On the biases affecting water ages inferred from isotopic data

Abstract: Groundwater age has become a fundamental concept in groundwater hydrology, but ages originating from isotopic analyses are still identified with a lack of clarity and using models that occasionally are unrealistic. If the effect of advection and dispersion on water ages has already been extensively identified, very few studies address the reliability of using radiometric ages as derived from isotopic data to estimate aquifer properties such as average velocities. Using simple one-dimensional and two-dimensional analytical solutions for single-site and two-sites mobile-immobile systems, we compare the radiometric ages to the mean ages (or residence times) as deduced from a direct, physically-based simulation approach (using the mean age equation), and show that the competition between isotope decay rate and dispersion coefficient can generate important discrepancies between the two types of ages. A correction for the average apparent velocity originating from apparent isotopic ages is additionally provided. The particular case of the Tritium age dating method is also addressed, and a numerical example is finally given for illustrating the analysis considering a more complex and heterogeneous aquifer system. Our results suggest that age definitions based on the radioactivity of isotopes may not be representative for the mean age of the sample or for the groundwater velocity at given locations, and may not always be suitable for constraining the calibration of hydrogeological models.