BOLBOT H., LUKIANETS O., GREBIN V. STRUCTURE OF THE TIME SERIES OF THE ANNUAL WATER RUNOFF OF THE RIVERS OF THE SIVERSKYI DONETS RIVER BASIN BASED ON THE STOCHASTIC ANALYSIS OF ITS LONG-TERM FLUCTUATIONS

To detect the cyclic component in time series of annual water runoff of rivers of the Siverskyi Donets River Basin structure, it is necessary to have a hydrological gauge, which closes a large river basin and has long continuous observations of water runoff. The only hydrological gauge that meets these conditions is the Siverskyi Donets – Lysychansk – the catchment area is 52,400 km2 and the beginning of observations of water runoff since 1892. The Siverskyi Donets – Lysychansk is quite intermittent. Analysis of annual water runoff data of “neighboring basins” for the Siverskyi Donets Basin showed that the longest series of continuous observations has a hydrological gauge the Desna River near Chernihiv – since 1895, the catchment area is 81400 km2. So, we can consider this basin as basic for calculations and determination of patterns of long-term variability of annual water runoff of rivers of the Siverskyi Donets Basin. The study used data from eight hydrological gauges: the Desna – Chernihiv, intermediate river basins – the Sula – Lubny, the Psel – Zapsillya, the Vorskla – Kobeliaky, within the study basin – the Siverskyi Donets – Lysychansk, the Bakhmut – Siversk, the Aidar – Novoselivka, as well as the Southern Bug – Oleksandrivka. To confirm the spatial consistency of the annual water runoff of the studied rivers, a correlation matrix between the time sequences of the water runoff of neighboring basins was determined. To obtain a generalized characteristic of long-term water runoff variability of the studied rivers, chronological graphs of changes in modular coefficients were constructed, initial time series smoothing was performed, graphical analysis of difference integral curves was applied. In the course of the research regularities in long-term variability of average annual water runoff of rivers of the Siverskyi Donets Basin were identified, cyclic component was identified, duration and nature of water runoff cycles and within them low and wet phases were identified. The prediction of annual water runoff in the near future was done. Joint autocorrelation and spectral analysis allowed to identify a mutually confirmed cycle lasting 19-24 years for the rivers of the Siverskyi Donets Basin. According to the criterion of series, it is established that the periods of low phase can be 9±2 years. Knowing the length of the cycles, water-specific phases were identified. As a result, for the rivers of the Siverskyi Donets Basin it was found that the runoff in the wet phase exceeds the norm by an average of 18%, and the runoff in the low phase is lower by an average of 17% than normal. Therefore, the difference in river water phases is ≈ 35%. Forecast estimates show that by 2020±2, the rivers of the basin will have a low phase, which began in 2008. A wet phase is expected from 2020±2 to 2029±2, then in the period 2030±2 – 2038±2 years – low phase. In the period from 2039±2 to 2048±2, we should expect an increase of annual water runoff of the rivers of the Siverskyi Donets Basin.