Mapping superficial and subsurface conditions play an important role in analysis and design of geotechnical structures and facilities. The mechanical parameters of sandy clays soil have significant spatial variability or heterogeneity due to the complex deposition process of soil, leading to the high uncertainty of the quantifications of its parameters. Therefore, understanding the spatial variability of the parameters is an important approach to reduce uncertainty. This paper deals with the characterization of spatial variability of soil interface in Japoma-Cameroon, using standard penetration test, cone penetration tests and pressuremeter test data. The vertical and horizontal variability was taken into account, and the data are analyzed within a heterogeneous layer using Bayesian kriging and random field theory, probability density function, coefficient of variation (COV) and scale of fluctuation (theta v). The study reveals that lognormal and normal distributions, respectively, fit the histograms of NSPT, qc, Em and Pl, with all parameters increasing with depth. The COV values, ranging from 30.10 to 88.75%, reflect significant site heterogeneity. Parameters N, qc and Em are adjusted by the exponential cosine autocorrelation function, while Pl uses the exponential sinusoidal function. Fluctuation scales theta v values vary from zero to 1.98 m vertically, with horizontal scales extending up to hundreds of meters, indicating the influence of geological anisotropy. Semivariograms show limitations in small-scale structure capture, potentially overestimating the nugget effect. The exponential model's predictions are reliable for site microzonation, subsoil bearing capacity analysis and other geotechnical parameters based on NSPT.

来源平台：INDIAN GEOTECHNICAL JOURNAL