Abstract:AbstractThe levels of 12 heavy metals (Ag, Ba, Be, Cd, Co, Cr, Cu, Ni, Pb, Tl, V, Zn) were considered in 229 soil samples in Semnan Province, Iran. To discriminate between natural and anthropogenic inputs of heavy metals, factor analysis was used. Seven factors accounting for 90.5 % of the total variance were extracted. The mining and agricultural activities along with geogenic sources have been attributed as the main causes of the levels of heavy metals in the study area. The partial least squares regression was utilized to predict the level of soil pollution index (SPI) considering the concentrations of 12 heavy metals. The eigenvectors from the first three PLS represented more than 98 % of the overall variance. The correlation coefficient between the observed and predicted SPI was 0.99 indicating the high efficiency of this method. The resultant coefficient of determination for three PLS components was 0.984 confirming the predictive ability of this method.

The Extent and Prediction of Heavy Metal Pollution in Soils of Shahrood and Damghan, Iran