A reliable estimate of recharge is needed for the sustainable management of groundwater resources. Water stable isotope (d 18 O and d 2 H) profiles in the unsaturated zone are frequently used to quantify groundwater recharge based on the seasonality of water isotopic compositions in precipitation. A very simple approach consists of integrating the soil water content between peak values of soil water isotopic composition, typically corresponding to precipitation signatures from warm and cold seasons. When precipitation isotopic compositions are available, a conceptual surface water isotopes budget and lumped parameter dispersion model can be computed. These models were applied on two field sites with similar permeable soils with grass cover but contrasting recharge regimes and seasonality, one in the Paris Basin (France) with continuous recharge from autumn to spring and the other in the St. Lawrence Lowlands (Quebec, Canada) with episodic recharge in fall and after snowmelt. For the two sites, the peak-topeak method and isotope surface budget led to comparable recharge intensities. At least at the Paris Basin site, evaporation was shown to slightly modify the average unsaturated zone and hence groundwater isotope composition. The proposed parameterization of isotope fractionation due to evaporation allows qualitative estimation of the fraction of evaporation, at least during the recharge seasons. In spite of its simplifications and limitations, the proposed parsimonious model can give estimates of recharge in a variety of sites even if they are not well characterized, as it benefits from the large availability of monthly isotopic compositions in precipitation.

Modeling water and solute transport in the vadose zone for groundwater resource management requires an accurate determination of soil hydraulic parameters. Estimating these parameters by inverse modeling using in situ observations is very common. However, little attention has been given to the potential of pore water isotope information to parameterize soil water transport models. By conducting a Morris and Sobol sensitivity analysis, we highlight the interest of combining water content and pore water isotope composition data in a multi-objective calibration approach to constrain soil hydraulic property parameterization. We then investigate the effect of the sampling frequency of the observed data used for model calibration on a synthetic case. When modeling is employed in order to estimate the annual groundwater recharge of a sandy aquifer, it is possible to calibrate the model without continuous monitoring data, using only water content and pore water isotopic composition profiles from a single sampling time. However, even if not continuous, multi-temporal data improve model calibration, especially pore water isotope data. The proposed calibration method was validated with field data. For groundwater recharge estimate studies, these results imply a significant reduction in the time and effort required, by avoiding long-term monitoring, since only one sampling campaign is needed to extract soil samples.