Air Pollution and COVID-19: Any Causal Link?

Globally, an improvement in air quality was observed during the lockdown period in 597 major cities in 76 countries [Feng Liu et al. 2021]. NO₂ showed the largest percentage of reduction, followed by PM₁₀, SO₂, PM_2.5, and CO. In addition, a 10–27% increase in O₃ was observed. The rate of expected premature deaths declined due to improved air quality in all countries studied. In major cities worldwide – New York, Los Angeles, Zaragoza, Rome, Dubai, New Delhi, Mumbai, Beijing, and Shanghai – PM_2.5 concentrations declined as a result of the lockdown effect, even though COVID-19 cases were high at that time. In 37 states of the US, a 25.5% reduction in NO₂ was observed during the pandemic compared to the pre-COVID-19 period [Jesse, Ebisu 2020]. The decline is mainly confined to urban areas, however; the rural regions remained largely unaffected.

In the top hotspots of Italy, France, and Spain in Europe and Wuhan in China, the reduction in NO_x was 56%. In the case of PM, there was a 42% reduction in Wuhan and an 8% decline in Europe [Pierre et al. 2020]. In the cities of Spain, the most important source of pollution is traffic, from internal combustion emissions from motor vehicles. During the pandemic, a reduction in NO₂ concentration in Barcelona and Madrid was observed compared to the previous two years considering all the meteorological conditions. The reduction in maximum hourly peak ratios was between 1.2 and 1.7 [Baldasano 2020]. In the metropolitan city of Milan, there was an appreciable reduction in CO due to reduced vehicular transmissions. At the same time, SO₂ dropped only in the city metropolis compared to adjacent areas due to a reduction in the use of fuel for heating [Collivinarelli et al. 2020]. The average value of PM_2.5 in Rome in March 2020 was 159% lower compared to January 2020 [[Chauhan, Singh 2020].

On the other hand, a short-term decline in pollution levels was observed in 15 central monitoring stations after the shutdown in New York City, whereas no significant difference had been observed for over five years from 2015 to 2020 over 17 weeks during the months of January through May. A major improvement was found only in regions with higher levels of air pollutants [Zangari et al. 2020]. In Ontario, Canada, PM_2.5 didn’t show any change in 12 out of 32 monitoring sites, because 56% of PM_2.5 emissions are from residential sources [Adams 2020].

With the largest population in the world, China is affected by heavy air pollution, especially with PM_2.5. The major contributor to PM_2.5 is the industrial sector [Shi et al. 2017]. The transport sector contributes 6–12% [Tao et al. 2017], while the residential sector contributed 39% of PM_2.5 emissions in China in 2010 [Li et al. 2017]. During the lockdown period, out of 366 urban areas investigated in China's mainland [Wang et al. 2020], there was a decline in PM_2.5 from 65 to 51.4 μg/m³ (micrograms per cubic millilitre) in 315 cities. The decline in PM_2.5 concentrations positively correlates with motor vehicle numbers and percentages of secondary industries. The reduction in PM_2.5 in Shanghai, Jinan, Shijiazhuang, Beijing, Taiyuan, Xian, Tianjin, Zhengzhou, and Wuhan varied from 7.78 to 22.08 μg/m³ [Pengfei et al. 2020]. However, a wet deposition of PM_2.5 occurred in Beijing and Guangzhou during the study period because of rare rains. The low PBL increased the atmospheric stability, and low wind speed impeded dispersal of the pollutants.

In Wuhan [Zheng et al. 2020], the first city to adopt lockdown measures, the average mass concentrations of PM_2.5 decreased from 72.9 μg/m³ to 45.9 μg/m³ compared to the previous year. The main chemical species were biomass, coal combustion, fireworks emissions, industrial processes, road dust, secondary inorganic aerosol, and vehicle emissions. PM_2.5 declined, except from biomass and combustion of fireworks and secondary inorganic aerosols. The primary emissions reduced PM_2.5, whereas secondary formation increased PM_2.5.

The atmospheric oxidation process with respect to aerosol formation could, however, lead to air pollution even if anthropogenic sources are suppressed. This was observed in Shanghai, China. During the quarantine period, two episodes of PM_2.5 exceeding 100 μg/m³ occurred [H. Chen et al. 2020]. In the three metropolises of Beijing, Wuhan, and Guangzhou, which are not geographically adjacent and have different climatic characteristics, NO₂, the major pollutant, decreased by 28%, 58%, and 46% in these cities, respectively, during the lockdown period compared to the previous year during the same period. The PM_2.5 in these places was either steady or increased/decreased. The travel restrictions imposed during the lockdown in 44 cities in Northern China reduced human mobility by 69.8% [Bao, Zhang 2020]. The concentrations of the major pollutants declined as a result of the travel ban. The reduction in air quality index (AQI), PM_2.5, and CO was partially mediated, and SO₂, PM₁₀, and NO₂ were completely mediated by human mobility.

In northern parts of peninsular southeast Asian countries, NO₂ and aerosol optical depths remained high. Even with no traffic and no industrial activities, achieving good air quality in this region is a challenge, especially during March and April due to extensive vegetation and peat fires, but in urban and industrial sites in Malaysia, a significant reduction in NO₂ is observed [Kanniah et al. 2020].

