Google has rolled out WeatherNext 3, its latest AI-powered weather forecasting model designed specifically for the energy industry. This new version zeroes in on variables that directly impact power generation estimates and grid risk: wind speed at 100 meters, cloud cover, and the amount of sunlight hitting the ground.

New Variables Tailored for Energy

Pinpointing wind speed at 100 meters is a game-changer for wind energy calculations, while tracking cloud cover and solar radiation is crucial for forecasting solar plant output and potential grid loads. This set of features aims to give grid operators and energy planners a sharper picture of the conditions that matter most for generation planning and risk management.

Higher Resolution and Hourly Updates

WeatherNext 3 delivers global forecasts updated every hour. For ground-level variables like temperature and humidity, the model offers fine-grained detail down to 5 km. Precipitation forecasts are resolved at 11 km. This level of spatial and temporal precision makes it easier for energy teams to plan ahead and respond quickly to changing weather conditions.

How It Stacks Up to the Previous Version

Compared to WeatherNext 2, which worked on a 25 km grid, the new model is far more detailed and responsive. The shift to a smaller grid and hourly updates means forecasts are much better suited for the energy sector, where local differences and real-time data can make all the difference.