University of TehranIranian Journal of Soil and Water Research2008-479X48520180121Estimation of saturated hydraulic conductivity by using gene expression programming and ridge regression (A case study in East Azerbaijan province)Estimation of saturated hydraulic conductivity by using gene expression programming and ridge regression (A case study in East Azerbaijan province)108710956476510.22059/ijswr.2018.218413.667555FAAbbasAhmadiUniversity of Tabriz0000-0002-6667-0221PeymanPalizvan ZandIslamic Azad University of TabrizHabibPalivan ZandUniversity of Tabriz0000-0002-6667-0221Journal Article20161024The hydraulic conductivity of soil is an important physical characteristic, which is used for water modeling and the modeling of solutes and pollutants transport. The direct measurement of soil hydraulic conductivity is a time-consuming and costly process, and due to experimental errors and soil heterogeneity, the results are sometimes unrealistic. Besides, it could be estimated by easily measurable soil properties. The purpose of this study is to develop genetic programming and linear regression models to estimate the saturated hydraulic conductivity of soil using readily available soil properties. With this purpose, 160 soil samples with different properties were gathered from various areas of East Azerbaijan province of Iran. Then some physical and chemical characteristics of soil such as the proportions of sand, silt and clay in the soil, and organic matter, bulk density, pH and EC values were measured. Then the data was divided into two different data sets, namely training (75% of data) and testing (25% of data) datasets. GeneXproTools 4.0 and Statistica softwares were used to calibrate Genetic programming and regression models, respectively. Six pedotranfer functions (PTFs) with a combination of different mathematical operators were designed by the genetic programming. Finally, one of the PTFs which was more accurate than the others was selected. Also, the ridge regression was utilized to develop regression PTFs. The accuracy and reliability of PTFs were determined by R<sup>2</sup>, RMSE, and MAE criteria. The research results showed that the genetic programming PTF (GP-PTF) is more accurate and reliable in comparison with the regression-PTF. In a way that the R<sup>2</sup>, RMSE (Cm h<sup>-1</sup>) and MAE (Cm h<sup>-1</sup>) of GP-PTF were 0.91, 1.82 and 1.23 for the training dataset, respectively, and for the test dataset, the values were 0.92, 2.27 and 1.59, respectively; whereas the values of the above mentioned criteria of regression-PTF for the training dataset were 0.70, 3.48 and 2.07, respectively, and for the test dataset were 0.76, 3.11 and 1.88, respectively.The hydraulic conductivity of soil is an important physical characteristic, which is used for water modeling and the modeling of solutes and pollutants transport. The direct measurement of soil hydraulic conductivity is a time-consuming and costly process, and due to experimental errors and soil heterogeneity, the results are sometimes unrealistic. Besides, it could be estimated by easily measurable soil properties. The purpose of this study is to develop genetic programming and linear regression models to estimate the saturated hydraulic conductivity of soil using readily available soil properties. With this purpose, 160 soil samples with different properties were gathered from various areas of East Azerbaijan province of Iran. Then some physical and chemical characteristics of soil such as the proportions of sand, silt and clay in the soil, and organic matter, bulk density, pH and EC values were measured. Then the data was divided into two different data sets, namely training (75% of data) and testing (25% of data) datasets. GeneXproTools 4.0 and Statistica softwares were used to calibrate Genetic programming and regression models, respectively. Six pedotranfer functions (PTFs) with a combination of different mathematical operators were designed by the genetic programming. Finally, one of the PTFs which was more accurate than the others was selected. Also, the ridge regression was utilized to develop regression PTFs. The accuracy and reliability of PTFs were determined by R<sup>2</sup>, RMSE, and MAE criteria. The research results showed that the genetic programming PTF (GP-PTF) is more accurate and reliable in comparison with the regression-PTF. In a way that the R<sup>2</sup>, RMSE (Cm h<sup>-1</sup>) and MAE (Cm h<sup>-1</sup>) of GP-PTF were 0.91, 1.82 and 1.23 for the training dataset, respectively, and for the test dataset, the values were 0.92, 2.27 and 1.59, respectively; whereas the values of the above mentioned criteria of regression-PTF for the training dataset were 0.70, 3.48 and 2.07, respectively, and for the test dataset were 0.76, 3.11 and 1.88, respectively.https://ijswr.ut.ac.ir/article_64765_4fd9e1649d8cd4ed7213772b4dfac65c.pdf