论文部分内容阅读
以新疆天山西部国有林管理局霍城林场果子沟林区的云杉为对象,利用2008-2013年的气象记录数据、雪枯病发生调查数据、森林资源规划设计调查(二类调查)数据和双重筛选逐步回归法、FISHER判别分析法,建立了云杉雪枯病发生期及发生区预测模型,发生期判别模型的相关系数分别为0.999 9,检验样本正判率均达到100%。发生区判别模型的相关系数为:0.817 4,检验样本中发病小班正判率75.00%。
Taking the spruce in Guozigou forest area of Huocheng Forest Farm in the western part of Tianshan Mountain of Xinjiang as an example, using the data of meteorological records, the data of occurrence of snow blight and the data of forest resources planning and design survey (second survey) from 2008 to 2013 and Double screening stepwise regression and FISHER discriminant analysis were used to establish the prediction model of the occurrence period and occurrence area of spruce-withered snow blight. The correlation coefficients of the discriminant model during the period were 0.999 9 and the positive rate of the test samples all reached 100%. The correlation coefficient of the discriminant model of occurrence area is 0.817 4, the positive rate of the small class in the test sample is 75.00%.