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 dicult to measure. A connectivity function is proposed de®ned by the 3D-Euler number. The goodness in the estimation of the Euler number using serial sections is investigated in subsamples of di€erent sizes and shapes. Then, a simple network model is generated which can be adapted to a prede®ned pore-size distribution and connectivity function. Network simulations of hydraulic properties are compared to independent measurements at the same soil material and the e€ect of topology on water ̄ow and solute transport is investigated. It is concluded that a rough estimation of pore-size distribution and topology de®ned by the connectivity function might be sucient to predict hydraulic properties. Ó 2001 Elsevier Science Ltd. All rights