Impact of COVID-19 lockdown and meteorology on the air quality of Srinagar city: A temperate climatic region in Kashmir Himalayas

The deadly transmission of the coronavirus forced all countries to implement lockdowns to restrict the transmission of this highly infectious disease. As a result of these lockdowns and restrictions, many urban centers have seen a positive impact on air quality with a significant reduction in air pollution. Therefore, in this study, the impact of COVID-19 lockdown vis-a-vis meteorological parameters on the ambient air quality of Srinagar city was examined. In this regard, we have evaluated the temporal variation of six different key air pollutants (PM10, PM2.5, SO2, NO2, O3, and NH3) along with meteorological parameters (relative humidity, rainfall, temperature, wind speed, and wind direction). The duration of the study was divided into three periods: Before Lockdown(BLD), Lockdown (LD), and Partial Lockdown(PLD). Daily average data for all the parameters was accessed from one of the real-time continuous monitoring stations of the central pollution control board (CPCB) at Rajbagh Srinagar. Some air pollutants have decreased, according to the results, while others have increased. The air quality index (AQI) decreases overall by 6.15 percent compared to before lockdown, and it never exceeds the "moderate" category. The AQI was in the following order for both lockdown and pre-lockdown periods: satisfactory > moderate > good. However, for partial lockdown, it was moderate > satisfactory > good. It was observed that the maximum decrease was seen in the concentration of NO2, NH3 with 75.11% and 69.18%. A modest decrease was observed in PM10 at 3.8%. While SO2 and O3 had an upward trend of 85.82% and 48.74%, The NO2 to SO2 ratio reveals that the emissions of NO2 have substantially decreased due to the complete restriction of transport systems. From principal component analysis for all three study periods, PM10 and PM2.5 were combined into a single component, inferring their shared behavior and source of origin. SO2 and O3 demonstrated identical behavior during the lockdown and partial lockdown periods of study. According to the findings of the study, it is beneficial for the government, environmentalists, and policymakers to impose rigorous lockdown measures, particularly during extreme air pollution events, in order to reduce the damage caused by automotive and industrial emissions.


