Deriving pedo‑transfer functions for estimating soil saturated hydraulic conductivity and its mapping in GIS in some semi‑arid soils

Abstract

The soil saturated hydraulic conductivity is an important parameter for designing and managing irrigation systems. Accurate prediction of soil saturated hydraulic conductivity by deriving PTFs (pedo-transfer functions) from easily to measure soil properties is of fundamental importance. The aims of this research were to derive some regression models including MLR (multiple linear regression)- and GMDH (group method of data handling)-based PTFs for estimating soil saturated hydraulic conductivity in a semi-arid region. A double-ring method at 84 studied sites was employed to measure water infiltration data. The soil saturated hydraulic conductivity data were calculated through optimizing the parameters of Green and Ampt infiltration model on measured infiltration data. Additionally, easily to measure soil properties including soil clay, silt, sand, soil moisture contents, saturation percent, soil calcium carbonate equivalent, gravel contents, and soil organic carbon data were determined for all sites. The results of the present study indicated that the GMDH-based PTFs performed better (ME = − 0.0149 cm min− 1, R2 = 0.614, E = 0.729, and RMSE = 0.0447 cm min− 1) than the MLR-based PTFs (ME = − 0.0122 cm min− 1, R2 = 0.361, E = 0.595, and RMSE = 0.0532 cm min− 1). This result can be a useful guide for predicting soil saturated hydraulic conductivity for implementing a suitable irrigation system under semi-arid conditions.