梅县低温霜冻的预报方法研究

Meixian Low-temperature Method of Forecasting Frost

选取1984～2007年24年内梅县气象站的霜冻发生日前一天14:00的本站气压、14:00的气温、日最低气温、14:00的相对湿度、14:00的风速、14:00的风向等资料作为预报因子。运用合成分析方法,从梅县全区整体角度出发,探讨梅县霜冻发生次数的规律。运用气象统计分析方法,结合SPSS软件确定预报量,并选取恰当的预报因子。根据预报因子计算回归系数标准方程组所包含的有关统计量(因子交叉积、矩阵协方差阵或相关阵,以及因子与预报量交叉积向量等)。通过解线性方程组定出的回归系数,建立梅县低温霜冻回归方程并进行显著性检验。

This paper from 1984 to 2007 weather station 24 years of Meixian frost occurred the day before the 14 o＇lock site of the depression, temperature of the 14 o＇clock, daily minimum temperature, relative humidity of 14 o＇clock, the wind speed of Ahe 14 o＇clock, wind direction, etc. 14：00 information as a prediction factor. Analysis of the use of synthetic methods, the whole region from the point of view of Meixian explore Meixian the law of the number of frost occurred. Statistical analysis of the use of meteorological methods, combined with SPSS software to determine the forecast amount, and select the appropriate forecasting factor. Prediction factor based on the standard regression coefficients calculated by equations of the statistics contained in（ Factor cross-plot, Covarianee matrix or array, as well as the factor and the volume of cross-product vector prediction, etc. ）. Through the solution of linear equations for the regression coefficient, the establishment of low-temperature frost Meixian regression equation and statistical significance test.