Geostatistical modeling of groundwater quality in the Moghan Plain based on the Wilcox index
Subject Areas : Pollution of water resources
1 - Professor, Department of Water Science and Engineering, Faculty of Agriculture and Natural Resources, University of Mohaghegh Ardabili, Ardabil, Iran.
Keywords: Assessment, Geostatistics, Water quality, GIS, Zoning,
Abstract :
This study evaluated the groundwater quality of the Moghan Plain based on the Wilcox Index using geostatistical methods. The evaluation results obtained using the cokriging method showed that the covariograms of the random variables electrical conductivity (EC) and sodium adsorption ratio (SAR) followed the Stable model. For the SAR variable, the covariogram had a nugget effect of 0.01 and a sill of 0.08, whereas these values were 0.02 and 1.07, respectively, for electrical conductivity. These findings indicate the dominance of the structured component of the covariogram over its unstructured component and confirm the strong spatial structure of both random variables. Furthermore, the results showed that geostatistical methods, particularly the simple cokriging model, provided accurate estimates of groundwater quality, with a coefficient of determination (R²) of 0.67 and a normalized root mean square error (NRMSE) of 0.34 for SAR, and an R² of 0.99 and an NRMSE of 0.007 for electrical conductivity. Overall, approximately 1% of the groundwater in the plain was classified as C2S1, indicating its suitability for agricultural use. Among the remaining 99% of the groundwater, which exhibited high salinity and was unsuitable for agriculture, nearly 14% also had excessive sodium levels, making it unsuitable for irrigation.