Анализ динамики уровня воды в Днестре, Припяти и реках Закарпатья

Abstract

In practice the property of fractal for hydrograph of many rivers is discovered. In connection with this feature of dynamics of water level the development of research methods for the structure of river hydrograph is actual. For fractal research of time series Hurst parameter H is used . It characterizes attitude of force of trend (determined factor) toward the level of noise (stochastically factor). The account of fractal of hydrograph will allow more exactly predicting a water level in a river, that will provide the receipt of the set indexes of quality of maintenance of hydrotechnical buildings. Research of structure of hydrograph of rivers for the exposure of it characteristic features is the purpose of this work. The following tasks decide in work: the closeness’s of distributing, autocovariance functions, power spectrums of the studied time series, fractal degree (Hurst parameter). Researches are executed in the Matlab environment. The studied time series are measuring of water level in rivers by the automated hydrological stations. In first case (river Dnestr) information is got in the Mogylev-Podol'skyy town, in the second case (the Prypyat river) – on the Lyubyaz post. The third case (river Uge) is got on a post Zarechnoy, fourth case (the Latorytsa river) – on a post in the Mukachevo town. The following results are got: 1) the studied time series do not submit to normal distribution; 2) the studied time series do not possess exponentially falling AСF, incident to the stochastic time series; 3) the studied time series are persistent and possess of long duration memory. The analysis of dynamics of water level of rivers can be perspective direction of further researches by the methods of nonlinear dynamics.