--- title: DeepMind’s hurricane breakthrough has surprised weather... description: Google DeepMind's WeatherNext AI model, published in Nature, can predict cyclones with unprecedented accuracy — giving forecasters on average one full day more... canonical: https://daily.dev/posts/deepmind-s-hurricane-breakthrough-has-surprised-weather-scientists-zcbbleayw twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: DeepMind’s hurricane breakthrough has surprised weather scientists | daily.dev og:description: Google DeepMind's WeatherNext AI model, published in Nature, can predict cyclones with unprecedented accuracy — giving forecasters on average one full day more... og:url: https://daily.dev/posts/deepmind-s-hurricane-breakthrough-has-surprised-weather-scientists-zcbbleayw og:image: https://api.daily.dev/og/posts/zcbBlEayw.png og:image:alt: DeepMind’s hurricane breakthrough has surprised weather scientists og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # DeepMind’s hurricane breakthrough has surprised weather scientists **[Ars Technica](https://daily.dev/sources/arstechnica)** · 2 min read · 25 upvotes · 5 comments ## Summary Google DeepMind's WeatherNext AI model, published in Nature, can predict cyclones with unprecedented accuracy — giving forecasters on average one full day more lead time than existing models. During Hurricane Melissa in October 2025, it predicted a Category 5 landfall in Jamaica with 80% confidence five days before impact. The model overcomes the challenge of sparse cyclone training data by training jointly on general weather data and cyclone data. Historically, gaining an extra day of forecast accuracy would take a decade of conventional research. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### How much earlier can DeepMind's WeatherNext model predict hurricanes compared to existing models? WeatherNext gives forecasters on average one full day more lead time than existing models — its three-day predictions are as accurate as previous models' two-day predictions. During Hurricane Melissa, it predicted a Category 5 landfall in Jamaica with 80% confidence five days before impact. Historically, gaining that extra day of forecast accuracy would have taken a decade of conventional research. _Teams working on applied ML for high-stakes domains track breakthroughs like this on daily.dev._ ### How did DeepMind train WeatherNext on cyclone data if extreme weather events are rare? WeatherNext was trained jointly on general weather data and cyclone-specific data. Because cyclone events are rare, there isn't enough cyclone data alone to train a reliable model. By combining abundant general weather data with cyclone data, the model learns both broad atmospheric patterns and the specific dynamics of extreme storm systems. _Developers applying ML to rare-event prediction problems find relevant research and discussion on daily.dev._ ## Community discussion Top comments from developers on daily.dev. **@bits\_and\_bytes** · 2 upvotes > finally helping to farmers **@hardik15** · 0 upvotes > Great news which can finally help in a much better way, A good use-case of SLM. **@justpew** · 0 upvotes > One extra day of lead time is huge when it's the difference between evacuating and not. **@agustinbarrientos** · 0 upvotes > When the model says 3%, does a storm actually happen about 3% of the time? **@tonygair** · 0 upvotes > Yes , so we need more AI and more climate change so we can predict even more devastating weather events, even some nice new categorys of weather incident. We all know which prediction matters here. Lets take note of that one. ## Similar posts on daily.dev - [AI breakthrough: WeatherNext predicts Hurricane Melissa](https://daily.dev/posts/ai-breakthrough-weathernext-predicts-hurricane-melissa-gbvitp2ed) · DeepMind · 0 upvotes · 0 comments - [Google’s new hurricane model was breathtakingly good this season](https://daily.dev/posts/google-s-new-hurricane-model-was-breathtakingly-good-this-season-zio9yzee7) · Ars Technica · 0 upvotes · 0 comments --- Tags: [#ai](https://daily.dev/tags/ai), [#machine-learning](https://daily.dev/tags/machine-learning) [View this post on daily.dev](https://daily.dev/posts/deepmind-s-hurricane-breakthrough-has-surprised-weather-scientists-zcbbleayw) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"DeepMind’s hurricane breakthrough has surprised weather scientists","url":"https://daily.dev/posts/deepmind-s-hurricane-breakthrough-has-surprised-weather-scientists-zcbbleayw","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/deepmind-s-hurricane-breakthrough-has-surprised-weather-scientists-zcbbleayw"},"datePublished":"2026-08-08T12:09:48.772Z","dateModified":"2026-08-08T12:10:13.295Z","description":"Google DeepMind's WeatherNext AI model, published in Nature, can predict cyclones with unprecedented accuracy — giving forecasters on average one full day more...