On the estimation of saturated hydraulic conductivity: from local to field scale

This thesis presents a comparison of Ks estimates obtained by three classical devices, namely the double ring infiltrometer (DRI), the Guelph version of the constant-head well permeameter (GP) and the CSIRO version of the ten- sion permeameter (CTP). A distinguishing feature of this study is the use of steady deep flow, obtained from controlled rainfall-runoff experiments, as a benchmark of Ks at "local" and field/plot scales to assess the reliability of the above methods. Spatially representative estimates of Ks are needed for simulating catchment scale surface runoff and infiltration. Classical methods for measuring Ks at the catchment scale are time-consuming. Important insights can be obtained by experiments aimed at understanding the controls of Ks in an agricultural setting and identifying the minimum number of samples required for estimating representative plot scale Ks values. This thesis presents results from a total of 131 double-ring infiltrometer measurements at 12 plots in a small Austrian catchment. Classical field techniques to determine Ks at the plot and catchment scales are complex and time-consuming, therefore the development of pedotransfer functions, PTFs, to derive Ks from easily available soil properties is of utmost importance. However, PTFs have been generally developed at the point scale, while application of hydrological modeling requires field scale estimates. In this thesis, values of field-scale saturated hydraulic conductivity, K ̄s , measured in a number of areas within the Austrian catchment, have been used to derive two PTFs by multiple linear regression (PTFMLR) and ridge regression (PTFR). Two alternative approaches have been used: (A) soil properties have been first interpolated with successive application of the PTFs, (B) the PTFs have been first applied in the sites where soil properties were available and then interpolated.