The positive impact of the Omicron pandemic lockdown on air quality and human health in cities around Shanghai

Abstract

The Omicron pandemic broke out in Shanghai in March 2022, and some infected people spread to some cities in the Yangtze River Delta (YRD) region. To achieve the dynamic zero-COVID target as soon as possible, Shanghai and nine cities that were heavily affected by Shanghai implemented the lockdown measures. This paper aims to quantify the impact of the lockdown on air quality and human health. A difference-in-difference (DID) model was first used to measure the impact of the lockdown on air quality in these ten cities. Based on the results of the DID model, we estimated the PM_2.5-related health and economic benefits using the concentration–response function and the value of statistical life method. Results showed that the lockdown has reduced the concentrations of PM_2.5, PM₁₀, SO₂, NO₂, and CO by 9.87 μg/m³, 17.31 μg/m³, 0.75 μg/m³, 9.03 μg/m³, and 0.07 mg/m³, respectively. The number of avoided premature deaths due to PM_2.5 reduction was estimated to be 35,342. The resulting economic benefits totaled 18.86 billion US dollars. We investigated the reasons for the air quality improvement in these ten cities and found the “3 + 11” policy has had a great impact on air quality. Compared with the first COVID-19 lockdown in early 2020, the effect of the lockdown in 2022 was smaller. These findings demonstrated that reductions in anthropogenic emissions would achieve substantial air quality improvement and health benefits. This paper re-emphasized continuous efforts to improve air quality are essential to protect public health.

Introduction

On March 1, 2022, Shanghai Municipal Health Commission announced a new confirmed COVID-19 case, which was later confirmed to be infected with the Omicron BA.2 variant. Since then, Shanghai experienced its worst COVID-19 outbreak since the pandemic began in early 2020. To achieve the dynamic zero-COVID target as soon as possible, strict control measures were gradually implemented. As the situation worsened, Shanghai decided to impose a two-phase citywide lockdown on March 28, 2022 for city-wide nucleic acid testing, but later switched to citywide static management in early April. More than 25 million people were ordered to stay at home. Because the Omicron variant BA.2 is extremely contagious, some infected people spread to other cities in the YRD region, despite Shanghai's strict prevention and control measures. In the affected cities, the governments have taken strict prevention and control measures such as the “3 + 11” and the “7 + 7” policies. The “3 + 11” policy required that people from other cities be placed in centralized isolation for 3 days before being monitored for health conditions for an additional 11 days. The “7 + 7” policy required that people from Shanghai be placed in centralized isolation for 7 days before being quarantined at home for 7 days. These two policies severely hampered social mobility and logistics, effectively paralyzing the supply chain in the YRD region. As a result, many factories in the YRD region were forced to halt operations. The YRD region's production and economic activities were severely harmed. For example, the value of Shanghai’s gross industrial output fell by 61.5% in April 2022 compared to the same month the last year. Industrial added value in Jiangsu Province in April 2022 fell by 12.3% compared to the same period last year.

As China’s economic hub, no doubt that the economic cost of pandemic containment in the YRD region is enormous. Nevertheless, there have been some unexpected environmental benefits. Comprehensive prevention and control measures resulted in huge and large-scale reductions in air pollutants concentrations, which would partially offset the cost of anti-COVID-19 measures. After the outbreak of the COVID-19 pandemic in 2020, various studies have reported significant reductions in air pollutants concentrations including PM_2.5, PM₁₀, SO₂, NO₂, and CO due to the restriction measures during the lockdown (Berman & Ebisu, 2020; Filonchyk et al., 2021; Ikhlasse et al., 2021; Jephcote et al., 2021; Kumar et al., 2021; Pandey et al., 2021; Song et al., 2021; Vega et al., 2021). Table 1 shows the overview of the previous studies which studied the impact of the COVID-19 lockdown on air quality and human health. Table 1 includes the characteristics such as case study areas, air quality parameters, the novelty of the study, and outputs. For instance, ground-based observations in California during the lockdown (March 19 to May 7) show a 38%, 49%, and 31% drop in NO₂, CO, and PM_2.5 concentrations, respectively, compared to the pre-lockdown period (January 26 to March 18) in 2020 (Liu et al., 2021b). Celik and Gul (2022) investigated how this pandemic affects the air quality in Istanbul, Turkey. They found the concentrations of PM₁₀, NO₂, NO, and NOx declined significantly compared to the normal times, and a non-homogenous trend for SO₂ and CO concentrations was observed for different air monitoring stations. In Kolkata, India, the pollutants like PM₁₀, PM_2.5, CO, NO₂ and SO₂ significantly decreased, while the average level of O₃ slightly increased (Bera et al., 2021). Ganguly et al. (2021) analyzed the impact of lockdown on the air quality in three Indian megacities and found the concentrations of PM₁₀ and NO₂ decreased by 30–60% and 52–80%, respectively. In another study, He et al. (2020) evaluated the impact of prevention and control measures on air quality index (AQI) and PM_2.5 in Chinese locked-down cities. They found that AQI and PM_2.5 reduced by 19.84 points and 14.07 μg/m³ within weeks relative to cities without the lockdown policy. A more important finding is that the lockdown effects are larger in colder, richer, and more industrialized cities. Similarly, Wang et al. (2021) believed that the effects of the COVID-19 prevention and control measures are heavily influenced by city characteristics such as traffic development level and industry structure. Furthermore, some works attempted to understand the role of meteorological factors in pollution levels and observed reductions (Khatri & Hayasaka, 2021; Petetin et al., 2020). There is even a study that suggested air quality improvements during the lockdown were due to seasonal variation and meteorological factors rather than restricted movements (Jakob et al., 2022). In addition, unfavorable meteorological conditions such as increased humidity and decreased planetary boundary layer heights could cause PM_2.5 pollution during the COVID-19 lockdown in Nanning, China (Mo et al., 2021).

Table 1.

