Abstract

The COVID-19 pandemic has significantly changed the air pollution of the world. The present study investigated the temporal and spatial variability in air quality in Xi’an, China, and its relationship with meteorological parameters during and before the COVID-19 pandemic. The outcomes of this study indicated that air pollutants, PM_2.5, NO₂, PM₁₀, CO, and SO₂ are likely to decrease during winter (25%, 50%, 30%, 40%, and 35%) to spring (30%, 55%, 38%, 50%, and 40%) and summer (40%, 58%, 60%, 55%, and 47%), respectively. However, the concentration of O₃-8h increased by 40%, 55%, and 65% during winter, spring, and summer, respectively. The values of the air quality index decreased during the COVID-19 period. Furthermore, significant positive trends were reported in PM_2.5, NO₂, PM₁₀, O₃, and SO₂, and no notable trends in CO during the COVID-19 pandemic. Both during and before the COVID-19 period, PM₁₀, NO₂, PM_2.5, CO, and SO₂ showed a negative correlation with the temperature and a moderately positive significant correlation between O₃-8h and temperature. The findings of this study would help understand the air pollution circumstances in Xi’an before and during the COVID-19 period and offer helpful information regarding the implications of different air pollution control strategies.

Keywords

Spatiotemporal Analysis, Air Quality Index, Meteorological Parameters, COVID-19

Share and Cite:

1. Introduction

In the 21st century, modernization and industrialization are at the highest point globally, and the risk of air pollution is increasing. Air pollution is one of the biggest threats to public health globally. The low-quality air is causing severe health issues through respiratory diseases, cardiovascular, cancer, intelligence quotient loss, premature mortality, asthma, heart attacks, chronic bronchitis, lung diseases, and shortness of breath (Alqasemi et al., 2021; Rahman et al., 2021). Besides, it has been estimated that 90% of the world’s population lives in areas where air pollution levels are above the safe threshold level for human health (WHO, 2016) and that these outcomes are causing about seven million deaths each year and shortening the average life expectancy by about two years (Wyche et al., 2021).

Due to rapid industrialization and uncontrolled population, automobile traffic pressure and the cities’ unscrupulous development are increasing daily, resulting in massive amounts of environmental air pollution (Rahman et al., 2021). There are five major air pollutants such as ozone (O₃), nitrogen dioxide (NO₂), sulfur dioxide (SO₂), and particulate matter i.e. PM_2.5, and PM₁₀, which help in the determination of air quality (Kumari & Toshniwal, 2020). Air quality is severely affected by anthropogenic emissions, and the primary sources of anthropogenic particulate matter emissions are various industries, the transportation sector, and households (Khorsandi et al., 2021). In addition, weather factors are the main determinants of air quality (Kayes et al., 2019; Manju et al., 2018) and are closely interconnected through atmospheric dynamic processes and chemical reactions (Radaideh, 2017). Therefore, air pollution also adversely affects the economy, climate, environment, and vegetation.

The COVID-19 outbreak has had a tremendous influence on society and caused unexpected changes in air quality. Governments all over the world have implemented various policies to promote social distancing, and it is the most efficient strategy to stop the virus’s spread (Chinazzi et al., 2020; Kraemer et al., 2020; Sun & Zhai, 2020; Tang et al., 2020; Tian et al., 2020). The Xi’an Municipal Government has also enacted an extreme lockdown from January 25, 2020, which includes the closure of all public transit, restaurants, enterprises, and construction sites. This unexpected quarantine significantly reduced pollutant emissions, offering vital information on the variables that could affect air quality (Pomponi et al., 2020; Tanzer-Gruener et al., 2020).

Recent research indicates that lockdown situations related to the COVID-19 outbreak significantly impact air quality in many cities and nations worldwide. The COVID-19 lockdown affected the world’s air quality, and the concentrations of PM_2.5, NO₂, PM₁₀, CO, and SO₂ decreased while those of O₃ increased during the COVID-19 pandemic (Liu et al., 2021). Many infectious diseases’ epidemiological dynamics depend on environmental factors (Shi et al., 2020). Generally, weather indicators can affect virus transmission (Raza et al., 2021). Firstly, Jahangiri et al. (2020) pointed out that temperature is a critical element in the spread of COVID-19. According to Chen et al. (2021), the spread of COVID-19 was also considerably impacted by temperature and humidity. Similarly, Tosepu et al. (2020) analyzed the relationship between weather indicators (rainfall, temperature) and the COVID-19 outbreak in Indonesia and found a significant positive association between COVID-19 and temperature. Furthermore, relative humidity and temperature significantly impact the dissemination of the COVID-19 outbreak in China (Wang & Su, 2020) and affected COVID-19 deaths in Wuhan (Ma et al., 2020; Yao et al., 2021). Few studies have investigated air pollutants’ spatial and temporal distribution during the COVID-19 pandemic. For example, Liu et al. (2021) investigated spatiotemporal distribution patterns and variations in air pollution before, during, and after the lockdown. Furthermore, Mehmood et al. (2021) pointed out the spatiotemporal variability of the COVID-19 pandemic concerning socioeconomic factors, climate, and air pollution in Pakistan. They found a significant relationship between humidity and COVID-19 cases.

