Analysis Of Storm Flood Variation Characteristics In Daqing River Basin

The condition of the underlayer in Haihe river basin has changed dramatically during past few years. This affected runoff and flood accordingly. Due to the complexity of hydrological phenomenon, the paper mainly employed mathematical statistics method, the theory of stochastic hydrology and information entropy method to describe hydrological phenomenon and to analyze storm flood characteristics quantitatively in Daqing river basin after the preparatory work of collecting rainstorm flood data in stations of Fuping, Zijingguan, Longmen, Hengshanling and Chengtouhui. The main work and fruit are as follows:(1) It did primary analysis of parameters representing the characteristics of rainstorm flood: Through calculation, it got numerical value of period rainstorm,period flood volume and flood peak, then it got the value of parameters including mean value, CV, percentage of direct runoff, duration of flood and lag time, which represented the characteristics of rainstorm flood. The results showed: the items of percentage of direct runoff and duration of flood didn't change too much, period rainstorm almost kept the same, whereas the value of period flood volume and flood peak dropped dramatically.(2) It analyzed hydrological time series of rainstorm flood: Adopting hydrological time series analysis methods ,the paper did trend analysis, aberrance point diagnosis, periodicity analysis to period rainstorm,flood volume and flood peak . It applied R/S analysis method in fractal theory to identify aberrance point and forecast the future trend. In the mean time, based on the fruit of spectral analysis, this paper used the maximum entropy method to do a periodicity analysis and achieved reasonable results.(3) The results showed: compared with years before1984, years after 1984 witnessed a downward trend of storm flood, among which the period flood volume and flood peak showed a much more obvious trend of decline ; The aberrance points of precipitation rain fall,flood volume and flood peak showed coherence around 1963; Using methods of harmonic wave, power spectrum estimation and maximum entropy, it was found that the series had the cycle of 13 or 22 years in large scale and the cycle of around 3 years in small scale.