Changes in air quality related to the control of coronavirus in China: Implications for traffic and industrial emissions.

Measures taken to control the disease (Covid-19) caused by the novel coronavirus dramatically reduced the number of vehicles on the road and diminished factory production. For this study, changes in the air quality index (AQI) and the concentrations of six air pollutants (PM2.5, PM10, CO, SO2, NO2, and O3) were evaluated during the Covid-19 control period in northern China. Overall, the air quality improved, most likely due to reduced emissions from the transportation and secondary industrial sectors. Specifically, the transportation sector was linked to the NO2 emission reductions, while lower emissions from secondary industries were the major cause for the reductions of PM2.5 and CO. The reduction in SO2 concentrations was only linked to the industrial sector. However, the reductions in emissions did not fully eliminate air pollution, and O3 actually increased, possibly because lower fine particle loadings led to less scavenging of HO2 and as a result greater O3 production. These results also highlight need to control emissions from the residential sector.

Introduction

In December 2019, a disease that was eventually linked to a beta coronavirus and named Covid-19 was reported in Wuhan, China (Zhu et al., 2020). In the weeks that followed, measures were taken to reduce large gatherings to control the spread of the disease (China State Council, 2020). For example, the Spring Festival holiday was extended beyond February 10, and during that time only essential enterprises involving people's immediate needs (such as health care, or providing food), were allowed to operate. In addition, the opening of schools after the holiday was postponed.

These measures led to a dramatically reduced number of vehicles on the road and a near total reduction in factory production (MEP, 2020; Wang et al., 2020). Pollution emissions from the transportation and industrial sectors were expected to decrease, and the following questions can be asked: Did the air quality improve during the control of Covid-19? If so, how was the improvement related to the reduced emissions from these two sectors? Also, what was the major cause for the improvements in air quality and which atmospheric species were most affected? Answering these questions is a way to evaluate the influences of traffic and industrial emissions on air quality during winter in China. The results of the evaluation can then be used to make recommendations regarding air pollution control during winter in China.

In this study, changes in the air quality index (AQI) and the concentrations of six air pollutants (PM2.5, PM10, CO, SO2, NO2, and O3) were evaluated, and the causes for these changes during the control period for Covid-19 were investigated. Correlation analyses were then used to analyze the relationships between socioeconomic factors (i.e., motor vehicle usage, percentage of secondary industries, and industrial emissions) and air quality. Finally, suggestions are given for air pollution control in China during winter.

Methods

Study area and period

A total of 366 urban areas in China's mainland were selected for study. These cities are widely distributed over China with the greatest coverage in eastern-southeastern parts of the country (Fig. 1 ). Taken together, the sites can be considered representative of the overall air quality in China (Kuerban et al., 2019; Zhao et al., 2019).

Fig. 1

Geographical locations of 366 urban areas included in the study.

The overall strategy for the study was to compare the concentrations of selected pollutants before and after the Covid-19 control measures were put in place. The period prior to the controls was from 1 to the 23 January 2020, while the Covid-19 control period was from 24 January to 9 February 2020. During the control period, a series of measures was undertaken to reduce gatherings of people (China State Council, 2020). Specifically, all non-essential factories were shut down, and schools were closed as were all entertainment venues and restaurants. A dramatic reduction in road traffic was observed during the control period. For example, the flow of commercial trucks and buses in the Beijing-Tianjin-Hebei region and its surrounding areas decreased by 77% and 39%, respectively, during the control period (MEP, 2020).

Data sources

The real-time monitoring data for AQI, PM2.5, PM10, SO2, NO2, O3, and CO in the 366 urban areas were obtained from China's National Environmental Monitoring Center. The data had a time resolution of 1 h (http://www.cnemc.cn). Previous studies have shown that data from China's National Environmental Monitoring Center are statistically reliable (Kuerban et al., 2019; Zhao et al., 2019). The AQI monitoring values calculated from the six priority pollutants (Bai et al., 2019) and the data for the individual pollutants themselves were averaged for the periods prior to and during the Covid-19 controls.

The geographical distributions of secondary industries (which convert raw materials produced by primary industries into goods and products) and industrial SO2 and NOx emissions were obtained from the 2018 China City Statistical Yearbook. The numbers of motor vehicles in operation were obtained from the China Statistic Yearbook. Notably, the socioeconomic data from seven provinces (Tianjin, Sichuan, Jilin, Heilongjiang, Gansu, Xinjiang, and Qinghai) where the holiday was not extended to February 10, were not used in this study (Table S1).

