🤖 AI Summary
To address the computational inefficiency of physics-based models (e.g., HEC-RAS) in real-time flood forecasting for large watersheds in South Florida, this study develops a suite of deep learning surrogate models—including MLP, RNN, LSTM, CNN, and RCNN—trained on historical water levels, real-time meteorological data, and auxiliary covariates to enable rapid stage prediction along the Miami River. For the first time, these models are rigorously evaluated under extreme precipitation events (e.g., tropical storms), demonstrating superior accuracy and robustness over HEC-RAS while accelerating inference by more than 500×. The work establishes a transferable methodological framework and engineering implementation for high-resolution, long-horizon, and low-latency hydrological forecasting in data-scarce, dynamically complex river systems.
📝 Abstract
Simulating and predicting the water level/stage in river systems is essential for flood warnings, hydraulic operations, and flood mitigations. Physics-based detailed hydrological and hydraulic computational tools, such as HEC-RAS, MIKE, and SWMM, can be used to simulate a complete watershed and compute the water stage at any point in the river system. However, these physics-based models are computationally intensive, especially for large watersheds and for longer simulations, since they use detailed grid representations of terrain elevation maps of the entire watershed and solve complex partial differential equations (PDEs) for each grid cell. To overcome this problem, we train several deep learning (DL) models for use as surrogate models to rapidly predict the water stage. A portion of the Miami River in South Florida was chosen as a case study for this paper. Extensive experiments show that the performance of various DL models (MLP, RNN, CNN, LSTM, and RCNN) is significantly better than that of the physics-based model, HEC-RAS, even during extreme precipitation conditions (i.e., tropical storms), and with speedups exceeding 500x. To predict the water stages more accurately, our DL models use both measured variables of the river system from the recent past and covariates for which predictions are typically available for the near future.
