Impact of the COVID-19 Event on Air Quality in Central China

In early 2020, the COVID-19 epidemic spread globally. This study investigated the air quality of three cities in Hubei Province, Wuhan, Jingmen, and Enshi, central China, from January to March 2017–2020 to analyze the impact of the epidemic prevention and control actions on air quality. The results indicated that in the three cities, during February 2020, when the epidemic prevention and control actions were taken, the average concentrations of atmospheric PM2.5, PM10, SO2, CO, and NO2 in the three cities were 46.1 μg m–3, 50.8 μg m–3, 2.56 ppb, 0.60 ppm, and 6.70 ppb, and were 30.1%, 40.5%, 33.4%, 27.9%, and 61.4% lower than the levels in February 2017–2019, respectively. However, the average O3 concentration (23.1, 32.4, and 40.2 ppb) in 2020 did not show a significant decrease, and even increased by 12.7%, 14.3%, and 11.6% in January, February, and March, respectively. This is because a lower concentration of NO2 resulted in constraints on the NO + O3 reaction, and the O3 could not be effectively further depleted. In addition, the average air quality index (AQI) for the three cities in January, February, and March 2020 were 32.2%, 27.7%, and 14.9% lower than the levels in 2017–2019, respectively. Based on the AQIs for the three cities, the combined proportions of Class I and Class II in January, February, and March 2020 increased by 27.9%, 24.8%, and 4.3%, respectively, while the combined proportion of AQI Classes III, IV, V, and VI was reduced from 34.8% to 15.8%. In addition, in the first three months of 2020, the indicatory air pollutants in the three cities for the AQIs were predominant in the following order: PM2.5 (72.0%), O3 (16.4%), PM10 (8.3%), NO2 (2.9%), and CO (0.4%). This study provides useful information for establishing a scientific air pollution control strategy and is a valuable reference for future research on improving urban air quality.

In order to eliminate the spread of the epidemic, the Chinese government immediately took effective prevention and control actions. On January 24, 2020, Hubei Province government, China, announced the launch of a first-level response to major public health emergencies, which included actions such as quarantining, traffic restrictions, and factory closures, which were immediately implemented. The above prevention and control actions were closely related to the air quality at the time.
The Air Quality Index (AQI) typically reflects the degree of air cleanliness or the level of air pollution and focuses on the assessment of a crowd breathing for a specific period of time (acute or chronic) and the effects of air pollution on their health.
Air quality has become a serious concern of both the Chinese government and the public. Many scholars have conducted air quality assessments in China, reflecting the current problems and proposing control strategies for improvements in air pollution (Wang et al., 2014;Hu et al., 2015;Tong et al., 2016;Lee et al., 2018).
In this study, the air quality of three cities (Wuhan, Jingmen and Enshi) in Hubei Province, central China, including the air pollutants PM 2.5 , PM 10 , SO 2 , NO 2 , CO, and O 3 , and AQIs from January to March 2017-2020 were investigated, compared, and discussed. In addition, the impact of the COVID-19 epidemic prevention and control actions on air quality was specifically addressed.
(1) AQI = max(I 1 , I 2 , …, I n ) IAQI P : the air quality sub index for air pollutant p; C P : the concentration of pollutant p; C low : the concentration breakpoint that is ≤ C P ; C high : the concentration breakpoint that is ≥ C P ; I low : the index breakpoint corresponding to C low ; I high : the index breakpoint corresponding to C high . Ambient air quality is closely related to the development of human society and has a major impact on human health. The AQI simplifies the concentrations of different pollutants into a single numerical value to reflect overall air quality. The daily AQI value is calculated from the 24-hour average concentrations of PM 2.5 , PM 10 , SO 2 , CO, NO 2 , and the daily maximum 8-hour concentration of O 3 .

