Calibration of a Distributed Hydrological Model (VIC-3L) Based on Global Water Resources Reanalysis Datasets

Abstract

The lack of observed ow datasets for calibration of rainfall-runoff models imposes fundamental problems for the applicability of them in most catchments, especially in poorly gauged or ungauged river basins. By developing satellite technologies over the past decades, several water balance components globally databases by remote sensing and data assimilation techniques have been developed and provided to researchers. This study performs various products of surface soil moisture (ASCAT, HTESSEL, HBV, and LISFLOOD), evapotranspiration (W3RA, GLEAM, HBV and Ensemble), and surface runoff (SURFEX-TRIP and Ensemble) to the calibration of the macro-scale hydrological model (VIC-3L) over the Se dRood basin (SRB) in Iran. Results show that using the distributional approach and applying LISFLOOD lead to an acceptable performance in the ow hydrograph simulation than others with Nash-Sutcliffe (NS) values of 0.68 and Root Mean Square Error (RMSE) values of 64.84 m 3 /s. Among the studied evapotranspiration sources, the GLEAM daily and monthly time scales with NS values of 0.59 and 0.51, respectively, have better performance than other products in the simulation of ow hydrograph.Based on the results, the application of evapotranspiration has been better than soil moisture and surface runoff in estimating low-ows. Finally, results indicated that the multi-objective approach (GLEAM and ASCAT) had better performance than single-objective calibration to estimate runoff volume with Relative Error = 9.1 %. This study encourages applying the re-analysis database to hydrological models' calibration process to estimate discharge in data-limited areas.