Comparative Analysis of Long-Term Variation Characteristics of SO2, NO2, and O3 in the Ecological and Economic Zones of the Western Sichuan Plateau, Southwest China

Sulfur dioxide (SO2), nitrogen dioxide (NO2), and ozone (O3) are important atmospheric pollutants that affect air quality. The long-term variations of SO2 and NO2 in 2008–2018 and O3 in 2015–2018 in the relatively less populated ecological and economic zones of Western Sichuan Plateau, Southwest China were analyzed. In 2008–2018, the variations in SO2 and NO2 in the ecological zone were not significant, but Ganzi showed a slight upward trend. SO2 decreased significantly in the economic zone, especially in Panzhihua, where NO2 changes were not obvious. From 2015 to 2018, the concentration of O3 in the ecological zone increased significantly, while the economic zone showed a downward trend. The rising trend of the concentration ratio of SO2 to NO2 in the ecological zone and the declining trend in the economic zone indicate that the energy consumption structure of these two zones is quite different. The lower correlation coefficients between NO2 and O3 in the Western Sichuan Plateau imply that the variations of O3 are mainly affected by the regional background. The effects of meteorological factors on SO2, NO2, and O3 were more obvious in the economic zone where there are high anthropometric emissions.


Introduction
In recent decades, China's economy has developed rapidly. Especially in the 21st century, the emission of artificial air pollutants has increased significantly, and hazy weather is frequent; therefore, the air pollution problem has attracted people's attention. The Sichuan Basin is one of the areas in China suffering from serious air pollution due to its special weather and topography [1][2][3][4]. Sulfur dioxide (SO 2 ), nitrogen dioxide (NO 2 ), and ozone (O 3 ) are important pollutants that not only affect the generation of atmospheric fine particles and the photochemical and atmospheric acidification processes but also threaten human health [5][6][7][8]. Many previous studies [9][10][11] have carried out quantitative analysis of SO 2 , NO 2 , and O 3 pollutants, discussed the diurnal and seasonal variation characteristics, and pointed out that SO 2 and NO 2 have the lowest concentration in summer due to The related statistical data of the basic situation of each city were collected from the statistical yearbook provided by the Sichuan provincial bureau (http://tjj.sc.gov.cn/tjnj). Table 1 shows the region category, area, resident population, urbanization rate, GDP, and possession of civil motor vehicles of the study area at the end of 2016. The urbanization rates of Aba, Ganzi, and Liangshan are relatively low, at 37.86%, 29.26%, and 33.04%, respectively. GDP and the number of vehicles in Liangshan and Panzhihua are obviously higher than in Aba and Ganzi, which means there are more man-made air pollutants in the economic zone (Liangshan and Panzhihua). Table 1. Resident population, urbanization rate, population density, gross domestic product (GDP), possession of civil motor vehicles and area in the ecological and economic zones of Western Sichuan Plateau.

Temporal Variations of SO2, NO2, and O3
The variations of SO2 and NO2 from 2008 to 2018 and O3 from 2015 to 2018 in the ecological and economic zones of Western Sichuan Plateau based on monthly average mass concentration are shown in Figure 2. The variations of SO2 and NO2 were divided into two periods before and after 2012. SO2 and NO2 in Aba showed a downward trend before and after 2012, and the downward trend was more The related statistical data of the basic situation of each city were collected from the statistical yearbook provided by the Sichuan provincial bureau (http://tjj.sc.gov.cn/tjnj). Table 1 shows the region category, area, resident population, urbanization rate, GDP, and possession of civil motor vehicles of the study area at the end of 2016. The urbanization rates of Aba, Ganzi, and Liangshan are relatively low, at 37.86%, 29.26%, and 33.04%, respectively. GDP and the number of vehicles in Liangshan and Panzhihua are obviously higher than in Aba and Ganzi, which means there are more man-made air pollutants in the economic zone (Liangshan and Panzhihua). Table 1. Resident population, urbanization rate, population density, gross domestic product (GDP), possession of civil motor vehicles and area in the ecological and economic zones of Western Sichuan Plateau.