Introduction
With the emergence of rapid globalization and urbanization, megacities in developing nations are facing severe health issues due to ambient air pollution ( Agarwal et al., 2020 ;Bherwani et al., 2021 ). Poor air quality has become a global concern; according to studies, over 4.2 million people die prematurely each year as a result of poor air quality exposure, which causes chronic respiratory disorders, cardiovascular diseases, asthma, premature mortality, and lung cancer ( Gautam, 2020 ;Srivastava et al., 2021 ;Kumari and Toshniwal, 2020 ;Praveen et al. , 2022 ). Urban air quality is a serious issue that is attracting attention in both developing and developed countries ( Sheikh and Najar, 2018 ). Due to the emission of air pollutants from transportation, industries, thermal power plants, construction and commercial activities, open burn-ing of municipal waste and agricultural residuals over many years, air quality in many Indian urban centers has deteriorated and is one of the most difficult issues to address Verma and Kamyotra, 2021 ). Controlling anthropogenic emissions can prevent serious air pollution events and the health hazards caused by air pollutants. An appreciable reduction was seen by some Indian studies associated with COVID-19 lockdown restrictions ( Maji et al., 2021 ).
The World Health Organization (WHO) declared SARS-COVID-19 a "global pandemic " on March 12, 2020. The virus originated in Wuhan, China. Ranjan et al., 2021 ;Hu et al., 2021 ;Jain and Sharma, 2020 ;Gouda et al., 2021 ). All affected countries enforce lockdowns and preventive measures to stop the virus's spread Gouda et al., 2021 ). In India, the first incidence of infection was reported in Kerala on January 30, 2020 ( Gautam et al., 2020 ;Chelani and Gautam 2022 ) country, including the Union Territory of Jammu and Kashmir, was under complete lockdown from March 25, up to the first week of June 2020. During the lockdown, there was a strict ban on all nonessential activities, including the movement of people, transportation (road, rail, and air), construction, and industries ( Sathe et al., 2021 ;Ravindra et al., 2021 ). which, according to studies, significantly reduces the concentration of all air pollutants caused by anthropogenic activities, especially those related to transportation ( Navinya et al., 2020 ;Naqvi et al., 2022 ;Wang et al., 2020 ;Gouda et al., 2021 ). As of December 2020, India recorded a total of 9.88 million cases and 1.62 million deaths (Covid19india.org; https://www.mygovin/covid-19). By the end of 2020 and starting of 2021 COVID-19 cases in India have seen a sharp dip with lowest reported cases of 8660 on 08 February 2021 and Union Territory of Jammu and Kashmir recorded 57 cases ( Covid19.india.healthdata.org/India ). As cases declined, the centre and state governments relaxed restrictions, despite new variants being identified. The picture changed drastically when several parts of the world, especially Europe and India, began experiencing a massive surge of COVID-19 cases and deaths ( Kar et al., 2021 ; www.nationalgeographic.com/science/article). From April 2021, Maharashtra, Punjab, and Tamil Nadu imposed weekend and night curfews. On April 28, 2021, Jammu and Kashmir reported 3023 new cases and 30 deaths (https://Indiatoday.in/India/jammuandkashmir). In response, UT imposed a lockdown from April 29 to May 30, 2021. Due to these restrictions once again anthropogenic emissions of air pollutants are believed to have decreased throughout India, including the Jammu and Kashmir . Several studies have been published on the impacts of COVID-19 on air pollution levels in India ( Kumari and Toshniwal, 2020 ;Gouda et al., 2021 ;Chelani and Gautam, 2021 ;Mor et al., 2021 ;Verma and Kamyotra, 2020;Rathore et al., 2021 ). According to the findings of the review, the COVID-19 pandemic shutdown has led to a reduction in pollution all across India. The aim of the study is (i) to assess the impact of the COVID-19 lockdown on the concentration of key air pollutants like PM 2.5 , PM 10 , NO 2 , SO 2 , O 3 and NH 3 over the city of Srinagar during three different phases: the before-lockdown (April 14 to May 4), the lockdown (5 to 25 May) and the partial lockdown (26 May to 14 June 2021) keeping meteorological parameters in view. (ii) The percentage change in key air pollutants during lockdown with respect to before lockdown is evaluated. (iii) The trend in the concentration of air pollutants has been evaluated in order to study phase-to-phase changes in air pollutant levels during the entire period of study. (iv) A principal component analysis was also performed to understand and identify similar behaving pollutants.

Site description
Srinagar, the summer capital of Jammu and Kashmir UT, is the largest city among all the Himalayan urban centers, situated on the river Jhelum ( Bhat et al., 2021 ). The city is located within the geographical coordinates of 33°18' to 34°45´N & 74°56 to 75°79´E with a mean elevation of 1585 m ( Sheikh and Najar, 2018 ), as depicted in Fig. 1 . The valley of Kashmir is enclosed by mountain ranges from all sides, like the Pir Panjal range from the Southwest, the Great Himalayas from the Northeast, and the Shamasbari range from the North. Kashmir has a temperate climate which differs significantly from the rest of the country with well-defined four seasons based on temperature and precipitation pattern: Spring (March-May), Summer (June-August), Autumn (September-November) and Winter (December-February). ( Zaz et al., 2019 ).
The city is the largest urban part of the Union Territory and is home to about 1.2 million inhabitants ( Census of India 2011 ). Unlike other urban cities in India, there is an absence of large-scale industrial activity ( Huma et al., 2016 ). Apart from vehicular emissions, road dust, biomass burning, and coal combustion are the main sources of particulate pollution in the city ( Hakim et al., 2018 ).

Data collection
In this study, we have considered the three periods of study, such as before-lockdown (14 April to 04 May 2021), lockdown (05 to 25 May 2021), and partial-lockdown (26 May to 14 June 2021), to assess the impact of lockdown on the ambient concentration of atmospheric pollutants like PM 2.5 , PM 10 , NO 2 , SO 2 , O 3 , and NH 3 over the city of Srinagar. These air pollutants are being monitored by the Central Pollution Control Board (CPCB) over 573 Continuous Ambient Air Quality Monitoring Stations (CAAQMS) in 240 Indian cities with the finest temporal resolution of 15 min ( Navinya et al., 2020 ). For this study, we procured ambient air quality and meteorological parameters for one of the CAAQMS located in Rajbagh, Srinagar from the CPCB website ( https://app.cpcbccr.com/ ) and data for rainfall was retrieved from ( https://www.wunderground.com/weather/in/srinagar ).

