Google launches weathernext 3, a high-resolution AI weather forecasting model

Google has unveiled WeatherNext 3, an advanced AI-driven weather forecasting model that delivers high-resolution, real-time forecasts using satellite data. The model provides hourly updates with five times greater spatial resolution than its predecessor, WeatherNext 2, offering improved accuracy for agriculture, renewable energy, and daily planning. It is now integrated into Google Search, Maps, and Cloud platforms. WeatherNext 3 leverages live satellite observations to generate forecasts every hour, enabling more precise predictions for rapidly changing weather conditions. It offers forecasts at 5-kilometer resolution for surface variables, 10-kilometer for other surface variables, and 25-kilometer for atmospheric variables. The model also includes specialized predictions for renewable energy, such as 100-meter wind speeds and solar radiation levels, aiding clean energy planning. Google DeepMind and Google Research claim the model significantly improves precipitation forecasting, with up to 60% better accuracy compared to existing benchmarks. The model is designed to provide reliable forecasts globally, particularly benefiting regions in Latin America, Africa, and Asia-Pacific that previously lacked high-resolution forecasting capabilities.