江西早稻高温热害等级动态判识及时空变化特征

Dynamic identification of double-early rice heat and its spatiotemporal characteristics in Jiangxi Province,China

构建考虑高温天气过程(包括发生时间、持续日数)的水稻高温热害指标,可以实现对水稻高温热害等级的动态判识,对精准监测、预警与评估水稻高温热害意义重大。以江西早稻为对象,利用气象资料、早稻高温热害灾情史料和生育期资料,反演历史早稻高温热害,采用K-S分布拟合检验和置信区间方法,构建基于高温天气过程的早稻高温热害动态指标,并采用预留的独立的早稻高温热害样本进行检验验证。在此基础上,计算江西各站点早稻高温热害指数(M),并分析水稻热害。结果表明:高温天气过程起始时间、持续日数是影响早稻高温热害发生程度的关键因子,其中,起始时间的影响大于持续日数。3~5 d早稻轻、中、重度高温热害的起始时间阈值分别为抽穗后第10~12、5~9、2~4天;6~8 d早稻轻、中、重度高温热害的起始时间阈值为抽穗后第11~18、8~10和1~7天;>8 d早稻轻、中、重度高温热害的起始时间阈值为抽穗后第12~18、8~11和0~7天。指标验证完全一致的吻合率为73.7%,完全一致及相差1级的吻合率为89.5%。1981-2015年,M的线性倾向率为0.04·a^-1,1999年左右发生由低到高突变;M高值区域主要位于江西中部和东北部,M>0.18;江西中部、东北部和南部地区M值呈显著增加趋势,线性倾向率均大于0.04·a^-1。总体来说,本文构建的指标实现了基于高温天气过程的早稻高温热害动态判识,江西中部和东北部是早稻高温热害的高风险区域。

Constructing evaluation indicator for rice heat damage based on hot weather process(occurring time of hot weather and its duration)can realize the dynamic identification of rice hightemperature heat damage level,which is of great importance to the precisely monitoring,warning and assessment of rice heat.Meteorological,historical disaster and phenological data on doubleearly rice in Jiangxi Province were integrated to retrieve the historical heat of double-early rice.The dynamic index of high temperature heat injury on early rice based on high temperature weather process was constructed based on K-S distribution fitting test and confidence interval method.The results were verified with reserved independent samples.A rice heat index(M)was calculated,with which rice heat risk was analyzed.The results showed that the starting time and duration of hot weather were key factors affecting the occurrence of rice heat damage,with the effect of starting time greater than the duration.Light,moderate,and severe rice heat for 3-5 d was identified at 10-12,5-9 and 2-4 d after heading respectively.Similarly,light,moderate and severe rice heat lasting for6-8 d and>8 d started at 11-18,8-10,1-7 d after heading and 12-18,8-11,0-7 d after heading respectively.The coincident rate of rice heat damage indicator was 73.7%,and that verified to be identical or one grade different was 89.5%.The linear tendency rate of M from 1981 to 2015 was 0.04·a^-1,with abrupt change from low to high around 1999.A high M(>0.18)was mainly found in the middle and the northeast part of the study area.Increasing trends of a high M occurred in the middle,northeast and south of Jiangxi,with tendency rates>0.04·a^-1.In general,the indicators constructed in this study realized the dynamic identification of process-based rice heat.The middle and northeast parts of Jiangxi Province were identified as high risk areas for double-early rice heat.