基于Copula函数的金沙江流域云龙与松华坝水库入库径流丰枯遭遇分析  

Analysis of Wet and Dry Encounter of Inflow Runoff of Yunlong andSonghuaba Reservoirs in Jinsha River Basin Based on Copula Function

作　　者：孙治才 阿长松 孙岩 杨荣赞 SUN Zhi-cai;A Zhangsong;SUN Yan;YANG Ron-zan(Yunnan Water Resources and Hydropower Survey and Design Institute Co.,Ltd,Kunming 650021,Yunnan Province,China;Yunnan Water Resources and Hydropower Survey and Design Institute,Kunming 650021,Yunnan Province,China;Yunnan University,Kunming 650021,Yunnan Province,China)

机构地区：[1]云南省水利水电勘测设计院有限公司,云南昆明650021 [2]云南省水利水电勘测设计研究院,云南昆明650021 [3]云南大学,云南昆明650021

出　　处：《中国农村水利水电》2024年第7期10-18,共9页China Rural Water and Hydropower

基　　金：云南省重点研发计划项目(202203AA080010)。

摘　　要：通过Copula函数进行建模,采用P-Ⅲ分布分别对金沙江流域承担昆明市供水的两座重要水库——云龙水库和松华坝水库入库径流在三大时间尺度(全年、汛期、非汛期)上进行边缘分布函数的拟合,对拟合的边缘分布函数进行联合分布函数的构建,并对联合分布函数采用K-S检验法进行拟合优度检验,采用图形分析计算法、AIC准则和BIC准则对联合分布函数进行拟合优选,最后通过优选出来的Copula函数对两座水库在三大时间尺度上进行丰枯遭遇分析。主要结论如下:两座水库全年与非汛期入库径流最优Copula函数均为Gumbel Copula函数、汛期为Frank Copula函数。三大时间尺度上两座水库的丰平枯同步概率之和(全年0.734,汛期0.541,非汛期0.728)均大于三大时间尺度上两座水库的丰平枯异步概率之和(全年0.266,汛期0.459,非汛期0.272);在所有丰平枯组合中,两座水库的同丰概率在所有丰平枯组合情况中最高(全年同丰0.310,汛期同丰0.232,非汛期同丰0.308),表明两座水库在所有丰平枯组合中同为丰水年的可能性最大;在全年和非汛期时间尺度上两座水库为枯枯的概率(全年同枯0.298,非汛期0.296)相差不大且均大于两座水库在汛期时间尺度上为枯枯的概率(汛期同枯0.232),因此在两座水库进行实际调度的过程中,特别是在枯水年应考虑两座水库的丰枯同步性。该研究结果可在金沙江流域的水库群联合优化调度中提供一定的决策参考依据。Copula function is used to model,and P-III distribution is used to fit the marginal distribution function of the inflow runoff of two important reservoirs,Yunlong Reservoir and Songhuaba Reservoir,which are responsible for Kunming water supply in the Jinsha River Basin on three time scales (annual, flood season and non-flood season). The joint distribution function of the fitted marginal distributionfunction is constructed, and the K-S test method is used to test the goodness of fit of the joint distribution function. Graphic analysis and cal⁃culation method, AIC criterion and BIC criterion are used to fit and optimize the joint distribution function. Finally, the optimal Copula func⁃tion is used to analyze the wet and dry encounters of the two reservoirs on three time scales. Main results are as follows:the optimal Copulafunction of the annual and non-flood season inflow runoff of the two reservoirs is Gumbel Copula function, and the optimal Copula function inflood season is Frank Copula function. The sum of the synchronous probability of the two reservoirs on the three time scales (0.734 in thewhole year, 0.541 in the flood season, and 0.728 in the non-flood season ) is greater than the sum of the asynchronous probability of the tworeservoirs on the three time scales (0.266 in the whole year, 0.459 in the flood season, and 0.272 in the non-flood season). Among all thewet/normal/dry combinations, the probability of wet/wet of the two reservoirs is the highest in all the combinations (0.310 in the whole year,0.232 in the flood season and 0.308 in the non-flood season ), indicating that the two reservoirs are the most likely to be wet years in all thecombinations. The probability of dry/dry of the two reservoirs on the time scale of the whole year and the non-flood season (0.298 in thewhole year and 0.296 in the non-flood season ) is not much different and both are greater than the probability that the two reservoirs are dryand dry on the time scale of the flood season (0.232 in the flood season ). Therefore,

关 键 词：金沙江流域 COPULA函数 丰枯遭遇 概率分析 

分 类 号：TV121[水利工程—水文学及水资源]