基于病穗率的小麦赤霉病发生等级气象预测模型  

Meteorological Forecast Model For Occurrence Grade Of Wheat SCAB Based On Disease Ear Rate

作　　者：孔令帅 江胜国 韩正英 蔡冬梅 钱俊 KONG Lingshuai(Tongcheng Meteorological Bureau,Tongcheng 231400,China)

机构地区：[1]桐城市气象局,安徽桐城231400

出　　处：《安徽农学通报》2020年第16期152-154,共3页Anhui Agricultural Science Bulletin

摘　　要：通过对桐城市1995—2019年共25年小麦赤霉病的发生资料与气象资料进行分析,选取影响小麦赤霉病发生关键期的气温、降水、相对湿度、日照等42个气象因子,基于年度病穗率,采用逐步回归统计方法,建立小麦赤霉病发生等级气象预测模型。结果表明,2月中旬平均相对湿度、3月上旬和中旬累计日照时数与年度病穗率呈明显的负相关,而2月中旬累计日照时数、2月下旬平均最低气温、3月上旬和中旬平均相对湿度则呈显著的正相关;交叉检验复相关系数达到0.63,预测结果同号率达88%,预测复相关系数达到0.68(α0.01=0.5052),均达到0.01的显著性相关水平;2017—2019年试报应用预测值相对误差控制在9%以内,发生等级预测完全正确。模型拟合率达100%,试报效果好,可靠性强,可投入业务应用。Based on the analysis of 25-year meteorological data and occurrence data of wheat scab in Tongcheng from 1995 to 2019,42 meteorological factors,including temperature,precipitation,relative humidity and Sunshine,were selected to influence the key period of wheat scab occurrence,based on the annual ear disease rate,the meteorological forecast model of wheat scab occurrence grade was established by using the stepwise regression statistical method.The results showed that the average relative humidity in the middle of February,the accumulated sunshine hours in the first ten days and the middle ten days of March were significantly negative correlated with the annual ear rate,and The cumulative sunshine hours in the middle of February,the average minimum temperature in the last ten days of February,and the average relative humidity in the first ten days and the middle ten days of March were significantly positively correlated.The multiple correlation coefficient of cross test was 0.63,the same sign rate of prediction results was 88%,and the multiple correlation coefficient of prediction was 0.68(α0.01=0.5052),all of which reached the significant correlation level of 0.01.In 2017-2019,the relative error of the predicted value of the trial application is controlled within 9%,and the occurrence level prediction is completely correct.The fitting rate of the model is up to 100%.The test result is very good and the reliability is strong.It can be put into business application.

关 键 词：病穗率 赤霉病 发生等级 预测模型 

分 类 号：S435[农业科学—农业昆虫与害虫防治]