中国农业气象 ›› 2015, Vol. 36 ›› Issue (05): 631-639.doi: 10.3969/j.issn.1000-6362. 2015.05.014

• 论文 • 上一篇    下一篇

典型区域小麦白粉病发生气象等级动态预警模型

张 蕾,霍治国,姜 燕,王 丽   

  1. 1. 中国气象科学研究院,北京 100081;2. 南京信息工程大学气象灾害预警预报与评估协同创新中心,南京 210044;3. 国家气象中心,北京 100081;4. 中国气象局应急减灾与公共服务司,北京 100081;5. 西安市人工影响天气办公室,西安 710016
  • 收稿日期:2014-12-30 出版日期:2015-10-20 发布日期:2015-10-19
  • 作者简介:张蕾(1987-),硕士,研究方向为农业和生物气象灾害监测预警与风险评估。E-mail:leizhang@cma. gov.cn
  • 基金资助:
    公益性行业(气象)科研专项经费项目(GYHY201006026)

Occurrence Grade Index Model of Dynamic Early Warning for Wheat Powdery Mildew in Typical Region

ZHANG Lei1, HUO Zhi-guo, JIANG Yan, WANG Li   

  1. 1.Chinese Academy of Meteorological Sciences, Beijing 100081, China; 2.Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters, Nanjing University of Information Science & Technology, Nanjing 210044; 3. National Meteorological Center, Beijing 100081; 4.China Meteorological Administration Department of Emergency Response, Disaster Mitigation and Public Services, Beijing 100081; 5.Xi'an Meteorological Bureau, Xi'an 710016
  • Received:2014-12-30 Online:2015-10-20 Published:2015-10-19

摘要: 通过分析历年小麦白粉病病情指数随时间的变化特点,将河南省南阳市的病情资料统一内插处理为以周为时间尺度的资料序列,将河北省正定县、辛集市、馆陶县和磁县的资料统一内插处理为以候为时间尺度的资料序列。利用2001-2010年小麦白粉病观测资料及相应时段的气象资料,采用秩相关分析和通径分析方法筛选分析影响白粉病发生等级的关键因子及因子贡献,基于Bayes准则建立南阳市和河北4县的区域小麦白粉病等级预测模型。结果表明:小麦白粉病病情指数随时间的变化符合Logistic曲线特点,内插处理效果较好;影响南阳市小麦白粉病发生的关键因子是前1周的病害发生等级、前1周至当周的相对湿度、前2周至当周的日照时数和前3周至当周的降雨系数,影响河北4县的关键因子是前1候的病害发生等级、前3候至当候的平均温度、前3候至当候的降水量和前3候至当候的降雨系数。利用指标模型对南阳市和河北4县病害发生等级进行检验,达到完全“一致”的准确率超过85%,“基本一致”的准确率超过90%,说明模型预警效果较好,可以用于相应区域小麦白粉病短时临近预报。

关键词: 小麦白粉病, 病情指数, 秩相关分析, 通径分析, Bayes

Abstract: As being prone to wheat powdery mildew, Henan and Hebei province are affected severely by wheat powdery mildew. Through analysis on variable characteristic of disease index for winter powdery mildew, the disease index data in Nanyang city was interpolated at week scale and interpolated at pentad scale in Zhengding, Xinji, Guantao and Cixian. Based on disease observation data for wheat powdery mildew from 2001 to 2010 and daily meteorological data for corresponding period, the key factors and key period affecting occurrence grade for wheat powdery mildew were selected using rank correlation analysis and path analysis. Occurrence grade index model of dynamic warning for wheat powdery mildew was built based on Bayes criterion. The results elucidated that the variation of disease index for winter powdery mildew conformed to Logistic curve, and interpolation effect was well. The key factors for wheat powdery mildew occurrence grade in Nanyang were actual occurrence grade of previous week, humidity from the previous one week to current week, sunshine hours from the previous two week to current week and rain coefficient from the previous three week to current week. Which for wheat powdery mildewoccurrence grade in Hebei were actual occurrence grade of previous pentad, mean temperature from the previous three pentad to current pentad, precipitation from the previous three pentad to current pentad, rain coefficient from the previous three pentad to current pentad. The entirely accurate rate of occurrence grade index model of dynamic warning for wheat powdery mildew was above 85%, and accurate rate of nearly same grade exceeded 90%. The model can be well applied in early warning for wheat powdery mildew at the short time scale. The results can provide useful information that contributes to a better understanding of occurrence grade for wheat powdery mildew in Nanyang and Hebei region and help for the policy formation of disease protection management.

Key words: Wheat powdery mildew, Disease index, Rank correlation analysis, Path analysis, Bayes