STATISTICAL EVALUATION OF THE VULNERABILITY OF GROUNDWATER WELLS, A CASE-STUDY OF DATA FROM THE STRASBOURG POLYGONE PUMPING FIELD

The case study presented in this paper illustrates how statistical met hods can help to understand the relationships between ground-water wel ls and pollutant sources in alluvial systems. The observation of a tim e series of chloride concentrations in a specific well in Strasbourg, France, influenced by the infiltration of chlorinated water from the R hine river, made it possible to derive a two-predictor model of the wa ter quality in the well. This model combines a linear dependence upon a time-lagged chloride concentration in the river, reflecting the stea dy-state water balance in the well, and a nonlinear dependence upon av erage lagged water output pumped from the well. This simple model expl ains 86% of the variance in well chloride concentration measured weekl y over a two year period.