变化环境下区域水资源变异与评价方法不确定性
|
摘要
人类一直依赖水文序列作为水文规律最主要和最直接的信息来源,并对水循环现象进行推理解释。但由于环境变化,包括气候变化和人类活动等的影响,水资源形成的过程发生了较大变化,引起水文变异的发生,使得用于水文频率分析的水文序列失去了“一致性”。如何从非一致性水资源序列中推求出水资源变异的规律、选择最适合某种水文变异形式的非一致性水文频率计算方法,并对所得到的结果进行不确定性分析,是目前区域水资源评价、管理中具有基础性和关键性的理论问题之一。针对上述问题,本文主要从三个方面开展了相关的研究工作,分别是区域时空尺度的水资源变异规律研究、非一致性水文频率计算方法的不确定性研究、非一致性水文频率计算方法的择优研究。
(1)研究综述
主要介绍了变化环境下区域水资源评价问题的研究背景及研究意义,在对国内外研究进展进行综述的基础上,指出当前研究所存在的问题,提出作者的研究思路,并介绍本文的主要内容。
(2)区域时空尺度水资源变异分析
在水文变异诊断系统对水资源序列进行综合诊断的基础上,依据变异结果,以变异形式及变异程度为纽带,提出了对区域水资源时间尺度、空间尺度以及年内分配方面,进行水资源变异规律分析的方法。在时间尺度方面,结合三级区永定河册田水库以上的丰镇(饮)和册田水库站不同时间尺度实测径流序列的变异诊断结果,得出影响年尺度、丰枯尺度和季尺度变异年份和变异形式的主要因素,取决于其所包含月尺度径流序列所占比例的大小。在年内分配方面,通过基尼系数GI、集中度Cd、集中期D和不均匀系数Cv的变异诊断结果得出,丰镇(饮)站和册田水库站径流序列年内分配情势趋向于不均匀化。在空间尺度方面,选择三级区永定河册田水库以上及其包含四级区的年降水序列和年径流序列进行变异规律研究,结果显示三级区永定河册田水库以上水资源序列的变异形式,主要受四级区大同和朔州水资源序列变异的影响。
(3)非一致性水文频率计算原理及方法
非一致性水文频率计算原理假设非一致性水文序列由相对一致的随机性成分和非一致的确定性成分两部分组成,确定性规律反映非一致性变化成分,随机性规律反映一致性变化成分。文中详细地介绍了基于该原理而发展起来的5种非一致性水文频率计算方法(IHFCM),并分析了各种方法的特点。
(4)线性趋势成分影响下IHFCM方法的不确定性
借鉴水文模型中对参数不确定性研究的方法,文中归纳了线性趋势成分影响下7种IHFCM方法的不确定性分析的方法；并在其基础上,进一步提出了7种IHFCM方法的优度判别方法。在大量统计试验结果的基础上得出,在过去条件下,对于含有线性趋势成分的非一致性时间序列,基于跳跃分析的IHFCM方法的不确定度最大,其它6种方法的不确定度非常接近。在现状条件下,受到确定性成分的影响,7种方法的不确定度均有所增加,跳跃分析方法的不确定度依然是最大的,其它6种方法的不确定度仍比较接近。同时,斜率K的变化并不会对IHFCM方法的不确定度区间造成太大的影响；
鉴于不确定度不能衡量方法间的优劣程度,提出基于90%置信区间上下限的择优度指标,即通过分别计算过去和现状条件下7种IHFCM方法的择优度再进行综合计算,对方法的优劣程度进行判断。7种IHFCM方法综合择优度显示,跳跃分析方法在线性趋势成分的条件下,表现始终不理想。而其它6种方法的综合择优度非常接近。通过汇总不同斜率条件下的综合择优度,文中给出了方法选择的择优表,在出现线性趋势变异的非一致性频率计算中,只要通过斜率和序列长度两个指标,即可选择出一种最好的IHFCM方法进行后续的频率计算。
结合发生了趋势下降中变异的三级区永定河册田水库以上1956～2000年实测水资源序列,选择基于指数趋势的IHFCM方法作为该区水资源频率计算的最优方法,并对该区过去和现状条件下的水资源量进行评价。结果显示,过去条件下所得到的频率计算结果,与真实情况的平均相对误差(ART)会以90%的概率落在误差区间(2.55%～18.40%),其不确定度为15.85%；现状条件下的频率计算结果,与真实情况的平均相对误差会以90%的概率落在误差区间(1.54%～20.75%),其不确定度为19.21%。
(5)非线性趋势成分影响下IHFCM方法的不确定性
为了更加符合水资源序列非平稳非线性的特点,本文对指数非线性趋势的不确定性及择优度进行了研究。统计试验结果表明,在过去条件下7种IHFCM方法的频率计算结果中,跳跃分析方法的不确定度最大,其它6种方法的不确定度非常接近；现状条件下,受到指数趋势成分的影响,7种方法的不确定度均有所增加,跳跃分析方法的不确定度依然是最大的,而希尔伯特-黄变换(HHT)和小波分析方法受到确定性成分的影响,其不确定度也有较大的变动,其它4种方法的不确定度非常接近；同时,指数参数B的变化并不会对IHFCM方法的不确定度区间造成太大的影响；
从7种IHFCM方法综合择优度之间的比较可以看出,跳跃分析方法在指数趋势成分的条件下,表现始终不理想,而其它6种方法的综合择优度非常接近。通过汇总不同指数参数条件下的综合择优度,本文给出了方法选择的择优表,可以通过指数参数和序列长度两个指标,选择出一种最好的IHFCM方法进行后续的频率计算。
结合发生了趋势下降中变异的三级区永定河册田水库以上1956～2000年实测水资源序列,选择基于线性趋势的IHFCM方法作为该区水资源频率计算的最优方法,并对该区过去和现状条件下的水资源量进行评价。结果显示,过去条件下所得到的频率计算结果,与真实情况的平均相对误差(ART)会以90%的概率落在误差区间(2.75%～21.35%),其不确定度为18.60%；现状条件下的频率计算结果,与真实情况的平均相对误差会以90%的概率落在误差区间(0.89%～21.58%),其不确定度为20.69%。
(6)跳跃成分影响下IHFCM方法的不确定性
跳跃成分与趋势成分有一定的区别,为了适应跳跃成分的特点,本文提出采用标准跳跃系数SJC对跳跃成分影响下7种IFHCM方法的不确定性进行分析,并在其基础上,进一步提出了7种IHFCM方法的优度判别方法。
统计试验结果表明,在过去条件下,总体而言,这7种方法的不确定度均没有超过15%,说明这些方法均较为稳定。其中在SJC=1%时,跳跃分析方法的不确定度最大,随着SJC的增长,该方法的不确定度迅速降低并维持在一个非常低的水平；而其它6种方法则相反,在SJC较小的时候,其不确定度也较小,随着SJC的增大,其不确定度呈现出波动上升的态势。
通过出现跳跃变异的非一致性水资源序列的SJC值,即可选择出一种综合择优度最高的IHFCM方法进行后续的频率计算。从7种IHFCM方法综合择优度之间的比较可以得出,当SJC>2%时,跳跃方法始终是最优的方法,当SJC≤2%时,指数趋势方法的综合择优度最高。
结合发生了跳跃上升变异的三级区吉木乃诸小河1956～2000年实测水资源序列,利用统计试验的结果,本文选择基于跳跃分析的IHFCM方法作为该区水资源频率计算的最优方法,并对该区过去和现状条件下的水资源量进行评价。结果显示,过去条件下所得到的频率计算结果,与真实情况的平均相对误差会以90%的概率落在误差区间(0.04%～0.61%),其不确定度为0.57%。现状条件下的频率计算结果,与真实情况的平均相对误差会以90%的概率落在误差区间(0.15%～1.93%),其不确定度为1.78%。
综合考虑线性趋势、非线性趋势和跳跃成分影响下,7种IHFCM方法的综合择优结果可以得出,7种IHFCM方法的综合择优结果和确定性成分的形式有着密切联系,最为突出的就是跳跃变异,其次是线性趋势,指数趋势也比较明显。
Hydrological series are the uppermost and the most direct resources of information for human to understand the hydrological mechanism. We rely on these information all along to infer and explain the phenomenon in hydrological cycle. But during the recent decades, the physical background has changed a lot duo to climate changes and human activities, and result in the hydrological alteration. The most direct effect of hydrological elements'spatiotemporal variation is that the hydrological series lost their consistency. How to uncover the regulations of hydrological alteration in the inconsistent hydrological series, select a best Inconsistent Hydrological Frequency Calculation Method (IHFCM) to the certain alteration condition, and analyze the uncertainty of the frequency calculation results, have become one of the fundamental and critical theoretical problems. The relevant research works were opterated from3aspects to solve the problem, first of all, the regional temporal and spatial scale hydrological alteration regulations were studied, then the uncertainty of IHFCM was analyzed, finally, the best chosen research of IHFCM was conducted.
(1) Summarize of research
The research background and meanings of regional water resources assessment problem in the changing environment were introduced, based on the internal and international study progress, the problem still exist and the research ideas was put forward, the contents of this paper was presented at the same time.
