Groundwater recharge estimation using surface modeling and validation with stable water isotopes

Accurate estimation of groundwater recharge is of paramount importance to guarantee sustainable groundwater management. However, quantifying recharge in aquifers is a challenge, particularly in irrigated agricultural environments. Indeed, recharge processes are deeply impacted by surface inputs, vegetation dynamics, and vadose zone processes whose effects on both the magnitude and timing of recharge are further influenced by heterogeneous land use and irrigation practices. To handle these complexities, it is highly recommended to adopt sound approaches to minimize uncertainties in recharge estimation.

This study tackles this challenge through the estimation of groundwater recharge in the Crau aquifer in southeastern France. The area is intensively irrigated, and the aquifer is recharged by both precipitation and irrigation water from the Durance River. Groundwater recharge was estimated using a spatially distributed soil water balance model combining three models according to land use. An empirical model was used for bare or sparsely vegetated soils. This model is based on observations measured by a flux tower located on a steppic land representative of the area. For irrigated woody crops and gardening, the recharge was computed using a Kc model that calculates evapotranspiration from the reference ET0 using a Kc crop coefficient. For field crop and irrigated grassland, the STICS crop model was used to depict seasonal variations in the water balance and the impact of agricultural practices on recharge. Implementing such a comprehensive groundwater model requires documenting the parameters of every spatial entity (>100000) by using soil, meteorological land use maps and an assessment of agronomic practices.

The surface model outputs were then validated using a stable isotope mass balance based on measurements of oxygen (δ¹⁸O) and hydrogen (δ²H) isotopes in groundwater, precipitation, and irrigation water. The strong contrast between the isotopic signatures of precipitation and irrigation water is interesting to delineate the different water flows between the surface and the groundwater. By combining surface modeling with isotope-based validation, this approach provides an independent means of validating the modeled recharge components in a context where recharge estimation is highly uncertain and contributes to better groundwater management under increasing climatic and anthropogenic pressures.