Soil pores form interconnected networks that directly or indirectly influence transport processes in soils. There is great interest in estimating pore properties such as size and shape from threedimensional (3D) images as a way of understanding linkages between structure and function. This work evaluates the applicability of an approach to measure poresize distributions (PSDs) from 3D images of four soils with contrasting physical properties. Images were obtained with a benchtop Xray microtomograph system at spatial resolutions of 4, 6, 12, and 30 ìm by varying the size of the samples from 0.5 cm to about 4 cm in diameter. Poresize distributions obtained at each image resolution were fitted with lognormal distribution functions with r2 ranging from 0.81 to 0.98. The four PSDs for a given soil were combined by identifying crossover points in the lognormal distributions. When pore sizes where imaged at more than one resolution, the criterion followed to eliminate redundant information was that large pore sizes are better represented in low resolution images of large sample sizes, and vice versa. Poresize data from four truncated PSDs were fitted with a lognormal distribution function to obtain a final (global) PSD for each soil. Global PSDs obtained with the imagebased approach correlated well with PSDs measured by mercury intrusion porosimetry and also with PSDs derived from soil water retention curves, showing potential for application of this imagebased approach to soils.
All Science Journal Classification (ASJC) codes
- Soil Science