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.