H196-0009
Is the Green-Ampt model a suitable approach for hydrodynamic rainfall-runoff simulations in ungauged catchments?
Is the Green-Ampt model a suitable approach for hydrodynamic rainfall-runoff simulations in ungauged catchments?
Wednesday, 16 December 2020
Poster
Abstract:
The Green-Ampt model is one of the most commonly used approaches to consider infiltration in rainfall-runoff models. The so-called Green-Ampt parameters and the initial water content are needed as model input. Often, these soil parameters are not directly measured in the field but determined by calibration against observed runoff data. However, when studying ungauged catchments or flash floods measured data is lacking, and the Green-Ampt parameters can only be estimated based on tabulated values from the literature. This study investigates the suitability of such tabulated Green-Ampt parameters through numerical simulations with a coupled shallow water flow and infiltration model. Several different small-scale test cases with measured runoff are investigated, and the performance of tabulated parameter values from two different sources are evaluated. Three of the test cases are laboratory experiments and one is a field experiment on a small plot in Senegal. Finally, a case study on flash floods in an ungauged region in Egypt is simulated and the plausibility of the results as well as the impact of surface crusts and bottom friction are discussed. The results show that in two of the three laboratory experiments, the infiltration rates based on average Green-Ampt parameters are underestimated, while for the field experiment in Senegal and the case study in Egypt - both with sandy soil - infiltration based on tabulated Green-Ampt parameters is strongly overestimated. In the latter case, the consideration of a thin surface crust with reduced hydraulic conductivity leads to more plausible results. Based on these two cases with sandy soil, a reduction of the hydraulic conductivity by about 90 – 100 % compared to the considered literature values or the implementation of a surface crust can be recommended for similar real-world applications in ungauged areas with sandy soil. Future research will focus on conducting plot-scale rainfall-runoff experiments with a rainfall simulator to better represent natural conditions during heavy rainfall. High-resolution digital elevation models of the plots will be obtained by using photogrammetry. This data will be used in coupled shallow water and infiltration simulations to further investigate the impact of surface crusts, bottom friction, and microtopography on runoff generation.