(2)Regional temporal and spatial scale hydrological alteration analysis
Based on the alteration formation and degree of hydrological series alteration diagnose results from the Hydrological Alteration Diagnose System (HADS), the general methods ofregional hydrological alteration regulations analysis at the scale of temporal, spatial and annual allocation were put forward. With the alteration results of difference temporal scale runoff series of Fengzhen and Cetianshuiku hydrological station located at the third grade region Upstream of Cetian Reservoir at Yongding River, the alteration regulations of temporal scale was uncovered, the results show that:the main factor that impact the altreration formation of annual, drought/flood and seasonal scale is the proportion of monthly runoff series included. The precipitation and runoff series of the third and fourth grade region in Upstream of Cetian Reservoir at Yongding River were chosen to study the spatial regulations, the results uncover that:the closest and most relevant fourth grade to the third grade is Datong and Shuozhou. Four indexes were chosen to study the annual allocation alteration regulations, they are Gini Coefficient GI, Consentration Degree Cd, Consentration Area D, Ununiform Coefficient Cv, the conclusion of the four indexes alteration results of Fengzhen and Cetianshuiku hydrological station show the uneven tendency of annual allocation.
(3)The principle and methods of inconsistent hydrological frequency calculation
The principle of inconsistent hydrological frequency calculation assumes the inconsistent hydrological series was conposed with two components, one is relatively consistent random component, and the other is inconsistent deterministic component. The inconsistent component reflected by the deterministic regulations, and the consistent component reflected by the random regulations. Five IHFCMs were introduced based on this principle, and the characteristics were analyzed in the paper.
(4)Uncertainty of IHFCM with the influence of linear tendency component
Enlightened by the parameters uncertainty research of hydrological model, the way to study the uncertainty of7IHFCMs under the influence of linear tendency componet was summarized, and based on that, the best chosen method of7IHFCMs was also put forward.
The conclusions of abundant experimental statistics as follow:at the past condition, to the inconsistent time series with linear tendency component, the uncertainty degree of IHFCM that based on the jump analysis was the biggest, the other IHFCMs'uncertainty degree were very close to each other. At the telquel condition, because of the impact from deterministic component, the uncertainty of all7IHFCMs increased, the uncertainty degree of IHFCM that based on the jump analysis was still the biggest, the other IHFCMs'uncertainty degree were still very close to each other. Meanwhile, the value change of slope K would not disturb the distribution area of7IHFCMs.
Considering uncertainty degree could not tell the good or bad of the7IHFCMs, the synthesized best chosen index based on the upper limit and lower limit of90%confidence interval was put forward in this paper. Synthesized best chosen index includes2parts; one is the best chosen index at the past condition, the other is the best chosen index at the telquel condition. The synthesized best chosen index results of the7IHFCMs show that, with the influence of linear tendency component, the IHFCM that based on the jump analysis was not good all the time, the other IHFCMs' synthesized best chosen indexes were very close to each other. Based on the synthesized best chosen indexes results, the table of best chosen method was summarized in this paper, from that a best hydrological frequency calculation method could be chosen by the series length and slope if the linear tendency alteration appears in the inconsistent hydrological series.
Because of the linear decrease tendency alteration happened in the hydrological series from the year of1956to2000, the third regionalization Upstream of Cetian Reservoir of Yongding River in the first regionalization Haihe River was chosen as an example. Based on the linear slope and series length, the IHFCM based on the exponent non-linear tendency method was chosen as the best method, to assess the water resources quantity at the past and telquel condition. The assessment results show that, the probability of Average Relative Tolerance (ART) between the calculation results and the reality value at the past condition in the interval from2.55%to18.40%was90%, the uncertainty degree was15.85%. While at the telquel condition, the probability of ART between the calculation results and the reality value in the interval from1.54%to20.75%was90%, the uncertainty degree was19.21%.
(5)Uncertainty of IHFCM with the influence of non-linear tendency component
In order to adjust the non-stationary and non-linear characteristics of hydrological series, the exponent non-linear tendency component was chosen to study the uncertainty and synthesized best chosen index. By the results of experimental statistics, among the hydrological frequency calculation results, the uncertainty degree of IFHCM based on jump analysis was biggest, the other IHFCMs' uncertainty degree were very close to each other. At the telquel condition, because of the impact from deterministic component, the uncertainty of all7IHFCMs increased, the uncertainty degree of IHFCM that based on the jump analysis was still the biggest, while the uncertainty degree of IHFCM that based on the Hilbert-Huang Transform (HHT) and Wavelet Analysis changed much, the other4IHFCMs' uncertainty degrees close to each other. Meanwhile, the value change of exponent parameter B would not disturb the distribution area of7IHFCMs.
Because of the linear decrease tendency alteration happened in the hydrological series from the year of1956to2000, the third regionalization Upstream of Cetian Reservoir of Yongding River in the first regionalization Haihe River was chosen as an example. Based on the linear slope and series length, the IHFCM based on the exponent non-linear tendency method was chosen as the best method, to assess the water resources quantity at the past and telquel condition. The assessment results show that, the probability of Average Relative Tolerance (ART) between the calculation results and the reality value at the past condition in the interval from2.75%to21.35%was90%, the uncertainty degree was18.60%. While at the telquel condition, the probability of ART between the calculation results and the reality value in the interval from0.89%to21.58%was90%, the uncertainty degree was20.69%.
(6)Uncertainry of IHFCM with the influence of jump component
There are some difference between jump and tendency component, in order to adjust the characteristics of jump component, the Standard Jump Coefficient (SJC) was put forward to analyze the uncertainty degree of7IFHCMs with the influence of jump alteration, and based on that, the synthesized best chosen indexes of7IFHCMs were also brought forward.
By the results of experimental statistics, the uncertainty degrees of7IFHCMs at the past condition were all less than15%, so the stability of all IFHCM is well. While the SJC equals to1%, the uncertainty degree of IFHCM based on the jump analysis was the biggest, but as the SJC increasing, its uncertainty degree recede rapidly and keep a low steady level. On the contrary, the uncertainty degrees of the other6IFHCMs were small when the SJC was little, as the growing of SJC, the uncertainty degrees mount up fluctuately.
By the value of SJC of inconsistent hydrological series with jump alteration, the biggest synthesized best chosen index of7IHFCMs could be confirmed to calculate the hydrological frequency. The results show that, if SJC≤2%, the IHFCM based on the exponent tendency is best, while SJC>2%, the IHFCM based on jump analysis is best all the time.
Because of the jump increase alteration happened in the hydrological series from the year of1956to2000, the third regionalization Jimunaizhu River in the first regionalization Northwesten Rivers was chosen as an example. Based on the value of SJC, the IHFCM based on the jump analysis was chosen to assess the water resources quantity at the past and telquel condition. The assessment results show that, the probability of ART between the calculation results and the reality value at the past condition in the interval from0.04%to0.61%was90%, the uncertainty degree was0.57%. While at the telquel condition, the probability of ART between the calculation results and the reality value in the interval from0.15%to1.93%was90%, the uncertainty degree was1.78%.
The synthesized best chosen index results of the7IHFCMs show that, with the influence of exponent non-linear tendency component, the IHFCM that based on the jump analysis was not good all the time, the other IHFCMs' synthesized best chosen indexes were very close to each other. Based on the synthesized best chosen indexes results, the table of best chosen method was summarized in this paper, from that a best hydrological frequency calculation method could be chosen by the series length and exponent parameter if the exponent tendency alteration appears in the inconsistent hydrological series.
