A new artificial intelligence weather forecasting model developed by Google DeepMind has demonstrated superior accuracy to current state-of-the-art systems in global medium-range weather predictions. The model, named GraphCast, outperformed the European Centre for Medium-Range Weather Forecasts’ high-resolution forecast (HRES) in over 90% of 1,380 test variables, including critical measurements for tracking severe weather events. GraphCast operates significantly faster than traditional numerical weather prediction models, generating a 10-day forecast in under one minute on a single Google TPU v4 machine, compared to the hours of computation required by conventional methods. The AI system was trained on decades of historical weather data from the ECMWF’s ERA5 reanalysis dataset to learn the cause-and-effect relationships that govern atmospheric physics. While promising, the model is currently designed as a research tool and is not intended to replace existing operational systems, which incorporate more observational data and human expertise. Read the full article at: https://technologyreview.com/2023/11/14/1083821/ai-weather-forecast-model-graphcast-beats-traditional-systems/