Some scholars went much more profoundly to analyze the resources of pollutants. They discovered that while transportation and industrial emissions decreased, ozone (O₃) concentrations increased, which implies that emissions from urban activities should be prioritized (Abdullah et al., 2020; Borhani et al., 2021; Dhahad et al., 2021; Islam et al., 2021; Mor et al., 2021; Latif et al., 2021; Wang et al., 2020b; Yang et al., 2020). Eliminating the lockdown of the COVID-19 pandemic at the national level has caused significant changes in air pollution worldwide (Singh et al., 2020). For example, during the COVID-19 pandemic, most of the emissions in Almaty, Kazakhstan, were from residential heating systems and coal power plants rather than vehicles (Kerimray et al., 2020). However, these studies have not examined the complex interplay between the decline in pollutant concentrations and the reduction in emissions during the COVID-19 lockdown. Therefore, this research investigated the connection between climatic variables, air quality, and the COVID-19 pandemic.

Xi’an is located west of the FenWei Plain graben. The Fenhe-Weihe Plain is one of China’s most polluted regions due to the high cost of coal, which represents 90% of local energy consumption and is 30% more expensive than the national average. In 2019, the overall air quality in Xi’an was “Unhealthy” with an average US AQI number of 56.6 (WHO, 2021). This area’s primary particulate matter (PM) sources include coal consumption, dust emission, industrial activities, lighting, and motor vehicles. Compared to the emission source analysis results of previous years, the proportion of industrial operations declined from 58% to 11.3% in 2020, coal burning decreased from 52% to 18.8%, and motor vehicles increased from 2% to 27.4% (Dai et al., 2018; Wang et al., 2014; Wang et al., 2019; Wang et al., 2020a). Other research focused on biological and human factors that contribute to O₃ levels. The primary anthropogenic sources of O₃ in Xi’an are industrial emissions, which account for most of the city’s O₃ emissions; motor vehicles, which account for most of the city’s volatile organic compounds (VOC) emissions (64.4%); organic solvents, and biomass combustion. Various creatures and plants release numerous VOCs in the Qinling Mountain region near Xi’an, which can combine with NOx to produce O₃ in the vicinity of Xi’an (Feng et al., 2016; Li et al., 2018; Sun & Zhai, 2020). Understanding the impact of different human and environmental factors on persistent outbreaks is critical for infection control policymaking, particularly in places where transmission risk is perhaps underestimated, such as wet and warm places. Effective mapping and analysis of city air quality are essential to identify the air quality situation and formulate effective policies (Mahato et al., 2020). These investigations demonstrated the complexities of Xi’an’s pollution contributions, but additional research is needed from the viewpoints of geospatial data and urban planning issues.

This research aims to discover the relationships between air quality, the COVID-19 pandemic, and meteorological variables in Xi’an. First, we noticed changes in the COVID-19 distribution from January 2020 to July 2022. Second, we show the meteorological and air quality parameter concentration distribution from 2019 to 2020. We also illustrated how the pollutants were distributed seasonally before and during COVID-19. We used concentrations from the previous full year for a more accurate comparison. We divided them into three seasons (spring, summer, and winter) to correspond to the 12 months of the COVID-19 year. Third, we use R software to conduct a correlation to determine the relationships between air pollutants, COVID-19, and meteorological parameters before and after the COVID-19 period.

2. Data Collection and Methodology

2.1. Stduy Area

Xi’an is the largest city in the central part of Northwest China, and it is the capital of Shaanxi Province. It is located south of the Qinling Mountains, north of the Weihe River, and in the middle of the Guangzhong plain of the Yellow River (Mokoena et al., 2019). The study area is located approximately between geo-coordinate 34˚40'0''N, 107˚40'00''E, and 33˚40'0''N, 109˚40'00''E (Figure 1). The land extent of Xi’an is 10,097 km², with over 8.7 million inhabitants (Mokoena et al., 2019). The climate of the study area is sub-mixed and temperate continental monsoon, and it has a unique model with four seasons each year. The terrain of the study area is primarily flat, and the mountains in the northeast and east are about 1800 meters above sea level. Xi’an, known as Changan in ancient times, is China’s most renowned cultural and historical city. It has a severe problem of air pollution and ranks among the top 10 worst cities polluted with air pollutants in China (Cao et al., 2018). The geography and meteorology surrounding Xi’an are major factors in the poor air quality because of anticyclonic airflows stopping before mountains and leeward slopes, which cause pollutants to collect, and northeasterly winds blowing at low or even zero speeds, which prevent pollutants from dispersing.

