摘要
利用湖北省2008—2017年4—9月逐小时降水量资料挑选出141个区域性暴雨个例,在此基础上运用OPTICS空间聚类算法识别出2586个降水聚集区并处理得到936个雨团,根据中尺度雨团定义识别出489个中尺度雨团,包括314个中尺度静止雨团和175个中尺度移动雨团,分析中尺度雨团特征得到:1)中尺度静止雨团的持续时间频数比随时间增加而减少,峰值是2 h,中尺度移动雨团的持续时间呈单峰型,多集中在3~6 h;2)中尺度静止雨团大值区为恩施自治州以及武汉市,次大值在宜昌市和咸宁市北部,中尺度移动雨团分布主要武汉市和孝感市交界处、恩施自治州、宜昌市;3)对中尺度移动雨团路径进行k-mean聚类得出,中尺度移动雨团路径可分为东部型、西南型和西北型三种,其多向东北东南方向移动。
Based on the hourly precipitation data from April to September of 2008-2017 in Hubei Province,the 141 regional storms were selected;on the basis of this,the OPTICS spatial clustering algorithm was used to identify 2586 precipitation clusters and process 936 rain masses.According to the definition of mesoscale rain clusters,489 mesoscale rain clusters were identified,including 314 mesoscale stationary rain clusters and 175 mesoscale mobile rain clusters;analysis of the characteristics of mesoscale rain clusters yielded:1)the duration frequency ratio of mesoscale stationary rain clusters decreases with time,the peak value is 2 h,and the duration of mesoscale mobile rain clusters is single-peak type,mostly concentrated in 3-6 h.2)The large value areas of mesoscale stationary rain masses are Enshi Autonomous Prefecture and Wuhan City,and the second largest value is in Yichang and northern Xianning.Meso-scale moving rain masses are mainly distributed at the junction of Wuhan and Xiaogan,Enshi Autonomous Prefecture,and Yichang.3)The k-mean clustering of mesoscale mobile rain cluster paths can be obtained.There are three types of mesoscale mobile rain cluster paths:eastern,southwest and northwest.And they mostly move to the northeast and southeast direction.
作者
林佩贤
田刚
李超
LIN Peixian;TIAN Gang;LI Chao(Chengdu University of Information Technology,Chengdu 610225;Wuhan Central Meteorological Observatory,Wuhan 430074)
出处
《暴雨灾害》
2021年第5期505-512,共8页
Torrential Rain and Disasters
关键词
区域性暴雨
雨团识别
聚类方法
雨团特征
regional rainstorm
identification of rain clusters
clustering method
characteristics of rain clusters
作者简介
第一作者:林佩贤,主要从事气象科学研究。E-mail:2547673879@qq.com。
