主因子分析在粮食作物产量气象预报中的应用

本文作者提出在农业气象产量预报中,当选用的预报因子较多,而且诸预报因子之间存在一定相关性的情况下,可采用主因子分析方法来浓缩这些预报因子中的信息量。本文简述了主因子分忻的原理和步骤;论述了用主因子回归预报作物产量的一般方法;并选用前期7个大与环流特征参数作为预报因子,建立了辽宁省粮食作物产量的主因子回归预报模式,经过历史概括率检验和试报检验,效果较好。

Application of Principal Factor Analysis in Meteorological Forecasting of Crop Yields

In this paper it is recommended that in agrometeorological forecasting of crop yields, when the number of forecasting factors adopted is great, and when there are certain correlations existing among these factors, principal factor analysis be used to condense the quantity of information carried by these factors. The paper describes the principle and procedures for princi- pal factor analysis and the common practice of forecastinpg crop yields with principal factor regression process. A principal factor regression mod- el is established for grain crops in Liaoning Province, using seven atmos- pheric circulation character parameters in the early stage of growing sea- son as forecasting factors. Historical probability tests and trial forecasting proved that the new method worked well.