Russian AI System Predicts Arctic Storms Quickly and Accurately
A new AI model developed by Russian researchers delivers highly detailed forecasts of dangerous Arctic winds and storms, performing up to 50 times faster than traditional physical models
Scientists from the Moscow Institute of Physics and Technology and the Russian Academy of Sciences’ Shirshov Institute of Oceanology have introduced a neural network designed to forecast extreme weather events in the Arctic. Tested in the Barents and Kara Seas, the AI system predicts the formation of hazardous vortices and storms with high spatial detail.
Despite its speed, the system maintains strong analytical quality. To train the model, researchers used hydrodynamic data from the WRF system spanning 2015 to 2021. They then validated the AI on more recent datasets, where it successfully captured overall weather dynamics and simulated specific dangerous events that the global ERA5 model failed to detect.
The development will support safer navigation and industrial activities across the Arctic. Russian scientists are also using advanced technologies to improve forecasts of melting sea ice and overall ice conditions in the region—crucial data for the global climate system, given the Arctic’s influence on Earth’s environmental balance.
