1 Introduction

The rapid transmission of the COVID-19 pandemic has become a major health threat globally. In late December 2019, an epidemic of pneumonia linked with novel coronavirus disease-2019 (COVID-19) (Huang et al. 2020a, b; Wu et al. 2020) that has its epicenter in Wuhan, China, has spread over 215 countries worldwide (WHO 2020a). The COVID-19 epidemic was declared a public health emergency at a global level by the World Health Organization (WHO) on Jan 30, 2020 (WHO 2020b). The COVID-19 cases burst at an alarming rate with significant mortality (Wang et al. 2020a, b) and have impacted human life and the world economy. The transmission of COVID-19 has mostly occurred between people via respiratory droplets and contact routes (Burke et al. 2020; Chan et al. 2020; Li et al. 2020; Liu et al. 2020a, b; WHO 2020c).

Breaking the chain of transmission of SARS-CoV-2 via social isolation is the only way to control the exponential spread of the COVID-19 pandemic. Social distancing is a prevention strategy to control infection by discouraging/decreasing contact with infected and non-infected people, contaminated surfaces, and also among the general public to slow down the rate of spread of disease. Many countries have imposed dramatic interventions like lockdown of the entire country, curtailing human interaction, restriction on public gatherings, and public transportations to adopt social distancing (He et al. 2020). The first lockdown was intruded in Wuhan, China, and later followed by other countries (Jing 2020). Indian government locked the country for fourteen hours in the name of “Janta Curfew” after Prime Minister’s call on March 22, 2020. After Janta Curfew, India went into complete lockdown on March 25 for 21 days followed by the next lockdown for 19 days from April 15 to May 3, the third phase from May 4 to May 17 and the fourth phase from May 18 to May 31, 2020. The lockdown restricted movement of the public and shut down commercial and industrial establishments that adversely affected the growth and economy of India, but on the flip side, our society and environment gained substantial benefits that would be advantageous to the layperson and moderately counterbalance the cost of this pandemic. Due to the lockdown, dramatic positive changes have happened in the air quality of the world that showed a glimpse of a cleaner world. As per media intelligence, a massive reduction in air pollution due to lockdown has been observed in numerous countries by satellite pictures. A few scientific researchers also reported a significant augmentation in air quality because of COVID-19 lockdown worldwide (Cadotte 2020; Dantas et al. 2020; He et al. 2020; 2020a, b; Isaifan 2020; Kerimray et al. 2020; Muhammad et al. 2020; Ogen 2020; Shrestha et al. 2020; Sicard et al. 2020; Wang and Su 2020; Wang et al. 2020a, b, Zhu et al. 2020) and in India (Chauhan and Singh 2020; Gautam 2020; Mahato et al. 2020; Saadat et al. 2020; Sharma et al. 2020).

Air quality has emerged as an international community health concern in the last few decades. Air pollution is a severe environmental health threat that causes premature mortality and attributes over seven million deaths per annum worldwide (WHO 2020d). Air pollution means the presence of harmful substances in the air such as airborne particulate matter (PM), fine particulate matter (PM2.5), coarse particulate matter (PM10), and gaseous pollutants like ozone (O3), nitrogen dioxide (NO2), volatile organic compounds (like benzene), carbon monoxide (CO), sulfur dioxide (SO2), etc. (Harrison et al. 2002; Newby et al. 2015). According to WHO, particulate pollution, ground-level O3, CO, sulfur dioxide (SO2), nitrogen oxides (NOx), and lead (Pb) are the six major air pollutants that adversely affect human health and the environment (Ghorani-Azam et al. 2016). Particulate matter (PM) pollutant is responsible for most of the pulmonary and cardiac diseases, and mortality (Sadeghi et al. 2015; Sahu et al. 2014). According to a study of Global Burden of Disease (GBD), PM2.5 is the 5th most vulnerable pollutant responsible for the death of humans worldwide, which triggered 4.2 million deaths in the year 2015 (Cohen et al. 2017). Ground-level ozone is a probable risk factor for respiratory diseases, predominantly asthma (Gorai et al. 2014), and carbon monoxide toxicity causes hypoxia, apoptosis, and ischemia (Akyol et al. 2014). Exposure of sulfur dioxide is linked with bronchospasm, pulmonary edema, pneumonitis, and acute airway blockage, and nitrogen oxides prominently boost the danger of respiratory problems (Chen et al. 2007). According to WHO, out of the ten most polluted cities worldwide, nine cities are from India. Delhi stands at 6th position in the polluted cities ranking list (Yuda 2019) and holds the 1st rank among all the PM10 polluted cities (Donkelaar et al. 2016; WHO 2018). In India, the concentration of air pollutants is beyond ambient air quality standards of the WHO and Central Pollution Control Board (CPCB), India (Garaga et al. 2018; Mukherjee and Agrawal 2018).