Temporal Variations of SO 2 , NO 2 , and O 3
The variations of SO 2 and NO 2 from 2008 to 2018 and O 3 from 2015 to 2018 in the ecological and economic zones of Western Sichuan Plateau based on monthly average mass concentration are shown in Figure 2. The variations of SO 2 and NO 2 were divided into two periods before and after 2012. SO 2 and NO 2 in Aba showed a downward trend before and after 2012, and the downward trend was more prominent after 2012. However, SO 2 and NO 2 in Ganzi increased before 2012 and decreased after 2012. 4 of 19 O 3 in the ecological zone of the Western Sichuan Plateau, Aba and Ganzi, showed a significant upward trend, with a rising rate of 0.303 µg·m −3 /month and 0.154 µg·m −3 /month, respectively. Ma et al. [28] reported an increasing trend of surface O 3 in a rural site north of eastern China and indicated that meteorological factors had little influence on long-term O 3 change, which was completely related to emissions. SO 2 and NO 2 in Liangshan showed a trend of first rising and then falling before and after 2012, and the decline of SO 2 was particularly significant after 2012. SO 2 in Panzhihua increased significantly before 2012 and decreased after 2012, while NO 2 showed a slight decline and increase trend before and after 2012. The declining trend of O 3 in Liangshan was very prominent, with a declining rate of 0.36 µg·m −3 /month, while the change range of O 3 in Panzhihua, with the highest urbanization rate, was not obvious. In general, the changes of SO 2 and NO 2 in the ecological zone of Western Sichuan Plateau were weak, while the decrease of SO 2 in the economic zone was very significant and the variation trend of O 3 in the ecological and economic zones was opposite.  [28] reported an increasing trend of surface O3 in a rural site north of eastern China and indicated that meteorological factors had little influence on long-term O3 change, which was completely related to emissions. SO2 and NO2 in Liangshan showed a trend of first rising and then falling before and after 2012, and the decline of SO2 was particularly significant after 2012. SO2 in Panzhihua increased significantly before 2012 and decreased after 2012, while NO2 showed a slight decline and increase trend before and after 2012. The declining trend of O3 in Liangshan was very prominent, with a declining rate of 0.36 μg·m −3 /month, while the change range of O3 in Panzhihua, with the highest urbanization rate, was not obvious. In general, the changes of SO2 and NO2 in the ecological zone of Western Sichuan Plateau were weak, while the decrease of SO2 in the economic zone was very significant and the variation trend of O3 in the ecological and economic zones was opposite. For further analysis, Figure 3 shows the annual variation trend of SO2, NO2, and O3 in the ecological and economic zones of Western Sichuan Plateau, based on the annual average mass concentrations of SO2 and NO2 from 2008 to 2018, and O3 from 2015 to 2018. The concentrations of SO2 and NO2 in the ecological area were relatively lower due to the small influence of human activity.  For further analysis, Figure 3 shows the annual variation trend of SO 2 , NO 2 , and O 3 in the ecological and economic zones of Western Sichuan Plateau, based on the annual average mass concentrations of SO 2 and NO 2 from 2008 to 2018, and O 3 from 2015 to 2018. The concentrations of SO 2 and NO 2 in the ecological area were relatively lower due to the small influence of human activity. The annual variation characteristics of SO 2  The biggest difference between annual variation characteristics of SO 2 and NO 2 in the economic and ecological zones was that the annual average concentration of SO 2 was almost always higher than that of NO 2 . In 2008-2016, the mean SO 2 concentration in Liangshan was always higher than that of NO 2 . The mean SO 2  In general, the annual variation of gas pollutants in the Western Sichuan Plateau, a relatively clean area, was different from that in large and medium-sized cities. The overall variations of SO 2 and NO 2 in the Western Sichuan Plateau were not very large, showing a trend of first increasing and then decreasing.
Overall, SO 2 and NO 2 in Aba, Ganzi, and Liangshan showed a small change range, and both showed a trend of first increasing and then decreasing in 2008-2018. However, after China's 11th Five-Year Plan (2006-2010), SO 2 concentration in large and medium-sized cities [14,29,30] showed an obvious downward trend, while NO 2 showed a fluctuating or rising trend. The significant decrease of SO 2 concentration in Panzhihua from 2012 to 2015 was also related to the emission reduction measures in China's 12th Five-Year Plan (2011)(2012)(2013)(2014)(2015). This shows that after the 11th Five-Year Plan, the emission reduction requirements for SO 2 and NO 2 proposed by the government had no obvious impact on the relatively clean Western Sichuan Plateau. The annual variations of O 3 in the ecological and economic zones of the Western Sichuan Plateau were like those in Eastern China. Li et al. [20] showed that O 3 increased significantly during the period of 2013-2017 in the North China Plain, the Pearl River Delta, Chengdu-Chongqing urban agglomeration, and Southeastern China, while the O 3 pollution in parts of the northwest, southwest, and northeast of the country was relatively light.      The average mass concentration ratio of SO2 and NO2 (SO2/NO2) can represent the energy structure characteristics of a local region [31]. For example, in regions with more developed industries, a large amount of sulfur-containing fuel is consumed and SO2 is emitted, thus, SO2/NO2 is larger. If an area has a mass of cars and emits more NOx, the ratio is relatively small. In the 1990s, SO2/NO2 in North America and some eastern and central European countries exceeded 1, indicating that these regions had developed industries and the emissions were mainly SO2 [31]. Figure 5 shows the change characteristics of SO2/NO2 based on monthly average concentration in the ecological and economic zones of Western Sichuan Plateau from 2008 to 2018. As can be seen, SO2/NO2 of Aba and Ganzi in the ecological zone showed an upward trend, while SO2/NO2 of Liangshan and Panzhihua in the economic zone showed a downward trend. From 2008 to 2018, SO2/NO2 of Aba fluctuated and showed a slight upward trend, with an average increase of 8.9 × 10 −5 /month. Overall, SO2/NO2 of Ganzi was smaller than that of Aba, and the overall SO2/NO2 fluctuated and had a more obvious upward trend, with an average increase of 0.002/month, and SO2/NO2 stayed below 1.0 most of the time. SO2/NO2 in both Liangshan and Panzhihua showed an obvious downward trend, with an  The average mass concentration ratio of SO 2 and NO 2 (SO 2 /NO 2 ) can represent the energy structure characteristics of a local region [31]. For example, in regions with more developed industries, a large amount of sulfur-containing fuel is consumed and SO 2 is emitted, thus, SO 2 /NO 2 is larger. If an area has a mass of cars and emits more NO x , the ratio is relatively small. In the 1990s, SO 2 /NO 2 in North America and some eastern and central European countries exceeded 1, indicating that these regions had developed industries and the emissions were mainly SO 2 [31]. Figure 5 shows the change characteristics of SO 2 /NO 2 based on monthly average concentration in the ecological and economic zones of Western Sichuan Plateau from 2008 to 2018. As can be seen, SO 2 /NO 2 of Aba and Ganzi in the ecological zone showed an upward trend, while SO 2 /NO 2 of Liangshan and Panzhihua in the economic zone showed a downward trend. From 2008 to 2018, SO 2 /NO 2 of Aba fluctuated and showed a slight upward trend, with an average increase of 8.9 × 10 −5 /month. Overall, SO 2 /NO 2 of Ganzi was smaller than that of Aba, and the overall SO 2 /NO 2 fluctuated and had a more obvious upward trend, with an average increase of 0.002/month, and SO 2 /NO 2 stayed below 1.0 most of the time. SO 2 /NO 2 in both Liangshan and Panzhihua showed an obvious downward trend, with an average decline of −0.007/month in Liangshan and a slightly higher rate in Panzhihua, −0.008/month. According to the annual average concentration ratio of SO 2 and NO 2 (not shown), before 2017, the annual SO 2 /NO 2 in Liangshan was greater than 1.0, and the maximum value (about 1.88) appeared in 2011, while in 2017 and 2018, SO 2 /NO 2 was less than 1.0 (0.73 and 0.80, respectively). The annual SO 2 /NO 2 in Panzhihua was higher than that in Liangshan; from 2008 to 2012, it showed a quasi-linear growth trend, increasing from 1.71 in 2008 to 2.17 in 2012, and then significantly decreasing, reaching the lowest value of about 0.99 in 2017.
