Author(s): Getachew Ewonetu Mamo; Davide Motta; Martin Crapper

Linked Author(s):

Keywords: River flooding; Flood modelling; Hydraulic model; Terrain modification

Abstract: Producing a reliable hydraulic model for river flooding prediction, able to reproduce the extent of historically observed overbank flooding, is especially challenging for low slope rivers characterized by complex flow patterns and limited data availability, especially regarding main channel cross-section geometry and bankfull carrying capacity. This paper presents the comparison of different hydraulic model setups for river flooding modelling: fully one-dimensional (1D), fully two-dimensional (2D) and coupled 1D/2D (1D in the river main channel and 2D in the floodplains), in all cases including techniques for main river channel bathymetry reconstruction (based on bankfull depth reconstruction). The comparison is made for the modelling of the seasonal flooding of a 196 km reach of the Baro River in the Gambela floodplain in Ethiopia. The river flow is characterized by complex patterns of interaction with the nearby Alwero River and the topographic information available is limited. The wet season of the year 1990 is considered for the study. Our results show that the 1D approach cannot capture the flooding patterns in the quasi-flat area downstream of Gambela City, while 2D and coupled 1D/2D models can. The 2D approach is, however, time-consuming while the coupled 1D/2D produces results in good agreement with the historically observed flood extent, while ensuring model stability and containing the model simulation time, making it a good tool for prediction of flood extent, magnitude, and duration in large-scale flood mapping.

DOI: https://doi.org/10.3850/IAHR-39WC2521711920221376

Year: 2022