辽东山区春季极端低温时空分布特征及发生概率预测  

Spatial and temporal distribution characteristics and occurrence probability prediction of extreme low temperature in spring in eastern Liaoning mountainous area

作　　者：蔡慧君 金磊 肇同斌 陈庆达 冯香程 李秀芬[1,5] CAI Huijun;JIN Lei;ZHAO Tongbin;CHEN Qingda;FENG Xiangcheng;LI Xiufen(Shenyang Agricultural University,Shenyang 110866,China;Northeast Air Traffic Management Bureau of CAAC,Shenyang 110043,China;Dalian Meteorological Bureau,Dalian 116001,China;Liaoning Provincial Meteorological Information Center,Shenyang 110001,China;Qingyuan Forest CERN,Chinese Academy of Sciences,Shenyang 110016,China)

机构地区：[1]沈阳农业大学,辽宁沈阳110866 [2]民航东北地区空中交通管理局气象中心,辽宁沈阳110043 [3]大连市气象局,辽宁大连116001 [4]辽宁省气象信息中心,辽宁沈阳110001 [5]中国科学院清原森林生态系统观测研究站,辽宁沈阳110016

出　　处：《自然灾害学报》2020年第3期173-185,共13页Journal of Natural Disasters

基　　金：国家自然科学基金项目(31870605)。

摘　　要：利用辽宁省东部山区13个气象站1960—2014年春季(3-5月)逐日最低气温等观测资料,采用EOF和Morlet小波分析法分析研究区春季极端低温的时空分布特征,在此基础上用Gumble-I型分布、Weibull分布和正态分布对各站极端低温序列进行了概率分布模型拟合,并检验筛选出最佳模型计算了不同重现期的极端低温及其发生的概率和空间分布。结果表明:(1)辽东山区春季极端低温空间分布主要表现为全区一致分布型、南北相反分布型和东西相反分布型;(2)全区域极端低温存在明显的5年和10年左右的周期变化;(3)所选13个站的极端低温概率分布模型拟合结果显示:Weibull分布为最适宜的分布模型。不同重现期极端低温预测与实测值对比结果表明,所选模型预测结果误差较小(绝对值<1℃,R 2=0.98);(4)北部区和高海拔区容易遭受春季低温的灾害,低温灾害发生的概率及分布区域范围以轻度为最大,中度次之,重度最小。综上所述,该研究结果可为该区域极端低温预测及春季回暖后再次遇到低温灾害的防御提供科学参考依据。The spatiotemporal characteristics are investigated by EOF method and Morlet wavelet analysis based on the spring daily minimum temperature of 13 representative weather stations in Eastern Liaoning mountainous region during 1960-2014.The probability distributions of each station’s extreme minimum temperature sequence are also modeled by the following methods:Gumble-I distribution,Weibull distribution,and Normal distribution.The best model was selected to calculate the extreme minimum temperature at different recurrence periods and occurrence probability and spatial distribution.We find that(1)The spatial distribution of spring extreme minimum temperature in the Eastern Liaoning mountainous region was mainly decomposed into the entire domain EOF mode,the north-south EOF mode,and the east-west EOF mode;(2)There were periods of 5 years and 10 years in the entire domain for extreme minimum temperature;(3)The Weibull distribution was the most appropriate model for the patterns of spring extreme minimum temperature in Eastern Liaoning mountainous region.The insignificant differences between predicted and measured extreme minimum temperature in different recurrence periods indicated minior errors(Absolute error<1°C,R 2=0.98);(4)The northern area and high-altitude area are vulnerable to spring low-temperature disasters.The probability and distribution of mild low-temperature disasters is the largest,the moderate follows,and the severe is the minimal.In conclusion,our results provide scientific reference for the prediction of extreme minimum temperature in this region,and the prevention of recurring low-temperature disaster after spring warm-up.

关 键 词：春季低温 时空变化 EOF 极值概率分布 

分 类 号：P42[天文地球—大气科学及气象学] X43[环境科学与工程—灾害防治]