Severe air pollution events not avoided by reduced anthropogenic activities during COVID-19 outbreak.

Due to the pandemic of coronavirus disease 2019 in China, almost all avoidable activities in China are prohibited since Wuhan announced lockdown on January 23, 2020. With reduced activities, severe air pollution events still occurred in the North China Plain, causing discussions regarding why severe air pollution was not avoided. The Community Multi-scale Air Quality model was applied during January 01 to February 12, 2020 to study PM2.5 changes under emission reduction scenarios. The estimated emission reduction case (Case 3) better reproduced PM2.5. Compared with the case without emission change (Case 1), Case 3 predicted that PM2.5 concentrations decreased by up to 20% with absolute decreases of 5.35, 6.37, 9.23, 10.25, 10.30, 12.14, 12.75, 14.41, 18.00 and 30.79 μg/m3 in Guangzhou, Shanghai, Beijing, Shijiazhuang, Tianjin, Jinan, Taiyuan, Xi'an, Zhengzhou, Wuhan, respectively. In high-pollution days with PM2.5 greater than 75 μg/m3, the reductions of PM2.5 in Case 3 were 7.78, 9.51, 11.38, 13.42, 13.64, 14.15, 14.42, 16.95 and 22.08 μg/m3 in Shanghai, Jinan, Shijiazhuang, Beijing, Taiyuan, Xi'an, Tianjin, Zhengzhou and Wuhan, respectively. The reductions in emissions of PM2.5 precursors were ~2 times of that in concentrations, indicating that meteorology was unfavorable during simulation episode. A further analysis shows that benefits of emission reductions were overwhelmed by adverse meteorology and severe air pollution events were not avoided. This study highlights that large emissions reduction in transportation and slight reduction in industrial would not help avoid severe air pollution in China, especially when meteorology is unfavorable. More efforts should be made to completely avoid severe air pollution.

Introduction

Along with the substantial development in the past decades, China is suffering severe air pollution, causing significant economic loss and health outcomes (Han et al., 2016a; Huang et al., 2012; Cheng et al., 2013; Chai et al., 2014; Zhao et al., 2013; Xu et al., 2015; Chen et al., 2016; Cao et al., 2011; Richter et al., 2005; Zhang et al., 2012b). With the expectation of controlling air pollution problem, the Chinese government released and implemented the Air Pollution Prevention and Control Action Plan (the “Action Plan”) in September 2013 (China, 2019b). From the plan, the concentrations of fine particulate matter (PM2.5) in Beijing-Tianjin-Hebei (BTH), Yangtze River Delta (YRD) and Pearl River Delta (PRD) area shall decrease by 25%, 20% and 15% by the end of 2017 compared with 2012, respectively. To fulfill this goal, tremendous effects have been made such as strengthen industrial and vehicle emission standards, closing small and polluting factories, and upgrades on industrial boilers, and air quality has been largely improved (Zhang et al., 2019; Ma et al., 2019; China 2019a). The severe particulate pollution days were decreased from 122 and 33 in 2013 to 31 and 25 in 2017 within BTH and YRD areas (Li et al., 2019).

Meteorology plays significant roles in air pollution formation, transport, deposition and transformation and unfavorable meteorological condition could bring severe pollution days even the total emission is reduced (Shi et al., 2018; Xu et al., 2016; Li et al., 2017c; Cai et al., 2018; Gui et al., 2019; Wang et al., 2019b). In the pollution period of December 20–26, 2015 in BTH region, about 34% increasing of PM2.5 are believed due to adverse meteorological condition (Ma et al., 2020). Wu et al. (2019) found a high relative humidity (RH) enhanced aqueous-phase oxidation of SO2 explosive growth was the main reason for the 7 days heavy PM2.5 pollution episode in North China Plain in December 2016. In the study of two haze episodes from January 6 to 16 in Chengdu Plain, Li et al. (2017a) indicated that the decreased Planetary boundary layer (PBL) height restrained the air pollutant dispersion and increased RH accelerated the transformation of secondary pollutants. Meteorological condition in winter are more unfavorable to pollutant dilution and dispersion than other three seasons (Yang et al., 2019). Compared with December 2015, the PM2.5 concentration increased by 36% in BTH region. Liu et al. (2017) attributed the increase to unfavorable meteorology of PBL, RH and wind.

