NASA's AI-enabled predictions can offer advance notice to prepare for solar storms

By utilising AI to analyse spacecraft measurements of the solar wind, the model can anticipate the location of an approaching solar storm with 30 minutes of advance notice. This valuable time allows for necessary preparations to eliminate potential damage to power grids and vital infrastructure. Solar wind is a continuous stream of charged particles, mainly electrons and protons, that sheds from the Sun's outer atmosphere. It carries energy and magnetic fields throughout the solar system.

Occasionally, the sun experiences intense disturbances called solar storms, which result from sudden release of energy. These storms have a significant impact on Earth’s space environment, causing disruption in satellite communication, power grids, and navigation systems. To better understand and predict solar storms, Nasa has harnessed the power of Artificial Intelligence(AI).

What is a solar storm?

The Sun continuously sheds solar material into space, including a steady stream called the "solar wind" and more powerful bursts from solar eruptions. When this material comes in contact with Earth's magnetic environment, known as the "magnetosphere," it can give rise to geomagnetic storms. These storms differ in intensity and can have disruptive effects on our technology-dependent world.

An international team of researchers, including NASA, the U.S. Geological Survey, and the U.S. Department of Energy, is using artificial intelligence (AI) to find links between the solar wind and disruptive geomagnetic disturbances. These disturbances can wreak havoc on our technology, and the team aims to better prepare for them. The researchers used "deep learning," an AI technique that trains computers to identify patterns using previous examples.

1/ AI-powered solar storm prediction model

DAGGER stands for Deep Learning Geomagnetic Perturbation.

Put simply it predicts solar storms 30 mins before they occur.

• Produces predictions in less than 1 second
• Predictions update every minute
• Combines AI & satellite data

— Alex Banks (@thealexbanks) April 8, 2023

DAGGER: The AI Forecaster

DAGGER, or Deep Learning Geomagnetic Perturbation, is a computer model that accurately forecasts worldwide geomagnetic disturbances with a 30-minute prior notice. The model generates predictions in less than a second and updates them every minute.

The DAGGER team conducted tests on the model using two geomagnetic storms that happened in August 2011 and March 2015. In both happenings, DAGGER successfully predicted the global impacts of the storms with accuracy. Alex Banks on Twitter: "Last week NASA released an AI system ... Earlier prediction models used AI to forecast local geomagnetic conditions for specific places on Earth. However, non-AI models offered global predictions that were not timely.

DAGGER is a groundbreaking prediction model that combines AI with real-time measurements from space and Earth to provide accurate and timely geomagnetic forecasts for locations worldwide and is very timely. The DAGGER model's computer code is available as open source. Upendran(Atmospheric scientist at NASA Goddard Space Flight Center) suggests that power grid operators, satellite controllers, telecommunications companies, and others can adopt the model and customise it to meet their specific requirements.

By utilising the predictions, they can receive advance warnings about solar storms and take necessary measures to safeguard their assets and infrastructure. These measures may include temporarily shutting down sensitive systems or repositioning satellites to minimise potential damage.

In the future, models like DAGGER could enable the development of solar storm sirens in power stations and satellite control centres worldwide. Similar to how tornado sirens warn about approaching severe weather, these solar storm sirens would provide an alarm system for potential solar storms approaching towards us.