摘要
利用500多种不同参数模拟计算结果对水文辗转相关插补延长有关问题进行了研究,结果表明∶在资料插补延长中,如果参证站X资料较长而设计站Z与参证站X同期观测资料较短,那么就无法采用设计站Z与参证站X同期观测资料建立回归方程,这时可选用一个中间参证站Y进行辗转相关插补延长,但该站必须具备条件(a)参证站Y与设计站Z、参证站Y与参证站X均具有较长的同期观测资料,(b)参证站Y分别与设计站Z和参证站X均具有较强的相关性,即相关系数RXY>0 9,RYZ>0 9,才能满足插补精度要求.同时还利用统计试验资料建立了辗转相关插补延长各相关系数之间的关系表达式.
Some problems in interpolation and extension of hydrological data by multivariate correlation are studied based on the simulated result of more than five hundred different parameters. The following conclusions are drawn: in hydrological data interpolation and extension, if the data series of reference station X is long, while the series of synchronous observation data of station Z to be designed and reference station X are short, it is difficult to set up a regression equation based on the data series, so the multivariate correlation method should be used for data interpolation and extension by introduction of an intermediate reference station Y. The station introduced should meet the following requirements: (a) longterm synchronous observation data should be available of stations Y and Z, and stations Y and X; (b) reference station Y is well correlated with station Z and X, the correlation coefficients RXY and RYZ being larger than 0.9, so as to ensure the accuracy of interpolation. Furthermore, expressions reflecting the relationships between correlation parameters are deduced based on statistical results.
出处
《河海大学学报（自然科学版）》
CAS
CSCD
北大核心
2003年第5期494-496,共3页
Journal of Hohai University(Natural Sciences)
关键词
水文计算
辗转相关
回归方程
相关系数
hydrological computation
multivariate correlation
regression equation
correlation coefficient