Can severe air pollution events be avoided by controlling anthropogenic emissions? The answer is true. China has succeeded in various emission reduction campaigns such as during the 2014 Asia-Pacific Economic Cooperation (APEC) meeting and the 2015 China Victory Day Parade (Xu et al., 2017; Wang et al., 2017; Xue et al., 2018; Li et al., 2016). In order to reduce anthropogenic emission, factories, industrial plants, construction sites, and gas stations was temporarily closed in and around Beijing, vehicles were kept off the roads, and 6-days mandatory holidays were brought to state-owned enterprises, local government offices and educational institutions (Sun et al., 2016; Wang et al., 2016; Li et al., 2017d; Huang et al., 2015). “APEC Blue” and “Parade Blue” are referred to the good air quality during the time afterward. The daily PM2.5 concentrations in Beijing were 47.53 µg/m3 during “APEC blue” and 17.07 µg/m3 during “Parade blue” (Lin et al., 2017).

Coronavirus disease 2019 (COVID-19) is an infectious disease initially identified in Wuhan, China in December 2019. COVID-19 has led to 2700 deaths by February 25, 2020 worldwide (WHO, 2020). Due to the contagion of COVID-19, the unofficial transit going in and out of Wuhan was shut down on January 23, 2020 and in the whole Hubei Province a couple of days after. The 2020 Spring Festival holiday started on January 24, and was extended to after February 10 nationwide because of the disease. Most transportation was prohibited and almost all avoidable outdoor human activities stopped all around the country. However, severe air pollution episodes still occurred in the North China Plain (NCP). It is of great concern for people who are limited to use mobile vehicles on heavy pollution days why severe air pollution events were not avoided when no one was on road.

Thus, this study used the Community Multi-Scale Air Quality (CMAQ) model to assess emission and meteorology conditions for air pollution during the outbreak of COVID-19. Different emission cases were simulated to see why the severe air pollution was not avoided.

Methods

The CMAQ V5.0.1 model was applied to simulate air pollution in China with a horizontal resolution of 36 km × 36 km during January 1 to February 12, 2020 (Fig. 1 ). The model has been updated to better predict air pollutants in China and details can be found in Hu et al. (2016). The meteorological inputs were generated using WRF v3.7.1 with initial and boundary conditions from the National Center for Environmental Prediction (NCEP) FNL Operational Model Global Tropospheric Analyses dataset (Zhang et al., 2012a; NCEP, 2000). Multi-resolution Emission Inventory for China of 2016 (MEIC) (http://www.meicmodel.org) was used for the monthly anthropogenic emissions from China. Emissions from other countries were obtained from MIX Asian emission inventory (Li et al., 2017b). The Model of Emissions of Gases and Aerosols from Nature (MEGAN) v2.1 was used for generating biogenic emissions (Guenther et al., 2012). The emissions from biomass burning were based on the Fire Inventory from the National Center for Atmospheric Research (NCAR) (FINN) (Wiedinmyer et al., 2011).

Fig. 1

Locations of domains and major cities (red point), Wuhan is marked as yellow star and Hubei province is in blue. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

In order to investigate the effect of reduced anthropogenic activities on air pollution, three simulation scenarios were performed and compared (Table 1 ). The base case scenario (Case 1) is using the original anthropogenic emission inventory of MEIC 2016 for the simulation period. Although the emissions keep changing since 2016, this does not affect the study much as emission inventories usually have larger uncertainties in comparison to changes in a few years (Zhao et al., 2011; Zheng et al., 2009; Fu et al., 2013; Chen et al., 2014). Compared with the Case 1, Case 2 had a decrease of 40% and 20% in transportation and industry emission, respectively. Emission from residential was lifted by 10% and emissions from agriculture and power remain unchanged. In Case 3, transportation, industry and agriculture emission in Hubei province are decreased by 80% while residential and power emission is kept the same with the Case 1. For areas out of Hubei province, 80% of transportation emission and 20% of industry emission are cut while all the rest are kept still. Since there is no official emission inventory released, emission changes suggested by the Chinese Research Academy of Environmental Sciences are being used to set emission reduction scenarios (CRAES 2020). An interview of Professor Kebin He from Tsinghua University refers to the high-pollution in BTH regions in February 2020 by China Central Television (CCTV) was also considered (CCTV, 2020). Many people came back for the Spring Festival and were trapped due to COVID-19 epidemic, the residential emission from cooking and heating was asummed to increase in rural areas including Hebei, Henan and Shandong. The notices released from goverment were the basis for emission reduction in Hubei Province (Hubei, 2020f, 2020e, 2020d, 2020c, 2020b). The governme announced that from 10 am in January 23, 2020, buses, subways, ferries, and long-distance passenger transportation in Wuhan were temporarily suspended, and outbound of railway and air planes were closed (Hubei, P. s. G. o. 2020g). All the enterprises are remaining closed no earlier than February 13 except required for public service operation, epidemic prevention and control, and residential life needs (Hubei, P. s. G. o. 2020a). The agriculture production was highly affected because there was nowhere to buy seeds and fertilizer (Daily, 2020).

