Impact of the COVID-19 pandemic and control measures on air quality and aerosol light absorption in Southwestern China

Abstract

China has been performing nationwide social lockdown by releasing the Level 1 response to major public health emergencies (RMPHE) to struggle against the COVID-19 (SARS-CoV-2) outbreak since late January 2020. During the Level 1 RMPHE, social production and public transport were maintained at minimal levels, and residents stayed in and worked from home. The universal impact of anthropogenic activities on air pollution can be evaluated by comparing it with air quality under such extreme conditions. We investigated the concentration of both gaseous and particulate pollutants and aerosol light absorption at different levels of (RMPHE) in an urban area of southwestern China. During the lockdown, PM_2.5, PM₁₀, SO₂, NO_x, and BC decreased by 30–50%, compared to the pre-Level 1 RMPHE period. Meanwhile, the decrease of NO_x caused the rise of O₃ by up to 2.3 times due to the volatile organic compounds (VOCs) limitation. The aerosol light absorption coefficient at multiple wavelengths decreased by 50%, and AAE decreased by 20% during the Level 1 RMPHE. BrC played essential roles in light absorption after the RMPHE was announced, accounting for 54.0% of the aerosol absorption coefficient at 370 nm. Moreover, the lockdown down-weighted the fraction of fossil fuel in BC concentrations to 0.43 (minima). This study characterizes air pollution at the most basic level and can provide policymakers with references for the “baseline.”

Graphical abstract

1. Introduction

The ongoing COVID-19 (also known as SARS-CoV-2) outbreak has caused a global pandemic since late December 2019 (Wu et al., 2020). COVID-19 has been classified as a Category B infectious disease but managed as a Category A disease. The Chinese Central Government has released policies for the control and prevention of the COVID-19 pandemic. More than 30 provincial governments released a Level 1 response to major public health emergencies (RMPHE) in late January. According to the Law of the People's Republic of China on the Prevention and Control of Infectious Disease, the Level 1 RMPHE demands a complete prohibition of social and other gatherings, a total shutdown of schools, industries, businesses and unnecessary traffic (civil aviation, private cars, and buses), and for residents to stay at home.

In most provinces, during the Level 1 RMPHE, only the activities that keep society running were permitted, such as power generation and the transportation of supplies. Residents were asked to maintain social distancing, stay in, and work from home, going out to buy supplies only when necessary. The universal impact of anthropogenic activities on air pollution can be evaluated by comparing it with air quality under such extreme conditions.

Air quality is sensitive to anthropogenic emissions. The strict lockdown measures reduced daily CO₂ emissions globally by up to 25% (Le Quéré et al., 2020). In European Economic Area countries, the lockdown was responsible for about 54% of nonmethane volatile organic compounds (NMVOC), 51% of NO_x, 30% of PM_2.5, and 25% of SO₂ emissions (Cheval et al., 2020). In China, according to satellite and ground-based observations, these specific emissions were reduced by up to 90% during the city lockdown period (Le et al., 2020).

On a global scale, pollutants such as NO₂ and PM₁₀ decreased in major cities in China, South Korea, Europe (Milan, Paris, Warsaw, etc.) and South and North America (Bauwens et al., 2020; Ma and Kang, 2020; Siciliano et al., 2020). However, with the remarkable reduction of emissions during the COIVD-19 outbreak, unexpected air pollution rose in multiple areas in China, such as the Beijing-Tianjin-Hebei and Yangtze River Delta regions (Chang et al., 2020; Le et al., 2020; Wang et al., 2020). Le et al. (2020) proposed that the air pollution was caused by the anomalously high humidity promoting heterogeneous aerosol chemistry, along with stagnant airflow and continuous emissions from power plants and petrochemical facilities, contributing to severe haze formation. Huang et al. (2020) also proposed that decreases in NO_x emissions from transportation increased ozone and nighttime NO₃ radical formation, and these increases in atmospheric oxidizing capacity facilitated the formation of secondary particulate matter.

Atmospheric aerosols impair visibility, change the earth's radiation balance, and affect human health (Ban-Weiss et al., 2011; Kim et al., 2015; Poschl, 2005). Light scattering and light absorption are the two basic optical properties of aerosols (Kim and Ramanathan, 2008; Rosenfeld, 1999). Light-scattering aerosols that can cool the atmosphere are mainly composed of ammonium, sulfate, nitrate, and most organics (Tao et al., 2017). Light-absorbing aerosols, however, contain the two major components of black carbon and brown carbon (Bond and Bergstrom, 2006).

