Google has updated WeatherNext 3, its global weather AI model, with real-time satellite data. The upgrade makes forecasts sharper, refreshed every hour, and more reliable for fast-changing weather events.
WeatherNext 3 now delivers higher-resolution forecasts: 5 km for temperature and moisture, 10 km for surface variables, and 25 km for atmospheric data. That’s five times sharper than before.
The model learns from satellite mosaics and limited weather station data, giving better coverage for regions in Latin America, Africa, and Asia-Pacific where detailed forecasts were lacking.
The update also improves clean energy planning. WeatherNext 3 can predict wind speeds at turbine height, solar radiation, and cloud cover. This helps renewable energy operators plan output and balance demand.
Rain forecasts are more accurate too. By training on NASA IMERG data and Google’s satellite radar reanalysis, WeatherNext 3 improves precipitation predictions by up to 60% compared to older baselines. This means better warnings for storms and heavy rain.
Google has rolled WeatherNext 3 into Search, Gemini, Maps, Google Maps Platform, and Cloud. Precipitation forecasts are now up to 50% more accurate.
Also Read: Google offers PH college students free Gemini AI Plus access for 1 year
This update brings forecasts closer to real-world conditions and expands access to high-quality weather data worldwide.