Table 1

Configuration of simulation scenarios.

	Transportation	Industry	Residential	Agriculture	Power
Case1	No change
Case2	−40%	−20%	+10%	No change	No change
Case3	−80%	−80%	No change	−80%	No change	Hubei
	−80%	−20%	No change	No change	No change	Other

Hourly PM2.5 observations at monitoring sites nationwide were downloaded from the publishing website of the China National Environmental Monitoring Center (http://www.mee.gov.cn/hjzl/dqhj/). The data were validated, and sanity check was conducted as previous studies (Hu et al., 2015; Wang et al., 2014). Meteorological observations were obtained from National Climate Data Center (NCDC), including precipitation, WS and wind direction (WD) at 10 m, as well as air temperature (T2) and RH at 2 m above the ground level.

To better understand the role of emissions changes and meteorological conditions, a rough method is proposed with similar idea of a previous study (Wang et al., 2019b; Liu et al., 2017; WEI et al. 2017; Li et al., 2015). The changes due to emission changes are calculated by subtracting the concentrations in Case 3 from that in Case 1, while the changes due to meteorology are calculated using concentrations in high-pollution days minus concentrations in low-pollution days, assuming that meteorology is the reason for severe pollution as the emissions are not changing day by day. It should be note that this is a rough estimate without considering the feedbacks of aerosols on meteorology.

Results

Model validation

The performance of meteorology simulation by WRF for the simulation period is validated including T2, WS, and WD at 10 m and RH, as statistic results are shown in Table S1. The benchmarks are suggested by Emery et al. (2001). WRF has slightly underpredicted T2 with MB of −0.9. Both T2 and WD went beyond the GE benchmark by 40% to 50%. The GE of WS is within the benchmark, and high RMSE of 2.5 indicated oscillations among prediction results. WD prediction results were acceptable with MB of 8.5 while RH was slightly under-predicted. It should be declared that out of the benchmark does not mean failure of the prediction as the benchmark value is constructed on the MM5 simulations for the eastern United States with finer grid resolutions (12 km and 4 km). The WRF model predictions are normally reliable since the model performance statistics are like other WRF studies applied over China (Zhang et al., 2012a; Qiao et al., 2019; Hu et al., 2016).

Tables S2 and S3 show the model performance statistics of PM2.5 from the two emission reduction cases. Mean observations, mean predictions, mean fractional bias (MFB), mean fractional error (MFE), mean normalized bias (MNB) and mean normalized error (MNE) were calculated for each city from January 02 to February 12, 2020. PM2.5 is well predicted in all cities in both cases except Wuhan in Case 2 with MFB exceeding the criteria suggested by EPA (EPA 2007). As activities in Hubei were much more strictly limited, Case 3 is a better estimation and is used in following discussion.

Fig. 2 shows the predicted and observed daily PM2.5 and its components in 10 major cities based in Case 3. Compared with observation results, the model presented great performance. In Wuhan, the model captured the changes of PM2.5 except for few overpredicted days. In all the other cities, the model successfully predicted peak and valley values not only in the most polluted cities like Xi'an, Zhengzhou but in relatively less polluted cities such as Shanghai and Guangzhou. The model also succeeded in reproducing the significant decrease of PM2.5 that occurred from 29 January to 6 February in all cities.

Fig. 2

Predicted daily PM2.5 and its components with observed daily PM2.5 in 10 major cities in Case 3. Units are µg/m³.

