| description abstract | Flood impacts threaten the socioeconomic conditions of peoples’ lives in the Mekong River Basin. In this study, the rainfall-runoff-inundation (RRI) model, capable of simulating rainfall runoff and flood inundation simultaneously, was used to enhance understanding of flooding characteristics in this region by using grid-satellite-based rainfall. Input data for the simulation included HydroSHEDS topographic data, APHRODITE precipitation data, MODIS land use data, and river cross sections. Moreover, the shuffled complex evolution developed at The University of Arizona (SCE-UA) global optimization method was integrated with the RRI model to calibrate sensitive parameters. In this study, the flood event in 2 has been selected in the Mekong River Basin. The simulation results were compared with observed discharges at monitoring stations along the river and an inundation map from Landsat 7 satellite imagery and the Mekong River Commission (MRC) data. The results indicated good agreement between the observed and simulated discharges, for example, with Nash-Sutcliffe efficiency (NSE)=.86 at the Stung Treng Station. The model predicted inundation extent with a success rate (SR)=67.5% and modified success rate (MSR)=74.53% compared with satellite Landsat 7, and SR=68.27% and MSR=75.11% compared with MRC data. Therefore, the RRI model was successfully used to simulate a large-scale inundation flood event in 2 using a grid precipitation data set in the Mekong River Basin. However, the underestimation might be due to the uncertainties of input data, river geometry, the large scale of the basin, coarse resolution of topographic data, and error in remote sensing image in detecting the flood extent. | |
