NASA is working with Amazon’s cloud computing services service AWS and Amazon Machine Learning Solutions Lab on a project to use artificial intelligence (AI) to provide early warning of supersolar storms,media reported. Solar storms occur when large amounts of radiation and charged particles from disturbances on the sun’s surface erupt at millions of kilometers per hour, affecting more than half of the world’s radio communications and possibly even destroying satellites and power grids.
This ultraviolet image from NASA’s Solar Dynamics Observatory shows a huge flare during the 2017 solar storm
However, humans are not unprepared for solar storms, and through solar observation satellites such as NASA’s Solar Dynamics Observatory and Advanced Explorer, we can warn of upcoming solar storms for a long time. to enable satellite operators and grid managers to take protective measures.
Now, to improve future risk assessment levels, NASA is using Amazon’s AWS analysis tools to filter up to 1,000 sets of data at a time and train computer models to identify signals of impending solar storms. NASA’s approach links solar wind drivers to the level of magnetic fields around the Earth to identify anomalies in the data.
“With Amazon’s help, we can get every data on supersolar storms and use the anomalies we find to improve models for predicting and effectively classifying superstorms,” said Janet Kozyra, a solar physicist at NASA’s Washington headquarters. “