Restricted Anthropogenic Activities and Improved Urban Air Quality in China: Evidence from Real-Time and Remotely Sensed Datasets Using Air Quality Zonal Modeling

Abstract

The study aims to examine the major atmospheric air pollutants such as NO₂, CO, O₃, PM_2.5, PM₁₀, and SO₂ to assess the overall air quality using air quality zonal modeling of 15 major cities of China before and after the COVID-19 pandemic period. The spatio-temporal changes in NO₂ and other atmospheric pollutants exhibited enormous reduction due to the imposition of a nationwide lockdown. The present study used a 10-day as well as 60-day tropospheric column time-average map of NO₂ with spatial resolution 0.25 × 0.25° obtained from the Global Modeling and Assimilation Office, NASA. The air quality zonal model was employed to assess the total NO₂ load and its change during the pandemic period for each specific region. Ground surface monitoring data for CO, NO₂, O₃, PM₁₀, PM_2.5, and SO₂ including Air Quality Index (AQI) were collected from the Ministry of Environmental Protection of China (MEPC). The results from both datasets demonstrated that NO₂ has drastically dropped in all the major cities across China. The concentration of CO, PM₁₀, PM_2.5, and SO₂ demonstrated a decreasing trend whereas the concentration of O₃ increased substantially in all cities after the lockdown effect as observed from real-time monitoring data. Because of the complete shutdown of all industrial activities and vehicular movements, the atmosphere experienced a lower concentration of major pollutants that improves the overall air quality. The regulation of anthropogenic activities due to the COVID-19 pandemic has not only contained the spread of the virus but also facilitated the improvement of the overall air quality. Guangzhou (43%), Harbin (42%), Jinan (33%), and Chengdu (32%) have experienced maximum air quality improving rates, whereas Anshan (7%), Lanzhou (17%), and Xian (25%) exhibited less improved AQI among 15 cities of China during the study period. The government needs to establish an environmental policy framework involving central, provincial, and local governments with stringent laws for environmental protection.

1. Introduction

Air pollution and its quality in cities are the major concerns worldwide and China is no exception. Air pollution has been a serious environmental issue in China since extensive industrial production and other anthropogenic activities exponentially increased the concentration of major atmospheric pollutants that damaged the environmental quality and harmfully impacted human health [1,2,3,4,5]. The World Health Organization (WHO) reiterates that environmental and health risk factors from outdoor air pollution are substantial. In 2016, 91% of the world population was living in places where the WHO air quality guidelines levels were not met [6]. New estimates in 2018 revealed that 9 out of 10 people breathe air containing high levels of pollutants [7]. Over the last four decades, rapid economic development, industrialization, and urbanization across China has caused serious air pollution problems [8,9,10]. China’s National Western Development Strategies (the ‘Go West’ movement) by the Chinese government in 1999 also contributed to the spatial expansion of industrialization and urbanization [11]. This has resulted in the GDP reaching around RMB 8270 billion, to which industrial production contributes 34% as the most significant contributing factor (http://www.stats.gov.cn/tjsj/ndsj/, accessed on 18 April 2020). Furthermore, due to rapid industrialization and urbanization air pollution has become the fourth primary health risk factor for all deaths in China after heart attack, dietary risk, and smoking [12]. The recent COVID-19 pandemic at the global and regional level has substantially affected the spatial and temporal characteristics of significant air pollutants, such as NO₂, CO, O₃, PM₁₀, PM_2.5, and SO₂ across China [13,14,15]. In this context, the present study aims to assess the overall air quality from ground-based monitoring and remotely sensed datasets using air quality zonal modeling in 15 major cities of China before and after pandemic lockdown periods. Here, we estimated the changes in pollutant concentration and the spatio-temporal variation characteristics for 60 days before and after the shutting down of the nation. Section 2 discusses data and methods where we mentioned data sources, how we analyzed the data, and what techniques we used. Next are the results in Section 3 where we analyzed the spatio-temporal distribution of NO₂ before and after the pandemic; total NO₂ load over the area using air quality zonal modeling, and trends of air pollutants in China before and after the pandemic. Section 4 presents the discussion where we critically evaluate the results and link our study with other relevant studies with reasons and evidence.

NO₂ is primarily released by anthropogenic emissions, which contain the industrial burning of fossil fuels such as coal, oil, natural gas, vehicle exhaust, biomass burning, and electricity generation [16,17,18]. Pollutants such as carbon monoxide (CO) and sulphur dioxide (SO₂) are mainly released from industrial plants and heating processes due to anthropogenic activities [19]. SO₂ is regarded as one of the major air pollutants in cities because of its negative effects on human health and the ecosystem [20]. Particulate Matter (PM₁₀, PM_2.5) is a common proxy indicator for air pollution and it affects more people than any other pollutant. The major components of PM are sulfate, nitrates, ammonia, sodium chloride, black carbon, mineral dust, and water [6]. Tropospheric O₃ is produced by emissions of nitrogen oxides (NO_x = NO + NO₂) and volatile organic compounds (VOCs) in the presence of sunlight [21]. The distribution of O₃ varies with space and time and is lower in urban polluted areas than elsewhere because it disappears when it reacts with other pollutants [22,23]. Among NO_x elements, NO₂ is the most important precursor and quencher of O₃ through NO_x titration particularly during wintertime [24]. The daytime relationship between NO₂ and O₃ concentration in Northern China demonstrates significant ozone titration. However, this titration effect has been considerably attenuated with increasing PM_2.5, which further reduces the incoming solar radiation, during the lockdown period [25]. It is not only harmful to human health but also poses adverse impacts on plants and ecosystems [1,3,26,27].

