Parameterisation of physically based solute transport models in sandy soils

Abstract

Immobile water fractions of up to 40% had been reported in sands, and it was therefore relevant to determine if the convection-dispersion equation (CDE) or mobile-immobile model (MIM) should be used as the basic physical model for field studies of solute transport in sandy soils. A review of literature data for granular media indicated that for steady state flow, the dispersion coefficient could be estimated from the grain Peclét number and the mass exchange coefficient from the pore water velocity, but the immobile water fraction was poorly predicted from soil properties.

In this study, performed on a sandy soil at Moora, Western Australia, the choice of model was determined after analysis of column effluent breakthrough curves (BTC), sequential tracer experiments, and single tracer experiments on the same core. The latter two methods have recently been introduced in an attempt to independently measure some of the MIM parameters (immobile water content and mass transfer exchange coefficient) in situ.

Low immobile water content was found in this sand, with very rapid exchange between the mobile and immobile regions. All 3 techniques gave measured immobile water contents around 10%, which was consistent with most literature values for granular media. The single tracer experiment does not give the mass exchange coefficient α (h^–1), but α determined by the sequential tracer technique could not be confirmed by the BTC technique due to the wide 95% confidence interval of the fitted parameter. Although the MIM behaviour was minor and inconsistent in the Moora sand, the choice of model may depend on the problem under consideration. For short column experiments, the CDE and MIM produced similar solute transport behaviour. However, for leaching below the root-zone, the MIM is recommended.