Environmental forensic approach towards unraveling contamination sources with receptor models: A case study in Nakdong River, South Korea

Jung, Jihyeun; Park, Jun-Young; Choi, Yongju; Choe, Jonggeun; An, Jinsung; Nam, Kyoungphile

doi:10.1016/j.scitotenv.2023.164554

Cited by 7 publications

(1 citation statement)

References 70 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As far as R 2 was concerned, the Unmix model has the highest reliability, followed by APCS-MLR and PMF. In terms of their respective diagnostic parameters, the KMO and Bartlett test results of the APCS-MLRU model show that the KMO value was 0.771, meeting the requirement of > 0.6, and the P value was 0.000, meeting the requirement of P < 0.001 30 , which was suitable for APCS-MLR analysis (Fig. 4 a).…”

Section: Resultsmentioning

confidence: 99%

Source identification and driving factor apportionment for soil potentially toxic elements via combining APCS-MLR, UNMIX, PMF and GDM

Gong,

Xia,

Lan

et al. 2024

Sci Rep

View full text Add to dashboard Cite

The contamination and quantification of soil potentially toxic elements (PTEs) contamination sources and the determination of driving factors are the premise of soil contamination control. In our study, 788 soil samples from the National Agricultural Park in Chengdu, Sichuan Province were used to evaluate the contamination degree of soil PTEs by pollution factors and pollution load index. The source identification of soil PTEs was performed using positive matrix decomposition (PMF), edge analysis (UNMIX) and absolute principal component score-multiple line regression (APCS-MLR). The geo-detector method (GDM) was used to analysis drivers of soil PTEs pollution sources to help interpret pollution sources derived from receptor models. Result shows that soil Cu, Pb, Zn, Cr, Ni, Cd, As and Hg average content were 35.2, 32.3, 108.9, 91.9, 37.1, 0.22, 9.76 and 0.15 mg/kg in this study area. Except for As, all are higher than the corresponding soil background values in Sichuan Province. The best performance of APCS-MLR was determined by comparison, and APCS-MLR was considered as the preferred receptor model for soil PTEs source distribution in the study area. ACPS-MLR results showed that 82.70% of Cu, 61.6% of Pb, 75.3% of Zn, 91.9% of Cr and 89.4% of Ni came from traffic-industrial emission sources, 60.9% of Hg came from domestic-transportation emission sources, 57.7% of Cd came from agricultural sources, and 89.5% of As came from natural sources. The GDM results showed that distance from first grade highway, population, land utilization and total potassium (TK) content were the main driving factors affecting these four sources, with q values of 0.064, 0.048, 0.069 and 0.058, respectively. The results can provide reference for reducing PTEs contamination in farmland soil.

show abstract

Section: Resultsmentioning

confidence: 99%