MIT Doubles Accuracy of 2-6 Week Weather Forecasts With ML Postprocessing of ECMWF Models
nordicinst · x · 2026-10-14
MIT researchers developed a machine-learning model that learns to correct errors in existing forecasting systems, doubling accuracy for the 2-6 week subseasonal window—a critical gap for emergency planning, demonstrated against events like Texas's 2021 grid failure. Built as postprocessing atop ECMWF models, the approach also improves extreme weather prediction and is open and low-cost, with potential global impact on energy planning, farming, and disaster prep.
Related event: MIT machine learning doubles accuracy of 2-6 week forecasts(2 posts)→
More from Research
- EMNLP paper: diffusion LMs internally encode latent denoising timestep representations — gsarti_ · 2026-10-23
- KU Leuven and VIB open two Group Leader professorships in AI for Biology, start 2027 — jpillowtime · 2026-10-23
- Subspace Tuning curbs LLM catastrophic forgetting by freezing principal singular directions — scottniekum · 2026-10-23
- Perceptic launches PercepticOS, an AI workspace for 'scientific superintelligence' — nathanbenaich · 2026-10-23
- EPFL's Bosselut to talk long context as test-time learning at ELLIS seminar — andre_t_martins · 2026-10-23
- DRATS: Adaptive Task Sampling Boosts Worst-Task Performance in Multi-Task RL — Cohere · 2026-10-23