Next Article in Journal
Comparison of Backscatter Coefficient at 1064 nm from CALIPSO and Ground–Based Ceilometers over Coastal and Non–Coastal Regions
Previous Article in Journal
Towards an Improved Inventory of N2O Emissions Using Land Cover Maps Derived from Optical Remote Sensing Images
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Atmosphere
Volume 11
Issue 11
10.3390/atmos11111189
atmosphere-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Impact of the COVID-19 Pandemic Lockdown on Air Pollution in 20 Major Cities around the World

by
Franck Fu
*,
Kathleen L. Purvis-Roberts
and
Branwen Williams
W.M. Keck Science Department, Claremont McKenna College, Pitzer College and Scripps College, Claremont, CA 91711, USA
*
Author to whom correspondence should be addressed.
Atmosphere 2020, 11(11), 1189; https://doi.org/10.3390/atmos11111189
Submission received: 2 October 2020 / Revised: 29 October 2020 / Accepted: 31 October 2020 / Published: 3 November 2020
(This article belongs to the Section Air Quality)
Browse Figures
Review Reports Versions Notes

Abstract

:
In order to fight against the spread of COVID-19, the most hard-hit countries in the spring of 2020 implemented different lockdown strategies. To assess the impact of the COVID-19 pandemic lockdown on air quality worldwide, Air Quality Index (AQI) data was used to estimate the change in air quality in 20 major cities on six continents. Our results show significant declines of AQI in NO2, SO2, CO, PM2.5 and PM10 in most cities, mainly due to the reduction of transportation, industry and commercial activities during lockdown. This work shows the reduction of primary pollutants, especially NO2, is mainly due to lockdown policies. However, preexisting local environmental policy regulations also contributed to declining NO2, SO2 and PM2.5 emissions, especially in Asian countries. In addition, higher rainfall during the lockdown period could cause decline of PM2.5, especially in Johannesburg. By contrast, the changes of AQI in ground-level O3 were not significant in most of cities, as meteorological variability and ratio of VOC/NOx are key factors in ground-level O3 formation.

1. Introduction

The majority of the world’s major cities suffer from serious air pollution issues, leading to more than two million deaths globally through damage to the lungs and the respiratory system [1]. The International Agency for Research on Cancer (IARC) has classified air pollution as Carcinogenic to Humans (Group 1), as studies show exposure to outdoor air pollution causes lung cancer [2]. The common pollutants of concern are nitrogen dioxide (NO2), sulfur dioxide (SO2), carbon monoxide (CO), ground-level ozone (O3), and particulate matter (PM2.5 and PM10). NO2, SO2, PM, and O3 are all associated with the development and/or aggravation of respiratory diseases that reduce lung function, particularly in vulnerable populations with pulmonary disease or asthma [3,4,5]. CO can cause subtle cardiovascular and neurobehavioral effects even at low concentrations [6]. Because of these severe health issues associated with air pollution, these six criteria pollutants are routinely measured in many countries. If the concentration of these air pollutants is high enough to impact human heath, this can lead to governmental action plans and policies to control pollutant discharge and improve the air quality.
Air pollution is released from a wide variety of natural and anthropogenic activities [7], most pollutants are both primary and secondary. NO2, SO2 and CO are mainly primary and originated from anthropogenic activities globally [8,9,10]. Among anthropogenic sources, transportation and combustion in power plants are the primary and secondary sources of NO2 [11,12,13] and CO [11,14]. However, the main source of SO2 is generally combustion of coal in power plants and the manufacturing industries [11,15]. Ground-level O3 is a secondary pollutant formed in the air by a series of photochemical reactions, the key factors are sunlight, NOx and a variety of volatile organic compounds (VOCs). Primary particulate matter is directly released into the atmosphere by natural and anthropogenic activities, while secondary particles are formed in the atmosphere from other precursor pollutants, such as SO2, NO2, NH3 and VOCs. In urban areas, anthropogenic sources of PM10 (aerodynamic diameter ≤ 10 µm) and PM2.5 (aerodynamic diameter ≤ 2.5 µm) dominate, and the major sources are residential combustion, large-scale combustion (i.e., power plants), industrial processes, agriculture and transportation (road and non-road) [16].
In December 2019, cases of pneumonia of “unknown etiology” were first identified in Wuhan, China [17]. On 11 February 2020, the World Health Organization (WHO) announced an official name coronavirus disease 2019 (COVID-19) for this epidemic disease. After the first outbreak in Wuhan, several community outbreaks occurred in February in countries outside of China, such as South Korea, Italy, Germany and Spain. In early 2020, the epicenter moved from Asia to Europe and then to the Americas by March. In order to mitigate the infection rate of COVID-19, many countries applied different lockdown strategies. Lockdown measures included partial or full closure of international borders, schools, non-essential business and citizen mobility restriction [18]. The reduction of transportation, commercial and industrial activities due to these lockdown strategies has the potential to reduce the emissions of primary pollutants and change the formation rate of secondary pollutants, due to the change of precursor emissions.
Recent studies suggested that lockdown measures contributed to improvements in air quality, especially in urban areas, where more anthropogenic activities are present (see the summary of recent studies on COVID-19 and air quality impacts [19]). Most studies assessed the impacts of COVID-19 on air quality within a single country. Few studies expanded world-wide, but in one study, remarkable declines of AQI were observed in NO2 (−44% to −13%), ozone (−20% to −2%) and PM2.5 (−28% to 10%) during the first two weeks of the lockdown in 27 countries [20]. Another focused on the global change of NO2, CO and AOD (Aerosol Optical Depth), they observed reduction of NO2 (0.00002 mol m−2), CO (<0.03 mol m−2) and AOD (~0.1–0.2) in the major hotspots of COVID-19 outbreak between February and March relative to 2019 [21]. Thus, in order to provide a more comprehensive analysis of the impact of lockdowns on all 6 critical air pollutants during the entire lockdown period and to assess the impact of different lockdown strategies on air pollution, AQI in 20 major cities worldwide was examined. Major cities were selected because they often have air pollution issues and more available air quality monitoring stations.

2. Materials and Methods

2.1. City Selection

Twenty major cities were selected (Figure 1) representing all continents except Antarctica, including seventeen cities in the most hard-hit countries: Wuhan, Beijing (China), Delhi (India), Tehran (Iran), Istanbul (Turkey) in Asia; Rome (Italy), Madrid (Spain), Paris (France), London (UK), Berlin (Germany) and Moscow (Russia) in Europe; Johannesburg (South Africa) in Africa and Los Angeles, New York City (USA), Mexico city (Mexico), Sao Paulo (Brazil) and Lima (Peru) in North and South America.

