Modeling saturated hydraulic conductivity based on soil characteristics

The saturated hydraulic conductivity (KS) of several roadside cuttingswith slope failure around Universiti Putra Malaysia were studied to see how the occurrence is related to the Ks. The study was to develop a model of Ks based on soil characteristics. Ks measurements were done by using Double Ring Infiltrometer and Guelph Permeameter. A model was then developed from the soil characteristics using SAS Stepwise (MAX-R) Analysis to predict the saturated hydraulic conductivity at the slopes. The soil parameters include particle size distribution, bulk density and soil moisture contents. The best regression model developed has 5 variables for prediction of soil Ks value with R2= 0.630. The variables are percent clay, silt and sand, bulk density and soil moisture content. The model performed relatively well as seen from the predicted and the measured Ks values being scattered along the 1:1 line. This model give good estimate of saturated hydraulic conductivity of clayey soils at the top of the slopes where there was occurrence of slope failure. The extrapolated terminal infiltration rate from Double Ring Infiltrometer also provide good estimate of Ks.