A validated fast algorithm for simulation of flooding events in plains

Abstract

Hydrological modelling in large plains is not a straightforward task owing to the crucial influence of several elements, namely, accumulation of water on local depressions, absence of an integrated drainage network, generalized meagre surface slopes, strong sensitivity of the soil water content before rainfall events, and variation of the soil infiltration capacity with numerous transient and local factors. It would then appear that distributed hydrological models (DHM) are prone to produce better results than lumped-parameter models. The trade-off in setting up a DHM, however, is how realistic its results are as compared to its simplicity. This paper presents a physically based DHM, named AQUA, which can handle large domains discretized in squared cells of typically 80 m a side. In order to address properly the flow resistance, it relies on a relaxation parameter (α) with different values for the watercourses and the terrain. Likewise, an infiltration function (I) regulates the downward movement of water. Moreover, the model is supported by an adequate description of the gentle topographical features, which was achieved by a digital elevation model (DEM) from radar interferometry. The testing phase took into account the Santa Catalina Creek Basin (158 km2, centre Buenos Aires Province in Argentina). The results were quite encouraging, as the model was able to reproduce the effect of various rainfall scenarios. Sensitivity analyses of the model parameters were consistent with the actual simulation results. AQUA was tested against the well-known HEC-1 model, with a fairly good match of their results.