For the determination of groundwater recharge processes in arid environments, vadose zone water fluxes and water storage should be considered. To better understand and quantify vadose zone processes influencing groundwater recharge, a soil column experimental setup has been developed that mimics arid atmospheric conditions and measures water and temperature fluxes in high temporal and spatial resolution. The focus of the experiment was on the determination of water infiltration, redistribution, evaporation and percolation under non-isothermal conditions. TDR rod sensors and a specific TDR ''Taupe'' cable sensor were used for water content measurements and allowed the infiltration fronts to be traced over the whole column length. Applying single irrigations of different amount and intensity showed the applicability of the experimental setup for the measurement of water movement in the unsaturated soil column.

In arid regions, groundwater resources are prone to depletion due to excessive water use and little recharge potential. Especially in sand dune areas, groundwater recharge is highly dependent on vadose zone properties and corresponding water fluxes. Nevertheless, vadose zone water fluxes under arid conditions are hard to determine owing to, among other reasons, deep vadose zones with generally low fluxes and only sporadic high infiltration events. In this study, we present an inverse model of infiltration experiments accounting for variable saturated nonisothermal water fluxes to estimate effective hydraulic and thermal parameters of dune sands. A subsequent scenario modeling links the results of the inverse model with projections of a global climate model until 2100. The scenario modeling clearly showed the high dependency of groundwater recharge on precipitation amounts and intensities, whereas temperature increases are only of minor importance for deep infiltration. However, simulated precipitation rates are still affected by high uncertainties in the response to the hydrological input data of the climate model. Thus, higher certainty in the prediction of precipitation pattern is a major future goal for climate modeling to constrain future groundwater management strategies in arid regions.

No abstract