The rising trend of SO 2 /NO 2 in the Western Sichuan Plateau ecological area and the declining trend in the economic area indicates that the energy consumption structure of these two areas is quite different. The rising trend of the ecological area indicates that industrial SO 2 emission of in the area have increased in the past 11 years compared with NO x emissions from transportation. However, the declining trend of SO 2 /NO 2 in the economic zone in the past decade is similar to that of some large and medium-sized cities in eastern China, indicating that the SO 2 emission reduction effect in this region was relatively obvious, while the NO x emission reduction effect was not significant. It is worth noting that China's 11th Five-Year Plan (2011-2015) put forward specific SO 2 and NO x emission reduction targets, since SO 2 /NO 2 in most Chinese cities showed a declining trend, and SO 2 /NO 2 in the ecological zone of Western Sichuan Plateau, which is cleaner than the large and medium-sized cities, showed an increasing trend. This suggests that the SO 2 pollution level in clean areas is relatively light, but its growth trend cannot be ignored. The rising trend of SO2/NO2 in the Western Sichuan Plateau ecological area and the declining trend in the economic area indicates that the energy consumption structure of these two areas is quite different. The rising trend of the ecological area indicates that industrial SO2 emission of in the area have increased in the past 11 years compared with NOx emissions from transportation. However, the declining trend of SO2/NO2 in the economic zone in the past decade is similar to that of some large and medium-sized cities in eastern China, indicating that the SO2 emission reduction effect in this region was relatively obvious, while the NOx emission reduction effect was not significant. It is worth noting that China's 11th Five-Year Plan (2011-2015) put forward specific SO2 and NOx emission reduction targets, since SO2/NO2 in most Chinese cities showed a declining trend, and SO2/NO2 in the ecological zone of Western Sichuan Plateau, which is cleaner than the large and medium-sized cities, showed an increasing trend. This suggests that the SO2 pollution level in clean areas is relatively light, but its growth trend cannot be ignored. Due to the lack of emissions data of SO2 and NOx in the ecological and economic zones of Western Sichuan Plateau, the energy consumption of unit added industrial value and the passenger and freight turnover of highways were selected to indirectly characterize the emission change characteristics of SO2 and NOx, respectively. As can be seen from Figure 6a, from 2008 to 2016, the energy consumption of unit added industrial value in Aba first increased and then decreased, while other regions showed a significant downward trend, indicating that industrial consumption of standard coal equivalent (SCE) in other regions except Aba declined year by year. Compared with the change characteristics of SO2 concentration from 2008 to 2018 (Figure 3), it was found that the variations of SO2 are not consistent with industrial emissions, indicating that in these relatively clean areas, other sources besides industrial emissions will also contribute significantly to SO2; for instance, biomass combustion in Qinghai Plateau [32] makes an important contribution to SO2. The passenger and freight turnover of highways (Figure 6b) shows an overall upward trend, first increasing and then decreasing before and after 2013, which is close to the changing trend of NO2 concentration Due to the lack of emissions data of SO 2 and NO x in the ecological and economic zones of Western Sichuan Plateau, the energy consumption of unit added industrial value and the passenger and freight turnover of highways were selected to indirectly characterize the emission change characteristics of SO 2 and NO x , respectively. As can be seen from Figure 6a, from 2008 to 2016, the energy consumption of unit added industrial value in Aba first increased and then decreased, while other regions showed a significant downward trend, indicating that industrial consumption of standard coal equivalent (SCE) in other regions except Aba declined year by year. Compared with the change characteristics of SO 2 concentration from 2008 to 2018 (Figure 3), it was found that the variations of SO 2 are not consistent with industrial emissions, indicating that in these relatively clean areas, other sources besides industrial emissions will also contribute significantly to SO 2 ; for instance, biomass combustion in Qinghai Plateau [32] makes an important contribution to SO 2 . The passenger and freight turnover of highways ( Figure 6b) shows an overall upward trend, first increasing and then decreasing before and after 2013, which is close to the changing trend of NO 2 concentration (Figure 3), indicating that motor vehicle emissions made a prominent contribution to the change in NO 2 concentration.

Effects of Meteorology on SO2, NO2 and O3
Emissions and meteorological conditions are predominant in the air pollutant trend; emissions exceeding atmospheric environmental capacity is the basic cause of air pollution [33,34]. In order to analyze the influence of meteorological factors on gaseous air pollutants in the ecological and economic zones of Western Sichuan Plateau, the relationships between SO2, NO2, O3, and routine meteorological parameters (mean monthly average temperature, maximum and minimum monthly average temperature, mean relative humidity, monthly total precipitation, monthly sunshine duration, mean wind speed, and maximum wind speed) were discussed by using monthly mean meteorological data from 2008 to 2017. Figure 7 shows these relationships in Aba, in the ecological zone. SO2 and NO2 concentrations were less dependent on meteorological parameters in the ecological zone with less anthropogenic emission. The relationship between O3 and meteorological parameters in Figure 7 implies that a high concentration of O3 was more likely to form under the meteorological conditions of high temperature, low relative humidity, long sunshine duration, and especially high wind speed.
