A new study demonstrates that artificial intelligence systems can generate accurate weather forecasts faster and with less computational power than conventional physics-based models. Researchers developed an AI model trained on decades of historical weather data, which learned to predict atmospheric conditions by identifying complex patterns. In tests, the AI system produced short-term forecasts for temperature, pressure, wind, and humidity that matched or exceeded the accuracy of leading numerical weather prediction models, while completing calculations in minutes instead of hours. The approach highlights machine learning’s potential to revolutionize meteorology by providing rapid, high-resolution forecasts. However, scientists note that long-term reliability and the AI’s ability to predict extreme, rare events require further validation. For the full details, read the complete article at https://sciencedaily.com/releases/2024/05/240521123456.htm.