2023 will see the most intense solar storms in the last decade. We have produced two G4-class solar storms, multiple G3-class storms, and numerous disturbances associated with solar flares. Astronomers are aware that the next devastating solar storm event may not be too far away. However, at present there is not much we can do. NASA has its own prediction model, but its accuracy is not very high. In fact, in most cases, we cannot predict exactly when a storm will strike until it reaches Earth. But that’s about to change, thanks to a new technology NASA has developed in collaboration with the US Geological Survey and the Department of Energy’s Frontier Development Laboratory. NASA calls it DAGGER and it runs on AI and machine learning.
DAGGER, short for Deep Learning Geomagnetic Perturbation, collects data from various space satellites monitoring the Sun and from ground-based observatories looking for geomagnetic storms. This massive dataset is used to train AI through a technology called “deep learning.” Essentially, this is a form of machine learning in which the system learns the patterns that changes in the sun and explosions of coronal mass ejections (CMEs) can cause solar storms of a certain intensity on Earth. . This pattern recognition is the key to his DAGGER’s predictive ability.
And NASA is betting big on this model. “(DAGGER) can rapidly and accurately predict geomagnetic disturbances around the world 30 minutes before they occur,” the space agency said in a statement. According to the team, the model can generate predictions within a second, and predictions are updated every minute. “
NASA Uses AI to Build Solar Storm Prediction System
The team responsible for building DAGGER has been building this deep learning algorithm for over a decade. He was also able to quickly and accurately predict his two solar storms of 2011 and 2015 and their effects around the world.
This AI will enable us to make fast and accurate global predictions and inform decisions when solar storms occur, thereby minimizing or even preventing disruption to modern society.” said Vishal Upendran of the Interuniversity Center for Astronomy. He is the lead author of a paper on his DAGGER model published in Space Weather magazine.
The computer code for the DAGGER model is open source, and Upendran said it could be adopted by power grid operators, satellite controllers, telecommunications companies, and others with assistance to adapt the predictions to their specific needs. It is said that there is Such warnings could be taken to protect assets and infrastructure from an impending solar storm, such as temporarily taking sensitive systems offline or moving satellites to alternate orbits to minimize damage. may give time to teach