Comparison for Air Pollutants
The average concentrations for PM 2.5 , PM 10 , SO 2 , CO, NO 2 , and O 3 in January, February and March 2017-2019 and those of 2020, respectively, are shown and compared in Figs. 2(A)-2(F), respectively. Atmospheric particulate matter is a mixture that includes organic components, inorganic ions, mineral dust, and so on. PM 2.5 refers to atmospheric particles with a particle size of less than 2.5 µm, which are easily accumulated in the human respiratory tract and seriously affect human health. Studies have shown that PM 2.5 pollution is closely related to human morbidity or mortality (Dockery et al., 1993;Pope III et al., 2002) and that pollution is more severe in winter (Tao et al., 2009;Xu et al., 2017;Ning et al., 2018;Zhao et al., 2018b).
It can be seen that the average concentration of PM 2.5 decreased from January to March 2020. There may be two reasons for this significant reduction in January. The first is that the second half of January 2020 is the lunar New Year holiday in China. As a result, most factories were closed, and the employees were on vacation, which resulted in a significant reduction in industrial emissions. Secondly, since January 23, 2020, Hubei Province has implemented strict epidemic prevention and control actions, which led to traffic stagnation and factory closures, thus greatly reducing both emissions from automobile exhaust and industrial production. The reason for the decrease in PM 2.5 concentration in February 2020 was due to strict epidemic prevention and control actions. In March, under these actions, PM 2.5 concentrations in Wuhan and Jingmen still showed a downward trend, but in Enshi, there was an increase of 1.7%. This may be because a better ecological environment in Enshi leads to lower PM 2.5 levels, and at the same time, the temperature rise in March was more conducive to the dispersion of air pollutants. Therefore, the epidemic prevention and control actions did not have a significant impact on the air quality in Enshi.
PM 10 refers to atmospheric particulate matter with a diameter of less than 10 microns, mainly derived from industrial exhaust emissions, fossil fuel combustion, motor vehicle exhaust, and dust entrainment (Kong et al., 2011). Airborne particulate pollution is extremely detrimental to human health and affects the human respiratory and nervous systems (Wang and Chen, 2016).
As shown in Fig. 2(B)(b), in Wuhan, Jingmen, and Enshi, during February 2017-2019, the PM 10 concentrations ranged between 13 and 211, 29 and 218, and 24 and 150 µg m -3 and averaged 88.2, 105.0, and 69.6 µg m -3 , respectively. Those during February 2020 ranged from 12-103, 13-122, and 21-101 µg m -3 , with averages of 46.0, 54.2, and 52.1 µg m -3 , respectively, which were 47.9%, 48.4%, and 25.1% lower  It can be seen that the concentration of PM 10 decreased significantly from January to March 2020, especially in February when strict prevention and control actions were taken. A decrease in PM 10 in January 2020 can be attributed to the reduction in construction dust and industrial production emissions during the Lunar New Year holidays. The reduction in February and March can be attributed to the strict implementation of epidemic prevention and control actions, which resulted in a substantial reduction in mobile exhaust emissions.

PM2.5 concentration
Artificially generated SO 2 mainly comes from the combustion of coal, petroleum, and chemical fuels. SO 2 in the air is easily oxidized into SO 3 , which then develops into acid rain, causing damage to surface vegetation and buildings (Kato et al., 2016).
It can be seen that the SO 2 concentration of the three cities significantly decreased from January to March in 2020, which shows that the suspension of production caused by the Chinese New Year holiday and factory closures due to epidemic prevention and control actions resulted in a significant reduction in SO 2 emissions.
Carbon monoxide (CO) is one of the important indicators of air pollutants that are mainly derived from incomplete combustion activities such as fuel combustion and automobile exhaust emissions related to human activities. High concentrations of CO pose a major threat to human health and can quickly cause hypoxia in humans, leading to dizziness and even death (Scharte et al., 2000;Li et al., 2017).
As shown in Fig In the first three months of 2020, the CO concentration also showed a significant decrease compared with the same period in the previous three years. This shows that the Lunar New Year holiday and the implementation of epidemic prevention and control actions greatly reduced both the burning of industrial fossil fuels and the traffic flow, thus reducing CO emissions.
