摘要
以2011—2013年每年的1—3月和10—12月的中国120个城市逐日空气污染指数(API)数据为基础,通过选取最优的空间插值方法获得雾霾易发生期间每月的API空间分布;采用面向对象的分形网络演化算法对API分布图进行图像分割,获取"污染场"基元,通过G*指数的空间聚集分析提取"污染岛"范围;构建城市污染效应比例指数来反映污染效应的时空变化趋势;利用小波变换分析城市大气污染时间序列的多尺度变化特征。结果表明:近3年来雾霾易发生期间API空间分布的高值区域面积呈现不断增大的态势,京津冀、山东半岛、长三角等地污染岛分布面积较大;在时间尺度上,全国七大地理分区的API强度呈现出一定的周期性,污染强度虽未表现出明显的时间差异,但1—3月西部的污染大于东部,10—12月北方的污染大于南方。
To analyze the causes and characteristics of urban air pollution during a large-scale haze event in China,the API spatial distributions of monthly air pollution index were obtained by using the optimal spatial interpolation method. The analysis was based on the daily air pollution index data of 120 major cities of China in winter and spring( 1,2,3,10,11,12 months) from 2011 to 2013. The urban pollution island( UPI) elements were extracted by adopting object-oriented fractal net evolution approach( FNEA) for the segmentation of API profiles. The G*index spatial aggregation analysis was used to estimate the UPI,and the urban pollution index was constructed to reflect the temporal trends in the UPI. The time series of the daily air pollution index was analyzed using the continuous Morlet wavelet transformation. The correlation between the API values and the meteorological factors were analyzed for the typical cities. The results showed that while the areas with high values of pollution showed an increasing trend in the API spatial distributions over nearly three years,stable compound air pollution characteristics were indicated for Beijing,Tianjin,the Yangtze River Delta,and the Pearl River Delta Region. The proportion of UPI in the national urban pollution index showed an unstable trend in spring and a gradually increasing trend in winter. Normalized trends for pollution islands were found in Beijing,Tianjin,Shandong Peninsula,the Yangtze River Delta,and some other places. The urban air pollution trend during the nationwide large-scale haze event showed a certain periodicity,with an overall deterioration of air quality in winter and then a gradual improvement in spring. Pollution intensity did not exhibit significant seasonal differences;however,the contamination in the western regions was greater than in the eastern regions in spring,and pollution in the northern parts was greater than in the southern regions in winter.
出处
《环境工程学报》
CAS
CSCD
北大核心
2016年第3期1340-1348,共9页
Chinese Journal of Environmental Engineering
基金
国家自然科学基金资助项目(41361040)
甘肃省自然科学基金项目(1506RJZA117)
甘肃省高校基本科研业务费项目(2014-63)
关键词
空气污染分布
雾霾
分形网络演化算法
GIS中国
air pollution distribution
fog / haze
fractal net evolution approach
GIS
China