A comparison of fractal methods and probability plots in identifying and mapping soil metal contamination near an...

Mining activities may contribute significant amounts of metals to surrounding soils. Assessing the potential effects and extent of metal contamination requires the differentiation between geogenic and additional anthropogenic sources. This study compares the use of conventional probability plots with two forms of fractal analysis (number-size and concentration-area) to separate geochemical populations of ore-related elements in agricultural area soils adjacent to Pb-Zn mining operations in the Irankuh Mountains, central Iran. The two general approaches deliver similar spatial groupings of univariate geochemical populations, but the fractal methods provide more distinct separation between populations and require less data manipulation and modeling than the probability plots. The concentration-area fractal approach was more effective than the number-size fractal and probability plotting methods at separating sub-populations within the samples affected by contamination from the mining operations. There is a general lack of association between major elements and ore-related metals in the soils. The background populations display higher relative variation in the major elements than the ore-related metals whereas near the mining operations there is far greater relative variation in the ore-related metals. The extent of the transport of contaminants away from the mine site is partly a function of the greater dispersion of Zn compared with Pb and As, however, the patterns indicate dispersion of contaminants from the mine site is via dust and not surface/groundwater. A combination of geochemical and graphical assessment, with different methods of threshold determination, is shown to be effective in separating geogenic and anthropogenic geochemical patterns.