“…In order to determine the similarities between the relationships between the selected morphometric features of the reservoirs, the average temperature of the water retained in them for the duration of ice phenomena occurrence (at a depth of 50 cm), the average thickness of the snow cover deposited on the ice (independent variables) and the average and maximum thickness of the ice (dependent variables) during the three winter seasons, the analysis of the discrimination function, principal components analysis (PCA) and the canonical analysis (RDA) were applied (Jolliffe, 1986;Krzanowski, 2000;Zuur et al, 2007;Topolski and Topolska, 2019;Topolski, 2019Topolski, , 2020a. Redundancy analysis (RDA) is the canonical form of principal component analysis (PCA) and is a method for reducing the dimensionality of multivariate data by introducing a Origin (G, dyke-type; N, subsidence bowl; P g , polygenetic; P e , former extraction pit; S, artificial bowl; Z, dammed-lake).…”