PMF receptor models and sequential Gaussian simulation to determine the quantitative sources and hazardous areas of potentially toxic elements in soils

Abstract

Identifying the quantitative sources and hazardous areas of potentially toxic elements (PTEs) is a pivotal role of soil management. A total of 1465 samples (0–20 cm) were collected from topsoils in Dongying City, eastern China, and the concentrations of As, Cd, Co, Cr, Cu, Hg, Mn, Ni, Pb, and Zn were measured. US-EPA positive matrix factorization (EPAPMF) and weighted alternating least squares positive matrix factorization (WALSPMF) were applied to evaluate the source apportionment of ten PTEs in soils. Hazardous areas of ten PTEs were delineated using sequential Gaussian simulation (SGS) with uncertainty analysis. Both EPAPMF and WALSPMF modeling identified three source factors with similar contributions, suggesting that WALSPMF could provide a reliable PMF solution for source apportionment of PTEs in soils. Ensemble-average source contributions were calculated based on the two PMF results. As, Co, Cr, Cu, Mn, Ni, and Zn originated from natural sources, with the contributions of parent materials ranging from 85.3% to 93.2%. Hg originated equally from parent materials (47.2%) and atmosphere deposition from human emissions (45.0%). Parent materials and atmosphere deposition from human emissions contributed to 68.7% and 20.5% of Pb concentrations. Cd was mainly controlled by parent materials (61.0%) and agricultural practices (37.4%). SGS indicated that parent materials controlled the spatial distribution patterns of As, Cd, Co, Cr, Cu, Mn, Ni, Pb, and Zn. Critical probability of 0.9 could generate a reliable joint probability of PTEs concentrations exceeding pollution thresholds of 1.5 times background values in Shandong Province, and was adopted to determine confidential hazardous areas. Approximately 4.5% and 1.6% of the total area for Cd and Hg were determined as hazardous areas.