Overview of the impact of the COVID-19 lockdown on air quality and human health

No.	Study	Study areas	Air quality parameters	Novelty of the study	Outputs
1	Pandey et al. (2021)	Delhi, India	NO2, PM2.5, and PM10	The impact of the COVID-19 pandemic on the air quality	The city's air quality had improved significantly during the COVID-19 lockdown
2	Vega et al. (2021)	Mexico City, London, and Delhi	PM10, PM2.5, NO2, O3 and CO	The air quality changes across these cities during the COVID-19 lockdown	Major reductions in PM10, PM2.5, NO2 and CO and increases in O3 during the lockdown occurred in London and Mexico City but not Delhi
3	Filonchyk et al. (2021)	Poland	PM2.5, PM10, NO2, and SO2	The impact on the air quality resulting from the pandemic control measures	Restrictions imposed to control the COVID-19 pandemic significantly improved Poland’s air quality
4	Berman and Ebisu (2020)	The USA	PM2.5 and NO2	Air quality changes during the COVID-19 pandemic in the continental USA	The concentrations of PM2.5 and NO2 statistically decreased
5	Song et al. (2021)	China	PM2.5, PM10, and NO2	The relationship between anti-epidemic measures and air quality	The concentrations of the PM2.5, PM10, and NO2 decreased by 19.01 μg/m3, 20.20 μg/m3, and 2.13 μg/m3, respectively
6	Kumar et al. (2021)	India (9 cities)	NO2, CO, and SO2	An assessment of air quality changes during the period of lockdown and unlocking	The concentrations of NO2, CO, and SO2 greatly decreased
7	Jephcote et al. (2021)	The UK	NO2, O3 and, PM2.5	The air quality changes across the UK during the period of most stringent travel restrictions	The lockdown reduced NO2 and PM2.5 concentrations by 38.3% and 16.5% and increases the O3 concentrations by 7.6%
8	Ikhlasse et al. (2021)	France	SO2, NO2, CO, O3, C6H6, NOX, PM2.5 and PM10	The impact of the COVID-19 on air quality in France	All pollutants concentrations decreased greatly except O3
9	Liu et al., (2021b)	California, USA	NO2, CO, and PM2.5	The spatiotemporal patterns and changes in air pollution before, during and after the lockdown	Ground-based observations around California show a 38%, 49%, and 31% drop in the concentrations of NO2, CO, and PM2.5
10	Celik and Gul (2022)	Istanbul, Turkey	PM10, SO2, CO, NO2, NO, NOx, and O3	How this pandemic affects the air quality of a metropolis	The concentrations of PM10, NO2, NO, and NOx declined significantly compared to the normal times
11	Bera et al. (2021)	Kolkata, India	PM10, PM2.5, O3, SO2, NO2, and CO	The impact of COVID‑19 lockdown on urban air pollution in Kolkata, India	The pollutants like PM10, PM2.5, CO, NO2 and SO2 significantly decreased, while the average level of O3 slightly increased
12	Ganguly et al. (2021)	Delhi, Kolkata, and Mumbai	PM10 and NO2	The impact of lockdown on the air quality in three megacities of India	The concentrations of PM10 and NO2 decreased in the ranges of 30–60% and 52–80%, respectively
13	He et al. (2020)	China	PM2.5	The impact of the COVID-19 prevention measures on air pollution	The PM2.5 concentrations reduced by 14.07 μg/m3 compared with the control group
14	Wang et al. (2021)	Northern China	NO2, PM10, PM2.5, and CO	The lockdown effect on air quality in the northern China	The concentrations of NO2, PM10, PM2.5, and CO were reduced by 37.8%, 33.6%, 21.5%, and 20.4%, respectively
15	Jakob et al. (2022)	Jakarta, Indonesia	PM2.5 and PM10	The impact of meteorological factors on air quality during the lockdown	Rainfall can explain changes in PM10 and PM2.5 parameters in the city of Jakarta during lockdown
16	Mo et al. (2021)	Nanning, China	PM2.5	The sources and causes of PM2.5 pollution that occurred in Nanning during the COVID-19 lockdown	Unfavorable meteorological conditions such as increased humidity and decreased planetary boundary layer heights could cause PM2.5 pollution during the lockdown in Nanning, China
17	Campbell et al. (2021)	The USA	O3	The impact of the COVID-19 economic slowdown on ozone concentrations	Decreases in NOx emissions led to widespread decreases in ozone concentrations in the rural regions that are NOx-limited
18	Zhang et al. (2021)	China	O3	The impact of the lockdown on NOx emission and subsequent influence on surface ozone	The large magnitude of decrease in NOx emission has large influence on surface O3 concentrations
19	Brancher (2021)	Vienna, Austria	NO2 and O3	The impact of Vienna’s first lockdown on the NO2 and O3 concentrations	The benefit of improved air quality was offset by amplified O3 pollution during the lockdown
20	Zhang and Stevenson (2022)	London	O3	The effects of lockdown on O3 and its precursors	The increase in O3 concentrations caused by the lockdown was closely related to the rapid decrease in NOx emissions
21	Grange et al. (2021)	European	O3 and NO2	The effect of the European COVID-19 lockdowns on NO2 and O3	The NO2 concentrations were 34% and 32% lower than expected for respective traffic and urban background locations, whereas O3 was 30% and 21% higher
22	Ye et al. (2021)	China	PM2.5, PM10, CO, NO2, O3, and SO2	The impact of control measures on air pollution and the subsequent consequences on health and the health-related economy	The COVID-19 lockdown caused substantial benefits in human health and health-related costs due to improved urban air quality
23	Liu et al., (2021a)	76 countries	PM2.5, PM10, SO2, CO, and O3	The impact of 8 types of lockdown measures on air pollution and human health in 76 countries	Expected averted premature deaths due to air pollution declines are around 99,270 to 146,649 among 76 countries during the COVID-19 lockdown

However, the O₃ concentration greatly increased during the lockdown (Brancher, 2021; Campbell et al., 2021; Zhang et al., 2021). Specifically, Zhang and Stevenson (2022) investigated the effects of lockdown on O₃ and its precursors in London (i.e., NO_x and VOCs) and found that the VOCs/NOx ratio was smaller than eight during the lockdown, indicating that O₃ formation was in the VOC-limited regime. Therefore, the rapid decrease in NO_x emissions was the main reason for the rapid increase in O₃ concentrations. In Wuhan, the O₃ concentration increased by 43% compared with the corresponding period in 2015–2019 due to a weakening of the titration effect of NO (Yin et al., 2021). Grange et al. (2021) represented a classic air quality data analysis in Europe and found that NO₂ concentrations decreased by 34% at roadside locations accompanied by increases in the O₃ concentration of a similar magnitude (30%).

