---

作 者 ：阳传和 杨旺 周仲铭

期 刊 ：林业科学 1992年 28卷 2期 页码：122-127

关键词：杨树溃疡病；预测；多元线型逐步回归模型；

Keywords:Poplar canker, Forecast, Multivariant linear stepwise model,

摘 要 ：根据杨树溃疡病的发生规律,在测报期间内,病情指数(Dis)与相应时期的菌量(F₁、F₂、F₃)、积温(CT)、温湿比(TH)、温雨比(TR)及湿雨比(HR)有着密切的相关关系。应用多元统计方法建立的多元线性回归预测模型:Dis=19.0-2.8(F₃)^(1/2)+0.1F₁+0.0015(F₂)³+4.15TH-0.042(CT)^(1/2)-0.4HR+3.5TR,可以预测一个月后的发病情况,模型经实际病情验证,有较高的准确性。

Abstract:According to the rules of development of poplar canker, caused by Dothiorella gregaria Sacc., the multivariant linear stepwise regression prediction model was established using method of multivatiant statistics with factors as disease index (Dis), pathogens (F₁, F₂, F₃), collect temperature(CT), ratio of temperature to relative humidity (TH), ratio of temperature to raina ll(TR) and ratio of relative humidity to rainfall (HR):Dis=19.0-F₃^1/2.8+0.1F₁+0.0015(F₂)³+4.15TH-CT^1/0.042-0.4HR+1.5TR.The disease index of the following a month could be predicted with this equation what was proved to be reliable by examination using actual disease index obtained in the survey.

---