2.2. Air Quality Data

After the outbreak of COVID-19, the World’s Air Pollution, Real-time Air Quality Index (WAQI project) started to provide a new dedicated, dataset specific for COVID-19 related research, covering approximately 380 major cities throughout the world (aqicn.org/data-platform/COVID19/). The World Air Quality project is a non-profit started in 2007 to provide air quality information for more than 100 countries, covering more than 30,000 stations (local governmental/professional monitoring network) in 1000 major cities. The data for each major city is based on the median of several stations. The dataset provides a statistical summary for each of the air pollutant species, all air pollutants are converted to an Air Quality Index (AQI) with the U.S. Environmental Protection Agency (EPA) standard calculation. The daily median AQI was used in this study for each of 20 major cities.
The AQI is a dimensionless index that quantitatively describes air quality conditions based on standards of each pollutant, which provides a comprehensive evaluation on the combined effects of the six criteria pollutants (NO2, SO2, CO, ground-level O3, PM2.5 and PM10). The AQI is classified into six grades calculated from the concentrations of various pollutants, i.e., AQI: 0–50 (Good), 51–100 (Moderate), 101–150 (Unhealthy for Sensitive Groups), 151–200 (Unhealthy), 201–300 (Very Unhealthy), and >300 (Hazardous).

2.3. Lockdown Data

In order to define the lockdown period for each of the cities of interest, sources including media and official government websites and research papers were used. It should be noted that it’s counted from the beginning of lockdown or strict social distancing policy to the date the lockdown started to ease (Table 1), as the reopening is a complex process and could include several steps. In order to compare AQI data during the lockdown period to historical data, the same period of data for 2017, 2018 and 2019 were taken into account. The average AQI of each pollutant was compared to the previous year (2019) and to the 3-year average of the same period to calculate the decrease or increase of each pollutant.

2.4. Climate Data

In order to assess the impact of weather condition on the air quality, the climate data from the Global Historical Climatology Network was analyzed. This daily (GHCN-Daily) dataset includes daily land surface observations from around the world. The GHCN-Daily was developed to meet the needs of climate analysis and monitoring studies that require data on a sub-monthly time resolution. The dataset includes observations from the World Meteorological Organization, Cooperative, and Community Collaborative Rain, Hail and Snow (CoCoRaHS) networks.
The daily rainfall and daily average temperature for the lockdown period of 2020 and the same period of 2019 were from the GHCN-Daily dataset. The average temperature was not available for Los Angeles (LA), New York City (NYC) and Sydney, so the maximum temperature was used, as the maximum temperature is important for ozone formation. For Moscow, Mexico City and Berlin, the climate data was not complete; for instance, Moscow had no data for April and May of 2020 and Mexico City was missing 15 days of data in May. The data from the weather underground website was used instead of GHCN data for those three locations. From both sources, there is no data for rainfall for Moscow, Mexico City and Berlin. The information about the location (name or/and number of the station) can be found in Supplementary Table S1.

3. Results

3.1. Primary Pollutants: NO2, SO2 and CO

NO2 AQI decreased for all cities during the lockdown period, relative to 2019 and the past 3 years average for the same period (Table 2, Figure 2). NO2 AQI declined in all cities with the highest decrease (−60%) in Delhi and the lowest (−11.1%) in Sydney in comparison to 2019 and decreased the most (−63.3%) in Wuhan and the least in Sydney (−15.5%) relative to an average of the past three years.
According to ANOVA and Tukey HSD tests with 95% confidence, the decreases were statistically significant for all cities, meaning the 2020 AQI was significantly different to each of past 3 years, expect for Berlin, Tokyo, Istanbul and Sydney. NO2 continuously decreased from 2017 to 2020 in all Asian cities: Beijing, Wuhan, Tokyo, Istanbul and in South American cities: Lima and Mexico City (Figure 2). This trend of decreasing NO2 AQI in these cities (that existed prior to 2019) contributed to the small difference between the 2017–2019 average and 2019 value. The SO2 AQI decreased significantly in 6 cities, compared to each of past 3 years (Table 2, Figure S1). The highest decrease (−54.8%) occurred in Madrid relative to 2019, due to a much higher SO2 level in 2019, for which the cause is unknown. However, the AQI in 2020 was not statistically different in comparison to those of 2017 and 2018, suggesting that the SO2 level was stable in Madrid. In addition, significant increases of SO2 AQI in Istanbul (+29.3%) and Berlin (+29.5%) were also observed (Table 2). For other cities, the changes were not statistically significant, or there was no available data.
Limited CO AQI data was available in the World Air Quality Index project data set. For cities with CO data, the CO AQI decreased significantly in eight of 15 cities (Table 2, Figure S2), with the maximum decrease in Lima (−60%) relative to 2019. In six other cities, the AQI changes were not statistically significant, compared to each of past 3 years. In Beijing, where a much higher increase of CO AQI relative to 2019 was observed (+92.4), by excluding the abnormally and unexplained low 2019 concentrations, the level of CO decreased by 26.2% relative to 2017 and 2018.

3.2. Secondary Pollutants: Ground-Level O3

Contrary to the trends of decreasing primary pollutants, ground-level O3 AQI increased (+2.9–+54.2%) in 12 cities and decreased (from −0.1 to −43.6%) in the other 7 (no data in Tehran) during the lockdown period for each city relative to 2019 (Table 2 and Figure 3). Comparing each of the past 3 years using an ANOVA with Tukey HSD test, the O3 significantly increased in Wuhan, Paris, London and Sao Paulo, and significantly decreased in Istanbul. For other cities, the changes were not statistically significant. Wuhan experienced the maximum increase (+54.2%) and Istanbul experienced the maximum decrease (−43.6%). Lima also experienced a large decrease (−42.9%) relative to 2019, because of the dramatically higher ozone concentration in 2019 than other years (Figure 3). However, the AQI level in 2020 is statistically insignificant in comparison to 2018′s level.

3.3. Particulate Matter: PM2.5 and PM10

PM2.5 AQI decreased in all cities (Table 2 and Figure 4), except in Rome with an increase of +8.3% during the lockdown period relative to 2019 and the 2017–2019 reference, with the maximum decrease in Johannesburg (−31.3%) relative to 2019. The decreases are statistically significant in 12 cities relative to 2019, and in 9 cities compared to each of the past 3 years. For other cities, the changes are not statistically significant (see Table 2 and Figure 4). PM10 also decreased in all cities, except in Seoul and Los Angeles (insignificant increases), with the maximum decrease in Delhi (−45.9%) relative to 2019. The decreases are statistically significant in 9 of 19 cities relative to 2019, and in 4 of 17 cities compared to each of the past 3 years (see Table 2, Figure 5).

3.4. Temperature and Rainfall

Temperatures in 2020 are significantly higher than that observed in 2019 for Moscow and Paris, and significantly lower in 2020 for Delhi, Tehran, NYC and Sao Paulo, with the largest temperature decrease of 2.5 °C in Sao Paulo (Figure 6). For other cities, the temperature changes in 2020 relative to 2019 were not statistically significant. Rainfall was more than two times higher in 2020 than 2019 in Beijing, Istanbul, Johannesburg, LA, Rome, Tokyo and Wuhan (Figure 6).

