Environmental risks often stem from contamination driven by chemical stressors introduced from multiple sources, either geogenic or anthopogenic. Differentiating between anthropogenic chemical anomalies and those inherent to the environment is crucial. This distinction is essential for defining feasible remediation objectives. This study applied univariate and multivariate statistical techniques to analyse geochemical data from over 7000 topsoil samples in Campania (Southern Italy), over an area of approximately 13,600 km². A key step in the methodology was applying Normal Score Transformation (NST), which stabilized the variance of the dataset, pulling the extreme outliers back to normal ranges, making it more suitable for multivariate analysis. Principal Component Analysis (PCA) was performed, and four components were selected; the spatialization of their scores revealed four primary independent sources controlling geochemical variability across the region. Specifically, two distinct volcanic districts were identified, plus a siliciclastic and an anthropogenic component. The integration of RGB composite maps further refined this differentiation, emphasising the coexistence or the predominance of one component over the other. The methodological approach demonstrated here provides valuable insights for environmental risk assessment and remediation planning in geochemically complex and anthropized regions.

The municipality of Solofra (Avellino, Italy) is a major center for the tanning industry, where long-term leather processing has generated significant environmental pressure. An abandoned industrial site used to treat organic tannery waste was found to contain illegally dumped materials. A characterization campaign conducted in 2016 involved drilling 68 boreholes to depths of about 10 meters and collecting 235 samples. Concentrations of 16 potentially toxic elements were determined, georeferenced, and processed into a three-dimensional database. A novel methodological framework combining multivariate statistics and three-dimensional geostatistical modelling was applied. Empirical Bayesian Kriging generated spatial models for individual elements, while Principal Component Analysis identified the main geochemical associations, distinguishing natural volcanic soil signatures (V, As, and Be) from anthropogenic inputs associated with tanning activities, particularly Cr, Sb, Sn, and Tl. Fuzzy Cluster Analysis applied to PCA scores further delineated two main compositional domains: a widespread cluster of tannery waste mixed with volcanic soils and a second cluster of relatively pure tannery waste enriched in chromium and associated elements. Chromium occurs predominantly as Cr(III), with Cr(VI) largely below detection limits. Its strong association with organic matter suggests the formation of stable Cr(III)–organic complexes that promote adsorption and partial immobilization within the waste matrix. The integration of multivariate statistics with three-dimensional modelling enabled reconstruction of the landfill’s internal geochemical architecture, providing a powerful tool for identifying contaminant hotspots and supporting targeted investigation and remediation in complex industrial waste sites.

The municipality of Solofra (Avellino, Italy) is a major center for the tanning industry, where long-term leather processing has generated significant environmental pressure. An abandoned industrial site used to treat organic tannery waste was found to contain illegally dumped materials. A characterization campaign conducted in 2016 involved drilling 68 boreholes to depths of about 10 m and collecting 235 samples. Concentrations of 16 potentially toxic elements were determined, georeferenced, and visualized in a three-dimensional model. A novel methodological framework combining multivariate statistics and three-dimensional geostatistical modelling was applied. Empirical Bayesian Kriging generated spatial models for individual elements. At the same time, Principal Component Analysis identified the main geochemical associations, distinguishing natural volcanic soil signatures (V, As, and Be) from anthropogenic inputs associated with tanning activities, particularly Cr, Sb, Sn, and Tl. Fuzzy Cluster Analysis applied to PCA scores further delineated two main compositional domains: a widespread cluster of tannery waste mixed with volcanic soils and a second cluster of relatively pure tannery waste enriched in chromium and associated elements. Chromium valence partitioning was inferred under assumed reducing conditions, with the data indicating Cr(VI) below detection limits. The observed correlation with organic matter is consistent with possible retention of Cr(III) via adsorption and organic complexation. However, this pattern should be interpreted as indirect evidence of inferred speciation rather than direct confirmation. The integration of multivariate statistics with three-dimensional modelling enabled reconstruction of the landfill's internal geochemical architecture, providing a powerful tool for identifying contaminant hotspots and supporting targeted investigation and remediation in complex industrial waste sites.