应用滑动秩和检验法与Mann-Kendall方法对渭河干支流主要的7个水文站点的水文序列变异点的位置进行诊断与识别。结合Hilbert-Huang变换，对各水文站点数据进行EMD多尺度分解，获得不同时间尺度的IMF信号和趋势过程，然后采用滑动秩和检验法对各分解信号进行变异识别，确定变异的响应尺度。最后，结合水文变异指标(IHA)，以变异点位置为依据，进行阶段划分，识别变异前后水文序列的变异领域，对水文情势变异进行细节解析，找出发生强烈变异的水文指标。结果表明：1) 渭河干流水文站点年径流序列的变异点主要发生在1971年和1994年，1971年的水文变异主要由人类活动驱动，而1994年的水文变异则由气候变化与人类活动共同影响。2) 水文变异主要影响渭河干支流径流序列的趋势过程，只有华县站6年时间尺度变化及状头站4~6年时间尺度变化过程受到水文变异的扰动。3) 发生变异最强烈的地方主要是渭河的中游，即林家村和咸阳站，综合变异指数超过0.66，属强变异；上游次之，下游与两条支流更次，皆属于中等变异。4) 渭河上游的强水文变异范围主要集中在枯水季月径流、最大流量、低脉冲次数与延时、下降率与逆转次数等方面，中游强变异领域较宽，包括丰枯月径流、最大最小流量、高脉冲次数、上升下降率等，下游则反映在汛前以及丰水期月径流、最大最小流量以及下降率与逆转次数方面。渭河支流中，泾河的水文强变异主要在最小流量、最大流量、基流以及最大流量发生时间，而北洛河则侧重于最小流量、基流、低脉冲延时与逆转次数等方面。综合以上结果，可得出如下结论，渭河干支流在水文变异较为强烈，人类活动和气候变化是其重要致因。水文变异范围较宽，空间分布的差异性大，变异程度也各不相同，主要集中于枯季径流、最大流量、变化率等方面。渭河水文变异的驱动力除在长程趋势上对年径流量产生向上或向下的扰动外，对华县和状头站径流的周期变化特征也产生了影响，未来某一些周期变化特征可能会因为变异驱动而改变或消失。
Hydrological Variation Analysis in Wei River Basin
Moving Mann-Whitney U test is an excellent method to identify the locations of hydrologic variation points. In this paper, it was used to look for these points from annual runoff series at seven main hydrologic stations in the Wei River basin. Mann-Kendall method was applied here, as a validation method. Moreover, for understanding the more detailed impact from hydrologic variation, Hilbert Huang transform was also introduced to do empirical mode decomposition of annual runoff series on these hydrologic stations. Through the process, several intrinsic mode function (IMF) signals with different time-scale characteristics and trend signal were obtained on every station. Then, by using the moving Mann-Whitney U test to identify the variation or not on IMF signals, we could confirm the impact from hydrologic aviation on a certain time-scale. Finally, based on the above hydrologic variation points, the hydrologic series can be divided into two or three stages. Selecting the stage before this variation point as reference, and the stage after it as evaluating series, then we can evaluate the variation scope and level after variation point relative to the before by range of variability approach (RVA). The result indicated that 1) hydrologic variation of annual runoff series in Wei River mainly took place in 1971 and 1994; hydrologic variation in 1971 was mainly driven by human activities, and the variation in 1994 was affected by climate change and human activities. 2) Hydrologic variation mainly affected the trend element of runoff series in the Wei River, and had hydrologic disturbance on 6-year time-scale change at Huaxian station and 4 - 6-year time-scale change at Zhuangtou station. 3) The strongest hydrologic variation was mainly on the middle reaches of the Wei River, i.e. Linjiacun and Xianyang stations, with a comprehensive variation index exceeding 0.66 which belonged to strong variation. The upstream reaches came second, two tributaries and the downstream took the third place, and they all belonged to medium variation. 4) On the upstream Wei River, the strong hydrological variation indexes mainly concentrated in the dry season monthly runoff, maximum flow and low pulse frequency and duration, the fall rate and reverse, etc.; Hydrologic index field with strong variation on the middle reaches was greatly wide, including monthly runoff in dry and wet seasons, maximum and minimum flow, high pulse frequency, etc.; Some strong variation indexes were reflected in the downstream, such as monthly runoff before flood and in wet season, maximum and minimum flow, fall rate and reverse. In two tributaries of the Wei River, strong hydrologic variation in Jing River was mainly on the minimum and maximum flow, base flow and maximum flow date, while in Beiluo River, they focused on minimum flow, base flow, low pulse duration and reverse, etc. Synthesizing the above results, it can be concluded that hydrological variation in the Wei River is relatively strong, for which human activities and climate change are two important causes. Hydrologic variation index range is greatly wide, with significant difference on the spatial distribution and non-identical variation degree on every district. The strong variation indexes in the whole Wei River mainly concentrated in the dry flow, maximum flow and rate of change, etc. The force driving the hydrological variation would take some disturbance on the runoff cycle change characteristics at Huaxian and Zhuangtou stations, in addition to mainly an impact on its long-term trend. In the future, these cycle characteristics may alter or disappear because of the variation driving.