Black carbon (BC) and brown carbon (BrC) are significant contributors of light-absorbing carbonaceous aerosols. BC has been recognized as the third most crucial climate-changing agent which absorbs solar radiation from the ultraviolet (UV) to infrared spectrum (Bond and Bergstrom, 2006; Bond et al., 2013). BrC can absorb radiation in the UV and visible ranges (Moosmüller et al., 2011; Peng et al., 2020b). Both BC and BrC can be emitted by anthropogenic sources due to incomplete combustion, and their characterization can be used as an indicator of economic activities among industry, traffic, and residential behaviors. The characterization of these pollutants can be used to evaluate the impact of the lockdown caused by the Level 1 REMPHE due to the outbreak of COVID-19.

The ongoing COVID-19 pandemic and following lockdown for control and prevention of the infectious disease severely influenced social activities, leading to impacts on air, water, and soil pollutions. The in-depth analysis of the impacts is required to understand air pollution in such an extreme condition that the anthropogenic emissions are ultimately diminished. Such studies are also important to improve our knowledge on the running of global societies, giving perspective on the impacts of anthropogenic activities on Earth's system. The aim of the study is to investigate the impact of COVID-19 pandemic and the lockdown on the air quality and aerosol light absorption in Southwestern China. The influence of different levels of RMPHE on air quality is also evaluated. The study also implies more sustainable ways on improving air quality after recovering from the COVID-19 outbreak.

2. Methods

2.1. Sampling site

The sampling site is located at the Chongqing Institute of Green and Intelligent Technology, Chinese Academy of Science (106.55°E, 29.81°N) at ground level. It is situated in the north of Chongqing. There are no power plants or coal-using industries nearby. The site is influenced by local biomass burning for heating or cooking during winter, as well as local traffic. This site was specifically chosen for background air quality observation. As shown in Figs. S1, S2, and S3, locally, the wind direction was predominantly from the north and northwest, and the air mass passing the sampling site was mainly from the east, north, and southwest. Therefore, the sampling site can represent a mixed background of cities, rural, and industrial areas in the Sichuan Basin region.

A comprehensive observation of light absorption was conducted from January 17th to April 1st. Aside from the multiple levels of RMPHE, there were also other important milestones. On February 9th, industries were allowed to reopen to produce medical supplies. Therefore, the period from January 24th (announcement of the Level 1 RMPHE) to February 9th can be considered the strictest lockdown period. As such, in Fig. 2, we separate the Level 1 RMPHE into two parts: January 24th to February 9th and February 10th to March 15th, and the results also indicate that the concentrations of most pollutants were minimized under the strictest conditions of the lockdown.

Fig. 2.

Radar plot of normalized air quality and light absorption data; data from before the Level 1 RMPHE is used as a baseline. After February 9th, part of the industry reopened for producing medical supply, the period from level 1 RMPHE (January 24th) to 9th February was under the most strict lockdown.

2.2. Instrumentation and data analysis

Air quality data including NO, NO₂, SO₂, O₃, and CO were collected using gas monitors (Thermo Inc., models 42i, 43i, 49i, and 48i respectively). The meteorological parameters and total solar radiation were recorded using a Lufft WS501-UMB.

A seven-band aethalometer (Magee Scientific Inc., Model AE-33) was used to determine aerosol light absorption at 370, 470, 520, 590, 660, 880, and 950 nm. The aethalometer was associated with a PM_2.5 inlet at a flow rate of 5 L min⁻¹. The methods of data acquisition, compensation of attenuation (ATN), and calculation of BC concentration and Absorption Ångström exponent (AAE) can be found in the literature (Qiu et al., 2019). The mass absorption cross-section (MAC) values used in this study are 18.47, 14.54, 13.14, 11.58, 10.35, 7.77, and 7.19 m² g⁻¹ for the wavelengths of 370, 470, 520, 590, 660, 880, and 950 nm respectively. The wavelength-dependent absorption coefficient (b_abs) at a specific λ is determined by:

$${\mathrm{b}}_{\mathrm{abs}}\left(\mathrm{\lambda }\right)=\mathrm{BC}\left(\mathrm{\lambda }\right)\bullet \mathrm{MAC}\left(\mathrm{\lambda }\right)$$ (1)

AAE can describe the wavelength dependence of aerosol light properties (Laskin et al., 2015). AAE can be affected by aerosol size distribution, chemical composition, and mixing state (Weingartner et al., 2003). A typical AAE for BC is ~1, but it can reach up to 9.5 for BrC (Lack and Langridge, 2013). The AAE of BrC varies in different locations and seasons worldwide. Light properties can also be used to quantitatively determine the sources of BC (Sandradewi et al., 2008b).

