A statistical and support vector machine approach for assessing heavy metal contamination in soil, dust and rainwater from the Arak city, Iran

This investigation was performed to provide heavy metal signatures of urban soils, dusts, rainwater and to evaluate potential sources in Arak, Markazi Province. Thirty-one samples of urban soils, 29 dust and16 rainwater samples were collected in Arak city. Then, Pb, Cu, Zn, As and Ni concentrations were determined by using atomic absorption spectrophotometry. Application of discriminant analysis, correlation matrix, cluster analysis and support vector machines showed that Pb, Cu, As and Ni in soils, dust and rainwater have similar source and Zn has different source. Contaminations in soils, dusts, rainwater were classified as enrichment factor (EF). EF values indicate that heavy metal pollution levels of soils, dusts, rainwater collected from industrialization sites are greater than those of distal parts of industrialization. Spreading of hazardous wastes from industrial activities in the study area is the main sources of soils, dusts, rainwater pollution. In addition, agricultural wastes and traffic-related metal pollution is also observed.