Handling Three Types of Correlation in Contaminated Land Site Assessment

This paper outlines the three types of correlation that will characterise most attempts to model land contamination and optimise site assessment strategies. The first type of correlation is that which exists between indicators of contamination. A potential methodology for handling these correlations is given. The second type of correlation is the spatial (auto) correlation between different locations on a site. This has been previously handled through methods such as Kriging interpolation. The possible benefits of integrating the analysis of the first two types of correlation are very briefly discussed. The third type of correlation is that which exists between the information from site samples. Examples of this type of correlation are given.