DeepMind Scientist Explains How AI Predicts Cyclones
Google published a post on its own site titled "Ask a Scientist: How do researchers use AI to predict a cyclone?", dated September 1, 2026. It is a Q&A with Ferran Alet, a research scientist at Google DeepMind who helped build the WeatherNext models. Everything below comes from it.
Transcript
A DeepMind scientist explained today how AI predicts cyclones. Forecasts once needed supercomputers as big as three story buildings.
Weather forecasting historically gained one day of accuracy per decade. Google says its AI models delivered that decade of progress at once.
Google says WeatherNext trained on fifty years of weather data. These faster models run on a single TPU instead of building sized machines.
Google says its predictions helped the National Hurricane Center warn that Melissa would intensify from category one to category five.
The extra day of warning helped Jamaica prepare. Agencies want forecasts extended from five days to seven. Google open sourced its WeatherNext models.
Inboxsmith builds AI phone and email agents for small business. Please like and subscribe for more news.
Sources
Every claim in this video comes from the top ranking coverage of this topic. The claims and where each one came from:
- We sat down with Ferran Alet, a research scientist at Google DeepMind who helped develop our WeatherNext models, to explore how this breakthrough technology is giving people and first responders the vital gift of time before a cyclone strikes.(Google's official announcement, Ask a Scientist: How do researchers use AI to predict a cyclone?, published September 1, 2026)
- Historically, forecasting cyclones relied on physics-based models powered by supercomputers the size of shipping containers, or even as big as three-story buildings.(Google's official announcement, Ask a Scientist: How do researchers use AI to predict a cyclone?, published September 1, 2026)
- The meteorological community historically gained about one day of forecast accuracy per decade, by building better models, launching more satellites and getting more computing power. By applying AI to the data from these physics-based models, we saw that same decade of progress in a single generation leap of WeatherNext models.(Google's official announcement, Ask a Scientist: How do researchers use AI to predict a cyclone?, published September 1, 2026)
- With WeatherNext, we trained AI on 50 years of historical weather data. Instead of having to rely on building-sized supercomputers, researchers can run these faster models on a single TPU.(Google's official announcement, Ask a Scientist: How do researchers use AI to predict a cyclone?, published September 1, 2026)
- What was notable about Melissa was that, while the hurricane initially appeared weak, the AI's high-confidence predictions helped the National Hurricane Center to issue a historic forecast, warning authorities of its rapid intensification from a Category 1 storm into a Category 5.(Google's official announcement, Ask a Scientist: How do researchers use AI to predict a cyclone?, published September 1, 2026)
- The additional day of warning allowed authorities in Jamaica to make life-saving emergency preparations.(Google's official announcement, Ask a Scientist: How do researchers use AI to predict a cyclone?, published September 1, 2026)
- Weather agencies are looking to extend their public forecast horizons from five days to seven, giving communities vital extra time to prepare before a cyclone hits.(Google's official announcement, Ask a Scientist: How do researchers use AI to predict a cyclone?, published September 1, 2026)
- We’ve open-sourced our WeatherNext 2 and WeatherNext cyclones models, and developed an interactive Weather Lab website to advance science for a global scientific community.(Google's official announcement, Ask a Scientist: How do researchers use AI to predict a cyclone?, published September 1, 2026)