MULTIVARIATE-ANALYSIS OF CHEMICAL-PHYSICAL PARAMETERS TO CHARACTERIZEAND DISCRIMINATE KARSTIC WATERS

In the Karst near Trieste two contiguous, apparently similar, springs were sampled during autumnal and spring periods. Twelve chemical-physi cal parameters were monitored, with the purpose of determining the com position and verifying seasonal variations of the waters. The methodol ogical approach for characterizing and discriminating karstic freshwat ers, based on a multivariate analysis of the analytical data, as PCA ( Principal Component Analysis) or EDA (Linear Discriminant Analysis), i s alternative to intrusive methods based on chemical, radioactive or b iological tracers. A discriminant model was proposed for these karstic waters, verifying the significance of the variables.