(1)研究综述
主要介绍了变化环境下区域水资源评价问题的研究背景及研究意义,在对国内外研究进展进行综述的基础上,指出当前研究所存在的问题,提出作者的研究思路,并介绍本文的主要内容。
(2)区域时空尺度水资源变异分析
在水文变异诊断系统对水资源序列进行综合诊断的基础上,依据变异结果,以变异形式及变异程度为纽带,提出了对区域水资源时间尺度、空间尺度以及年内分配方面,进行水资源变异规律分析的方法。在时间尺度方面,结合三级区永定河册田水库以上的丰镇(饮)和册田水库站不同时间尺度实测径流序列的变异诊断结果,得出影响年尺度、丰枯尺度和季尺度变异年份和变异形式的主要因素,取决于其所包含月尺度径流序列所占比例的大小。在年内分配方面,通过基尼系数GI、集中度Cd、集中期D和不均匀系数Cv的变异诊断结果得出,丰镇(饮)站和册田水库站径流序列年内分配情势趋向于不均匀化。在空间尺度方面,选择三级区永定河册田水库以上及其包含四级区的年降水序列和年径流序列进行变异规律研究,结果显示三级区永定河册田水库以上水资源序列的变异形式,主要受四级区大同和朔州水资源序列变异的影响。
(3)非一致性水文频率计算原理及方法
非一致性水文频率计算原理假设非一致性水文序列由相对一致的随机性成分和非一致的确定性成分两部分组成,确定性规律反映非一致性变化成分,随机性规律反映一致性变化成分。文中详细地介绍了基于该原理而发展起来的5种非一致性水文频率计算方法(IHFCM),并分析了各种方法的特点。
(4)线性趋势成分影响下IHFCM方法的不确定性
借鉴水文模型中对参数不确定性研究的方法,文中归纳了线性趋势成分影响下7种IHFCM方法的不确定性分析的方法；并在其基础上,进一步提出了7种IHFCM方法的优度判别方法。在大量统计试验结果的基础上得出,在过去条件下,对于含有线性趋势成分的非一致性时间序列,基于跳跃分析的IHFCM方法的不确定度最大,其它6种方法的不确定度非常接近。在现状条件下,受到确定性成分的影响,7种方法的不确定度均有所增加,跳跃分析方法的不确定度依然是最大的,其它6种方法的不确定度仍比较接近。同时,斜率K的变化并不会对IHFCM方法的不确定度区间造成太大的影响；
鉴于不确定度不能衡量方法间的优劣程度,提出基于90%置信区间上下限的择优度指标,即通过分别计算过去和现状条件下7种IHFCM方法的择优度再进行综合计算,对方法的优劣程度进行判断。7种IHFCM方法综合择优度显示,跳跃分析方法在线性趋势成分的条件下,表现始终不理想。而其它6种方法的综合择优度非常接近。通过汇总不同斜率条件下的综合择优度,文中给出了方法选择的择优表,在出现线性趋势变异的非一致性频率计算中,只要通过斜率和序列长度两个指标,即可选择出一种最好的IHFCM方法进行后续的频率计算。
结合发生了趋势下降中变异的三级区永定河册田水库以上1956～2000年实测水资源序列,选择基于指数趋势的IHFCM方法作为该区水资源频率计算的最优方法,并对该区过去和现状条件下的水资源量进行评价。结果显示,过去条件下所得到的频率计算结果,与真实情况的平均相对误差(ART)会以90%的概率落在误差区间(2.55%～18.40%),其不确定度为15.85%；现状条件下的频率计算结果,与真实情况的平均相对误差会以90%的概率落在误差区间(1.54%～20.75%),其不确定度为19.21%。
(5)非线性趋势成分影响下IHFCM方法的不确定性
为了更加符合水资源序列非平稳非线性的特点,本文对指数非线性趋势的不确定性及择优度进行了研究。统计试验结果表明,在过去条件下7种IHFCM方法的频率计算结果中,跳跃分析方法的不确定度最大,其它6种方法的不确定度非常接近；现状条件下,受到指数趋势成分的影响,7种方法的不确定度均有所增加,跳跃分析方法的不确定度依然是最大的,而希尔伯特-黄变换(HHT)和小波分析方法受到确定性成分的影响,其不确定度也有较大的变动,其它4种方法的不确定度非常接近；同时,指数参数B的变化并不会对IHFCM方法的不确定度区间造成太大的影响；
从7种IHFCM方法综合择优度之间的比较可以看出,跳跃分析方法在指数趋势成分的条件下,表现始终不理想,而其它6种方法的综合择优度非常接近。通过汇总不同指数参数条件下的综合择优度,本文给出了方法选择的择优表,可以通过指数参数和序列长度两个指标,选择出一种最好的IHFCM方法进行后续的频率计算。
结合发生了趋势下降中变异的三级区永定河册田水库以上1956～2000年实测水资源序列,选择基于线性趋势的IHFCM方法作为该区水资源频率计算的最优方法,并对该区过去和现状条件下的水资源量进行评价。结果显示,过去条件下所得到的频率计算结果,与真实情况的平均相对误差(ART)会以90%的概率落在误差区间(2.75%～21.35%),其不确定度为18.60%；现状条件下的频率计算结果,与真实情况的平均相对误差会以90%的概率落在误差区间(0.89%～21.58%),其不确定度为20.69%。
(6)跳跃成分影响下IHFCM方法的不确定性
跳跃成分与趋势成分有一定的区别,为了适应跳跃成分的特点,本文提出采用标准跳跃系数SJC对跳跃成分影响下7种IFHCM方法的不确定性进行分析,并在其基础上,进一步提出了7种IHFCM方法的优度判别方法。
统计试验结果表明,在过去条件下,总体而言,这7种方法的不确定度均没有超过15%,说明这些方法均较为稳定。其中在SJC=1%时,跳跃分析方法的不确定度最大,随着SJC的增长,该方法的不确定度迅速降低并维持在一个非常低的水平；而其它6种方法则相反,在SJC较小的时候,其不确定度也较小,随着SJC的增大,其不确定度呈现出波动上升的态势。
通过出现跳跃变异的非一致性水资源序列的SJC值,即可选择出一种综合择优度最高的IHFCM方法进行后续的频率计算。从7种IHFCM方法综合择优度之间的比较可以得出,当SJC>2%时,跳跃方法始终是最优的方法,当SJC≤2%时,指数趋势方法的综合择优度最高。
结合发生了跳跃上升变异的三级区吉木乃诸小河1956～2000年实测水资源序列,利用统计试验的结果,本文选择基于跳跃分析的IHFCM方法作为该区水资源频率计算的最优方法,并对该区过去和现状条件下的水资源量进行评价。结果显示,过去条件下所得到的频率计算结果,与真实情况的平均相对误差会以90%的概率落在误差区间(0.04%～0.61%),其不确定度为0.57%。现状条件下的频率计算结果,与真实情况的平均相对误差会以90%的概率落在误差区间(0.15%～1.93%),其不确定度为1.78%。
综合考虑线性趋势、非线性趋势和跳跃成分影响下,7种IHFCM方法的综合择优结果可以得出,7种IHFCM方法的综合择优结果和确定性成分的形式有着密切联系,最为突出的就是跳跃变异,其次是线性趋势,指数趋势也比较明显。
Hydrological series are the uppermost and the most direct resources of information for human to understand the hydrological mechanism. We rely on these information all along to infer and explain the phenomenon in hydrological cycle. But during the recent decades, the physical background has changed a lot duo to climate changes and human activities, and result in the hydrological alteration. The most direct effect of hydrological elements'spatiotemporal variation is that the hydrological series lost their consistency. How to uncover the regulations of hydrological alteration in the inconsistent hydrological series, select a best Inconsistent Hydrological Frequency Calculation Method (IHFCM) to the certain alteration condition, and analyze the uncertainty of the frequency calculation results, have become one of the fundamental and critical theoretical problems. The relevant research works were opterated from3aspects to solve the problem, first of all, the regional temporal and spatial scale hydrological alteration regulations were studied, then the uncertainty of IHFCM was analyzed, finally, the best chosen research of IHFCM was conducted.
(1) Summarize of research
The research background and meanings of regional water resources assessment problem in the changing environment were introduced, based on the internal and international study progress, the problem still exist and the research ideas was put forward, the contents of this paper was presented at the same time.
(2)Regional temporal and spatial scale hydrological alteration analysis
Based on the alteration formation and degree of hydrological series alteration diagnose results from the Hydrological Alteration Diagnose System (HADS), the general methods ofregional hydrological alteration regulations analysis at the scale of temporal, spatial and annual allocation were put forward. With the alteration results of difference temporal scale runoff series of Fengzhen and Cetianshuiku hydrological station located at the third grade region Upstream of Cetian Reservoir at Yongding River, the alteration regulations of temporal scale was uncovered, the results show that:the main factor that impact the altreration formation of annual, drought/flood and seasonal scale is the proportion of monthly runoff series included. The precipitation and runoff series of the third and fourth grade region in Upstream of Cetian Reservoir at Yongding River were chosen to study the spatial regulations, the results uncover that:the closest and most relevant fourth grade to the third grade is Datong and Shuozhou. Four indexes were chosen to study the annual allocation alteration regulations, they are Gini Coefficient GI, Consentration Degree Cd, Consentration Area D, Ununiform Coefficient Cv, the conclusion of the four indexes alteration results of Fengzhen and Cetianshuiku hydrological station show the uneven tendency of annual allocation.
(3)The principle and methods of inconsistent hydrological frequency calculation
The principle of inconsistent hydrological frequency calculation assumes the inconsistent hydrological series was conposed with two components, one is relatively consistent random component, and the other is inconsistent deterministic component. The inconsistent component reflected by the deterministic regulations, and the consistent component reflected by the random regulations. Five IHFCMs were introduced based on this principle, and the characteristics were analyzed in the paper.
(4)Uncertainty of IHFCM with the influence of linear tendency component
Enlightened by the parameters uncertainty research of hydrological model, the way to study the uncertainty of7IHFCMs under the influence of linear tendency componet was summarized, and based on that, the best chosen method of7IHFCMs was also put forward.
The conclusions of abundant experimental statistics as follow:at the past condition, to the inconsistent time series with linear tendency component, the uncertainty degree of IHFCM that based on the jump analysis was the biggest, the other IHFCMs'uncertainty degree were very close to each other. At the telquel condition, because of the impact from deterministic component, the uncertainty of all7IHFCMs increased, the uncertainty degree of IHFCM that based on the jump analysis was still the biggest, the other IHFCMs'uncertainty degree were still very close to each other. Meanwhile, the value change of slope K would not disturb the distribution area of7IHFCMs.
