Estimating the permeability of soils under different tillage practices and cropping systems: Roles of the three percolating pore radii derived from X-ray CT

Soil permeability associated with pore characteristics is a critical factor affecting water and nutrient transport in agricultural fields. However, which soil pore characteristics determine soil permeability remains inconclusive due to high heterogeneity of soil pore structure affected by soil and crop management practices. The objective of this study was to compare the pore characteristics of soils under various field management and assess the performance of different pore radii in estimating soil permeability. In this study, we utilized X-ray computed tomography (XCT) to quantify the pore characteristics of soil cores from three treatments: no-tillage under continuous maize cropping (NT-MM), no-tillage under continuous soybean cropping (NT-SS) and conventional tillage under continuous maize cropping (CT-MM). Subsequently, we calculated the permeability with pore-scale numerical simulations directly based on XCT images. Our results revealed that compared to the NT-MM treatment, the NT-SS and CT-MM treatments significantly increased soil porosity of > 200 mu m pore classes (soil depths of 2-8 cm and 2-14 cm), as well as pore compactness, specific surface area, fractal dimension, and global connectivity. Among the three XCT-derived pore radii, the critical pore radius (CR) exhibited excellent performance in estimating permeability for samples from the NT-SS and CT-MM treatments, surpassing the hydraulic radius (HR) and mean pore radius of the limiting layer (MRLL). For samples from the NT-MM treatment, the CR and MRLL had similar estimation capabilities (R2 = 0.429, P = 0.110), but both were not as good as the HR (R2 = 0.539, P = 0.060). Furthermore, when estimating the permeability of all the aforementioned soil samples, the CR demonstrated the best estimation performance (R2 = 0.838, P < 0.001). In conclusion, utilizing CR as a predictor for estimating permeability in naturally structured soils is a good choice, which provides valuable insights into improving the accuracy of estimating soil permeability.