HGS RESEARCH HIGHLIGHT – Fully integrated and physically-based approach for simulating water flows in a large-scale, heavily agricultural and low-instrumented watershed
Tagnon, B. O., Oularé, S., Kouamé, K. I., Kouassi, K. A., Kouadio, Z. A., & Savané, I. (2020). Fully integrated and physically-based approach for simulating water flows in a large-scale, heavily-agricultural and low-instrumented watershed. In Journal of Hydrology (Vol. 586, p. 124781). Elsevier BV. https://doi.org/10.1016/j.jhydrol.2020.124781
“The BRW’s physically-based, fully integrated surface–subsurface [HydroGeoSphere] model practically required minimal calibration efforts and the results obtained are quite conclusive.”
With climate change impacting water sources, water resource managers need mitigation and adaptation strategies. This study by researchers at the University of N. Abrogoua, the University of F. Houphouet-Boigny, and the University of Lorougnon G. Daloa looks at the limitations in modelling a large-scale watershed with very little data. Fully integrated modelling has seen an increase in availability and increased demand, since the computational resources required for this type of advanced modelling is becoming more and more available. Most integrated modelling has been at a small scale and in well-instrumented watersheds; however, this paper reviews the HydroGeoSphere modelling approach for the Boubo River Watershed, which has an area of 4,957 sqkm. The struggle with modelling this watershed is the lack of data. There are only a few stations set up to collect hydrometric data within the region, making it difficult to produce the statistical relationships which have traditionally been used to predict future trends. However, the lack of data is not a barrier for entirely physics-based simulation platforms that minimize the use of empirical relationships, which is why HydroGeoSphere is such a powerful tool in poorly instrumented watersheds.
HydroGeoSphere was chosen to model this watershed, which was set up with only three boundary conditions: precipitation, potential evapotranspiration, and surface water outflow (I.e. critical depth). The model calibrated in two stages. First a quasi-steady state was achieved with long-term average climatic forcing data and potential evapotranspiration rates were adjusted to match observed water table elevations. The model was then run for two decades under transient monthly climatic forcing conditions, and during this phase surface flow parameters (e.g. Manning’s roughness) were adjusted to match surface flow rates at the Bobokon gauging station. There are many factors affecting the hydrology of this watershed including forest cover loss, a large agriculture sector, flooding, and pollution. Additionally, during the dry season, there is limited drinking water availability. One of the desired outcomes of this study is to have a better understanding of the water cycle in the area to better manage it in the future.
The results of the model were accurate with one exception. The surface water depths were consistently overestimated, and surface flow rates were underestimated as a result of the large model scale, poor discretization/low model resolution, and poor representation of surface water features. Nevertheless, the conceptual model made the simulation of the water cycle of the Boubo River Watershed much simpler and has the potential to help manage water resources in the future.
This study is an excellent example of how a physics-based approach to simulating integrated hydrology allows researchers to overcome the limitations of data scarcity. Allowing water to flow naturally (or as ‘naturally’ as possible for a digital environment) also simplifies the calibration process, as a well conceptualized watershed scale model should be able to accurately represent the integrated hydrology of the watershed inherently.
CLICK HERE TO READ THE ARTICLE.
Abstract:
Recent advances in fully integrated and physically-based hydrologic models, as well as computational resources, have increased interest for these models, which seem most realistic for assessing the impacts of anthropogenic activities and climate change on water resources. However, their applications, for most of the cases, are carried out on relatively small-scale (<1,000 km2) and well-instrumented watersheds. The present study uses the fully integrated and physically-based approach to evaluate surface and subsurface hydrodynamics, in the Boubo river watershed (BRW), located in the southern part of Côte d'Ivoire. The BRW is a large-scale area (4,957 km2), lacking data and where local and transversal issues impacts' on the hydrosystem are simultaneous and significant. The conceptual model uses an equivalent porous medium for the 3D subsurface domain and the integrated modelling approach necessitated only three boundary conditions: precipitation, potential evapotranspiration and surface water outflow. The simulations were conducted using two main calibration phases: quasi-steady-state and transient conditions. Minor calibration effort was required and overall, the results obtained seem quite plausible and would reflect the field reality. However, abnormally high surface water depths (>15 mm), observed at the land surface and the underestimation of simulated surface water flows rates, compared to measured ones, seem rather problematic. These problematic results could be explained by a defect of drainage in the model. This could be due to the weak representation in the model of particularly complex local surface features and low mesh resolution, in relation to the large-scale of the study area. The transient phase simulation allowed observation of model response with respect to changes in forcing data. This is encouraging in the context of assessing the possible impacts of climate change on water resources.