Regional changes

The changes of predicted PM2.5 and its major components between two emission reduction cases (Case 2 and Case 3) and Case 1 are shown in Fig. 3 . The total PM2.5 is decreasing mainly in south China in both cases (Case 2 and Case 3), while an increasing (up to 4 µg/m3) is observed in Northeast China in Case 2 as residential activities were increased possible due to more people were staying at home. More significant PM2.5 decrease (averaged 10–20%) is observed in Case 3 in Central China, while for Case 2 the averaged decreased rate is less than 10%. Among all the components, NO3 shares the largest reduction since transportation is the key factor in both emission reduction cases. Except EC and POA, all the species were decreasing in Case 2. A 5% increase in EC is shown in Northeast and Southwest China (Fig. S3). POA increases by 10% on average in large areas of China with peak values in Northeast, Henan, Hebei and the Sichuan Basin as residential emissions were increased.

Fig. 3

Predicted PM2.5 and its major components and the changes between cases in unit of µg/m³ during January 20 to February 12, 2020. SO4 is sulfate, NO3 is a nitrate, NH4 is ammonium, EC is elemental carbon, POA is primary organic aerosol, SOA is secondary organic aerosol and OTHER is the sum of inexplicit components.

All the country experiences reduction of PM2.5 by 10~20% due to the significant emission reductions in Case 3. Most significant decreases of PM2.5 and its components (30–50%) of SO4, NO3, NH4, and SOA are observed in Hubei (Fig. S3). Based on the emission scenario, the falling of SO4 concentration mostly results from the emission reduction of industry release. The decrease of NO3 is likely due to the reduction of transportation and the drop of agriculture emission explains the NH4 decrease. The amount of SOA was fell because of the lack of precursors due to the emissions reduction of all three. Overall, the emission reductions due to COVID-19 did result in reductions of PM2.5 by 10–20% nationwide.

Changes in different cities

Fig. 4 shows the averaged predicted and observed PM2.5 concentrations with components in 10 major cities of these three scenarios from January 21 to February 12, 2020. The positive effect of emission reduction on decreasing PM2.5 concentrations can be clearly observed. At least 5 μg/m3 (Guangzhou) decrease was found in each city while Wuhan had the largest decrease up to 30.79 μg/m3. Compared with the Case 1, the decreases of PM2.5 in Case 3 in Beijing, Tianjin, Shijiazhuang, Jinan, Taiyuan, Zhengzhou, Xi'an, and Shanghai were 9.23, 10.30, 10.25, 12.14, 12.75, 18.00, 14.41, and 6.37 μg/m3, respectively.

Fig. 4

Predicted daily PM2.5 with components and observed PM2.5 in 10 major cities of three scenarios during January 20 to February 12, 2020. Units are µg/m³.

Fig. 5 shows the predicted and observed PM2.5 during the high-pollution days and low-pollution days in theses 10 cities (separated based on the standard of the second level of Chinese NAAQS, 75 μg/m3 of observation values). The PM2.5 reductions in high-pollution days in Case 3 were 13.42, 22.08, 14.42, 11.38, 7.78, 13.64, 16.95, 14.15, and 9.51 μg/m3 in Beijing, Wuhan, Tianjin, Shijiazhuang, Shanghai, Taiyuan, Zhengzhou, Xi'an, and Jinan. All the changes were larger than 10% compared with PM2.5 concentrations in Case 1 and the largest decrease of 19.04% was found in Wuhan. During low-pollution days, the reductions were 6.55, 32.63, 4.94, 7.07, 6.07, 11.60, 20.43, 14.81, and 14.57 μg/m3 in above cities with the same consequence, separately. With a 29.20% decrease in low-pollution days, Wuhan ranked first. Tianjin and Shijiazhuang had the decreases of 9.5% and 9.4%, separately. The reduction percentage in all other 6 cities were larger than 10%. Basically, secondary components of SOA, SO4 and NO3 contributed the most both in high and low-pollution days. This indicates that the role of chemistry and meteorological condition in secondary pollutants formation should not be neglected.

Fig. 5

Averaged predicted daily PM2.5 with components and observed PM2.5 in 10 major cities of three scenarios on high-pollution days (>75 µg/m³, left column) and low-pollution days (right column). The number of days in high (H) and low (L) pollution days are marked after the city names. Units are µg/m³.

Comparison of changes in emission and concentrations

As we can see from above results that the concentrations of PM2.5 did reduce within the emission reductions cases. But it is important to know how efficient it was. Fig. 6 shows decreasing ratios of predicted PM2.5, EC, OC, SO4, NO3, and SOA concentrations by Case 3 and the emission rates of related precursors in Beijing, Wuhan, Guangzhou and Shanghai during the simulation period. All species decreased as the emissions were reduced in a monotonic but non-linear way. This is consistent with the nonlinearity of atmospheric processes as confirmed by previous studies (Cai et al., 2017; Zhao et al., 2017; Han et al., 2016b; Wang et al., 2011).