Results and discussion

Air quality during the control of Covid-19

Significant differences (p < 0.01) were found in the AQI and the concentrations of six air pollutants (PM2.5, PM10, CO, SO2, NO2, and O3) of 366 urban areas before and during the control of Covid-19, suggesting the air quality changed during the control period. The AQI averaged over all stations decreased by 20%, from 89.6 before the control period for Covid-19 to 71.6 during the controls (Fig. 2 ), demonstrating an overall improvement in air quality from the control measures. A total of 322 of the 366 cities studied experienced a decline in the AQI. The highest AQI reductions occurred in the Ningxia, Shandong, and Henan Provinces (Fig. S1), which normally have high numbers of vehicles in use (Shandong and Henan) (Fig. S2a) and many secondary industries in operation (Ningxia, Shandong, and Henan) (Fig. S2b). The lowest AQI reductions occurred in the Yunnan, Guizhou, and Hainan Provinces (Fig. S1), where motor vehicle usage and secondary industries are relatively low (Fig. S2). Therefore, the improvement of air quality apparently was related to the number of motor vehicles in use and the percentages of secondary industries in each province. The AQI in 366 urban areas had the highest correlation coefficient (r2 of 0.99) with the PM2.5 before and during the control of Covid-19, which reflects the fact that fine particles were the major air pollutant throughout the study.

Fig. 2

Comparison of air quality indices (AQI) before and during the control of Covid-19.

The concentrations of the PM2.5, PM10, SO2, NO2, and CO, but not O3, decreased during the Covid-19 control period (Fig. S3) compared with those before the controls were put in place. The concentration of PM2.5, decreased from 65.0 μg m−3 to 51.4 μg m−3 (Fig. S3), and 315 of the 366 cities experienced a decrease in PM2.5. The regions with the highest and lowest reductions in PM2.5 concentrations were the same as those for the AQI (Figs. S1, S4), again, because fine particles were most often the major air pollutant.

After the controls were implemented, the concentrations of CO, SO2, and NO2 decreased by 0.23 mg m−3, 2.2 μg m−3, and 19.4 μg m−3, respectively (Fig. S3). A total of 331 (309, 366) of the 366 cities studied experienced reductions in CO (SO2, NO2). The highest reductions of CO and NO2 occurred in the Shandong and Henan Provinces (Figs. 3 , S5) where there are large numbers of motor vehicles and many secondary industries (Fig. S2). This suggests that the reduced emissions from the transportation and industrial sectors were what caused the concentrations of these two gases to decrease. The largest decreases in SO2 were found in Shanxi and Jiangxi Provinces (Fig. 4 ), which have low numbers of vehicles but many secondary industries (Fig. S2). This indicates that the reduced SO2 concentration was probably caused by the lower emissions from secondary industries during the control period.

Fig. 3

Change in NO2 mass concentrations (∆NO2) between the pre-control and Covid-19 control period. ∆NO2 refers to the average NO2 mass concentrations before the control of Covid-19 in each province minus the average NO2 mass concentrations during the control period.

Fig. 4

Changes in the concentrations of SO2 (∆SO2) during the control period for Covid-19. ∆SO2 refers to the average SO2 mass concentrations before the control of Covid-19 in each province minus the average SO2 mass concentrations during the control period.

The amplitude of the concentration variation (ACV) was calculated using the equation, ACV = (y − x) / x × 100%, where x and y are the mass concentrations of a substance of interest before and during the control period for Covid-19, respectively. The air pollutant that showed the largest decrease with the Covid-19 controls was NO2 (ACV = −54%) (Fig. S3) while SO2, showed the smallest decline (ACV = −16%). The SO2/NO2 ratio is an indicator of the relative contributions of air pollutants from stationary versus mobile sources (Aneja et al., 2001), and higher values occur when there are greater influences from stationary sources. The SO2/NO2 ratio averaged over the two sets of samples increased from 0.39 to 0.70 after the controls were in place (Fig. S6), suggesting an increase in the relative importance of stationary sources (Song et al., 2017). The AVCs for PM2.5 and PM10 were −21% and −27%, respectively (Fig. S3), and the PM2.5/PM10 ratio increased from 0.76 to 0.82 (Fig. S6); these are signs of either decreased impacts from dust sources or enhanced secondary aerosol formation during the Covid-19 control period (Song et al., 2017; Zhao et al., 2018).

Factors driving the improvements in air quality

Significantly positive relationships were found between the numbers of motor vehicles and the reduced AQIs (R2 = 0.11, p < 0.1; Fig. 5 ) and between the percentages of secondary industries and the change in AQIs (R2 = 0.25, p < 0.05; Fig. 5). With the people largely confined to their homes, the provinces with higher numbers of vehicles should have had greater reductions in vehicle emissions during the control period, and the same should have been true for the secondary industries. The decreases in AQIs were more strongly correlated with the percentages of secondary industries than with motor vehicle numbers (Fig. 5), suggesting that the changes in industrial emissions were more responsible for the improvements air quality, especially fine particles, than motor vehicle usage.