Figure 1. Geographic location of Xi’an city in the capital of Shaanxi Province in the central part of Northwest China.

2.2. Data Collection

In this study, two years (2019 and 2020) of daily meteorological parameters (humidity, precipitation, temperature, and wind speed) data from NASA POWER (https://power.larc.nasa.gov/) were used. Data on the daily concentration of air pollutants such as O₃, NO₂, PM₁₀, PM_2.5, SO₂, CO, and the Air Quality Index, as well as mean values of NO₂, PM_2.5, PM₁₀, CO, and SO₂, and the 90th percentile concentration of monthly O₃ data were collected from August 2019 to August 2020 using the National Meteorological Information Center (http://www.nmc.cn/). COVID-19 case data was obtained from Xi’an Center for Disease Control and Prevention (http://www.xiancdc.com/). The complete methodology is shown in Figure 2.

Figure 2. The workflow of data collection, processing, and output in this study.

2.3. Spatial Variability of Air Quality

A geographic information system (GIS) is a robust tool for managing, storing, analyzing, and presenting spatial or geographic data. Under the expert judgment of GIS users or analysts, it can generate remedies to spatial difficulties. Researchers have used GIS technology to investigate pollutants’ temporal and spatial distribution (Jensen et al., 2001; Kumar et al., 2016). This study used spatial analysis to detect the spatial distribution patterns of air pollutants in the study area. The data regarding the monthly mean concentration of contaminants, such as NO₂, O₃, PM₁₀, and PM_2.5 from January to August 2019-20, were analyzed and visualized as a color-coded map regarding their spatial distribution in the study area.

2.4. Temporal Variability and Daily Trend of Air Pollutants

Time series data are ubiquitous, and time series analysis aims to deeply understand the phenomenon, discover repetitive patterns and trends, and predict future trends (van Wijk & van Selow, 1999). This study emphasized determining the temporal characteristics of Air quality index values. For this purpose, daily air quality index data from January 01 to December 31 during 2019-20 were collected. As well as visualized the temporal variation of the air quality index as the data calendar.

2.5. Mann-Kendall Trend Test

This test was used to understand the trend of air pollutants in over 13 districts of Xi’an. Mann-Kendall analysis can be applied to data that are not normally distributed (Meals et al., 2011). The Mann-Kendall analysis assumes that one value can always be declared lower than, greater than, or equal to another: the data are independent; data distribution in the original or converted units remains unchanged. Since Mann-Kendall analysis statistics are not changed to log transformation (test statistics have the same value as the original and log transformation data), it could be applied to many cases. To achieve a Mann-Kendall test, compute the difference between the later-measured value and all earlier-measured values (Y_j − Y_i), where j > i, and assign the integer value of 1, 0, or −1 to positive differences, no differences, and negative differences, respectively. The test statistic (S) is then computed as the sum of the integers using the following equation.

$S={\displaystyle {\sum }_{i=1}^{n=1}{\displaystyle {\sum }_{j=i+1}^n\mathrm{sgn}\left(Y_j-Y_i\right)}}$ (1)

When S is a large positive number, the value measured in the later stage is often larger than that in the early stage and indicates an upward trend (Meals et al., 2011). When S is a large negative number, the later value is often less than the earlier value and indicates a downward trend. No trend will be displayed when the absolute value of S is small. Test statistics τ can be calculated as follows:

$\tau =\frac{s}{n\left(n-1\right)/2}$ (2)

The values of τ range from −1 to +1, similar to the correlation coefficient in regression analysis. When S and τ significantly differ from zero, refuse to deny the null hypothesis of any trend (Meals et al., 2011). If a significant trend is established, the rate of change can be calculated using the Sen Slope estimator (Helsel & Hirsch, 1992; Meals et al., 2011) as given below.

${\beta }_1=\text{median}\left(\frac{Y_j-Y_i}{X_j-X_i}\right)$ (3)

For all, i < j and i = 1, 2, …, n − 1 and j = 2, 3, …, n; in other words, computing the slope for all pairs of data that were used to calculate S. The median of those slopes is the Sen Slope estimator (Meals et al., 2011).

2.6. Relationship between Weather Parameters and Air Pollutants

A scatter diagram is a tool to analyze the relationship between two variables. One variable is drawn on the vertical axis, and the other is on the horizontal axis. Their interesting point patterns can be displayed graphically concerning patterns (Chow & Yeung, 2006). Using a scatter diagram, this study analyzed the O₃, NO₂, CO, SO₂, PM_2.5, and PM₁₀ with weather indicators such as wind speed, air temperature, precipitation, and relative humidity between January to December 2019 and 2020.