4. Discussion

4.1. Changes in AQI of Main Pollutants

A significant impact of the COVID-19 lockdown on air quality was found in 20 major cities, including significant decreases in AQI levels for NO2, SO2, CO and PM, and increases of ground-level O3 AQI in most cities (Table 2). The decrease of primary pollutants NO2, SO2, CO during the lockdown period relative to 2019 and to the average of 2017–19 for the same period are due to the reduction of emissions from anthropogenic activities. The transportation reduction is responsible for the declines of NO2 and CO, due to the restriction of human mobility (e.g., automobile use decreased in all cities during lockdown [22]), and the reduced electricity consumption is responsible for the decrease of SO2 due to the restrictions of industrial and commercial activities [23] (see Section 4.2 for a discussion).
For the secondary pollutant ground-level O3, the AQI level depends on the O3 formation rate through complex photochemical reactions, in which the three determinizing factors are sunlight, NOx and VOCs. However, the chemistry of O3 formation is highly nonlinear, and the effects of precursor concentrations on O3 production rate can be characterized as either NOx-sensitive or VOC-sensitive [24,25,26]. Under a VOC-sensitive regime (low ratio of VOC/NOx), an increase in NOx concentration causes a decrease of ozone with low concentrations of VOCs. On the other hand, under the NOx-sensitive regime (higher ratio of VOC/NOx), the reduction of NOx emissions will lead to an increase in ozone concentrations. As VOC data was not available, it’s difficult to define the regime of NOx-VOC-O3 sensitivity. However, O3 levels increased during the lockdown in most cities (Figure 3). Even in the case of Istanbul and Lima, by excluding the data from 2019, the level from 2020 was also higher than 2017 and/or 2018. Thus, ozone likely formed under a NOX-sensitive regime in most of cities. Weather conditions are also another important factor, especially solar radiation or temperature.
The decline in PM reflects changes in both primary and/or secondary particles emissions and reactions. The primary PM is from natural and anthropogenic processes including road traffic within urban areas [27,28]. The two main species for secondary PM formation are sulfate and nitrate, formed in air from precursor pollutants: SO2 and NO2. Thus, the decline of emissions of primary pollutants SO2 and NO2 recorded here also indirectly reduced the formation of secondary PM.

4.2. Impact of Lockdown Strategy on Air Quality

Lockdown policies reduced transportation and electricity demand, reflecting restricted human mobility, industry and commercial activities [22,23]. Since the primary and secondary sources of NO2 are transportation and combustion in power plants, this led to a reduction in NO2 [11,12,13]. The percentages of decrease in NO2 AQI in 20 major cities were compared to assess the impact of lockdown policy on air pollution (Figure 7), as NO2 showed the most significant changes during lockdown. The car driving data [22] was also used to assess the restriction of human mobility, as transportation is the most important source of NO2.
In Wuhan, the first epicenter of COVID-19, experienced the strictest lockdown policy (Table 2). All public transport was suspended. The residents of Wuhan were not allowed to leave the city without permission from authorities, and only one person from each household was allowed to leave their house or apartment for two hours every second day for essentials [29]. A very strict lockdown policy was also applied in Delhi: public transportation was also suspended during the lockdown, and only those who worked for “essential services” could move freely. Due to the strictest lockdown policies, NO2 reduced the most in Delhi and Wuhan (60.0%, 58.3%, respectively) (Figure 7). Driving in automobiles dropped about 88% in Delhi during the lockdown [22], also the highest reduction in mobility among all cities, consistent with the strictest lockdown policy and the highest reduction of NO2 among all cities.
As the second epicenter of COVID-19, strict lockdown policies were implemented in hard-hit European countries, which aimed to reduce human mobility, especially in Italy, Spain, France and the UK [30,31,32,33]. For instance, people could not leave home without a permit in France, citizen mobility reduced by 68% and 79% in the Paris region, respectively, within and leaving the region [32]. In the UK, people could only go outside to buy food, to exercise once a day, or go to work if they absolutely could not work from home [33]. The reduction of NO2 in Madrid (51.6%), Rome (36.4%), London (46.4%) and Paris (37.8%) (Figure 7), agrees with the medium-strict lockdown policies implemented in those countries. Driving of automobiles declined 65–85% in these 4 European cities [22], consistent with the strictness of lockdown policies and the reduction of NO2.
The Peruvian government implemented the longest lockdown (108 days, Table 2) in the world. The lockdown policy restricted citizen mobility based on gender. Only men could go out on Monday, Wednesday and Friday and only women on Tuesday, Thursday and Saturday. Otherwise, no Peruvian citizens were allowed to leave their homes, and the police and the army were deployed to enforce the lockdown [34]. Lima experienced fourth highest reduction (51.8%) of NO2 relative to 2019 for the same period.
The U.S. federal government did not issue a national lockdown, but a national emergency instead. Most states issued their own stay-at-home orders. For example, stay at home orders started on March 17th in California and 22nd in New York. About 60% of the decrease in automobile use [22] was observed during stay-at-home periods in NYC and LA, which is lower than most of European cities and Delhi, meaning less strict policies applied in the U.S. than those in hard-hit European cities. The stay-at-home orders were not legally enforced. This voluntary social distancing led to a 27.5% (LA) and 24.2% (NYC) decrease of NO2 AQI. This decrease was less than most of the European cities, consistent with the less strict lockdown policies in the U.S. and the lower decrease of human mobility.
The smallest declines in automobile use were observed in Sydney (50%), Berlin (50%), Seoul (45%) and Tokyo (30%) among all cities [22]. The least strict lockdown policies were implemented in Seoul and Tokyo. The Japanese government declared a “state of emergency” for seven prefectures, including Tokyo, Osaka, and Fukuoka, and implemented a Japanese version of a social-distancing policy. The government did not impose severe lockdown regulations but encouraged self-restraint on the part of the public and businesses [35]. The Korean government handled the COVID-19 crisis by applying intensive testing and contact tracing rather than enhanced social distancing, and since these measures worked to control the COVID-19 spread, a lockdown was never mandated [36]. Germany implemented a nation-wide social distancing and contact restriction on 22 March, in contrast to most other European countries, the stringency of measures differs substantially between states [37]. For example, in Berlin, people were only allowed to go out alone or with a person from same household. Australia also implemented similar lockdown policies in comparison to most European countries. All Australians were strongly advised to leave their homes only for limited essential activities and public gatherings were limited to two people. Berlin (17.6%) and Sydney (11.1%) experienced the least decreases of NO2 AQI.
Among all countries, the Turkish government issued a particular policy to fight against the COVID-19 outbreak: weekend curfew for residents under the age of 20, and aged 65 and above, without full or partial lockdown [38]. However, automobile use dropped ~60%, higher than Berlin and Sydney, where statewide lockdown with human mobility restrictions were implemented. As a consequence, the decrease of NO2 AQI was also one of the lowest in Istanbul, with a 19.9% decrease, also higher than those in Berlin and Sydney.
In conclusion, the decline of NO2 related to the drop of automobile usage in most cases, as transportation is the main source of NO2. The decrease of car use depended on the strictness of lockdown policy, especially the restrictions on human mobility and the forces that were used to control it. However, other social factors also could impact the effectiveness of lockdown policy, such as the voluntariness of citizen on social distancing [39] and the trust in government [40].

