Estimation of flow regime for a spatially varied Himalayan watershed using improved multi-site calibration of the Soil and Water Assessment Tool (SWAT) model

Due to the influential role in global climate, the hydrologic modeling of the watersheds in Himalayan mountain range is critically important for the socioeconomy and livelihood of surrounding regions. As these watersheds usually have snow-driven hydrology and abrupt changes in orography, the challenges in hydrologic model are acknowledged in scientific community. In this study, we addressed this challenge by implementing an improved multivariable and multi-site approach to calibration and validation of the Soil Water Assessment Tool (SWAT) model for determining its ability to mimic flow regime of the watershed. In the improved multi-site approach, the model was successfully calibrated using 1980–1985 streamflow data and validated using 1990–1995 data for the daily time steps. Combination of different performance metrics indicates that the improved method increased the efficiency of daily prediction of Karnali River discharge. We utilized 30 years of streamflow data from four discharge stations to explore the changes in flow pattern at the decadal scales. Groundwater flow decreased in monsoon season compared to other season where the changes in flow regimes were insignificant within the decadal scale. The proposed calibration method can be used to any other large mountainous watershed to improve the estimation of the hydrologic processes.