Soil Science Society of America journal, 2012, Vol.76(5), pp.1509-1517
The gas diffusion coefficient (Ds,g) and solute diffusion coefficient (D(s,l)) and their dependencies on fluid content (κ) (equal to soil–air content θ for D(s,g) and soil–water content ɛ for D(s,l)) are controlling factors for gas and solute transport in variably saturated soils. In this study, we propose unified, predictive models for D(s,g)(ɛ) and D(s,l)(θ) based on modifying and extending the classical Maxwell model at fluid saturation with a fluid-induced reduction term including a percolation threshold (ɛ(th) for D(s,g) and θ(th) for D(s,l)). Different percolation threshold terms adopted from recent studies for gas (D(s,g)) and solute (D(s,l)) diffusion were applied. For gas diffusion, ɛth was a function of bulk density (total porosity), while for solute diffusion θ(th) was best described by volumetric content of finer soil particles (clay and organic matter), FINES(vol). The resulting LIquid and GAs diffusivity and tortuosity (LIGA) models were tested against D(s,g) and D(s,l) data for differently-textured soils and performed well against the measured data across soil types. A sensitivity analysis using the new Maxwell’s Law based LIGA models implied that the liquid phase but not the gaseous-phase tortuosity was controlled by soil type. The analyses also suggested very different pathways and fluid-phase connectivity for gas and solute diffusion in unsaturated soil. In conclusion, the commonly applied strategy of using the same, soil-type-independent model for gas and solute diffusivity in analytical and numerical models for chemical transport and fate in variably-saturated soils appears invalid, except for highly sandy soils. The unified LIGA model with differing percolation thresholds for diffusion in the liquid and gaseous phases solves this problem. ; p. 1509-1517.
Clay ; Sandy Soils ; Bulk Density ; Solutes ; Mathematical Models ; Organic Matter ; Porosity ; Diffusivity
View full text (authentication may be required)