Air quality index (AQI)
One of the most effective ways to express the status of air quality in an area is the Air Quality Index (AQI), which represents the weighted value of an individual air pollutant into a single value matrix with its possible environmental and human health impacts ( Sathe et al., 2021 ). The Indian Air Quality Index (IAQI) is categorized into six categories and is calculated by a weighted average of 8 criteria pollutants with a single value ( Verma and Kamyotra, 2021 ). The Central Pollution Control Board provides real-time NAQI of 231 CAAQMS which can be retrieved from the CPCB website ( https://app.cpcbccr.com/AQI-India/ ). Table S1 shows the different categories of AQI with range and color coding as per CPCB ( Sheikh and Najar, 2018 ). In this study, we have analyzed the AQI of Srinagar city for before-lockdown, lockdown, and partial-lockdown periods.

Data analysis
The impact of COVID-19 lockdown on air pollution levels in Srinagar city was inspected for three different periods. A comparison was made for the time period from 14 April to 04 May 2021 (Before-lockdown) with 5 to 25 May (Lockdown) and 26 May to 14 June (Partial-lockdown). Time series data was analyzed to see the effect of meteorological parameters like relative humidity, rainfall, temperature, and wind speed on air pollution.

Percentage change in concentration of pollutants
The percentage change between before-lockdown and lockdown was computed. The relative change in concentration in lockdown with respect to before-lockdown is computed using the Eq. (1) below: Where, Conc. BLD, is the pollutant concentration average over the Before-lockdown period (April 14 to May 04-2021) and Conc. LD, is the pollutant concentration averaged over the lockdown period (May 5 to 25, 2021).

Principal component analysis
To identify the source of the pollutants, a statistical technique called principal component analysis (PCA) is applied using SPSS software. With the help of PCA, it is possible to simplify a data set having a large number of variables to a new small data set having fewer variables (i.e., Principal Components), which represents the variability of the original dataset having linear combinations between them. Principal component analysis reduces the number of dimensions of variables without losing much information .

Variations in the criteria air pollutants
The PM 2.5 , PM 10 , NO 2 , SO 2 , O 3 and NH 3 average values were compared to pre-lockdown conditions. PM 10 , NO 2 , and NH 3 decreased during lockdown, whereas overall they increased during partial lockdown. Throughout this period, SO 2 and O 3 have been increasing. NO 2 concentrations decreased significantly during lockdown and partial lockdown compared to before-lockdown levels. Fig. 2 shows the variability in the average concentration of pollutants. PM 2.5 averages concentration were 31.0 g/m 3 , 31.44 g/m 3 , and 47.86 g/m 3 before, during, and partiallockdown. Fig. 2 shows no violation of NAAQS (60 g/m 3 ) for PM 2.5 before and after lockdown. However, it has been violated for 5 days of partial lockdown. Table S3(c). PM 10 concentrations fell during lockdown  and then increased after partial lockdown relaxation ( Fig. 2 ). PM 10 averages concentration were 70.35 g/m 3 , 67.64 g/m 3 and 99.49 g/m 3 before, lockdown, and partial lockdown. Hakim et al. (2018) identified bio-fuel burning and transportation as major sources of particulate matter in Srinagar. Hence, a reduction in traffic due to the COVID-19 lockdown has reduced average PM 10 concentrations and also stabilized the increasing trend of PM 2.5 . Fig. 3 (Boxplots) shows the variation in daily average concentrations of NO 2 , SO 2 , O 3 , and NH 3 for three different periods. Because major emission sources like transportation, construction, and agricultural activities were allowed to resume operations after being restricted during the lockdown, nearly all of the pollutants, including PM 2.5 , PM 10 , SO 2 , NH 3 , and O 3 , showed a significant increase during the partial period of lockdown. It has been reported that automobile emissions have been identified as the primary sources of atmospheric pollution in Srinagar ( M.A. Bhat and Gaga, 2022 ;Savio et al., 2022 ).
NO 2 ranged between 2.84 to 33.32 g/m 3 (before-lockdown), 5.45 to 5.47 g/m 3 (lockdown) and 0.72 to 9.59 g/m 3 (Partial-lockdown). SO 2 and O 3 kept on increasing with 1.34, 2.49, 6.29 g/m 3 and 5.54, 8.24, 11.43 g/m 3 respectively during all three period of study ( Fig. 3 ,  D&E). A substantial increase in SO 2 could be due to domestic biomass burning, including cooking, as people stayed at home during lockdown. Contrarily, NH 3 decreased during lockdown compared to before lockdown and then increased during partial lockdown. The overall air quality index dropped during lockdown and subsequently increased after limitations were relaxed during partial lockdown Fig. S1.

