NASA’s new AI model can predict when a solar storm may strike
1 min read
Summary
IBM and NASA have unveiled Surya, an open-source machine learning model trained on a decade’s worth of NASA solar data, to improve scientific understanding of the Sun’s physics and weather patterns.
The model aims to predict potentially dangerous solar flares, which see the sun eject energy and particles into space and can disrupt radio signals and expose astronauts to bursts of radiation.
Although the timing and intensity of solar flares are challenging to predict, early tests of Surya have successfully predicted some flares two hours before they occur.
IBM claims this doubles the warning time possible with existing methods.
The model could also be used to understand the relationship between solar storms and Earth weather.
Furthermore, Surya could lead to new insights into the behaviour of other celestial bodies.