A Physically Based Method for Soil Evaporation Estimation by Revisiting the Soil Drying Process

Abstract

AbstractWhile numerous models exist for soil evaporation estimation, they are more or less empirically based either in the model structure or in the determination of introduced parameters. The main difficulty lies in representing the water stress factor, which is usually thought to be limited by capillarity‐supported water supply or by vapor diffusion flux. Recent progress in understanding soil hydraulic properties, however, have found that the film flow, which is often neglected, is the dominant process under low moisture conditions. By including the impact of film flow, a reexamination on the typical evaporation process found that this usually neglected film flow might be the dominant process for supporting the Stage II evaporation (i.e., the fast falling rate stage), besides the generally accepted capillary flow‐supported Stage I evaporation and the vapor diffusion‐controlled Stage III evaporation. A physically based model for estimating the evaporation rate was then developed by parameterizing the Buckingham‐Darcy's law. Interestingly, the empirical Bucket model was found to be a specific form of the proposed model. The proposed model requires the in‐equilibrium relative humidity as the sole input for representing water stress and introduces no adjustable parameter in relation to soil texture. The impact of vapor diffusion was also discussed. Model testing with laboratory data yielded an excellent agreement with observations for both thin soil and thick soil column evaporation experiments. Model evaluation at 15 field sites generally showed a close agreement with observations, with a great improvement in the lower range of evaporation rates in comparison with the widely applied Priestley and Taylor Jet Propulsion Laboratory model.