ABSTRACT

Numerical groundwater models, which are calibrated with the help of head data are often nonunique. A fit to head data alone is necessary but in general not sufficient. The basic problem in flow modelling consists of the fact that hydraulic conductivity is in general poorly known. Since boundary fluxes, recharge rates, and fluxes to or from rivers cannot be measured directly, they have usually to be estimated independently. Both types of information are uncertain, and therefore, the resulting flow model suffers from uncertainty. Model calibration cannot remove this uncertainty as there is usually no unique solution to the parameter estimation problem. Environmental tracers can be of help in this situation. Environmental tracers can for example indicate that there is recent recharge, they allow estimating the age of groundwater or its residence times, they can yield streamline information, ratios of fluxes including recharge rates, and effective porosity values. Indeed, examples from the literature show that various environmental tracers were successfully used in the past to check or to improve flow models, to select among several alternative scenarios or conceptual models, and to determine solute transport parameters. Expectations should, however, be modest concerning their ability to improve the characterisation of hydraulic conductivity or the accuracy of fluxes. This review suggests that environmental tracer information can primarily be used for improving ratios of parameters and constraining the conceptual model, but not necessarily for getting higher numerical accuracy. In the Bayesian sense, the integration of environmental tracers in groundwater models can reduce the range of possible alternative interpretations consistent with all observations.