AAE is acquired via the following fitting (Moosmüller et al., 2011):

$${\mathrm{\sigma }}_{\mathrm{abs}}\left(\mathrm{\lambda }\right)\propto {\mathrm{\lambda }}^{-\mathrm{AAE}}$$ (2)

Assuming that BrC has negligible absorption at long wavelengths such as 880 nm and 950 nm, the contribution of BrC to total light absorption can be calculated by subtracting the light absorption of BC from the total absorption at all wavelengths (Liu et al., 2015).

$$B_{\mathit{abs},370\mathit{nm},\mathit{BC}}=B_{\mathit{abs},370\mathit{nm},\mathit{BC}}{\left(\frac{370}{880}\right)}^{-\mathit{AAE}}$$ (3)

$$B_{\mathit{abs},370\mathit{nm},\mathit{BrC}}=B_{\mathit{abs},370\mathit{nm},\mathit{total}}-B_{\mathit{abs},370\mathit{nm},\mathit{BC}}$$ (4)

A similar procedure is performed at 520 nm, 590 nm and 660 nm to obtain a fit for the light absorption of BrC.

3. Results and discussion

3.1. Impact of COVID-19 pandemic and its prevention on air quality

Chongqing Municipality announced the Level 1 RMPHE on January 24th, 2020, the same day as the Chinese Spring Festival. A sector of the industry manufacturing supplies for COVID-19 prevention was allowed to open on February 9th, 2020. Level 2 and 3 RMPHE were announced on March 15th and March 24th respectively. After March 24th, all business was restored.

Fig. 1 shows the time series of PM_2.5, gases, b_abs, and AAE data in different phases of the COVID-19 pandemic from January 17th to April 1st. Generally, SO₂, NO, PM₁₀, PM_2.5, and CO were 1.2 ± 1.9 ppb, 2.5 ± 2.2 ppb, 5.3 ± 3.5 ppb, 48.3 ± 22.5 μg m⁻³, 35.9 ± 31.0 μg m⁻³, and 594.0 ± 150.8 ppb (Table 1 ). The concentrations of gases, PM_2.5, and PM₁₀ before the Level 1 RMPHE were set as the baseline to normalize the concentration data at different levels of RMPHE, as shown in Fig. 2 . During the strictest control measures from the Level 1 RMPHE to February 9th, 2020, PM_2.5, PM₁₀, NO₂, SO₂, and BC decreased to 58%, 52%, 49%, 74%, and 61% respectively; the concentrations were then restored to 72%, 74%, 80%, 90%, and 82% between February 10th and March 15th (when the Level 2 RMPHE was announced) (Fig. 2). After the Level 2 RMPHE, SO₂, NO₂, and PM₁₀ rose to 220%, and 105% compared with before the Level 1 RMPHE. In short, the prevention and control measures caused a remarkable decrease in air pollutants. During the Level 1 RMPHE, public transport and private automobiles were halted. Only heavy-duty vehicles delivering essential food and supplies were allowed, providing a unique case for evaluating traffic emissions between the “normal” mode and this “minimal” mode. Since there was no considerable difference in meteorological conditions before and during the Level 1 RMPHE, the conclusion can be drawn that NO_x emissions reduced to 58% between the normal and minimal modes in the Chongqing region (Fig. 3 ).

Fig. 1.

Overview of the impact of the Level 1 RMPHE on air quality (PM_2.5, NO₂, SO₂, BC, and O₃) and light absorption (AAE, b_abs).

Table 1.

Summary of gas, meteorological, and light absorption data during winter 2020.

