Spatial Variability of Saturated Hydraulic Conductivity of the Subsoil of Two Forested Watersheds

Abstract

AbstractSubsurface flow has been shown to be greater in soils with spatially dependent hydraulic properties. The objective of this study was to quantify the spatial dependency in saturated hydraulic conductivity, Ksat, of two potential shallow waste disposal sites in eastern Tennessee. The Guelph Permeameter was used to determine Ksat of the subsoil of two forested watersheds, Walker Branch and Melton Branch. The simultaneous equations approach resulted in more than 50% negative Ksat values and was concluded to be invalid for these forested subsoils. As an alternative, flux density was used as an index of the saturated hydraulic conductivity for spatial correlation analysis and for determination of scaling factors. the spatial realizations of hydraulic conductivity for both watersheds were found to be intrinsic random functions of order zero, i.e., stationary, with a generalized covariance function for Walker Branch consisting of a pure nugget and for Melton Branch a nugget with a linear term. Semivariograms revealed no spatial dependency for separation distances >4.2 m at Walker Branch. Spatial dependency at Melton Branch for separation distances <30 m was described by an exponential model. The hydraulic conductivity of Walker Branch subsoils had a greater arithmetic mean (1.50 vs. 0.76 × 10−5 m s−1) and greater random variability than Melton Branch subsoils. These findings suggest that soils common to Walker Branch Watershed are hydrologically preferable for shallow waste disposal.