PRINCIPAL COMPONENT ANALYSIS AND SPATIAL CORRELATION - ENVIRONMENTAL ANALYTICAL SOFTWARE TOOLS

Using principal component analysis and geostatistics, the authors have reexamined the results of a study of chlorinated dibenzodioxins and dibenzofurans in Hamburg (Germany) Harbor sediments. These methods identified the same anomalous samples as the original study, but without recourse to the more difficultly obtained isomer-specific chemical analyses. In addition, they revealed in the remaining data a trend highly suggestive of two spatially and chemically discrete sources for nearly all (>90%) of the subject compounds found in the study, once again using no isomer-specific data. All of the above was done with readily available, public-domain software.