Hurricane Center Uses Google AI Forecast Model
The National Hurricane Center has used Google DeepMind's AI model in operations since late June 2025, and the model was later added to some track and intensity consensus aids in 2026.
Updated on Nov. 11, 2026 in Forecasts
Live Poll
Do you think AI's benefits for weather forecasting outweigh its energy costs?
The U.S. National Hurricane Center began using Google DeepMind's AI model for hurricane forecasting in late June 2025. The model was added to some track and intensity consensus aids in 2026, and the center is working with Google on developing its successor.
Why it matters
A DeepMind paper published in Nature in August found the model gave forecasters on average an extra day's worth of predictive accuracy. The center also relies on AI models from the European Centre for Medium-Range Weather Forecasts and the National Weather Service.
The Nature paper reports an average gain of one day of predictive accuracy, with 3-day forecasts comparable to earlier models' 2-day forecasts. The model recognized Hurricane Isaias could become a tropical cyclone a full week before it formed, though broader performance across all storms is not detailed in the source.
The players
National Hurricane Center
The National Hurricane Center is the U.S. federal agency that tracks and forecasts tropical cyclones, operating under the National Weather Service.
Google DeepMind
Google DeepMind is Google's artificial intelligence research lab, which developed the weather forecasting model used by the center.
Dan Brown
Dan Brown serves as Operations Division Chief at the National Hurricane Center.
The details
The model is the same one Google uses for everyday weather, pushed to its limits for extreme weather. It generates 1,000 possible scenarios for each cyclone and forecasts rainfall alongside track and intensity.
Timeline
Late June 2025: National Hurricane Center begins operational use of the AI model.
August 2025: DeepMind publishes accuracy findings in Nature.
2026: Model added to NHC track and intensity consensus aids.
Seasonal Patterns
The adoption of Google DeepMind's everyday weather model for extreme hurricane forecasting extends its use beyond routine conditions. The change comes as the Pacific recorded a record number of hurricanes in 2026 during an intense El Niño.
Residents in hurricane-prone areas can rely on official National Hurricane Center forecasts for storm tracks and intensity. The center's AI-assisted models aim to give forecasters more lead time before storms form and strengthen.
The takeaway
Improved AI forecasting gives forecasters earlier warning of developing storms, which can help communities prepare. Following official National Hurricane Center updates remains the most reliable way to track storms.
Further reading
Explore more in the Forecasts section for context on how these models are changing storm predictions.
Live Poll
Do you think AI's benefits for weather forecasting outweigh its energy costs?
