Google DeepMind’s AI model adds five days to hurricane warnings
Google DeepMind’s WeatherNext Cyclones model won the 2026 Gizmodo Science Fair for predicting tropical cyclone tracks and intensity with unprecedented accuracy. During the 2025 Atlantic hurricane season, it gave forecasters an extra day of warning for Hurricane Melissa, reaching 80% confidence five days ahead and nearly 100% three days before landfall. A study in Nature confirmed the model’s…
Key points
- WeatherNext Cyclones predicts hurricane tracks and intensity with 80% confidence five days ahead, per a 2026 *Nature* study
- Model provides an extra day of warning on average, matching a decade’s progress in traditional forecasting
- Open-sourced and accessible via GitHub, requiring no supercomputer, for global forecasters
Developed by DeepMind researchers like Tom Andersson and Ferran Alet, the model addresses long-standing challenges: balancing track and intensity predictions, and handling rare but extreme storms. Unlike physical models, WeatherNext Cyclones runs on standard hardware and is open-sourced via GitHub. The team aims to expand global partnerships beyond the U.S. and Caribbean, citing climate-driven increases in storm intensity as a priority. The model’s ability to generate 1,000 scenarios also helps mitigate the ‘gray swan’ problem—underrepresented extreme events—critical for vulnerable regions.
Google DeepMind’s Hurricane Forecast Model Provides a Whole Extra Day of Warning
aol.com · 28 September 2026
Loading the full article…
This text was published by aol.com and written by AOL. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Research
All →- OpenAI claims new model solved Navier-Stokes and 100+ math problems · 3 src
- ConTP uses contrastive learning to predict transporter substrate specificity · 1 src
- German researchers introduce ReImaGin for image-based chain-of-thought reasoning · 1 src
- OpenAI contractors fired for using banned AI tools on ChatGPT review tasks · 1 src
- Zhipu automates infrastructure with GLM-5.3’s outer RSI loop in under two weeks · 1 src
Comments
via GitHub Discussions