Considering uncertainty degree could not tell the good or bad of the7IHFCMs, the synthesized best chosen index based on the upper limit and lower limit of90%confidence interval was put forward in this paper. Synthesized best chosen index includes2parts; one is the best chosen index at the past condition, the other is the best chosen index at the telquel condition. The synthesized best chosen index results of the7IHFCMs show that, with the influence of linear tendency component, the IHFCM that based on the jump analysis was not good all the time, the other IHFCMs' synthesized best chosen indexes were very close to each other. Based on the synthesized best chosen indexes results, the table of best chosen method was summarized in this paper, from that a best hydrological frequency calculation method could be chosen by the series length and slope if the linear tendency alteration appears in the inconsistent hydrological series.
Because of the linear decrease tendency alteration happened in the hydrological series from the year of1956to2000, the third regionalization Upstream of Cetian Reservoir of Yongding River in the first regionalization Haihe River was chosen as an example. Based on the linear slope and series length, the IHFCM based on the exponent non-linear tendency method was chosen as the best method, to assess the water resources quantity at the past and telquel condition. The assessment results show that, the probability of Average Relative Tolerance (ART) between the calculation results and the reality value at the past condition in the interval from2.55%to18.40%was90%, the uncertainty degree was15.85%. While at the telquel condition, the probability of ART between the calculation results and the reality value in the interval from1.54%to20.75%was90%, the uncertainty degree was19.21%.
(5)Uncertainty of IHFCM with the influence of non-linear tendency component
In order to adjust the non-stationary and non-linear characteristics of hydrological series, the exponent non-linear tendency component was chosen to study the uncertainty and synthesized best chosen index. By the results of experimental statistics, among the hydrological frequency calculation results, the uncertainty degree of IFHCM based on jump analysis was biggest, the other IHFCMs' uncertainty degree were very close to each other. At the telquel condition, because of the impact from deterministic component, the uncertainty of all7IHFCMs increased, the uncertainty degree of IHFCM that based on the jump analysis was still the biggest, while the uncertainty degree of IHFCM that based on the Hilbert-Huang Transform (HHT) and Wavelet Analysis changed much, the other4IHFCMs' uncertainty degrees close to each other. Meanwhile, the value change of exponent parameter B would not disturb the distribution area of7IHFCMs.
Because of the linear decrease tendency alteration happened in the hydrological series from the year of1956to2000, the third regionalization Upstream of Cetian Reservoir of Yongding River in the first regionalization Haihe River was chosen as an example. Based on the linear slope and series length, the IHFCM based on the exponent non-linear tendency method was chosen as the best method, to assess the water resources quantity at the past and telquel condition. The assessment results show that, the probability of Average Relative Tolerance (ART) between the calculation results and the reality value at the past condition in the interval from2.75%to21.35%was90%, the uncertainty degree was18.60%. While at the telquel condition, the probability of ART between the calculation results and the reality value in the interval from0.89%to21.58%was90%, the uncertainty degree was20.69%.
(6)Uncertainry of IHFCM with the influence of jump component
There are some difference between jump and tendency component, in order to adjust the characteristics of jump component, the Standard Jump Coefficient (SJC) was put forward to analyze the uncertainty degree of7IFHCMs with the influence of jump alteration, and based on that, the synthesized best chosen indexes of7IFHCMs were also brought forward.
By the results of experimental statistics, the uncertainty degrees of7IFHCMs at the past condition were all less than15%, so the stability of all IFHCM is well. While the SJC equals to1%, the uncertainty degree of IFHCM based on the jump analysis was the biggest, but as the SJC increasing, its uncertainty degree recede rapidly and keep a low steady level. On the contrary, the uncertainty degrees of the other6IFHCMs were small when the SJC was little, as the growing of SJC, the uncertainty degrees mount up fluctuately.
By the value of SJC of inconsistent hydrological series with jump alteration, the biggest synthesized best chosen index of7IHFCMs could be confirmed to calculate the hydrological frequency. The results show that, if SJC≤2%, the IHFCM based on the exponent tendency is best, while SJC>2%, the IHFCM based on jump analysis is best all the time.
Because of the jump increase alteration happened in the hydrological series from the year of1956to2000, the third regionalization Jimunaizhu River in the first regionalization Northwesten Rivers was chosen as an example. Based on the value of SJC, the IHFCM based on the jump analysis was chosen to assess the water resources quantity at the past and telquel condition. The assessment results show that, the probability of ART between the calculation results and the reality value at the past condition in the interval from0.04%to0.61%was90%, the uncertainty degree was0.57%. While at the telquel condition, the probability of ART between the calculation results and the reality value in the interval from0.15%to1.93%was90%, the uncertainty degree was1.78%.
The synthesized best chosen index results of the7IHFCMs show that, with the influence of exponent non-linear tendency component, the IHFCM that based on the jump analysis was not good all the time, the other IHFCMs' synthesized best chosen indexes were very close to each other. Based on the synthesized best chosen indexes results, the table of best chosen method was summarized in this paper, from that a best hydrological frequency calculation method could be chosen by the series length and exponent parameter if the exponent tendency alteration appears in the inconsistent hydrological series.
引文
[1]谢平,陈广才,雷红富,等.变化环境下地表水资源评价方法[M].北京：科学出版社,2009.
[2]王浩,王建华,秦大庸等.现代水资源评价及水资源学学科体系研究[J].地球科学进展.2002,17(1)：12-17.
[3]陈家琦,钱正英.关于水资源评价和人均水资源量指标的一些问题[J].中国水利.2003(11)：42-44.
[4]梅亚东.水资源评价与管理[M].武汉：武汉大学出版社,2002.
[5]陈家琦,王浩,杨小柳.水资源学[M].北京：科学出版社,2002.
[6]中华人民共和国水利部.水利水电工程水文计算规范(SL278-2002)[S].水利部长江水利委员会水文局,2002.
[7]李艳.水文特征变异的尺度研究——理论、指标体系和评价方法[D].广州：中山大学,2007.
[8]黑马.科学家：人类已经改变了47%的地球表面[N].北京青年报,2002.8.29.
[9]谢平,陈广才,雷红富,等.水文变异诊断系统[J].水力发电学报.2010(1)：8591.
[10]左其亭,窦明,马军霞.水资源学教程[M].北京：中国水利水电出版社,2008.
[11]左其亭,王树谦,刘廷玺.水资源利用与管理[M].郑州：黄河水利出版社,2009.
[12]谢平,窦明,朱勇,等.流域水文模型——气候变化和土地利用/覆被变化的水文水资源效应[M].北京：科学出版社,2010.
[13]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.
[14]谢平,陈广才,雷红富.变化环境下基于跳跃分析的水资源评价方法[J].干旱区地理.2008,31(4)：588-593.
[15]谢平,陈广才,雷红富.变化环境下基于趋势分析的水资源评价方法[J].水利发电学报.2009,28(2)：14-19.
[16]谢平,陈广才,陈丽.变化环境下基于降雨径流关系的水资源评价方法[J].资源科学.2009,31(1)：6974.
[17]Xiong L H, Guo S L. A two-parameter monthly water balance model and its application[J]. Journal of Hydrology.1999,216:111-123.
[18]王浩,王建华,秦大庸,等.基于二元水循环模式的水资源评价理论方法[J].水利学报.2006,37(12)：1496-1502.
[19]夏军.现代水文学的发展与水文复杂性问题的研究[C].第二届中国水问题论坛学术研讨会论文集—水问题复杂性与不确定性研究与进展,北京：中国水利水电出版社,2004.
[20]Vorosmarty C, Lettenmaier D, Leveque C, et al. Humans Transforming the Global Water System[R]. Eos, Transactions, American Geophysical Union,2004.
[21]中华人民共和国国务院.国家中长期科学和技术发展规划纲要(2006-2020年)[R].2006.
[22]Ipcc. Fourth Assessment Report:Climate Change 2007-The Scientific Basic. The Fourth Assessment Report of the Intergovernmental Panel on Climate Change[R]. 2007.
[23]国家自然科学基金委员会地球科学部.地球科学“十一五”发展战略[M].北京：气象出版社,2006.
[24]陈家琦.水安全保障问题浅议[J].自然资源学报.2002,17(3)：276-279.
[25]陈广才.变化环境下非一致性水文频率计算方法及应用[D].武汉：武汉大学,2009.
[26]王志民,任宪韶,郭宏宇.面向21世纪的海河水利[M].天津：天津科学技术出版社,2000.
[27]Richter B D, Baumgartner J V, Wigington R, et al. How much water does a river need[J]. Freshwater Biology.1997,37:231-249.
[28]Richter B D, Baumgartner J V, Braun D P, et al. A spatial assessment of hydrologic alteration within a river network[J]. Regulated Rivers:Research and Management.1998,14(4):329-340.
[29]Perreault L, Bernier J, Bobee B, et al. Bayesian Change-point Analysis in Hydrometeorological Time Series[J]. Journal of Hydrology.2000,235:221-263.
[30]刘昌明,陈志恺.中国水资源现状评价和供需发展趋势分析[M].北京：中国水利水电出版社,2001.
[31]Burn D H, Hag Elnur M A. Detection of hydrologic trends and variability[J]. Journal of Hydrology.2002,255(1-4):107-122.
[32]肖宜,夏军,申明亮,等.差异信息理论在水文时间序列变异点诊断中的应用[J].中国农村水利水电.2001(11)：28-30.
[33]李订芳,胡文超,章文.基于小波包的时间序列变点探测算法[J].控制与决策.2005,20(5)：521-524.