China, a rapidly developing country, is currently dealing with severe environmental issues, particularly air pollution. Meanwhile, air pollution has posed a great threat to public health. However, when some international events were held in China such as Olympic Games, Asian Games, and political summits, the government would implement a series of control measures such as production suspension and traffic restrictions to ensure good air quality during these major events. Studies have shown that control measures taken during major events not only improve local air quality but also result in significant health benefits (Finch, 2016; Guo et al., 2021). During the 2010 Guangzhou Asian Games, the mean PM_2.5 concentration in Guangzhou decreased by 3.5 μg/m³ and avoided 106 premature deaths, 1869 cases of hospital admissions, and 20,026 cases of outpatient visits (Ding et al., 2016). Like these major events, the COVID-19 pandemic can be viewed as an unprecedented quasi-artificial experiment. One of the most significant outcomes of this unprecedented experiment is the avoidance of premature deaths associated with air pollution reductions. For example, compared with the business-as-usual scenario, 1239 PM_2.5-related deaths, 2777 PM₁₀-related deaths, 1587 CO-related deaths, 4711 NO₂-related deaths, 215 O₃-related deaths, and 1088 SO₂-related deaths were avoided in 367 Chinese cities during February and March 2020 (Ye et al., 2021). Similarly, in the YRD region, the number of avoided premature deaths associated with PM_2.5 reduction during the lockdown is estimated to be 42.4 thousand. At the global level, the health benefits due to air pollution declines caused by COVID-19 prevention and control measures are even more evident. Liu et al., (2021a) studied air pollution changes from January 1, 2020, to July 5, 2020, in 76 countries and found the expected avoided premature deaths due to air pollution reduction are around 99,270 to 146,649.

Currently, the YRD region is working hard to reduce air pollution. This wave of the Omicron pandemic in the YRD region provided a special opportunity to observe the response of air pollutants concentrations to the short-term reduction in anthropogenic emissions. Meanwhile, the Omicron lockdown could provide references and insights into formulating environmental regulations during the recovery period. Air pollution is inextricably linked to human health. The reduction in pollutant concentrations will cause substantial health effects, avoiding premature deaths from related diseases. To provide more knowledge, this paper quantified the impact of the Omicron pandemic control measures on air quality and corresponding health benefits. This paper also provided some back-of-the-envelope calculations on the expected health benefits from the apparent air quality improvement, which contributed to the literature on the health costs of changes in air pollution concentrations. The key challenge in this paper is to quantify the impact of control measures on air quality. Some studies compared air pollutants concentrations before and after the outbreak of the COVID-19 pandemic or compared pollutant concentrations during the lockdown with the corresponding period in the previous years (Anil & Alagha, 2021; Mashayekhi et al., 2021; Singh & Chauhan, 2020; Spohn et al., 2022). This method did not consider the influence of climate factors and social factors (e.g., weekends) on pollutant concentrations, which will bring some deviations to the research results. Some studies predicted pollutant concentrations in the business as usual (BAU) scenario using the multiple linear regression (MLR) and random forest (RF) and compared the predicted results with the observed concentrations during COVID-19 lockdown (Cao et al., 2022; González-Pardo et al., 2022). However, the predicted results are not very accurate, which may result in inaccurate conclusions. Besides, some scholars employed the Community Multi-scale Air Quality model (CMAQ) to simulate pollutant concentrations during the COVID-19 lockdown period and compared them with observed concentrations (Liu et al., 2021c; Wang et al., 2020). However, the simulated data are smaller than the observed data, which may amplify the impact of the COVID-19 pandemic on air quality. To address these above-mentioned shortcomings, the difference-in-difference (DID) model was used in this paper because it could take meteorological factors and weekends as control variables to avoid their influence on the research results. Furthermore, the DID model can control the systematic differences between the treatment and control groups, thus ensuring the accuracy of the research results. Based on changes in PM_2.5 concentrations, the associated health and economic benefits were estimated by a widely used concentration–response function and the value of the statistical life method. Figure 1 describes the steps of implementation in this paper.

Fig. 1.

The flowchart of the complete study

Materials and methods

Study area and temporal samples

As one of the regions with the fastest economic development in the world, the YRD region consists of three provinces (Zhejiang, Jiangsu, and Anhui) and one municipal city (Shanghai), with a total of 41 cities. The YRD region nevertheless contributes 16.7% and 24.1% of the total population and GDP in 2021 despite having a relatively small area (3.7% of China's total area). If the YRD region was considered an independent economy, it could be ranked among the top five economies in the world. However, massive amounts of energy consumption, particularly the burning of fossil fuels, have had a negative impact on the local air quality and public health along with the economy's quick development. Therefore, it is highly necessary to explore effective air pollution control measures. In the past several years, the Chinese government has been promoting integrated development of the YRD region. The integrated regional development plan of the YRD region was approved by the State Council of China in 2019. While integrated development has greatly contributed to economic development, it has also aided the virus spread in the COVID-19 pandemic era. Although Shanghai has adopted strict prevention and control measures since the COVID-19 pandemic initially broke out in March 2022, some infected people have spread to other cities in the YRD region. This paper selected Shanghai and nine cities that were severely affected by Shanghai as the spatial samples. Figure 2 shows the locations of these ten cities. Six cities are in Jiangsu Province, namely Nanjing (NJ), Suzhou (SZ), Wuxi (WX), Nantong (NT), Changzhou (CZ), and Zhenjiang (ZJ). Three cities are in Zhejiang Province, namely Hangzhou (HZ), Jiaxing (JX), and Huzhou (HUZ).

Fig. 2.

The locations of the spatial samples in this paper

To prevent the spread of the COVID-19 pandemic and achieve the dynamic zero-COVID target as soon as possible, Shanghai decided to impose a lockdown in two phases for city-wide nucleic acid testing on March 28, 2022. Therefore, this paper takes March 28, 2022, as the start of the lockdown. On April 11, 2022, Shanghai started to adjust containment measures and divided residential units into three categories, including lockdown areas, controlled areas, and prevention areas. This meant shanghai was gradually returning to normal life. As the pandemic situation improved, Shanghai began to gradually resume production at the end of April. In mid-April 2022, the other nine cities gradually resumed logistics and production. Therefore, the research period of this paper ends on April 30, 2022.

Data sources

The daily PM_2.5, PM₁₀, SO₂, NO₂, CO, and O₃ concentrations from February 1, 2022, to April 30, 2022, were collected from the National Air Quality Real-time Publishing Platform. Meteorological data used in this paper including temperature, relative humidity, wind speed, atmospheric pressure, and precipitation, were collected from the National Meteorological Data Center. Table 2 shows the Summary statistics of key variables.

Table 2.