The relationships of SO2, NO2, O3, and meteorological parameters in Panzhihua of the economic zone of Western Sichuan Plateau are shown in Figure 8. In the economic zone, with more man-made emissions than the ecological zone, gaseous pollutants are more significantly affected by meteorological parameters with a higher coefficient of determination. SO2 and NO2 decreased with increased temperature, humidity, precipitation, and wind speed. O3 was positively correlated with temperature, sunshine duration, and wind speed and negatively correlated with relative humidity. The dependence of gaseous pollutants on meteorological factors was more prominent in the economic zone than in the ecological zone, which may have to do with local emissions and special topography [21].

Effects of Meteorology on SO 2 , NO 2 and O 3
Emissions and meteorological conditions are predominant in the air pollutant trend; emissions exceeding atmospheric environmental capacity is the basic cause of air pollution [33,34]. In order to analyze the influence of meteorological factors on gaseous air pollutants in the ecological and economic zones of Western Sichuan Plateau, the relationships between SO 2 , NO 2 , O 3 , and routine meteorological parameters (mean monthly average temperature, maximum and minimum monthly average temperature, mean relative humidity, monthly total precipitation, monthly sunshine duration, mean wind speed, and maximum wind speed) were discussed by using monthly mean meteorological data from 2008 to 2017. Figure 7 shows these relationships in Aba, in the ecological zone. SO 2 and NO 2 concentrations were less dependent on meteorological parameters in the ecological zone with less anthropogenic emission. The relationship between O 3 and meteorological parameters in Figure 7 implies that a high concentration of O 3 was more likely to form under the meteorological conditions of high temperature, low relative humidity, long sunshine duration, and especially high wind speed.
The relationships of SO 2 , NO 2 , O 3 , and meteorological parameters in Panzhihua of the economic zone of Western Sichuan Plateau are shown in Figure 8. In the economic zone, with more man-made emissions than the ecological zone, gaseous pollutants are more significantly affected by meteorological parameters with a higher coefficient of determination. SO 2 and NO 2 decreased with increased temperature, humidity, precipitation, and wind speed. O 3 was positively correlated with temperature, sunshine duration, and wind speed and negatively correlated with relative humidity. The dependence of gaseous pollutants on meteorological factors was more prominent in the economic zone than in the ecological zone, which may have to do with local emissions and special topography [21].

Figure 7.
Relationships between SO2, NO2, O3 and routine meteorological parameters (mean monthly average temperature, maximum and minimum monthly average temperature, mean relative humidity, monthly total precipitation, monthly sunshine duration, mean wind speed, and maximum wind speed) in Aba of the ecological zone of Western Sichuan Plateau.  Figure 7. Relationships between SO 2 , NO 2 , O 3 and routine meteorological parameters (mean monthly average temperature, maximum and minimum monthly average temperature, mean relative humidity, monthly total precipitation, monthly sunshine duration, mean wind speed, and maximum wind speed) in Aba of the ecological zone of Western Sichuan Plateau. Figure 8. Relationships between SO2, NO2, O3, and routine meteorological parameters (mean monthly average temperature, maximum and minimum monthly average temperature, mean relative humidity, monthly total precipitation, monthly sunshine duration, mean wind speed, and maximum wind speed) in Panzhihua of the economic zone of Western Sichuan Plateau. MaM wind speed (m s -1 ) Figure 8. Relationships between SO 2 , NO 2 , O 3 , and routine meteorological parameters (mean monthly average temperature, maximum and minimum monthly average temperature, mean relative humidity, monthly total precipitation, monthly sunshine duration, mean wind speed, and maximum wind speed) in Panzhihua of the economic zone of Western Sichuan Plateau. Figure 9 shows the multi-year average mass concentrations of SO 2 , NO 2 , and O 3 in the Western Sichuan Plateau ecological and economic zones. The average mass concentrations of O 3 in Aba, Ganzi, Liangshan, and Panzhihua were significantly higher than that of SO 2 and NO 2 . Liangshan suffered the most serious O 3 pollution of the four regions, with a mean mass concentration of 96.25 ± 28.17 µg·m −3 , followed by Panzhihua, 83.96 ± 29.17 µg·m −3 . O 3 pollution in the ecological region, at higher altitude, was lighter than in the economic zone, with mean mass concentrations of 79.99 ± 23.41 µg·m −3 and 72.61 ± 28.28 µg·m −3 , respectively, in Aba and Ganzi. The mean mass concentrations of SO 2 and NO 2 in the ecological zone were lower than in the economic zone. The mean mass concentration of SO 2 in Aba and Ganzi was similar, at 12.39 ± 8.31 µg·m −3 and 15.02 ± 8.51 µg·m −3 , respectively, and in Panzhihua, it was significantly higher than in Liangshan, at 60.05 ± 35.25 µg·m −3 and 32.73 ± 18.02 µg·m −3 , respectively. Among the four regions, Panzhihua has the most serious SO 2 pollution, and Aba has the least, the multi-year average