[34]金菊良,魏一鸣,丁晶.基于遗传算法的水文时间序列变点分析方法[J].地理科学.2005,25(6)：720-723.
[35]张少文,张学成,王玲,等.黄河天然径流长期丰枯状态变化特性研究[J].人民黄河.2005,27(5)：9-13.
[36]杨晓玉.南水北调西线一期工程调水区域径流特性及其丰枯遭遇分析[D].天津：天津大学,2008.
[37]牛军宜.供水枢纽工程水环境系统安全管理问题的研究[D].天津：天津大学,2010.
[38]李剑锋,张强,陈晓宏,等.考虑水文变异的黄河干流河道内生态需水研究[J].2011,66(1)：99-110.
[39]汤奇成,程天文,李秀云.中国河川月径流的集中度和集中期的初步研究[J].地理学报.1982,37(4)：383-393.
[40]杨远东.河川径流年内分配的计算方法[J].地理学报.1984,39(2)：218-227.
[41]汤奇成,曲耀光,周聿超.中国干旱区水文及水资源利用[M].北京：科学出版社,1992.
[42]冯国章,李瑛,李佩成.河川径流年内分配不均匀性的量化研究[J].西北农业大学学报.2000,28(3)：50-53.
[43]郑红星,刘昌明.黄河源区径流年内分配变化规律分析[J].地理科学进展.2003,22(6)：585-590.
[44]燕华云,杨贵林,汪青春.长江源区径流年内分配时程变化规律分析[J].冰川冻土.2006,28(4)：526-529.
[45]刘新有,史正涛,彭海英,等.基于“基尼系数”的降水时间分布均匀度变化研究[J].气象研究与应用.2007,28(2)：46-49.
[46]Xie P, Xu B, Hu C X. Alteration of Inner-Annual Runoff Distribution and Attribution Analysis of Longchuan Station at Dongjiang River[Z]. Institute of Electrical and Electronics Engineers, Inc,2011:572-575.
[47]贾宇平.地统计学在地理因子空间变异研究中的应用[J].太原师范学院学报(自然科学版).2003,2(3)：78-81.
[48]胡克林,李保国,陈德立.区域浅层地下水埋深和水质的空间变异性特性[J].水科学进展.2000,11(4)：408-414.
[49]李毅,门旗,罗英.土壤水分空间变异性对灌溉决策的影响研究[J].干早地区农业研究.2000,18(2)：80-85.
[50]李丽娟,王娟,李海滨.无定河流域降雨量空间变异性研究[J].地理研究.2002,21(4)：434-440.
[51]许义和,魏晓妹,蔡明科,等.基于地统计学的宝鸡市区地下水位空间变异特征研究[J].水土保持研究.2011,18(1)：201-214.
[52]朱芮芮,李兰,王浩,等.降水量的空间变异性和空间插值方法的比较研究[J].中国农村水利水电.2004(7)：25-28.
[53]徐冰.基于分形地质统计学的草地土壤空间变异[D].内蒙古：内蒙古农业大学,2009.
[54]仇亚琴.水资源综合评价及水资源演变规律研究[D].北京：中国水利水电科学研究 院,2006.
[55]熊立华,郭生练,付小平,等.两参数月水量平衡模型的研制和应用[J].水科学进展.1996,7(S1)：80-86.
[56]杨井,郭生练,王金星等.基于GIS的分布式月水量平衡模型及其应用[J].武汉大学学报(工学版).2002,35(1)：22-26.
[57]李建峰.基于GIS的流域水资源数量评价方法及其应用研究[D].郑州：郑州大学,2005.
[58]徐斌.动态分质水资源评价方法研究及其应用[D].南京：河海大学,2006.
[59]Strupczewski W G, Singh V P, Feluch W. Non-stationary approach to at-site flood frequency modelling I. Maximum likelihood estimation[J]. Journal of Hydrology. 2001,248(1-4):123-142.
[60]Strupczewski W G, Singh V P, Feluch W. Non-stationary approach to at-site flood frequency modelling II. Weighted least squares estimation[J]. Journal of Hydrology.2001,248(1-4):143-151.
[61]Strupczewski W G, Singh V P, Feluch W. Non-stationary approach to at-site flood frequency modelling. III. Flood analysis of Polish rivers[J]. Journal of Hydrology.2001,248(1-4):152-167.
[62]Cunderlik J M, Burn D H. Non-stationary pooled flood frequency analysis[J]. Journal of Hydrology.2003,276(1-4):210-223.
[63]Zhang L. Multivariate Hydrological Frequency Analysis and Risk Mapping[D]. Louisiana:Louisiana State University,2005.
[64]Renard B, Garreta V, Lang M. An application of Bayesian analysis and Markov chain Monte Carlo methods to the estimation of a regional trend in annual maxima[J]. Water Resources Research.2006,42 (12):422-439.
[65]Leclerc M, Ouarda T. Non-stationary regional flood frequency analysis at ungauged sites [J]. Journal of Hydrology.2007,343(3-4):254-265.
[66]陆中央.试论年径流量系列的还原计算[J].河北水利科技.2001,22(1)：1-5.
[67]唐继业,吴俊秀.水量还原方法及合理性检查[J].东北水利水电.2003(11)：4-5.
[68]王巧平.天然年径流量系列一致性修正方法的改进[J].水利规划与设计.2003(2)：38-40.
[69]乔云峰,夏军,王晓红.投影寻踪法在径流还原计算中的应用研究[J].水利发电学报.2007,26(1)：6-10.
[70]乔云峰.环境变化条件下的地表水资源评价理论方法及应用研究[D].武汉：武汉大学,2004.
[71]叶寿征,叶健.论水资源系列的修正及其他——关于对第二次水资源评价中若干认识的存疑[J].山西水利科技.2006(4)：5-9.
[72]叶寿征,叶健.论水资源系列的修正及其他(续)——关于对第二次水资源评价中若干认识的存疑[J].山西水利科技.2007(1)：4-7.
[73]金新芽.径流还原实用方法研究[D].南京：河海大学,2006.
[74]雷强.河道水量还原计算技术方法及其应用研究——以山西省坪上水库流域为例[D].太原：太原理工大学,2006.
[75]谢平,许斌,章树安,等.变化环境下区域水资源变异问题研究[M].北京：科学出版社,2012.
[76]詹道江,叶守泽.工程水文学[M].北京：中国水利水电出版社,2000.
[77]宋松柏,康艳,荆萍.水文频率曲线参数优化估计研究[J].西北农林科技大学学报(自然科学版).2008,36(4)：193-198.
[78]Tung Y K, Mays L W. Risk models for flood levee design[J]. Water Resources Research.1981,18(4):833-841.
[79]Loukas A. Flood frequency estimation by a derived distribution procedure[J]. Journal of Hydrology.2002,255(1-4):69-89.
[80]Cheng K S, Chiang J L, Hsu C W. Simulation of probabi lity distributions commonly used in hydrological frequency analysis[J]. Hydrological Processes.2007, 21(1):51-60.
[81]Cameron D, Beven K, Tawn J, et al. Flood frequency estimation by continuous simulation with likelihood based uncertainty estimation[J]. Hydrology and Earth System Sciences.2000,4(1):23-34.
[82]Kuo J T, Yen B C, Hsu Y C, et al. Risk analysis for dajn overtopping Feitsui Reservoir as a case study [J]. Journal of Hydraulic Engineering.2007,133(27): 955-963.
[83]Vrugt J A, Gupta H V, Bouten W, et al. Effective and efficient algorithm for multiobject ive optimization of hydro logic models[J]. Water Resources Research. 2003,39(8):12-14.
[84]方乐润.红水河径流的随机模拟研究综述[J].河海科技进展.1991,11(4)：34-39.
[85]方乐润.径流系列的人工合成[J].黑龙江水专学报.1998,25(3)：1-5.
[86]金光炎,朱湖根.水文频率计算中统计参数的不确定性[J].合肥工业大学学报(自然科学版).1999,22(1)：78-82.
[87]张明,周润娟,金菊良.水文频率分析的未确知数随机模拟模型[J].水力发电学报.2012,31(5)：14-18.
[88]Beven K. A manifesto for the equifinality thesis[J]. Journal of Hydrology. 2006,320(10):18-36.
[89]Beven K, Freer J. Equifinality, data assimilation, and uncertainty estimation inmechanisticmodelling of complex environmental systems using the GLUE methodology[J]. Journal of Hydrology.2001,249(1-4):11-29.
[90]Beven K, Binley A. The future of distributedmodels-model calibration and uncertainty prediction[J]. Hydrological Processes.1992,6(3):279-298.
[91]刘艳丽,梁国华,周惠成.水文模型不确定性分析的多准则似然判据GLUE方法[J].四川大学学报(工程科学版).2009,41(4)：89-96.
[92]Tang W H. Bayesian frequency analysis[J]. Journal of the Hydraulics Division. 1980(106):1203-1218.
[93]梁忠民,李磊,王军,等.考虑参数和线型不确定性的水文设计值估计的贝叶斯方法[J].天津大学学报.2010,43(5)：379-384.
