IN SITU MEASUREMENT OF FIELD-SATURATED HYDRAULIC CONDUCTIVITY, SORPTIVITY, AND THE α-PARAMETER USING THE GUELPH PERMEAMETER

The Guelph permeameter method is used in 0.02-m- and 0.03-m-radius wells to measure in situ the field-saturated hydraulic conductivity (K10) and matric flux potential (φm) of a heterogeneous, anisotropic, structured loam soil. The K10, estimates, obtained using both Richards (GP-R) and Laplace (GP-L) analyses, are compared with saturated hydraulic conductivity (K.) estimates obtained using vertically and horizontally oriented, “undisturbed” soil cores. The φm values, obtained using both GP-R and the Gardner (GP-G) analyses, are used in estimating soil sorptivity (S) and the α-parameter of the exponential hydraulic conductivity-pressure head relationship for infiltration. The K10 values are found to effectively average the vertical and horizontal K. values. In accordance with the theoretical prediction, the GP-L calculation of K10, which neglects capillarity, significantly overestimates the GP-R calculation, which accounts for capillarity. This implies that earlier well permeameter analyses that neglect capillarity can also be expected to overestimate K10 (in the absence of overriding factors, such as smearing of the well wall). The GP-G calculation of φm, which neglects field-saturated flow, overestimates the corresponding GP-R estimate, which takes it into account. This also agrees with theoretical prediction. The mean S estimates are plausible for the soil type and condition, though the mean α values are high relative to some published values. It is speculated that the α values may, in fact, be appropriate considering that they apply to transient, 3-D infiltration into a structured soil. The 0.03-m-radius wells, which were dug using a “Riverside” bucket auger, tended to produce higher K10 and φm values than the 0.02-m-radius wells dug with a screw-type auger. The difference, however, is attributed to heterogeneity and the macroporous structure of the soil, rather than to auger type. The K10 and φm estimates from the GP-R analysis are considered valid and accurate if they are both positive and both lower than their corresponding GP-L and GP-G estimates.