A conceptual model for infiltration in two-layered soils with a more permeable upper layer: From local to field scale

SummaryA conceptual model with a semi-analytical formulation for estimating the expected field-scale infiltration in two-layered soils is proposed here. At the local scale, an upper layer with a vertically homogeneous saturated hydraulic conductivity, K1s, much greater than that of subsoil, K2s, is considered. The basic element of the local model is the relatively simple mathematical representation of infiltration as a function of time. This representation depends on whether the dynamic wetting front is entirely within the upper layer when surface saturation occurs, or if the front has moved past the interface into the subsoil at the time to ponding. This local-scale infiltration model is found to be very accurate when compared to the Richards equation. Moreover, the local model clearly identifies if infiltration is controlled by rainfall, the top soil layer, or the bottom soil layer. For upscaling to field-scale, K1s is represented by a spatial horizontal random field while K2s for the subsoil is assumed constant. Areally-averaged infiltration is obtained by integrating the local infiltration equations. Monte Carlo simulations over a vertical soil profile representative of a sandy loam top soil and a clay loam subsoil are performed for constructing the ensemble averages of field-scale infiltration to be used for model testing. The simulations reveal that the proposed model shows promise for field-scale infiltration computations with errors referred to the aforementioned numerical benchmark typically less than 10% and 5% for infiltration rates and cumulative infiltration, respectively. Conditions when spatial variability in infiltration properties of the top soil can be neglected are identified.

► A conceptual model for local infiltration in two-layered soils is proposed.
► An extension from local to field-scale is presented.
► The upper soil layer is considered much more permeable than the subsoil.
► The model relies on algebraic equations for local and areal infiltration.
► The model involves considerable accuracy and low computational effort.