Prediction of near saturated soil hydraulic properties by using CT images and network model

Near-saturated soil hydraulic properties mainly depend on the presence of structural macropores, which in turn, determine the preferential pathways of water flow and solute flux in soil. To characterize the effect of macropore geometry and topology on effective hydraulic properties, a pore-scale network model was constructed by using measured pore morphology. Digital images of sequential soil sections were obtained by computerized tomography (CT) and pore-size distribution and connectivity parameters were determined by digital image analysis. A spatially-correlated network model was then set up to simulate the pore scale water retention and flow processes, and to predict the hydraulic properties near saturation. Result indicate that, although random network agrees well with hydraulic properties measured on laboratory repacked samples, the dual-porosity nature of undisturbed soil can only be realistically reproduced by the correlated network.