Air quality index is an extensively used tool that describes the severity of air pollution (Nagendra et al. 2007). The air quality index is a single value indicator for the air quality assessment to measure the effects on human health (Bortnick et al. 2002; Kyrkilis et al. 2007; Murena 2004; Thom and Ott 1976).

Hence, in this study, the effects of the lockdown on the air quality of the eight metropolitan cities of India were analyzed. The data of six major criteria air pollutants—PM2.5, PM10, CO, NO2, SO2, and O3—and air quality index were collected from the official website of the CPCB, India during the lockdown period from March 25 to May 3, 2020, and the same period of the previous year 2019. For assessing the consequences of the lockdown on air quality and daily concentrations of air pollutants, the data of the COVID-19 lockdown period were compared with the data of the same period of the previous year 2019 that helps in better understanding of the enhancement in air quality as a result of COVID-19 lockdown.

2 Materials and methods

2.1 Selection of the metropolitan cities

Eight major metropolitan cities—Delhi, Mumbai, Kolkata, Chennai, Bengaluru, Hyderabad, Jaipur, and Lucknow—were selected to estimate the overall air quality status based on (1) total population, (2) previous year’s history of air quality status, (3) urban and industrial development, (4) geographical region, and (5) availability of air quality data (Shrestha et al. 2020). The details of region, population, population density, and number of air quality monitoring stations are shown in Table 1, and locations of the selected metropolitan cities are shown in Fig. 1.

Table 1 Details of selected metropolitan cities of India
Full size table
Fig. 1
figure 1

Source: https://www.google.com/maps/@24.1580341,82.7349726,5z

Locations map of eight metropolitan cities of India.

Full size image

2.2 Selection of air pollutants

Air quality status can be described by calculating the atmospheric concentrations of six criteria pollutants, i.e., PM2.5, PM10, carbon monoxide (CO), sulfur dioxide (SO2), nitrogen dioxide (NO2), and ozone (O3) (EPA 2014). The main source of all criteria pollutants is the combustion of fuel in the automobiles and industries, while O3 is being formed by the reaction of primary pollutants in the presence of sunlight. Therefore, to assess the outcome of lockdown on air quality of Indian metropolitan cities, the air pollutant parameters, viz. PM2.5, PM10, CO, NO2, SO2, and O3, along with meteorological parameters such as temperature and relative humidity during the COVID-19 lockdown period from March 25 to May 3, 2020, were studied and compared with the air quality data of the same period of the previous year 2019.

2.3 Data sources

The data of 24 hourly concentrations of PM2.5, PM10, NO2, SO2 and 8 hourly concentrations of CO and O3 along with meteorological parameters such as temperature and relative humidity for eight metropolitan cities were downloaded from the online portal of CPCB, India (https://app.cpcbccr.com/ccr/#/caaqm-dashboard-all/caaqm-landing) from 80 air quality monitoring stations spread across the selected eight metropolitan cities. Data were collected for the COVID-19 lockdown period of the year 2020 from March 25 to May 3 and the same period of the previous year 2019. To compress the large data sets, average concentrations were calculated for each pollutant at all stations of each metropolitan city for each day.

2.4 Air quality index (AQI)

Air quality index (AQI) is a tool that represents the weighted values of individual air pollutants into a single value. AQI is computed in the subsequent two steps (CPCB 2015):

  1. (1)

    Calculation of subindices (for each pollutant): It is calculated by 24 h average concentrations of PM2.5, PM10, SO2, NO2, NH3, and 8 hourly concentrations of CO and O3, and health breakpoint concentration range. For calculation, the minimum three of the above pollutants are obligatory, out of which one has to be either PM2.5 or PM10.

