Abstract:

Based on the PM_{2. 5} data published by 114 cities of China in winter（ from December 2013 to February 2014）,differences in spatial distribution of PM_{2. 5} concentration in winter between the cities and their affecting factors were studied by means of spatial auto-correlation analysis,Kriging interpolation and stepwise regression analysis. Results show that during this period of time,the spatial distribution of PM_2.5 concentration was characterized by clustering of high values,clustering of low values and clustering of low values in high-value domains; the overall Moran＇s I was 0. 27; PM_2.5 concentration displayed a rising first and then declining trend from North to South and from the inland to the coast; and high PM_2.5 concentration was mainly distributed in North China Plain,the middle-lower reaches of the Yangtze River and Guanzhong Plain of Shaanxi,where the mean concentrations were all beyond 150 μg·m^- 3and reached as high as 250 μg·m^{- 3}. Multiple factors stepwise regression analysis shows that the impact of anthropogenic activities was significant on distribution of high PM_2.5 concentrations（ 〉150 μg·m^{- 3}）,but not on that of low concentrations（ ≤75 μg·m^{- 3}）. The dense population and GDP of the secondary industry in the cities were the two main factors affecting distribution of high PM2. 5concentrations,while the area of built districts in the city area,the total population at end of the year of the whole city and the area of paved roads in the city area,were the major factors causing differences between cities in distribution of PM_2.5 concentration.

Key words: PM2.5, spatial structure, anthropogenic influencing factor, GIS