Calibration of a physically based distributed hydrological model: a case study of the Alaknanda River basin

The unique topographical and climatic conditions of the Himalayan region along with fewer hydrometeorological stations, add complexity to the hydrological modeling of this region. This study addresses the constraints of calibrating a hydrological model in a data-scarce watershed by substituting reanalysis surface runoff data (RSRD) for discharge data. In this study, we calibrated a distributed hydrological model, WATFLOOD, using ERA-5 surface runoff data. We assess water balance components in nine land use classifications across different locations in the study area. We also evaluate six separate water balance components over time using simulated data from different land use classes. Our results indicate that the WATFLOOD model can successfully replicate the water balance in the Himalayan region (CC = 0.8 and NSE = 0.75). Since the observation data is not easily accessible, RSRD can be utilized to calibrate the hydrological model. We validated our results with observed data available at one station for a short period. Our results reveal the drawbacks of calibrating a hydrological model with RSRD. The annual fluctuation of water balance components above and below ground in each land cover class varies in response to wet and dry years. The less variation of Total Upper Zone Storage (TUZS) and Lower Zone Storage (LZS) in dry and wet years shows that these water balance components are independent of the rainfall in the region. The proposed hydrological model has the potential to help manage water resources in highland locations with limited data.

Keywords: WATFLOOD model, Reanalysis Surface Runoff, Alaknanda River basin, Western Himalayan region