UNDERSTANDING THE ROLE OF LAND USE-LAND COVER IN RUNOFF GENERATION IN THE UPPER KOSI RIVER CATCHMENT USING SWAT MODEL

Abstract

The land use-land covers (LULCs) alterations in the Himalayan regions are leading to different environmental issues such as change in the hydrological behaviour of the small catchments. The LULC changes in the regions are caused by different factors such as continuous deforestation, changes in climatic patterns, and anthropogenic activities such as the conversion of the forest area into agricultural land and agricultural area into pastureland. This study analysed the role of different LULCs on the water yield in different sub-basins of a Himalayan River, Upper Kosi, using a semi-distributed hydrological model, namely Soil and Water Assessment Tool (SWAT). The area of the Upper Kosi River catchment is 1469 km2 and the catchment was divided into 21 sub-basins and 75 Hydrological Response Units (HRUs) for modelling purposes. The SWAT model was calibrated and validated using SWAT-Calibration and Uncertainty Program (CUP) based Sequential Uncertainty Fitting (SUFI-2) algorithm. Out of the limited four years of data, the first two years (2014-15) were used for model calibration, and the next two years (2016-17) were used for model validation. A parameter sensitivity test was performed, and the most sensitive parameters were identified as GW_DELAY.gw, CH_N1.sub, GWQMN.gw, and RCHRG_DP.gw. The performance of the model was evaluated using standard metrics, namely, Nash Sutcliffe Efficiency (NSE), Coefficient of Determination (R2), and Percent-bias (PBIAS). For the calibration period, the performance metrics were 0.45 (NSE), 0.37 (R2), and 2.9 (PBIAS), and during the validation period, performance metrics were 0.60 (NSE), 0.75 (R2), and 2.07 (PBIAS). Despite the limited training model, the model performance was reasonably good during the calibration and validation periods, as indicated by performance metrics. The sub-basin scale analysis has shown that the upper Kosi River catchment receives a higher water yield contribution from the downstream sub-basins compared to the midstream and upstream sub-basins. Further, the water yield was analysed for the relative contribution of individual land covers to the watershed. It was found that urban areas have the highest contribution to surface runoff, followed by agricultural and forest lands, respectively. The contribution of mixed and evergreen forests to surface runoff was found to be approximately the same.