Modeling of experimental data on trace elements and organic compounds content in industrial waste dumps

The main objective of the study presented was to investigate the differences between 20 mine waste dumps located in the Silesian Region of Poland and Czech Republic, in terms of trace elements and polycyclic aromatic hydrocarbons contents. The Principal Component Analysis and Hierarchical Clustering Analysis were applied in exploration of the studied data. Since the data set was affected by outlying objects, the employment of a relevant analysis strategy was necessary. The final PCA model was constructed with the use of the Expectation-Maximization iterative approach preceded by a correct identification of outliers. The analysis of the experimental data indicated that three mine waste dumps located in Poland were characterized by the highest concentrations of dibenzo(g,h,i)anthracene and benzo(g,h,i)perylene, and six objects located in Czech Republic and three objects in Poland were distinguished by high concentrations of chrysene and indeno (1.2.3-cd) pyrene. Three of studied mine waste dumps, one located in Czech Republic and two in Poland, were characterized by low concentrations of Cr, Ni, V, naphthalene, acenaphthene, fluorene, phenanthrene, anthracene, fluoranthen, benzo(a) anthracene, chrysene, benzo (b) fluoranthene, benzo (k) fluoranthene, benzo(a)pyrene, dibenzo(g,h,i) anthracene, benzo(g,h,i)perylene and indeno (1.2.3-cd) pyrene in comparison with the remaining ones. The analysis contributes to the assessment and prognosis of ecological and health risks related to the emission of trace elements and organic compounds (PAHs) from the waste dumps examined. No previous research of similar scope and aims has been reported for the area concerned. (C) 2016 Elsevier Ltd. All rights reserved.