Source apportionment of heavy metals in sediments and soils in an interconnected river-soil system based on a...

Heavy metal pollution has been a global concern and key points of environmental pollution prevention and control due to the growing problems of urbanization and industrialization. Rapidly and correctly apportioning sources of heavy metal is still a great challenge because of the stability of source fingerprint and complex interaction of multiple contaminants and sources. In this study, we perform a combination of optimization of pollution source fingerprint and source apportionment through jointly utilizing two machine classification and screening methods for characterizing the pollution sources of heavy metal in the sediments of an urban river and its surrounding soils. Dominance-based rough set model (DRS), content optimization tools, and multivariate curve resolution-alternating least squares model (MCR-WALS) were employed to screen representative pollution source samples, optimize pollution source fingerprint, and apportion the potential sources of heavy metals, respectively. Further, Support vector machine (SVM) was adopted to correspondence analysis results and pollution fingerprint based on the factor characteristics for achieving source apportionment accurately. Results showed that the pollution source pollution source fingerprints optimized by DRS and optimization tools are more representative and stable, and the results obtained by SVM and MCR-WALS are more accurate comparing with traditional methods. As whole, source apportionment suggested that printing and dyeing, chemical, electroplating, metal processing were the main origins of heavy metals in this area and the proportions of them in sediment and soil pollution sources were 67.05% and 28.43%, respectively. Besides, coal combustion was also the main sources of heavy metal pollution in soils, accounting about 34.16%. Results of the study can advance our knowledge to better understand the characterization of heavy metal pollution in the peri-urban ecosystem and to design effective targeted strategies for reducing heavy metal pollution diffusion.