Percentage change in air pollutants
To explore the real impact of lockdown like conditions, we have also analyzed the percentage change in air pollutants during lockdown with respect to before-lockdown. The overall decrease during lockdown can be understood by Fig. 4 . There was a significant reduction in PM 10 , NO 2 , NH 3 , and AQI, with percentage changes of − 3.85%, − 75.11%, − 69.18%, and − 6.15% respectively, during lockdown ( Table 1 ). The maximum reduction was seen in NO 2 and NH 3 . in comparison to a study conducted in the national capital of Delhi during lockdown, which reported a decrease of 47% in PM 2.5 and 57% in NO 2 concentrations. A similar trend was also reported for NO 2 and NH 3 Mor et al., 2021 ). From Table 1 , it is found that the maximum decrease in NO 2 concentration was reported at Mumbai ( − 75%) followed by Kolkata ( − 69%) and the minimum decrease was reported at Singrauli ( − 12.5%). As NO 2 and NH 3 are primarily associated with the combustion of fuel like gasoline and diesel in vehicles and industries Rathore et al., 2021 ). This maximum reduction in NO 2 and NH 3 could be attributed to the sudden shutdown of the transportation sector, except for emergency services, as transport is the major source of NO 2 emissions ( Hakim et al., 2018 ). An increase in the percentage change of + 1.32%, + 85.8% and + 48.74% was recorded for PM 2.5 , SO 2 and O 3 respectively. The increase in SO 2 concentration was also reported at Singrauli (11.8%) and Chandigarh (1.4%) which could be attributed to the non-ceasing of coal-based industrial activities ( Kumari and Toshniwa., 2020;Mor et al., 2021 ). O 3 showed an increment of + 48.74% in its average concentration during lockdown as compared to before-lockdown, which can be supported by other studies ( Jain and Sharma, 2020 ;Singh and Chuhan, 2020 ). An overall increase in O 3 concentration throughout the period could be due to a continuous increase in solar radiation as the study proceeded. Hence, by the action of sunlight, oxygen from NO 2 becomes free, which can react with O 2 to form O 3 ( Kumari and Toshniwal, 2020 ;Mor et al., 2021 ). Illustrating that natural mechanisms like deposition, diffusion, and dispersion of airborne particles and gasses are sufficient to remove contaminants from the atmosphere when anthropogenic emissions are controlled ( Kolluru et al., 2021 ). However, society must return to its pre-pandemic normal state, which began with a limited easing of travel and economic restrictions known as "unlock. "

NO 2 /SO 2 ratio
The nitrogen dioxide to sulfur dioxide ration in the ambient air is usually calculated to see the relative contribution of mobile emission sources (vehicular emissions) and stationary emissions (industrial activities). The NO 2 /SO 2 ratio calculated over the study period was to understand the relative contribution of mobile and stationary sources for before-lockdown, lockdown, and partial lockdown periods, which was found to be in the order of 17.01 > 5.17 > 0.79 respectively. A comparison was also made across different cities, like Delhi, Noida, Jind, and Srinagar, to evaluate how NO 2 and SO 2 emissions were affected by lockdowns. Since transportation services were completely suspended during the lockdown periods, it is evident from Fig. 5 that the contribution of mobile emission sources, such as buses, trucks, vehicles, and motorcycles, has significantly decreased compared to stationary sources, such as industrial and domestic biomass burning, including cooking and heating activities, for all the cities ( Nguyen et al., 2021 ).