4.3. Impact of Meteorological Conditions on Air Quality

Year-to-year variations in meteorological conditions impacts air pollution, particularly the formation of secondary pollutants. Higher temperature/solar radiation favors the formation of ozone, and temperature correlates with ground-level ozone [5,41]. Higher temperatures can promote the formation of ground-level O3. The significant increase of O3 in Paris and insignificant increase in Moscow in 2020 (Figure 3) could be partially due to the higher temperature (Figure 7). The significantly lower temperature in NYC during lockdown relative to 2019 could be responsible for the slight decrease of O3. However, in Delhi and Sao Paulo, despite the significantly lower temperature during the lockdown, the O3 AQI increased significantly in Sao Paulo and insignificantly in Delhi, meaning the increases of O3 could be higher, if the temperature was the same as in past years.
Enhanced rainfall can wash air pollutants out of the atmosphere, especially particulate matter and water-soluble pollutants, such as NO2 and SO2. The decrease in primary pollutants and PM in Beijing, Istanbul, Johannesburg, LA, Rome, Tokyo and Wuhan, where rainfall was more than twice as high in 2020 than 2019, could be partially caused by this higher rainfall during the lockdown. This is especially true in Johannesburg where 4.9 times more rain fell in the 2020 lockdown period than 2019 and could explain the largest decrease of PM2.5 (31.3%).

4.4. Impact of Environmental Policy on Air Quality

Prior to COVID-19, many countries imposed action plans or policies limiting emissions to the atmosphere to mitigate the impact of air pollution on public health. For instance, the Action Plan on Prevention and Control of Air Pollution in China 2013, the National Clean Air Program (NCAP) in India, the Clean Air Act in the U.S., and the National Air Pollution Control Program in the E.U. all aim to lower air pollution in their country or region. In addition to the lockdown policies and meteorological conditions, these action plans and policies could also play a role in air pollution change. This role can be assessed by observing the inter-annual variation of each pollutant prior to the occurrence of COVID.
NO2 decreased significantly from 2017 to 2019 in Wuhan, Istanbul, Lima and Mexico City (Figure 2) and SO2 declined significantly in Beijing, Wuhan and Mexico City (see Supplementary Figure S1). Significant decreases in PM2.5 occurred in Delhi, Wuhan, Tokyo and Lima (Figure 4). These inter-annual declining trends could show the continuous reduction of pollutant emissions due to numerous local environmental policies implemented in these countries, especially in China and Mexico.
China and India, major developing countries in the world with the largest populations, have considerable air pollution issues especially in major cities [42,43], due to the high growth in urban population and the increased demand for energy and transportation. After the implementation of different policies to control air pollution, China experienced significant decreases (21–59%) of PM2.5, SO2 and NOx, since the Action Plan on Prevention and Control of Air Pollution was implemented in 2013 [44,45]. For the decrease of air pollutants during the lockdown in 2020, pre-existing decreasing trends of air pollutants should be taken into account. The pre-existing decreasing trends of NO2, SO2 and PM2.5 were reported previously in the literature and also observed in this study (Figure 2, Figure 4 and Figure S1), especially in Wuhan. In contrast, India’s SO2 and NO2 levels increased by more than 100% and 50% from 2005 to 2015 respectively, due to the high growth of coal power plants and smelters [46]. The significant increases of NO2 (Figure 2) and SO2 (Figure S1) in 2019 relative to 2017 and 2018 confirm this increasing trend in Delhi. In January 2019, India launched a National Clean Air Program aimed to reduce particulate matter pollution by 20–30% by 2024 relative to 2017 levels [47]. It is too early to observe the results of this program. The increasing pollution trends of past years could cause an underestimation in the decrease of air pollution during the lockdown period.
In other major developing countries: significant pre-existing decreasing trends of NO2 and SO2 were also found in Mexico City during 2017–2019. Since the 1990s, the Mexican government developed and implemented successive air pollution programs that combined regulatory actions with technological changes that resulted in significant improvement to air quality. PM2.5 (60%) NO2 (40%) and SO2 (90%) decreased dramatically since 1990 to 2018 (Molina et al., 2019). This decreasing trend could partially be responsible for the lower concentrations of NO2 and SO2 during 2020 lockdown in Mexico City. In Turkey, according to the regulations, every Provincial Directorate of Environment and Urban Planning has to prepare a clean air plan. The concentration of PM10 and SO2 has decreased by 50% and 98% respectively since 1990s to 2014 [48], due to numerous measures included in a clean air action plan. Similar to Mexico City, the decrease of NO2 could be underestimated in considering the pre-existing decreasing trend of NO2 in Istanbul. However, a significant increasing trend of SO2 was observed from 2017 to 2020 (Figure S1), signifying the increase of SO2 during 2020 lockdown was a continuous trend, but not specifically caused by the COVID-19 lockdown.
In contrast with Asian and South American countries, air pollution concentrations in European countries and the United States remain stable and at a relatively lower level compared to most Asian and South American countries. This is due to earlier urbanization and implementation of air pollution action plans. In the E.U. countries, the Convention on Long-Range Transboundary Air Pollution (LRTAP) was signed in 1979, aiming to mitigate the air pollution transmitted over long distances by reducing emissions and pollution prevention [49]. Since 1980, numerous directives on the limitations of air pollution concentrations have been implemented [50], and the air quality has been improved in many European countries. For instance, since 2000 in the Paris region the PM2.5 concentration is lower than the World Health Organization suggested limit (25 μg/m3). NO2 and SO2 concentrations also remain stable over the past 4 years [12]. The U.S. implemented the Clean Air Act (CAA) in 1970, which dramatically improved air quality in the U.S. nationally, concentrations of air pollutants in 2019 dropped significantly compared to 2000: 92% (SO2), 62% (NO2), and PM2.5 (43%) [51]. However, in the last 4–5 years, the national annual average of air pollutants is stable. Due to the relatively stable concentrations of air pollutants, especially NO2, SO2 and PM2.5 in those countries, the impact of environmental policies on short-term air quality should be too low to observe.

5. Conclusions

Significant decreases in the AQI of NO2, SO2, CO, PM2.5 and PM10 were observed during the lockdown period in most of 20 megacities in the world relative to 2019 and to the 2017–2019 average for the same period. For the primary pollutants: SO2, NO2 and CO, the significant decreases were directly due to the reduction of emissions caused by lockdown, as citizen mobility was restricted. The difference of NO2 reduction between cities was mainly due to the various lockdown policies, and Wuhan and Delhi exhibited the highest decrease of NO2 due to the strictest lockdown policies. For the secondary pollutants, O3 increased in most of cities, due to photochemical reactions promoting ozone formation under a potential VOC-sensitive regime. PM2.5 and PM10 decreased in 19 and 17 of all cities respectively, but the decrease was less than its precursor gases, especially NO2, as the sources of PM are complex.
Meteorological variability also plays a role in air pollutant concentration: significantly higher rainfall during the lockdown period in Johannesburg could explain the largest decline of PM2.5, and the lower temperature and higher rainfall in Istanbul and Tokyo could explain the exceptional decrease of ozone. In addition, environmental policy regulations, especially in Asian cities, such as Beijing, Wuhan, Seoul and Tokyo, reduced pollutant emissions leading to decreasing NO2, SO2 and PM2.5 concentrations through the past three years prior to the lockdown. Globally, despite the non-negligible impacts of meteorological variability and preceding environmental policy, lockdown explains the large reductions in air pollution.

