Hydraulic conductivity variability for two sandy soils

Reliable estimates of saturated hydraulic conductivity are a prerequisite for accurate estimations of water flow and chemical transport through soil profiles. While informative, mean estimates of flow and transport have been shown to be inadequate for modeling purposes. Distributions and variances along with mean predictions provide a better form of model input. The spatial variability of field-measured saturated hydraulic conductivity (K fs ) was investigated on a transect across a 1-ha field in west-central Georgia. The field includes two soil types, a Pine Flat loamy sand (coarse-loamy, siliceous, thermic Typic Paleudult) and a Troup sand (loamy, siliceous, thermic Grossarenic Kandiudult). Estimates of K fs were calculated using data collected with a constant-head permeameter. Mean, variance, and distribution data were determined for each soil and four selected horizons. Calculated values in the top 2 m of the soil profiles ranged from 0.07 to 117 mm h ' for the Pine Flat loamy sand and from 23 to 406 mm h ' for the Troup sand. Statistical differences between K fs data calculated for the lower soil profiles coincide with soil textural differences observed with depth and across the plot. Semivariances for In K fs data were calculated and models fit to the semivariograms. Random behavior, with little spatial correlation, dominated the In K fs data for the upper 0.5 m of the Pine Flat soil and the 2.0-m measurement for the Troup soil. Observed correlation ranges varied from 2 to 166 m. Large statistical differences in K fs were observed for the adjacent soils.