<ul data-eligibleForWebStory="true"><li>This paper introduces a convolutional neural network model called LUPIN for precipitation nowcasting that integrates data-driven learning with physics-based domain knowledge.</li><li>LUPIN utilizes a Lagrangian Double U-Net architecture with components for generating motion fields, extrapolation, and capturing precipitation evolution.</li><li>The model is fully differentiable and GPU-accelerated, enabling end-to-end training and inference with a data-driven Lagrangian coordinate system transformation.</li><li>Evaluation results show that LUPIN performs comparably or better than existing AI models in extreme event scenarios, demonstrating the potential of Lagrangian machine learning approaches.</li></ul>

Fully Differentiable Lagrangian Convolutional Neural Network for Physics-Informed Precipitation Nowcasting

Discover more