A new study demonstrates that artificial intelligence can generate weather forecasts with accuracy comparable to traditional physics-based models, but in a fraction of the time. Researchers developed a machine learning model trained on decades of historical weather data. The AI system produces detailed forecasts for variables like temperature and pressure up to ten days ahead. While the current model operates at a lower resolution than top-tier conventional systems, its speed—generating forecasts in under a minute on a single computer—highlights a significant efficiency gain. Experts note that AI models do not replace the need for understanding atmospheric physics but could serve as powerful, rapid supplemental tools. Further development is needed to improve resolution and integrate more complex atmospheric processes. Read the full article: https://sciencedaily.com/releases/2024/10/241008150927.htm