    The subindex (Ii) for a given pollutant concentration (Cp) is calculated as:

    $$I_{\text{i}} = \, \left[ {\left\{ {\left( {I_{\text{HI}} - \, I_{\text{LO}} } \right)/\left( {B_{\text{HI}} - B_{\text{LO}} } \right)} \right\} \, * \, \left( {C_{\text{p}} - B_{\text{LO}} } \right)} \right] + \, I_{\text{LO}}$$

    where BHI = breakpoint concentration greater or equal to given concentration; BLO = breakpoint concentration smaller or equal to given concentration; IHI = AQI value corresponding to BHI; ILO = AQI value corresponding to BLO; and Cp = pollutant concentration.

  2. (2)

    Aggregation of subindices to get an overall AQI: Maximum operator system aggregates AQI, after the calculation of subindices (Ott 1978).

    $${\text{AQI}} = {\text{Max }}\left( {I_{1} , \, I_{2} , \, I_{3} , \, \ldots , \, I_{n} } \right)$$

Indian air quality index (IND-AQI) is categorized into six categories to represent air quality status and its effects on human health (Table 2).

Table 2 Indian air quality index (IND-AQI) category and range
Full size table

2.5 Prominent pollutants

The “prominent pollutant” is determined every day for all metropolitan cities to measure what pollutant is primarily responsible for the air quality deterioration and identified by the AQI system. The pollutant that has maximum AQI values is identified as a prominent pollutant on that particular day (Cheng et al. 2007).

2.6 Statistical analysis

The Pearson correlation analysis was conducted among the six criteria air pollutants (PM2.5, PM10, CO, NO2, SO2, and O3), the meteorological parameter (temperature and relative humidity), air quality index, and population density to investigate the interrelationship among them. The Pearson correlation coefficient (R) is ranged from − 1 to + 1, where R ≤ .35 corresponds to a weak correlation, R = .36–.67 exhibits moderate correlation, and R ≥ .68 represents strong correlation, and it can be positive or negative (Mason et al. 1983; Weber and Lamb 1970).

3 Results and discussion

3.1 Overview of air pollutants

The average concentrations of six criteria air pollutants monitored in eight metropolitan cities are summarized in Table 3. The average concentration of PM2.5 ranged from 17.03 ± 8.14 µg/m3 (Chennai) to 51.66 ± 13.05 µg/m3 (Lucknow) during the assessment period. The variation in the average concentration of PM2.5 was observed in Delhi (− 47%), Mumbai (+ 1%), Kolkata (− 38%), Chennai (− 48%), Bengaluru (− 52%), Hyderabad (− 23%), Jaipur (− 47%), and Lucknow (− 49%) during the COVID-19 lockdown in comparison with the same period of the previous year 2019 (Table 3; Fig. 2). The data of the average concentration of PM10 were available for six cities except for Chennai and Lucknow, ranged from 47.88 ± 10.82 µg/m3 (Bengaluru) to 98.51 ± 34.03 µg/m3 (Delhi) during the lockdown period. The downturn in the average concentration of PM10 was recorded in Delhi (− 57%), Mumbai (− 27%), Kolkata (− 47%), Bengaluru (− 54%), Hyderabad (− 41%), and Jaipur (− 52%) (Table 3; Fig. 2).

Table 3 Mean concentrations, standard deviation, maximum, minimum, and % variation in six criteria air pollutants, temperature, relative humidity, and air quality index of selected metropolitan cities of India
Full size table
Fig. 2
figure 2

Daily variation and average difference in mean PM2.5, PM10 and CO concentrations during March 25 to May 3, 2020 (COVID-19 lockdown period), and March 25 to May 3, 2019

