New AI Model Predicts Cloud Patterns Hours in Advance for Weather Forecasting

Researchers have developed KM-UNet, a hybrid deep-learning system that forecasts cloud movements and formations with improved accuracy by combining two advanced neural network architectures. The breakthrough could enhance weather prediction for aviation, renewable energy planning, and climate monitoring—industries where accurate short-term forecasts directly impact operations and costs.

Originaltitel: KM-UNet: A KAN-Mamba Hybrid Network with Direction-Sensitive Attention and Multi-Scale Fusion for Cloud Mask Nowcasting