[94]尚晓三,王振龙,王栋.基于贝叶斯理论的水文频率参数估计不确定性分析——以P-Ⅲ型分布为例[J].应用基础与工程科学学报.2011,19(4)：554-564.
[95]刘攀,郭生练,田向荣,等.基于贝叶斯理论的水文频率线型选择与综合[J].武汉大学学报(工学版).2005,38(5)：36-40.
[96]张继国,刘新仁.水文水资源中不确定性的信息熵分析方法综述[J].河海大学学报.2000,28(6)：32-37.
[97]Singh V P, Rajagopal A K. A new method of parametes estimation for hydrologic frequency analysis[J]. Hydrol Science and Technology.1987,2(3):33-40.
[98]Singh V P, Guo H. Parameters estimation for 2-parameter Parete distribution by POME[J]. Water Resour Manage.1995(9):81-93.
[99]Singh V P, Guo H. Parameters estimation for 2-parameter generalized Parete distribution by POME[J]. Stochastic Hydrology and Hydraulics.1997(11): 211-227.
[100]Singh V P, Guo H. Parameters estimation for 3-parametergeneralized Parete distribution by POME[J]. Hydrological Sciences Journal.1995(4):165-181.
[101]Singh V P, Guo H, Yu F X. Parameters estimation for 2-parameter log-logistic Parete distribution (LLD3) by maximum entropy[J]. Civil Engineering Systems. 1995(12):343-357.
[102]Singh V P, Guo H. Parameters estimation for 3-parameter log-logistic Parete distribution (LLD2) by POME[J]. Stochastic Hydrology and Hydraulics.1993(7): 163-177.
[103]谢平,陈广才,雷红富,等.论变化环境下的地表水资源评价方法[J].水资源研究.2007,28(3)：1-3.
[104]谢平,陈广才,雷红富.基于Hurst系数的水文变异分析方法[J].应用基础与工程科学学报.2009,17(1)：32-39.
[105]庄常陵.相关系数检验法与方差分析一致性的讨论[J].高等函授学报.2003,16(4)：11-14.
[106]潘承毅,何迎晖.数理统计的原理与方法[M].上海：同济大学出版社,1992.
[107]周芬Kendall检验在水文序列趋势分析中的比较研究[J].人民珠江.2005(2)：35-37.
[108]丁晶.洪水时间序列干扰点的统计推断[J].武汉水利水电学院学报.1986(5)：36-41.
[109]Lee A F S, Heghinian S M. A shift of the mean level in a sequence of independent normal random variable:A Bayesian Approach[J]. Technometrics.1977,19(4): 503-506.
[110]孙山泽.非参数统计讲义[M].北京：北京大学出版社,2000.
[111]陈广才,谢平.水文变异的滑动F检验与识别方法[J].水文.2006,26(5)：57-60.
[112]盛骤,谢式千,潘承毅.概率论与数理统计[M].北京：高等教育出版社,2001.
[113]丁晶,邓育仁.随机水文学[M].成都：成都科技大学出版社,1998.
[114]夏军,穆宏强,邱训平,等.水文序列的时间变异性分析[J].长江职工大学学报.2001,18(3)：1-4.
[115]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)：1012.
[116]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.
[117]Mann H B. Non-parametric test against trend[J]. Econometic.1985(13): 245-259.
[118]熊立华,周芬,肖义,等.水文时间序列变点分析的贝叶斯方法[J].水电能源科学.2003,21(4)：39-41.
[119]徐绪松,马莉莉,陈彦斌.R/S分析的理论基础：分数布朗运动[J].武汉大学学报(理学版).2005,50(5)：547-550.
[120]焦利明,冯世立,邓长江.基于向量相似度赋权法的C3I系统效能评估[J].火力与指挥控制.2005(30)：202-204.
[121]张录军,钱永甫.长江流域汛期降水集中程度和洪涝关系研究[J].地球物理学报.2004,47(4)：622-630.
[122]李如忠,舒琨.基于基尼系数的水污染负荷分配模糊优化决策模型[J].环境科学学报.2010,30(7)：1518-1525.
[123]王丽琼.基于公平性的水污染物总量分配基尼系数分析[J].生态环境.2008,17(5)：1796-1801.
[124]黄英,刘新有.水电开发对河流水沙年内分配的影响分析[J].水科学进展.2010,21(3)：386-387.
[125]韩娜娜.山西桑干河流域水环境承载力研究[D].西安：西安理工大学,2005.
[126]沈宏.天然径流还原计算方法初步探讨[J].水利规划与设计.2003(2)：38-40.
[127]郑泽权,谢平,蔡伟.小波变换在非平稳水文时间序列分析中的初步应用[J].水电能源科学.2001,19(3)：49-51.
[128]谢平,陈广才,李德,等.水文变异综合诊断方法及其应用研究[J].水电能源科学.2005,23(2)：11-14.
[129]曲小红,石生新,荣丰涛.降雨-径流系列的一致性分析[J].山西水利科技.1999(1)：34-37.
[130]薛联青,崔广柏,陈凯麒.非平稳时间序列的动态水位神经网络预报模型[J].湖泊科学.2002,14(1)：19-24.
[131]张学成,王玲,高贵成,等.黄河流域降雨-径流关系动态分析[J].水利水电技术.2001,32(8)：1-5.
[132]谢平,朱勇,陈广才,等.考虑土地利用/覆被变化的集总式流域水文模型及应用研究[J].山地学报.2007,25(3)：257-264.
[133]谢平,郑泽权.水文频率计算有约束加权适线法[J].武汉水利电力大学学报.2000,33(1)：49-52.
[134]周志革,黄文振,张利.一种计算随机变量函数均值和标准差的方法[J].机械强度.2001,23(1)：107-110.
[135]崔维成,徐向东,邱强.一种快速计算随机变量函数均值与标准差的新方法[J].船舶力学.1998(6)：50-60.
[136]谢平,陈丽,唐亚松,等.变化环境下基于非线性趋势分析的干旱评价方法[C].大连：大连理工大学出版社,2009.
[137]王文圣,丁晶,李跃清.水文小波分析[M].北京：化学工业出版社,2005.
[138]桑燕芳,王栋.水文序列小波分析中小波函数选择方法[J].水利学报.2008,39(3)：295-300.
[139]Huang N E. The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis [J]. Proceedings of the Royal Society Lond A.1998(454):903-995.
[140]安怀志.希尔伯特黄变换理论和应用的研究[D].哈尔滨：哈尔滨工程大学,2008.
[141]王恩荣,耿鸿江.黑龙江省主要江河水文要素的周期分析[J].水文.1995(1)：42-47.
[142]赵利红.水文时间序列周期分析方法的研究[D].南京：河海大学,2007.
[143]张友权,张赛珍,徐文斌,等.年月径流序列的随机特性分析[J].云南工学院学报.1994(2)：47-53.
[144]刘静君.水文变异与非一致性测度研究[D].武汉：武汉大学,2012.
[145]黄忠恕.波谱分析方法及其在水文气象学中的应用[M].北京：气象出版社,1983.
[146]木黑亚提·加列力.阿勒泰水文水资源特点及开发利用存在问题[J].农村科技.2012(11)：69-70.
[2]王浩,王建华,秦大庸等.现代水资源评价及水资源学学科体系研究[J].地球科学进展.2002,17(1)：12-17.
[3]陈家琦,钱正英.关于水资源评价和人均水资源量指标的一些问题[J].中国水利.2003(11)：42-44.
[4]梅亚东.水资源评价与管理[M].武汉：武汉大学出版社,2002.
[5]陈家琦,王浩,杨小柳.水资源学[M].北京：科学出版社,2002.
[6]中华人民共和国水利部.水利水电工程水文计算规范(SL278-2002)[S].水利部长江水利委员会水文局,2002.
[7]李艳.水文特征变异的尺度研究——理论、指标体系和评价方法[D].广州：中山大学,2007.
[8]黑马.科学家：人类已经改变了47%的地球表面[N].北京青年报,2002.8.29.
[9]谢平,陈广才,雷红富,等.水文变异诊断系统[J].水力发电学报.2010(1)：8591.
[10]左其亭,窦明,马军霞.水资源学教程[M].北京：中国水利水电出版社,2008.
[11]左其亭,王树谦,刘廷玺.水资源利用与管理[M].郑州：黄河水利出版社,2009.
[12]谢平,窦明,朱勇,等.流域水文模型——气候变化和土地利用/覆被变化的水文水资源效应[M].北京：科学出版社,2010.
[13]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.
[14]谢平,陈广才,雷红富.变化环境下基于跳跃分析的水资源评价方法[J].干旱区地理.2008,31(4)：588-593.
[15]谢平,陈广才,雷红富.变化环境下基于趋势分析的水资源评价方法[J].水利发电学报.2009,28(2)：14-19.
[16]谢平,陈广才,陈丽.变化环境下基于降雨径流关系的水资源评价方法[J].资源科学.2009,31(1)：6974.
[17]Xiong L H, Guo S L. A two-parameter monthly water balance model and its application[J]. Journal of Hydrology.1999,216:111-123.
[18]王浩,王建华,秦大庸,等.基于二元水循环模式的水资源评价理论方法[J].水利学报.2006,37(12)：1496-1502.
