{"version":1,"type":"story","url":"https://digestai.news/story/microsoft-releases-open-source-tool-to-correct-weather-model-bias","json":"https://digestai.news/story/microsoft-releases-open-source-tool-to-correct-weather-model-bias.json","markdown":"https://digestai.news/story/microsoft-releases-open-source-tool-to-correct-weather-model-bias.md","slug":"microsoft-releases-open-source-tool-to-correct-weather-model-bias","headline":"Microsoft Research introduces Probabilistic Bias Correction for longer-term weather forecasts","summary":"A team led by Microsoft Research published a study in *Nature* on October 14, 2026, introducing **probabilistic bias correction (PBC)**, a machine learning framework that improves weather forecasts for two- to six-week lead times. The authors claim PBC reduces systematic errors in probabilistic forecasts by learning corrections from historical data, rather than altering raw observations. The study reports gains of 26–37% in temperature accuracy and 16–32% in sea-level pressure forecasts when applied to ECMWF’s dynamical ensemble over 2016–2024, with 98% of grid cells showing improved precipitation predictions.\n\nThe team also tested PBC on AI-based models like ECMWF’s AIFS-SUBS and POET, doubling the latter’s mean skill score across variables. In the 2025 *AI Weather Quest* competition, PBC-powered **MicroDuet** (a combination of PBC-ECMWF and PBC-POET) ranked first for all variables and lead times, outperforming operational models from six forecasting centers. The authors highlight PBC’s ability to anticipate extreme events—such as a 2025 cold air outbreak in the U.S.—and improve flood forecasting accuracy by 86–150% for severe events. The code is open-sourced under an MIT license, with replication data available on GitHub and Zenodo.","keyPoints":["PBC improves weather forecasts 2–6 weeks ahead by correcting systematic errors in probabilistic models, per Microsoft Research study","Applied to ECMWF’s dynamical ensemble, PBC boosted temperature accuracy by 26–37% and precipitation skill in 98% of grid cells","MicroDuet, a PBC-powered ensemble, won the 2025 *AI Weather Quest* competition and outperformed six operational forecasting centers"],"whyItMatters":"PBC could reduce forecast errors in critical sectors like agriculture, energy, and disaster response by leveraging ML to correct biases in existing models without overhauling them. Its open-source release lowers the barrier for meteorological centers to adopt it.","category":{"slug":"research","name":"Research","url":"https://digestai.news/category/research"},"entities":{"companies":["Microsoft Research","Microsoft","MIT","Harvard College","University of Toronto","IMPA"],"models":["PBC (Probabilistic Bias Correction)","Debias++","Persistence++","AIFS-SUBS","POET","FuXi-S2S"],"people":["Hannah Guan","Soukayna Mouatadid","Judah Cohen","Lester Mackey"]},"firstPublishedAt":"2026-10-14T09:05:00Z","updatedAt":"2026-10-14T16:32:10Z","sourceCount":2,"hasPrimarySource":false,"sources":[{"outlet":"Unite.AI","title":"Microsoft Research Publishes AI Weather Forecasting Study in Nature","url":"https://unite.ai/microsoft-research-publishes-ai-weather-forecasting-study-in-nature","publishedAt":"2026-10-14T16:32:10Z","type":"press","primary":false,"lead":true},{"outlet":"theverge.com","title":"Microsoft has a new machine learning tool to improve weather forecasts","url":"https://theverge.com/science/1011127/ai-weather-forecast-microsoft","publishedAt":"2026-10-14T09:05:00Z","type":"press","primary":false,"lead":false}],"sourceNotes":null,"discussions":[],"thread":null,"cite":{"text":"Digest AI, \"Microsoft Research introduces Probabilistic Bias Correction for longer-term weather forecasts\", 14 October 2026, https://digestai.news/story/microsoft-releases-open-source-tool-to-correct-weather-model-bias","publisher":"Digest AI","title":"Microsoft Research introduces Probabilistic Bias Correction for longer-term weather forecasts","datePublished":"2026-10-14T09:05:00Z","url":"https://digestai.news/story/microsoft-releases-open-source-tool-to-correct-weather-model-bias"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}