Supplementary Materials

The following are available online at https://www.mdpi.com/2073-4433/11/11/1189/s1, Figure S1: SO2 AQI in selected worldwide cities during the lockdown period in 2020 compared to the same period of 2017, 2018 and 2019. Figure S2: CO AQI in selected worldwide cities during the lockdown period in 2020 compared to the same period of 2017, 2018 and 2019. Table S1: Name and number of stations for Meteorological data for 20 selected major cities.

Author Contributions

All authors made significant and extensive contributions to the work presented in this manuscript. F.F., K.L.P.-R. and B.W. conceived the idea, all authors designed and discussed data. F.F. conducted the data analysis and wrote the manuscript with guidance from K.L.P.-R. and B.W., all co-authors contributed critical revisions. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by Envirolab Asia and the Henry Luce Foundation.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Shah, A.S.V.; Langrish, J.P.; Nair, H.; McAllister, D.A.; Hunter, A.L.; Donaldson, K.; Newby, D.E.; Mills, N.L. Global association of air pollution and heart failure: A systematic review and meta-analysis. Lancet 2013, 382, 1039–1048. [Google Scholar] [CrossRef] [Green Version]
  2. International Agency for Research on Cancer (IARC). Outdoor Air Pollution: A Leading Environmental Cause of Cancer Deaths. Available online: https://www.iarc.fr/news-events/iarc-outdoor-air-pollution-a-leading-environmental-cause-of-cancer-deaths/ (accessed on 17 October 2013).
  3. Halonen, J.I.; Lanki, T.; Yli-Tuomi, T.; Tiittanen, P.; Kulmala, M.; Pekkanen, J. Particulate air pollution and acute cardiorespiratory hospital admissions and mortality among the elderly. Epidemiology 2009, 20, 143–153. [Google Scholar] [CrossRef]
  4. World Health Organization (WHO). Update and Revision of the Air Quality Guidelines for Europe: Meeting of the Working Group on Classical Air Pollutants; WHO Regional Office for Europe: Copenhagen, Denmark, 1995. [Google Scholar]
  5. Lin, C.-Y.C.; Jacob, D.J.; Fiore, A.M. Trends in exceedances of the ozone air quality standard in the continental United States, 1980–1998. Atmos. Environ. 2001, 35, 3217–3228. [Google Scholar] [CrossRef]
  6. Raub, J.A.; Mathieu-Nolf, M.; Hampson, N.B.; Thom, S.R. Carbon monoxide poisoning—A public health perspective. Toxicology 2000, 145, 1–14. [Google Scholar] [CrossRef]
  7. Poschl, U. Atmospheric aerosols: Composition, transformation, climate and health effects. Angew. Chem. Int. Ed. Engl. 2005, 44, 7520–7540. [Google Scholar] [CrossRef] [PubMed]
  8. Huang, T.; Zhu, X.; Zhong, Q.; Yun, X.; Meng, W.; Li, B.; Ma, J.; Zeng, E.Y.; Tao, S. Spatial and Temporal Trends in Global Emissions of Nitrogen Oxides from 1960 to 2014. Environ. Sci. Technol. 2017, 51, 7992–8000. [Google Scholar] [CrossRef] [PubMed]
  9. Chin, M.; Savoie, D.L.; Huebert, B.J.; Bandy, A.R.; Thornton, D.C.; Bates, T.S.; Quinn, P.K.; Saltzman, E.S.; De Bruyn, W.J. Atmospheric sulfur cycle simulated in the global model GOCART: Model description and global properties. J. Geophys. Res. Atmos. 2000, 105, 24671–24687. [Google Scholar] [CrossRef]
  10. Zhong, Q.; Huang, Y.; Shen, H.; Chen, Y.; Chen, H.; Huang, T.; Tao, S. Global estimates of carbon monoxide emissions from 1960 to 2013. Environ. Sci. Pollut. Res. 2016, 24, 864–873. [Google Scholar] [CrossRef]
  11. United States Environmental Protection Agency (US EPA). National Annual Emissions Trend. Available online: https://www.epa.gov/sites/production/files/2018-04/national_tier1_caps.xlsx (accessed on 31 July 2020).
  12. Airparif. Inventaire Régional des Emissions en Ile-de-France Année de Référence 2012—Eléments Synthétiques Edition mai 2016. Available online: https://www.airparif.asso.fr/_pdf/publications/inventaire-emissions-idf-2012-150121.pdf (accessed on 31 July 2020).
  13. Greater London Authority (GLA). London Atmospheric Emissions Inventory 2017. Available online: https://data.london.gov.uk/dataset/london-atmospheric-emissions-inventory-2013 (accessed on 31 July 2020).
  14. Gurjar, B.; Nagpure, A. Indian megacities as localities of environmental vulnerability from air quality perspective. J. Smart Cities. 2015, 1, 15–30. [Google Scholar] [CrossRef]
  15. EMEP/CEIP. Officially Reported Emission Data for 2019. Available online: https://www.ceip.at/webdab-emission-database/reported-emissiondata (accessed on 31 July 2020).
  16. Klimont, Z.; Kupiainen, K.; Heyes, C.; Purohit, P.; Cofala, J.; Rafaj, P.; Borken-Kleefeld, J.; Schöpp, W. Global anthropogenic emissions of particulate matter including black carbon. Atmos. Chem. Phys. 2017, 17, 8681–8723. [Google Scholar] [CrossRef] [Green Version]
  17. Lu, H.; Stratton, C.W.; Tang, Y.W. Outbreak of pneumonia of unknown etiology in Wuhan China: The mystery and the miracle. J. Med. Virol. 2020, 92, 401–402. [Google Scholar] [CrossRef] [Green Version]
  18. Pepe, E.; Bajardi, P.; Gauvin, L.; Privitera, F.; Lake, B.; Cattuto, C.; Tizzoni, M. COVID-19 outbreak response: A first assessment of mobility changes in Italy following national lockdown. medRxiv 2020. [Google Scholar] [CrossRef] [Green Version]
  19. Kumar, P.; Hama, S.; Omidvarborna, H.; Sharma, A.; Sahani, J.; Abhijith, K.V.; Debele, S.E.; Zavala-Reyes, J.C.; Barwise, Y.; Tiwari, A. Temporary reduction in fine particulate matter due to ‘anthropogenic emissions switch-o’ during COVID-19 lockdown in Indian cities. Sustain. Cities Soc. 2020, 62, 102382. [Google Scholar] [CrossRef] [PubMed]
  20. Venter, Z.S.; Aunan, K.; Chowdhury, S.; Lelieveld, J. COVID-19 lockdowns cause global air pollution declines with implications for public health risk. medRxiv 2020. [Google Scholar] [CrossRef] [Green Version]
  21. Lal, P.; Kumar, A.; Kumar, S.; Kumari, S.; Saikia, P.; Dayanandan, A.; Adhikari, D.; Khan, M.L. The dark cloud with a silver lining: Assessing the impact of the SARS COVID-19 pandemic on the global environment. Sci. Total Environ. 2020, 732, 139297. [Google Scholar] [CrossRef]
