Stepwise calibration of a lumped hydrological model for Xun River basin, China, considering the heterogeneity of soil moisture

Study region: Xun River in the Han River basin of China. Study focus: This study aims to examine the potential of soil moisture products in the temporal calibration of lumped hydrological models, as well as soil moisture data for improving hydrological simulations by considering heterogeneity. The spatial calibration of semi- or fully distributed models has seldom been carried out for conceptual lumped models since it is more common for gridded products to be used. New hydrological insights: The stepwise calibration scheme considering spatial heterogeneity had the highest average soil moisture correlation coefficient (r = 0.54), followed by the simultaneous calibration scheme considering spatial heterogeneity (r = 0.49), and lastly the simultaneous calibration scheme without considering spatial heterogeneity (r = 0.38). Additionally, the stepwise calibration scheme exhibited lower streamflow accuracy compared with that of the simultaneous calibration scheme. This apparent inconsistency arises because the stepwise approach, while achieving the highest average soil moisture correlation coefficient through spatial heterogeneity consideration, demonstrates compromised performance in streamflow simulation due to the neglected scale-dependent factors. These results reveal that the spatial heterogeneity of soil moisture is of great significance to the calibration results, and it is useful to consider it in the lumped hydrological model by using local soil moisture instead of average soil moisture.