A comprehensive experimental and numerical analysis of water flow and travel time in a highly heterogeneous vadose zone

Abstract

The assessment of water travel time through the vadose zone is known to be critical for a proper estimation of the hydrologic response time of water bodies to evolutions in land use management and global changes. In this study, the hydraulic properties of fifteen samples displaying contrasted lithologies (soil, powdery limestone, calcareous sand, limestone rock) and extracted from three cored boreholes drilled throughout the vadose zone of a vulnerable limestone aquifer were first determined in the laboratory. Three 23 m-deep vadose zone profiles were then reconstituted with the HYDRUS-1D software for the numerical simulation of water flow and the estimation of the water travel time using unimodal and bimodal approaches. The measured hydraulic properties, meteorological and water table level data were used as input for a virtual bromide tracing experiment undertaken over a period of 55 years (1966–2020). The results showed that the experimental hydraulic properties of the samples were more accurately described with a dual porosity approach, since the latter allowed a precise representation of the bimodal characteristic of most of the samples. The water flow and travel time simulated using unimodal or bimodal models for describing the vadose zone hydraulic properties were largely different. The impact of the vadose zone lateral heterogeneities on the simulated water flow and the estimated travel time was relatively limited compared to the influence of the vadose zone vertical heterogeneities and meteorological conditions. The mean travel time of the first concentration, peak concentration and last concentration of bromide simulated with the bimodal model at the maximum water table level of the aquifer was 13.8, 20.9 and 31.5 years, respectively. Increase in travel time was clearly identified since the late 1970s, and could be a consequence of global warming. These results also pointed out the need for conducting extensive studies at larger scales to take into account possible fast transfers of water that might occur through open fractures and karst networks.