Improving the representation of river–groundwater interactions in land surface modeling at the regional scale: Observational evidence and parameterization applied in the Community Land Model
Groundwater is an important component of the hydrological cycle, included in many land surface models to provide a lower boundary condition for soil moisture, which in turn plays a key role in the land–vegetation–atmosphere interactions and the ecosystem dynamics. In regional-scale climate applications land surface models (LSMs) are commonly coupled to atmospheric models to close the surface energy, mass and carbon balance. LSMs in these applications are used to resolve the momentum, heat, water and carbon vertical fluxes, accounting for the effect of vegetation, soil type and other surface parameters, while lack of adequate resolution prevents using them to resolve horizontal sub-grid processes. Specifically, LSMs resolve the large-scale runoff production associated with infiltration excess and sub-grid groundwater convergence, but they neglect the effect from loosing streams to groundwater. Through the analysis of observed data of soil moisture obtained from the Oklahoma Mesoscale Network stations and land surface temperature derived from MODIS we provide evidence that the regional scale soil moisture and surface temperature patterns are affected by the rivers. This is demonstrated on the basis of simulations from a land surface model (i.e., Community Land Model – CLM, version 3.5). We show that the model cannot reproduce the features of the observed soil moisture and temperature spatial patterns that are related to the underlying mechanism of reinfiltration of river water to groundwater. Therefore, we implement a simple parameterization of this process in CLM showing the ability to reproduce the soil moisture and surface temperature spatial variabilities that relate to the river distribution at regional scale. The CLM with this new parameterization is used to evaluate impacts of the improved representation of river–groundwater interactions on the simulated water cycle parameters and the surface energy budget at the regional scale. âº Observed soil moisture analysis suggest that rivers affect large scale groundwater distribution. âº This process can be parameterized as a modified TOPMODEL approach in CLM. âº Brightness temperature analysis of model data confirms this evidence. âº Rivers effect on the hydrological cycle cannot be neglected in regional climate models.