Effect of Model Setup Complexity on Flood Modeling in Low‐Gradient Basins

The flat topography, large natural storage features, backwater effects, and boundary conditions all play different roles in the flood response of rivers in low‐gradient environments. The combined effects of these factors result in frequent episodes of reverse flows, slow recession of flood waters, and complex flow interactions. This study investigates the value of varying degrees of model complexity and setup features on the model ability to reproduce some of the unique flooding characteristics in low‐gradient basins. The study focuses on (1) effect of streamflow routing techniques; (2) effect of incorporating large natural storage areas; (3) effect of model dimensionality; and (4) effect of downstream boundary conditions. The study assessed six different model setups for the Vermilion River in south Louisiana, during a series of flood‐inducing storm events in May–June 2014. A successful simulation of the repetitive reverse flows in the river was only possible after incorporating the large swamp areas within the basin. The slow recession of the flood peaks was accurately reproduced with the use of a two‐dimensional representation in characterizing the swamp areas. The results of this study have implications for understanding flood dynamics in low‐gradient basins, and for guiding the development of reliable flood models that take advantage of available technologies and information without adding unwarranted complexities that require extensive, yet typically unavailable calibration data. Research Impact Statement: Incorporating large natural storage areas into flood models of low‐gradient rivers is critical to capture complex flow patterns, including reverse flows and prolonged recession of flood water.