Google's New AI Weather Model Sharpens Storm Forecasts
Google has rolled out WeatherNext 3, the latest version of its AI-driven weather forecasting system, claiming major gains in accuracy and detail. The company says the model can now generate global forecasts at what it calls "unprecedented resolution," roughly five times sharper than the previous WeatherNext model.
The improvement comes from training the model on a wider range of real-time observational data rather than relying solely on the historical simulation data most AI weather systems use. According to Google researcher Samier Merchant, this lets WeatherNext 3 capture atmospheric patterns more precisely, which should translate into better predictions for tricky events like rainfall and snowfall timing and intensity.
Google plans to integrate the model into its own weather-related products and make it available to researchers and forecasting partners, continuing its push to out-compete traditional physics-based forecasting systems with machine learning.