In the eastern province of Saudi Arabia, the lockdown measures reduced PM₁₀ between 21% and 70% in most of the sites. The CO and NO₂ emissions were produced mainly from transportation, so they were reduced substantially during the lockdown phase. SO₂ reduction was insignificant since the eastern province is a low SO₂-emitting region because of stringent air pollution regulations. Nevertheless, there was an increase in O₃ concentrations [Anil, Alagha 2020].

In Karachi, Pakistan, a 50% reduction in PM₁₀ was observed during partial lockdown compared to pre-lockdown levels, and a 60%–70% reduction in SO₂, NO₂, and CO was recorded during the lockdown period. The overall air quality improved by 40% to 50% just four days after the commencement of lockdown [Shareef, Hashmi 2020].

In Almaty, Kazakhstan, the traffic-free conditions didn’t reduce air quality as a result of the domination of several primary emission sources, although PM_2.5 declined by 21% during lockdown compared to the two previous years. O₃ increased because of high levels of solar activity, and coal combustion cause SO₂ and the toluene/benzene ratio to increase. The toluene/benzene ratio was elevated during the lockdown, which suggests that these are emitted from coal-related sources like burning garbage, household use, and use by power plants [Aiymgul et al. 2020].

In Sale City, Morocco, before and during the lockdown, the difference in concentrations of NO₂, SO₂, and PM₁₀ were 96%, 49%, and 75%. SO₂ is a major precursor for nucleation formation of new particles. The variation in SO₂ is less than that of the other pollutants because the fuel and lubricant oils used have low sulphur content, and also because of the closure of industrial activities and commercial ship transport [Anas et al. 2020].

In the two Indian cities of New Delhi and Mumbai, which have very high pollution levels, a decline of 35% and 14%, respectively, in PM_2.5 is observed compared to the previous year. Power generation plants in India were a significant source of NO₂ during March 2020. In two major Indian cities, Mumbai and New Delhi, energy consumption declined. Ship transportation was also affected by the lockdown. These two cities experienced a substantial reduction in NO₂ of 40% to 50% compared to the same period in the previous year [Shehzad et al. 2020]. In New Delhi, particulate matter concentrations declined to half what they were before the lockdown. On the second and fourth day of lockdown, the air quality had improved by about 40%–60% [Mahato et al. 2020].

In India's four most polluted cities, Agra, Noida, Gurugram, and Delhi, the pollutants PM_2.5, PM₁₀, NO₂, and CO were reduced significantly. PM_2.5 and PM₁₀ declined by more than 40% at all the sites where it was recorded [Abhishek, Shani 2021]. A reduction of 85.1% in the PM_2.5 concentration at Ghaziabad, the highest polluting city in India during the lockdown period was remarkable when compared to the concentration over the previous three months. Other pollutants, like CO, NO₂, and PM₁₀, also declined significantly [Lokhandwala, Gautam 2020].

Indrajit Chowdhuri et al. (2020) observed a reduction of 51% in PM₁₀, 68% in NO₂, 40% in SO₂, and 42% in O₃ in Kolkata. Kolkata is a region where thunderstorms are frequent, and as a result, the reduction of surface pollution in this environment is strongly related to lightning activity. As the pollutants and aerosol concentrations declined by 40% during lockdown the overall occurrence of lightning events also declined by 49.6%.

S. Selvam et al. (2020) investigated the effect of lockdowns on air quality during lockdowns in different zones of the state of Gujarat, India. In the zone where the power plants were functioning, CO reduction was lower than that of SO₂ and NO₂. In the zone where the textile industry, pharmaceutical companies, petroleum companies, beverage manufacturing, and steel recycling are predominant, a reduction of 47% in PM₁₀, 58% in SO₂, 82% in NO₂, 38% in CO, and 38% in PM_2.5 was observed, and O₃ increased by 38%. The restrictions imposed on the navigation activities of large cargo ships, ferries, and cruises led to a 78% reduction in PM_2.5, 80% in PM₁₀, 58% in SO₂, 50% in NO₂, 40% in CO, and an increase in O₃ by 48%. The lack of dominant industry and transport activity in zone 4 meant minimal changes occurred in CO and O₃ because the population is small and only a small number of industries are located there.

The main precursors of ozone are NO₂ and VOCs. Steady emissions of formaldehyde (HCHO) in urban areas helps in the formation of O₃. Hence, regardless of the cities’ reduction in NO₂ during lockdown, O₃ increased as a result of the HCHO concentration, a proxy for VOC. Thus, the lockdown had little influence on ozone levels [Pei et al. 2020]. The increase in O₃ was due to a higher airborne benzene concentration in Milan even though VOCs from vehicular and industrial combustion were minimized. The other VOCs may act similarly to benzene [Collivignarelli et al. 2020].