  22. Apple Mobility Trends Reports. Available online: https://www.apple.com/covid19/mobility (accessed on 31 July 2020).
  23. International Energy Agency (IEA). Global Energy Review 2020: The Impacts of the COVID-19 Crisis on Global Energy Demand and CO2 Emissions, Flagship Report—April 2020. Available online: https://www.iea.org/reports/global-energy-review-2020/electricity (accessed on 31 July 2020).
  24. Sillman, S. Tropospheric ozone and photochemical smog. Treatise Geochem. 2020, 9, 407–431. [Google Scholar] [CrossRef]
  25. Kim, M.J.; Park, R.J.; Kim, J.J. Urban air quality modeling with full O3-NOx-VOC chemistry: Implications for O3 and PM air quality in a street canyon. Atmos. Environ. 2012, 47, 330–340. [Google Scholar] [CrossRef]
  26. Simon, H.; Reff, A.; Wells, B.; Xing, J.; Frank, N. Ozone trends across the United States over a period of decreasing NOx and VOC emissions. Environ. Sci. Technol. 2015, 49, 186–195. [Google Scholar] [CrossRef] [Green Version]
  27. Charron, A.; Harrison, R.M.; Quincey, P. What are the sources and conditions responsible for exceedances of the 24 h PM10 Limit Value (50 μg m–3) at a heavily trafficked London site? Atmos. Environ. 2007, 41, 1960–1975. [Google Scholar] [CrossRef]
  28. Lenschow, P.; Abraham, H.J.; Kutzner, K.; Lutz, M.; Preu, J.D.; Reichenbacher, W. Some ideas about the sources of PM10. Atmos. Environ. 2001, 35, 23–33. [Google Scholar] [CrossRef]
  29. Wang, Y. China’s ongoing battle against the coronavirus: A scholar-practitioner’s experiences and reflections. Socio. Ecol. Pract. Res. 2020, 2, 181–183. [Google Scholar] [CrossRef]
  30. Tobías, A. Evaluation of the lockdowns for the SARS-CoV-2 epidemic in Italy and Spain after one month follow up. Sci. Total Environ. 2020, 725, 138539. [Google Scholar] [CrossRef]
  31. Mitjà, O.; Arenas, À.; Rodó, X.; Tobias, A.; Brew, J.; Benlloch, J.M. Experts’ request to the Spanish government: Move Spain towards complete lockdown. Lancet 2020, 395, 1193–1194. [Google Scholar] [CrossRef]
  32. Pullano, G.; Valdano, E.; Scarpa, N.; Rubrichi, S.; Colizza, V. Population mobility reductions during covid-19 epidemic in France under lockdown. medRxiv 2020. [Google Scholar] [CrossRef]
  33. Hampton, M.; Clark, M.; Baxter, I.; Stevens, R.; Flatt, E.; Murray, J.; Wembridge, K. The effects of a UK lockdown on orthopedic trauma admissions and surgical cases: A Multicentre comparative study. Bone Joint Open 2020, 1, 137–143. [Google Scholar] [CrossRef]
  34. Perez-Brumer, A.; Silva-Santisteban, A. COVID-19 policies can perpetuate violence against transgender communities: Insights from Peru. AIDS Behav. 2020, 24, 2477–2479. [Google Scholar] [CrossRef]
  35. Morita, H.; Nakamura, S.; Hayashi, Y. Changes of Urban Activities and Behaviors Due to COVID-19 in Japan. Available online: http://dx.doi.org/10.2139/ssrn.3594054 (accessed on 6 May 2020).
  36. Aum, S.; Lee, S.Y.T.; Shin, Y. COVID-19 Doesn’t Need Lockdowns to Destroy Jobs: The Effect of Local Outbreaks in Korea. National Bureau of Economic Research, Working Paper 27264. Available online: https://doi.org/10.3386/w27264 (accessed on 2 October 2020).
  37. Armbruster, S.; Klotzbücher, V. Lost in Lockdown? COVID-19, Social Distancing, and Mental Health in Germany, Diskussionsbeiträge, No. 2020-04. Available online: http://hdl.handle.net/10419/218885 (accessed on 2 October 2020).
  38. Kaskun, S.; Ulutas, K. The Effect of COVID-19 Pandemic on Air Quality Caused by Traffic in Istanbul. Res. Square 2020, 5. [Google Scholar] [CrossRef]
  39. Brzezinski, A.; Kecht, V.; Van Dijcke, D. The Cost of Staying Open: Voluntary Social Distancing and Lockdowns in the US. Working Paper. 2020. Available online: http://dx.doi.org/10.2139/ssrn.3614494 (accessed on 2 October 2020).
  40. Bargain, O.; Aminjonov, U. Trust and Compliance to Public Health Policies in Times of COVID-19. IZA 2020, DP 13205. Available online: http://ftp.iza.org/dp13205.pdf (accessed on 2 October 2020).
  41. Jacob, D.J.; Winner, D.A. Effect of climate change on air quality. Atmos. Environ. 2009, 43, 51–63. [Google Scholar] [CrossRef] [Green Version]
  42. Fan, H.; Zhao, C.; Yang, Y. A comprehensive analysis of the spatio-temporal variation of urban air pollution in China during 2014–2018. Atmos. Environ. 2019, 220, 117066. [Google Scholar] [CrossRef]
  43. Sharma, R.; Kumar, R.; Sharma, D.K.; Son, L.H.; Priyadarshini, I.; Pham, B.T.; Bui, D.T.; Rai, S. Inferring air pollution from air quality index by different geographical areas: Case study in India. Air Qual. Atmos. Health 2019, 12, 1347–1357. [Google Scholar] [CrossRef]
  44. Zhai, S.; Jacob, D.J.; Wang, X.; Shen, L.; Li, K.; Zhang, Y.; Gui, K.; Zhao, T.; Liao, H. Fine particulate matter (PM2.5) trends in China, 2013–2018: Separating contributions from anthropogenic emissions and meteorology. Atmos. Chem. Phys. 2019, 19, 11031–11041. [Google Scholar] [CrossRef] [Green Version]
  45. Zheng, B.; Tong, D.; Li, M.; Liu, F.; Hong, C.; Geng, G.; Li, H.; Li, X.; Peng, L.; Qi, J.; et al. Trends in China’s anthropogenic emissions since 2010 as the consequence of clean air actions. Atmos. Chem. Phys. 2019, 18, 14095–14111. [Google Scholar] [CrossRef] [Green Version]
  46. Krotkov, N.A.; McLinden, C.A.; Li, C.; Lamsal, L.N.; Celarier, E.A.; Marchenko, S.V.; Swartz, W.H.; Bucsela, E.J.; Joiner, J.; Duncan, B.N.; et al. Aura OMI observations of regional SO2 and NO2 pollution changes from 2005 to 2015. Atmos. Chem. Phys. 2016, 16, 4605–4629. [Google Scholar] [CrossRef] [Green Version]
  47. Ministry of Environment, Forest & Climate Change of India (MoEFCC). National Clean Air Programme. Available online: http://moef.gov.in/wp-content/uploads/2019/05/NCAP_Report.pdf (accessed on 2 October 2020).
