DAGGER - short for Deep Learning Geomagnetic Perturbation - is an AI model that predicts geomagnetic disturbances across the whole planet roughly 30 minutes before they occur. It was developed by an international team at the Frontier Development Lab, a partnership that includes NASA, the US Geological Survey and the US Department of Energy, and the results were published in 2023 in the journal Space Weather.
The model uses deep learning to connect solar-wind measurements from heliophysics missions such as ACE, Wind, IMP-8 and Geotail to the geomagnetic disturbances recorded at ground stations on Earth. According to NASA, it can produce a prediction in under a second and refresh that prediction every minute, and it was tested against major storms in August 2011 and March 2015. The code is open source, so power-grid operators, satellite controllers and telecom companies could in principle use the half-hour of lead time to protect their infrastructure from a severe solar storm.