摘要
以普遍推广的A0文件(即基本观测资料)为基础,经过一定的程序处理,得到各气象要素的气候值,并形成建立方程所需自变量数据文件和预报所需的自变量数据文件。计算自变量与因变量的相关系数,将相关系数较大的因子筛选出来,通过人机交互的方式进行第2次筛选,用相关显著的自变量与因变量建立回归方程。由于输出的自变量相对因变量来说都是前期因子,根据回归方程和预报所需的自变量,就可以得出预报结果。
By using the specifically processed basic meteorological observation data, the climatic parameters of various meteorological elements are obtained, and the data files of various independent variables needed for establishing the regression equation and forecasting （each independent variable corresponding to a dependent variable is a predictor） are established. The correlation coefficients of independent variables and dependent variable are computed. Through selecting greater correlation coefficient independent variables in a man-machine interactive way, the regression equations are established. According to the regression equation, forecasts can be obtained from the needed independent variables.
出处
《气象科技》
2008年第1期119-122,共4页
Meteorological Science and Technology
关键词
A0文件
资料处理
逐步回归
自动化
basic meteorological data, data processing, stepwise regression, automation
