Jorge Arevalo, Xubin Zeng, and Pieter Hazenberg
Department of Hydrology and Atmospheric Sciences
The University of Arizona
Earth System Models (ESMs) are evolving to resolutions of a few kilometers. In this context, one challenging task is to update the parameterizations used to represent the subgrid processes. In particular, Land Surface Models (LSM) included in the ESMs, traditionally have been represented as 1D-columns accounting for fluxes inside the column just in the vertical direction. The subgrid heterogeneity of the soil properties greatly influences the boundary fluxes, hence Hazenberg et al. (2015) developed a computationally efficient hybrid 3D hillslope hydrological model (h3D) for rainfall-runoff processes at hillslope scale, coupling a 1D-column with a pseudo-2D lateral flow (for overland and the saturated zone). The h3D model brings more realistic results for subsurface and overland water flows than the LSMs currently used in the ESMs. This study presents a sensitivity analysis (for an idealized study case) of the h3D model to the initial conditions and to several soil characteristics, including Hydraulic conductivity, rooting depth, soil anisotropy, hillslope shape and slope among others. Quantification of the sensitivity of the h3D model will allows to better conducts the efforts in the implementation of this model into a full LSM, the development of its initialization datasets, and the adjustment of its parameters.