Estimating Saturated Hydraulic Conductivity from Soil Porosity

Measuring the tempospatial variability of saturated hydraulic conductivity (Ks ) is time consuming, expensive,and encounters many uncertainties. This work aimed to develop a new model (REPM, Relative Effective Porosity Model)that estimates Ks from relative effective porosity (er ) and then compare it with a model (EPM, Effective Porosity Model)that estimates it from effective porosity (e ). The effective porosity (e ) is defined as the total porosity minus field capacity(FC), and the relative effective porosity (er ) is defined as effective porosity (e ) divided by FC. Both er and e can beestimated from FC and bulk density (Bd ). Data from 11 homogeneous texturalclass mean soils and several international andAmerican soils were used to evaluate REPM and EPM. For the 11 texturalclass mean soils, log (Ks ) was highly correlatedto log (er ) as well as to log (e ). For the international soils, log (Ks ) was highly correlated to log (er ) (r 2 = 0.77), but thecorrelation was less pronounced between log (Ks ) and log ( e ) (r 2 = 0.58). The saturated hydraulic conductivity of soils froman international database was more accurately predicted by REPM (RMSE of 539 cm d 1 ) than by EPM (RMSE of 733 cmd 1 ), while both of them performed as well for American soils. The slope and the intercept of REPM and the slope of EPMwere independent of soil. These results suggest that our new model gives reasonable estimates of Ks for different soils.