concentration of SO 2 in Panzhihua is about five times that in Aba. For NO 2 , the average mass concentration was highest in Panzhihua, at 37.81 ± 11.98 µg·m −3 and lowest in Aba, at 14.37 ± 7.07 µg·m −3 . The level of NO 2 pollution in Ganzi and Liangshan was moderate, and the average concentration of NO 2 was similar, 24.99±11.59 µg·m −3 and 23.9 ± 7.17 µg·m −3 , respectively. Zhao et al. [21] analyzed the characteristics of six criteria for air pollutants in the Sichuan basin in 2015-2017, and showed that the average SO 2 concentration in Chengdu, the provincial city, was close to that in the ecological zone due to effective desulfurization measures and strict SO 2 emission reduction requirements, while the average NO 2 and O 3 concentration in Chengdu was significantly higher than that in the ecological and economic zones in this study.  Figure 9 shows the multi-year average mass concentrations of SO2, NO2, and O3 in the Western Sichuan Plateau ecological and economic zones. The average mass concentrations of O3 in Aba, Ganzi, Liangshan, and Panzhihua were significantly higher than that of SO2 and NO2. Liangshan suffered the most serious O3 pollution of the four regions, with a mean mass concentration of 96.25 ± 28.17 μg·m −3 , followed by Panzhihua, 83.96 ± 29.17 μg·m − 3. O3 pollution in the ecological region, at higher altitude, was lighter than in the economic zone, with mean mass concentrations of 79.99 ± 23.41 μg·m −3 and 72.61 ± 28.28 μg·m −3 , respectively, in Aba and Ganzi. The mean mass concentrations of SO2 and NO2 in the ecological zone were lower than in the economic zone. The mean mass concentration of SO2 in Aba and Ganzi was similar, at 12.39 ± 8.31 μg·m −3 and 15.02±8.51 μg·m − 3, respectively, and in Panzhihua, it was significantly higher than in Liangshan, at 60.05 ± 35.25 μg·m − 3 and 32.73 ± 18.02 μg·m − 3, respectively. Among the four regions, Panzhihua has the most serious SO2 pollution, and Aba has the least, the multi-year average concentration of SO2 in Panzhihua is about five times that in Aba. For NO2, the average mass concentration was highest in Panzhihua, at 37.81 ± 11.98 μg·m − 3 and lowest in Aba, at 14.37 ± 7.07 μg·m −3 . The level of NO2 pollution in Ganzi and Liangshan was moderate, and the average concentration of NO2 was similar, 24.99±11.59 μg·m −3 and 23.9 ± 7.17 μg·m −3 , respectively. Zhao et al. [21] analyzed the characteristics of six criteria for air pollutants in the Sichuan basin in 2015-2017, and showed that the average SO2 concentration in Chengdu, the provincial city, was close to that in the ecological zone due to effective desulfurization measures and strict SO2 emission reduction requirements, while the average NO2 and O3 concentration in Chengdu was significantly higher than that in the ecological and economic zones in this study.   The average concentration of SO 2 in Ganzi and Liangshan was the same, with a seasonal distribution of winter > autumn > spring > summer. The average concentration was 17.4 ± 9.7 µg·m −3 and 13.94 ± 7.28 µg·m −3 in Ganzi and was 38.46 ± 21.88 µg·m −3 and 28.87 ± 15.15 µg·m −3 in Liangshan in winter and summer, respectively. The maximum concentration of SO 2 in Aba occurred in winter, while the minimum concentration occurred in autumn, at 13.84 ± 9.74 µg·m −3 and 11.24 ± 7.38 µg·m −3 ; the seasonal distribution was winter > spring > summer > autumn. The mean concentration of SO 2 in Panzhihua was the highest in winter, 75.44 ± 45.67 µg·m −3 , and the lowest in spring 49.99 ± 25.32 µg·m −3 .

Characteristics of SO2, NO2, and O3
Ganzi, Liangshan, and Panzhihua had the same NO 2 seasonal distribution of winter > autumn > spring > summer. The difference in NO 2 mean concentration in Ganzi and Liangshan between the four seasons was not significant; the maximum and minimum concentrations in Ganzi were 27.16 ± 13.5 µg·m −3 and 22.66 ± 8.68 µg·m −3 , respectively, and the difference between winter and summer was about 4.5 µg·m −3 . The maximum and minimum NO 2 mean concentrations in Liangshan were 27.72 ± 8.93 µg·m −3 and 21.37 ± 5.31 µg·m −3 , respectively, and the difference in concentration between winter and summer was about 6.3 µg·m −3 . The average concentration of NO 2 in Panzhihua in the four seasons was significantly different; the maximum concentration was 47.14 ± 11.88 µg·m −3 in winter, the minimum concentration was 30.89 ± 8.13 µg·m −3 in summer, and the difference was about 16.2 µg·m −3 . The seasonal distribution of NO 2 in Aba was different from the other regions, with an average concentration of 15.39 ± 7.57 µg·m −3 in spring and the lowest concentration of 12.71 ± 6.19 µg·m −3 in autumn. The average concentration of SO2 in Ganzi and Liangshan was the same, with a seasonal distribution of winter > autumn > spring > summer. The average concentration was 17.4 ± 9.7 μg·m −3 and 13.94 ± 7.28 μg·m −3 in Ganzi and was 38.46 ± 21.88 μg·m −3 and 28.87 ± 15.15 μg·m −3 in Liangshan in winter and summer, respectively. The maximum concentration of SO2 in Aba occurred in winter, while the minimum concentration occurred in autumn, at 13.84 ± 9.74 μg·m −3 and 11.24 ± 7.38 μg·m −3 ; the seasonal distribution was winter > spring > summer > autumn. The mean concentration of SO2 in Panzhihua was the highest in winter, 75.44 ± 45.67 μg·m −3 , and the lowest in spring 49.99 ± 25.32 μg·m −3 .