Nitrogen dioxide (NO 2 ) is an important pollutant and oxidant in the atmosphere, mainly derived from hightemperature combustion of fossil fuels, thermal power generation, industrial emissions, and automobile exhaust (Burnett et al., 2004;Jaeglé et al., 2005). Ambient NO 2 is associated with a variety of health hazards. High concentrations of NO 2 can lead to the formation of acid rain and nitrate aerosols, and are also important precursors for O 3 production (Biswas et al., 2019).
From January to March 2020, the decrease in NO 2 concentration is of great significance, and the reduction rate is much higher than other pollutants. Especially in February and March when the epidemic prevention and control actions were taken, the NO 2 concentration decreased by more than 60% compared with the same period in 2019-2019. This may have been because during the epidemic prevention and control actions, industrial production activities and transportation were greatly restricted, resulting in a sharp reduction in the emission of NO 2 from both industrial production and vehicle exhaust.
Ozone (O 3 ) is an important gas in the process of atmospheric chemical reactions and is also one of the key greenhouse gases. Due to the rapid development that has occurred in the past few decades, power plant emissions, industrial exhaust gas, and the burning of fossil fuels have indirectly caused increased ambient O 3 pollution (Logan et al., 1981;Ryerson et al., 2001). Solar radiation and higher air humidity, as well as increased NO x and VOCs (volatile organic compounds) in the environment, promote the photochemical reaction that produces O 3 .
It is worth noting that the pattern of O 3 concentration was completely opposite to the pattern of the other five air pollutants. As shown in Fig. 2(F)(a), in Wuhan, Jingmen, and Enshi, during January 2017-2019, the O 3 concentrations ranged between 2.80 and 42.0, 6.53 and 49.9, and 1.87 and 34.5 ppb and averaged 20.4, 26.8, and 14.2 ppb, respectively. Those during January 2020 ranged from 2. 33-51.3, 9.80-61.1, and 2.80-30.8 ppb and averaged 22.3, 31.0, and 16.1 ppb, respectively, which was an increase of 9.6%, 15.5%, and 13.1% compared with that of during January 2017-2019. Based on the data from the three cities, during January 2020, the average O 3 rose by 12.7% compared with that in January 2017-2019.
As shown in Fig Based on the data from the three cities, during February 2020, the average O 3 rose by 11.6% compared with that in February 2017-2019.
It can be seen that from January to March 2020, the average concentration of O 3 was significantly higher than the average concentration in the same period in the previous three years (2017-2019), which may have been caused by a lower level of NO x in 2020. According to previous studies, the level of O 3 is closely related to NO 2 and VOCs. When the NO x concentration is low, NO x promotes the formation of O 3 , and the concentration of VOCs has little effect on O 3 . When the VOC concentration is low, NO x concentration is negatively correlated with O 3 production (Chameides et al., 1992). It can be seen that from January to March 2020, the O 3 concentration and the NO 2 concentration in the three cities are inversely proportional.
Under sufficient intensity of solar radiation, NO 2 acts as a precursor in photochemical reactions and is first dissociated into NO and O ( 3 P): It can be seen that NOx is one of the important precursors for O 3 production, and NO is the direct cause of O 3 depletion. Lower levels of NO 2 in the atmosphere will cause a reduction in NO, which reduces the possibility of NO reacting with O 3 , resulting in the accumulation of O 3 . In general, urban NO x and O 3 have negative correlation characteristics, which is particularly obvious in colder winter. This is because in the summer, due to the intense solar radiation and the dominant photochemical reactions that occur at this time, the environment is more suitable for the accumulation of O 3 . In winter, the photochemical reaction during the day is relatively weak, so a higher NO 2 concentration within a specific range is beneficial to the consumption of O 3 , but a lower NO 2 concentration causes more O 3 to be generated during the day that cannot be further effectively converted (Zhao et al., 2018a;Biswas et al., 2019). This is a good explanation for the significant increase in O 3 concentration in the three cities in February and March when the epidemic prevention and control actions were taken.

AQI Distribution
In order to investigate the impact of the COVID-19 outbreak on the Air Quality Index (AQI) in central China, the AQI distribution for Wuhan, Jingmen, and Enshi in January, February and March 2017-2019 and those of 2020 are shown in Fig. 3.
