*2.2. PM2.5 Concentration Data and Spatial Characterization Method*

The hourly PM2.5 concentration data of the state-controlled air-quality-monitoring sites in SCB from December 2015 to November 2019 were obtained from China National Environmental Monitoring Centre (https://air.cnemc.cn:18007/ (accessed on 1 December 2019)). The concentrations of all the sites in the same city were averaged to represent the urban PM2.5 concentration in this city. The annual PM2.5 concentration of a certain city was calculated by averaging the hourly data from January to November of the current year and December of the previous year. Correspondingly, winter was defined as January and February in the current year and December in the previous year in this study. Spring, summer and autumn were defined as March to May, June to August and September to November, respectively.

Two metrics were used to analyze the characteristics of spatial distribution of PM2.5 concentration in SCB. The first one was the coefficient of variation (CV) used to quantify the spatial disparity of PM2.5 concentration between 18 cities [12]. The CV was defined as CV = σ/Cm, in which σ and Cm were the standard deviation and the average of PM2.5 concentrations in 18 cities, respectively. During a certain period, the CV represented how uniformly the PM2.5 concentration distributed in the cities of SCB. Smaller CV indicated more obvious regional characteristics of PM2.5 pollution. The CV can characterize the variations in PM2.5 homogeneity but cannot reveal how these variations evolved. Hence, the relative spatial anomalies (RA) of average PM2.5 concentration for each city were used. RA for a certain city was calculated by RA = (Ci−Cm)/Cm, in which Ci and Cm were average PM2.5 concentration in this city and all cities, respectively. Therefore, larger RA meant relatively higher PM2.5 concentration in the relevant city compared to other cities. The inter-annual variation in CVs could provide the change in PM2.5 spatial disparity and the RAs could better show the spatial distribution of PM2.5 varying trend.