3. Results and Discussions

3.1. Spatio-Temporal Analysis of Air Quality

3.1.1. Spatial Variability of Air Quality over the 13 Districts of Xi’an

To distinguish the spatial distribution pattern of pollutants may include the seasonal component. In this study, the spatial distribution pattern of the pollutants before and during the COVID-19 period was compared during winter (from January to February), spring (from March to May), and summer (from June to August). Figure 3 illustrates the spatial distribution pattern of air pollutants in the winter season with mean and 90th percentile values of air pollutant concentrations of districts before and during COVID-19. During the winter season (before the COVID-19 period), the highest PM₁₀ value i.e. 200.0 to 250.0 µg∙m⁻³ was reported in Zhoushi District. However, during the COVID-19 winter season, the concentration of PM₁₀ was significantly decreased over the study area, and its concentration ranged from 50.0 to 200.0 µg∙m⁻³.

Similarly, the concentration of NO₂ and PM_2.5 was significantly decreased during the COVID-19 winter season compared to that recorded before the COVID-19 winter season (Figure 3). The reported values during the COVID-19 winter season ranged from 50.0 to 150.0 µg∙m⁻³ (Figure 3). Before COVID-19 in January, the concentration of PM_2.5 had the highest value ranging from 150.0 to 200.0 µg∙m⁻³ (Figure 3). Earlier, the studies conducted in China revealed a reduction of 29% - 34% and 26% - 48% in the concentration of PM₁₀, and PM_2.5, respectively, during the COVID-19 period from January to March 2020 compared to that recorded in the same period before COVID-19 (Li et al., 2020).

The concentration of O₃ in Xi’an was distributed evenly both before and during the COVID-19 period in January, with values between 50.0 to 150.0 µg∙m⁻³. An upward distribution pattern only in February was reported, which has a higher value than before the pandemic, and it varied between 100.0 to 150.0 µg∙m⁻³ than before the COVID-19 period. In February (before COVID-19), a variation of the distribution of O₃ among all districts of the study area can be identified (Figure 4). According to Figure 4, Zhouzhi, Lantian, Linton, Yanta, and Xincheng districts reported a concentration of O₃ between 50.0 and 100.0 µg∙m⁻³. However, the concentration of O₃ was higher during the COVID-19 period than in the previous period i.e. the concentration of O₃ ranged between 100.0 to 150.0 µg∙m⁻³ for the whole city of Xi’an during the COVID-19 period.

Air pollutants in winter season before COVID-19 period

Air pollutants in winter season during COVID-19 period

Figure 3. Spatial distribution pattern of air pollutants in winter season before and during COVID-19 period.

Air pollutants in winter season before COVID-19 period

Air pollutants in winter season during COVID-19 period

Figure 4. Mean values for PM₁₀, PM_2.5, and NO₂ concentrations and 90th percentile value of ozone (O₃) concentrations in winter before and during the COVID-19 period.

Figure 5 illustrates the spatial distribution pattern of air pollutants in the district before and during the COVID-19 period in the spring season with their mean and 90th percentile concentration values. In the spring season, the contaminants of PM₁₀ fluctuated among all the districts of Xi’an, and values ranged between 50.0 - 150.0 µg∙m⁻³. During the spring season of 2019, the concentration of PM₁₀ ranged from 50.0 to 150.0 µg∙m⁻³, while lower concentrations were noted before and during the COVID-19 period in the winter season. NO₂ and PM_2.5 values ranged from 0.001 to 100.0 µg∙m⁻³, varying from district to district monthly. During May, the lowest NO₂ levels were recorded in all districts, which ranged between 0.001 to 50.0 µg∙m⁻³. However, the O₃ level showed a steady increase during this season. Accordingly, the highest value of O₃ concentration was recorded in the Huyi district in May, with a value range of 200.0 to 250.0 µg∙m⁻³. Overall, the distribution of air pollutants decreased in the spring season (before the COVID-19 period) compared to the winter season (before the CO- VID-19 period).

In the spring season of 2020, the PM₁₀ values were evenly distributed over three months, with values ranging from 50.0 to 100.0 µg∙m⁻³ in all the districts of Xi’an (Figure 5). A significant decrease in PM₁₀ concentrations was noted during the COVID-19 spring season. In addition, PM_2.5 and NO₂ showed a steady decline. It showed the lowest values in districts of Xi’an from the beginning of spring (March) to the end of spring (May). During the last month of the spring

Air pollutants in spring season before COVID-19 period

Air pollutants in spring season during COVID-19 period

Figure 5. Spatial variation of air pollutants in spring season before and during COVID-19 period.

season i.e. May, the reported values ranged from 0.0 to 50.0 µg∙m⁻³ in all districts. However, their levels steadily increased over the months of this season. The highest value of O₃ concentration was recorded during May in Xi’an, with values ranging from 200.0 to 250.0 µg∙m⁻³. The reported values of O₃ concentration were higher than other air pollutants (PM_2.5 and PM₁₀). Hence, the distribution pattern of air pollutants (NO₂, PM₁₀, and PM_2.5) showed a decreasing trend during the spring (during the COVID-19 period) compared to that observed during the spring before the COVID-19 period. Figure 6 shows this variability.