Summary statistics of key variables

Variables	Obs	Units	Min	Max	Mean	SD
PM2.5	2225	μg/m3	2	125	32.24	18.15
PM10	2225	μg/m3	3	271	57.23	31.00
SO2	2225	μg/m3	1	17	6.51	2.42
NO2	2225	μg/m3	5	71	25.19	11.33
CO	2225	mg/m3	0.2	1.27	0.61	0.16
O3	2225	μg/m3	18	157	75.48	22.40
Temperature	2225	°C	-0.7	26.6	12.14	6.32
Relative humidity	2225	%	23	100	69.34	14.19
Atmospheric pressure	2225	hpa	988	1036	1017.02	8.41
Wind speed	2225	m/s	0.4	6.7	2.08	0.90
Precipitation	2225	mm	0	84.6	3.42	9.39

Methods

This paper aims to quantify the air quality improvement and associated health and economic benefits caused by the Omicron pandemic prevention and control measures in ten cities in the YRD region. We estimated the impact of the Omicron pandemic lockdown on air quality using the DID model. The DID model is commonly used to evaluate the effect of policy implementation. For instance, Qiu and He (2017) used the DID model to investigate how a specific green traffic policy affects air quality in China. Chen et al. (2021) employed the DID model to examine the relationship between the driving restrictions policy and air quality in 173 Chinese cities. The lockdown can be regarded as a policy adopted by the government to control the Omicron pandemic. Unintentionally, this policy greatly improved air quality. Therefore, the DID model was applicable in this paper to measure the impact of the lockdown on air quality in these ten cities. In the DID model, a key point is the selection of an appropriate control group. In this paper, 15 cities in Jiangsu and Zhejiang provinces were selected as the control group because they were not severely affected by the Omicron pandemic. Table 3 displays the cities in the treatment and control groups. The specific DID model is as follows:

$$Y_{\mathit{it}}={\alpha }_0+{\alpha }_1\text{Loc}\ast \text{Treat}+{\alpha }_2\text{Weekend}+{\alpha }_3{X}_{\mathit{it}}+{\mu }_i+{\pi }_t+{\epsilon }_{\mathit{it}}$$ 1

where Y_it is the air pollutants concentrations in city i on day t (PM_2.5, PM₁₀, SO₂, NO₂, CO, and O₃). Loc is a dummy variable that is set to 1 when it is after March 28, 2022, or 0 otherwise. Treat is a group dummy variable that is set to 1 if it is in the treatment group and set to 0 for the control group. Weekend is set to 1 if day t is a weekend, or 0 otherwise. X_it is a series of control variables including temperature, relative humidity, wind speed, atmospheric pressure, and precipitation. µ_i is individual fixed effects. π_t indicates date fixed effects. ε_it is the stochastic error term.

Table 3.

The treatment group and the control group in this paper

Group	Shanghai	Jiangsu province	Zhejiang province
Treatment group	Shanghai	Nanjing, Suzhou, Wuxi, Changzhou, Zhenjiang, Nantong	Hangzhou, Huzhou, Jiaxing
Control group		Yancheng, Suqian Taizhou, Huai’an, Xuzhou, Yangzhou, Lianyungang,	Ningbo, Lishui, Wenzhou, Jinhua, Quzhou, Zhoushan, Taizhou, Shaoxing

During the Omicron lockdown period, many cities in the YRD region implemented the “3 + 11” policy to prevent the spread of the COVID-19 pandemic. The policy required that people from other cities must be placed in centralized isolation for 3 days before being monitored for health conditions for an additional 11 days. This policy severely restricted intercity mobility, leading to supply chain disruptions in the YRD region in April 2022. We collated the Move-In (MI) and Move-Out (MO) indices from Baidu Maps for these ten study cities from February 2, 2022, to April 30, 2022. The changes in MI and MO indices reflected the impact of the “3 + 11” policy on intercity traffic and greatly affected air quality. To quantify the impact of the “3 + 11” policy on the air quality, we incorporated the MI and MO indices into the explanatory model as independent variables:

$$ln(Y_{\mathit{it}})={\beta }_0+{\beta }_{1}MI+{\beta }_{2}MO+{\beta }_{3}Weekend+{\beta }_4{X}_{\mathit{it}}+{\mu }_i+{\pi }_t+{\epsilon }_{\mathit{it}}$$ 2

where ln (Y_it) is the logarithms on the air pollutants concentrations. Other variables have the same meaning as in Eq. (1).

This is the second time that Shanghai has taken lockdown measures since the COVID-19 pandemic started in early 2020. Compared with the first COVID-19 lockdown in early 2020, the prevention and control measures are more stringent and thorough. Therefore, we hypothesized that the air quality improvement during the 2022 lockdown is expected to be more significant than that during the 2020 lockdown. To verify our hypothesis, we used the DID model again to quantify the air quality changes in these ten cities during the 2020 lockdown. The specific DID model is as follows:

$$Y_{\mathit{it}}={\gamma }_0+{\gamma }_{1}Loc\ast Year+{\gamma }_{2}Weekend+{\gamma }_3{X}_{\mathit{it}}+{\mu }_i+{\pi }_t+{\epsilon }_{\mathit{it}}$$ 3

where Y_it is the air pollutants concentrations in city i on day t (PM_2.5, PM₁₀, SO₂, NO₂, CO, and O₃). Loc is a dummy variable that is set to 1 when it is after January 23, 2020, or 0 otherwise. Year is a dummy variable that is set to 1 if it is in the treatment group (2020) and set to 0 for the control group (2019). Weekend is set to 1 if day t is a weekend, or 0 otherwise. X_it is a series of control variables including temperature, relative humidity, wind speed, atmospheric pressure, and precipitation. µ_i is individual fixed effects. π_t indicates date fixed effects. ε_it is the stochastic error term.

In the past few years, PM_2.5 has been the primary pollutant in China. PM_2.5 would cause serious damage to the respiratory and cardiovascular systems. Based on the DID model results according to Eq. (1), this paper attempted to measure the avoided premature deaths caused by the reduction in PM_2.5 concentrations during the lockdown period. A widely used concentration–response function is employed in this paper:

$$\mathrm{\Delta }H=Y_k(1-e^{-{\beta }_k\mathrm{\Delta }C})\ast Pop$$ 4

$${\beta }_k=\frac{ln(R{R}_k)}{\mathrm{\Delta }P{M}_{2.5}}$$ 5

where $\mathrm{\Delta }H$ is the number of avoided premature deaths due to reduction in PM_2.5 concentrations. The health endpoints selected for this paper included ischemic heart disease (IHD), cerebrovascular disease (CVD), lung cancer (LC), and respiratory disease (RD). Y_k is the baseline incidence rate and the values are obtained from the China Health Statistical Yearbook 2021 compiled by the National Health Commission of the People's Republic of China. RR_k is the relative risk and the values are obtained from previous studies (shown in Table 4). For example, if the epidemiological study reported that “An increase of 10 μg/m³ of PM_2.5 was associated with 9% (95% CI: 6–12%) increases in the risk of mortality from IHD”, that means △PM_2.5 is 10 and the value of RR is 1.09 (95% CI: 1.06–1.12). β_k is the PM_2.5 concentration–response coefficient and can be calculated according to Eq. (5).$\mathrm{\Delta }C$ is the net change in PM_2.5 concentrations. Pop is the population of these ten cities and the value is obtained from the seventh National Census.

