**4. Discussion**

During the study period, the average concentrations of PM2.5, PM10, SO2, NO2, CO, and O3, decreased in 92.5%, 96.2%, 92.5%, 64.5%, 88.7%, and 11.3% of the 53 cities in NWC. Based on the results above, we concluded that strict environmental regulations have significantly improved the air quality in NWC between 2015 and 2018 [3,20–23]. PM2.5 mainly originates from industrial activities, coal consumption, power generation, biomass burning, automobile exhausts, construction activities, road dust, etc. [28–30,50–52]. From 2015 to 2018, the average concentration of PM2.5 decreased in all the cities except a few cities (Changji, Ili, Shihezi, Wujiaqu) in the northern part of XJ. Higher pollution in XJ, SN, and GS, PM2.5 hotspots [53,54], is because of increased coal-based industry, vehicular emission, civil heating, construction activities, natural sources (dust storms), and adverse meteorology [55–60]. PM10 mainly originates from natural sources, e.g., sand storms, haze events, etc., as well as from anthropogenic sources, e.g., developmental activities, industrial emissions, traffic emissions, road dust, etc. [15,26–28]. The highest pollution levels occurred in XJ, followed by SN, NX, GS, and QH. Elevated pollution levels in southern Xinjiang (Kashgar) indicate the influences of emissions from natural sources, e.g., Taklimakan deserts, dust storms, haze events, etc. [61–64]. Similarly, higher particulate pollution in Shaanxi (FWP region) is associated with increased anthropogenic emissions, e.g., industrial activities, construction activities, etc. [55]. All the cities of NWC experienced a reduction in PM10 except Shihezi and Wujiaqu in northern XJ, indicating the influence of both manmade and natural emission sources. This decrease is associated with strict environmental regulations, e.g., Chinese Ambient Air Quality Standards (CAAQS) (GB 3095–2012) [17], Atmospheric Pollution Prevention and Control Action Plan (APPCAP, 2013) [18], technical regulation on ambient air quality index (HJ 633–2012) [19], the establishment of nationwide air quality monitoring stations, etc. [3,20–23].

In the case of gaseous pollutants, the average concentration of SO2, NO2, and CO decreased by 41.9%, 6.19%, and 27.3%, respectively, in NWC. Industrial emissions, coal burning, fossil fuel burning, power generation, traffic exhausts, etc. are major sources of SO2, NO2, and CO [65–69]. The concentration of SO2 decreased in all the cities except Ili, Hami (XJ), Jiuquan (GS), and Yushu (QH), indicating the influence of increased coal combustion, vehicular exhaust, and industrial emission [27,28,57,70]. In the case of NO2, the average concentration decreased in 71.7% of the cities of NWC between 2015 and 2018. The highest pollution level occurred in the provincial capitals, e.g., Xi'an (SN), Urumqi (XJ), Lanzhou (GS), etc., and major cities (Ili, Hami, Jiuquan, Yushu, etc.), indicating increased fossil fuel combustion, e.g., automobile exhaust, industrial emission, etc. [34,63,65–69]. Similarly, the highest concentration of CO occurred in SN, followed by XJ, QH, GS, and NX. Higher CO indicates incomplete combustion of fossil fuels, biomass burning, industrial emission, and causes multiple health disorders, e.g., hypoxia, major heart and neural disorders [28,67,68]. In contrast to other pollutants, O3 increased by 3.69% in NWC during the study period. All the cities experienced an increase in O3 concentration except Ankang, Shanglou, Longnan, Haibei, Hainan, and Haixi. This increase in O3 is associated with a decrease in PM2.5 and other pollutants, which slows down the sink of hydroperoxy radicals and helps in the accumulation of ozone [55].

In terms of seasonality, PM2.5, PM10, SO2, NO2, and CO experienced the same seasonal variation, e.g., highest in winter and lowest in summer. Higher pollution in winter is associated with increased coal combustion, civil heating, power generation, fossil fuel burning, industrial activity, vehicular exhausts, and adverse/stagnant meteorology [13–15,26–30,50–52,71–75]. In the case of PM10, higher pollution levels also occurred from March to May (spring) due to haze events [76,77]. In contrast to other pollutants, the concentration of O3 was highest in summer and lowest in winter [14,28,30,78]. Ozone is a secondary pollutant, formed due to a photochemical reaction between VOCs and NOx [43,44,79]. The concentration of ozone in the summer was approximately twice that in winter due to lower NOx levels in winter as NOx levels decrease the O3 depletion and enhance the accumulation of O3 [80]. Similarly, higher temperatures, e.g., in summer, favor the accumulation of ozone [81–83].