Ganzi, Liangshan, and Panzhihua had the same NO2 seasonal distribution of winter > autumn > spring > summer. The difference in NO2 mean concentration in Ganzi and Liangshan between the four seasons was not significant; the maximum and minimum concentrations in Ganzi were 27.16 ± 13.5 μg·m −3 and 22.66 ± 8.68 μg·m −3 , respectively, and the difference between winter and summer was about 4.5 μg·m −3 . The maximum and minimum NO2 mean concentrations in Liangshan were 27.72 ± 8.93 μg·m −3 and 21.37 ± 5.31 μg·m −3 , respectively, and the difference in concentration between winter and summer was about 6.3 μg·m −3 . The average concentration of NO2 in Panzhihua in the four seasons was significantly different; the maximum concentration was 47.14 ± 11.88 μg·m −3 in winter, the minimum concentration was 30.89 ± 8.13 μg·m −3 in summer, and the difference was about 16.2 μg·m −3 . The seasonal distribution of NO2 in Aba was different from the other regions, with an average concentration of 15.39 ± 7.57 μg·m −3 in spring and the lowest concentration of 12.71 ± 6.19 μg·m −3 in autumn.   The maximum and minimum concentrations of O 3 in Panzhihua occurred in April and December, at 109.65 ± 20.66 µg·m −3 and 56.17 ± 16.49 µg·m −3 , respectively. The differences between maximum and minimum monthly concentrations in the economic zone were significantly higher than in the ecological zone, among which the difference in Panzhihua was the largest, about 53.48 µg·m −3 , and the difference in Ganzi was the smallest, about 38.5 µg·m −3 , indicating that the monthly change of O 3 with a longer residence time was more significant in the area with more artificial emissions than in the clean area with less artificial emissions.  −3 , respectively. The differences between maximum and minimum monthly concentrations in the economic zone were significantly higher than in the ecological zone, among which the difference in Panzhihua was the largest, about 53.48 μg·m −3 , and the difference in Ganzi was the smallest, about 38.5 μg·m −3 , indicating that the monthly change of O3 with a longer residence time was more significant in the area with more artificial emissions than in the clean area with less artificial emissions. The monthly mean concentrations of SO2 in the ecological zone did not change significantly. The maximum and minimum concentrations in Aba appeared in December and September, respectively, and the difference was about 4.76 μg·m −3 , while the maximum and minimum concentrations in Ganzi appeared in December and June, respectively, and the difference was about 4.9 μg·m −3 . Artificial emissions were more prominent in the economic zone, and the monthly variations of average concentrations of SO2 were more obvious. The maximum and minimum SO2 concentration in Liangshan occurred in January and May, with a difference of 14.43 μg·m −3 , and in Panzhihua occurred in January and May, respectively, at 86.01 ± 49.78 μg·m −3 and 44.96 ± 23.54 μg·m −3 , with a difference of 41.05 μg·m −3 . The range of variation of monthly average NO2 concentration in Aba was the lowest in the ecological and economic zones; the maximum and minimum concentrations appeared in June and September, with a difference of 6.05 μg·m −3 , and the maximum and minimum concentrations in Ganzi appeared in December and February, with a difference of 9.82 μg·m −3 , respectively. The monthly average NO2 concentration in Panzhihua showed the most prominent variation characteristics, with a mean maximum of 50.82 ± 12.79 μg·m −3 in January, minimum of 30.21 ± 9.79 The monthly mean concentrations of SO 2 in the ecological zone did not change significantly. The maximum and minimum concentrations in Aba appeared in December and September, respectively, and the difference was about 4.76 µg·m −3 , while the maximum and minimum concentrations in Ganzi appeared in December and June, respectively, and the difference was about 4.9 µg·m −3 . Artificial emissions were more prominent in the economic zone, and the monthly variations of average concentrations of SO 2 were more obvious. The maximum and minimum SO 2 concentration in Liangshan occurred in January and May, with a difference of 14.43 µg·m −3 , and in Panzhihua occurred in January and May, respectively, at 86.01 ± 49.78 µg·m −3 and 44.96 ± 23.54 µg·m −3 , with a difference of 41.05 µg·m −3 . The range of variation of monthly average NO 2 concentration in Aba was the lowest in the ecological and economic zones; the maximum and minimum concentrations appeared in June and September, with a difference of 6.05 µg·m −3 , and the maximum and minimum concentrations in Ganzi appeared in December and February, with a difference of 9.82 µg·m −3 , respectively. The monthly average NO 2 concentration in Panzhihua showed the most prominent variation characteristics, with a mean maximum of 50.82 ± 12.79 µg·m −3 in January, minimum of 30.21 ± 9.79 µg·m −3 in June, and a difference of 20.61 µg·m −3 . The maximum and minimum monthly average concentrations of NO 2 in Liangshan occurred in December and August, respectively, with a difference of 9.01 µg·m −3 .

Correlation between SO 2 , NO 2 , and O 3
It is generally believed that CO and NO x pollutants are discharged by mobile sources, while SO 2 and NO x pollutants are discharged by point sources [35]. Since mobile sources do not discharge a large amount of SO 2 and point sources produce both SO 2 and NO 2 , if the correlation between SO 2 and NO 2 is high, the point source is more prominent. On the other hand, if the correlation is low, the mobile source is more prominent. Figure 12 shows the correlation between SO 2 and NO 2 in Aba, Ganzi, Liangshan, and Panzhihua. SO 2 and NO 2 in the ecological and economic zones show a positive correlation with a significance test of 99% confidence. The correlation between SO 2 and NO 2 in Ganzi is the highest, with a Pearson coefficient r 0.55, followed by Panzhihua, with r = 0.47, which indicates that the emission characteristics of SO 2 and NO 2 in these two regions are relatively similar, SO 2 and NO 2 are discharged mainly from point sources and point sources contribute more to local air pollutants than mobile sources. The lowest correlation between SO 2 and NO 2 is found in Aba, with a correlation coefficient of 0.27, while that in Liangshan is slightly higher, with r = 0.31, indicating that the contribution of point sources to SO 2 and NO 2 pollution in these two areas is not very prominent. In general, although SO 2 and NO 2 in the ecological and economic zones were responsible less pollution than in other large and medium-sized cities in China, the emission source characteristics of SO 2 and NO 2 in this region were different from those of the other cities. Mao et al. [36] reported a strong positive correlation between SO 2 and NO 2 in Chongqing, Wuhan, and Nanjing, three metropolises along the Yangtze River, indicating the similar origins and elimination processes of SO 2 and NO 2 . μg·m −3 in June, and a difference of 20.61 μg·m −3 . The maximum and minimum monthly average concentrations of NO2 in Liangshan occurred in December and August, respectively, with a difference of 9.01 μg·m −3 .