Table 4.

Health endpoints specific relative risk values (RR) and concentration–response coefficients (β_k)

Health endpoints	Gender	Yk (‰)	RRk (95% CI)	βk (95% CI)	References
IHD	Male	1.2992	1.09 (1.06, 1.12)	0.0086 (0.0058, 0.0113)	Yin et al., (2017)
	Female	1.2380	1.16 (1.09, 1.22)	0.0148 (0.0086, 0.0199)	Hayes et al., (2020)
CVD	Male	1.4987	1.064 (1.022, 1.108)	0.0062 (0.0022, 0.0103)	Cheng et al., (2022)
	Female	1.2002	1.076 (1.028, 1.127)	0.0073 (0.0028, 0.0120)	Cheng et al., (2022)
LC	Male	0.6653	1.21 (1.01, 1.44)	0.0191 (0.0010, 0.0365)	Huang et al., (2017)
	Female	0.2959	1.15 (1.12, 1.18)	0.0140 (0.0113, 0.0166)	Huang et al., (2017)
RD	Male	0.6715	1.087 (1.020, 1.159)	0.0083 (0.00198, 0.0148)	Nayebare et al., (2019)
	Female	0.4320	1.093 (1.012, 1.175)	0.0089 (0.00119, 0.0161)	Nayebare et al., (2019)

To evaluate the economic benefits associated with avoided premature deaths, the value of statistical life (VSL) is used. VSL is a measure of the willingness of residents to use the money to reduce the risk of death. VSL is affected by factors such as residents' income levels and inflation in different years. The VSL value in this article is calculated by the following equation:

$$VS{L}_{\mathit{TY}}=VS{L}_{\mathit{BY}}\times \frac{CP{I}_{\mathit{TY}}}{CP{I}_{\mathit{BY}}}\times {\left(\frac{Incom{e}_{\mathit{TY}}}{Incom{e}_{\mathit{BY}}}\right)}^e$$ 6

This paper calculated the VSL based on the previous research results. VSL_BY is the VSL value of Shanghai in 2017, which can be found in (Dai et al., 2019). CPI_BY and CPI_TY are the Consumer Price Index in 2017 and 2021, respectively. Income_BY and Income_TY are the per capita disposable annual income in 2017 and 2021, respectively. e is the income elasticity, which is assumed to be 0.8 (Nie et al., 2021).

The corresponding economic benefits can be calculated by the following equation:

$$\mathrm{\Delta }E=\mathrm{\Delta }H\ast VS{L}_{\mathit{TY}}$$ 7

where $\mathrm{\Delta }E$ is the economic benefits.$\mathrm{\Delta }H$ is the number of avoided premature deaths from Eq. (4). VSL_TY is the result of Eq. (6).

Results

Empirical results

Table 5 shows the DID model results for the sample from February 1, 2022, to April 30, 2022. We found that the implementation of the Omicron lockdown reduced all pollutant concentrations except O₃, indicating that air quality was greatly improved. The concentrations of PM_2.5, PM₁₀, SO₂, NO₂, and CO decreased by 9.87 μg/m³, 17.31 μg/m³, 0.75 μg/m³, 9.03 μg/m³, and 0.07 mg/m³, respectively. However, the concentration of O₃ increased by 9.66 μg/m³. The results also showed that weekends could significantly reduce the pollutant concentrations. Weekends reduced the concentrations of PM_2.5, PM₁₀, SO₂, and NO₂ by 1.91 μg/m³, 4.57 μg/m³, 0.44 μg/m³, and 3.40 μg/m³. This phenomenon was called the “weekend effect”, which has been studied in previous papers (Elansky et al., 2020; Qin et al., 2004). Therefore, it is highly necessary to take the “Weekend” as a control variable in Eq. (1) to avoid its impact on the research results.

Table 5.

Regression results according to Eq. (1)

Variables	PM2.5	PM10	SO2	NO2	CO	O3
Loc*Treatment	− 9.87***	− 17.31***	− 0.75***	− 9.03***	− 0.07***	9.66***
	(1.19)	(1.88)	(0.12)	(0.68)	(0.01)	(1.17)
Temperature	0.16	0.45**	0.06***	0.02	− 0.001*	2.95***
	(0.14)	(0.22)	(0.01)	(0.08)	(0.001)	(0.14)
Relative humidity	− 0.02	− 0.71***	− 0.07***	− 0.26***	− 0.001***	− 0.22***
	(0.03)	(0.05)	(0.004)	(0.02)	(0.0003)	(0.03)
Wind speed	− 5.21***	− 3.99***	− 0.56***	− 4.62***	− 0.04***	3.26***
	(0.46)	(0.73)	(0.05)	(0.26)	(0.004)	(0.45)
Atmospheric pressure	− 0.14	− 0.90***	− 0.04***	− 0.41***	− 0.007***	1.49***
	(0.12)	(0.19)	(0.01)	(0.07)	(0.001)	(0.11)
Precipitation	− 0.40***	− 0.52***	− 0.002	− 0.04*	− 0.002***	− 0.05
	(0.04)	(0.07)	(0.004)	(0.02)	(0.0003)	(0.04)
Weekend	− 1.91***	− 4.57***	− 0.44***	− 3.40***	− 0.01**	0.68
	(0.74)	(1.16)	(0.07)	(0.42)	(0.006)	(0.72)
R2	0.165	0.269	0.373	0.306	0.169	0.471
Observations	2225	2225	2225	2225	2225	2225

*P < 0.05; **P < 0.01; ***P < 0.001

Parallel trend test

The application of the DID model requires the satisfaction of the parallel trend assumption. In this paper, the DID model requires the air quality in the treatment group and the control group have the same trend before March 28, 2022. Following He et al. (2020), this paper adopted the event study method to investigate how the trends in air quality between the treatment and control groups evolve before the implementation of the lockdown. We used weekly air pollution data in this case to avoid being influenced by the high volatility of daily air pollution data. To verify the parallel trend assumption, the interaction term Week_j*Treat is added to the model. The specific model is as follows:

$$Y_{\mathit{it}}={\alpha }_0+{\alpha }_j\sum _{-5}^{1}Wee{k}_j\ast Treat+{\alpha }_{j+1}{X}_{\mathit{it}}+{\mu }_i+{\pi }_t+{\epsilon }_{\mathit{it}}$$ 8

where Week_j is a dummy variable that is set to 1 on the jth week before the implementation of lockdown. Treat is a group dummy variable that is set to 1 if it is in the treatment group and set to 0 for the control group. Other variables have the same meaning as in Eq. (1). The parameter of interest α_j represents the systematic difference between the treatment group and the control group. If α_j is not significant before March 28, 2022, the parallel trend assumption is met. Figure 3 shows the regression results according to Eq. (6). The estimated coefficients before the implementation of the lockdown are statistically insignificant because the 95% confidence intervals (CI) include 0. Therefore, the parallel trend assumption is met in this paper.

