Quantifying soil heterogeneity from solute dispersion experiments

The complicated morphology of the pore space of sandy/clayey/silty soils is characterized by multi-scale heterogeneities. The quantification of soil heterogeneity is a difficult task associated with a high uncertainty. For this purpose, miscible displacement experiments were performed on undisturbed and heterogeneous soils. The solute concentration averaged over various cross-sections was determined by monitoring the electrical conductivity with pairs of rod electrodes. The solute concentration breakthrough curves were measured on three cross-sections of two undisturbed soil columns at various flow rates. The datasets of all experiments were fitted successfully using a multi-region model in which the heterogeneous medium is regarded as a system of parallel homogeneous regions quantified by two parameters: (i) the longitudinal dispersivity aL, a measure of the macro-heterogeneity, which reflects the intensity of preferential flow paths; (ii) the standard deviation σk⁎, of the region permeability distribution, a measure of the micro-heterogeneity related to variations of the effective pore radii. Macroscopic simulations show that the estimated region log-permeability distribution shifts to lower values and its standard deviation increases compared to the actual permeability distribution.