CHEMOMETRIC CHARACTERIZATION OF SOIL DEPTH PROFILES

The application of multivariate-statistical methods to the description of the metal distribution in soil depth profiles is shown. By means o f cluster analysis, it is possible to get a first overview of the main differences in the metal status of the soil horizons. In case of anth ropogenic soil pollution or geogenic enrichment, cluster analysis was able to detect the extent of the polluted soil layer or the different geological layers. The results of cluster analysis can be confirmed by means of multidimensional variance and discriminant analysis. Methods of discriminant analysis can also be used as a tool to determine the optimum number of variables which has to be measured for the classific ation of unknown soil samples into different pollution levels. Factor analysis yields an identification of not directly observable relations hips between the variables. With additional knowledge about the orogra phic situation of the area and the probable sources of emission the fa ctor loadings give information on the immission structure at the sampl ing location.