Fig. 5

Scatterplot of the change in air quality index (∆AQI) and socioeconomic factors (number of motor vehicles, secondary industry share) in provinces with air quality improvements. ∆AQI on the Y-axis refers to the average AQI before the control of Covid-19 in each province minus the average AQI during the control period.

The decreased PM2.5 concentrations were positively correlated with motor vehicle numbers (R2 = 0.11, p < 0.1; Fig. 6 ) and percentages of secondary industries (R2 = 0.28, p < 0.05; Fig. 6). Therefore, the reduction in the PM2.5 is best explained by lower emissions from the transportation and industrial sector. The reduced NO2 concentrations were positively correlated with both vehicle numbers (R2 = 0.44, p < 0.001; Fig. 6) and industrial NOx emissions (R2 = 0.36, p < 0.01; Fig. 6), indicating that decreased emissions from both the transportation and industrial sectors led to improvements in NO2. As the reduced NO2 concentrations were more strongly correlated with vehicle population than with the industrial NOx emissions, transportation probably was more responsible for the decrease in NO2 concentrations. The SO2 concentrations showed a significant positive relationship with industrial SO2 emissions (R2 = 0.16, p < 0.1; Fig. 6) but not with motor vehicle numbers. Therefore, the reduced SO2 concentrations were only linked to the industrial sector. The CO concentrations showed a significant positive correlation with both the vehicle numbers (R2 = 0.17, p < 0.05; Fig. 6) and percentages of secondary industries (R2 = 0.29, p < 0.01; Fig. 6), but the stronger correlation with the latter (Fig. 6) suggests a reduction in industrial emissions was more responsible for the decrease.

Fig. 6

Concentrations of the changes in pollutant gases (∆PM2.5, ∆NO2, ∆SO2 and ∆CO) during the Covid-19 control period versus socioeconomic factors (vehicle population, industrial NOx emissions, industrial SO2 emissions, secondary industry share). The mass concentrations of pollutant gases (∆PM2.5, ∆NO2, ∆SO2 and ∆CO) on the Y-axis refer to the average mass concentrations before the control of Covid-19 in each province minus the average mass concentrations during the control period.

Policy implications

Reducing air pollution

The air quality improved during the control period for Covid-19, and that apparently was mainly caused by reductions in emissions from the transportation and industrial sectors. The transportation sector was the major factor for the reduction of NO2 mass concentrations, indicating the control measures greatly reduced the pollution emissions caused by the movement of people. As a major pollutant emitted from the coal heating in winter (Kuerban et al., 2019), the mass concentration of SO2 decreased the least, suggesting the emissions from coal heating activities were probably little affected by the control measures.

Although the air quality improved, the average AQIs in 84 of the 169 cities in northern China were >100 after the controls were implemented, suggesting that the air pollutants in many cities were still at harmful levels. This means that even though the reduced emissions from the transportation and industrial sectors did lead to improvements in air quality, the concentrations of some pollutants were still at unhealthy levels.

The transportation sector is not generally thought to be the major source for PM2.5 during winters in northern China (Huang et al., 2014; Elser et al., 2016). Rather, this source has been shown to contribute to 6%–22% of the PM2.5 mass concentration (Tao et al., 2017) and 5%–21% of the organic aerosol mass (Wang et al., 2019). The total number of motor vehicles in China increased from 5.5 million in 1990 to 327 million in 2019 (Wu et al., 2017; MPSC, 2019), which is a 60-fold increase over 30 years. However, increasingly stringent emission standards, electric vehicle subsidies, and the promotion and development of the public transportation have limited the impacts from mobile emissions (Wu et al., 2017). In fact, those measures have prevented increases in the vehicular emissions of air pollutants (except for NOx) since 2010 (Wu et al., 2017). As a result, the contribution of the transportation sector to air pollution has not increased in parallel with the rising numbers of vehicles on the roads (Van et al., 2017).

Industrial emissions are the major contributor to PM2.5 pollution in China (Shi et al., 2017), but the reduced emissions from that sector did not prevent air pollution during the control period. In fact, essential industries, some of which emit large amounts of pollutants, did not curtail operations during the control period for Covid-19 (MEP, 2020). For perspective, under normal circumstances, thermal power generation contributes 20.1% of the total SO2 emissions and 32.6% of the total NOx in China (Huang et al., 2017). These critical industries must operate continuously (Huang et al., 2017; MEP, 2020), and therefore, reducing their impacts on air quality obviously should be a central element of pollution control efforts.

