Predicting Pesticide Attenuation in a Fractured Aquifer Using Lumped‐Parameter Models

Neighboring springs draining fractured-rock aquifers can display large differences in water quality and flow regime, depending on local variations of the connectivity and the aperture size distribution of the fracture network. Consequently, because homogeneous equivalent parameters cannot be assumed a priori for the entire regional aquifer, the vulnerability to pollution of such springs has to be studied on a case by case basis. In this paper, a simple lumped-parameter model usually applied to estimate the mean transit time of water (or tracer) is presented. The original exponential piston-flow model was modified to take land-use distribution into account and applied to predict the evolution of atrazine concentration in a series of springs draining a fractured sandstone aquifer in Luxembourg, where despite a nationwide ban in 2005, atrazine concentrations still had not begun to decrease in 2009. This persistence could be explained by exponentially distributed residence times in the aquifer, demonstrating that in some real world cases, models based on the groundwater residence time distribution can be a powerful tool for trend reversal assessments as recommended for instance by current European Union guidelines.