首页 | 本学科首页   官方微博 | 高级检索  
     检索      

模糊C均值聚类法在新疆年降水气候分区中的应用
引用本文:王新萍,杨青,李秦,古丽格娜.海力力.模糊C均值聚类法在新疆年降水气候分区中的应用[J].新疆气象,2013,7(6):30-35.
作者姓名:王新萍  杨青  李秦  古丽格娜.海力力
作者单位:中国气象局乌鲁木齐沙漠气象研究所,中国气象局乌鲁木齐沙漠气象研究所,新疆气象信息中心,新疆气象信息中心
基金项目:国家重点基础研究发展计划(973计划)
摘    要:利用新疆98个气象站1960-2011年的年降水量资料,采用模糊C均值聚类法,对新疆年降水量进行分区研究;同时利用线性趋势、累积距平、M-K检验、t检验相结合的方法,对新疆年降水量在不同区域上的变化趋势以及突变时间等进行了对比诊断分析,得到了新疆地区年降水量分布的空间特征。结果表明:(1)新疆年降水的分布大致可分为7个区域;(2)根据模糊C均值方法所分区域能够体现出由于地形差异导致的降水分布不均匀的特点,这一结果与其他研究干旱区降水分布差异产生的原因相同。分区结果合理,说明该聚类方法适用于区域气候区划;(3)新疆大部分区域的降水量在20世纪80年代中后期以前偏少,低于多年平均值,自80年代中后期以后才开始偏多;不同区域的降水量依次从70年代初、80年代中后期、90年代初开始增加;整个南疆盆地、阿勒泰地区、准噶尔盆地降水的突变时间较为接近;(4)新疆降水量整体异常表现为降水一致多(或少)、北多南少、西多东少。

关 键 词:模糊C均值聚类法  降水气候分区  变化趋势  突变分析
收稿时间:2013/1/15 0:00:00
修稿时间:2013/3/24 0:00:00

Fuzzy C means clustering method for climatic regionalization about precipitation in Xinjiang
wangxinping,yang qing,liqin and guligena.hailili.Fuzzy C means clustering method for climatic regionalization about precipitation in Xinjiang[J].Bimonthly of Xinjiang Meteorology,2013,7(6):30-35.
Authors:wangxinping  yang qing  liqin and guligenahailili
Institution:Institute of Desert Meteorology, CMA,Institute of Desert Meteorology, CMA,Meteorological Information Center of Xinjiang,Meteorological Information Center of Xinjiang
Abstract:Based on the annual data of precipitation from 100 meteorological stations in Xinjiang during 1960-2011, spatial characteristics of precipitation for all the region and sub-regions are studied by means of RPCA (rotated principal component analysis) and fuzzy C means clustering method. At the same time, by using combined methods of linear trend, cumulative departure, M-K test, t-test, the variation trend of the precipitation in different sub-regions and also the abrupt changes it underwent are compared and analyzed to find the spatial characteristics of the annual precipitation in Xinjiang. The result showed: (1)It can be roughly divided into 7 sub-regions according to its precipitation in Xinjiang. (2)The sub-regions divided by fuzzy C means clustering method reflects the features showed in the uneven distribution of precipitation caused by terrain differences. In this regard it is the same as the results gained by other methods. The reasonable climatic regionalization achieved by in this way shows that the fuzzy C means clustering method can be applied to climatic regionalization. (3) Comparatively speaking, the precipitation in most of Xinjiang was lower before middle or late 1980s, but it showed an increase since late 1980s. The precipitation in different sub-regions witnessed an increase from early 1970s, middle or late 1980s, early 1990s respectively. The abrupt changes of precipitation of the whole Southern basin, Altay Region and Zhunger basin mainly took place in the same time, and they showed synchronicity. (4)The whole abnormal structures of precipitation in Xinjiang manifest itself in that dry or wet in whole, wet in the North and dry in the South, or wet in the west and dry in the east.
Keywords:rotated principal component analysis  fuzzy c means clustering method  climatic regionalization about precipitation  change trend  abrupt change analysis
本文献已被 维普 等数据库收录!
点击此处可从《新疆气象》浏览原始摘要信息
点击此处可从《新疆气象》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号