渭河流域年尺度降水序列非一致性分析及重构

Analysis and reconstruction of inconsistency of annual-scale precipitation series of Weihe River Basin

受全球气候变化和人类活动的影响,流域的下垫面情况和产汇流特性发生了显著改变,破坏了水文序列的一致性,给工程水文分析带来了挑战。针对水文序列非一致性问题,为了达到工程水文分析的要求,选取渭河流域的12个站点,应用线性回归、滑动平均、Mann-Kendall趋势检验法以及Lee-Heghinian、滑动T、滑动秩和6种方法,对1960—2020年的年降水序列的趋势性、突变性进行了检验,识别了主要的变异点及影响变异的主要突变类型;提出了水文序列一致性重构方法,并对其进行了验证。结果表明:(1)渭河流域降水的趋势性并不显著,5个站点的趋势呈增加趋势,但未通过显著性检验;7个站点呈下降趋势,除华山站外均未通过显著性检验。(2)渭河流域12个站点中10个站点存在显著突变,其中8个站点的突变年份出现在1970年、1990年前后;对站点年降水序列进行突变影响因素识别,其中5个站点的突变成分是均值,4个站点是偏态系数,1个站点是变差系数,因此可得渭河流域降水的突变类型以均值突变和偏态系数突变为主。(3)经序列重构,渭河流域存在突变的10个站点中9个站点的年降水量通过了滑动秩和突变检验,验证了重构方法对均值、偏态系数等导致的突变因素具有较好的适用性。

Under the impacts from global climate change and human activities, the condition of underlying surface as well as the runoff yield and confluence characteristics of river basin are significantly changed;which destroy the consistency of hydrological series and bring challenges to engineering hydrological analysis. Aiming at the problem of the inconsistency of hydrological series and meeting the requirement for the relevant engineering hydrological analysis, 12 hydrological stations within Weihe River Basin are selected and 6 methods, i.e. linear regression, sliding average, Mann-Kendall trend test, Lee-Heghinian, sliding T and sliding rank sum are applied to test both the trend and mutation of the annual precipitation series from 1960 to 2020 and identify the main variation points and the main mutation types affecting the variation, while a method for reconstructing the consistency of hydrological series is proposed and tested herein. The results show that(1) the trend of precipitation in Weihe River Basin is not so significant, in which the trends of five hydrological stations exhibit increasing trends, but all fail to pass through the relevant significance test, while 7 hydrological stations exhibit decreasing trends, but all fail to pass through the relevant significance test except Huashan Hydrological Station;(2) 10 in 12 hydrological stations in the Weihe River Basin have significant mutations, in which the mutation years of 8 hydrological stations occurred around 1970 and 1990. After the identification of the factors affecting the mutations of the precipitation series of the hydrological stations, the values of the mutation components are the mean values for 5 hydrological stations, the skewness coefficients for 4 hydrological stations and the variation coefficient for 1 hydrological station, thus the mutation types of the precipitation in Weihe River Basin are mainly dominated by the mean value mutation and the skewness coefficient mutation;(3) through the reconstruction of the series, the annual precipitations of 9 in 10 hydrological stations with mutation in the river basin are passed through the sliding rank sum mutation test, which verifies that the reconstruction method has better applicability to the mutation factors caused by the mean value and the skewness coefficient, etc.