Air pollutants in spring season before COVID-19 period

Air pollutants in spring season during COVID-19 period

Figure 6. Mean values for PM₁₀, PM_2.5, and NO₂ concentrations and 90th percentile value of ozone (O₃) concentrations in spring before and during the COVID-19 period.

In the summer, the concentration of studied air pollutants was low compared to other seasons (Figure 7). In the summer of 2019, PM₁₀ concentration varied from district to district, ranging from 0.0 to 100.0 µg∙m⁻³. During June, July, and August, PM₁₀ spread equally in Zhouzhi, Huyi, and Changan districts with ranges between 0.0 to 50.0 µg∙m⁻³. The highest PM₁₀ value was reported from the Yanliang district throughout the summer season before the COVID-19 period. During the summer season (before the COVID-19 period), PM_2.5 and NO₂ spread throughout all districts in Xi’an without fluctuation, and their concentrations ranged from 0.000 to 50.000 µg∙m⁻³ for the whole city of Xi’an (Figure 7). The concentration of O₃ was higher than other air pollutants (PM₁₀, NO₂, and PM_2.5) during the summer season before the COVID-19 period, and its values ranged from 150.0 to 250.0 µg∙m⁻³ (Figure 7). The amount of O₃ also varied from district

Air pollutants in summer season before COVID-19 period

Air pollutants in summer season during COVID-19 period

Figure 7. Spatial distribution pattern of air pollutants in summer season before and during COVID-19 period.

to district during this season. In August, a downward distribution pattern in the concentration of O₃ was recorded in the Zhouzhi district, and it ranged from 100.0 to 150.0 µg∙m⁻³.

Figure 7 illustrates that the PM₁₀ level declined from the beginning to the end of the summer season during COVID-19. The concentration of PM₁₀ ranged from 50.0 to 100.0 µg∙m⁻³ during the June 11 districts of the study area. Nevertheless, it was not reported in any district in August, and its concentration ranged from 0.0 to 50.0 µg∙m⁻³. During the COVID-19 period in the summer season, a similar concentration range of PM_2.5 and NO₂ was recorded. Its spread was uniform throughout all the districts of the study area. The concentration of PM_2.5 and NO₂ ranged from 0.0 to 50.0 µg∙m⁻³ for the whole city of Xi’an during the COVID-19 period. Not only that, but also this distribution pattern was also the same as before the COVID-19 period.

However, the concentration of O₃ in all the districts of Xi’an was higher in the summer than in other seasons, and its values ranged from 200.0 to 250.0 µg∙m⁻³ during the COVID-19 period. June and August showed the same distribution pattern of O₃, while the highest O₃ value was recorded in the Huyi and Changan districts during the three months of the summer season. The reported PM₁₀, PM_2.5, and NO₂ values were lower in summer than in winter and spring seasons. In a similar study, Singh et al. (2020) reported that a decreasing trend in PM_2.5 can be observed across all regions because of the winter-spring-summer transition. However, the concentration of O₃ was reported to be higher than in other seasons. The bar charts included in Figure 8 showed this variability very clearly.

Air pollutants in summer season before COVID-19 period

Air pollutants in summer season during COVID-19 period

Figure 8. Mean values for PM₁₀, PM_2.5, and NO₂ concentrations and 90th percentile value of ozone (O₃) concentrations in summer before and during the COVID-19 period.

Air pollutants fluctuate annually, seasonally, monthly, and diurnally in different environments (Singh et al., 2020). Meteorology and local emissions control changes in a specific geographic location. Due to changes in the weather conditions, the primary pollutants emitted during winter to spring and summer may decrease, respectively, while O₃ concentrations tend to increase during the transition from winter, spring, and summer due to photochemical production (Singh et al., 2020). However, some pollutants were highest in other months, possibly due to anthropogenic activities such as transportation, industries, thermal power plants, and residential biomass burning in that area. During the COVID-19 period, the concentration of air pollutants in all the districts showed a significant decrease compared to before the COVID-19 period, which might be due to the reduction of industrial production, strict traffic, and people’s lack of willingness to travel, thermal power plants, and residential biomass burning due to the pandemic in the study area. In a similar study, Kumar (2020) applied air quality in India during this COVID-19 period. His conclusion showed that due to the strict lockdown in India, all public transport, industries, and individual activities were shut down, reflected in air quality and aerosols all over India. The aerosols have decreased sharply in India, in contrast to the average value of AOD over the last three years. As well as it’s very clear from the plot that the concentration of NO₂ is reduced after the lockdown compared to the three-year average value of NO₂. Finally, they found an apparent reduction in all the pollutants for all cities during the lockdown period.