The PM2.5/PM10 ratio reflects air quality, pollution sources, and origin, e.g., a higher PM2.5/PM10 ratio indicates the increased proportion of PM2.5, mainly emitted from anthropogenic activities, and a lower ratio indicates an increased proportion of PM10, mainly from natural activities [28,61–64]. During the study period (2015–2018), the PM2.5/PM10 ratio slightly decreased by 0.43% and in 50.9% of the cities of NWC. This decrease is associated with a reduction in PM2.5 over time. In general, the PM2.5/PM10 ratio was higher in winter (low temperature) as compared with summer (high temperature) due to increased anthropogenic activities that release a significant amount of PM2.5 and stable atmospheric conditions that help the accumulation of pollution [26–29,53,54,84].

During the study period (2015–2018), the AQI improved by 21.3%, and 86.8% of the cities of NWC experienced AQI improvement. This improvement is associated with a reduction in the criteria pollutants over time. AQI crossed the threshold value of 100 in 10 cities, out of which seven cities are in Xinjiang (Kashgar, Hotan, Aksu, Wujiaqu, Crete, Urumqi, Turpan) and three cities are in Shaanxi (Xiangyang, Xi'an, Weinan) (Table S1). The higher AQI values in Xinjiang and Shaanxi are associated with the increased coal-based industry, civil heating, and vehicular exhaust [3,14,30,51,52,56,69,85]. In the case of seasonal variations, the highest AQI occurred in winter due to increased anthropogenic emissions and stable atmospheric conditions [15,58,84]. In NWC, the proportion of AQI "Class I" improved by 114.1%, while the proportion of Class II, Class III, Class IV, Class V, and Class VI decreased by 1.08%, 55.2%, 29.5%, 69.2%, and 58.3%, respectively, during 2015 to 2018. The proportion of AQI "Class I" improved from 12.9% in 2015 to 27.6% in 2018. Similarly, the proportion of AQI "Class I" improved in all the provinces, e.g., SN, XJ, GS, NX, and QH, in all the seasons, and improved by 2.53 times, 2.33 times, 1.63 times, and 2.79 times in spring, summer, autumn, and winter in 2018 as compared with 2015, which indicates improvement in air quality [28,57].

During the study period (2015–2018), the proportion of days with PM10, PM2.5, SO2, and CO as a major pollutant decreased by 35%, 38%, 52%, and 90%, respectively, due to strict environmental legislation [20–23]. PM2.5 was a major pollutant in winter (42%), indicating

anthropogenic emissions, e.g., coal burning, civil heat, industrial emissions, and vehicular emissions [28,56–58]. Similarly, O3 was a major pollutant (44.7%) in summer due to lower NOx levels as lower NOx levels prevent ozone depletion and the higher temperature in summer favors ozone generation and accumulation [78,80,82,83,86–88], while PM10 was a major pollutant in autumn (41.3%), spring (39.1%), and the second major pollutant in winter (34.3%). The number of days with PM10 as a major pollutant was higher in southern Xinjiang due to emissions from natural sources, e.g., the Taklimakan desert, sand storms [61– 64,89]. Any day with one or more pollutants exceeding CAAQS (Grade II) is considered as a non-attainment/pollution day [2,30]. During the study period, the proportion of nonattainment days decreased by 61.8% in NWC. Similarly, the proportion of non-attainment days decreased in all the provinces, e.g., SN, XJ, GS, NX, and QH, and experienced a reduction of 79.6%, 63.1%, 60.4%, and 37.6% in spring, summer, autumn, and winter, respectively, which clearly indicates that the ambient air quality improved significantly.

Rapid economic development, industrialization, haze events, dust storms, and adverse meteorological conditions play a crucial role in air quality deterioration [2,4–10]. The Chinese government is working proactively to combat the pollution levels by revising and implementing strict environmental regulations [3,20,21]. According to this study, the concentration of PM2.5, PM10**,** SO2, NO2, and CO, AQI, the proportion of AQI "Class I", and pollution days decreased significantly in NWC between 2015 and 2018.
