Artificial intelligence (AI) could have predicted the powerful solar storm that impacted Earth in May last year, triggered by the highly active region AR13664 on the Sun, according to a new study. The team of researchers at the University of Genoa, led by Sabrina Guastavino, stated that by training AI on historical solar events, it could identify patterns that precede coronal mass ejections (CMEs).
This early warning system is crucial because traditional methods, which rely on human analysis of solar images and data, often provide less accurate and timely forecasts.
“Despite advances in observational capabilities and model development, so far there remains a substantial uncertainty in both flare forecasting and CME travel time predictions, with the latter averagely amounting to approximately 12 hours,” the study stated.
The AI model works by processing vast amounts of data from solar imagery and other space weather parameters. It looks for subtle cues in the solar atmosphere’s behaviour, such as changes in magnetic field strength, solar wind speed, and the appearance of solar flares.
“The May 2024 event also underscores the broader implications of AI-driven reverse engineering for space weather science,” the study highlighted.
“The ability to predict CME travel times with such precision suggests that AI can furthermore serve as a diagnostic tool for testing and refining existing models of CME propagation,” it added.
What is coronal mass ejection?
Solar storms or powerful CMEs are massive expulsions of plasma and magnetic fields from the Sun’s corona. The huge clouds of electrified gas travel at speeds of hundreds of miles per second and can affect power grids, communications, GPS navigation, air travel and satellites.
As per ISRO, last year’s geomagnetic storm was the most intense since 2003, causing disruptions to communication and GPS systems.
Also Read | Solar Storm That Recently Hit Earth Was Most Intense Since 2003: ISRO
Solar flares
Solar activity follows a pattern with peaks and lows occurring every 11 years. Scientists describe these cycles as solar maximum and solar minimum, which are driven by the Sun’s magnetic field.
Currently, Solar Cycle 25 is underway which is expected to reach a solar maximum around July 2025. This cycle has shown greater activity than anticipated by NASA and the NOAA (National Oceanic and Atmospheric Administration), with the official prediction of around 115 sunspots at the peak.
Despite having a greater amount of data at disposal, scientists are unsure why the Sun has been more active than expected. Further observation is necessary to improve future predictions and increase our understanding of the Sun’s internal processes.