According to this study, the difference in atmospheric pollutant concentrations during the COVID-19 period and before the COVID-19 period in Xi’an indicates that the emergency response mechanism caused by the COVID-19 pandemic had a significant effect on changing the air pollutant concentration in Xi’an.

3.1.2. Temporal Variability of Air Quality Index (AQI) over the Study Area

The Air Quality Index (AQI) before and during COVID-19 has been visualized through two data calendar maps; one calendar map in 2019 with month and day variation, and the other for 2020 with month and day variation. Daily air quality values have been presented as a calendar map, which is helpful for an intuitive inspection of the severity of pollutants (Figure 9 and Figure 10).

In 2019, before COVID-19, the highest air quality values were recorded compared to those observed in 2020 (Figure 9). The pollutant concentrations were extremely high during January, May, and December. These three months had severely polluted days with AQI values of 301 - 500. The lowest AQI was reported during September 2019, with an air quality index of 0 - 50, and it had 11 excellent days. During 2019, around 225 days were excellent/good days (air quality value less than 100) in selected months, and 140 days were polluted with AQI values greater than 101.

Figure 9. Xi’an’s air quality index and air quality level before the COVID-19 period (January to December 2019).

Figure 10. Xi’an’s air quality index and air quality level of Xian, during the COVID-19 period (January to December 2020).

Figure 10 illustrates that pollutant concentrations were extremely high during January and December 2020, but it did not include severely polluted days as the AQI values were not greater than 300. The lowest air quality index was reported for August, with an air quality value of 0 - 50, and it has 11 excellent days. Not only that but no day in 2020 that was severely polluted was also reported in all the months. In 2020, 250 days can be identified as excellent/good days with air quality values less than 100 in the selected months, and 116 days were polluted with AQI values greater than 101.

According to Figure 9 and Figure 10, sixteen more excellent days were reported during COVID-19 in 2020 compared to before COVID-19 in 2019, which was 41% more than in 2019. Ten more good days were reported during COVID-19 in 2020 than in 2019. Air quality levels of heavy pollution days decreased in 2020 and 2019 by 31.8%, and eleven lightly polluted days decreased in 2020 compared to 2019. Hence, no severely polluted days were reported in 2020, and air quality was 100% decreased, which was a good sign for the environment during COVID-19. In a similar study, Mahato et al. (2020) documented that stated earlier the lockdown started on the 24th of March, and just after 1 day of the beginning of the lockdown (i.e. 25th of March), there was a significant upgrade in air quality in comparison to that of the pre-lockdown phase in New Delhi, India. There are several reasons behind the variations in air quality indices of the study area during the pre-and-post COVID-19 period. Xi’an is the largest city and the provincial capital in Central China. It is one of the most attractive places for tourists in China and has a market and trade center. Hence, it has rapid urbanization, a high population growth rate, and increasing traffic volume daily. The above factors led to increasing the pollution indices before the COVID-19 period. During the COVID-19 period, Xi’an restricted most of its economic and human activities, except for the utilities of primary products and facilities such as health, food, medicine, human and animal well-being, banks, families, public transport, construction, and emergency services, and energy (electricity and fuel). As a result, the primary source of emissions changed to a low level. Not only that, but the metrological conditions lead to a change in air pollution indices. For example, heavy and severe pollution days increased more in the winter (December, January, and February) than in other seasons before and during COVID-19. The researchers observed similar changes regarding air quality in different countries (Gouda et al., 2021; Khorsandi et al., 2021). Overall, the COVID-19 period leads to comparatively healthier air quality than before the COVID-19 period.

3.1.3. Daily Trend of Air Pollutants over the Study Area

To examine the trends (decreasing or increasing) of air pollutants’ variation over time, Mann-Kendall’s (M. K.) trend analysis was performed in this study. Figure 11 represents the Mann-Kendall test results for the daily PM_2.5, PM₁₀, NO₂, SO₂, CO, O₃-8h (µg∙m⁻³) in Xi’an from January 01 to December 31, 2019 (before the COVID-19 period) and January 01 to December 31, 2020 (during COVID-19 period).

Figure 11. Daily PM_2.5, PM₁₀, NO₂, SO₂, CO, O₃-8h (µg∙m⁻³) before COVID-19 and during the COVID-19 period.

In the Mann-Kendall analysis of air pollution data, the overall results of the M. K. trend analysis are presented in Table 1. If the p-value was less than the significance level i.e. α = 0.05, H0 was rejected. Rejecting H0 indicates a trend in the time series while accepting H0 indicates no trend was detected. When rejecting the null hypothesis, the results have been said to be statistically significant.

Table 1. Summary of results in Mann Kendall Test-before COVID-19 period.

H0 = There is no trend in the study period.

