A new study suggests that machine learning algorithms can accurately predict solar flares and coronal mass ejections (CMEs), offering hope for improved forecasting of geomagnetic storms. The research, led by Sabrina Guastavino from the University of Genoa, used decades of solar activity data to identify patterns associated with increased solar activity in the region AR13664.
Coronal Mass Ejections are massive bursts of plasma ejected into space due to disruptions in the Sun’s magnetic field. These events can travel at speeds up to several thousand kilometers per second and interact with Earth’s magnetosphere, triggering geomagnetic storms that disrupt satellite communications, GPS systems, and power grids.
The study applied artificial intelligence to predict solar activity associated with a significant storm in May 2024, including an X8.7 flare. The results showed ‘unprecedented accuracy’ in forecasting the occurrence of solar flares, CME production, and geomagnetic storms. The impact is profound, offering potential solutions to power grid outages, communication issues, and satellite problems caused by these events.
For sky watchers, this advance could also lead to better forecasts of auroral activity. With improved solar forecasting, astronomers aim to reduce uncertainties associated with traditional methods, providing a safer and more reliable connection between space and Earth’s magnetic field.
Source: https://www.sciencealert.com/ai-can-predict-incredible-solar-storms-before-they-strike