The residential sector contributed 39% of the total PM2.5 emissions in China in 2010 (Li et al., 2017), and emissions from residences were the most likely cause for air pollution during the Covid-19 control period. The industrial sector was largest contributor to fine PM in 2013 (Shi et al., 2017), but the emissions from this source decreased from 2013 to 2017 (Zhang et al., 2019), which caused the relative importance of the other pollution sources to increase proportionately. Indeed, the residential sector became the major contributor (>50%) to PM2.5 in representative cities of northwestern China during the winter of 2016–2017 (Yang et al., 2020). Therefore, plans for controlling emissions from residences should be developed and implemented to combat air pollution during the winter in northern China.

Ozone pollution control

Although the reduced emissions from transportation and secondary industrial sectors did not eliminate all air pollution, NO2, a precursor for O3, did decrease dramatically during the control period (Figs. 3, S3). The emission factors for NOx from the transportation and industrial sectors are much higher than those for other sectors (Li et al., 2017), making these the major contributors to NOx in China (Van et al., 2017). In fact, the Multi-resolution Emission Inventory for China (MEIC) (http://www.meicmodel.org/) indicated that 64% of the NO2 emissions originated from these two sectors.

In the troposphere, NOx reacts with volatile organic compounds (VOCs) to form O3, and therefore, one would expect that O3 would have decreased along with NOx during the control period for Covid-19; however, this was not the case. In fact, O3 averaged over all stations increased by 20.1 μg m−3 from 39.0 μg m−3 to 59.1 μg m−3 (Fig. S3), and the AVC for the O3 concentration was +51% (Fig. S3). For perspective, NOx concentrations have been declining in China since 2011 (De Smedt et al., 2015), while O3 has increased over the same time period (Li et al., 2019).

The observed increases in O3 during the Covid-19 controls may be related to the observed decreases in fine particles. That is, the lower PM2.5 would be a less effective sink for hydroperoxy radicals (HO2), which would increase peroxy radical-mediated O3 production (Li et al., 2019). In addition, decreased NOx would have little or no effect on O3 in any urban areas where O3 production is VOC limited (Jin et al., 2015; Wang et al., 2017; He et al., 2019; Lyu et al., 2019). Simultaneously reducing both NOx and VOCs would be the most effective way to reduce surface ozone in China (Li et al., 2019), but the sources for VOCs during winters in China are complex (H. Zheng et al., 2018; B. Zheng et al., 2018; Song et al., 2019; Huang et al., 2020). Indeed, controlling the NOx emissions from transportation and secondary industries may not reduce VOC concentrations to the extent that O3 is greatly affected.

Conclusions

The conclusions for the study can be summarized as answers to the questions posed in the introduction:

Did the air quality improve during the control period for Covid-19? Yes, the overall air quality improved: the AQI decreased from 89.6 prior to the control period to 71.6 during the control period. In fact, a large majority (322 of 366) of the cities studied experienced a decline in the AQI. The concentrations of five air pollutants (PM2.5, PM10, CO, SO2, NO2) decreased during the control period for Covid-19, but O3 did not decrease.

Which emission sources were responsible for the reduced concentrations of atmospheric species during the Covid-19 control period? Lower emissions from motor vehicles and secondary industries most likely were responsible for the observed decreases in PM2.5, NO2, and CO concentrations during the control period. Lower emissions from the transportation sector were the main cause for NO2 reduction, while the industrial sector was responsible for the PM2.5 and CO reductions. The reduction in SO2 was only related to reduced emissions from the industrial sector.

What are the policy implications? Despite restrictions placed on motor vehicles and secondary industries during the control period, air pollution continued to be problematic in large parts of northern China. The concentrations of O3 actually increased during the control period, possibly because lower fine particle loadings led to less scavenging of HO2 and as a result greater O3 production. These results illustrate the importance of reactions that can occur between gaseous and particulate pollutants, but clearly, lowering the emissions of both NOx and VOCs will be needed to control O3. Results of our study highlight the importance of emissions from the residential sector for wintertime pollution in northern China, and they show that this source must be taken into account in developing pollution mitigation plans.

CRediT authorship contribution statement

Yichen Wang: Conceptualization, Methodology, Writing - original draft. Yuan Yuan: Formal analysis. Qiyuan Wang: Writing - review & editing. ChenGuang Liu: Writing - review & editing. Qiang Zhi: Writing - review & editing. Junji Cao: Conceptualization, Writing - review & editing.