Based on the results for the three cities under observation, in January 2017-2019, the average AQIs ranged from 33-348 and averaged 124.2, while those during January 2020 ranged between 23-195 and averaged 84.2, which was 32.2% lower than that in January 2017-2019. In February 2017-2019, the average AQIs ranged from 25-229 and averaged 92.6, while those during February 2020 ranged from 20-185 and averaged 66.9, which was 27.7% lower than that in February 2017-2019. However, in March 2017-2019, the average AQIs in the three cities ranged between 23 and 272 and averaged 71.0, while those during March 2020 ranged between 28 and 125 and averaged 60.4, which was 14.9% lower than that in March 2017-2019.
It can be seen that the air quality during January-March 2020 was significantly improved compared with the same period in 2017-2019. The improvement in air quality in January 2020 can be attributed to the reduction in industrial production and construction activities during the Lunar New Year holidays. The reduction in February and March 2020 can be attributed to the strict implementation of epidemic prevention and control actions, which has resulted in a substantial reduction in transportation and industrial emissions. In March, as the atmospheric temperature increased, the vertical dilution and dispersion of air pollutants were accelerated, and the air quality significantly improved, so the impact of epidemic prevention and control actions on air quality weakened.

Distribution of Six AQI Classes
This study also made a statistical analysis for the distribution of six AQI classes in the three cities in January, February, and March 2017-2019 and those in 2020, respectively.
It can be seen from Fig. 4(A)(a), in Wuhan, in January 2017-2019, the proportions of classes I, II, III, IV, V, and VI were 3.2%, 37.6%, 36.6%, 16.1%, 6.5%, and 0%, respectively. While during January 2020, the proportions of AQI classes I, II, III, IV, V, and VI were 16.1%, 51.6%, 32.3%, 0%, 0%, and 0%, respectively. It can be seen that during January 2020, the combined proportions of Class I and Class II increased from 40.8% to 67.7%, while the combined proportions of classes IV, V, and VI decreased from 22.6% to zero, which indicates that the air quality had greatly improved.
For Enshi (Fig. 4(A)(a)), which had the greatest improvement in air quality, in January 2017-2019, the proportions of classes I, II, III, IV, V, and VI were 9.7%, 47.3%, 29.0%, 8.6%, 5.4%, and 0%, respectively. In January 2020, the same AQI class proportions were 51.6%, 48.4%, 0%, 0%, 0%, and 0%. The combined proportion of Class I and Class II increased from 57.0% to 100%, respectively, while the combined proportions of classes Ⅲ, VI, V, and VI decreased from 43.0% to zero. It can be seen that the air quality of the three cities in January 2020 improved very significantly compared with the same period in the previous three years. Fig.4 (A)(b) shows the distribution of AQI classes for the three-city combination in January 2017-2019 and in January 2020, respectively. It can be seen that from 2017 to 2019, the AQI class distribution for classes I, II, III, IV, V, and VI in the three cities was 4.3%, 36.6%, 32.3%, 16.1%, 9.7%, and 1.1%, respectively, but in January 2020, it was 25.8%, 43.0%, 21.5%, 9.7%, 0%, and 0%, respectively. The combined proportion of classes I and Ⅱ increased from 40.9% to 68.8%, while that of classes VI, V, and VI decreased from 26.9% to 9.7%, respectively. At the same time, classes Ⅴ and Ⅵ did not appear in the three cities. Based on the data from the three cities, it can be seen that the air quality in Hubei Province in January 2020 improved from the same period in the previous three years. This may have been due to the production stagnation caused by the Chinese New Year holiday in late January and the closure of factories caused by the COVID-19 epidemic prevention and control actions.
As shown in Fig. 4(B)(a), in Wuhan, in February 2017-2019, the proportions of classes I, II, III, IV, V, and VI were 6.1%, 56.1%, 32.9%, 3.7%, 1.2%, and 0%, respectively. In February 2020 when comprehensive epidemic prevention and control actions were taken, the proportions of classes I, II, III, IV, V, and VI were 46.4%, 50.0%, 3.6%, 0%, 0%, and 0%, respectively. This indicated that during the epidemic control period, the combined proportion of Class I and Class II increased from 62.2% to 96.4 %, while the combined proportion of classes IV, V, and VI decreased from 4.9% to zero, which indicates that the air quality improved significantly.
