首页 | 本学科首页   官方微博 | 高级检索  
     

主成分分析与聚类分析在青岛夏季气温变化研究中的应用
引用本文:屈家安,曹杰. 主成分分析与聚类分析在青岛夏季气温变化研究中的应用[J]. 南京气象学院学报, 2014, 0(4): 517-520
作者姓名:屈家安  曹杰
作者单位:南京信息工程大学,江苏南京210044
基金项目:国家自然科学基金资助项目(71101073)
摘    要:选用青岛站1951-2010年每年6-8月各月平均气温资料,通过SAS软件进行了主成分分析和聚类分析,分析了近60 a夏季气温的年际气候变化.主成分分析的结果表明,第一主成分反映青岛夏季气温距平,其正(负)方向反映夏季气温的正(负)距平,其强度反映气温偏高(低)的程度;第二主成分则反映同一年内夏季各月间气温的差异,其绝对值越大,表示各月气温差异越大.聚类分析的结果表明,青岛站夏季月平均气温的变化可以分为3类:1)6月、7月气温较低,在8月升温;2)7月平均气温最高,6月、8月相对较低;3)6月气温低,7、8两月气温较高.其中1993、2003年为第一类,2005年为第二类,其余为第三类.

关 键 词:主成分分析  聚类分析  SAS软件  气象应用

Application of principal component analysis and cluster analysis in a study on the change of summer temperature in Qingdao
QU Jia-an,CAO Jie. Application of principal component analysis and cluster analysis in a study on the change of summer temperature in Qingdao[J]. Journal of Nanjing Institute of Meteorology, 2014, 0(4): 517-520
Authors:QU Jia-an  CAO Jie
Affiliation:(Nanjing University of Information Science & Technology ,Nanjing 210044,China)
Abstract:Interannual change of summer temperature in Qingdao was analyzed based on the monthlyaverage temperature data during June and August in 1951--2010 by using the method of principal com-ponent analysis and cluster analysis through SAS software.According to principal component analysis,the first principal component was the summary of summer temperature departure in Qingdao.The posi-tive (negative) direction reflected the positive (negative) anomaly of temperature and the strength re-flected its degree. The second principal component reflected the temperature difference among eachmonth of the same year.The larger the absolute value of the second principal component was, the greaterthe temperature difference among each month was. Cluster analysis showed that the change of averagemonthly temperature in summer in Qingdao could be divided into three categories. The first type wasthat the temperature was low in June and July and warmed up in August.The second type was that theaverage temperature in July was the highest while in June and August it was relatively low. The thirdcategory was that the temperature in June was low and it got high in both July and August.The year of1993 and 2003 were of the first type,2005 of the second type and the other years of the third type.
Keywords:principal component analysis  cluster analysis  SAS software  applications in meteorology
本文献已被 维普 等数据库收录!
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