城市流通蔬菜区域污染的空间自相关分析

Spatial Autocorrelation Analysis of Urban Circulated Vegetable Area Pollution

以2012年11月16日-30日广州市从化、花都、增城、白云和番禹五个抽检监测区为研究单元,根据蔬菜实际抽检数据,建立了区域蔬菜污染监测指标体系,在此基础上利用主成份分析法得到区域蔬菜污染指数,并区域蔬菜污染指数进行空间自相关性研究。全局Moran'sⅠ结果显示:2012年11月28日蔬菜污染的全局Moran'sⅠ=0.28,Z(I)=2.02〉1.96,说明该天五个监测区域的蔬菜污染状况呈现显著空间正自相关性,即蔬菜污染出现集聚现象,其余时间五个监测区域的蔬菜污染全局空间不相关。局部Moran'sⅠ结果表明,11月19日从化的蔬菜污染出现空间差异,而增城的蔬菜污染呈现空间聚集性;11月25日花都的蔬菜污染呈现出显著的空间正自相关性,其余时间五个区域的蔬菜污染局部空间自相关性均不显著。

Based on the five vegetable sampling areas of Conghua,Huaduo,Zengchen,Baiyun and Panyu of Guangzhou city from 16 Nov.to 30 Nov.2012 as the study units,this paper used the vegetable actual sampling data to establish the area vegetable pollution index system,and got the area vegetable pollution index by means of the principal component analysis,then the spatial auto-correlation analysis was performed in the area vegetable pollution index.The results of global Moran＇s I indicated that,with the global Moran＇s I of vegetable pollution as 0.28 and Z（I） more than 1.96 on 28 Nov.2012,the vegetable pollution state in these five monitoring areas appeared significant spatial plus autocorrelation which mean the vegetable pollution state was gathering phenomenon,and the rest time of global spatial were not autocorrelation.The results of local Moran＇s I showed that the vegetable pollution in Conghua appeared spatial divergence,while the vegetable pollution in Zengchen emerged spatial gathering on 19 Nov.The vegetable pollution in Huaduo appeared significant spatial plus autocorrelation on 25 Nov.and the other time there was no significant local spatial autocorrelation in the five vegetable sampling areas.