EN
The hydrographic network of the young glacial areas is quite specific in view of the connections of rivers and lakes, and the role they play in the shaping of runoff. The young age of this network is evidenced by the significant share of lentic segments in the river courses, periodicity, significant share taken in the watershed structure by the areas with no surface outflow, and the longitudinal gradient stili not having stabilized in the evolution. The manner of development of the hydrological network within the confines of a catchment area can be represented with the method of network analysis, using Horton-Strahler classification for this purpose. The organisation of the thus ordered basin flows’ network model is ruled, in particular, by the law of the number of flows (expressed through the bifurcation index RB) and the law of mean flow length (expressed through the indicator of average flow length RL). The shape of the river network in a basin and the number of flows forming this network are not random, and so a certain regularity appears, along with a self-similarity with respect to the shape of the fluvial networks. This observation allows to propose that the river networks have a fractal dimension, which can be determined either on the basis of the Horton indices, or with the method of boxes. The Horton analysis of selected fluvial systems (Table 1) showed that all the systems considered satisfy the Horton’s laws of flow number and mean length, that the bifurcation indices RB and RL display a regional differentiation, and that some of the quantitative characteristics, describing the river systems analysed, are correlated among themselves at the significance level of p<0.05 (Figs. 1 and 2). Horton’s model does not describe, usually, the real fluvial systems, and so calculation of the fractal dimension on the basis of the statistical Horton’s laws may lead to erroneous results. The reliable results of fractal dimensioning of the river networks can be obtained with the box method (Table 1). Despite the fact that the analysed hydrographic networks are in the initial stage of development, we can already observe regional differentiation of the dependence between the fractal dimension value of the network and the catchment’s surface area covered by the network (Figs. 4 and 5). The fractal dimension of the fluvial networks considered is significantly statistically correlated with the characteristic discharges SSQ and WWQ. In case of SNQ this dependence is not statistically significant (Fig. 3). The association between the fractal dimension of the river network and the characteristic discharges is stronger in the Masurian Lake District than in the Lithuanian Lake District (Figs. 4 and 5). This would indicate a higher degree of hydrological consistency of the river network of the Masurian Lake District, its better adaptation to channelling of runoff.