中国农业气象 ›› 2015, Vol. 36 ›› Issue (03): 364-371.doi: 10.3969/j.issn.1000-6362.2015.03.015

• 论文 • 上一篇    下一篇

江西省稻瘟病发生潜势的气象预报模型

郭瑞鸽,刘文英,吴昊   

  1. 江西省气象灾害应急预警中心,南昌330046;九江市气象局,九江332000
  • 收稿日期:2014-10-08 出版日期:2015-06-20 发布日期:2015-10-20
  • 作者简介:郭瑞鸽(1984-),女,河北临城人,工程师,主要从事农业气象研究。Email:65736370@qq.com
  • 基金资助:

    江西省科技支撑计划“双季稻气象灾害预警信号研究”(20133BBF60047)

Meteorological Forecasting Model of Rice Blast Occurrence Potential in Jiangxi  Province

GUO Rui ge, LIU Wen ying, WU Hao   

  1. 1. Meteorological Disaster Emergency Warning Center of Jiangxi, Nanchang330046,China; 2 Jiujiang Meteorological Bureau,Jiujiang332000
  • Received:2014-10-08 Online:2015-06-20 Published:2015-10-20

摘要: 基于1981-2010年江西省市(县)级稻瘟病监测资料和同期气象资料,利用符号相关法,分苗瘟、叶瘟、穗瘟分析各时期病害发生程度与气象因子的关系,并分别建立病害发生潜势气象预报模型。结果表明:与病害高峰同期的气象因子之间,早稻苗瘟与4月份的平均气温、雨量、日照时数呈正相关,与相对湿度呈负相关;早稻叶瘟与5月中旬-6月上旬的平均气温、雨量、日照时数呈负相关,与相对湿度呈正相关;早稻穗瘟与6月上旬-7月上旬的平均气温、日照时数呈负相关,与雨量、相对湿度呈正相关。晚稻苗瘟偏重发生的概率随着苗期高湿天数和雨日的增加而增大;晚稻叶瘟与8月中旬-9月上旬的平均气温、日照时数呈负相关,与雨量、相对湿度呈正相关;晚稻穗瘟与9月平均气温、相对湿度呈正相关,与雨量、日照时数呈负相关。通过相关关系和符号相同、相反的概率差值确定各气象因子距平值的量化值和权重系数,计算稻瘟病发生潜势气象指数,建立稻瘟病发生潜势气象预报模型。历史回代检验表明,模型准确率为60%~75%,且对偏重级别的准确率高于偏轻级别。

关键词: 符号相关法, 稻瘟病, 发生潜势, 气象预报模型, 江西

Abstract: Based on rice blast monitoring data and observed meteorological data in Jiangxi province from 1981 to 2010,the relationship between rice blast degree and meteorological factors was analyzed and the meteorological forecasting model of rice blast occurrence was established in different stages, by using sign correlation method. The results showed that there was positive correlation between seedling blast of early rice and mean temperature, precipitation, sunshine hours in April, while negative correlation between that and air relative humidity. However, there was negative correlation between leaf blast of early rice and mean temperature, precipitation and sunshine hours from second ten days of May to the first ten days of June, while positive correlation between that and air relative humidity. There was negative correlation between neck blast of early rice and mean temperature, sunshine hours from the first ten days of June to the first ten days of July, while positive correlation between that and precipitation, air relative humidity. The occurrence probability of late rice seedling blast increased with the increase of high humidity days and rainy days. There was negative correlation between leaf blast of late rice and mean temperature, sunshine hours from second ten days of August to the first ten days of September, while positive correlation between that and precipitation, air relative humidity. However, there was positive correlation between neck blast of late rice and mean temperature, air relative humidity in September, while negative correlation between that and precipitation, sunshine hours. Based on the correlation among rice blast occurrence and meteorological factors and the absolute difference value between the same sign rate and the opposite sign rate, the rice blast occurrence potential meteorological index were calculated and the meteorological forecasting model was established. The accuracy rate was 60%-75% by testing substitution, and the accuracy rates of severe grade was higher than slight grade.

Key words: Sign correlation, Rice blast, Occurrence potential, Meteorological forecasting model, Jiangxi province