<![CDATA[Accurate assessment of soil salinity is essential for managing salt-affected soils and sustaining agricultural productivity. This study evaluated the potential of cokriging spatial interpolation for estimating the electrical conductivity of saturated paste extract (ECₑ), using soil electrical conductivity (EC) measured at 1:2.5 and 1:5 soil-to-water ratios. The objectives included identifying suitable scatter plot and cross-variogram models and assessing mapping accuracy. A total of 300 topsoil samples (0 to 30 cm) were collected from three salt-affected soil classes in Muang Pia Sub-district, Khon Kaen Province, Northeastern Thailand. Spatial modelling and cross-variogram analyses were performed using GS+ software to evaluate estimation accuracy across different sample sizes. The results showed that EC measurements at a 1:5 ratio exhibited the strongest correlation with across all soil classes, with coefficient of determination (R2) values reaching 0.98 in Class 1 and Class 2, and 0.85 in Class 3, despite a minimum sample size (n = 25). Gaussian and spherical models best described these relationships. Higher R2 values were consistently associated with lower mean error (ME) and root mean square error (RMSE), in almost all sample sizes and classes, indicating the robustness and reliability of the model across varying salinity conditions. Larger sample sizes (n = 100) yielded more consistent estimation performance, while smaller sample sizes maintained acceptable accuracy, particularly for EC 1:5. This study indicates that soil EC water ratios, especially 1:5, can serve as practical surrogates for ECₑ estimation using cokriging spatial interpolation. The proposed approach offers a cost-effective solution for salinity mapping in salt-affected soil areas, with implications for soil monitoring, land management, and sustainable agriculture under limited sampling conditions.]]>
