FLOOD FORECASTING USING FGM MODEL IN CHINDWIN RIVER BASIN

Abstract

The present research work deals with flood forecasting in Chindwin River basin which is situated in Northern West of Myanmar under available hydro-meteorological data. Flood is one of the natural disasters which occur in Myanmar every year. Flood forecasting and issues of flood warnings are the effective ways to reduce damages. The goal of the study has been to initiate operational flood forecasting. This research applied Flussgebietsmodell (FGM) which is originally developed by the Institute of Hydrology and Water Resources Planning (IHW) of the University of Karlsruhe, Germany. FGM model is an event-based rainfall-runoff model. Model parameters are runoff coefficient, unit hydrograph parameters and routing parameters. Unit hydrograph is determined using linear cascade model. The effective precipitation is routed to the outlet through a linear transfer function that is assumed to be time invariant. Flood from each subbasin is calculated by means of effective rainfall convoluted with a unit hydrograph. Flood routing is done by Kalinin - Milyukov method. The two important parameters, when predicting a flood hydrograph, are the time to peak discharge and the magnitude of the peak discharge. It was found that the FGM model has been able to predict this information with acceptable accuracy. The model performs quite well especially for the floods where relation between rainfall and runoff is good. The numerical verification criteria used in model calibration are Nash efficiency and coefficient of variation of the residual of error. Model efficiencies obtained in calibration periods are good efficiency. In this study, it is seen that the diagnosis performs well. Therefore, the FGM model is generally considered to be suitable for flood forecasting in Myanmar catchments.