Comparison of determination methods for saturated soil hydraulic conductivity with Guelph infiltrometer

Soil hydraulic properties are key factors controlling water and solute movement in soil. Soil hydraulic properties estimated from a laboratory experiment use commonly on relatively small soil cores, and they are often not representative of field condition. Ideally, the estimation of soil hydraulic parameters for the simulation of flow dynamics should be based as far as possible on field measurements. Reliable determination of saturated hydraulic conductivity is critical for modeling and predicting soil moisture pattern in soil under various field conditions. In this study, soil saturated hydraulic conductivity (Ks) of a forest land and a sloping farmland, located at Yanting Agro-ecological Experimental Station of Purple Soil (105?27?E，31?16?N), Sichuan, Southwest China, was determined with a Guelph infiltrometer using different methods, including single head methods (e.g., Laplace analysis, updated single head analysis) and two-head methods (e.g., two-head analysis, Bohne analysis). Differences in measured saturated hydraulic conductivity between in situ Guelph constant head methods and laboratory falling head method were also compared. We aimed to select a reliable method for determining Ks for the soils in this studied region. The Ks values obtained for the forest land at the 0-20 cm depth were lower than those obtained at the 20-40 cm depth; however, the results obtained from the sloping farmland were contrary to those measured in the forest land. For the same layer, the measured Ks value was related to the method adopted. Results showed that: 1) the Ks values using the two-head analysis were the greatest (except for the 20-40 cm layer of the sloping farmland). The Ks value in the forest land at the 0-20 cm and 20-40 cm depth was 0.134 and 0.266 mm/min, respectively. In the sloping farmland, the Ks value at the 0-20 cm depth was 0.86 mm/min. However, the updated single head analysis under a 5cm head was the smallest values. The Ks value at the 0-20 cm and 20-40 cm depth was 0.015 and 0.022 mm/min, respectively, in the forest land, 0.040 and 0.022 mm/min, respectively, in the sloping farmland; 2) Using the updated single head analysis, Ks measured at water head of 10 cm was higher than that obtained at 5 cm. The obtained Ks value was 0.031 and 0.056 mm/min at the 0-20 cm and 20-40 cm depth, respectively, in the forest land, 0.211 and 0.031 mm/min, respectively, in the sloping farmland; 3) The values of saturated hydraulic conductivity obtained by the Guelph infiltrometer using the Bohne analysis and Updated single head analysis at water head of 10 cm were greater than those obtained in the laboratory by falling head method on ring samples, probably due to the possible destruction of soil macropores and the smaller soil volume represented by the latter. Also since the updated single head analysis has the advantage of simple operation, the method at water head of 10 cm is therefore recommended in the purple soil region.