Quantitative source apportionment and associated driving factor identification for soil potential toxicity elements via combining receptor models, SOM, and geo-detector method

• DOI: 10.1016/j.scitotenv.2022.154721
• Publication Date: 2022-03-24T23:06:41Z
• Authors: Guanghui Guo; Kai Li; Degang Zhang; Mei Lei
• Keywords: China; 04 agricultural and veterinary sciences; 15. Life on land; Risk Assessment; 6. Clean water; Soil; Lead; 13. Climate action; Metals, Heavy; Soil Pollutants; 0401 agriculture, forestry, and fisheries; Cadmium; Environmental Monitoring

Quantitative source apportionment of soil potential toxicity elements (PTEs) and associated driving factor identification are critical for prevention and control of soil PTEs. In this study, 421 soil samples from a typical area in southeastern Yunnan Province of China were collected to evaluate the pollution level of soil PTE using pollution factors, pollution load index, and enrichment factors. Positive matrix factorization (PMF), absolute principal component score/multiple line regression (APCS/MLR), edge analysis (UNMIX) and self-organizing map (SOM) were applied for source apportionment of soil PTEs. The geo-detector method (GDM) was used to identify the driving factor to PTE pollution sources, which assisted in source interpretation derived from receptor models. The results showed that the geometric mean of As, Cd, Cu, Cr, Ni, Pb, and Zn were 94.94, 1.02, 108.6, 75.40, 57.14, 160.2, and 200.3 mg/kg, which were significantly higher than their corresponding background values (P < 0.00). Particularly, As and Cd were 8.71 and 12.75 times higher than their corresponding background values, respectively. SOM yielded four clusters of soil PTEs: AsCd, PbZn, CrNi, and Cu. APCS/MLR was regarded as the preferred receptor model for source apportionment of soil PTEs due to its optimal performance. The results of ACPS/MLR revealed that 36.64% of Pb and 38.30% of Zn were related to traffic emissions, Cr (92.64%) and Ni (82.51%) to natural sources, As (85.83%) and Cd (87.04%) to industrial discharge, and Cu (42.78%) to agricultural activities. Distance to road, lithology, distance to industries, and land utilization were the respective major driving factor influencing these four sources, with the q values of 0.1213, 0.1032, 0.2295 and 0.1137, respectively. Additionally, GDM revealed that nonlinear interactions between anthropogenic and natural factors influencing PTEs sources. Based on these results, comprehensive prevention and control strategies should be considered for pollution prevention and risk controlling.