基于判别分析的湖北省马铃薯晚疫病气象风险等级预报模型

Prediction Model of Meteorological Risk Grade for Potato Late Blight in Hubei Province Based on Discriminant Analysis

开展马铃薯晚疫病发病率气象风险等级中长期预报方法研究,以期为科学防控晚疫病提供依据。对湖北省2011—2019年平原低山丘陵区马铃薯晚疫病发病率与气象因子进行相关性分析,筛选出对马铃薯晚疫病发生流行影响较大的气象因子,并利用2011—2017年数据及判别分析方法,结合人为防治因子建立了平原低山丘陵区马铃薯晚疫病发病率气象风险等级预报模型。结果表明:晚疫病发病率与3月中旬最低气温、4月上旬日照时数以及4月下旬风速相关性较好(P<0.05),晚疫病发病率模型自身回代及交叉验证准确率分别为75.9%和75.3%,2018—2019年试报准确率达到61.7%。晚疫病发病率气象风险等级预报模型能够基本满足业务服务需求,通过预测马铃薯晚疫病当年发病情况可以为农技人员提前开展晚疫病防控提供参考。

The long-term prediction methods of meteorological risk grade for incidence of potato late blight were studied to provide a basis for scientific prevention and control of the disease.With the data in the low hilly and plain area of Hubei Province from 2011 to 2019,the correlations between incidence of potato late blight and meteorological factors were analyzed,and the meteorological factors which had great influence on the occurrence and epidemics of potato late blight were selected.By using discriminant analysis method and data of the meteorological factors selected above and artificial control factor from 2011 to 2017,prediction model of meteorological risk grade for incidence of potato late blight was established.The results showed that the incidence of late blight was correlated with the mean minimum temperature in middle March,sunshine duration in early April and wind speed in late April obviously(P<0.05).It was also indicated that the accuracy of the prediction model was 75.9%in self-validation and 75.3%in cross-validation,and the accuracy of the prediction model was 61.7%,which was tested with data from 2018 to 2019.It proved that the prediction model of meteorological risk grade for incidence of potato late blight could basically meet the needs of meteorological service,and it could provide reference for agricultural technicians to carry out late blight prevention and control in advance by predicting the incidence of potato late blight of the year.