[19]夏军.现代水文学的发展与水文复杂性问题的研究[C].第二届中国水问题论坛学术研讨会论文集—水问题复杂性与不确定性研究与进展,北京：中国水利水电出版社,2004.
[20]Vorosmarty C, Lettenmaier D, Leveque C, et al. Humans Transforming the Global Water System[R]. Eos, Transactions, American Geophysical Union,2004.
[21]中华人民共和国国务院.国家中长期科学和技术发展规划纲要(2006-2020年)[R].2006.
[22]Ipcc. Fourth Assessment Report:Climate Change 2007-The Scientific Basic. The Fourth Assessment Report of the Intergovernmental Panel on Climate Change[R]. 2007.
[23]国家自然科学基金委员会地球科学部.地球科学“十一五”发展战略[M].北京：气象出版社,2006.
[24]陈家琦.水安全保障问题浅议[J].自然资源学报.2002,17(3)：276-279.
[25]陈广才.变化环境下非一致性水文频率计算方法及应用[D].武汉：武汉大学,2009.
[26]王志民,任宪韶,郭宏宇.面向21世纪的海河水利[M].天津：天津科学技术出版社,2000.
[27]Richter B D, Baumgartner J V, Wigington R, et al. How much water does a river need[J]. Freshwater Biology.1997,37:231-249.
[28]Richter B D, Baumgartner J V, Braun D P, et al. A spatial assessment of hydrologic alteration within a river network[J]. Regulated Rivers:Research and Management.1998,14(4):329-340.
[29]Perreault L, Bernier J, Bobee B, et al. Bayesian Change-point Analysis in Hydrometeorological Time Series[J]. Journal of Hydrology.2000,235:221-263.
[30]刘昌明,陈志恺.中国水资源现状评价和供需发展趋势分析[M].北京：中国水利水电出版社,2001.
[31]Burn D H, Hag Elnur M A. Detection of hydrologic trends and variability[J]. Journal of Hydrology.2002,255(1-4):107-122.
[32]肖宜,夏军,申明亮,等.差异信息理论在水文时间序列变异点诊断中的应用[J].中国农村水利水电.2001(11)：28-30.
[33]李订芳,胡文超,章文.基于小波包的时间序列变点探测算法[J].控制与决策.2005,20(5)：521-524.
[34]金菊良,魏一鸣,丁晶.基于遗传算法的水文时间序列变点分析方法[J].地理科学.2005,25(6)：720-723.
[35]张少文,张学成,王玲,等.黄河天然径流长期丰枯状态变化特性研究[J].人民黄河.2005,27(5)：9-13.
[36]杨晓玉.南水北调西线一期工程调水区域径流特性及其丰枯遭遇分析[D].天津：天津大学,2008.
[37]牛军宜.供水枢纽工程水环境系统安全管理问题的研究[D].天津：天津大学,2010.
[38]李剑锋,张强,陈晓宏,等.考虑水文变异的黄河干流河道内生态需水研究[J].2011,66(1)：99-110.
[39]汤奇成,程天文,李秀云.中国河川月径流的集中度和集中期的初步研究[J].地理学报.1982,37(4)：383-393.
[40]杨远东.河川径流年内分配的计算方法[J].地理学报.1984,39(2)：218-227.
[41]汤奇成,曲耀光,周聿超.中国干旱区水文及水资源利用[M].北京：科学出版社,1992.
[42]冯国章,李瑛,李佩成.河川径流年内分配不均匀性的量化研究[J].西北农业大学学报.2000,28(3)：50-53.
[43]郑红星,刘昌明.黄河源区径流年内分配变化规律分析[J].地理科学进展.2003,22(6)：585-590.
[44]燕华云,杨贵林,汪青春.长江源区径流年内分配时程变化规律分析[J].冰川冻土.2006,28(4)：526-529.
[45]刘新有,史正涛,彭海英,等.基于“基尼系数”的降水时间分布均匀度变化研究[J].气象研究与应用.2007,28(2)：46-49.
[46]Xie P, Xu B, Hu C X. Alteration of Inner-Annual Runoff Distribution and Attribution Analysis of Longchuan Station at Dongjiang River[Z]. Institute of Electrical and Electronics Engineers, Inc,2011:572-575.
[47]贾宇平.地统计学在地理因子空间变异研究中的应用[J].太原师范学院学报(自然科学版).2003,2(3)：78-81.
[48]胡克林,李保国,陈德立.区域浅层地下水埋深和水质的空间变异性特性[J].水科学进展.2000,11(4)：408-414.
[49]李毅,门旗,罗英.土壤水分空间变异性对灌溉决策的影响研究[J].干早地区农业研究.2000,18(2)：80-85.
[50]李丽娟,王娟,李海滨.无定河流域降雨量空间变异性研究[J].地理研究.2002,21(4)：434-440.
[51]许义和,魏晓妹,蔡明科,等.基于地统计学的宝鸡市区地下水位空间变异特征研究[J].水土保持研究.2011,18(1)：201-214.
[52]朱芮芮,李兰,王浩,等.降水量的空间变异性和空间插值方法的比较研究[J].中国农村水利水电.2004(7)：25-28.
[53]徐冰.基于分形地质统计学的草地土壤空间变异[D].内蒙古：内蒙古农业大学,2009.
[54]仇亚琴.水资源综合评价及水资源演变规律研究[D].北京：中国水利水电科学研究 院,2006.
[55]熊立华,郭生练,付小平,等.两参数月水量平衡模型的研制和应用[J].水科学进展.1996,7(S1)：80-86.
[56]杨井,郭生练,王金星等.基于GIS的分布式月水量平衡模型及其应用[J].武汉大学学报(工学版).2002,35(1)：22-26.
[57]李建峰.基于GIS的流域水资源数量评价方法及其应用研究[D].郑州：郑州大学,2005.
[58]徐斌.动态分质水资源评价方法研究及其应用[D].南京：河海大学,2006.
[59]Strupczewski W G, Singh V P, Feluch W. Non-stationary approach to at-site flood frequency modelling I. Maximum likelihood estimation[J]. Journal of Hydrology. 2001,248(1-4):123-142.
[60]Strupczewski W G, Singh V P, Feluch W. Non-stationary approach to at-site flood frequency modelling II. Weighted least squares estimation[J]. Journal of Hydrology.2001,248(1-4):143-151.
[61]Strupczewski W G, Singh V P, Feluch W. Non-stationary approach to at-site flood frequency modelling. III. Flood analysis of Polish rivers[J]. Journal of Hydrology.2001,248(1-4):152-167.
[62]Cunderlik J M, Burn D H. Non-stationary pooled flood frequency analysis[J]. Journal of Hydrology.2003,276(1-4):210-223.
[63]Zhang L. Multivariate Hydrological Frequency Analysis and Risk Mapping[D]. Louisiana:Louisiana State University,2005.
[64]Renard B, Garreta V, Lang M. An application of Bayesian analysis and Markov chain Monte Carlo methods to the estimation of a regional trend in annual maxima[J]. Water Resources Research.2006,42 (12):422-439.
[65]Leclerc M, Ouarda T. Non-stationary regional flood frequency analysis at ungauged sites [J]. Journal of Hydrology.2007,343(3-4):254-265.
[66]陆中央.试论年径流量系列的还原计算[J].河北水利科技.2001,22(1)：1-5.
[67]唐继业,吴俊秀.水量还原方法及合理性检查[J].东北水利水电.2003(11)：4-5.
[68]王巧平.天然年径流量系列一致性修正方法的改进[J].水利规划与设计.2003(2)：38-40.
[69]乔云峰,夏军,王晓红.投影寻踪法在径流还原计算中的应用研究[J].水利发电学报.2007,26(1)：6-10.
[70]乔云峰.环境变化条件下的地表水资源评价理论方法及应用研究[D].武汉：武汉大学,2004.
[71]叶寿征,叶健.论水资源系列的修正及其他——关于对第二次水资源评价中若干认识的存疑[J].山西水利科技.2006(4)：5-9.
[72]叶寿征,叶健.论水资源系列的修正及其他(续)——关于对第二次水资源评价中若干认识的存疑[J].山西水利科技.2007(1)：4-7.
[73]金新芽.径流还原实用方法研究[D].南京：河海大学,2006.
[74]雷强.河道水量还原计算技术方法及其应用研究——以山西省坪上水库流域为例[D].太原：太原理工大学,2006.
[75]谢平,许斌,章树安,等.变化环境下区域水资源变异问题研究[M].北京：科学出版社,2012.
[76]詹道江,叶守泽.工程水文学[M].北京：中国水利水电出版社,2000.
[77]宋松柏,康艳,荆萍.水文频率曲线参数优化估计研究[J].西北农林科技大学学报(自然科学版).2008,36(4)：193-198.
[78]Tung Y K, Mays L W. Risk models for flood levee design[J]. Water Resources Research.1981,18(4):833-841.
[79]Loukas A. Flood frequency estimation by a derived distribution procedure[J]. Journal of Hydrology.2002,255(1-4):69-89.
[80]Cheng K S, Chiang J L, Hsu C W. Simulation of probabi lity distributions commonly used in hydrological frequency analysis[J]. Hydrological Processes.2007, 21(1):51-60.
