RÉSUMÉ
Information about soil properties helps the farmers to adopt e?ective and e?cient farming practices, which can increase higher yields with optimum usage of farm resources. An attempt has been made in this paper to predict soil properties using geospatial kriging approaches. This study mainly focuses on predicting soil pH using different kriging methods. Soil pH dramatically affects many other soil processes, such as nitrification and denitrification, mineralization, precipitation, and dissolution of soil organic matter. Total of seven kriging semivariogram models, namely spherical, circular, exponential, Gaussian, and linear, while two models of universal kriging, such as linear with linear drift and linear with quadratic drift, have been taken to interpolate the spatial soil pH. The performances of these entire models have been validated using mean error, and root mean square error. Spatial analysis revealed that Universal kriging outperformed ordinary kriging with less mean error and root mean square error, 0.016 and 0.52, respectively. The spatial analysis of soil mapping can be instrumental in adopting real-time and on-the-go soil precision practices.