A new study published in Science demonstrates that artificial intelligence weather models can now match or exceed the accuracy of traditional physics-based models for global weather forecasting. The research, led by scientists from the European Centre for Medium-Range Weather Forecasts (ECMWF) and several universities, evaluated multiple AI systems including Google’s GraphCast and Huawei’s Pangu-Weather. These AI models, trained on decades of historical weather data, were able to predict key atmospheric variables like temperature, wind, and humidity up to 10 days ahead with remarkable precision, often at a fraction of the computational cost of conventional models. The findings suggest a significant shift in meteorological science, where AI could soon become a standard tool for operational forecasting, though experts caution that integrating these systems with established methods and ensuring their reliability for extreme events remains a critical next step. Read the full article at: https://sciencedaily.com/releases/2024/07/240711215456.htm