Improving the realism of distributed hydrological models in mountainous catchments using remotely sensed observations

Over the past decades, remote sensing products have contributed with additional information on various components of the water cycle, especially in sparsely monitored areas. Although the inclusion of spatial patterns derived from satellite products can improve the performance of distributed hydrological models, simulating streamflow at interior points remains a challenge. In this study, we characterize the added value of incorporating remotely sensed soil moisture, fractional snow covered area, evapotranspiration and land surface temperature in the calibration of a distributed hydrological model. To this end, we configure the variable infiltration capacity (VIC) model at a 5-km horizontal resolution in two catchments located in Central and Southern Chile, and conduct calibration experiments with only streamflow data, and combining streamflow with remotely sensed spatial patterns. Specifically, we examine: (i) the effects at interior “ungauged” points, (ii) the benefits of adding gauging points in the calibration process, and (iii) the benefits of including additional variables. Previous results show that including spatial patterns in the calibrations allows a better representation of interior ungauged points, similar to including more streamflow gauges at interior locations.