海流兔河流域径流变化的周期和趋势分析

以荒沙草滩区典型河流海流兔河为例,利用连续小波变换、多时相趋势分析,并联合利用离散小波变换和Mann-Kndall秩次检验法对1957—2011年的年均径流分别进行周期、趋势以及趋势结构分析。结果表明:1海流兔河年均径流量在时间序列上主要存在着12~21 a周期变化,第一主周期为17 a,且径流量偏少阶段将从2010年持续至2016年左右,之后将进入偏多阶段。2海流兔河流域径流量呈现显著下降趋势,与整个黄河中游总体径流量减少趋势相一致。3联合利用离散小波变换和Mann-Kendall秩次检验法可以很好地解释趋势的基本结构,8~16 a的周期性事件对海流兔河径流的趋势影响最大。水文时间序列周期部分的周期性越强,其对序列趋势的影响越大。4序列趋势的显著性和方向取决于其长度以及在整个序列中的位置,主要是因为序列中长周期丰枯交替的存在,多时相趋势分析一定程度上解决了由于序列长周期丰枯交替的存在造成的序列趋势判断的不确定性问题。

Period and trend analysis of streamflow change of Hailiutu River Basin

In this study, the Hailiutu River Basin was chosen as a typical case of a semi-arid sand and grass land area, and the methods of continuous wavelet transform, multi-temporal trend analysis, and discrete wavelet transform and Mann-Kendall test were jointly used to investigate the periods, trends, and basic structure of the trends of the annual mean streamflow in the basin during the period from 1957 to 2011. The results are as follows: (1)There is a main period change in the process of the annual mean streamflow, with a period varying from 12 to 21 years. The first main period is 17 years and the dry period will last from 2010 to 2016 or so, and then the Hailiutu River Basin will enter a wet period. (2)The streamflow of the Hailiutu River Basin exhibits a significantly downward trend, which is consistent with the overall trend of streamflow in the middle reaches of the Yellow River. (3)The results of joint utilization of the discrete wavelet transform and Mann-Kendall test, which can explain the basic structure of the trends, suggest that 8 to 16 years of periodic events account for the observed trend of the annual mean streamflow in the Hailiutu River Basin. The stronger the periods presented in the hydrologic time series are, the greater the influence that the period components exert on the trends observed in the series will be. (4)The trend magnitude and direction depend on the length of the time series and the position of the sub-series within the whole series, because of the existence of the long-term alternating dry and wet periods that probably lead the trend test toward misleading results, and the multi-temporal trend test provides a good way to solve that problem.