水利学报2017，Vol.48Issue(9)：1039-1046,8.DOI:10.13243/j.cnki.slxb.20161280
论水文计算中的相关性分析方法
Modeling dependence and correlation in hydrological calculation
摘要
Abstract
Correlations widely exist in the hydrological phenomena.Correlation analysis is an important content of hydrological calculation.It is involved in nearly every field of hydrological calculation.It is of great help to solving the practical problems by finding a suitable correlation analysis method.Dependence and correlation modeling in hydrological calculation were reviewed thoroughly.The measurement and analysis methods of the correlation,as well as the relevant matters needing attention were summarized.The theoretical obstacles encountered in hydrological calculation of linear regression and auto regression methods were analyzed.The feasibility of using the Copula-based methods to solve the nonlinearity and non-normality of the hydrological variables was studied.When the long-term dependence of the increments of the reservoir storage was considered,the application of R/S analysis method in flood risk analysis was prospected.A new way to calculate the synthesis index for flood event using the thought of principal components analysis was proposed.Then the severity of the flood event can be ranked synthetically.The reliability of the results analyzed by the Copula function was studied.The distortion problem that may occur in the application of this method was also pointed out.Finally,the pseudo correlation problem which is often encountered in the hydrological calculation was analyzed.关键词
相关性/Copula函数/回归分析/自回归分析/R/S分析/主成分分析/假相关Key words
correlation/Copula function/regression/auto regression/R/S analysis/principal components analysis/pseudo correlation分类
天文与地球科学引用本文复制引用
闫宝伟,潘增,薛野,段美壮..论水文计算中的相关性分析方法[J].水利学报,2017,48(9):1039-1046,8.基金项目
国家重点研发计划项目(2016YFC0402708) （2016YFC0402708）
中央高校基本科研业务费资助(HUST:2017KFYXJJ195) （HUST:2017KFYXJJ195）
