Font Size: a A A

Mutation Detection And Mutation Precursory Signal Of Upper Fenhe River Hydrological Time Series

Posted on:2021-06-19Degree:MasterType:Thesis
Country:ChinaCandidate:C R QinFull Text:PDF
GTID:2480306110498354Subject:Hydraulic engineering
Abstract/Summary:PDF Full Text Request
Under the influence of natural factors and human activities,the hydrological sequence is subject to mutation or even variation,resulting in frequent occurrence of a series of extreme hydrological events such as drought and flood disasters.In this paper,the precipitation and runoff sequences of four hydrological stations in the upper reaches of Fenhe River from 1956 to 2000(45 years in total)were used as the basic data of the study.First,the variation analysis of the hydrological sequence is divided into three steps.The first step: the R/S analysis method and DFA detrend analysis method are used to make a preliminary judgment on whether the hydrological sequence has variation as a whole.The Step 2: use the multi-scale linear fitting method,Lyapunov index method and sliding dynamics autocorrelation factor method to analyze in detail and detect the mutation point of hydrological sequence.The Step 3: comprehensively analyze the variation of hydrological sequence from three aspects: time scale,variation type and variation degree.Secondly,the contribution rate of precipitation and human activities to the variation of runoff sequence was quantitatively analyzed by using the slope change rate comparison method of precipitation and runoff accumulation volume.Finally,the evolution law of four characteristic variables(system recovery rate,recovery force,variance and autocorrelation factor)of critical slowing of hydrological sequence system was described to study the early warning signals of hydrological sequence mutation.The main research contents and achievements are as follows:(1)Preliminarily judge whether there is variation in the hydrological sequence.The Hurst index value and scale index R value of the annual precipitation and annual runoff sequence were both greater than 0.5 by R/S analysis and DFA detrend analysis,indicating that the annual precipitation and annual runoff sequences of the four hydrological stations had overall variation.(2)Detailed detection of hydrological sequence mutation points.By using the multi-scale linear fitting method,Lyapunov index method and sliding dynamics autocorrelation factor index method to detect the hydrologicalsequence mutation points in detail,we obtained that the mutation point of the annual precipitation sequence of Shangjingyou station was in the late 1970 s,and the mutation point of the annual runoff sequence was in the late 1970 s and early1980 s.The abrupt mutation point of annual precipitation sequence of Fenhe reservoir station,Zhaishang station and Lancun station was around 1960,and the abrupt mutation point of annual runoff sequence was around 1966.(3)The variation of runoff sequence under different time scales was analyzed synthetically.According to the comprehensive analysis on the annual scale,only trend variation occurred at the mutation point of Shangjingyou station,Fenhe reservoir station and Zhaishang station,while compound variation occurred at the mutation point of Lancun station.According to the comprehensive analysis on the month-by-month scale,it was found that the trend variation occurred at the mutation point of Shangjingyou station and Fenhe reservoir station,the mean variation occurred at the mutation point of Lancun station,and the degree of variation was not obvious at Zhaishang station.According to the comprehensive analysis on the wet-dry scale,it was found that the trend variation occurred at the mutation point of the Shangjingyou station,the compound variation occurred at the Fenhe reservoir station,and the degree of variation was not obvious at Zhaishang station and Lancun station.During the dry season,trend variation occurred in Shangjingyou station and Fenhe reservoir station,while mean variation occurred in Zhaishang station and Lancun station.According to the comprehensive analysis on the quarterly scale,it was found that the trend variation occurred in the first,third and fourth quarters of the station,and the degree of variation was not significant in the second quarter.The degree of variation was not significant in the first and second quarters of Fenhe reservoir station,trend variation occurred in the third quarter,and compound variation occurred in the fourth quarter.The mean value variation occurred in the first quarter,and the trend variation occurred in the second,third and fourth quarters.In the first quarter of Lancun station,mean variation occurred,trend variation occurred in the third quarter,and compound variation occurred in the second and fourth quarters.(4)Attribution analysis of runoff time seriesThe slope change rate comparison method of precipitation and runoff accumulation volume was used to quantitatively analyze the contribution rate of precipitation and human activities to the variation of runoff time series.It was found that precipitation was the main factor affecting the mutation of runoff sequence in Shangjingyou station,and human activities were the main factor affecting the mutation of runoff sequence in Fenhe reservoir station,Zhaishang station and Lancun station.(5)Study on mutation precursor signals based on critical slowing downBased on the critical slowing down phenomenon principle,according to the evolution law of the recovery rate,recovery force,variance and autocorrelation coefficient of the hydrological sequence system,the mutation precursor signals of the annual precipitation and annual runoff sequence are obtained in advance of the actual mutation years within ten years,which has a reliable warning effect on the mutation.
Keywords/Search Tags:Fenhe River upstream, hydrologic time series, mutation detection, mutation analysis, the early precursor warning signals
PDF Full Text Request
Related items