Powerful solar storms could have damaging effects on our infrastructure on Earth, affecting satellites and probes as well as communications and navigation systems, power grids, oil pipelines, and even our Internet and drinking water supplies. Strong evidence of this comes from a recent geomagnetic storm that caused the crash of 40 of the 49 recently launched Starlink satellites.
This solar activity is monitored by a number of satellites and space probes. However, current early warning systems take too long to warn of potentially dangerous solar storms, leaving little time to take appropriate precautionary measures on Earth. So far, the problem is not the lack of data or AI models to calculate them, but the challenges lie in the speed of the calculations. This has so far prevented timely indication of exactly which regions could be affected by an impending solar storm.
Recognizing these risks, Nasa has now developed its own deep-learning model using existing data that provides 30 minutes of warning before a potentially dangerous solar storm hits us. This new model, called Nasa-Dagger, produces a prediction every minute. According to Nasa researchers, the algorithm can predict the strength and direction of a solar storm in less than a second.
The researchers presented the exact results and how their Dagger model works in the study, “Global Geomagnetic Perturbation Forecasting Using Deep Learning,” published in the online journal Space Weather.
According to Nasa, this algorithm provides accurate advance warning for any location on Earth. Especially because solar activity is currently reaching its maximum again, such accurate forecasts are of great importance.
To our knowledge, the largest recorded solar storm occurred more than 14,000 years ago and struck Earth and its magnetic field. A total of nine such major events, known as Miyake events, are known to have occurred, although they have never been measured with modern instruments. The last known Miyake event is believed to have occurred about 1,000 years ago. Research and development in this area therefore take a high priority to protect our infrastructure and technology from such potentially catastrophic events.
