Characterising the benefit of remote sensing soil moisture for calibrating a spatially distributed hydrological model

Remotely sensed (RS) data have additive value in calibrating a hydrological model. In this regard, RS soil moisture data were mostly assimilated into conceptual hydrological model by changing the entity of the product. In this study, raw remotely sensed (RS) surface soil moisture (SM) is used as the only calibration variable in the Soil water assessment tool (SWAT) model. This means the SM values were not transformed to another variable (e.g. soil water index, root zone soil moisture index). This study uses the CCI SM 2.0 merged soil moisture product from European Space Agency (ESA). For a nested catchment, calibration solely based on SM, improved streamflow predictions for some of the gauging stations compared to the un-calibrated model. However, this was only in parts of the catchments where the SM signal directly influenced the flow distribution. The seasonal breakdown indicates that the SM signal is more useful for calibration in wetter seasons and in areas with higher variations in elevation. The results identified SM thresholds of the modelled catchment and improved the simulation of the general rainfall-runoff response by optimizing the spatial distribution of SM. Furthermore, catchment characteristics (e.g., landuse, elevation, soil types, precipitation) regulating the SM variation in different seasons are identified. This study provides further opportunity to improve predictive capacity of distributed hydrological modelling by better model parameterization.