On the value of benchmarking a fully coupled surface-subsurface model with spatially distributed sap flow measurements

Integrated hydrological models (IHM) are often used for a better understanding of the hydrological fluxes at the surface and in the subsurface. IHMs can also be coupled to land surface models to study the interactions between vegetation and hydrological processes. At our study site, the Weierbach catchment (43 ha), Luxembourg, sap flux observations showed high spatial variability in observed transpiration due to many factors such as DBH, landscape characteristics and position, and tree species. However, this spatial variability is often not captured by land surface models at small scale and transpiration fluxes are usually treated as an integrated flux. In our study, we employ the coupled integrated surface-subsurface hydrological model Parflow coupled with the land surface model (CLM) to simulate water and energy fluxes in the forested Weierbach catchment in Luxembourg. Our objectives are twofold. First, we evaluate the coupled physically-based model and land surface model with discharge, groundwater level and soil moisture, and spatiotemporal sap flow data. In addition to that, we are exploring whether simulated and observed transpiration for three different hillslope positions (plateau, midslope, hillfoot) are driven by atmospheric demand or water availability in the subsurface. Our main result was that model has captured discharge, groundwater fluxes and the average transpiration well. However, the modelled transpiration showed a much smaller spatial variability compared to the spatial variability derived from sapflow observation. For the three different hillslope positions, we found that the fluxes were mainly driven by atmospheric demand and the model captured this dominance well. Our results demonstrate that there is a limitation of the model in reproducing the spatial variability of transpiration in the heterogeneous forest and future modelling work at small scale needs to better parameterize the spatial characteristics of vegetation.