An integrated approach to identify quantitative sources and hazardous areas of heavy metals in soils.

Identifying quantitative sources and hazardous areas of heavy metals is a crucial issue for soil management. For this purpose, an integrated approach composed of finite mixture distribution modeling (FMDM), positive matrix factorization (PMF) and sequential Gaussian co-simulation (SGCS) was proposed. FMDM was used to establish background standards and pollution thresholds. PMF supported by FMDM background standards was applied to estimate the source apportionment. Hazardous areas of single metals were delineated using SGCS with FMDM pollution thresholds and uncertainty analysis, and overall hazardous areas were defined by the presence of multiple metals. This integrated approach was applied to a dataset of seven metals as a case study. FMDM indicated that the distributions of Cr, Cu, Ni, and Zn were fitted to two-dimensional mixture distributions, representing a background distribution and a moderately polluted distribution. The distributions of Cd, Hg, and Pb were composed of a three-component lognormal mixture distribution, corresponding to the background, moderate, and high pollution distributions. Three sources were apportioned. Cr, Cu, Ni, and Zn were dominated by parent materials. Parent materials contributed 52.6%, 45.8%, and 81.9% of Cd, Hg, and Pb concentrations, respectively. Human emissions from coal combustion, industrial work and traffic had significant influences on Hg, Cd, and Pb, with contributions of 49.8%, 26.9%, and 15.6%, respectively. Agricultural practices were exclusively associated with 20.5% of Cd. Overall, hazardous areas exceeding moderate pollution thresholds covered 17.4% of the total area, corresponding to urban areas and industrial sites, whereas overall hazardous areas above high pollution thresholds were limited to 0.01% of the total area.