[81]Cameron D, Beven K, Tawn J, et al. Flood frequency estimation by continuous simulation with likelihood based uncertainty estimation[J]. Hydrology and Earth System Sciences.2000,4(1):23-34.
[82]Kuo J T, Yen B C, Hsu Y C, et al. Risk analysis for dajn overtopping Feitsui Reservoir as a case study [J]. Journal of Hydraulic Engineering.2007,133(27): 955-963.
[83]Vrugt J A, Gupta H V, Bouten W, et al. Effective and efficient algorithm for multiobject ive optimization of hydro logic models[J]. Water Resources Research. 2003,39(8):12-14.
[84]方乐润.红水河径流的随机模拟研究综述[J].河海科技进展.1991,11(4)：34-39.
[85]方乐润.径流系列的人工合成[J].黑龙江水专学报.1998,25(3)：1-5.
[86]金光炎,朱湖根.水文频率计算中统计参数的不确定性[J].合肥工业大学学报(自然科学版).1999,22(1)：78-82.
[87]张明,周润娟,金菊良.水文频率分析的未确知数随机模拟模型[J].水力发电学报.2012,31(5)：14-18.
[88]Beven K. A manifesto for the equifinality thesis[J]. Journal of Hydrology. 2006,320(10):18-36.
[89]Beven K, Freer J. Equifinality, data assimilation, and uncertainty estimation inmechanisticmodelling of complex environmental systems using the GLUE methodology[J]. Journal of Hydrology.2001,249(1-4):11-29.
[90]Beven K, Binley A. The future of distributedmodels-model calibration and uncertainty prediction[J]. Hydrological Processes.1992,6(3):279-298.
[91]刘艳丽,梁国华,周惠成.水文模型不确定性分析的多准则似然判据GLUE方法[J].四川大学学报(工程科学版).2009,41(4)：89-96.
[92]Tang W H. Bayesian frequency analysis[J]. Journal of the Hydraulics Division. 1980(106):1203-1218.
[93]梁忠民,李磊,王军,等.考虑参数和线型不确定性的水文设计值估计的贝叶斯方法[J].天津大学学报.2010,43(5)：379-384.
[94]尚晓三,王振龙,王栋.基于贝叶斯理论的水文频率参数估计不确定性分析——以P-Ⅲ型分布为例[J].应用基础与工程科学学报.2011,19(4)：554-564.
[95]刘攀,郭生练,田向荣,等.基于贝叶斯理论的水文频率线型选择与综合[J].武汉大学学报(工学版).2005,38(5)：36-40.
[96]张继国,刘新仁.水文水资源中不确定性的信息熵分析方法综述[J].河海大学学报.2000,28(6)：32-37.
[97]Singh V P, Rajagopal A K. A new method of parametes estimation for hydrologic frequency analysis[J]. Hydrol Science and Technology.1987,2(3):33-40.
[98]Singh V P, Guo H. Parameters estimation for 2-parameter Parete distribution by POME[J]. Water Resour Manage.1995(9):81-93.
[99]Singh V P, Guo H. Parameters estimation for 2-parameter generalized Parete distribution by POME[J]. Stochastic Hydrology and Hydraulics.1997(11): 211-227.
[100]Singh V P, Guo H. Parameters estimation for 3-parametergeneralized Parete distribution by POME[J]. Hydrological Sciences Journal.1995(4):165-181.
[101]Singh V P, Guo H, Yu F X. Parameters estimation for 2-parameter log-logistic Parete distribution (LLD3) by maximum entropy[J]. Civil Engineering Systems. 1995(12):343-357.
[102]Singh V P, Guo H. Parameters estimation for 3-parameter log-logistic Parete distribution (LLD2) by POME[J]. Stochastic Hydrology and Hydraulics.1993(7): 163-177.
[103]谢平,陈广才,雷红富,等.论变化环境下的地表水资源评价方法[J].水资源研究.2007,28(3)：1-3.
[104]谢平,陈广才,雷红富.基于Hurst系数的水文变异分析方法[J].应用基础与工程科学学报.2009,17(1)：32-39.
[105]庄常陵.相关系数检验法与方差分析一致性的讨论[J].高等函授学报.2003,16(4)：11-14.
[106]潘承毅,何迎晖.数理统计的原理与方法[M].上海：同济大学出版社,1992.
[107]周芬Kendall检验在水文序列趋势分析中的比较研究[J].人民珠江.2005(2)：35-37.
[108]丁晶.洪水时间序列干扰点的统计推断[J].武汉水利水电学院学报.1986(5)：36-41.
[109]Lee A F S, Heghinian S M. A shift of the mean level in a sequence of independent normal random variable:A Bayesian Approach[J]. Technometrics.1977,19(4): 503-506.
[110]孙山泽.非参数统计讲义[M].北京：北京大学出版社,2000.
[111]陈广才,谢平.水文变异的滑动F检验与识别方法[J].水文.2006,26(5)：57-60.
[112]盛骤,谢式千,潘承毅.概率论与数理统计[M].北京：高等教育出版社,2001.
[113]丁晶,邓育仁.随机水文学[M].成都：成都科技大学出版社,1998.
[114]夏军,穆宏强,邱训平,等.水文序列的时间变异性分析[J].长江职工大学学报.2001,18(3)：1-4.
[115]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)：1012.
[116]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.
[117]Mann H B. Non-parametric test against trend[J]. Econometic.1985(13): 245-259.
[118]熊立华,周芬,肖义,等.水文时间序列变点分析的贝叶斯方法[J].水电能源科学.2003,21(4)：39-41.
[119]徐绪松,马莉莉,陈彦斌.R/S分析的理论基础：分数布朗运动[J].武汉大学学报(理学版).2005,50(5)：547-550.
[120]焦利明,冯世立,邓长江.基于向量相似度赋权法的C3I系统效能评估[J].火力与指挥控制.2005(30)：202-204.
[121]张录军,钱永甫.长江流域汛期降水集中程度和洪涝关系研究[J].地球物理学报.2004,47(4)：622-630.
[122]李如忠,舒琨.基于基尼系数的水污染负荷分配模糊优化决策模型[J].环境科学学报.2010,30(7)：1518-1525.
[123]王丽琼.基于公平性的水污染物总量分配基尼系数分析[J].生态环境.2008,17(5)：1796-1801.
[124]黄英,刘新有.水电开发对河流水沙年内分配的影响分析[J].水科学进展.2010,21(3)：386-387.
[125]韩娜娜.山西桑干河流域水环境承载力研究[D].西安：西安理工大学,2005.
[126]沈宏.天然径流还原计算方法初步探讨[J].水利规划与设计.2003(2)：38-40.
[127]郑泽权,谢平,蔡伟.小波变换在非平稳水文时间序列分析中的初步应用[J].水电能源科学.2001,19(3)：49-51.
[128]谢平,陈广才,李德,等.水文变异综合诊断方法及其应用研究[J].水电能源科学.2005,23(2)：11-14.
[129]曲小红,石生新,荣丰涛.降雨-径流系列的一致性分析[J].山西水利科技.1999(1)：34-37.
[130]薛联青,崔广柏,陈凯麒.非平稳时间序列的动态水位神经网络预报模型[J].湖泊科学.2002,14(1)：19-24.
[131]张学成,王玲,高贵成,等.黄河流域降雨-径流关系动态分析[J].水利水电技术.2001,32(8)：1-5.
[132]谢平,朱勇,陈广才,等.考虑土地利用/覆被变化的集总式流域水文模型及应用研究[J].山地学报.2007,25(3)：257-264.
[133]谢平,郑泽权.水文频率计算有约束加权适线法[J].武汉水利电力大学学报.2000,33(1)：49-52.
[134]周志革,黄文振,张利.一种计算随机变量函数均值和标准差的方法[J].机械强度.2001,23(1)：107-110.
[135]崔维成,徐向东,邱强.一种快速计算随机变量函数均值与标准差的新方法[J].船舶力学.1998(6)：50-60.
[136]谢平,陈丽,唐亚松,等.变化环境下基于非线性趋势分析的干旱评价方法[C].大连：大连理工大学出版社,2009.
[137]王文圣,丁晶,李跃清.水文小波分析[M].北京：化学工业出版社,2005.
[138]桑燕芳,王栋.水文序列小波分析中小波函数选择方法[J].水利学报.2008,39(3)：295-300.
[139]Huang N E. The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis [J]. Proceedings of the Royal Society Lond A.1998(454):903-995.
[140]安怀志.希尔伯特黄变换理论和应用的研究[D].哈尔滨：哈尔滨工程大学,2008.
[141]王恩荣,耿鸿江.黑龙江省主要江河水文要素的周期分析[J].水文.1995(1)：42-47.
[142]赵利红.水文时间序列周期分析方法的研究[D].南京：河海大学,2007.
[143]张友权,张赛珍,徐文斌,等.年月径流序列的随机特性分析[J].云南工学院学报.1994(2)：47-53.
[144]刘静君.水文变异与非一致性测度研究[D].武汉：武汉大学,2012.
[145]黄忠恕.波谱分析方法及其在水文气象学中的应用[M].北京：气象出版社,1983.
[146]木黑亚提·加列力.阿勒泰水文水资源特点及开发利用存在问题[J].农村科技.2012(11)：69-70.