The use of size-normalisation techniques in interpretation of soil contaminant distributions

Analyses of contaminant concentrations in soils are frequently carried out in urban regions because soils can act as both a sink and a secondary source of contaminants to the environment. Total soil concentration data are required to evaluate soil quality and the ability of soils to support life, however to determine the source and dispersion of contaminants and an accurate depiction of spatial distributions, normalised data are required. The effectiveness and economics of three normalisation methods for soils are tested in the current study. Size normalisation undertaken by physical separation of the fine fraction (<62.5 mu m) provides consistent spatial information, but is time consuming and expensive and does not capture anthropogenic contributions associated with the coarse fraction of the soil, which is not generally analysed. Elemental normalisation only utilises a single analysis of the total soil, but requires that the normalising element acts conservatively, and as a proxy for the fine fraction, which are requirements not always met. A new post-extraction normalising procedure ( PEN) also requires an analysis of the total soil only, but is not dependent on a normalising element and can be used to normalise any type of contaminant. In this approach, the residue after acid digestion is assumed to be the major diluent controlling the confounding effects of variable grain size. A strong relationship between the residue mass and soil size supports the use of this residue material as a normalising agent. Results indicate that the PEN method is equally efficient at reducing the confounding effects of variable grain size as elemental normalisation, but that size normalisation produces the most consistent spatial information and should be used in detailed studies of source and dispersion.