Arctic signals could help predict winter weather

Researcher Judah Cohen links Arctic conditions to winter patterns and is combining those diagnostics with new AI tools to improve subseasonal (two-to-six-week) forecasts; a team he worked with won the 2025 AI WeatherQuest fall subseasonal competition.

Judah Cohen, a research scientist in MIT's Department of Civil and Environmental Engineering, tracks Arctic conditions to inform winter outlooks across Europe, Asia, and North America. His work on Siberian snow cover dates back to postdoctoral research with Professor Dara Entekhabi. This season he highlights Arctic indicators alongside a new generation of artificial intelligence tools to build a subseasonal atmospheric picture. Cohen notes that El Niño-Southern Oscillation is relatively weak this year, which raises the relative importance of high-latitude drivers. Key facts: - Cohen monitors October Siberian weather and snow cover, Arctic sea-ice extent, early-season temperature shifts, and polar vortex stability as subseasonal diagnostics. - This autumn Siberia was colder than normal with early snowfall, even as much of the Northern Hemisphere experienced an unusually warm October. - Warm Barents-Kara Sea temperatures and an easterly phase of the quasi-biennial oscillation are among factors that suggest a potentially weaker polar vortex in early winter. - A research team working with Cohen won first place in the fall 2025 AI WeatherQuest subseasonal forecasting competition run by the European Centre for Medium-Range Weather Forecasts (ECMWF). - The winning model combined machine-learning pattern recognition with Arctic diagnostics and showed gains in multi-week temperature forecasting, including an early signal for a possible mid-December cold surge on the U.S. East Coast. - Cohen's outlook indicates a greater chance of colder-than-normal conditions across parts of Eurasia and central North America later in the winter. Summary: If the AI-enhanced performance holds across multiple seasons, researchers say it could extend the lead time for predicting some impactful winter events. Cohen will continue to update his outlook through the season; broader validation across seasons is undetermined at this time.