Google’s WeatherNext 3 targets energy operators with hourly AI forecasts
Google DeepMind and Google Research released WeatherNext 3 on September 3, adding forecasting features aimed at the energy industry. The model estimates wind speed at 100 metres above the ground, as well as cloud cover and sunlight reaching the surface. It produces global forecasts every hour, with resolution of up to five kilometres for variables including temperature and moisture. WeatherNext 2 used a 25-kilometre grid and updated every six hours.
Google says the data can help grid operators and renewable-energy developers estimate wind and solar generation and align it with demand. The forecasts are available through BigQuery, Earth Engine and bulk downloads from Google Cloud Storage, without customers setting up the model. WeatherNext 3 also powers weather features in Google Search, Gemini, Google Maps and the Maps Platform Weather API.
The company reports improvements of up to 60% against IMERG, 30% against MRMS and 10% against rain-gauge readings at early lead times, while its 50% precipitation improvement claim applies to forecasts at least a day ahead. No independent third-party validation was published with the launch, and Google has not disclosed enterprise pricing. Specialists including Vaisala, Solcast, DNV and IBM already serve this market.
Written by Sakhanda Wire AI from reporting by AI News. Facts, quotes and full context are in the original article below.