Fig. 6

Relative changes in concentrations of primary PM2.5 (PPM) and secondary components, and in emissions of related precursors in Case 3 to Case 1 in four cities.

In Beijing, emissions of primary PM2.5, EC and OC were decreased by 20%, 50% and 10%, while the concentrations of them were decreased by 10%, 25% and 5%. SO2, NOx, and VOCs emissions were decreased by 20%, 50%, and 30% while concentrations of SO4, NO3, and SOA were decreased all by less than 20%. Similar cases are observed in other cities. For Wuhan and Guangzhou, the reductions in concentrations were about half of the decreases in emissions. While for Shanghai, the emission reduction ratios were only slightly higher than the concentration reduction ratios. Although nonlinear chemical processes are noneligible in this situation, effects of meteorology should not be neglected.

Role of meteorology

Through the emission reduction process, the decreases in pollutant concentrations were observed. The ratios of PM2.5 and precursors reductions in emission are larger than the reductions in PM2.5 and its components in four selected cities. This indicates that meteorological conditions were favorable for pollution formation. Fig. S4 showed the time series of meteorological parameters in Beijing and Guangzhou from January 21 to February 12, 2020. There was seldom wet deposition of PM in Guangzhou and Beijing due to the rare of rains during the simulation period. The low PBL increased atmospheric stability, the low wind speed made it worse for difficult dispersion of air pollutants (Wang et al., 2015; Liu et al., 2017). High RH and temperature usually accelerate secondary PM formation by speeding up chemical reactions (Li et al., 2018; Wang et al., 2019a; Rahman et al., 2019). Thus, it is very important to quantify the contribution of meteorological conditions when deciding controlling efficiency.

The contributions of emission reduction and meteorological conditions are estimated and shown in Fig. 7 . Emission reductions led to the reductions of pollutants in all cities with the largest in Wuhan (−22.08 μg/m3) and the lowest in Shanghai (−7.78 μg/m3) and Jinan (−9.51 μg/m3). All other cities experienced a >10 μg/m3 contributions, with the exactly values of −13.42, −14.42, −11.38, −13.64, −16.95, and −15.15 μg/m3 in Beijing, Tianjin, Shijiazhuang, Taiyuan, Zhengzhou, and Xi'an, separately. In contrary, unfavorable meteorology resulted in increases of PM2.5, with the largest value of 69.38 μg/m3 in Tianjin. For Wuhan and Jinan, unfavorable meteorology had less effect compared with emission changes. In Xi'an, the unfavorable meteorology brought 14.24 μg/m3 increase of PM2.5, which almost fully counteracts the contribution of emission reductions. In all other cities, the increases caused by meteorological conditions timely overstepped the positive influences from emission reductions.

Fig. 7

Estimated contributions of meteorology and anthropogenic emissions changes to averaged PM2.5 concentrations in selected cities during January 20 to February 12, 2020. Units are µg/m³. Guangzhou is not included as it had no high-pollution days during the episode.

Conclusions

In this study, the influences of emission reductions due to reduced anthropogenic activities during the COVID-19 outbreak in China on air pollution were investigated. It is concluded that anthropogenic emission decreases, mainly on transportation and industry, contributed to the decreases of PM2.5 concentrations. The decreases of PM2.5 in Beijing, Shanghai, Guangzhou, and Wuhan were 9.23, 6.37, 5.35, and 30.79 μg/m3, respectively. However, this is not enough to avoid severe air pollution events in most areas. The reduction ratios of PM2.5 concentrations were smaller than the reduction ratios of precursor emissions, partially due to the unfavorable meteorological conditions. This study highlights the importance of understanding the role of chemistry and meteorology in designing emission control strategies.

Author statement

Pengfei Wang: Methodology, Simulation, Data collection, Visualization, Writing-Original draft preparation

Kaiyu Chen: Data collection, Post-processing, Visualization

Shengqiang Zhu: Data collection, Visualization

Peng Wang: Methodology, Writing-Reviewing and Editing

Hongliang Zhang: Conceptualization, Methodology, Writing-Reviewing and Editing.

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.