Impact of meteorological parameters
When it comes to the dispersion of air pollutants, meteorological factors including wind speed, relative humidity, temperature, and precipitation are important. Hu et al., 2021 ). Fig. 6 . presents time series analysis of all the key air pollutants with meteorological parameters like wind speed, relative humidity, and precipitation (RF) for whole period ( Jain and Sharma, 2020 ). Precipitation, wind speed, and temperature are the important parameters that favor the dispersion of pollutants . PM 10 and PM 2.5 followed a completely inverse relationship with relative humidity and rainfall, as with an increase in the peaks of rainfall and relative humidity, the peaks in PM 10 and PM 2.5 decreased. Fig. 6 . Rainfall plays a key role in washing away particle pollution . Pre-lockdown experienced higher rain episodes (0.18 inches/day) than lockdown and partial lockdown (0.09 and 0.03 inches/day) ( Table.S3 a, b, c), which could explain the somewhat lower PM 2.5 values before-lockdown compared to lockdown and partial lockdown. Fig. 6 . shows that low relative humidity (60-70%) encouraged pollutant accumulation with PM 10 peaking at  100 g/m 3 and PM2.5 at 60 g/m 3 . Temperature has shown an increasing trend from before-lockdown to partial-lockdown with mean values of 12.83 °C, 15.18 °C and 21.03 °C.

Impact on air quality index
The Central Pollution Control Board (CPCB) provides a daily bulletin of the average air quality index on its website for more than 100 cities across India. We have retrieved the air quality index data for Rajbagh Srinagar station for the whole of the period ( https://app.cpcbcr.com/AQI_India ). The purpose of the air quality index is to communicate the status of the air quality index (AQI) to the public so that they can take early and necessary steps to protect their health and the environment ( Hu et al., 2021 ;Verma and Kamyotra, 2021 ). Fig. S1. depicts the overall air quality index decreasing during lockdown and subsequently increasing during the partial-lockdown phase. whereas Fig. 7 . (a, b, c) shows the distribution of different categories of AQI for three different periods of study ( Sathe et al., 2021 ). The AQI before-lockdown is mainly distributed as "Good " (25%), "Satisfactory " (65%), and "Moderate " (10%). The air quality index never touched the 'Poor' category throughout the period, which can be seen. The air quality improved during lockdown with respect to before-lockdown, which can be seen from the percentage contribution of different categories ( Fig. 7. b), such as 25% (Good), 70% (satisfactory), and only 5% of the days with breathing discomfort for people with lung diseases, i.e., moderate category. However, with the partial relaxation on lockdown having increased, the air quality index has increased, which could be understood from an increase of 5% to 47% of the contribution by the moderate category, and a decline is seen in the minimal impact category (Good) from 25% to 11%.

Principal component analysis
Data for all three periods was used to perform PCA on air pollutants Table 2 . Before-lockdown had the highest cumulative variance of 91.9 percent of all three periods , with PC1 having high NO 2 , O 3, and NH 3 loadings indicating traffic pollution. PC2 has substantial PM 10 and PM 2.5 loading . PC3 showed a variance of 19.4% with SO 2 loadings associated with fossil fuel and biomass burning (Before-lockdown). During the lockdown period, PCA extracted three components which explain a cumulative variance of 88.6%, where PC1 is significantly loaded with PM 10 and PM 2.5 explaining the variance of 33.3%. PC2 during lockdown had a significant load of SO 2 and O 3 . Whereas PC3 was loaded with NO 2 and NH 3 representing their common source.
For the study period of partial-lockdown, PCA extracted three factors with a cumulative variance of 87.58%. PC1 explains 33.3% of the variation in PM 10 and PM 2.5 loading. PC2 with significant SO 2 NO 2 and NH 3 levels indicates a significant source of fossil fuel combustion . PC3 showed about 23.5% of variance with significant factor loading of O 3 showing its photochemical production in the atmosphere. For all three study periods, PM 10 and PM 2.5 were combined into a single component to infer their shared behavior and source of origin. But during the lockdown and partial lockdown periods of the study, SO 2 and O 3 behaved the same.