Fig. 3.

Results of the parallel trend test according to Eq. (6). The error bars represent the 95% confidence intervals. “Current” is the date of the implementation of the Omicron pandemic lockdown

Robustness test

The robustness of our results was verified by repeating the DID model (Eq. (1)) using the sample from February 1, 2021, to April 30, 2021. There was no outbreak of the COVID-19 pandemic in the YRD region during the same time period in 2021. If the air quality in these ten cities did not improve in 2021 or did not improve as significantly as it did in 2022, it indicated that the air quality improvement in 2022 was caused by the Omicron lockdown. Table 6 shows the results of the DID model based on air pollution data from February 1, 2021, to April 30, 2021. During the same time period in 2021, the concentrations of PM_2.5, SO₂, and NO₂ decreased by 3.92 μg/m³, 0.27 μg/m³, and 1.67 μg/m³, respectively. There were no statistically significant changes in other air pollutants. It indicated that air quality did not greatly improve during the same period in 2021. This demonstrated that our results regarding the impact of the Omicron lockdown on air quality are robust.

Table 6.

Regression results of the DID model for the sample for February 1, 2021, to April 30, 2021

Variables	PM2.5	PM10	SO2	NO2	CO	O3
Loc*Treatment	− 3.92***	1.07	− 0.27**	− 1.67**	− 0.03	0.28
	(1.16)	(3.47)	(0.11)	(0.73)	(0.04)	(1.08)
Temperature	− 0.34**	− 1.53***	0.06***	0.81***	0.00001	1.95***
	(0.15)	(0.44)	(0.01)	(0.09)	(0.002)	(0.14)
Relative humidity	0.03	− 0.87***	− 0.06***	− 0.02	0.002***	− 0.54***
	(0.03)	(0.08)	(0.003)	(0.02)	(0.0002)	(0.02)
Wind speed	− 2.60***	2.04**	− 0.35***	− 3.81***	− 0.03***	4.49***
	(0.42)	(1.24)	(0.04)	(0.26)	(0.004)	(0.39)
Atmospheric pressure	− 0.77***	− 3.39***	− 0.04***	0.03	− 0.006***	0.56***
	(0.10)	(0.30)	(0.01)	(0.06)	(0.001)	(0.09)
Precipitation	− 0.59***	− 0.77***	− 0.002	0.15***	− 0.003***	− 0.19***
	(0.06)	(0.20)	(0.006)	(0.04)	(0.0006)	(0.06)
Weekend	− 0.96	− 6.87***	− 0.05	− 2.24***	0.01	− 0.93
	(0.72)	(2.16)	(0.07)	(0.45)	(0.008)	(0.68)
R2	0.182	0.150	0.341	0.184	0.104	0.444
Observations	2225	2225	2225	2225	2225	2225

PM_2.5-related health benefits and associated economic benefits during the lockdown

During the past several years, PM_2.5 has been the primary pollutant in most Chinese cities, and the YRD region was no exception. Therefore, this paper attempted to estimate avoided premature deaths due to PM_2.5 concentration reductions using the concentration–response function. Then, the value of the statistical life method was used to evaluate the economic benefits associated with PM_2.5-related avoided premature death. The avoided premature deaths and associated economic benefits during the lockdown are shown in Fig. 4. During the Omicron pandemic lockdown, the total number of avoided premature deaths attributed to PM_2.5 concentrations reductions is estimated to be 35,342 in these ten cities. IHD showed the maximum health benefits compared with CVD, LC, and RD. IHD contributed to 14,034 avoided premature deaths, with 5417 (95% CI: 3602–7214) male and 8617 (95% CI: 4853–11,890) female, totally accounting for 39.7% of PM_2.5-related avoided premature deaths. CVD contributed to 8419 avoided premature deaths, with 4451 (95% CI: 1548–7548) male and 3968 (95% CI: 1488–6678) female. LC contributed to 8436 avoided premature deaths, with 6496 (95% CI: 311–13,850) male and 1940 (95% CI: 1545–2331) female. RD contributed to 4453 avoided premature deaths, with 2698 (95% CI: 624–4971) male and 1755 (95% CI: 226–3293) female. According to Eq. (6), we calculated that the VSL value in 2022 is 533,773 U.S. dollars (USD). The economic benefits associated with PM_2.5-related avoided premature deaths are shown in Fig. 3. During the Omicron lockdown period, the estimated economic benefits from IHD, CVD, LC, and RD were 7.49 billion USD, 4.49 billion USD, 4.50 billion USD, and 2.38 billion USD, respectively.

Fig. 4.

The PM_2.5-related avoided premature deaths and associated economic benefits due to IHD, CVD, LC, and RD

Discussion

In this part, we will firstly discuss the reasons for the air quality improvement during the Omicron pandemic lockdown period. Next, we will explore the effect of the commonly adopted “3 + 11” policy on air quality. Finally, we will compare the effects of the 2022 Omicron lockdown with that of the COVID-19 lockdown in early 2020.

Some reasons for the air quality improvement

The reasons for air quality improvement during the Omicron lockdown are as follows. As suggested by previous studies, traffic and industrial activities were the dominant contributors to PM_2.5 and PM₁₀ in the YRD region (Hu et al., 2014; Wang et al., 2013). NO₂ is produced by a variety of sources, mainly vehicle emissions and fuel combustion. Therefore, the reasons for the reductions in PM_2.5, PM₁₀, and NO₂ concentrations during the lockdown are most likely to be restrictions on industrial activities and vehicles. Consistent with previous studies (Liu et al., 2021c; Sicard et al., 2020), the concentration of O₃ increased by 9.66 μg/m³. At the city level, the O₃ formulation depends on the VOCs/NOx ratio (Pusede & Cohen, 2012). A high ratio represents that more O₃ would be produced. This ratio is usually low in urban areas because of the high NOx emissions. Strict control measures reduced more NOx emissions than VOCs emissions, resulting in a higher VOC/NOx ratio and increased O₃ production. Furthermore, Li et al. (2017) and Liu et al., (2013) found that the O₃ concentration would increase when PM_2.5 and PM₁₀ concentrations decreased. During the lockdown period, the higher amount of solar radiation reached the near-surface air, thereby accelerating the photochemical reactions involved in O₃ production, so more O₃ was produced.

