Spatial Variability Analysis of Soil Salinity and Alkalinity in Chat Region, Golestan Province

Knowledge of important soil properties such as salinity and alkalinity plays a central role in soil and water management. The objectives of this paper were to investigate the spatial variability of soil electrical conductivity (EC)، exchangeable sodium percentage (ESP)، and acidity (pH)، and to map these parameters in soil surface layer (0-24 cm depth). The study was performed using 152 soil samples taken from Chat in Golestan province. The estimation methods used were geostatistical approaches of ordinary kriging، log-normal kriging، and co-kriging، and inverse distance weighting. The performance of the selected methods was evaluated through cross-validation technique with comparison criteria of root mean square error (RMSE)، mean absolute error (MAE)، and mean bias error (MBE). Geostatistical analyses showed that all parameters mouseover=«function onmouseover() {msoCommentShow (''_anchor_1''،''_com_1'')}» onmouseout=«function onmouseout() {msoCommentHide (''_com_1'')}» href= «#_msocom_1» > [D1] had moderate spatial correlation over the study area. The best model of semivariogram was exponential for EC، while ESP and pH followed a spherical structure. The smallest correlation distance (about 3000 m) belonged to EC and the largest correlation distance (about 6000 m) was found for ESP and pH. Based on the cross-validation results and kriging ability to predict estimation uncertainty، ordinary kriging was the most appropriate method for estimating soil EC (MAE= 8. 886 dS/m)، pH (MAE= 0. 183) and ESP (MAE= 13. 19). The generated maps showed that the majority of the study area suffered from salinity and alkalinity problem. However، EC and ESP showed a similar increasing drift from the east to the west of the region. Therefore، lands located in the west of the studied area need more attention in terms of management practices such as proper cropping pattern and soil reclamation.