文章引用: 张洪波 , 顾 磊 (2014) 渭河流域水文变异识别与初步解析。 水资源研究， 3， 1-8. doi: 10.12677/JWRR.2014.31001
谢平, 陈广才, 雷红富等. 变化环境下地表水资源评价方法[M]. 科学出版社, 北京, 2009.
XIE Ping, CHEN Guangcai, LEI Hongfu, et al. The assessment methods of surface water resources in changing environments. Science Press, Beijing, 2009. (in Chinese)
谢平, 陈广才, 夏军. 变化环境下非一致性年径流序列的水文频率计算原理[J]. 武汉大学学报(工学版), 2005, 38(6): 6-9.
XIE Ping, CHEN Guangcai and XIA Jun. Hydrological frequency calculation principle of inconsistent annual runoff series under changing environments. Engineering Journal of Wuhan University, 2005, 38 (6): 6-9. (in Chinese)
王孝礼, 胡宝清, 夏军. 水文时序趋势与变异点的R/S分析法[J]. 武汉大学学报(工学版), 2002, 35(2): 10-12.
WANG Xiaoli, HU Baoqing and XIA Jun. R/S analysis method of trend and aberrance point on hydrological time series. Engineering Journal ofWuhan University, 2002, 35(2): 10-12. (in Chinese)
谢平, 陈广才, 雷红富, 武方圆. 水文变异诊断系统[J]. 水力发电学报, 2010, 29(1): 82-91.
XIE Ping, CHEN Guangcai, LEI Hongfu and WU Fangyuan. Hydrological alteration diagnosis system. Journal of Hydroelectric Engineering, 2010, 29(1): 82-91. (in Chinese)
张一驰, 周成虎, 李宝林. 基于Brown-Forsythe检验的水文序列变异点识别[J]. 地理研究, 2005, 24(5): 741-748.
ZHANG Yichi, ZHOU Chenghu and LI Baolin. Brown-Forsythe based method for detecting change points in hydrological time series. Geographical Research, 2005, 24(5): 741-748. (in Chinese)
 HONG H., HU B. Q., et al. Change-point analysis of hydrological time series using grey relational method. Journal of Hydrology, 2006, 324(1-4): 323-338.
雷红富, 谢平, 陈广才等. 水文序列变异点检验方法的性能比较分析[J]. 水电能源科学, 2007, 25(4): 36-40.
LEI Hongfu, XIE Ping, CHEN Guangcai, et al. Comparison and analysis on the performance of hydrological time series change point testing methods. Water Resources and Power, 2007, 25(4): 36-40. (in Chinese)
魏风英. 现代气候统计诊断与预测技术[M]. 气象出版社, 北京, 1999.
Wei Fengying. Technology of statistical diagnosis and prediction of modern climate. Meteorological Press, Beijing, 1999.
张珏. 基于非线性理论的石泉和安康水文站径流及洪水规律挖掘[D]. 西安理工大学, 2008.
ZHANG Jue. Runoff and flood law mining based on nonlinear theory in Shiquan and Ankang hydrological stations. Xi’an University of Technology, 2008. (in Chinese)
冯平, 丁志宏, 韩瑞光. 基于EMD的洮河年径流量变化多时间尺度分析[J]. 干旱区资源与环境, 2008, 22(12): 73-76.
FENG Ping, DING Zhihong and HAN Ruiguang. Time-scale analysis on annual runoff time series of the Taohe River based on EMD. Journal of Arid Land Resources and Environment, 2008, 22(12): 73-76. (in Chinese)
刘莉红, 郑祖光, 琚建华. 基于EMD方法的我国年气温和东部年降水量序列的振荡模态分析[J]. 高原气象, 2008, 27(5): 1061-1065.
LIU Lihong, ZHENG Zhuaguang and JU Jianhua. Vacillations of the time series of annual temperature and precipitation in China based on EMD method. Plateau Meteorology, 2008, 27(5): 1061-1065. (in Chinese)
 RICHTER, B. D., BAUMGARTNER, J. V., BRANN, D. P. and POWELL, J. A spatial assessment of hydrologic alter ation within a river network. Regulated River: Research and Management, 1998, 14(4): 329-340.
刘燕, 李小龙, 胡安焱. 河川径流对降水变化的响应研究——以渭河为例[J]. 干旱区地理, 2007, 30(1): 49-52.
LIU Yan, LI Xiaolong and HU Anyan. Response of runoff to precipitation changes—Taking a case of Weihe. Arid Land Geo-graphy, 2007, 30(1): 49-52. (in Chinese)
粟晓玲, 康绍忠, 魏晓妹等. 气候变化和人类活动对渭河流域入黄径流的影响[J]. 西北农林科技大学学报(自然科学版), 2007, 35(2): 153-159.
SU Xiaoling, KANG Shaozhong, WEI Xiaomei, et al. Impact of climate change and human activity on the runoff of Wei River basin to the Yellow River. Journal of Northwest A&F University (Natural Science Edition), 2007, 35(2): 153-159. (in Chinese)
刘俊萍, 田峰巍, 黄强等. 基于小波分析的黄河河川径流变化规律研究[J]. 自然科学进展, 2003, 13(4): 383-387.
LIU Junping, TIAN Fengwei, HUANG Qiang, et a1. The change rule of run-off the Yellow River based on wavelet transform. Progress in Nature Science, 2003, 13(4): 383-387. (in Chinese)