nasa's new solar ai
Embedded Files
IBM and NASA have recently released a new open-source artificial intelligence model. The purpose of this model is to forecast solar activity (basically the Sun's weather) and its subsequent impact on Earth and space technology.
Surya, named after the Sanskrit word for the Sun, represents a huge step in using AI to interpret high-res solar imagery and improve predictions of space weather. It's the first model of its kind which has been trained specifically on detailed solar observation data, offering scientists new ways to analyse and somewhat predict the Sun's surface activity.
Preliminary results show that Surya is making significant progress in solar flare forecasting, a long-standing challenge in heliophysics. Early tests indicate that it has outperformed existing benchmarks by an impressive 16%!
Examples of solar imagery used for training Surya. Source: NASA
The model’s success relies heavily on the extensive dataset collected by NASA’s Solar Dynamics Observatory (SDO). Since its launch in 2010, SDO has continuously recorded the Sun in high resolution for nearly 15 years, capturing images every 12 seconds across multiple wavelengths and measuring its magnetic field with precision.
These images compare the ground-truth observations (right) with the model’s output (center) for solar flares, the primary drivers of space weather. Surya’s prediction closely matches the actual event, indicating that it has captured key aspects of solar flare dynamics (in this example). These early results suggest that Surya can infer the structure and evolution of a flare from its initial phase. Source: NASA
published: 31/08/25 by kaan evcimen
Google Sites
Report abuse