基于Kruskal-Wallis检验的地下水污染质分布的研究——以长春地区为例  

Study on the distribution of groundwater contamination based on Kruskal-Wallis Test MEthod:taking Changchun area a example

作　　者：高彦伟[1] 陈殿友[1] 

机构地区：[1]吉林大学数学学院,长春130012

出　　处：《地球物理学进展》2011年第6期2249-2254,共6页Progress in Geophysics

基　　金：科技部农业科技成果转化资金项目(2009GB2B100085);吉林省科技发展重点项目(20080207)资助

摘　　要：为准确评价地下水污染质空间分布,本文以长春地区地下水重要组分NO_3^-为例,利用统计学的熵—矩检验法对长春地区2008年56个水质采样点的检测数据的概率分布进行了检验,得出数据不服从正态分布的结论.在此基础上,选择不要求总体服从正态分布的非参数Kruskal-Wallis检验方法对长春地区所属的各个行政子区域地下水中的NO_3^-质量浓度的差异性进行了检验,发现7个区域中城区、环城、双阳之间NO_3^-质量浓度没有显著性差异,而与其它4个子区域地下水中的NO_3^-质量浓度存在显著性差异,从而按照各个子区域之间的差异性分区绘制了NO_3^-的单项评价等值线图.结果表明在多样本比较的基础上分区绘制等值线的方法能够反映出多因素作用下地下水污染质空间分布的差异性.To exactly assess the spatial distribution ofduoyinsu groundwater contaminants, in this paper, we take the important component of groundwater, NO~ as example and test with the Entropy-moment Method in statistics the probability distribution of the data collected in 2008 from the 56 groundwater sample points in Changchun area. We conclude that the data don＇t obey the normal distribution. Based on this, we choose in this paper the nonparametric Kruskal-Wallis Method, which doesn＇t require population to obey the normal distribution, to test the difference of mass concentration of NO3- in the groundwater of sub-regions in Changchun and conclude that the mass concentration of NO3 in city area and around and in Shuangyang, among the seven regions tested, isn＇t quite different but is obviously different from that of the rest four sub-regions. Thus, we draw contour maps of NO3 regionally for uni-parameter evealuation according to the difference among the studied sub-regions. The results suggest the method of drawing regional contour maps, based on the comparison among multiple samples, can reflect the difference of the spatial distribution of groundwater contaminants under the influence of multiple factors.

关 键 词：熵-矩检验法 Kruskal-Wallis检验 长春地区 地下水 等值线图 

分 类 号：P631[天文地球—地质矿产勘探]