AI predicts solar storms that could trigger complete blackouts on Earth

The Sun may appear steady, but its activity is unpredictable. Powerful solar storms can disrupt satellites, GPS, and power grids, posing risks to modern technology. A new study suggests artificial intelligence can significantly improve space weather predictions, offering earlier warnings for potential disruptions on Earth. Scientists used AI to analyse past solar activity, revealing patterns that traditional methods often miss.

AI detects solar flares before they happen
The Sun is a churning ball of plasma, constantly shifting. Powerful eruptions, known as coronal mass ejections (CMEs), send charged particles hurtling through space. These can interfere with satellites, GPS, and power grids when they reach Earth.

A team led by Sabrina Guastavino from the University of Genoa applied AI to study past solar storms. Their model analysed data from the May 2024 solar storm, successfully predicting solar flares and CMEs with high precision. The AI system detected patterns in solar activity that were difficult to identify using conventional methods.

Better predictions for future space weather
The AI model accurately forecasted an X8.7-class flare and predicted the storm’s impact on Earth. The results were significantly more precise than traditional forecasting techniques, reducing uncertainties in CME travel times and geomagnetic storm intensities.