Storm surge in color: RGB-encoded physics-aware deep learning for storm surge forecasting

Storm surge forecasting plays a crucial role in coastal disaster preparedness, yet existing machine learning approaches often suffer from limited spatial resolution, reliance on coastal station data, and poor generalization. Moreover, many prior models operate directly on unstructured spatial data, making them incompatible with modern deep learning architectures. In this work, we introduce a novel approach that projects unstructured water elevation fields onto structured Red Green Blue (RGB)-encoded image representations, enabling the application of Convolutional Long Short Term Memory (ConvLSTM) networks for end-to-end spatiotemporal surge forecasting. Our model further integrates ground-truth wind fields as dynamic conditioning signals and topo-bathymetry as a static input, capturing physically meaningful drivers of surge evolution. Evaluated on a large-scale dataset of synthetic storms in the Gulf of Mexico, our method demonstrates robust 48-hour forecasting performance across multiple regions along the Texas coast and, critically, exhibits strong generalization to previously unseen coastal areas without retraining. This RGB encoding approach achieves sub-second inference times while maintaining median R² > 0.9 for most forecast hours, a speed well suited to rapid ensemble-based uncertainty quantification. By combining a structured representation, physically grounded forcings, and proven spatial transferability, this study advances operational storm surge forecasting capabilities.