Full size image

The decrease in the concentration of PM10 along with PM2.5 was observed in a similar pattern in Delhi, Kolkata, Bengaluru, Hyderabad, and Jaipur during the lockdown period; it indicated that a substantial fraction of PM10 is determined by PM2.5 concentration in most aforementioned cities (> 60%) (Wang et al. 2014; Zhang and Cao 2015) (Fig. 2). The vehicular exhaust, industrial emission, dust fallout, and construction-demolition activities are the major source of PM10 and PM2.5 pollution. The reduction in the concentration of PM2.5 and PM10 was due to a ~ 80% drop in vehicular flux (Cyberlab 2020) and shut down of the industrial and constructional activities during the lockdown period. In Wuhan, fossil fuel combustion and vehicles on roads are the main contributors to PM2.5 (Wang et al. 2017). In Barcelona, particulate matter emission is added in the atmosphere by road transport (+ 18%) and non-road transport (+ 21%) (Karamchandani et al. 2017). Both these reports support the findings of our current study that transportation activities are major contributors of particulate matter (PM) in the environment, which were suspended during the lockdown period, which consequently declined the concentration of PM2.5 and PM10 in the environment. The reduction in the concentration of PM2.5 and PM10 in the present study is also supported by a similar study by Sicard et al. (2020). They also observed 36% and 49% reduction in PM2.5 and PM10 concentration, respectively, in Wuhan, China, during the lockdown period in 2020. Huang et al. (2020a, b) exhibited 9% to 34% decrease in PM2.5 concentration in China during the lockdown period in 2020.

The average concentration of CO ranged from .41 ± .09 mg/m3 (Kolkata) to .99 ± .04 mg/m3 (Lucknow) during the lockdown period. A decrease in the average concentration of CO in Delhi (− 43%), Mumbai (− 75%), Kolkata (− 22%), Chennai (− 32%), Bengaluru (− 28%), Hyderabad (− 23%), Jaipur (− 46%), and Lucknow (− 28%) was observed during the lockdown period as compared to the same period of 2019 (Table 3; Fig. 2). The minimum and maximum value of the average concentration of NO2 was ranged 7.92 ± 2.53 µg/m3 (Chennai) and 21.42 ± 4.71 µg/m3 (Hyderabad), respectively. The significant decline in the average concentration of NO2 was exhibited in Delhi (− 59%), Mumbai (− 59%), Kolkata (− 68%), Chennai (− 32%), Bengaluru (− 64%), Hyderabad (− 37%), Jaipur (− 62%), and Lucknow (– 66%) (Table 3; Fig. 3).

Fig. 3
figure 3

Daily variation and average difference in mean NO2, SO2, and O3 concentrations during March 25 to May 3, 2020 (COVID-19 lockdown period), and March 25 to May 3, 2019

Full size image

Carbon monoxide and NO2 both have also exhibited significant decline during the COVID-19 lockdown period (Fig. 2 and 3). The emission of CO and NO2 is primarily associated with the combustion process of fuel like gasoline and diesel in vehicles and industries. COVID-19 lockdown majorly shut down vehicular movement, industrial activities that diminished the concentration of CO and NOx in the environment. The transport sector (road transport: 39% and non-road transport: 8%) largely emits NOx pollutants in Europe (EEA 2019). In China, industrial activities were stopped due to the COVID-19 lockdown dropped 30% and 25% of NO2 and carbon emission, respectively (Isaifan 2020). During the lockdown period, Sicard et al. (2020) noticed a similar reduction in mean NO2 concentration in urban stations of all European cities (– 57%) and traffic stations of Wuhan (– 65%); Huang et al. (2020a, b) reported a more than 60% reduction in NO2 and 13% to 41% reduction in CO concentration in China; Dantas et al. (2020) reported 30.3% to 48.5% diminutions in CO concentration in Rio de Janeiro, Brazil.

The average concentration of sulfur dioxide (SO2) ranged from 5.54 ± 1.79 µg/m3 (Chennai) to 14.38 ± 2.55 µg/m3 (Delhi) during the lockdown period. The variation in the average concentration of SO2 was observed in Delhi (− 32%), Mumbai (− 48%), Kolkata (+ 25%), Chennai (− 22%), Bengaluru (+ 9%), Hyderabad (− 9%), Jaipur (− 9%), and Lucknow (− 16%) (Table 3; Fig. 3).

