Water Retention Model Research Articles

Analysis of the inverse problem for transient unsaturated flow

The inverse problem of determining unsaturated soil hydraulic properties from one‐dimensional, transient infiltration and redistribution events is analyzed. Hydraulic properties are assumed to be described by an extension of van Genuchten's (1980) model which allows for hysteresis in the retention function and air entrapment. Unknown parameters in the model are estimated from observed water contents and heads during transient flow by numerical inversion of the unsaturated flow equation. The inverse problem is formulated as a weighted least squares problem and solved using an efficient Levenberg‐Marquardt algorithm. The flow event consists of ponded infiltration followed by gravity drainage with evaporation at the soil surface. Sensitivity analyses indicate that observations during ponded infiltration should be made near the position of the wetting front. The location of observation points during the drying stage is less critical than during the infiltration stage, but for the relatively high imposed evaporative flux sensitivity of pressure head is highest near the soil surface. Large differences in sensitivity are observed among the various model parameters. Unknown evaporative fluxes are approximated in the inverse solution as an equivalent first‐type boundary condition requiring only periodic measurements of surface water content during the drying stage. Little error is incurred provided accurate measurements are possible. Corruption of input data with random error is shown to have a larger effect on the predicted conductivity function than on the retention function and more effect on the wetting branch of the hysteretic retention function than on the drying. When measurements are subject to error and the assumed parametric model for water retention and conductivity relations is not exact, it may no longer be possible to detect hysteresis in the retention function.

Soil Survey Interpretations from Water Retention Data: II. Assessment of Soil Capability Ratings and Crop Performance Indices

AbstractA water retention model developed for Ontario soils is used to estimate the preseason plant‐available moisture‐holding capacity (Θva) of the soil root zone in equilibrium with a phreatic surface and the air‐filled porosity (ε) of the surface 30 cm. Preliminary soil capability class ratings are then assessed from a set of regional guidelines based on class intervals of Θva and ε. Crop performance indices are determined for capability classes 1 to 5 inclusive using a statistical corn productivity model which relates long‐term mean grain corn yields in Ontario to a seasonal soil moisture deficit or surplus. Total seasonal field water supply is budgeted against potential crop evapotranspiration to estimate the extent to which plant water stress from excess or deficit soil water availability limits the attainment of regional agro‐climatic potential yields. Advantages in using a water retention model as a basis for agronomic interpretations of soil survey data for land evaluation purposes are discussed.

Soil Survey Interpretations from Water Retention Data: I. Development and Validation of a Water Retention Model

AbstractThe prediction of spatial and temporal variations in the soil water regime provides the basis for a proposed soil survey interpretive system. A water retention model estimates the boundary desorption curve from saturation to pF 4.2 for mineral soils in situ using limited soil physical data (standard error of estimate <2.2% g g−1). The total porosity (structural parameter) and clay content (adsorptive parameter) define the position of the moisture characteristic in logarithmic coordinates and the silt content its configuration. Organic matter affects the structural parameter by reducing both the dry bulk and particle densities but only significantly influences the adsorptive parameter at organic matter contents of 5% or greater. Model validation shows that the state and quantity of water retained by field soils near equilibrium is dependent on the position of the phreatic surface. Under field conditions, the ψ(θ) relationship establishes an effective equilibrium midway between the predicted wetting and drying curves of the hysteresis loop.