Google Unveils AI Model That Surpasses Leading Weather Forecasting Systems
Google’s DeepMind team unveils AI Model named GenCast that surpasses leading weather forecasting systems.
In a study published in Nature, researchers revealed that GenCast outperforms the European Centre for Medium-Range Weather Forecasts’ (ECMWF) ENS, widely regarded as the world’s leading operational forecasting system.
In a blog post, DeepMind explained the technology in simpler terms: unlike their earlier deterministic model that provided a single best estimate of future weather, GenCast generates an ensemble of 50 or more predictions, each representing a potential weather trajectory. This approach creates a complex probability distribution of future weather scenarios.
To evaluate GenCast’s performance, the team trained the model on weather data up to 2018 and compared its forecasts for 2019 against ENS. Remarkably, GenCast proved more accurate 97.2% of the time.
Google is integrating GenCast into its suite of AI-driven weather tools, including Google Search and Maps. Additionally, it plans to release real-time and historical forecasts from GenCast, enabling researchers and developers to incorporate its predictions into their own models and studies.
