摘要
随着梯级水库和水库群的建设,不仅需要单站的信息,而且需要流域内各站的综合信息,迫切需要发展多站的随机模拟模型。基于多变量分析理论提出了一种新的多站随机模拟方法。采用季节性自回归模型模拟主站的日流量序列,建立多维Copula函数描述主站和从站之间的时空相关性特征,依据Copula函数,生成从站的日流量序列。以长江上游为研究对象,建立了长江宜昌站、金沙江屏山站、岷江高场站、沱江李家湾站、嘉陵江北碚站和乌江武隆站的多站随机模拟模型,并与现有的主站法进行比较分析。结果表明,模拟与实测序列的统计特征值差别不大,模拟效果优于现有的主站法。
New muhi-site daily discharge simulation method need be developed, because more information is needed accompanying with the rapidly increasing construction of reservoirs. In this study, a new multi-site daily flow simula- tion model was proposed, in which the daily flow of key gauging station was simulated by an autoregressive model. The copula function was built to describe the temporal and spatial dependence between the key gauging station and oth- ers. At last, the daily flows of other sites were obtained using the copulas. The upper Yangtze River was selected as a case study. The daily flow of six sites, including the Yichang in the Yangtze River, Pingshan in the Jinshajiang River, Gaochang in the Minjiang River, Lijiawan in the Tuojiang River, Beibei in the Jialingjiang River and Wulong in the Wujiang River, were simulated. The results indicated that the difference of characteristic values between simulated and observed series were small and the proposed method is superior to the current methods. This study provides a new way for stochastic simulation of muhisite daily flows.
出处
《水科学进展》
EI
CAS
CSCD
北大核心
2013年第4期504-512,共9页
Advances in Water Science
基金
国家自然科学基金资助项目(51079100
51190094)~~
关键词
多站日流量
随机模拟
COPULA函数
长江上游
multi-site daily flow
stochastic simulation
Copula function
upper Yangtze River