H1 = There is a trend in the study period.

The M. K. test gave fascinating insights into air pollutants data before and during COVID-19. The M. K. test Statistic (S) indicated a decreased trend for the five air pollutants such as PM_2.5, PM₁₀, NO₂, SO₂, CO, and O₃-8h before COVID-19 (Table 1). However, an increasing trend in O₃-8h and NO₂ air pollutants was recorded during COVID-19. The null hypothesis (H0) was rejected for PM_2.5, PM₁₀, and CO air pollutants before the COVID-19 period, while it was accepted for the pollutants i.e. O₃-8h, NO₂, and SO₂ before the COVID-19 period. This trend suggests that a significant decreasing trend can be identified for PM_2.5, PM₁₀, and CO, and no trend was identified for O₃-8h, NO₂, and SO₂ before COVID-19. During the COVID-19 period 2020, the null hypothesis (H0) was rejected for PM_2.5, PM₁₀, NO₂, SO₂, and O₃-8h while it was accepted for CO only. This suggests that a significant decreasing trend can be identified for PM_2.5, PM₁₀, NO₂, SO₂, and O₃-8h, with no significant trend for CO during the COVID-19 period. Among the pollutants, the concentrations of five major pollutants showed a decreasing trend. In contrast, CO had no significant trend during the COVID-19 period.

3.2. Relationship between Weather Parameters and Air Pollutants in the Study Area before and during the COVID-19 Pandemic

The scatter plot method was applied to study the relationship between the weather and air pollutants data from January to December 2019 and 2020 (Figures 12-15). The details about each relationship are given below:

Before the COVID-19 period

During the COVID-19 period

Figure 12. The relationship between air pollutant and temperature before and during the COVID-19 period.

Before the COVID-19 period

During the COVID-19 period

Figure 13. The relationship between air pollutants and precipitation before and during the COVID-19 period.

Before the COVID-19 period

During the COVID-19 period

Figure 14. The relationship between air pollutants and relative humidity before and during the COVID-19 period.

Before the COVID-19 period

During the COVID-19 period

Figure 15. The relationship between air pollutants and wind speed before and during the COVID-19 period.

3.2.1. Air Pollutants and Temperature

A negative correlation was recorded between different air pollutants (PM_2.5, NO₂, PM₁₀, CO, and SO₂) and the temperature in 2019 with correlation coefficient (R2) values i.e. 0.4592, 0.234, 0.2853, 0.4867, 0.4342, respectively (Figure 12). These results reveal that air pollutants such as PM_2.5, NO₂, PM₁₀, CO, and SO₂ reduce when the temperature increases. However, a moderately positive correlation was recorded between O₃-8h and temperature during 2019. Its correlation coefficient value was R2 = 0.6344, which showed that O₃-8h in the air increases with an increase in temperature.

During the COVID-19 period in 2020, a negative correlation between different air pollutants (PM_2.5, NO₂, PM₁₀, CO, and SO₂) and temperature was recorded with R2 = 0.4269, 0.0264, 0.2251, 0.3359, 0.3815, respectively (Figure 12). These results display that the concentration of air pollutants such as PM_2.5, NO₂, PM₁₀, CO, and SO₂ decreased as the temperature increased. However, a moderately positive correlation (R2 = 0.4411) between O₃-8h and temperature was observed during 2020, which showed that the O₃-8h in ambient air escalates with increasing temperature.

Comparing the correlation results between air pollutants and temperature before and during COVID-19, the R2 values were lower than during COVID-19. It has been found that the unpropitious impact of environmental ozone pollution on humans’ well-being increases with increasing temperature (Jacob & Winner, 2009). The natural cover of an urban environment composed of wetlands, vegetation, waterbody, bare soil, and open spaces is usually substituted by the characteristics of diverse land-use types, such as roads, sidewalks, buildings, and other artificial structures made of different materials (such as asphalt), metals and bricks, which themselves can absorb heat during the day and then release heat at night; All these together lead to the rise of surface temperature in an urban environment, which in turn contributes to the development of Urban Heat Island (UHI) phenomenon (Ayanlade, 2016; Oke, 1973; Sampson et al., 2021) and it supports the increase in air pollution. Jiang et al. (2020) conducted a similar study in Fuzhou, China, and recorded similar results i.e. the temperature was negatively correlated with PM₁₀, PM_2.5, and NO₂ and positively correlated with O₃-8h.