Correlation between SO2, NO2, and O3
It is generally believed that CO and NOx pollutants are discharged by mobile sources, while SO2 and NOx pollutants are discharged by point sources [35]. Since mobile sources do not discharge a large amount of SO2 and point sources produce both SO2 and NO2, if the correlation between SO2 and NO2 is high, the point source is more prominent. On the other hand, if the correlation is low, the mobile source is more prominent. Figure 12 shows the correlation between SO2 and NO2 in Aba, Ganzi, Liangshan, and Panzhihua. SO2 and NO2 in the ecological and economic zones show a positive correlation with a significance test of 99% confidence. The correlation between SO2 and NO2 in Ganzi is the highest, with a Pearson coefficient r 0.55, followed by Panzhihua, with r = 0.47, which indicates that the emission characteristics of SO2 and NO2 in these two regions are relatively similar, SO2 and NO2 are discharged mainly from point sources and point sources contribute more to local air pollutants than mobile sources. The lowest correlation between SO2 and NO2 is found in Aba, with a correlation coefficient of 0.27, while that in Liangshan is slightly higher, with r = 0.31, indicating that the contribution of point sources to SO2 and NO2 pollution in these two areas is not very prominent. In general, although SO2 and NO2 in the ecological and economic zones were responsible less pollution than in other large and medium-sized cities in China, the emission source characteristics of SO2 and NO2 in this region were different from those of the other cities. Mao et al. [36] reported a strong positive correlation between SO2 and NO2 in Chongqing, Wuhan, and Nanjing, three metropolises along the Yangtze River, indicating the similar origins and elimination processes of SO2 and NO2. O3 is an important secondary gaseous pollutant in the urban atmosphere, and NOx plays a very important role in its formation as the main precursor. The formation of surface O3 is highly dependent on solar radiation intensity, the absolute concentration of O3 and volatile organic compounds (VOCs), and the ratio of O3 to VOCs, while the local surface concentration of O3 is affected by meteorological elements, local precursor emissions, and the close and long-distance transport of O3 and precursor [37][38][39]. As shown in Figure 13a,b, the correlations between NO2 and O3 in the ecological and economic zones of Western Sichuan Plateau were analyzed by using the daily mass concentration from 2015 to 2018. The correlations between NO2 and O3 in the four regions were negatively correlated, and all of them passed the significance test of 99% confidence. Among the four regions, Ganzi has the highest Pearson correlation coefficient (−0.34), indicating that this region is where NO2 contributes the most to O3 generation as a precursor. Aba, Liangshan, and Panzhihua showed a relatively low correlation between NO2 and O3, with a Pearson correlation coefficient of −0.24, −0.22, O 3 is an important secondary gaseous pollutant in the urban atmosphere, and NO x plays a very important role in its formation as the main precursor. The formation of surface O 3 is highly dependent on solar radiation intensity, the absolute concentration of O 3 and volatile organic compounds (VOCs), and the ratio of O 3 to VOCs, while the local surface concentration of O 3 is affected by meteorological elements, local precursor emissions, and the close and long-distance transport of O 3 and precursor [37][38][39]. As shown in Figure 13a,b, the correlations between NO 2 and O 3 in the ecological and economic zones of Western Sichuan Plateau were analyzed by using the daily mass concentration from 2015 to 2018. The correlations between NO 2 and O 3 in the four regions were negatively correlated, and all of them passed the significance test of 99% confidence. Among the four regions, Ganzi has the highest Pearson correlation coefficient (−0.34), indicating that this region is where NO 2 contributes the most to O 3 generation as a precursor. Aba, Liangshan, and Panzhihua showed a relatively low correlation between NO 2 and O 3 , with a Pearson correlation coefficient of −0.24, −0.22, and −0.26, respectively. The correlations between NO 2 and O 3 in the metropolitan areas of China, such as the Beijing-Tianjin-Hebei region, Yangtze River Delta, Pearl River Delta, and Chengdu-Chongqing city cluster [2,36,40] are significantly higher than in the ecological and economic zones of the Western Sichuan Plateau in our study, indicating that NO 2 contributes more to O 3 in large and medium-sized urban areas with high NO 2 emissions. The contribution of local precursors of O 3 can better explain the formation of O 3 , while in regions with less NO 2 emission, O 3 may be more affected by the regional background. de Souza et al. [41] also indicated that when the correlation between NO 2 and O 3 is high, the local precursor plays a leading contributory role, and when the correlation is low, the variation of O 3 concentration is mainly affected by the O 3 concentration in the regional background. and −0.26, respectively. The correlations between NO2 and O3 in the metropolitan areas of China, such as the Beijing-Tianjin-Hebei region, Yangtze River Delta, Pearl River Delta, and Chengdu-Chongqing city cluster [2,36,40] are significantly higher than in the ecological and economic zones of the Western Sichuan Plateau in our study, indicating that NO2 contributes more to O3 in large and medium-sized urban areas with high NO2 emissions. The contribution of local precursors of O3 can better explain the formation of O3, while in regions with less NO2 emission, O3 may be more affected by the regional background. de Souza et al. [41] also indicated that when the correlation between NO2 and O3 is high, the local precursor plays a leading contributory role, and when the correlation is low, the variation of O3 concentration is mainly affected by the O3 concentration in the regional background.

