基于空气质量指数AQI的空间自相关特征分析--以福建省为例

Spatial autocorrelation analysis of air quality index--A case study of Fujian Province

以福建省9个地级市为研究对象,运用时间序列模型、全局空间自相关分析、K-medoids聚类及相关分析等研究方法,探索福建省各地级市空气质量指数的空间自相关特征。研究结果显示,相比秋冬两季,各地级市春夏两季的空气质量相关性较低,具体表现为,春夏两季各地级市莫兰指数(Moran’s I)显著较低。以空气质量指数AQI的6个空气质量分指标作为聚类标准进行研究,发现福建省地级市的空气相关性区分为两大类。进一步对AQI进行相关性分析,发现AQI与PM10、PM2.5、NO2分指标高度相关,由此说明,空气微粒与NO2是福建空气污染的主要来源。空气质量季节上的分布差异,可以为政府准确地制定治理方案提供有力证据。

Sampling prefecture-level cities in Fujian Province and using the time-series model,global spatial autocorrelation method,k-medoids clustering method,and correlation analysis method,we explore the spatial distribution characteristics of the air quality index(AQI)in Fujian Province.The results show that,compared to autumn and winter,Moran′s I is significantly lower in spring and summer,and there is lower correlation between Moran′s I and the air quality impact factors of prefecture-level cities.Taking six air quality sub-indices of AQI as clustering criteria,we divided the nine prefecture-level cities in Fujian Province into two categories,in which natural factors and human factors for the same type of cities have similar characteristics.Further correlation analysis of AQI shows that AQI is highly correlated with PM10,PM2.5,and NO2 than the other indicators,demonstrating that the treatment of air particulates and NO2 is important in improving the air quality of Fujian.This study revealed the differences in the seasonal distribution of air quality,providing strong evidence for the government to accurately formulate a governance plan.