Last year, Earth faced its most powerful geomagnetic storm in 20 years, showcasing stunning auroras and pointing the value of preparation against solar events. While such storms threaten energy providers, improvements meant no major disruptions this time.
NASA and IBM have introduced an open-source AI model, Surya, named after the Sanskrit for the Sun. Described by Juan Bernabe-Moreno as an “AI telescope for the Sun,” Surya uses machine learning on images from NASA’s Solar Dynamics Observatory (SDO) to forecast solar weather.
Surya shows a 16% accuracy boost over previous systems in predicting solar flares, providing visual forecasts up to two hours ahead, an advantage for modern infrastructure providers.
There’s irony in its release amid potential NASA budget cuts under President Trump’s proposal, threatening operations like SDO, despite bipartisan resistance.
Surya’s presence underscores NASA’s vital scientific role, emphasizing advances beyond the agency. IBM and NASA have enabled Surya to live on as an open-source model accessible on platforms like Hugging Face, extending its impact.
For more information about Surya, visit Hugging Face.