*3.1. Spatial and Temporal Variation of Six Criteria Pollutants*

During the study period (2015–2018), the average concentration of PM2.5, PM10, SO2, NO2, and CO decreased by 28.2% (14.2%, 25.8%, 31.1%, 34.4%, 47.1%) (Figure 1a), 32.7% (18.5%, 32.5%, 31.8%, 32.7%, 46.3%) (Figure 1b), 41.9% (30.6%, 43.8%, 46.8%, 48.7%, 32.3%) (Figure 1c), 6.2% (+3.59%, 4.44%, 18.8%, +3.72%, 8.45%) (Figure 1d), and 27.3% (32.7%, 21%, 38.4%, 17.2%, 16%) (Figure 1e), respectively, in NWC (SN, XJ, GS, NX, QH). In contrast to the other pollutants, the ozone levels increased by 3.69% (5.4%, 6.04%, 1.32%, 19.3%, 5.66%) in NWC (SN, XJ, GS, NX, QH) between 2015 and 2018 (Figure 1f). The annual average concentration of PM2.5 and PM10 failed to comply with CAAQS Grade II standards (35 μg/m3 and 70 μg/m3, annual mean) and exceeded them by 25% and 31.9%, respectively, and exceeded them by 3.37 and 3.61 times, respectively, for the WHO air quality guidelines (10 μg/m3 and 20 μg/m3, annual mean) in NWC. PM2.5 and PM10 failed to comply with CAAQS Grade II standards in SN, XJ, GS, NX, and NWC (Figure 1a,b), while SO2 and NO2 complied with CAAQS Grade II standards (60 μg/m3 and 40 μg/m3, annual mean) in SN, XJ, GS, NX, QH, and NWC (Figure 1c,d). CO and O3 do not have annual standards under CAAQS; CO decreased in SN, XJ, GS, NX, QH, and NWC, while O3 decreased in GS and QH during 2018 as compared with 2015 (Figure 1e,f). During the study period, the highest concentration of PM2.5, PM10, SO2, NO2, CO, and O3 occurred in SN, XJ, NX, SN, SN, and QH, respectively. Figure 2 explains the spatial distribution of the criteria pollutants in 53 cities of NWC during 2015 to 2018, obtained by the inverse distance weighted (IDW) interpolation technique. The obtained results from spatial interpolation were quite similar to the actual values. In the case of spatial distribution, 92.5%, 96.2%, 92.5%, 64.5%, 88.7%, and 11.3% of the cities of NWC experienced a reduction in PM2.5 (Figure 2a–d), PM10 (Figure 2e–h), SO2 (Figure 2i–l), NO2 (Figure 2m–p), CO (Figure 2q–t), and O3 (Figure 2u–x), respectively, during 2018 as compared 2015. Similarly, 66%, 72.5%, and 13.2% of the cities failed to meet the CAAQS Grade II for PM2.5, PM10, and NO2 (35 μg/m3, 70 μg/m3, 40 μg/m3, annual mean), respectively (Table S1). Most of the cities that were not complying with the CAAQS are cities with a larger population and increased industrial activities.

**Figure 1.** Annual variation of PM2.5 (**a**), PM10 (**b**), SO2 (**c**), NO2 (**d**), CO (**e**), O3 (**f**), PM2.5/PM10 (**g**), and AQI (**h**) in five provinces (Shaanxi (SN), Xinjiang (XJ), Gansu (GS), Ningxia (NX), and Qinghai (QH)) of northwest China (NWC) and NWC as a whole between 2015 and 2018. Descriptions are as follows: light blue bar (2015), orange bar (2016), grey bar (2017), yellow bar (2018), blue bar (four-year average (FYA)), parrot line with dots (CAAQS Grade II standards, annual mean), and dark blue with dots (WHO standards). The abbreviations are as follows: PM2.5 (fine particulate matter), PM10 (coarse particulate matter), SO2 (sulfur dioxide), NO2 (nitrogen dioxide), CO (carbon monoxide), O3 (ozone), PM2.5/PM10 (ratio of PM2.5 with PM10), and AQI (air quality index).

**Figure 2.** The spatial distribution of PM2.5 (**a**–**d**), PM10 (**e**–**h**), SO2 (**i**–**l**), NO2 (**m**–**p**), CO (**q**–**t**), and O3 (**u**–**x**) between 2015 and 2018 in northwest China (NWC). Color represents the different pollution levels, e.g., green (good), yellow (moderate), orange (unhealthy for the sensitive group), red (unhealthy for all), purple (very unhealthy), and maroon (hazardous). The abbreviations are as follows: PM2.5 (fine particulate matter), PM10 (coarse particulate matter), SO2 (sulfur dioxide), NO2 (nitrogen dioxide), CO (carbon monoxide), and O3 (ozone).
