The application of Principal Component Analysis to reveal the contributions of various natural and anthropogenic sources to the chemical composition of soil.

Environmental risks are often linked to contamination processes driven by various chemical stressors introduced into the environment from specific sources. These sources may be anthropogenic, stemming from human activities, or natural, associated with the geolithological context and geological weathering processes. It is crucial to distinguish between chemical anomalies resulting from anthropogenic inputs and those arising from the inherent compositional characteristics of the natural environment, which may not be remediable. This differentiation is essential for establishing reliable and practical remediation objectives.

When anthropogenic activities release waste products into the environment in airborne, liquid, or solid forms, these materials typically possess distinct chemical compositions. Such compositions frequently involve associations of substances that can disrupt environmental equilibria. This study employed both univariate and multivariate statistical techniques to analyze geochemical data from over 7,000 topsoil samples collected in the Campania region of Southern Italy. The objective was to develop an operational model for assessing environmental risks by identifying their sources. The database encompasses the concentrations of 52 chemical elements for each sample, with data georeferenced to facilitate spatial analysis, delineate geochemical patterns, and correlate anomalies with known human or natural sources.

The complex geological setting of the Campania region, combined with the diverse sources of anthropogenic contamination, rendered Principal Component Analysis (PCA) an especially effective method for identifying element associations that predominantly influence the variability of the geochemical data. PCA was conducted using a selection of 21 variables, resulting in the identification of four significant principal components (PCs) that account for the majority of the observed data variability:

- PC1 (42% of total variance), characterized by Th, Be, As, U, V, and Bi.

- PC2 (16% of total variance), characterized by Sb, Zn, Hg, Pb, Sn, and Cd.

- PC3 (10% of total variance), characterized by Mn, Ni, Cr, and Co.

- PC4 (9% of total variance), characterized by Ba, Cu, and Sr.

The scores of the components for each sample were spatially plotted and classified to enhance their interpretability. The legend for the component scores was centered at zero, indicating the minimal contribution of the covered areas to overall component variability; higher absolute values identified areas where the featured elemental association was more significant.

The analysis effectively differentiated soils predominantly influenced by natural contributions, such as loose materials from the volcanic centers of Campania (e.g., Mt. Roccamonfina, Campi Flegrei, and Mt. Somma-Vesuvius) (PC1 and PC4) and weathering products from the region's siliciclastic deposits (PC3). Furthermore, the PCA found areas subjected to considerable anthropogenic pressure concerning the association of Sb, Zn, Hg, Pb, Sn, and Cd (PC2). These findings underscore the effectiveness of multivariate analysis, particularly PCA, in discriminating between geogenic and anthropogenic processes and, further, in distinguishing among various anthropogenic sources. This methodological approach offers valuable insights for managing environmental risks and prioritizing remediation efforts.