3.3.2. The Impacts of Synoptic Patterns on PM2.5 Spatial Distribution

Figure 6 shows the average PM2.5 concentrations and their CVs in SCB for six synoptic patterns. The PM2.5 concentrations of different synoptic patterns were almost the same in summer, around 26 μg/m3. In other three seasons, the concentrations of Type 1 and Type 2 were relatively higher than those of Type 3. In winter, PM2.5 concentrations exceeded 80 μg/m3 in Type 1 and Type 2 and only 64 μg/m3 in Type 3. The PM2.5 concentrations of Type 1, Type 2 and Type 3 were 50μg/m3, 43μg/m3 and 36μg/m3 in spring, and 44 μg/m3, 45 μg/m3 and 33 μg/m3 in autumn, respectively. In addition, the concentration of Type 6 also exceeded 80 μg/m3 in winter and the concentrations of Type 5 in spring and Type 4 in autumn were relatively low, about 38 μg/m3 and 32 μg/m3, respectively. Therefore, Type 1, Type 2 and Type 6 were more conducive to the formation of PM2.5 pollution, and the air quality was relatively better in Type 3, Type 4 and Type 5. From the perspective of the spatial disparity, the largest CVs were 0.18 (in Type 2), 0.20 (in Type 5), 0.21 (in Type 4) and 0.22 (in Type 1) in winter, spring, summer and autumn, respectively. This indicated that the CVs were not related to concentrations.

**Figure 6.** Seasonal average PM2.5 concentrations and their CVs for six synoptic patterns in SCB.

The spatial distribution of RAs in four seasons for different synoptic patterns is shown in Figure 7. In winter, the distribution of RAs for Type 1, Type 3 and Type 6 was consistent to the average feature, as shown in Figure 4. However, in Type 2, the RAs in the northwestern basin (CD, DY and MY) were higher than those in the southern basin and the RA in GY showed a relatively low value compared with the other types. The ranges in wintertime RAs were relatively large in Type 2 and small in Type 6. In spring, Type 1 and Type 5 presented more prominent regional characteristics, with higher (lower) RAs in the western and southern (northern) basin. Comparatively, Type 2 and Type 3 showed relatively higher RAs in the northern basin and lower RAs in the western and southern basin. The largest and smallest variation ranges of RAs occurred in Type 5 and Type 2, respectively. In summer, Type 1 showed a similar RA distribution to Type 1 in spring. Type 2 presented a distinct distribution feature with lower RAs in the southern basin and higher RAs in the eastern and northeastern basin. The RA was only 0.08 in ZG and actually negative in YB and LS. The highest RAs were distributed in CQ, DZ and NC. Type 4 and Type 5 showed similar RA distribution characteristics but with higher (lower) RAs in the southern (northern) basin compared with Type 2. In autumn, the RA distribution feature of Type 1 was similar to that of Type 5 in spring, and the distribution features of Type 3 and Type 4 were similar to that of Type 5 in summer. Type 2 showed a different distribution. The RA of ZG was the largest, exceeding 0.40, but the RAs of other southern cities were lower than 0.10. RAs in the western basin were larger than the remaining regions, but the difference was small compared with other types in autumn.

**Figure 7.** The distribution of RAs of seasonal average PM2.5 concentrations for different synoptic patterns in SCB.

3.3.3. The Mechanisms of the Impacts of Synoptic Patterns on PM2.5 Spatial Distribution

PBLH and horizontal wind were key meteorological factors to measure the vertical and horizontal diffusion ability of air pollutants [43]. Hence, the PBLH and 10 m wind fields over SCB in four seasons were extracted from ERA5 reanalysis data and these

fields for different synoptic patterns are presented in Figures 8–11. In winter (Figure 8), the main meteorological feature was the relatively high PBLH area covering the central basin, including the eastern parts of MY, DY, CD and MS, the western parts of ZY and SN, and the southern part of GY. The covering regions and PBLH values, collocating with horizontal wind fields and emissions, determined the distribution of PM2.5. Type 3 presented the highest PBLH and strongest winds, and correspondingly, the average PM2.5 concentration was the lowest. In Type 2, the low PBLH and calm winds were conducive to the accumulation of pollutants. Hence, the distribution of emissions was the dominant factor influencing PM2.5 distribution, which made the RAs vary in relatively larger ranges in Type 2. In Type 1 and Type 6, the regions with massive emissions, such as the western and southern basin, were controlled by the high-PBLH area and relatively strong winds. This made the diffusion condition in higher emission areas better than other areas. As a result, the spatial disparity of PM2.5 concentrations in Type 1 and Type 6 was lower.

In spring, the PBLH in the basin was significantly higher than those in winter. Hence, the wind-induced transportation of air pollutants might be the main factor determining PM2.5 distribution. In Type 5 and Type 3, northerly winds invaded the basin from GY, BZ and DZ, blew straight southwards and converged in the southern and southeastern basin. This flow field could transport air pollutants to the south and aggravate the pollution in the southern basin. Meanwhile, the downwind regions confronted low-PBLH conditions. Consequently, these factors led to the relatively high RAs in the southern basin. In Type 2, the wind speeds were the lowest and pollutant transport was limited. Additionally, the relatively high PBLH in the western and southern basin promoted the diffusion of pollutants in massive emission areas. Therefore, the difference in PM2.5 concentrations in SCB was relatively low. The winds in Type 1 blew from east to west and turned south near the western edge of the basin, which caused higher RAs in the western and southern basin.

**Figure 8.** The PBLH and 10 m wind fields of different synoptic patterns in winter.

In summer, the wind fields were also the dominant factors influencing PM2.5 distribution because of the relatively uniformly distributed PBLH. Similar wind fields were observed in Type 1 compared to those of Type 1 in spring, and resulted in higher RAs in the western and southern basin. The weak southerly winds prevailed in Type 2, with relatively larger velocities in the northeastern basin (CD, DY and MY) and the low PBLH controlled the eastern and southern basin. These made the higher RAs distribute in the northeastern basin (CD, DY, MY) and eastern basin (DZ and CQ). Type 4 and Type 5 presented similar

cyclonic circulation in SCB, which led to calm winds in the southern and eastern basin and resulted in higher RAs in these regions. Furthermore, the air masses from north entered SCB through western pathways and the northerly winds in the western basin were stronger in Type 5. Hence, the RAs in the western basin were lower in Type 5 than those in Type 4.

**Figure 9.** The PBLH and 10 m wind fields of different synoptic patterns in spring.

**Figure 10.** The PBLH and 10 m wind fields of different synoptic patterns in summer.

The wind fields were the dominant factors in autumn due to the relatively low and uniformly distributed PBLH. In Type 1, easterly winds invaded SCB from the northeast basin corner and caused prevailing northwesterly winds in the basin. The transport of air pollutants caused the relatively higher RAs in the western and southern basin. Extreme stagnation conditions occurred in Type 2 and made the high RAs distribute in massive emission areas. Type 3 and Type 4 presented similar northerly invasion air flows and caused cyclonic circulation, converging in the eastern basin. The pollutant transport and calm-wind-induced stagnation led the higher RAs in the southern basin and eastern basin,

respectively. The relatively stronger winds in Type 4 more thoroughly transported the pollutants to the downwind areas, so lower RAs were observed in LS and YB compared with those in Type 3.

**Figure 11.** The PBLH and 10 m wind fields of different synoptic patterns in autumn.
