HGS RESEARCH HIGHLIGHT – Upscaling Hydrological Processes for Land Surface Models with a Two-Hydrologic-Variable Model: Application to the Little Washita Watershed
Picourlat, F., Mouche, E., & Mügler, C. (2022). Upscaling Hydrological Processes for Land Surface Models with a Two‐Hydrologic‐Variable Model: Application to the Little Washita Watershed. In Water Resources Research. American Geophysical Union (AGU). https://doi.org/10.1029/2021wr030997
CLICK HERE TO READ THE ARTICLE.
Sometimes HydroGeoSphere isn’t the focus of a study, but rather provides important baseline data against which to evaluate other methodology. Actually, using HGS model results as the reference simulations in a study is a huge vote of confidence in the capabilities of HGS. This is exactly the case in this new study by researchers at the Laboratoire des Sciences du Climat et de l’Environnement at Université Paris-Saclay.
The authors have used a 3D HydroGeoSphere model of a heavily studied sub-catchment (the Little Washita Watershed, Oklahoma) as a reference point to test the validity of much simpler modelling approaches. Results of the 3D HydroGeoSphere model are compared against a simpler 2D hillslope model, also constructed using HydroGeoSphere. Finally, some simplifying assumptions are applied to identify two variables controlling water balance in the hillslope: the water table slope and location of the seepage face. These variables were developed into a very simple conceptual model to elucidate water balance information for the experimental hillslope. The authors find that the simple conceptual model can successfully capture the water balance of hillslopes of any geometry, when compared to both 2D and 3D HydroGeoSphere models.
This is an interesting study that shows HydroGeoSphere is a great option to validate the results of other, simpler models. In this case, the conceptual hillslope model could provide an improved strategy for modelling water balances which can be incorporated into Land Surface Models.
CLICK HERE TO READ THE ARTICLE.
Abstract:
Land Surface Models (LSMs) are key components of Earth System Models (ESMs), which the Intergovernmental Panel on Climate Change (IPCC) relies on in many of their studies. However, these models either neglect or oversimplify basin-scale hydrological processes that produce the land surface water balance. Bringing the dominant physical processes from the local scale up to the LSMs presents a significant challenge for improving the models. This article presents an upscaling (or bottom-up) approach to identify the basin-scale driving variables that need to be exported into LSMs. The approach is developed on the Little Washita Watershed (Ok, USA) using 20-year hydrology (1993-2013). A physically-based 3D model built with HydroGeoSphere first produces a reference simulation. An equivalent hillslope model is then able to capture both 3D simulated water balance and local water table dynamics with reasonable accuracy. Physical analysis of the water balance in the different hillslope compartments leads to the identification of two driving variables: seepage face extension and water table slope. The two variables are then implemented in a conceptual model. Results show a good capacity of this model to capture the water balance of hillslopes having different lengths and slopes. Moreover, the model is able to capture the water balance of the reference simulation with reasonable accuracy. The proposed approach thus reduces the 3D watershed model to a two-variable conceptual model that constitutes a basis for developing an improved LSM hydrology.
Key Points:
An upscaling approach reduces model dimensionality from 3D watershed to 2D equivalent hillslope, and then to a two-hydrologic-variable model
The seepage face extension and the water table slope are identified as two key variables
The approach is validated on the 20-year hydrology (1993-2013) of the Little Washita Watershed (OK, USA)