For Enshi (Fig. 4(B)(a)), in February 2017-2019, the AQI proportions of classes I, II, III, IV, V, and VI were 25.0%, 57.1%, 16.7%, 1.2%, 0%, and 0%, respectively. In February 2020, the AQI proportions were 35.7%, 51.1%, 7.1%, 0%, 0%, and 0%. This was similar to both Wuhan and Jingzhou. However, in Enshi, the combined proportion of classes I and II increased from 82.1% to 86.8%, while the combined proportion of classes IV, V, and VI decreased from 1.2% to zero. Even though the improvement in air quality was not a big step, it is very clear that the epidemic prevention and control action had a good effect on air quality. Fig. 4(B)(b) shows the distribution of AQIs for three-city combination in February 2017-2019 and in February 2020, respectively. It can be seen that from 2017 to 2019, the AQI distribution of classes I, II, III, IV, V, and VI in the three cities was 10.5%, 52.8%, 29.4%, 6.0%, 1.2%, and 0%, respectively, but in February 2020, the distribution was 33.3%, 54.8%, 9.5%, 2.4%, 0%, and 0%, respectively. The combined proportion of classes I and II increased from 63.3% to 88.1%, while the combined proportion of classes IV, V, and VI decreased from 7.2% to 2.4%. According to the data for the three cities, it is clear that in February 2020, the air quality of the three cities improved significantly compared with that in February 2017-2019 (non-epidemic period). This is because in February 2020, Hubei Province implemented strict epidemic prevention and control actions including quarantines, industrial plant closures, and traffic restrictions that greatly reduced air pollutant emissions.
It can be seen in Fig. 4(C)(a), that in Wuhan, in March 2017-2019, the AQI proportions of classes I, II, III, IV, V, and VI were 8.6%, 77.4%, 11.8%, 1.1%, 1.1%, and 0%, respectively. During March 2020 when comprehensive epidemic prevention and control action were taken, the AQI proportions of classes I, II, III, IV, V, and VI were 38.7%, 61.3%, 0%, 0%, 0%, and 0%, respectively. This indicated that during the epidemic control period, the combined proportion of Class I and Class II increased from 86.0% to 100%, which revealed that the air quality had greatly improved.
For Enshi, in March 2017-2019, the AQI proportions of classes I, II, III, IV, V, and VI were 40.9%, 59.1%, 0%, 0%, 0%, and 0%, respectively. In March 2020, the AQI proportions were 48.4%, 45.2%, 6.5%, 0%, 0%, and 0%. This indicated that Enshi had good air quality in March 2017-2020. The forest coverage in Enshi accounts for about 70% of the area, so after the temperature rose in March, the vertical convention and dispersion of air pollutants was further increased, and the air quality was improved, so the control actions of the epidemic did not display a significant impact. Fig. 4(C)(b) shows the distribution of AQI Class for the three-city combination in March 2017-2019 and in March 2020, respectively. It can be seen that from 2017 to 2019, the AQI distribution of classes I, II, III, IV, V, and VI in the three cities was 18.3%, 73.1%, 7.9%, 0.4%, 0.4%, and 0%, respectively, but in March 2020, it was 34.4%, 61.3%, 4.3%, 0%, 0%, and 0%, respectively. The combined proportions of classes I and II increased from 91.4% to 95.7%, while the Xu et al.,Aerosol and Air Quality Research,xxxx 11 Fig . 4(B). The distribution of six AQI classes (a) for Wuhan, Jingmen, and Enshi in February 2017-2019 and February 2020, respectively and (b) for the three cities under observation. combined proportion of classes IV, V, and VI decreased from 0.8% to zero. Due to the strict epidemic prevention and control actions, the air quality in Hubei Province in March 2020 improved.

Indicatory Air Pollutants
The indicatory air pollutants for the different AQI classes from January to March 2017-2019 and that of 2020, respectively are shown in Table 1.