Soil Water Retention Curves and Their Impact on Evaporation

An accurate description of soil moisture dynamics in both the liquid and vapor phases is crucial to properly estimate soil evaporation. Soil moisture dynamics are largely dependent on the soil water retention. In the most commonly used models the water retention curve diverges at residual water content, the value below which liquid connectivity is lost and liquid flow stops. Not only this is physically unrealistic but results in incorrect evaporation modeling near dry conditions since the water vapor role is limited. We consider two of the main parametrizations that allow vapor flux below residual water content (modified models): one consists in a re-fit of the standard retention curve with zero residual water content, the other, supported by some laboratory measurements, considers a linear extension (on a semi-log plot) of the standard retention curve in the dry region. For a medium-textured sand and a loam we numerically investigate the effects of both the modified and the standard Van Genuchten models on the liquid and vapor transport during the simulated drying process, with and without surface radiative forcing. In the isothermal case, we show how all the models almost identically describe the capillary-dominated evaporative regime whereas when vapor diffusion is the dominant evaporative mechanism the modified models yield larger and longer sustained vapor fluxes, significantly increasing soil water removal. In the presence of diurnal radiative forcing at the soil surface, we focus on the effects of temperature fluctuations on soil water retention. The impact on liquid and vapor fluxes is analyzed in order to assess whether temperature-dependent and dry-extended retention curves may 'fill the gap' or not between theory and some still debated field experimental evidences (e.g. the midday moisture content rise) without the need of introducing any questionable and ad-hoc empirical terms such as vapor enhancement and/or liquid gain factors.