipating extreme solar activity with greater accuracy.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!