In addition to the reasons mentioned above, regional pollution transmission also had a significant impact on air quality during the Omicron lockdown. Table 7 presents the correlation coefficient between pairs of cities for PM_2.5 and O₃. All correlations were moderate to strong and statistically significant (p < 0.01, two-tailed). However, closer inspection reveals that some correlations were stronger than others. For example, PM_2.5 in Shanghai was highly correlated with PM_2.5 in most cities except ZJ, NJ, and HZ. The correlation weakens as the distance between cities grows. Analogous differences exist in the correlation of O₃. This indicates that there exists strong pollution transmission between cities. Figures 5 and 6 show the concentrations of PM_2.5 and O₃ during the Omicron lockdown, respectively. Shanghai, the epicenter of this wave of the Omicron pandemic, implemented complete lockdown measures to control the pandemic. Therefore, Shanghai naturally had the lowest PM_2.5 concentration among these ten cities. Meanwhile, the closer a city is to Shanghai, the lower the PM_2.5 concentration. As previously stated, the O₃ concentration tended to show the opposite trend to the PM_2.5 concentration. The O₃ concentration in Shanghai was the highest among these ten cities. In contrast to PM_2.5, the closer a city is to Shanghai, the higher the O₃ concentration. The phenomenon that pollutant concentrations in cities adjacent to Shanghai were close to those in Shanghai during the Omicron pandemic lockdown was primarily caused by regional pollution transmission.

Table 7.

Values of Pearson correlation coefficient between pairs of cities for PM_2.5 (cells below the diagonal) and for O₃ (cells above the diagonal)

	SH	SZ	WX	CZ	ZJ	NJ	NT	HZ	JX	HUZ
SH	1	0.92	0.86	0.79	0.72	0.72	0.85	0.72	0.88	0.76
SZ	0.92	1	0.95	0.90	0.81	0.82	0.82	0.81	0.94	0.90
WX	0.85	0.98	1	0.97	0.88	0.88	0.80	0.78	0.91	0.88
CZ	0.85	0.95	0.97	1	0.93	0.91	0.76	0.74	0.88	0.85
ZJ	0.67	0.79	0.83	0.90	1	0.93	0.69	0.66	0.79	0.76
NJ	0.76	0.85	0.85	0.88	0.88	1	0.63	0.67	0.82	0.79
NT	0.89	0.89	0.90	0.88	0.76	0.72	1	0.71	0.77	0.70
HZ	0.79	0.87	0.84	0.82	0.65	0.79	0.74	1	0.82	0.89
JX	0.86	0.94	0.90	0.86	0.65	0.79	0.78	0.92	1	0.89
HUZ	0.80	0.92	0.91	0.89	0.74	0.82	0.75	0.90	0.93	1

Fig. 5.

The PM_2.5 concentration during the Omicron lockdown

Fig. 6.

The O₃ concentration during the Omicron lockdown

Effect of the “3 + 11” policy on air quality

During the lockdown period, these ten cities implemented the “3 + 11” policy to prevent the spread of the Omicron pandemic. This policy greatly restricted mobility between cities, which improved air quality. We investigated the changes in MI and MO after the outbreak of the Omicron pandemic in these ten cities (Fig. 7). The first confirmed case was found on March 1, 2022. Since then, the values of MI and MO started to drop as shown in Fig. 4. On March 28, 2022, Shanghai decided to impose a lockdown in two phases for city-wide nucleic acid testing. The values of MI and MO dropped to a nadir at the end of March and remained there until the end of April. Table 8 shows the regression results according to Eq. (2). MI had a negative impact on PM_2.5, PM₁₀, and CO concentrations. A one-unit increase in MI in these 10 cities decreased the concentrations of PM_2.5, PM₁₀, and CO by 1–5%. However, there was no significant impact on the concentrations of SO₂, NO₂, and O₃. MO had a positive impact on all air pollutant concentrations except O₃. A one-unit increase in MO in these 10 cities increased the air pollutants concentration except O₃ by 4–20% and decreased the O₃ concentrations by 3.6%. During this wave of Omicron pandemic lockdown period, almost all intercity traffic was disrupted except for some food trucks. All cars can only run locally, leading to higher intra-city travel intensity. This is the reason for the MO had a positive impact on air pollutant concentrations.

Fig. 7.

Changes in MI and MO in the study cities from February 1, 2022, to April 30, 2022

Table 8.

Regression results according to Eq. (2)

Variables	ln (PM2.5)	ln (PM10)	ln (SO2)	ln (NO2)	ln (CO)	ln (O3)
MI	− 0.040***	− 0.048***	0.0003	0.011	− 0.007*	0.005
	(0.011)	(0.010)	(0.005)	(0.009)	(0.005)	(0.005)
MO	0.171***	0.195***	0.018**	0.128***	0.040***	− 0.036***
	(0.020)	(0.018)	(0.010)	(0.016)	(0.009)	(0.009)
Temperature	0.002	0.009	0.012***	− 0.003	− 0.005*	0.048***
	(0.006)	(0.006)	(0.003)	(0.005)	(0.003)	(0.003)
Relative humidity	− 0.003*	− 0.015***	− 0.010***	− 0.012***	− 0.0001	− 0.002***
	(0.002)	(0.001)	(0.001)	(0.001)	(0.0007)	(0.0007)
Wind speed	− 0.192***	− 0.086***	− 0.081***	− 0.175***	− 0.078***	0.045***
	(0.019)	(0.018)	(0.010)	(0.016)	(0.009)	(0.010)
Atmospheric pressure	− 0.011**	− 0.017***	− 0.003	− 0.025***	− 0.014***	0.026***
	(0.005)	(0.005)	(0.003)	(0.004)	(0.002)	(0.002)
Precipitation	− 0.014***	− 0.012***	− 0.0005	− 0.0007	− 0.003***	− 0.003***
	(0.002)	(0.002)	(0.0008)	(0.0013)	(0.0007)	(0.0008)
Weekend	− 0.033	− 0.057**	− 0.060***	− 0.129***	− 0.030**	0.014
	(0.032)	(0.029)	(0.016)	(0.025)	(0.015)	(0.015)
R2	0.374	0.509	0.489	0.428	0.206	0.516
Observations	890	890	890	890	890	890

*P < 0.05; **P < 0.01; ***P < 0.001

Compared with the effect of the COVID-19 pandemic lockdown in early 2020

Because the Omicron strain is more transmissible than all previous variants, the government has implemented more stringent control measures than ever before. Therefore, we hypothesized that the air quality improvement during the 2022 lockdown is expected to be more significant than that during the 2020 lockdown. However, the results disappointed us. Table 9 shows the regression results according to Eq. (3). During the 2020 lockdown period, the concentrations of PM_2.5, PM₁₀, SO₂, NO₂, and CO decreased by 19.73 μg/m³, 29.81 μg/m³, 2.44 μg/m³, 30.50 μg/m³, and 0.24 mg/m³, respectively. The O₃ concentration increased by 30.94 μg/m³. The impact of the 2020 lockdown measures on air quality in these ten cities was greater than the impact of the 2022 lockdown measures.