3.2.2. Air Pollutants and Precipitation

Figure 13 shows a weak negative correlation between air pollutants i.e. PM_2.5, NO₂, PM₁₀, CO, SO₂, and O₃-8h, and the precipitation before and during COVID-19. These results show that air pollutants such as PM_2.5, NO₂, PM₁₀, CO, SO₂, and O₃-8h increase as the precipitation decreases. However, this reduction is the same in magnitude as that detected during a similar time during the period of COVID-19 in this study. Moreover, the values of R2 during the COVID-19 period were lower than before the COVID-19 period. Earlier, it has been found that precipitation can have a variable effect on air pollutants by removing gaseous pollution and particulate matter deposition through atmospheric chemical processes (Shukla et al., 2008). Similarly, Liu et al. (2020) pointed out that the exclusion impact of precipitation on aerosol particles was associated with raindrop diameter, aerosol particle size, and precipitation intensity. They suggested that air pollutants may not be correlated with precipitation.

3.2.3. Air Pollutants and Relative Humidity

Air pollutants such as PM₁₀, NO₂, SO₂, and O₃-8h were negatively correlated with relative humidity before and during COVID-19 (Figure 14). These results show that the concentration of air pollutants such as PM₁₀, NO₂, SO₂, and O₃-8h decreases with increasing relative humidity. Kayes et al., (2019) found that the concentration of most air pollutants had a negative relationship with relative humidity. Relative humidity affects the movement of particles and may deposit particles on the ground surface. Consequently, air pollutants decrease with the rise in relative humidity (Giri et al., 2008; Kumar, 2020). A weak but positive correlation between PM_2.5, CO, and relative humidity was recorded before and during COVID-19.

3.2.4. Air Pollutants and Wind Speed

A weak and negative correlation between air pollutants (PM_2.5, NO₂, SO₂, CO, and O₃-8h) and wind speed was recorded before and during the COVID-19 period in the study area (Figure 15). These results showed that air pollutants such as PM_2.5, NO₂, SO₂, CO, and O₃-8h increase as the wind speed decreases. The wind significantly impacts pollution (Liu et al., 2020). Dincer et al., (2010) mentioned that wind speed negatively correlates with the concentration of air pollutants data. However, in this study, PM₁₀ and wind speed have shown a weak but positive correlation during COVID-19. The relationship between air pollutants and wind speed can supply significant details about air pollution, as wind speeds can move air pollutants from distant sources (Dincer et al., 2010).

The densely populated capital of the Shanxi province of Xi’an had the highest air pollutant emissions before the COVID-19 period. However, activities of air pollutant emission sources, such as houses, coal-based thermal power plants, residential biomass burning, industries (sulfur-contained fuel used in different industries), and construction activities, were reduced during COVID-19. As a result, the correlation coefficient values regarding various parameters studies decreased during the COVID-19 period compared to before the COVID-19 period.

4. Conclusions and Recommendations

The COVID-19 pandemic (SARS-COV-2) has led to an improvement in global air quality. This study investigated the spatial distribution pattern of air pollutants, the temporal distribution of air quality indices, and the trends in air pollutant patterns. Finally, it analyzed the relationship between meteorological indicators (temperature, precipitation, humidity, and wind speed) and air pollutants (PM₁₀, NO₂, PM_2.5, CO, SO₂, and O₃-8h), in Xi’an, China, before and during the COVID-19 period. Based on the experimental analyses, this study concluded that the concentration of air pollutants such as PM_2.5, NO₂, PM₁₀, CO, and SO₂ decreases in winter, spring, and summer. In contrast, the concentration of O₃-8h increases from winter to summer due to the variation in weather conditions during COVID-19. No severely polluted days were reported during COVID-19, and the air quality index decreased by 100 percent. A significant trend was identified in air pollutants (PM_2.5, PM₁₀, and CO), while no significant trend was observed in the case of O₃-8h, NO₂, and SO₂ before COVID-19.

The results recorded in this study will help recognize Xi’an’s air pollution during and before COVID-19. The results may also provide helpful evidence of the benefits of different air pollution control approaches before and during COVID-19. In addition, these analyses are an addition to identifying the literature before and during the impact of COVID-19 on air pollution and, more generally, identifying air quality changes related to specific causes. In this study, outcomes provide information for further research to categorize the status of exposure performance and level of air pollution, particularly from the studies with the robust procedure usual for essential confounders. The advantages of air quality measures during the COVID-19 lockdown were performed to be an inimitable chance for pollution control strategies. Finally, this study would aid in various clean air programs and air pollution modeling in the future globally.

Acknowledgements

This research was funded by the Postdoctoral Startup Research Fund of Henan University, number (CJ3050A0671293).

List of Abbreviations

Particulate Matter (PM), Nitrogen Dioxide (NO₂), Carbon Monoxide (CO), Sulfur Dioxide SO₂, Ozone O₃, Air Quality Index (AQI), Volatile Organic Compounds (VOC), Nitrogen Oxide (NO_x), particulate matter (PM), National Aeronautics and Space Administration (NASA), Prediction of Worldwide Energy Resources (POWER), Standard Deviation (SD), Geographic Information System (GIS).

Conflicts of Interest

The authors declare no conflicts of interest regarding the publication of this paper.

References