A new artificial intelligence weather forecasting model developed by researchers has demonstrated superior accuracy compared to conventional physics-based systems in global medium-range forecasts. The model, named Aurora, was trained on over a million hours of diverse weather and climate data using a foundation model approach. In benchmark tests, it significantly outperformed the European Centre for Medium-Range Weather Forecasts (ECMWF) system—considered the gold standard—in predicting global weather patterns up to ten days in advance. The AI system achieves high-resolution forecasts in under a minute, a task that takes traditional supercomputers hours to complete. Researchers highlight its potential for rapid, efficient predictions of extreme weather events, though note it is designed to complement rather than replace existing physical models. The full details of the model’s architecture and performance metrics are available in the research publication. Read the full article at https://technologyreview.com/2024/05/13/ai-weather-model-aurora-outperforms-traditional-forecasts.