Risk based characterisation of contaminated industrial site using multivariate and geostatistical tools.

Human and ecological risk assessment requires the sources, distribution, mobility and environmental behaviour of contaminants to be investigated on a site-specific basis. It often deals with data sets which are relatively small and affected by sampling gaps. In the case of a polycyclic aromatic hydrocarbon (PAH) contaminated industrial site, Kriging interpolation of spatial data and principal component analysis (PCA) proved useful for extracting additional value from the data set. Kriging was adopted for assessing the horizontal and vertical distribution and transport of PAHs in soil. PCA was applied to PAH concentration and relative abundance in soil samples and interpreted on the basis of the PAH physico-chemical and bio-degradation properties. It revealed correlation with the products of a neighbouring factory and the weathering of the lighter PAHs. The geo- and multivariate statistical results were coupled with the previous hydrogeological characterisation of the site to develop a site-conceptual model for use in the exposure scenario modelling for risk assessment.