Evaluation of pedotransfer Functions in estimating saturated water content of limy soils

Estimating the soil moisture curve has an important role in modeling water movement and solute transport in the soils. Saturated water content is one of the important parameters in soil studies which is used to estimate the soil water retention curve and unsaturated hydraulic conductivity. Pedotransfer functions are as undirected methods which estimate soil time consuming parameters from readily measured parameters. The multi-linear regression and artificial neural network methods were used to develop the pedotransfer functions. In this research، soil texture، bulk density، soil particle density، organic material percent and lime content percent as readily measured parameters and saturated water content as time consuming parameter were considered. In this study، using soil readily measured parameters in 136 soil samples، 14 models of multi-linear regression and 6 models of artificial neural network were evaluated in order to estimate saturated water content. Finally، measured and estimated values of soil saturated water content were compared and each model ability was evaluated by statistical indices. The indices of Geometric Mean Error Ratio (GMER)، Akaike’s Information Criterion (AIC) and Root Mean Square Error (RMSE) showed that Minasny et al and shinoset et al models had better estimation of saturated water content. The results showed that low content of organic materials had the significant effect on the accuracy of neural network models estimation but lime percent had not the significant effect on the so called models.