Various studies have been carried out to investigate the spatial and temporal distribution patterns of NO₂, CO, O₃. PM₁₀, PM_2.5, and SO₂ at national, regional, and local levels [2,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42] along with their driving forces in the recent past [27,43,44,45,46]. Numerous studies have shown a strong correlation between level of exposure to atmospheric pollutants and human health [2,4,5]; increased mortality [47,48,49]; sleeping disorders among elderly [50]; body weight and obesity [51]. To mitigate air pollution, the Chinese government has imposed stringent air quality standards [52] and strengthened emission controls of major atmospheric pollutants [53,54,55,56,57].

The understanding of the spatial distribution of air pollution in China has been recently improved by the application of advanced assessment tools, such as satellite remote sensing. The satellite-retrieved products have many advantages, including global coverage, high spatial-temporal resolution, and historical datasets [41,42,47]. Amid the coronavirus pandemic in China, NASA and European Space Agency (ESA) pollution monitoring satellites have detected significant decreases in nitrogen dioxide (NO₂) over mainland China [13,14,15]. There is evidence that the change is at least partly related to the economic slowdown following the outbreak of the COVID-19 virus. According to NASA scientists, the reduction in NO₂ pollution was first apparent near Wuhan, but eventually broke out across the country [58]. In line with the previous studies of holiday effects on pollution [59,60,61,62], the main purpose of this work is to quantify spatial and temporal changes in primary pollutants such as NO₂, CO, O₃, PM₁₀, PM_2.5, and SO₂ over China’s mainland due to shutdown of anthropogenic activities.

2. Data and Methods

2.1. Study Region

This study focuses on COVID-19′s footprint on air quality in 15 cities of China’s mainland namely Guangzhou, Xiamen, Xian, Wuhan, Beijing, Nanjing, Xinjiang, Lanzhou, Anshan, Shanghai, Jinan, Harbin, Chongqing, Zhengzhou, and Chengdu from where 145 random grid samples were taken for micro-level assessment with the help of air quality zonal modeling (Figures 3 and 6).

2.2. Data Source

The time-average map of NO₂ for consecutive 10- and 60-day average tropospheric columns (molecules/cm²) and time series daily area-average of NO₂ tropospheric columns (30% cloud screened) with spatial resolution 0.25 × 0.25° from 24 November 2019 to 22 March 2020 were collected from Giovanni interface [63,64,65], Global Modeling and Assimilation Office, NASA. The hourly and daily average concentrations of CO, O₃, NO₂, PM₁₀, PM_2.5, SO₂, and AQI data for 50 consecutive days before and 50 days after lockdown were collected from ground monitoring stations provided by the Ministry of Environmental Protection of China (MEPC), China Ministry of Ecology and Environment (https://aqicn.org/map/china/cn/, accessed on 20 March 2020) for the same study period.

2.3. OMI Data Retrieval

Nitrogen dioxide (NO₂) is a very important atmospheric gas in both the troposphere where it is a precursor to ozone production and in the stratosphere where it plays the main role in ozone chemistry (https://disc.gsfc.nasa.gov/datasets/OMNO2d_003/summary, accessed on 25 March 2020). NO₂ in the troposphere is produced in various combustion processes and lightning, which is an indicator of poor air quality [66]. NO₂ product (OMNO2d) is a daily global Level-3 gridded data with a spatial resolution of 0.25 × 0.25° [67,68] which contains time-average tropospheric column (30% cloud screened) and area-average total column [31]. We further retrieved the dataset from optional tools (use smoothing, equidistant cylindrical projection system) with linear scaling and a maximum of 65 palettes were used for this study. In this process, we ignored negative values especially for the 10-day average of the NO₂ total column map [31,69].

2.4. Data Analysis Method

As mentioned before, OMI data have been used for detecting the total column of NO₂ levels in the atmosphere [31,70]. Previous studies [31,70,71,72] have also used time-average OMI tropospheric total column maps and area-average OMI tropospheric total column datasets for assessing the air quality, especially for macro scale. We selected China’s mainland for this study and assessed the total column of NO₂ before and after the outbreak of COVID-19. This study divided China’s mainland into 145 grids (dimension 255 × 255 km²) for the micro-level study. While in the process, the work extracted the exact value of NO₂ tropospheric total column of each grid for 15 cities using spatial attribute tools in ArcMap and then we divided 145 grids into 5 equal categories based on the maximum and minimum range. In the last stage, we calculated the total area of each category and compared the extracted datasets with before and after 60- and 10-day area-average columns of NO₂ in total and in percentage.