  48. Sevimoglu, O. Assessment of Major Air Pollution Sources in Efforts of Long-Term Air Quality Improvement in Istanbul. Sakarya Univ. J. Sci. 2020, 24, 389–405. [Google Scholar] [CrossRef] [Green Version]
  49. Convention on Long-Range Transboundary Air Pollution (CLRTAP). Protocols to the Convention on Long-Range Transboundary Air Pollution; United Nations Economic Commission for Europe: Geneva, Switzerland; Available online: http://www.unece.org/fileadmin//DAM/env/lrtap/status/lrtap_s.htm (accessed on 2 October 2020).
  50. Kuklinska, K.; Wolska, L.; Namiesnik, J. Air quality policy in the U.S. and the EU—A review. Atmos. Pollut. Res. 2015, 6, 129–137. [Google Scholar] [CrossRef]
  51. United States Environmental Protection Agency (US EPA). National Air Quality: Status and Trends of Key Pollutants. Available online: https://www.epa.gov/air-trends/) (accessed on 31 July 2020).
Atmosphere 11 01189 g001 550
Figure 1. Twenty major cities impacted by COVID-19 that applied lockdown or social distancing policies around the world.
Figure 1. Twenty major cities impacted by COVID-19 that applied lockdown or social distancing policies around the world.
Atmosphere 11 01189 g001
Atmosphere 11 01189 g002 550
Figure 2. NO2 AQI in 20 worldwide cities during the lockdown period in 2020 compared to the same period of 2017, 2018 and 2019. Line within the box: the median, box: first and third quartiles, whiskers: non-outlier range, dot: outliers. Years sharing the same letter mean AQI are not significantly different (p > 0.05). The grey box for 2020: the change in 2020 is significant relative to every previous year.
Figure 2. NO2 AQI in 20 worldwide cities during the lockdown period in 2020 compared to the same period of 2017, 2018 and 2019. Line within the box: the median, box: first and third quartiles, whiskers: non-outlier range, dot: outliers. Years sharing the same letter mean AQI are not significantly different (p > 0.05). The grey box for 2020: the change in 2020 is significant relative to every previous year.
Atmosphere 11 01189 g002
Atmosphere 11 01189 g003 550
Figure 3. Ground-level ozone AQI in selected worldwide cities during the lockdown period in 2020 compared to the same period of 2017, 2018 and 2019. Line within the box: the median, box: first and third quartiles, whiskers: non-outlier range, dot: outliers. Years sharing the same letter mean not significantly different (p > 0.05). The grey box for 2020: the change in 2020 is significant relative to every previous year.
Figure 3. Ground-level ozone AQI in selected worldwide cities during the lockdown period in 2020 compared to the same period of 2017, 2018 and 2019. Line within the box: the median, box: first and third quartiles, whiskers: non-outlier range, dot: outliers. Years sharing the same letter mean not significantly different (p > 0.05). The grey box for 2020: the change in 2020 is significant relative to every previous year.
Atmosphere 11 01189 g003
Atmosphere 11 01189 g004 550
Figure 4. PM2.5 AQI in selected worldwide cities during the lockdown period in 2020 compared to the same period of 2017, 2018 and 2019. Line within the box: the median, box: first and third quartiles, whiskers: non-outlier range, dot: outliers. Years sharing the same letter mean not significantly different (p > 0.05). The grey box for 2020: the change in 2020 is significant relative to every previous year.
Figure 4. PM2.5 AQI in selected worldwide cities during the lockdown period in 2020 compared to the same period of 2017, 2018 and 2019. Line within the box: the median, box: first and third quartiles, whiskers: non-outlier range, dot: outliers. Years sharing the same letter mean not significantly different (p > 0.05). The grey box for 2020: the change in 2020 is significant relative to every previous year.
Atmosphere 11 01189 g004
Atmosphere 11 01189 g005 550
Figure 5. PM10 AQI in selected worldwide cities during the lockdown period in 2020 compared to the same period of 2017, 2018 and 2019. Line within the box: the median, box: first and third quartiles, whiskers: non-outlier range, dot: outliers. Years sharing the same letter mean not significantly different (p > 0.05). The grey box for 2020: the change in 2020 is significant relative to every previous year.
Figure 5. PM10 AQI in selected worldwide cities during the lockdown period in 2020 compared to the same period of 2017, 2018 and 2019. Line within the box: the median, box: first and third quartiles, whiskers: non-outlier range, dot: outliers. Years sharing the same letter mean not significantly different (p > 0.05). The grey box for 2020: the change in 2020 is significant relative to every previous year.
Atmosphere 11 01189 g005
Atmosphere 11 01189 g006 550
Figure 6. Meteorological condition changes in selected cities during lockdown period in 2020 and the same period in 2019. (A) Average temperature, blue boxes mean significant change of average temperature: p < 0.05 according to ANOVA and Tukey HSD tests. (B) Total rainfall, no rainfall data for Berlin and Moscow from either source.
Figure 6. Meteorological condition changes in selected cities during lockdown period in 2020 and the same period in 2019. (A) Average temperature, blue boxes mean significant change of average temperature: p < 0.05 according to ANOVA and Tukey HSD tests. (B) Total rainfall, no rainfall data for Berlin and Moscow from either source.
Atmosphere 11 01189 g006
Atmosphere 11 01189 g007 550
Figure 7. NO2 reduction percentage during lockdown period in 2020 relative to 2019 for the same period in 20 major cities around the world. Red: cities with strictest lockdown policy (human mobility restriction and stopped public transportation), grey: cities with similar strict lockdown policy (full or partial lockdown, with different social distancing measures), blue: less strict lockdown policy or no lockdown.
