AI reconstructs 3D solar magnetism to sharpen storm forecasts

Summary

Researchers at the University of Hawaiʻi have developed an AI system that can infer the Sun’s magnetic field in full 3D, potentially transforming how we forecast solar storms. Writing in Gizmodo en Español, Thomas Handley explains that the model, called Haleakalā Disambiguation Decoder, doesn’t just crunch data blindly—it bakes in basic physical rules, like the fact that magnetic field lines form closed loops, to resolve long‑standing ambiguities in solar measurements. By combining this physics‑aware learning with high‑resolution observations, the system can reconstruct the direction and height of magnetic structures in active regions and sunspots far more accurately than classic techniques. That, in turn, should improve predictions of when powerful eruptions and coronal mass ejections will hit Earth, events that can disrupt satellites, GPS, communications and power grids. It’s a good example of AI augmenting—not replacing—traditional physics, giving space‑weather forecasters a richer 3D picture of a star we still don’t fully understand.