In coarse resolution hydrological modeling we face the problem of subgrid variability, the effects of which are difficult to express and are often hidden in the parameterization and calibration. We present a numerical experiment with the physically based hydrological model ParFlow‐CLM with which we quantify the effect of subgrid heterogeneities in headwater catchments within the cell size typically used for regional hydrological applications. We simulate homogeneous domains and domains with subgrid heterogeneities in topography or soil thickness for two climates and soil types. The presence of side slope is the main error source, leading to large underestimation of runoff, and marginally also of evapotranspiration. The spatial distribution of soil saturation in the presence of subgrid variability in topography also leads to underestimation of landslide risk. Soil thickness is the second influential subgrid property, affecting soil moisture distribution and surface runoff formation. Results are consistent for the climates and the soil types considered. The topographic wetness index approach is tested as a way to downscale soil moisture simulations within the domain. Although this method is successful in reproducing some spatial variability and patterns, it fails when the coarse grid mean soil saturation is inaccurate or subgrid topography does not represent subsurface flow paths accurately. We conclude that ignoring subgrid variability in topography and soil thickness in coarse‐scale hydrological models may lead locally to underestimation of runoff and slope instability. Users of such models should be aware of these biases and consider ways to include subgrid effects in coarse‐scale hydrological predictions.
|Publication Subtype||Journal Article|
|Title||Numerical analysis of the effect of subgrid variability in a physically based hydrological model on runoff, soil moisture, and slope stability|
|Series title||Water Resources Research|
|Publisher||American Geophysical Union|
|Contributing office(s)||Geologic Hazards Science Center|
|Description||e2020WR027326, 16 p.|
|Google Analytic Metrics||Metrics page|