Google’s AI model GenCast has been found to outperform the European Centre for Medium-Range Weather Forecasts’ ENS system in weather forecasting, with accuracy rates reaching 97.2% of the time. Developed by the DeepMind team, GenCast uses an ensemble approach of multiple predictions to create a complex probability distribution of future weather scenarios.
Unlike its previous model, which provided a single estimate, GenCast offers a more comprehensive view of possible weather trajectories. The team trained GenCast on weather data up to 2018 and compared its forecasts for 2019, showing significant improvements over ENS.
GenCast is part of Google’s suite of AI-based weather models, set to be incorporated into Google Search and Maps. The model will also provide real-time and historical forecasts that can be used by researchers and developers.
Source: https://techcrunch.com/2024/12/07/google-says-its-new-ai-model-outperforms-the-top-weather-forecast-system