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/b8a2f374ee30ce886d0be488860f8bf2?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/b8a2f374ee30ce886d0be488860f8bf2?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Ars Technica","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Ars Technica","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/80883e0e48a34b5ebcf93777016cb3fe","url":"https://daily.dev/sources/arstechnica"},"commentCount":5,"discussionUrl":"https://daily.dev/posts/deepmind-s-hurricane-breakthrough-has-surprised-weather-scientists-zcbbleayw","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":25},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":5}],"keywords":"ai,machine-learning","timeRequired":"PT2M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Ars Technica","item":"https://daily.dev/sources/arstechnica"},{"@type":"ListItem","position":3,"name":"DeepMind’s hurricane breakthrough has surprised weather scientists"}]} {"@context":"https://schema.org","@type":"WebPage","@id":"https://daily.dev/posts/deepmind-s-hurricane-breakthrough-has-surprised-weather-scientists-zcbbleayw","comment":[{"@type":"Comment","text":"finally helping to farmers","datePublished":"2026-08-09T07:40:27.738Z","url":"https://daily.dev/posts/zcbBlEayw#c-OU91wwfne","author":{"@type":"Person","name":"bits_007","url":"https://daily.dev/bits_and_bytes","image":"https://avatars.githubusercontent.com/u/182549945?v=4"},"interactionStatistic":{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":2}},{"@type":"Comment","text":"Great news which can finally help in a much better way, A good use-case of SLM.","datePublished":"2026-08-09T15:19:53.287Z","url":"https://daily.dev/posts/zcbBlEayw#c-b582ZOGQ5","author":{"@type":"Person","name":"Hardik","url":"https://daily.dev/hardik15","image":"https://media.daily.dev/image/upload/s--O0TOmw4y--/f_auto/v1715772965/public/noProfile"}},{"@type":"Comment","text":"One extra day of lead time is huge when it’s the difference between evacuating and not.","datePublished":"2026-08-10T08:12:58.103Z","url":"https://daily.dev/posts/zcbBlEayw#c-aJ0qxi2LF","author":{"@type":"Person","name":"Chris","url":"https://daily.dev/justpew","image":"https://lh3.googleusercontent.com/a/ACg8ocKoF76b4yIw_SQXJfF-jVBZeyeIZCmmUm1e4cc9fKDQoY52O2qN=s96-c"}},{"@type":"Comment","text":"When the model says 3%, does a storm actually happen about 3% of the time?","datePublished":"2026-08-16T03:25:27.346Z","url":"https://daily.dev/posts/zcbBlEayw#c-gJRZZTaDy","author":{"@type":"Person","name":"Agustin Barrientos","url":"https://daily.dev/agustinbarrientos","image":"https://media.daily.dev/image/upload/s--5ayxQnqn--/f_auto/v1783272917/avatars/avatar_wQYYVe5Tbj0NJ7C7qPoa8?_a=BAMAMicg0"}},{"@type":"Comment","text":"Yes , so we need more AI and more climate change so we can predict even more devastating weather events, even some nice new categorys of weather incident. We all know which prediction matters here. Lets take note of that one.","datePublished":"2026-08-17T07:05:57.831Z","url":"https://daily.dev/posts/zcbBlEayw#c-mKLTha5q5","author":{"@type":"Person","name":"tony gair","url":"https://daily.dev/tonygair","image":"https://lh3.googleusercontent.com/a/ACg8ocIaTFCRMIfYQlrq0BQ6KjY42IcYv1m5aIUf1MoOQGX_k_52n3M=s96-c"}}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/deepmind-s-hurricane-breakthrough-has-surprised-weather-scientists-zcbbleayw#faq","mainEntity":[{"@type":"Question","name":"How much earlier can DeepMind's WeatherNext model predict hurricanes compared to existing models?","acceptedAnswer":{"@type":"Answer","text":"WeatherNext gives forecasters on average one full day more lead time than existing models — its three-day predictions are as accurate as previous models' two-day predictions. During Hurricane Melissa, it predicted a Category 5 landfall in Jamaica with 80% confidence five days before impact. Historically, gaining that extra day of forecast accuracy would have taken a decade of conventional research. Teams working on applied ML for high-stakes domains track breakthroughs like this on daily.dev."}},{"@type":"Question","name":"How did DeepMind train WeatherNext on cyclone data if extreme weather events are rare?","acceptedAnswer":{"@type":"Answer","text":"WeatherNext was trained jointly on general weather data and cyclone-specific data. Because cyclone events are rare, there isn't enough cyclone data alone to train a reliable model. By combining abundant general weather data with cyclone data, the model learns both broad atmospheric patterns and the specific dynamics of extreme storm systems. Developers applying ML to rare-event prediction problems find relevant research and discussion on daily.dev."}}]} ```