REAL TIME FLOOD FORECASTING OF BAGMATI RIVER BASIN IN NEPAL

Abstract

Bagmati River Basin of Nepal is subjected to frequent flooding even washing away of the water level gauging stations. Estimation of surface runoff in the basin due to precipitation and quantifying discharge at desired location is important aspect of flood forecasting and management plan. The study area of Bagmati River Basin has been taken up to Bhorleni gauging site having the catchment area of 1694 sq.km. In this study, HEC-HMS hydrological model of version 4.2.1 is used to simulate rainfall-runoff process and the basin is taken as the single basin. The Arc Map-Arc-GIS version 10.4.1 and HEC-GeoHMS version 4.2 are used to process different types of spatial data required as input for the HEC-HMS model applications. All temporal data required for running HEC-HMS model are processed and utilized as input in the prescribed format required for the model. The model is applied with recession method for base flow estimation and initial and constant rate method for loss model. Two different transform methods viz. Clark and Snyder transform model are used in the study. Model is calibrated and validated using the available hourly rainfall and runoff data of three storm events observed in flood season of year 2018. The model is calibrated with two storm events and validated with one storm event by comparing observed and simulated discharges at the Bhorleni gauging site. Various performance criteria viz. peak value, total volume generated and Nash-Sutcliffe efficiency are used for evaluating the models. NSE values obtained from the Clark and Snyder transform models are 0.791 and 0.755 respectively. Comparison of result of Clark with Snyder shows that Clark method is more satisfactory in terms of higher Nash-Sutcliffe Efficiency. The calibrated and validated Clark model yielded 16.59% of error in peak, 25.00% of error in time to peak and 1.68% error in discharge volume. The results show the suitability of HEC-HMS in simulating rainfall-runoff process of Bagmati River Basin. Further, the validated storm event is used to generate unit hydrograph for Bagmati river basin for flood forecasting. The representative unit hydrograph for the basin derived from the Clark’s model is applied for formulation of the real time flood forecast for the event of 11 July 2018-14 July 2018, considering the blocks of hourly excess rainfall available up to the period of forecast. The peak discharge of the forecasted unit hydrograph for different block periods have been derived. While comparing, it is observed that the forecasted flood hydrographs slightly fit the observed flood hydrograph.