The average SO2 concentration also exhibited mixed trends during the lockdown and showed a significant decline in six metropolitan cities—Delhi, Mumbai, Chennai, Hyderabad, Jaipur, and Lucknow—while there was an increase in two metropolitan cities—Kolkata and Bengaluru (Fig. 3). Burning coal is a prime source of SO2 emission in the environment, and during the lockdown period, coal power plants might not have been shut down. Therefore, it would be a possible reason for a slight variation in SO2 concentration during the lockdown as compared to the non-lockdown period of the previous year. Huang et al. (2020a, b) also observed 15–42% diminution in SO2 concentration in China during the lockdown in 2020, which is similar to our results.

The average concentration of O3 ranged from 23.69 ± 8.10 µg/m3 (Mumbai) to 49.55 ± 6.85 µg/m3 (Jaipur) during the lockdown period. The variation in the average concentration of O3 was observed in Delhi (− 06%), Mumbai (− 02%), Kolkata (+ 63%), Chennai (+ 51%), Bengaluru (− 25%), Hyderabad (− 29%), Jaipur (− 17%), and Lucknow (− 28%) (Table 3; Fig. 3).

A mixed trend is also observed in O3 concentration during the lockdown, a slight decline in six metropolitan cities (Delhi, Mumbai, Bengaluru, Hyderabad, Jaipur, and Lucknow), while there is an increase in two metropolitan cities (Kolkata, Chennai) (Fig. 3). A slight variation in O3 concentration compared to other criteria pollutants like PM2.5, PM10, CO, and NO2 during the lockdown period might be due to three reasons: (1) Owing to the lockdown of vehicular movement and industrial activities, the reduction in NO concentration may decrease the consumption of ozone molecules (titration, NO + O3 = NO2 + O2), thus increasing the O3 concentration. (2) The concentration of PM10 and PM2.5 decreases drastically during the lockdown period, which may allow extra sunlight through the atmosphere, accelerating more photochemical activities and thereby increasing ozone formation (Dang and Liao 2019; Li et al. 2019). (3) From March to August, Sun migrates in the north that causes augmentation of insolation and temperature in the northern hemisphere and leads to acceleration in ozone production (Gorai et al. 2017). Sicard et al. (2020) also reported similar results. They recorded 24%, 14%, 27%, 2.4%, and 36% rise in O3 concentration at urban stations of Nice, Rome, Turin, Valencia, and Wuhan, respectively, during the COVID-19 lockdown in 2020.

3.2 Overview of meteorological parameters

The average temperature ranged from 24.27 ± .69 °C (Bengaluru) to 31.31 ± 1.04 °C (Mumbai) during the lockdown period. The variation in average temperature was observed in Delhi (− 7%), Mumbai (+ 2%), Kolkata (+ 1%), Chennai (− 3%), Bengaluru (− 10%), Hyderabad (+ 3%), Jaipur (− 7%), and Lucknow (− 5%) during lockdown as compared to the same period of the previous year 2019 (Table 3; Fig. 4). The average relative humidity ranged from 34.09 ± 14.05% (Jaipur) to 73.88 ± 5.25% (Mumbai) during the lockdown period. The variation in average relative humidity was observed in Delhi (+ 36%), Mumbai (+ 6%), Kolkata (− .30%), Chennai (+ 14%), Bengaluru (+ 9%), Hyderabad (+ 9%), Jaipur (+ 53%), and Lucknow (+ 16%) during lockdown as compared to the non-lockdown period of 2019 (Table 3; Fig. 4).

Fig. 4
figure 4

Daily variation and average difference in mean temperature and relative humidity during March 25 to May 3, 2020 (COVID-19 lockdown period), and March 25 to May 3, 2019

Full size image

3.3 Air quality index

In India, the air quality index is categorized among six categories: good (0–50), satisfactory (51–100), moderate (101–200), poor (201–300), very poor (301–400), and severe (401–500) (Table 2). The details of daily air quality index and prominent pollutants of selected eight metropolitan cities are depicted in Table 4. The average air quality index of the eight metropolitan cities is summarized in Table 3. The average air quality index ranged from 48.50 ± 6.53 (Chennai) to 103.43 ± 28.74 (Delhi) during the lockdown period. The reduction in average air quality index was observed in Delhi (− 51%), Mumbai (− 30%), Kolkata (− 19%), Chennai (− 32%), Bengaluru (− 52%), Hyderabad (− 35%), Jaipur (− 42%), and Lucknow (− 52%) during the lockdown as compared to the same period of 2019 (Table 3). In the comparison of daily air quality index of the same period of 2019, out of eight metropolitan cities, three cities like Chennai, Bengaluru and Hyderabad exhibited a consistent reduction throughout the lockdown period, while daily air quality index of Delhi, Mumbai, Kolkata, Jaipur, and Lucknow increased on a few days (Fig. 5).