	Before level 1a	Level 1c	Level 2b	Level 3d	Overall
SO2 (ppb)	1.5 ± 1.2	1.2 ± 1.9	3.2 ± 3.1	2.4 ± 1.5	1.8 ± 1.6
NO (ppb)	9.6 ± 8.8	2.5 ± 2.2	6.0 ± 5.3	8.8 ± 6.6	5.4 ± 3.3
NO2 (ppb)	7.6 ± 6.4	5.3 ± 3.5	12.7 ± 11.1	14.4 ± 11.9	9.3 ± 4.6
NOx (ppb)	17.2 ± 12.1	7.8 ± 5.9	18.7 ± 8.8	23.2 ± 21.4	14.5 ± 12.1
O3 (ppb)	10.1 ± 8.1	15.1 ± 14.5	23.2 ± 15.7	15.8 ± 14.3	16.0 ± 12.9
PM10 (μg m−3)	72.2 ± 41.0	48.3 ± 22.5	75.7 ± 61.4	46.7 ± 41.9	54.1 ± 42.6
PM2.5 (μg m−3)	53.2 ± 21.9	35.9 ± 31.0	40.7 ± 40.5	24.6 ± 21.1	35.5 ± 23.2
BC (μg m−3)	5.1 ± 2.2	3.3 ± 1.7	4.2 ± 3.4	2.9 ± 1.9	3.8 ± 3.0
CO (ppb)	600.1 ± 176.4	594.0 ± 150.8	698.0 ± 105.8	638.9 ± 221.9	616.7 ± 261.1
babs, 370 (M m−1)	150.2 ± 71.9	114.6 ± 81.7	116.2 ± 21.3	73.0 ± 51.5	113.3 ± 31.4
babs, 880 (M m−1)	35.8 ± 34.1	22.1 ± 11.0	27.7 ± 21.6	19.4 ± 21.7	23.0 ± 23.3
babs,370, BC (M m−1)	75.3 ± 27.6	52.3 ± 24.0	65.2 ± 24.5	37.4 ± 29.8	57.5 ± 27.1
babs,370, BrC (M m−1)	73.9 ± 39.2	61.9 ± 61.6	48.5 ± 30.3	19.3 ± 18.7	50.9 ± 54.7
AAE 370–880	2.0 ± 0.4	2.0 ± 0.5	1.8 ± 0.3	1.6 ± 0.1	1.9 ± 0.3

The values are mean ± standard deviation. The AE33-measured BC can be defined as equivalent BC (eBC) unless otherwise stated (Petzold et al., 2013). The selected wavelengths for BC were typical among BC studies using aethalometers (Harrison et al., 2013; Wang et al., 2011; Zhu et al., 2017).

^aFrom January 17th to January 24th.

^bFrom January 25th to March 15th.

^cFrom March 15th to March 24th.

^dFrom March 24th to April 1st.

Fig. 3.

Comparisons of fractional changes of pollutants in different levels of RMPHE to those in the same periods of 2019, the prior year.

NO_x and O₃ acted uniquely during different levels of RMPHE. As shown in Table 1, NO_x decreased from 17.1 ± 12.1 ppb to 7.8 ± 5.9 ppb under the Level 1 RMPHE. As there was no significant change in meteorological conditions, these results indicate that NO_x emitted from traffic was at 58% in the urban areas of Chongqing, while the average daily O₃ increased to 15.1 ± 14.5 ppb during Level 1 RMPHE and 23.2 ± 15.7 ppb during Level 2 RMPHE. We also used the maximum total solar radiation to normalize O₃ production. We normalized the effects of solar radiation using the maxima of the total solar radiation. O₃ production was at 66.8 ppb m²/w, 62.3 ppb m²/w, 82.3 ppb m²/w, 103.9 ppb m²/w, and 90.0 ppb m²/w in those stages, showing an increasing trend during the depletion of NO_x. Chongqing, similar to other regions such as Beijing-Tianjin-Hebei and the Yangtze River Delta, also observed an ozone-boosting case during the ongoing COVID-19 outbreak due to the reduction of NO_x (Huang et al., 2020; Le et al., 2020).

3.2. Impact of COVID-19 on aerosol light absorption

The average b_abs (370 nm) was 113 ± 31.4 M m⁻¹, with 150.2 ± 71.9 M m⁻¹, 114.6 ± 81.7 M m⁻¹, 116.2 ± 21.3 M m⁻¹, and 73.0 ± 51.5 M m⁻¹ before Level 1, at Level 1, at Level 2, and at Level 3 RMPHE respectively (Table 1). The b_abs (370 nm) was higher than in the winter in Guangzhou (56.0 M m⁻¹) (Qin et al., 2018) and Lhsha 53 ± 46 M m⁻¹ (Zhu et al., 2017). The average b_abs (880 nm) was 23.0 ± 23.3 M m⁻¹, with 35.8 ± 34.1 M m⁻¹, 22.1 ± 11.0 M m⁻¹, 27.7 ± 21.6 M m⁻¹, and 19.4 ± 21.7 M m⁻¹ before Level 1, at Level 1, at Level 2, and at Level 3 RMPHE respectively. Also, the RMPHE lockdown measures reduced both b_abs (370 nm) and b_abs (880 nm) significantly to 67% and 55% respectively until February 9th, with total solar radiation increasing from 35.5 w m⁻² to 53.6 w m⁻². The average AAE (370–880 nm) was 1.9 ± 0.3, showing a decreasing trend from 2.0 to 1.6 from before Level 1 to Level 3 RMPHE.