In top hotspots, Italy, France, Spain, and Wuhan, China, lockdowns caused O₃ to increase in all cities studied. The relative increase in O₃ is higher than the weekend effect [Pierre et al. 2020]. In the Yangtze River delta of China, O₃ rebounded [Li et al. 2020]. In 31 provincial cities in China, an increase in O₃ and a reduction in NO₂ was observed [Nie et al. 2021].

At two stations in Rio de Janeiro, where atmospheric chemistry is under controlled conditions and nonmethane hydrocarbons (NHMC) are available, O₃ increased as a result of an increase in the NHMC/NO_X ratio. NHMC emissions are caused by light-duty vehicles and motorcycles. Therefore, NHMC monitoring is essential and must be included in air quality standards [Siciliano et al. 2020].

While most studies reported an increase in O₃ concentrations, they were lower at 12 out 32 monitoring sites in Ontario, Canada, than they had been in the previous five years because of reduced transportation and a reduction in NO₂ and NO_x [Adams 2020]. Kolkata is a region with frequent thunderstorms, and the reduction of surface pollution in this environment is strongly related to lightning activity. The heat generated by lightning is ten times higher than radiant solar insolation and produces large amounts of NO₂ and O₃. The average number of lightning flash events decreased by 49.6%, reducing nitrate oxide and O₃. A 42% reduction in O₃ was found [Chowdhuri et al. 2020]. In the north Indian cities of Agra, Noida, Gurugram, and Delhi, a 98% increase in O₃ was observed in Agra, while a significant reduction was found in other sites. The O₃ variation at Agra is positively correlated with temperature. The positive correlation is because radiation controls the temperature and, therefore, has higher photolytic effectiveness [Abhishek, Shani 2021]. Thus, ozone declined wherever factors other than NO₂ and VOCs are predominant.

AQI represents the magnitude of pollution in air. Hao Xu et al. (2020) investigated the association of AQI with confirmed COVID-19 cases in 33 locations in China. During the outbreak period, COVID-19 transmission associated with AQI was higher when the temperature ranged between 10 °C and 20 °C and relative humidity ranged between 10% and 20%.

Air pollutant levels are also one of the indicators to assess vulnerability to COVID-19. A strong short-term correlation between NO₂, PM_2.5 and COVID-19 fatality is observed in London and borough cities of the United Kingdom. Either long- or short-term exposure to pollutants adversely affects lung function and increases the risk of dying from COVID-19 [Sasidharan et al. 2020]. A 1-μg/m³ increase in long-term PM_2.5 increases the risk of COVID19 mortality by 15% [Jesse, Keita 2020].

Exposure to high levels of NO₂ and PM_2.5 exacerbates the symptoms of COVID-19. A strong correlation between the severity of COVID-19 and air pollutants was investigated in 21 Italian regions. The highest number of cases of COVID-19 were recorded in Italy's most polluted regions. The correlation between mean PM_2.5 with the total number of cases, ICU admissions, deaths, and hospitalized patients was 0.6, 0.65, 0.53, and 0.62, respectively. Exposure to the pollutants PM_2.5 and NO₂ may increase the viral load in patients, because NO₂ causes pulmonary damage and chronic exposure to PM_2.5 leads to alveolar ACE-2 receptor overexpression [Frontera et al. 2020]. In 13 out of 31 provinces in Iran, it was observed that more cases coincided with higher concentrations of CO and NO₂ in the air. SO₂ was negatively correlated with COVID-19 caseload [Bigdeli et al. 2020].

Compared to other countries studied, Europe (Italy, Spain, France and Germany) exhibited a higher concentration of NO₂ accompanied a build-up of NO₂ over the surface caused by downward airflow. Long-term exposure to NO₂ causes diabetes, hypertension, and cardiovascular diseases. This may be one of the contributors to fatalities caused by the COVID-19 virus in these regions. The five areas in the two main hotspots of COVID-19, Northern Italy and Central Spain, have higher concentrations of NO₂ combined with downward airflow [Ogen 2020].

In 120 cities in China, a positive correlation of COVID-19 with PM_2.5, PM₁₀, NO₂, O₃, and CO and a negative correlation with SO₂ with confirmed cases was found. The negative correlation of SO₂ is due to veridical property. It was observed that a 10 μg/m³ increase in PM_2.5, PM₁₀, NO₂, and O₃ and a 1 mg/m³ in CO was associated with 2.24%, 1.76%, 6.94%, 4.76%, 15.11% increases, respectively, in daily counts of COVID-19 confirmed cases [Zhu et al. 2020].

Temperature also plays a vital role in determining the epidemic profile. Out of the ten most affected provinces in China, three exhibited a positive relationship between temperature and COVID-19; two provinces recorded a negative relationship, while the remaining displayed mixed trends. The COVID-19 incidence in Hubei, the most affected province, was positively correlated with temperature [Shahzad et al. 2020]. In Turkey, nine cities were included in the study. The correlation between COVID-19 and weather parameters (temperature, dew point, humidity, and wind speed) resulted a finding of a high correlation between wind speed with the number of cases. The average temperature shows a correlation, but minimum/maximum temperature, rainfall, and humidity are not significantly correlated with COVID-19 incidence. Some have claimed that high temperature and low humidity eliminate the virus's viability [Sahin 2020].