Evaporation From Deep Aquifers in Arid Regions: Analytical Model for Combined Liquid and Vapor Water Fluxes

Abstract

Evaporation is a significant part of the water cycle in hyper‐arid environments. The subsurface of these deserts is characterized by deep groundwater with negligible recharge, whereby water flows from the water table to the surface and evaporates. We propose an analytical model to predict the evaporation rate and the position of the evaporative front. The model accounts for water table depth, atmospheric conditions, and soil hydraulic properties. We consider steady state flow, with two distinct regions separated by an evaporative front, liquid‐phase flow from the water table to the front and vapor‐phase flow from the front toward the surface. The driving forces are pressure head gradients for Darcian liquid flow, and thermal and relative humidity gradients for Fickian diffusive vapor flow. Evaporation rates are predicted for different soil types. The impact of constitutive models applied for characterizing these soils, groundwater depth, and atmospheric conditions are evaluated. Evaporation increases as groundwater levels are shallower, and as atmospheric temperatures increase and/or relative humidity values decrease. Evaporation decreases exponentially with groundwater depth, approaching a constant value of about 0.02 mm per year under typical atmospheric conditions and water table depths below 500 m. The impact of soil type and other related uncertainties are important when groundwater is shallower than 300 m. The relative portion of the liquid phase region increases compared to that of the vapor one as evaporation rates increase. The actual size of the liquid phase flow region, however, reaches its maximum when the water flux approaches zero at hydrostatic conditions.

Liquid and Vapor Water in Vadose Zone Profiles Above Deep Aquifers in Hyper‐Arid Environments

“…What are the characteristics of the water regime in deep desert vadose zones when there is practically no recharge and the system is very dry? This question was addressed by several studies (Kamai & Assouline, ; Ripple et al, ; Ross, ; Selker, , ; Shokri, ; Walvoord et al, ). Ross () concluded that “ downward flow of liquid will result from a thermally driven upward diffusion of vapor .” Walvoord et al () presented a modeling approach that “ reconciles the paradox between the recognized importance of plants, upward driving forces, and vapor flow processes in desert vadose zones and the inadequacy of the downward‐only liquid flow assumption of the conventional chloride mass balance approach.” Their work has shown that water transport in thick desert vadose zones at steady state is usually dominated by upward vapor flow.…”

“…It has been suggested that the main component ruling the upward vapor flow is the geothermal gradient (Selker, ; Walvoord et al, ). Kamai and Assouline () have shown the relative importance of the liquid and vapor phase portions of the profile, the depth of the water table, and the soil type on estimating the evaporation rate from deep groundwater.…”

“…The studies of Ripple et al () and Kamai and Assouline () consider that, in the case of evaporation from a deep water table, continuity of liquid water cannot be sustained over the entire profile. Water in liquid phase can flow only partway from the water table toward the surface, up to an evaporative front located at some depth above the water table.…”

“…The part of the water vapor that can potentially condensate depends on the amount of water in the vapor phase and on its temperature. This conceptual model was presented in Kamai and Assouline (), where the focus was on the evaporation rates and the partitioning between the vapor‐ and liquid‐phase regions as function of soil hydraulic properties and water table depth. In this study, we first revisit the model to update the location of the evaporative front and then consider two different hydrological scenarios.…”

Reply to Comment by N. Shokri on “Analytical Estimation Show Low Depth‐Independent Water Loss Due to Vapor Flux From Deep Aquifers”

“…S18 starts with the notion that the system is at steady state , as did the other recent discussion of this problem by Kamai and Assouline (, hereafter KA18). S18 finds a steady state surface flux of 0.012–0.059 mm/year (apparently skewed strongly toward the lower value based on S18 Figure 2).…”