Riverbank filtration within the context of river restoration and climate change
Drinking water derived by riverbank filtration is generally of high quality and is an important source of drinking water in several European countries. In the future however, riverbankfiltration systems will face two major challenges – river restoration and climate change. The goal of this Ph.D. Thesis was to deepen the understanding of physical and biogeochemical processes that occur during riverbank filtration and develop new tools in order to facilitate the assessment of potential adverse effects of river restoration and climate change on the quality of river-recharged groundwater.
River restoration measures can lead to shorter residence times between the river and the pumping well and therefore can increase the risk of drinking water contamination by bacteria or pollutants. Numerical groundwater models provide quantitative information on groundwater flow paths and residence times, but require a rigorous definition of the spatial and temporal river water level distribution. In this thesis, two new interpolation methods were developed to generate time-varying 1D and 2D river water level distributions. The methods were implemented at the partly restored Niederneunforn field site at the peri-alpine Thur River (NE-Switzerland), and were applied to a 3D groundwater flow and transport model. The results confirmed the method’s suitability for accurately simulating groundwater flow paths and residence times.
The increased occurrence of heat waves due to climate change likely favors the development of anoxic conditions in the infiltration zone, which may significantly deteriorate the quality of river-recharged groundwater. Results from field sampling campaigns and column experiments suggest that particulate organic matter (POM) degradation mainly accounted for the variability of dissolved oxygen (DO) consumption during riverbank filtration. Furthermore, DO consumption was found to positively correlate with temperature and discharge. The latter was attributed to an enhanced trapping of POM within the riverbed during high-discharge conditions. To quantify the temperature and discharge dependence of DO consumption during riverbank filtration, a new semi-analytical model was developed and successfully applied to the Niederneunforn field site. The modeling approach can be transferred to other riverbankfiltration systems to efficiently estimate groundwater DO concentrations under various climatic and hydrologic conditions and, hence, to assess the risk of arising anoxic conditions.