Spatial Prediction of Soil Index Properties Using GIS and Empirical Bayesian Kriging

The purpose of this study is to assess the possible use of Empirical Bayesian Kriging (EBK) combined with Geographic Information Systems (GIS) to map and analyze geotechnical index properties in Thi Qar Province in southern Iraq. The aim of this objective is to describe the spatial variability of soil limits of the consistency and define areas with expansive soils, which may influence infrastructure development. Data on 550 boreholes and 862 observations per soil property, including Liquid Limit (LL), Plastic Limit (PL), and Plasticity Index (PI), were analyzed. To test the predictive accuracy of the EBK model and thus assure its statistical validation, RMSE, MSD, RMSSD, and correlation coefficients were used to test the model. The findings show that the LL was between 32% and 69%, the PL between 9% and 36%, and the PI between 1% and 39%, with most of the soils being CL and CH, which signifies moderate-high plasticity. The results indicate that there are good spatial patterns, and plasticity is more dense in the north and central areas. The originality of this work is the use of EBK to create detailed digital soil maps of a semi-arid area, where the available geotechnical data is sparse, which is used to form a dependable base to support engineering design, land-use planning, and regional geotechnical modelling.