摘要
以中国南方设施番茄为研究对象,利用1990—2019年3—9月359个气象站点的气象资料、温室小气候实测资料以及高温控制试验资料,通过BP神经网络模拟南方塑料大棚内日最高气温,结合高温控制试验资料,采用相关性分析和主成分分析方法,构建适用于中国南方设施番茄高温热害等级指标体系,开展设施番茄高温热害风险区划。结果表明:1990—2019年高温热害发生频率增加趋势不显著,轻度高温热害发生频率最高,其次是中度高温热害,各等级高温热害发生频率变化趋势均不显著,且年际变化较大。南方设施番茄高温热害高风险区主要分布在广东西部和东部、广西东部和西部以及云南北部、中部和南部;次高风险区分布在湖南南部、广西大部、广东中北部、江西南部以及福建;中度风险区分布在湖南中北部、江西北部、浙江、安徽、湖北、重庆;其他地区为低风险区。
With the intensification of global warming,high temperature and heat damage of spring-summer occurs frequently in recent years,which seriously affects the growth of greenhouse tomato and reduces agricultural economic efficiency.The high temperature and heat damage risk zoning of tomato is especially important,but the research is not sufficient.Based on meteorological data of 359 stations from 1990 to 2019 and greenhouse microclimate measured data,the highest daily temperature in the southern greenhouse is simulated by the BP neural network.Combining high temperature control test data using correlation analysis and principal component analysis methods,the main indicators are screened out,and a high temperature stress index model is constructed.The mean-standard deviation method is used to divide three levels of high temperature stress,and then the temperature with duration corresponding to the critical value of different levels of high temperature stress index are determined respectively,and the risk index system is established.Using geographic information system(GIS)software and other mathematical statistical methods,the characteristics of the year-over-year changes in the frequency of high temperature and thermal damage of tomatoes in the research area are analyzed,and a comprehensive risk index is established to assess the trend of risk development of high temperature and heat damage to greenhouse tomato in the past30 years.The results show that the frequency of mild and severe high temperature and heat damage is increasing from 1990 to 2019,while the frequency of moderate high temperature and heat damage shows a different pattern(insignificant decrease trend).Among them,the frequency of mild high temperature and heat damage is the highest,followed by moderate high temperature and heat damage.The frequency of heat damage of various grades varies greatly from year to year.There are obvious differences in the distribution of the risk of high temperature and heat damage in space.High-risk areas of heat damage in southern greenhouse tomato are mainly distributed in the western and eastern parts of Guangdong,the eastern and western parts of Guangxi,and the northern,central,and southern parts of Yunnan.The second highrisk areas are mainly distributed in southern Hunan,most areas of Guangxi,central and northern Guangdong,southern Jiangxi,and most areas of Fujian.Areas with moderate heat damage are mainly distributed in north-central Hunan,northwestern Jiangxi,Zhejiang,Anhui,Hubei,and Chongqing.And the risk in western region is low.
作者
郑艳姣
杨再强
王琳
杨世琼
Zheng Yanjiao;Yang Zaiqiang;Wang Lin;Yang Shiqiong(Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters,Nanjing University of Information Science&Technology,Nanjing 210044;Binjiang College,Nanjing University of Information Science&Technology,Wuxi 214000)
出处
《应用气象学报》
CSCD
北大核心
2021年第4期432-442,共11页
Journal of Applied Meteorological Science
基金
国家重点研发计划(2019FYD1002202)
国家自然科学基金项目(41775104)。
关键词
设施番茄
高温热害
BP神经网络
高温胁迫指数
greenhouse tomato
high temperature and heat damage
BP neural network
high temperature stress index
