AUTHORS: OLIVER S. SCHILLING, CHRISTOPH GERBER, DANIEL J. PARTINGTON, ROLAND PURTSCHERT, MATTHIAS S. BRENNWALD, ROLF KIPFER, DANIEL HUNKELER, PHILIP BRUNNER
To provide a sound understanding of the sources, pathways, and residence times of groundwater water in alluvial river-aquifer systems, a combined multi-tracer and modeling experiment was carried out in an important alluvial drinking water wellfield in Switzerland. 222Rn, 3H/3He, atmospheric noble gases, and for the first time, the novel 37Ar -method were used to quantify residence times and mixing ratios of water from different sources. With a half-life of 35.1 days, 37Ar allowed to successfully close a critical observational time gap between 222Rn and 3H/3He for residence times of weeks to months. Covering the entire range of residence times of groundwater in alluvial systems revealed that, to quantify the fractions of water from different sources in such systems, atmospheric noble gases and helium isotopes are tracers suited for end-member mixing analysis. An updated illustration of available tracer methods, now including the new 37Ar method, is provided below:
The tracer-based mixing ratios were subsequently compared to mixing ratios simulated with the fully-integrated, physically-based flow model HydroGeoSphere. In order to simulate the propagation of the different sources of water in HydroGeoSphere, the Hydraulic Mixing Cell method (HMC) of Partington et al. (2013) was extended for this study to incorporate not only streamflow separation, but also tracking in the subsurface. Simulations of three different scenarios (poorly permeable, permeable and strongly permeable riverbed) are illustrated below. Both the exchange flux pattern on the surface and the fraction of surface water in the subsurface are illustrated.
The comparison between tracer-based and simulated mixing ratios revealed that models, which are only calibrated against hydraulic heads, cannot reliably reproduce mixing ratios or residence times of alluvial river-aquifer systems. However, the tracer-based mixing ratios allowed the identification of an appropriate flow model parametrization. Consequently, for alluvial systems, we recommend the combination of multi-tracer studies that cover all relevant residence times with fully-coupled, physically-based flow modelling using HGS and HMC-based flow tracking to better characterize the complex interactions of river-aquifer systems.