Both BC and BrC were the most important light absorbers at multiple wavelengths. b_{abs, BrC} (370 nm) was 73.9, 61.9, 48.5, and 19.3 M m⁻¹ during the different RMPHE. The attributions to BC and BrC at multiple wavelengths between 370 and 880 are shown in Fig. 4 . BrC accounted for 49.4% of total absorption at 370 nm before the Level 1 RMPHE, increased up to 54.3% during the Level 1 RMPHE, then decreased to 42.6% during the Level 3 RMPHE. The fraction of b_{abs, BrC} (370 nm) was much higher than in Guangzhou (Qin et al., 2018), Lhasa (Zhu et al., 2017), Nanjing (Wang et al., 2018), Chengdu, and Chongqing (Peng et al., 2020a; Peng et al., 2020b).

Fig. 4.

Contributions of BC and BrC to the total light absorption coefficient at different wavelengths.

3.3. Source apportionment of fossil fuel burning using aerosol light absorption

Assuming that carbonaceous aerosols were mainly emitted from fossil fuel (ff) and wood-burning (wb), the contribution of b_abs (λ) can be given by (Sandradewi et al., 2008a; Sandradewi et al., 2008b):

$${\mathrm{b}}_{\mathrm{abs}}\left(\mathrm{\lambda }\right)={\mathrm{b}}_{\mathrm{abs},\mathrm{ff}}\left(\mathrm{\lambda }\right)+{\mathrm{b}}_{\mathrm{abs},\mathrm{ff}}\left(\mathrm{\lambda }\right)$$ (5)

b_abs,ff (λ) and b_abs,wb (λ) are the aerosol absorption coefficients of BC from the emissions of fossil fuel combustion and wood burning. If these absorption coefficients are proportional to λ^−α, then we have:

$$\frac{b_{\mathit{abs},\mathit{ff}\left(470\right)}}{b_{\mathit{abs},\mathit{ff}\left(880\right)}}={\left(\frac{470}{880}\right)}^{-{\alpha }_{\mathit{ff}}}$$ (6)

$$\frac{b_{\mathit{abs},\mathit{wb}\left(470\right)}}{b_{\mathit{abs},\mathit{wb}\left(880\right)}}={\left(\frac{470}{880}\right)}^{-{\alpha }_{\mathit{wb}}}$$ (7)

If α_ff = 2 and α_wb = 1 are applied in accordance with the literature (Kirchstetter and Thatcher, 2012; Prasad et al., 2018), then the contribution from fossil fuel is determined according to the different levels of RMPHE, as shown in Fig. 5 .

Fig. 5.

Source apportionment of fossil fuel in BC.

As shown in Fig. 5, the average ratio of BC (BC_ff) from fossil fuel was 0.47 before the Level 1 RMPHE due to the prevailing biomass burning in the region (Chen et al., 2017). Therefore, during the Level 1 RMPHE, the mean ratio of BC_ff decreased to 0.43 after February 10th due to the minimum level of traffic caused by the strict control measures. Then, during the Level 2 RMPHE, BC_ff exceeded the ratio of before the Level 1 RMPHE due to the restoration of traffic. The average of BC_ff was 0.67, with a maximum fraction of 0.95 during the Level 3 RMPHE.

The area is subtropical, much warmer than northern China, and residential heating is not common in rural areas. In urban areas, electrical air conditioning is widely used for heating. Most rural residents work in cities and return to their hometowns before the Chinese Spring Festival. The lockdown made them stay longer than in previous years. It is reasonable to extrapolate that household emissions increased. In southwestern China, wildfires are the major source of biomass burning, along with wood-burning, as shown in Fig. S4. According to these fire maps, the wildfire biomass burning spots between different RMPHE were identified as a background. Therefore, household cooking and heating increased the fraction of wood-burning, causing a corresponding decrease in fossil fuel contribution. Moreover, the decrease of fossil fuel for power plants and industries could also decrease the fraction of BC_ff (Le Quéré et al., 2020).