Table 4 Air quality index and prominent pollutant during March 25 to May 3, 2020 (COVID-19 lockdown period) and March 25 to May 3, 2019
Full size table
Fig. 5
figure 5

Daily variation in percentage and average difference in air quality index during March 25 to May 3, 2020 (COVID-19 lockdown period) and March 25 to May 3, 2019

Full size image

During the lockdown period, the air quality index on most of the days was observed to be good and satisfactory in all the metropolitan cities (Fig. 6) that indicated an improvement in air quality due to a reduction in vehicular and industrial emissions. The air quality index of Delhi was detected to be in a good, satisfactory, and moderate category by 2.5%, 52.50%, and 45% days, respectively, during lockdown days, while it was satisfactory, moderate, poor, and very poor by 2.5%, 40%, 55%, and 2.5% days, respectively, during the same period of 2019. The air quality index of Mumbai was found good and satisfactory by 10% and 90% days, respectively, during the lockdown period, while it was satisfactory and poor by 87.5% and 12.5% days, respectively, in the non-lockdown period of 2019. The air quality index of Kolkata was noticed good (30% days), satisfactory (47.5% days), and moderate (22.5% days) during lockdown period, while it was good (10% days), satisfactory (65% days), moderate (20% days), and poor (5% days) in the same period of 2019. The air quality of Chennai was observed good (35% days) and satisfactory (65% days) in lockdown period, while it was good (10% days), satisfactory (77.5% days), and moderate (12.5% days) in the previous year. The air quality index of Bengaluru was detected good and satisfactory by 17.5% and 82.5% days, respectively, during lockdown days, while it was satisfactory and moderate by 12.5% and 87.5% days, respectively, in the non-lockdown period of 2019. The air quality index of Hyderabad was observed to be good and satisfactory by 12.5% and 87.5% days, respectively, during lockdown days, while it was satisfactory and moderate by 57.5% and 42.5% days, respectively, on the same dates of the previous year. The air quality of Jaipur was good (5% days), satisfactory (82.5% days), and moderate (12.5% days) in lockdown period, while it was satisfactory (17.5% days), moderate (77.5% days), poor (2.5% days), and very poor (2.5% days) during the same period of the year 2019. The air quality index of Lucknow was noticed satisfactory (62.5% days), moderate (35% days), and poor (2.5%) during the lockdown period, while it was satisfactory (5% days) and moderate (42.5% days), poor (45%), and very poor (7.5% days) in the same period of the previous year.

Fig. 6
figure 6

Categorywise variation in air quality index during March 25 to May 3, 2020 (COVID-19 lockdown period), and March 25 to May 3, 2019

Full size image

The order of air quality index during lockdown for all metropolitan cities was satisfactory > moderate > good > poor, while it was moderate > satisfactory > poor > good > very poor during the same period of the previous year 2019.

3.4 Prominent pollutants

Table 4 demonstrates the prominent pollutants on all days during the lockdown in all metropolitan cities, and Fig. 7 illustrates the frequency of prominent pollutants in percentage during the study period.

Fig. 7
figure 7

Frequency of prominent pollutant during March 25 to May 3, 2020 (COVID-19 lockdown period) and March 25 to May 3, 2019

Full size image

Throughout the lockdown period, it was observed that ozone was the most prominent air pollutant in all metropolitan cities. The frequency of ozone as a prominent pollutant for Delhi, Mumbai, Kolkata, Chennai, Bengaluru, Hyderabad, Jaipur, and Lucknow was 75%, 42.50%, 97.50%, 80%, 80%, 32.50%, 62.50%, and 25%, respectively, while it was 25%, 0%, 7.50%, 10%, 62.50%, 32.50%, 27.50%, and 2.5%, respectively, during the same period of 2019.