Discussions
An attempt is made to disentangle the effects of lockdown and meteorology on the temporal variation of key air pollutants in an urban city in the northwestern Himalayas. The Kashmir valley is a "lacustrine basin " of the intermountain depression located between the Greater and Lesser Himalayas, distinguished by numerous aquatic ecosystems of immense environmental and economic significance. Kashmir is a highaltitude destination with a pristine environment that has become popular with vacationers. In contrast, these places are perceived to be ecologically deteriorating due to ever-increasing travellers and related pollution, as a result of which Srinagar's vehicular industry has also grown tremendously. Subsequently, the rapid increase in the vehicle population has resulted in huge amounts of pollution ( M.A. Bhat and Gaga, 2022 ). The topographical position of the landscape also plays an important role, which may lead to long-distance movement of pollutants ( M.A. Bhat and Gaga, 2022 ). Therefore, the area for the current study was chosen based on its different climatic representativeness and pollution sources, as Kashmir has a different climate with respect to the rest of India.
The concentrations of NO 2 and NH 3 were much lower during lockdown than before lockdown, which could be attributable to the sudden shutdown of the transportation sector during the lockdown period. PM 10 and PM 2.5 , on the other hand, have not changed significantly. However, SO 2 and O 3 show the greatest increases. The AQI was in the following order for both lockdown and pre-lockdown periods: satisfactory > moderate > good. However, for partial lockdown, it was moderate > satisfactory > good.
The review shows that the COVID-19 pandemic shutdown has resulted in reduced pollution across India. For example, a study conducted by Jain and Sharma, 2020 in Mumbai found that PM 2.5 , PM 10 , NO 2 , and O 3 decreased by 33%, 47%, 75%, and 8%, respectively. Singrauli, Madhya Pradesh, showed an increase in PM 2.5 , PM 10 , SO 2 , and O 3 but a decrease in NO 2 ( Kumari and Toshniwal, 2020 ). During the lockdown, PM 2.5 ( − 37.7%), NO 2 ( − 49%), SO 2 ( − 10%), and O 3 ( − 15.6%) dropped due to fewer industrial and transportation operations in Bangalore ( Gouda et al., 2021 ). According to an analysis ( Chelani and Gautam, 2021 ), lockdown reduced PM2.5 ( − 47%) and NO2 ( − 57%) concentrations over Delhi. All major pollutants except O 3 and SO 2 decreased during lockdown compared to non-lockdown in 2020 over Chandigarh . PM 10, PM 2.5 , and NO 2 levels were reduced by more than 50%. Lockdown raised O 3 ( Verma and Kamyotra, 2020 ). Almost all pollutants have dropped 9 to 41%. ( Rathore et al., 2021 ). The study's findings will assist the scientific community and policymakers in identifying and developing sustainable solutions to reduce the high levels of pollution in urban centers. The study includes both strengths and limitations. To our knowledge, this is the first study of its kind to examine the impact of the COVID-19 lockodown on air quality in a region of India with a temperate climate. However, the unavailability of data on air quality for previous years, which would have allowed for more accurate comparisons and drawn more conclusive conclusions, is the most significant limitation of the study. Future research will be directed towards the seasonal source apportionment studies of ambient air quality and in order to understand the pattern of seasonal changes in key air pollutants and to identify their principal sources.

Conclusions
The rapid shutdown of the transport sector during the lockdown has reduced NO 2 and NH 3 concentrations. Whereas, PM 10 and PM 2.5 have not shown any significant change. However, SO 2 and O 3 show maximum increments. The O 3 steadily increased as the lockdown stages progressed, which is consistent with the observed decrease in NO 2 and increasing ground solar radiation. The NO 2 /SO 2 ratio indicates that mobile sources of air pollution have contributed less over time due to restrictions on vehicular activity. The overall air quality index was also reduced during lockdown and then increased for partial lockdown. PM 10 and PM 2.5 were merged for all three study periods to infer comparable behaviors and sources. During lockdown and partial lockdown, SO 2 and O 2 behaved similarly. The study suggests that the government, environmentalists, and policymakers should impose rigorous shutdown measures, especially during severe air pollution events.

Funding
This work is not funded by any government or private agency.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.