Decoding the Arctic to predict winter weather
AI Summary: Judah Cohen, a research scientist at MIT, has developed a new AI-based subseasonal forecasting model that enhances winter weather predictions by integrating Arctic climate indicators. His model, which won first place in the 2025 AI WeatherQuest competition, combines machine-learning techniques with traditional Arctic diagnostics, demonstrating significant improvements in forecasting temperature patterns over two to six weeks. This year's findings suggest that colder temperatures and early snowfall in Siberia could lead to increased cold air masses affecting Europe and North America. If the model's performance is consistent across seasons, it could enable earlier warnings for extreme weather events, providing critical lead time for preparation by utilities and public agencies.