聚类分析在人工增雨效果检验中的应用  

Application of Cluster Analysis to the Statistical Assessment of the Effect of Artifical Rain Enhancement

作　　者：房彬[1] 肖辉[2] 王振会[1] 孙海燕[1] 黄美元[2] 

机构地区：[1]南京信息工程大学电子工程系,江苏南京210044 [2]中国科学院大气物理研究所,北京100029

出　　处：《南京气象学院学报》2005年第6期739-745,共7页Journal of Nanjing Institute of Meteorology

基　　金：国家自然科学基金资助项目(40333033);国家自然科学基金资助项目(40175001);国家"十五"科技攻关计划项目(2001BA610A-06-05);"奥运科技"专项(2001BA904B09);中国科学院大气物理研究所创新项目(8-4605)

摘　　要：在前人研究工作的基础上,对非随机区域历史回归试验进行了改进,提出了一种新的试验方案———基于聚类的浮动对比区历史回归人工增雨效果统计检验方法(CA-FCM方法),并与浮动对比区历史回归统计检验方法(FCM方法)进行了比较。通过对河南省春季6次作业进行评估得出:由于CA-FCM方法采用了聚类分析,提高了对比区和影响区的相关系数,及引入了整层大气可降水量作为协变量,所以提高了作业区自然降水量估计值的准确性,比FCM方法有更高的效果评估效率。In this paper a new method of evaluating the efficiency of non-randomized artificial precipitation enhancement-Cluster-Analysis-based Floating Control historical regression Method （CA- FCM） is presented, and compared with the traditional Floating Control historical regression Method （FCM） which does not base on cluster analysis. It is shown from six examples of effect evaluations for cloud seeding operation cases carried out in the central region of Henan Province, that the CA- FCM can more effectively evaluate the efficiency of cloud seeding operations than the FCM, because the CA-FCM adopts cluster analysis which highly improves the correlativity between rainfall distributions in the control area and target area, and chooses the atmospheric precipitable water as a covariant, thus improving the inferential accuracy of natural rainfall in the cloud seeding operation area.

关 键 词：聚类分析 浮动对比区 历史回归 整层大气可降水量 增雨效果 

分 类 号：P481[天文地球—大气科学及气象学]