Impact of river water levels on the simulation of stream–aquifer exchanges over the Upper Rhine alluvial aquifer (France/Germany)

Abstract

This study aims to assess the sensitivity of river level estimations to the stream-aquifer exchanges within a hydrogeological model of the Upper Rhine alluvial aquifer (France/Germany), characterized as a large shallow aquifer with numerous hydropower dams. Two specific points are addressed: errors associated with digital elevation models (DEMs) and errors associated with the estimation of river level. The fine-resolution raw Shuttle Radar Topographic Mission dataset is used to assess the impact of the DEM uncertainties. Specific corrections are used to overcome these uncertainties: a simple moving average is applied to the topography along the rivers and additional data are used along the Rhine River to account for the numerous dams. Then, the impact of the river-level temporal variations is assessed through two different methods based on observed rating curves and on the Manning formula. Results are evaluated against observation data from 37 river-level points located over the aquifer, 190 piezometers, and a spatial database of wetlands. DEM uncertainties affect the spatial variability of the stream-aquifer exchanges by inducing strong noise and unrealistic peaks. The corrected DEM reduces the biases between observations and simulations by 22 and 51% for the river levels and the river discharges, respectively. It also improves the agreement between simulated groundwater overflows and observed wetlands. Introducing river-level time variability increases the stream-aquifer exchange range and reduces the piezometric head variability. These results confirm the need to better assess river levels in regional hydrogeological modeling, especially for applications in which stream-aquifer exchanges are important.

The AquiFR hydrometeorological modelling platform as a tool for improving groundwater resource monitoring over France: evaluation over a 60-year period

“…It would also demand big resources of computational power. An attempt was made by Vergnes et al (2012) to simulate a single integrated model groundwater over the French metropolitan territory. Even though the results obtained were good enough to be used for large-scale climate applications, it was not fitted for operational water management: only one layer was defined at the coarse resolution of about 10 km; no pumping was defined; no calibration was achieved; and very simplified parameterizations were used.…”

“…However, the representation of these aquifer systems is rather coarse and is mostly used to have a realistic representation of the river base flow (Rousset et al, 2004) rather than provide consistent information on groundwater resources. developed a hydrogeological model dedicated to climate modelling that was first applied over France and on a global scale (Vergnes and Decharme, 2012). However, only a single layer at the resolution of approximately 10 km over France was considered.…”

Influence of multidecadal variability on high and low flows: the case of the Seine basin

“…Unlike MODCOU, the EauDyssée hydrogeological model implemented in Aqui-FR includes the simulation of river water levels, allowing for a better estimation of water exchanges between groundwater and rivers (Saleh et al, 2011). The exchanges between groundwater and rivers are computed daily based on the gradient head between the aquifer and the river, and a transfer coefficient that accounts for the river bed characteristic (see Vergnes and Habets, 2018, for details. The water transfers in the unsaturated zone are improved (Philippe et al, 2011), which may be important considering that the unsaturated zone can reach 60 m in the Seine basin, implying a delay in the groundwater recharge.…”

The AquiFR hydrometeorological modelling platform as a tool for improving groundwater resource monitoring over France: evaluation over a 60 year period

“…Groundwater horizontal flows are two-dimensionals while vertical flows through aquitards are represented by leakage.Therefore, unconfined and confined aquifers can be represented. SAM was successfully used to predict groundwater and surface water flows in different basins 25 of various scales and hydrogeological contexts: the Seine basin (Viennot, 2009), the Somme basin , the Loire basin (Monteil et al, 2010) or the Rhine basin (Thierion et al, 2012;Vergnes and Habets, 2018).…”

“…The results were difficult to analyse due to the differences in the way the surface water balance was calculated (either lumped-parameters model or soil-vegetation-20 atmosphere scheme), in the initialization methods, and in the way the evolutions were estimated. Moreover, in the meantime, several regional groundwater models were developed independently by research institutions in close relationship with the stakeholders for regional water management purposes or climate impact studies (Amraoui et al, 2014;Croiset et al, 2013;Douez, 2015;Habets et al, 2010;Monteil et al, 2010;Vergnes and Habets, 2018).…”