Google unveils high-resolution AI weather model with real-time satellite data
Google launched WeatherNext 3, an AI-driven forecasting system that uses live satellite observations to deliver sharper, hourly predictions, especially for rain and snow.
Google announced the rollout of WeatherNext 3, its latest AI weather model that creates global forecasts five times sharper than the earlier WeatherNext 2, which ran on a 25-kilometre grid every six hours. Leveraging real-time satellite observations, the model updates hourly and can display temperature and moisture at a 5-kilometre scale, improving short-term rain and snowfall predictions. Google claims precipitation forecasts a day out can be up to 50 percent more accurate, and the model now includes wind speed estimates at 100 metres for renewable-energy applications.
The technology has been embedded across Google Search, Maps, Gemini and other products, and the company is working with the US National Hurricane Center and Asian weather agencies to enhance forecasting. While AI predictions are faster, they will continue to be used alongside traditional physics-based models.
Why it matters
More precise, timely forecasts can improve safety, agriculture and renewable-energy planning worldwide.
In this story
