The value of distributed snow cover and soil moisture data for multi-objective calibration of a conceptual hydrologic model

Abstract

The latest advances and availability of satellite observations have great potential for improving hydrological model simulations of the water cycle. The recent study by Tong et al. (2021) showed that satellite observations of snow cover and soil moisture could improve river runoff simulations of conceptual hydrologic models with lumped model parameters. Still, the value and potential of spatial patterns of satellite observations for hydrologic model parametrization need to be better understood. This study aims to evaluate and compare different multiple-objective calibration strategies that use model inputs and satellite observations for the model calibration in lumped, spatially distributed and stepwise ways. We aim to test the potential of daily MODIS (Moderate Resolution Imaging Spectroradiometer) snow cover and ASCAT (Advanced Scatterometer) soil water index images observed over 204 Austrian catchments in 2000-2014. Results show that stepwise calibration strategies that first calibrate the snow model parameters to satellite snow cover data followed by calibrating the remaining model parameters outperform (particularly in lowlands) the classical calibration strategies estimating model parameters in one single calibration step. The use of distributed snow cover and soil moisture patterns in model calibration improves the snow and soil moisture simulation performance of the model. The use of MODIS snow cover data has a more significant contribution to the overall improvement in model performance than ASCAT soil moisture data.