基于PCA和地统计的西南烟田土壤重金属源解析

为定性、定量地分析西南某地区烟田土壤中重金属污染来源及源贡献得分的空间分布特征,用四分法采集研究区内256个表层土壤样品,利用数理统计方法相关性分析、主成分分析(PCA)和PMF受体模型],定量分析研究区域内Cd、Hg、Ni、Pb和Zn重金属的来源;并结合地统计方法对PMF得到的三种源的成分因子贡献率进行克里格空间插值,验证数理统计源解析结果。结果表明:重金属Hg、Pb、Ni、Zn的平均值(0.19、41.99、38.43、106.66 mg·kg^-1)均未超过农用地质量标准的风险筛选值(1.80、90.00、70.00、200.00 mg·kg^-1),Cd的平均浓度(0.66 mg·kg^-1)与当地背景值(0.66 mg·kg^-1)持平,但超过了该金属对应的农用地风险筛选值(0.30 mg·kg^-1),此外,Cd和Hg的空间变异性(81.21%、84.21%)较大,在土壤中积累较明显;数理统计分析可得,五种重金属归因为三个主要来源,其中Pb、Ni和Zn主要来源于交通源,Cd主要来源于成土母质自然源,Hg主要来源于工业源;地学统计分析表明,PC1(Pb、Ni和Zn)主要积累在研究区交通量多的东部地区,PC2(Cd)分布较为均匀,PC3(Hg)主要积累在研究区工矿业密集的东北和东南方向。研究区烟田土壤中重金属整体污染程度较轻,但需要对Hg和Cd采取适当措施降低其风险,多元统计与地统计学的结果相印证土壤中重金属的来源主要是工矿业开采、交通和地质背景,受体模型与地质统计学相结合,可以有效地估计土壤中重金属的来源及分布。

Principal component analyses and geostatistical analyses to estimate source apportionment of heavy metal (loids)in tobacco-growing soils in the Southwest region of China

To qualitatively and quantitatively analyze the spatial distribution characteristics and contributions of heavy metal (loids)in tobacco-growing soil in the Southwest region, 256 topsoil samples were collected and mathematical methodscorrelation analysis, principal component analysis (PCA), and positive matrix factorization (PMF)] were used to analyze the sources of Cd, Hg, Ni, Pb, and Zn; previously obtained results were verified by geostatistical methods. The results showed the following. The average contents of Hg, Pb, Ni, and Zn (0.19, 41.99, 38.43, and 106.66 mg·kg^-1, respectively)in tobacco-growing soils were lower than their risk screening values (1.8, 90.0, 70.0, and 200.0 mg·kg^-1, respectively). For Cd, the average concentration (0.66 mg·kg^-1)was equal to the local background value (0.66 mg·kg^-1)but exceeded the risk screening value (0.30 mg·kg^-1). The variable coefficients of Cd and Hg (81.21%, 84.21%)were big, clearly indicating the accumulation of these metals. The three main sources of the metals were parent materials (Cd), transportation (Pb, Zn, and Ni), and industrial sources (Hg). PC1 metals (Pb, Ni, and Zn)were mainly accumulated in the east of the study area, which had a high volume of traffic; the PC2 metal (Cd)was distributed uniformly, and the PC3 metal (Hg)was accumulated mainly in the northeast and southeast of the study area, which have a high concentration of industry and mining operations. The degree of overall pollution of the heavy metal (loids)in tobacco-growing soils was negligible; however, it was necessary to take appropriate measures to control the risk of Hg and Cd contamination. The systematic combination of GIS with multivariate statistical analysis proved to be an efficient method for determining the anthropogenic and natural sources of heavy metal (loids).