Tech And Science

A new AI tech can anticipate a solar storm 30 minutes in advance

Artificial Intelligence training water consumption
Source: Pixabay

In the midst of rising solar storms, a group of worldwide academics, including Indians, has collaborated with NASA to create a new computer model that combines artificial intelligence (AI) and satellite data to warn of potentially hazardous space weather.

Scientists forecast an increase in solar storms since the Sun is approaching a peak of activity that occurs every 11 years and is projected to occur around 2025.

The consequences of these magnetic storms can range from minor to severe, but in an increasingly technological society, their effects are becoming increasingly disruptive.

The new model use artificial intelligence to analyse spacecraft observations of the solar wind (a never-ending stream of particles from the Sun) and estimate where an imminent solar storm would strike, anywhere on Earth, with 30 minutes’ notice.

This may be just enough time to prepare for big storms and avoid major damage to power systems and other important infrastructure.

“With this AI, it is now possible to make rapid and accurate global predictions and inform decisions in the event of a solar storm, thereby minimising — or even preventing — devastation to modern society,” said Vishal Upendran of the Inter-University Centre for Astronomy and Astrophysics (IUCAA) in India.

The researchers at the Frontier Development Lab, a public-private collaboration comprised of NASA, the US Geological Survey, the US Department of Energy, and the IUCAA, used an AI technology known as “deep learning” to create a computer model known as DAGGER (formally, Deep Learning Geomagnetic Perturbation).

DAGGER, as described in the journal Space Weather, can anticipate geomagnetic disturbances worldwide 30 minutes in advance. It can make predictions in less than a second and updates them every minute.

According to the researchers, solar storm sirens like DAGGER might one day sound an alert at power plants and satellite control centres throughout the world.

The DAGGER team put the model through its paces against two geomagnetic storms that occurred in August 2011 and March 2015. In each case, DAGGER was able to predict the storm’s global impact immediately and precisely.

DAGGER is the first to integrate AI’s rapid analysis with real-world data from space and around the globe to create regularly updated forecasts that are both immediate and exact for sites throughout the world.

According to Upendran, the computer code in the DAGGER model is open source, and it may be adopted with assistance by power grid operators, satellite controllers, telecommunications corporations, and others to adapt the predictions to their unique needs.

Such warnings might provide businesses with enough time to take precautionary measures to protect their assets and infrastructure from an oncoming solar storm, such as temporarily shutting down key systems or shifting satellites to alternative orbits to reduce damage.