抄録
Warning and preparedness are two key measures for flood disaster reduction. Flood warning with adequate lead-time greatly help in reducing loss of life and economic damage. Preparedness calls for estimating flood risk and taking action to mitigate flood damage through both structural and non-structural measures. In the present study we explore a comprehensive warning and risk assessment scheme in a two stage approach considering data availability and operational ease. In the first stage, river flow is forecasted with sufficient lead-time to warn of flows that could exceed the safe river capacity. The second stage takes the forecasted river flows at the city as input to a limited area model that covers the city and simulates the inundation extent incorporating embankment overtopping.
A distributed hydrological model, which simulates all the hydrological processes, is used in the system. In stage 1, where river flow is forecasted, dynamic wave model is used to solve for river network but inundation simulation is switched off to improve computational efficiency. In stage 2, diffusive wave implicit solution scheme for both inundation modeling and river network solution is used. The modeling domains for stage 1 and stage 2 are setup such that flooding from embankment overtopping would not occur upstream of the common boundary covered in stage 1. For river flow forecasting, the river network derived from the GTOPO30 global data set whereas for stage 2 model is setup using a high resolution local DEM. The results show that for the application in Yom River basin, lead times about 14 hours can be achieved with the dynamic equation for river discharge forecasts. The discharge hydrographs are successfully incorporated as inputs in inundation modeling and employing 50m or 100m horizontal resolution grids sizes help differentiate inundations resulting from different river inflows.