Figure 7. NO2 reduction percentage during lockdown period in 2020 relative to 2019 for the same period in 20 major cities around the world. Red: cities with strictest lockdown policy (human mobility restriction and stopped public transportation), grey: cities with similar strict lockdown policy (full or partial lockdown, with different social distancing measures), blue: less strict lockdown policy or no lockdown.
Atmosphere 11 01189 g007
Table
Table 1. Start and end of lockdown period and lockdown policies in 20 selected major cities.
Table 1. Start and end of lockdown period and lockdown policies in 20 selected major cities.
CityCountryLockdownLockdown Policy
FromToTotal/PartialOther Local Actions
WuhanChina23-Jan7-AprTotalStrictest human mobility restrictions, public transportation stopped
DelhiIndia24-Mar31-MayTotalOnly those who work for “essential services” can move freely, public transportation stopped
LimaPeru15-Mar30-JunTotalGender-based mobility restriction
MadridSpain14-Mar4-MayTotalOutdoor physical exercise banned
TehranIran13-Mar17-AprTotalShops, streets and roads cleared
MoscowRussia30-Mar8-JunTotalDigital pass required for car or public transport use
RomeItaly9-Mar3-MayTotalOnly allowed to go out alone near home
ParisFrance17-Mar10-MayTotalPermit required for going out
LondonU.K.23-Mar10-MayTotalOnly go outside to buy food, to exercise once a day, or go to work if they absolutely cannot work from home
JohannesburgSouth Africa26-Mar30-AprTotalSevere restrictions on travel and movement
SydneyAustralia23-Mar27-AprTotalStay-at-home except for essential outings
BeijingChina10-Feb27-MarPartialOrdering residential communities and villages to limit access for outsiders
New York CityU.S.A22-Mar7-JunPartialStatewide stay-at-home order
Los AngelesU.S.A19-Mar7-MayPartialStatewide stay-at-home order, social pressure for violations
Mexico CityMexico23-Mar30-MayPartialClosing gyms, museums and clubs, banning big gathering, restricting mobility to areas less affected
Sao PauloBrazil24-Mar10-MayPartialSocial distancing measures
BerlinGermany17-Mar19-AprPartialRules differing across states. Only go out alone or with a person from same household.
SeoulSouth Korea24-Feb6-May--No strict lockdown, social distancing applied, no movement restriction
TokyoJapan7-Apr24-May--State of emergency, encouraged social distancing
IstanbulTurkey21-Mar10-May--Weekend curfew
Note: full lockdown: national lockdown, partial lockdown: regional/statewide lockdown, --: no official lockdown, only social distancing or curfew measurement.
Table
Table 2. Percentage (%) change in AQI for NO2, SO2, CO, ground-level O3, PM2.5, and PM10 during lockdown period in 2020 compared to average of 2017–2019, and to 2019 single year for the same period in 20 major cities in the world. Bold red: the change is statistically significant (p < 0.05) relative to 2019 or to at least one of the 3 past years, according to ANOVA and Tukey HSD tests. +: increase of AQI, −: decrease of AQI.
Table 2. Percentage (%) change in AQI for NO2, SO2, CO, ground-level O3, PM2.5, and PM10 during lockdown period in 2020 compared to average of 2017–2019, and to 2019 single year for the same period in 20 major cities in the world. Bold red: the change is statistically significant (p < 0.05) relative to 2019 or to at least one of the 3 past years, according to ANOVA and Tukey HSD tests. +: increase of AQI, −: decrease of AQI.
ContinentCityCountryNO2SO2COGround-level O3PM2.5PM10
To 17–19To 19To 17–19To 19To 17–19To 19To 17–19To 19To 17–19To 19To 17–19To 19
AsiaWuhanChina−63.3−58.3−28.6−11.2−17.3−13.8+45.6+54.2−26.2−23.2−31.9−30.4
Beijing−41.8−33.7−60.3−33.5−7.1+92.4+10.7+5.1−15.2−7.2+0.1−14.8
SeoulSouth Korea−28.0−25.8−28.3−20.8−14.5−14.6+10.9+10.2−21.2−19.1−8.8+19.5
TokyoJapan−25.8−19.5−37.6−28.1−3.1+7.3−5.3−6.7−22.9−11.4−24.1−11.0
DelhiIndia−57.7−60−23.7−31.7−22.8−34.0+19.3+36.3−31.0−27.6−47.9−45.9
Tehran *IranNA−35.2NANANANANANANA−21.9NA−37.9
IstanbulTurkey−36.5−19.9+53.8+29.3+48.5+2.9−8.3−43.6−19.1−3.4−22.4−19.0
EuropeMoscow †Russia−35.8−39.8−25.5−25.8−18.8−18.7NA+6.4−12.7−25.9−29.6−42.5
BerlinGermany−45.0−17.6NA+29.5NANA+16.4+3.9−27.4−22.5−15.1−10.3
RomeItaly−45.0−36.4−8.6−3.8NANA+4.4+2.9+0.1+8.3−6.6−1.8
ParisFrance−47.8−46.4−20.5+13.5NANA+24.0+26.8−4.5−13−14.7−22.3
MadridSpain−56−51.6+7.1−54.8NANA−4.9−9.9−2.4−1.6−17.2−19.8
LondonU. K−39.6−37.8+1.3+0.6−35.0−53.5+47.7+48.0−8.8−14−4.7−10
North AmericaNew York City #U. S−33.7−27.5NANA−22.0−20.7+1.0−6.3−30.8−27.6NANA
Los Angeles−29.0−24.2NANA−36.97.2−11.7−3.5−20.4−13.7−18.1+2.0
Mexico CityMexico−35.2−24.9−14.9+2.2−1.2+4.0+7.9−0.1−3.7−8.3−8.8−15.5
South AmericaLima ‡Peru−63.4−50.5−18.1−35.2−58.6−61.8−28.5−42.9−27.1−19.4−42.3−31.7
Sao PauloBrazil−36.0−37.8−27.3−23.2−31.8−32.7+33.9+24.9−11.3−18.3−5.4−12.9
AfricaJohannesburg §South AfricaNA−23.0NA−13.9NA5.0NA+9.0NA−31.3NA−33.1
OceaniaSydneyAustralia−15.0−11.1NANA−25.5−24+5.2+5.6−34.7−29.2−19.7−17.0
*: No available data in 2019, the comparison was made with 2018. †: the start date is 3 April instead of 30 March according to the data availability. ‡: the start date is 29 March instead of 15 March according to the data availability. §: no available data in 2017 and 2018, only compared to 2019. #: Manhattan area of New York City.
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Fu, F.; Purvis-Roberts, K.L.; Williams, B. Impact of the COVID-19 Pandemic Lockdown on Air Pollution in 20 Major Cities around the World. Atmosphere 2020, 11, 1189. https://doi.org/10.3390/atmos11111189

AMA Style

Fu F, Purvis-Roberts KL, Williams B. Impact of the COVID-19 Pandemic Lockdown on Air Pollution in 20 Major Cities around the World. Atmosphere. 2020; 11(11):1189. https://doi.org/10.3390/atmos11111189

Chicago/Turabian Style

Fu, Franck, Kathleen L. Purvis-Roberts, and Branwen Williams. 2020. "Impact of the COVID-19 Pandemic Lockdown on Air Pollution in 20 Major Cities around the World" Atmosphere 11, no. 11: 1189. https://doi.org/10.3390/atmos11111189

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop