A few variants of hierarchical cluster analysis (CA) as tool of assessment of multidimensional similarity in environmental dataset
are compared. The dataset consisted of analytical results of determination of metals (Na, K, Ca, Sc, Fe, Co, Zn, As, Br, Rb, Mo,
Sb, Cs, Ba, La, Ce, Sm, Hf and Th) in ambient air dried and kept alive, by the means of hydroponics, moss baskets collected in
12 locations on the area of Tricity (Poland). The selected urban locations covered sites of various anthropogenic impact (cargo
transit, heavy traffic, apartments, shipyard, etc.). Factor analysis in the variant of Principal Component Analysis (PCA) is offered
as a validating method for CA. The results of three scaling variants are presented: column scaling, row scaling and column
scaling followed by row scaling. Moreover, the results of two linkage methods with squared Euclidean distance as the similarity
measure are compared: Ward’s linkage and complete linkage. The cluster significance is indicated basing on Sneath’s index.
Some urban pollution profiles (fertilizer plant emission, fossil fuel burning, municipal waste dump, mineral soil dust) were
identified and the possibility of their application by the environmental decision makers or managers is pointed out. The complete
environmental data exploration procedure is recommended.