Characterization and source apportionment of heavy metal pollution in soil around red mud disposal sites using absolute principal component scores-multiple linear regression and positive matrix factorization models
Wenwen Cui,Xiaoqiang Dong,Jiajiang Liu,Fan Yang,Wei Duan,Mingxing Xie
DOI: https://doi.org/10.1007/s10653-024-02267-x
2024-11-09
Environmental Geochemistry and Health
Abstract:In recent years, industrial waste and agrochemicals have reduced soil fertility and productivity, significantly impacting food security and ecosystems. In China, areas near red mud deposits from the aluminum industry show severe heavy metal contamination. This study examines agricultural soil near a red mud site in Shanxi Province, analyzing Cd, Cr, Hg, Ni, Pb, As, Cu, and Zn levels and distribution. Geostatistical methods and GIS are utilized to assess heavy metal pollution using the single factor index, the Nemerow integrated index, and the Hakanson potential ecological risk index. Absolute Principal Component Scores-Multiple Linear Regression (APCS-MLR) and Positive Matrix Factorization (PMF) models are used for quantitative analysis of pollution sources. Research indicates that the average concentrations of eight heavy metals exceed the natural background values of Shanxi, placing them at a severe pollution level with moderate ecological risk. Specifically, indices for As, Pb, and Cr are 3.79, 3.38, and 3.26, indicating severe pollution; Cd, Cu, and Hg at 2.36, 2.62, and 3.00 suggest moderate pollution; Ni at 1.87 shows mild pollution, while Zn at 0.97 is not polluted. Hg presents the highest ecological risk with a coefficient of 120.00, followed by Cd (70.69) and As (37.92). Spatial analysis shows significant correlations among Pb, Zn, Cu, and Ni, while Cr, Cd, Hg, and As show greater variability and weaker correlations. Both models identify five main sources: industrial activities, agricultural fertilizers, red mud leachate, energy combustion, and natural geological backgrounds, with respective contribution rates in the APCS-MLR model at 27.7%, 24.6%, 18.1%, 15.2%, and 14.4%, and in the PMF model at 29.2%, 21.5%, 16.9%, 16.7%, and 15.7%. This study offers a scientific basis for controlling soil pollution in the region, filling a literature gap.
environmental sciences,engineering, environmental,water resources,public, environmental & occupational health