During the lockdown period, the second and third most frequent prominent pollutants were PM10 and PM2.5, respectively, while PM2.5 and PM10 were the second and third most prominent pollutants, respectively, during 2019. The frequency of PM10 was observed in Delhi (55%), Mumbai (100%), Kolkata (37.5%), Bengaluru (77.5%), Hyderabad (92.5%), and Jaipur (77.5%) during the lockdown period. The frequency of PM2.5 was observed in Delhi (65%), Mumbai (0%), Kolkata (37.5%), Chennai (55%), Bengaluru (72.5%), Hyderabad (57.5%), Jaipur (2.5%), and Lucknow (100%) during the lockdown period.

Carbon monoxide was the fourth prominent pollutant during the lockdown period. The frequency of carbon monoxide for Delhi, Mumbai, Kolkata, Chennai, Bengaluru, Hyderabad, Jaipur, and Lucknow during the lockdown was 10%, 22.5%, 2.5%, 75%, 60%, 5%, 2.5%, and 17.5%, respectively, while it was 0%, 70%, 2.50%, 85%, 10%, 0%, 0%, and 15%, respectively, during non-lockdown period of the previous year. The occurrence of NO2 and SO2 as a prominent pollutant during the lockdown as well as the non-lockdown period of 2019 was very less frequent.

The overall frequency order of prominent pollutants during lockdown was O3 > PM10 > PM2.5 > CO > NO2 > SO2, while it was PM2.5 > PM10 > CO > O3 > NO2 > SO2 during the same period of 2019 in all metropolitan cities. A diminution in the concentration of PM2.5 and PM10 during the lockdown period could also induce the evolution reaction of O3 by allowing more solar radiation in the troposphere (Heuss et al. 2003; Li et al. 2017; Liu et al. 2013; Murphy et al. 2007; Wolff et al. 2013) that may be a possible reason for O3 to become a prominent pollutant instead of PM2.5 and PM10.

3.5 Correlation among criteria air pollutants, temperature, relative humidity, air quality index, and population density

Pearson correlation coefficient (R) was computed among the six criteria air pollutants (PM2.5, PM10, CO, NO2, SO2, and O3), meteorological parameters (temperature and relative humidity) and air quality index for all the eight metropolitan cities during the lockdown and the same period of the previous year 2019 (Tables 5 and 6).

Table 5 Correlation analysis of six criteria air pollutants, temperature, relative humidity, and air quality index of selected metropolitan cities of India during March 25 to May 3, 2020 (COVID-19 lockdown period)
Full size table
Table 6 Correlation analysis of six criteria air pollutants, temperature, relative humidity, and air quality index of selected metropolitan cities of India during March 25 to May 3, 2019
Full size table

During the lockdown period, PM2.5 and PM10 were strongly correlated (R ≥ .68) in Delhi, Mumbai, Kolkata, Hyderabad, and Jaipur, while they were strongly correlated (R ≥ .68) in Delhi, Mumbai, Hyderabad, and Jaipur during the same period of the year 2019. The strong correlation between PM2.5 and PM10 during the lockdown period and the same period of 2019 clarified that PM2.5 contributes to the formation of PM10 (Wang et al. 2014, 2017; Zhang and Cao 2015).

A strong correlation (R ≥ .68) between PM2.5 and NO2 in Delhi, Kolkata, and Lucknow, and PM10 and temperature in Delhi, Kolkata, and Jaipur was observed during COVID-19 lockdown period, while CO and NO2 in Delhi, Mumbai, Kolkata, and Hyderabad, and CO and SO2 in Delhi, Kolkata, and Hyderabad were strongly correlated (R ≥ .68) during the same period of the previous year.

Air quality index and PM2.5 were strongly correlated (R ≥ .68) in Delhi, Mumbai, Kolkata, Chennai, Hyderabad, Jaipur, and Lucknow during the lockdown period. It was also strongly correlated (R ≥ .68) in Delhi, Mumbai, Kolkata, Chennai, Hyderabad, Jaipur, and Lucknow during the same period of the year 2019. The correlation between air quality index and PM10 was found strong (R ≥ .68) in Delhi, Mumbai, Kolkata, Hyderabad, and Jaipur during the lockdown period and it was also strong (R ≥ .68) in Delhi, Mumbai, Hyderabad, and Jaipur during the same period of the previous year. The strong positive correlation between air quality index and PM 2.5, and AQI and PM10 during the lockdown period and non-lockdown period of 2019 indicated that PM2.5 followed by PM10 dominantly influences the air quality. Yan et al. (2016) also observed that PM2.5 was chiefly degrading the air quality in Beijing and China.

