Correlating pore space morphology with numerically computed soil gas diffusion for structured loam and sand,...

Area of Science:

• Soil Science
• Geophysics
• Biogeochemistry

Background:

• Biogeochemical soil processes, like nutrient transport, are fundamentally linked to soil structure.
• The soil's pore network dictates crucial properties such as diffusivity and permeability.
• Quantitative analysis of soil morphology and numerical simulations can bridge microscopic structure and macroscopic properties.

Purpose of the Study:

• To validate and extend existing parametric regression formulas for predicting soil properties.
• To assess the applicability of these formulas to structured soils beyond finely sieved samples.
• To explore the use of artificially generated soil structures for formula validation.

Main Methods:

• Utilized 3D CT (Computed Tomography) data from six soil samples (three loam, three sand).
• Conducted extensive structural analysis of the soil samples.
• Performed numerical simulations to analyze the relationship between soil structure and effective macroscopic properties.

Main Results:

• Most regression formulas showed improved performance when parameters were adjusted for specific soil structures.
• Certain formulas, particularly those based on tortuosity, reliably predicted diffusion in structured soils without parameter adjustment.
• Artificially generated soil structures show potential for validating and improving regression formulas.

Conclusions:

• Regression formulas for predicting soil properties can be adapted for structured soils.
• Some existing formulas demonstrate robustness across different soil structures.
• Stochastic 3D microstructure models offer a promising avenue for future soil structure research and model validation.