Quantitative morphology and network representation of soil pore structure

Pore-network models are attractive to relate pore geometry and transport processes in soil. In this contribution a 'morphological path' is presented to generate a network model based on quantitative morphological investigations of the 3D pore geometry in order to predict soil hydraulic properties. The 3D-geometry of pores larger than 0.04 mm in diameter is obtained using serial sections through impregnated samples. Beside pore-size distribution an important topological aspect of pore geometry is the spatial connectivity of the pore space which is difficult to measure. A connectivity function is proposed defined by the 3D-Euler number. The goodness in the estimation of the Euler number using serial sections is investigated in subsamples of different sizes and shapes. Then, a simple network model is generated which can be adapted to a predefined pore-size distribution and connectivity function. Network simulations of hydraulic properties are compared to independent measurements at the same soil material and the effect of topology on water flow and solute transport is investigated. It is concluded that a rough estimation of pore-size distribution and topology defined by the connectivity function might be sufficient to predict hydraulic properties. (C) 2001 Elsevier Science Ltd. All rights reserved.