Table 9.

Regression results according to Eq. (3) for the sample from December 1, 2019, to February 29, 2020, in these 10 cities

Variables	PM2.5	PM10	SO2	NO2	CO	O3
Loc*Year	− 19.73***	− 29.81***	− 2.44***	− 30.50***	− 0.24***	30.94***
	(1.66)	2.01	(0.39)	(0.96)	(0.02)	(0.98)
Temperature	− 1.24***	0.46	0.46***	1.21***	− 0.01***	− 0.05
	(0.25)	(0.31)	(0.06)	(0.15)	(0.002)	(0.15)
Relative humidity	− 0.43***	− 0.95***	− 0.13***	− 0.39***	− 0.002***	− 0.42***
	(0.06)	(0.07)	(0.01)	(0.03)	(0.0005)	(0.04)
Wind speed	− 8.19***	− 9.72***	− 0.78***	− 7.29***	− 0.05***	3.34***
	(0.64)	(0.78)	(0.15)	(0.37)	(0.006)	(0.38)
Atmospheric pressure	− 1.79***	− 1.66***	0.07*	− 0.44***	− 0.02***	0.19*
	(0.17)	(0.20)	(0.04)	(0.10)	(0.002)	(0.10)
Precipitation	− 1.32***	− 1.55***	− 0.01	− 0.18***	− 0.01***	0.26***
	(0.12)	(0.14)	(0.03)	(0.07)	(0.001)	(0.07)
Weekend	− 0.37	− 1.96	− 0.12	− 1.30*	0.0005	− 0.75
	(1.24)	(1.51)	(0.29)	(0.72)	(0.011)	(0.73)
R2	0.315	0.387	0.136	0.525	0.282	0.471
Observations	1810	1810	1810	1810	1810	1810

*P < 0.05; **P < 0.01; ***P < 0.001

After careful consideration, this paper believed that it was mainly due to the fact that the COVID-19 pandemic in 2020 coincided with the Lunar New Year. Many migrant residents returned to their hometowns to celebrate the Lunar New Year with their families, which is particularly common in the YRD region. Normally, most factories would suspend their operation during the Lunar New Year holiday. Meanwhile, the Chinese government decided to extend the holiday to February 2, 2020. This resulted in people being required to stay at home and some factories suspended their operations for an extra week. This further improved the air quality. Therefore, the air quality improvement is more significant during the 2020 lockdown due to the dual influence of the lockdown measures and the Lunar New Year.

Conclusions and implications

To control the Omicron pandemic, Shanghai and nine other cities severely affected by Shanghai have taken strict lockdown measures, such as restrictions on traffic, production, mobility, and social activities. First, this paper quantified the impact of lockdown measures on air quality in these ten cities by building a DID model. The concentrations of PM_2.5, PM₁₀, SO₂, NO₂, and CO decreased by 9.87 μg/m³, 17.31 μg/m³, 0.75 μg/m³, 9.03 μg/m³, and 0.07 mg/m³, respectively. However, the concentration of O₃ increased by 9.66 μg/m³. Based on the results of the DID model, we estimated the associated PM_2.5-related health and economic benefits using the concentration–response function and the VSL method. The total number of avoided premature deaths attributed to PM_2.5 concentration reductions is estimated to be 35,342 due to IHD, CVD, LC, and RD. The associated economic benefits were 18.86 billion US dollars. Second, we measured the impact of the “3 + 11” policy on air quality. The results showed MI had a negative impact on PM_2.5, PM₁₀, and CO concentrations. A one-unit increase in MI in these ten cities decreased the concentrations of PM_2.5, PM₁₀, and CO by 1–5%. MO had a positive impact on all air pollutant concentrations except O₃. A one-unit increase in MO increased the air pollutant concentrations except O₃ by 4–20% and decreased the O₃ concentrations by 3.6%. Finally, we used the DID model to assess how the 2020 lockdown measures affected air quality in these ten cities. Compared with the impact of the 2022 lockdown on air quality, the impact of the 2020 lockdown measures on air quality was greater. This paper believed that it is mainly because the COVID-19 pandemic in 2020 coincided with the Lunar New Year.

There is no doubt that the Omicron pandemic is a tragedy. This article is by no means stressing the Omicron pandemic is beneficial to air quality and human health. Instead, we can learn some lessons to guide the design of future environmental policies. Some findings of this paper are important for future air pollution control. First, the Omicron pandemic lockdown achieved substantial air quality improvement and health benefits. These results re-emphasized the importance of emission reductions in mitigating the adverse health effects related to air pollution. However, the lockdown is not a sustainable way to reduce anthropogenic emissions. In the future, a flexible and effective air pollution control strategy should be explored rather than a one-size-fits-all approach. Moreover, inconsistent with the decline trends of the other air pollutants, the concentration of O₃ increased by 9.66 μg/m³. The imbalanced reductions of NOx and VOCs resulted in a higher VOCs/NOx ratio, which caused the increase in O₃ concentration (Akritidis et al., 2021; Fu et al., 2020). To control O₃ pollution, the reduction ratio of the VOCs to NOx should not be lower than 0.73 (Zeng et al., 2018). Future O₃ pollution control strategies should consider the synergistic reduction of NOx and VOCs. Second, the reduction in MI would decrease the intra-city travel intensity and improve air quality. The reduction in MO would increase the intra-city travel intensity and make the air quality worse. For economic development, intercity traffic cannot be restricted for a long time. Therefore, efforts to control air pollution caused by intercity traffic should not rely solely on traffic restrictions, but should mainly focus on upgrading transport vehicles. In the future, the elimination of high-polluting vehicles that do not meet national emission standards should be accelerated. Other measures, such as raising national vehicle emission standards and strictly prohibiting vehicles that do not meet national emission standards from driving on the road, should also be implemented.

Funding

The author declares that no funds, grants, or other support were received during the preparation of this manuscript.

Data availability

The datasets generated during the current study are available from the corresponding author on reasonable request.

Declarations

Conflict of interest

The authors declare that they have no conflict of interest.

Ethics approval and consent to participate

Not applicable.

Consent to publication

Not applicable.