基于动态聚类分析和模糊模式识别法的北京城区汛期降雨时空分布规律研究  

Study on Temporal Distribution of Precipitation in Beijing City during Flood Period Based on Dynamic Cluster Analysis and Fuzzy Pattern Recognition

作　　者：刘媛媛[1] 王毅[2] 刘洪伟[2] 杜龙刚 刘舒[1] 柴福鑫[1] LIU Yuanyuan;WANG Yi;LIU Hongwei;DU Longgang;LIU Shu;CHAI Fuxin(China Institude of Water Resources and Hydropower Research,Beijing 100038,China;Office of Flood Control and Drought Relief Headquarters of Beijing Municipal People’s Government,Beijing 100038,China;Beijing Hydrological Center,Beijing 100089,China)

机构地区：[1]中国水利水电科学研究院,北京100038 [2]北京市人民政府防汛抗旱指挥部办公室,北京100038 [3]北京市水文总站,北京100089

出　　处：《水文》2019年第1期74-77,73,共5页Journal of China Hydrology

基　　金：国家重点研发计划"一体化综合减灾智能服务系统"项目(2016YFC0803109)

摘　　要：以北京城区近十几年汛期50场暴雨过程为样本,利用动态聚类分析和模糊模式识别方法对单站雨型和降雨整体空间分布特征进行了研究,得到北京城区汛期最容易发生的雨型以及时空分布规律。结果表明,北京城区汛期主要发生前单峰型雨,峰值集中且极值雨量大,容易造成内涝积水。发生双峰型雨的次数较少,且极值雨量较小,对内涝积水影响相对较小。Dynamic cluster analysis and fuzzy pattern recognition method were used to study the spatial distribution characteristics of rain pattern and rainfall at single station,based on the 50 typical rainstorm processes in Beijing City during the flood periods in recent years.The results show that the unimodal type is mainly temporal pattern.The maximum precipitation is concentrated and the extreme value rainfall is large,which is easy to cause waterlogging.The bimodal type is seldom.

关 键 词：动态聚类 模糊模式识别 时空分布规律 

分 类 号：P333[天文地球—水文科学]