Bivariate probability plots: A method for delineating different populations in soil geochemical data

• Published in: The Science of the total environment 2019-06, Vol.671, p.1047-1055
• Main Authors: Geranian, Hamid, Mokhtari, Ahmad Reza, Cohen, David R.
• Format: Article
• Language: English
• Subjects: Anthropogenic contamination; Geogenic anomaly; Irankuh; Probability plots
• ISSN: 0048-9697; 1879-1026
• DOI: 10.1016/j.scitotenv.2019.03.295

Bivariate probability plots (BPP) have been developed as an extension to univariate probability plots (UPP) to characterize and separate geochemical populations. The two approaches have been applied to Pb, Zn, As and Cd in soils collected from the vicinity of the Irankuh Pb and Zn mines to compare their capacity to separate geogenic patterns from anthropogenic contamination in nearby agricultural land. Separating populations by UPP and comparing thresholds with the Iranian regulatory limits for metal concentrations in agricultural land indicates that the mining operations and adjacent open ground areas are heavily contaminated. Separating the samples into different populations using BPP, based on typical element associations of Pb with Zn and As with Cd, demonstrates the extent of anthropogenic contamination is relatively higher for Pb and Zn than the other two elements. This is attributed to dispersion of dust from the mining site. Outside this area elevated metal concentrations appear related to geogenic factors, and which is attributed to weathering and erosion of underlying mineralized lithologies. [Display omitted] •Pb, Zn, As and Cd toxic elements are studied in soil samples near Pb-Zn mine site.•Univariate and bivariate probability plots are used to determine contamination area.•Joint probability plot is introduced as a multivariate graphical clustering method.•Geogenic and anthropogenic contamination areas are separated graphically.