Large-scale identification of riverbank filtration wells using an isotopic and geochemical approach

This study proposes a simple isotopic framework for the identification of riverbank filtration wells, using time-series of water’s stable isotopes and electrical conductivity at the watershed scale. Riverbank-filtration (RBF) is a widely used managed aquifer recharge method where the infiltration of surface water is induced by pumping groundwater at proximity of a river or lake, and where the quality of the raw water is controlled by the interaction between ground and surface water bodies. As inventories of RBF sites are rare, they limit the development of specific source protection strategies. In the framework of a project funded by the Ministry of the Environment and the Fight against Climate Change of Québec (MELCC, Canada), 40 municipal wells located at less than 500 meters from a surface water body were sampled for 18 months on a monthly to weekly basis. The underlying hypothesis was that a significant contribution of infiltrated surface water to a pumping well would propagate the temporal variations of the tracers observed in surface water. Results highlighted that 25% of the wells pumped a significant contribution of infiltrated surface water: 15% established a continuous connection during the whole hydrological year, while the other 10% revealed a seasonal connection linked to spring floods. All those wells were located less than 120 meters from the surface water and drilled less than 40 meters deep in a granular aquifer. As the number and strength of spring floods may increase in the future, the seasonally connected wells are particularly vulnerable to global change and might prefigure increasing climate forcing on drinking water supply. The remaining 75% were either evidenced as groundwater only (50%) or lacked continuity in the data acquisition (25%). A simple abacus based on the standard deviation of the tracer distribution was then proposed to facilitate the interpretation and make the methods accessible to all. Based on an affordable and easy to perform sampling protocol for water managers, this framework helps in the assessment of specific risks associated to mixed sources of water, and in anticipating variations in quality of the abstracted drinking water.