【作者】 康锡言； 徐建芬；

【Author】 Kang Xiyan1 Xu Jianfen 2 (1.Hebei Meteorology Science Institute, Shijiazhuang 050021; 2.Hengshui Meteorological Bureau,Hengshui 053000)

【机构】 河北省气象科学研究所； 衡水市气象局；

【摘要】 本文以燕山山区的迁西县为例,阐述了在农业气候区划中,应用因子分析建立年降水量模型的步骤,并对模型的适宜性进行了评价。结果表明:因子分析在建立年降水量模型中,适宜范围有一定的限制。即:当自变量与因变量线性相关较差,且公共因子代表的主要自变量与因变量的相关系数相差较小时,适宜建立因子回归模型;当自变量与因变量线性相关较好,且公共因子代表的主要自变量与因变量的相关系数相差较大时,不适宜建立因子回归模型。更多还原

【Abstract】 Taking Qianxi County in the Yan Mountains as an example, the author expatiates the approach of building a model for annual precipitation by factor analysis method and evaluates the models suitability in regional agro-climate programming. The results show that factor analysis is limited in certain regions in the model building of regional annual precipitation, namely factor regression model could be applied when the linear correlation between independent variable and dependent variable is lower, and lesser is the discrepancy of the correlation coefficient for main independent variable and dependent variable represented by jointly factor. Whereas the model is not suitable when the linear correlation between independent variable and dependent variable is higher, and biggish is the discrepancy of the correlation coefficient for main independent variable and dependent variable represented by jointly factor更多还原