During the lockdown period, the air quality index was strongly correlated (R ≥ .68) with CO in Delhi, Mumbai, and Kolkata. A weak correlation (R ≤ .35) between SO2 and PM2.5 in Mumbai, Bengaluru, Hyderabad, and Jaipur; SO2 and PM10 in Mumbai, Bengaluru, and Jaipur; SO2 and CO in Bengaluru, Jaipur, and Lucknow; SO2 and air quality index in Chennai, Bengaluru, and Jaipur; ozone (O3) and PM2.5 in Delhi, Mumbai, and Jaipur; O3 and PM10 in Mumbai, Kolkata, and Hyderabad; O3 and air quality index in Mumbai, Chennai, Bengaluru, Hyderabad, and Jaipur; CO and NO2 in Mumbai, Chennai, Hyderabad, and Jaipur; CO and temperature in Mumbai, Bengaluru, Hyderabad, Jaipur, and Lucknow were observed during the lockdown period.

A negative strong correlation (R ≥ − .68) was observed between temperature and relative humidity in Delhi, Kolkata, and Jaipur and a negative moderate correlation (R = − .36 to − .67) between temperature and PM2.5 in Mumbai, Chennai, and Bengaluru; relative humidity and PM2.5 in Delhi, Bengaluru, Hyderabad, and Jaipur; relative humidity and PM10 in Delhi, Hyderabad, and Jaipur, and O3 and NO2 in Mumbai, Hyderabad, Jaipur, and Lucknow were observed during the lockdown period.

During the lockdown period, negative weak correlation (R ≤ − .35) was exhibited between CO and PM2.5 in Bengaluru, Hyderabad, and Lucknow; CO and O3 in Mumbai, Bengaluru, and Jaipur; relative humidity and CO in Mumbai, Bengaluru, and Jaipur; relative humidity and SO2 in Mumbai, Chennai, Bengaluru, and Lucknow; relative humidity and O3 in Chennai, Bengaluru, Jaipur, and Lucknow; air quality index and NO2 in Mumbai, Chennai, Bengaluru and Hyderabad; air quality index and temperature in Chennai, Bengaluru and Hyderabad; and air quality index and relative humidity in Mumbai, Chennai, Bengaluru, and Lucknow.

Pearson correlation coefficient (R) was also computed between air quality index and population density. During the lockdown, a negative moderate correlation (R = − .36 to − .67) was observed between air quality index and population density, while a negative strong correlation was observed (R ≥ − .68) between air quality index and population density during the same period of the year 2019.

4 Conclusion

The COVID-19 lockdown saved tens of thousands lives and controlled the transmission of SARS-CoV-2 primarily in many of the countries. Subsequently, it also restored the air quality of the world in a very impressive way. Major Indian metropolitan cities also re-established the degraded air quality during the COVID-19 lockdown period. It is concluded that out of six criteria pollutants, the concentration of four criteria pollutants like PM2.5, PM10, CO, and NO2 drastically decreased, while the concentration of SO2 and O3 declined slightly in most of the metropolitan cities during the COVID-19 lockdown period. The order of air quality index during lockdown for all metropolitan cities was satisfactory > moderate > good > poor, while it was moderate > satisfactory > poor > good > very poor during the previous year. It is concluded that the frequency order of prominent pollutants during lockdown was O3 > PM10 > PM2.5 > CO > NO2 > SO2, while it was PM2.5 > PM10 > CO > O3 > NO2 > SO2 during the same period of 2019 in all metropolitan cities. Overall air quality was improved during the lockdown period in Indian metropolitan cities. Further, based on the results of this study it is suggested that India should implement the lockdown once in a month to repair the damage of the air quality caused by heavy vehicular and industrial emissions.