4. Discussion

After the announcement of the Level 1 RMPHE, an unprecedented lockdown was performed all over China. Although there was a massive reduction in energy use, PM_2.5 increased by 24% compared to the same period in 2019. The results of this study and others commonly showed that the extreme lockdown inevitably harmed the economy, but was not efficiently helpful in improving air quality (Chang et al., 2020; Cheval et al., 2020; Huang et al., 2020; Le Quéré et al., 2020; Le et al., 2020; Siciliano et al., 2020). In multiple studies, the enhanced secondary pollution of aerosols was caused by the increase in oxidative capacity due to the increase in ozone (Chang et al., 2020; Huang et al., 2020). In the Beijing-Hebei-Tianjin region, the increase of heterogeneous aerosol chemistry promoted aerosol pollution under high relative humidity conditions (Le et al., 2020).

This study also provides implications on air quality management and sustainable development. The impacts of COVID-19 on air quality, carbon emissions, and socio-ecological systems form comprehensive topics that are useful for evaluating more sustainable ways to restart society after recovering from the COVID-19 outbreak. Our results suggest that a social shut down at any cost does not effectively improve air quality in Southwestern China, and the promoted ozone and secondary aerosol levels were unexpected (Wang et al., 2020). Therefore, future mitigation measures of air pollutants in China will eventually depend on the innovation of reducing emissions from industries, traffic, residential activities, and energy consumption. The cake-cutting shutdown of particular sources is not an efficient or effective way to improve air quality. Due to the nature of VOCs-limited areas, the decrease of NOx emission caused severe ozone pollutions. The policymakers should carefully consider the interaction between the pollutants of natural and anthropogenic emissions. Also, the “Global Human Confinement Experiment” provides a unique case for the study of sustainability and the corresponding policy analysis (Bates et al., 2020). Additionally, combined with emission reduction, cases on a global scale can be used to test and improve the current models of both air quality and climate.

This study focuses on essential pollutants, the light-absorbing properties, and sources of BC aerosols. A year-on-year comparison was also performed to understand the pandemic's impact and its control in Southwestern China. However, there are still limitations. In this study, the sampling site can represent the air quality on a regional scale. Air quality is also important in urban areas such as Chongqing and Chengdu, which are megacities with populations of over 10 million. Furthermore, detailed research of PM_2.5 chemical composition and source apportionment is also necessary to explore the changes of primary and secondary contributions due to the arising of ozone during the COVID-19 pandemic. We are preparing works on these aspects in the next steps.

5. Conclusions

In order to control the ongoing COVID-19 pandemic and block the transmission of the virus, strict lockdown measures have been in place since late January 2020. The total lockdown dramatically impacted production, transportation, and consumption in Chinese society. We have analyzed the impact on air quality and aerosol light absorption in a typical urban area of southwestern China. Due to the lockdown, pollutants such as PM_2.5, PM₁₀, SO₂, NO_x, and BC decreased by 30–50%. Meanwhile, the decrease of NO_x caused the rise of O₃ due to VOC-limited areas. We conclude that basic power generation, industry, and transport provided ~47% of NO_x and 68% of PM_2.5. Due to the decrease of PM_2.5, the aerosol light absorption coefficient at multiple wavelengths decreased by 50%, and AAE decreased by 20%. BrC played essential roles in light absorption after the RMPHE was announced, accounting for 54.0% of b_abs (370 nm). Moreover, the lockdown down-weighted the fraction of fossil fuel in BC concentration from 0.47 to 0.43 (minima). Accompanied by the reopening, the ratio of BC_ff increased to 0.95. In conclusion, the extreme conditions of air quality caused by COVID-19 and its control measures are providing us with a unique opportunity to evaluate anthropogenic emissions and their subsequent influence.

CRediT authorship contribution statement

Yang Chen: Conceptualization, Formal analysis, Writing - original draft, Writing - review & editing, Project administration. Shumin Zhang: Methodology, Formal analysis, Writing - original draft. Chao Peng: Methodology, Formal analysis. Guangming Shi: Methodology, Formal analysis. Mi Tian: Conceptualization, Methodology, Formal analysis, Writing - review & editing. Ru-Jin Huang: Investigation, Validation. Dongmei Guo: Investigation. Huanbo Wang: Investigation. Xiaojiang Yao: Investigation. Fumo Yang: Methodology, Investigation.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Editor: SCOTT SHERIDAN

Appendix A. Supplementary data

Supplementary figures