摘要： 以太行山东麓的赞皇县为对象,选取19712000年该县水文站和周围气象站的年平均降水量,以及各站的经度、纬度、海拔高度等资料,采用回归分析方法,建立了年降水量的不同形式的空间推算模型,并利用3σ规则分析了各模型的海拔高度适用范围。结果表明,X对数型、双对数型模型适用于海拔高度小于1028 975m的山区;线性、Y对数型、Y平方根型适用于海拔高度小于408 482m的山区;X平方根型、双平方根型适用于海拔高度小于518 481m的山区;双曲线(1)型、双曲线(2)型基本上不适用于山区降水量的推算。各模型中以双对数型模拟、推算效果最佳。

关键词: 正态性, 线性回归, 非线性回归, 山区, 推算模型

Abstract: Take Zanhuang County in the eastern part of Taihang Mountains, as an example, the author chose the data of the annual mean precipitation by the hydrometric station of this county and the meteorological sta- tions nearby from 1971 to 2000 and the longitude, latitude and altitude of each station as the study objects1 Us- ing regression analysis, the author built the forms of the spatial predicted model of the annual precipitation and also determined the applicable altitude zone of each model through 3R rule1 The result shows that the models of X logarithms and double logarithms are applied in the mountainous area with the altitude of less than 10281975 m; the modelsof linearity,Ylogarithms andYsquare root are applied in the mountain area with the altitude of less than 4081482 m; the models ofXsquare root and double square root are applied in the mountain area with the altitude of less than 5181481 m; while the models of hyperbola ( 1) and hyperbola ( 2) are generally not ap- plied to the prediction of precipitation in mountainous areas1 The model of double logarithms is the best in both simulating and predicting among all these models1

Key words: Normal distribution, Linear regression, Non-linear regression, Predicted model, Mountainous areas

中图分类号: 

• S16