Conclusions
Based on daily average mass concentration data of SO2 and NO2 from 2008 to 2018 and O3 from 2015 to 2018 in Aba, Ganzi, Liangshan, and Panzhihua in Western Sichuan Plateau, the characteristics and change trend of the three gas pollutants and their correlations in the relatively clean ecological and economic zones were analyzed. The main conclusions are as follows: (1) On the whole, the change ranges of SO2 and NO2 in Aba, Ganzi, and Liangshan were not large, and they all showed a trend of first increasing and then decreasing. The change range of SO2 in Panzhihua was the most obvious, and the decrease range was very significant from 2012 to 2015.
During the period 2015-2018, except for Aba, O3 showed a trend of first declining and then rising in the other regions. SO2/NO2 in the economic zone of Western Sichuan Plateau showed a decreasing trend, while in the relatively clean ecological zone this ratio showed an increasing trend. Although the SO2 pollution level in the clean areas was relatively low, the growth trend could not be ignored. The dependence of SO2 and NO2 on routine meteorological parameters in the ecological zone was low. The influence of meteorological parameters on gaseous pollutants was more significant in the economic zone with high anthropometric emissions. (2) O3 pollution in the Western Sichuan Plateau with less artificial emission was prominent, with the highest annual mean concentration in Liangshan, 96.25 ± 28.17 μg·m −3 , and the lowest annual mean concentration in Ganzi, 72.61 ± 28.28 μg·m −3 , indicating that O3 formation is promoted by strong solar radiation induced by smaller particle concentrations and more cloud-free days [42,43]. The pollution levels of SO2 and NO2 in the ecological zone were lower than those in the economic zone. The annual average concentration of SO2 in Panzhihua and Aba was the highest and lowest, at 12.39 ± 8.31 μg·m −3 and 60.05 ± 35.25 μg·m −3 , respectively. Panzhihua had the highest annual average NO2 concentration, 37.81 ± 11.98 μg·m −3 , while Aba had the lowest annual average concentration, 14.37 ± 7.07 μg·m −3 . The seasonal and monthly variations of O3 in the four regions were more obvious than those of SO2 and NO2. The seasonal and monthly

Conclusions
Based on daily average mass concentration data of SO 2 and NO 2 from 2008 to 2018 and O 3 from 2015 to 2018 in Aba, Ganzi, Liangshan, and Panzhihua in Western Sichuan Plateau, the characteristics and change trend of the three gas pollutants and their correlations in the relatively clean ecological and economic zones were analyzed. The main conclusions are as follows: (1) On the whole, the change ranges of SO 2 and NO 2 in Aba, Ganzi, and Liangshan were not large, and they all showed a trend of first increasing and then decreasing. The change range of SO 2 in Panzhihua was the most obvious, and the decrease range was very significant from 2012 to 2015. During the period 2015-2018, except for Aba, O 3 showed a trend of first declining and then rising in the other regions. SO 2 /NO 2 in the economic zone of Western Sichuan Plateau showed a decreasing trend, while in the relatively clean ecological zone this ratio showed an increasing trend. Although the SO 2 pollution level in the clean areas was relatively low, the growth trend could not be ignored. The dependence of SO 2 and NO 2 on routine meteorological parameters in the ecological zone was low. The influence of meteorological parameters on gaseous pollutants was more significant in the economic zone with high anthropometric emissions. (2) O 3 pollution in the Western Sichuan Plateau with less artificial emission was prominent, with the highest annual mean concentration in Liangshan, 96.25 ± 28.17 µg·m −3 , and the lowest annual mean concentration in Ganzi, 72.61 ± 28.28 µg·m −3 , indicating that O 3 formation is promoted by strong solar radiation induced by smaller particle concentrations and more cloud-free days [42,43]. The pollution levels of SO 2 and NO 2 in the ecological zone were lower than those in the economic zone. The annual average concentration of SO 2 in Panzhihua and Aba was the highest and lowest, at 12.39 ± 8.31 µg·m −3 and 60.05 ± 35.25 µg·m −3 , respectively. Panzhihua had the highest annual average NO 2 concentration, 37.81 ± 11.98 µg·m −3 , while Aba had the lowest annual average concentration, 14.37 ± 7.07 µg·m −3 . The seasonal and monthly variations of O 3 in the four regions were more obvious than those of SO 2 and NO 2 . The seasonal and monthly variations in the economic zone were higher than those in the ecological zone, indicating that man-made emissions have an important impact on the temporal distribution characteristics of air pollutants.
(3) The correlation between SO 2 and NO 2 in Ganzi was the highest, followed by Panzhihua; the emission characteristics of SO 2 and NO 2 in these two regions were similar, and point sources contributed more to local air pollutants than mobile sources. The correlation between SO 2 and NO 2 was the lowest in Aba, indicating that the contribution of point sources to SO 2 and NO 2 pollution was not prominent. Ganzi showed the highest correlation coefficient, suggesting that NO 2 contribute the most to O 3 in this region in the Western Sichuan Plateau, followed by Panzhihua, Aba, and Liangshan. The correlation between NO 2 and O 3 is significantly higher in large and medium-sized cities in China than in the ecological and economic zones of Western Sichuan Plateau, which indicates that local precursors contribute more to the formation of O 3 in areas with high NO 2 emissions, while in areas with low NO 2 emissions, O 3 is more significantly affected by the regional background.