Correlation between air permeability and saturated hydraulic conductivity: Unburned and burned soils

Prior research has shown a log-log correlation between ex situ air permeability (k_a) measured at a matric potential of -50 and -100 cm H₂O and saturated hydraulic conductivity (K_sat) in agricultural soils. We examined the applicability of a previously established log k_a vs. log K_sat predictive relation in unburned and burned woodland-chaparral and coniferous soils. Specifically, we measured in situ k_a in the field using the soil corer air permeameter (SCAP) and measured laboratory K_sat on soil cores. To determine postfire soil physical changes, K_sat was transformed into water permeability, k_w. Then k_a and k_w were compared in unburned and burned soils. The median k_a was lower and median k_w was higher in burned than unburned woodland-chaparral soils. The median k _a and k_w were both lower in burned than unburned coniferous soils. Despite these effects, the 95% prediction interval of the predictive relation includes 80% of the burned soils and 83% of all desert soils. The RMSE values of the predicted log K_sat and measured log K_sat were 0.371, 0.552, 0.588, and 0.511 m d^-1 for unburned and burned woodland-chaparral and unburned and burned coniferous soils, respectively. In comparison, the RMSE for the predicted and measured log K _sat was 0.654 m d^-1 for an agricultural data set. The overall RMSE for unburned and burned soils including the agricultural soils was 0.551 m d^-1. This suggests that the predictive relation is applicable for unburned and burned desert soils for in situ k_a measurements.