We calculated the daily mean concentration of CO, O₃, NO₂, PM₁₀, PM_2.5, SO₂, and AQI from ground monitoring datasets provided by MEPC. We also calculated the 50 days before and after average concentration of each pollutant in 14 major cities of China obtained from real-time datasets. This study conducted a compound bar graph and trend line analysis for monitoring the change in air quality before and after the lockdown of 14 cities of China, in terms of total and percentage.

Air Quality Zonal Modeling

We conducted air quality zonal modeling for predicting zonal change in air quality. In general, the air quality zonal modeling represents the overall pollution concentration in any specific region based on several pollutant particles obtained from remotely sensed raster datasets [73]. In other words, the air quality zonal model is a compilation of total air pollution load in a particular zone accumulated by various air pollution particles which is represented through individual grids over a specific time period. This spatial and temporal model were created through different phases using various statistical analyses of GES-DISC time-averaged map and area-averaged time-series datasets obtained from MODIS- Terra, MERRA-2, OMI, and AIRS.

For this study, we conducted various sets of statistical analyses for interpreting remotely sensed data. In the first phase, we divided China’s mainland into 145 grids, each grid representing 65,025 km². After that, we extracted each pollutant’s average concentration from an individual grid using the raster calculator in ArcMap. The Moving Mean (MM) of 10 days total column of NO₂ of 15 cities was interpreted using equation 1. A specific grid value was ranked using exploratory analysis of ranking given by Alvo and Philip [74] (Equation (2)). The average concentration of each air pollutant particle of 145 grids was categorized based on the maximum and minimum range of datasets which is computed in the second phase (Equation (3)). After that, we compiled total pollution concentration in a specific grid using composite indexing and principal component analysis (PCA); subsequently, we conducted Spearman’s model for factor analysis for assessing the maximum and minimum reduction [75] in specific air pollutants which is compiled in the third phase (Equations (2) and (3)). The 60-day average total air pollution load was calculated using Equation (4); here, we excluded zero factors of different raster datasets. The sum of the 60-day average concentration of different air pollutants and their percentage change was calculated using Equation (5). Zonal indexing of specific grids was formulated for extracting the exact concentration of air pollution (Equation (6)) and in the last phase, we calculated area/grid-wise total air pollution load (Equations (7) and (8)) through spatial analysis tools in GIS using air quality zonal modeling.

$$mm=\frac{\sum 10}{n}\times 6$$

(1)

where mm = moving mean, $\sum 10$ = sum of 10 days average, n = number, and 6 = number of raster data.

$$m=\sum _{j=1}^{\mathit{dx}} \frac{\left(n_j{v}_j\right)\times i}{n}$$

(2)

where $m=$ mean rank, $dx=$ dimension, $i$ = th entry equals, $v_j, j=1$ represents all possible rankings of the $dx$, $n_j$ = frequency of ranking, $v_j\left(i\right)$ = rank score, and n = number of raster data sets.

$$x=\left[n_1 F_1+n_2 F_2+\right]\dots \left[+n_{mm} F_m+s_m F_m\right]$$

(3)

where x = variation in maximum and minimum range of air pollutants with zero mean, $n_1, n_2$ = loading factors of specified air pollutants and $n_{mm}$ = moving mean, $F_1,F_2$ = common factors and $F_m$ = moving mean of the common factor with zero mean, and $s_m$ = specific factor of the individual mean.

$$\mu p=\left[\frac{\sum \left(x1+x2+x3+x4+x5+x6\right)}{6}\right]\cup \left[na\right]$$

(4)

where $\mu$p = 60 days average of pollution load, $x$ = sum of 10 days average of pollution load, 6 = number of datasets, and $\cup \left[na\right]$ = excluding factor of zero (data not available).

$$p=\frac{\left(\sum 60a-\sum 60b\right)}{\sum 60b}\times 100$$

(5)

where $p$ = change in percentage, $\sum 60$ = sum of 60 days average concentration, a = during lockdown, and b = before.

$$a=\sqrt{\frac{\sum wi{\left(xi-x\right)}^2}{\sum wi}}$$

(6)

where $a$ = area of individual cells, w = weights, i = an index over all the data points being averaged, and xi = individual pollutant variables.

$$t=\sum taop-\left[na\right]$$

(7)

where t = total pollution counts in individual grids, taop = total air pollution load, $\left[na\right]$= excluding ‘data not available’ girds.

$$a=n\times dx$$

(8)

where a = total area of each category, n = number of countable grids, and dx = dimension of different variables/ grids (255 × 255 km²).

3. Results

3.1. Spatio-Temporal Distribution of NO₂ before and after the Pandemic

The spatial distribution of 10-day time-average maps of NO₂ tropospheric columns before and after the COVID-19 pandemic exhibited that the cities experienced a low concentration of NO₂ and improved quality of air. Among these, the period 2 February to 11 February observed the lowest level of NO₂ in all 15 major cities (Figure 1).

The post-lockdown maps portray that NO₂ concentration has increased from 22 February to 12 March and then again started decreasing from 13 March to 22 March. The 10-day average NO₂ tropospheric column (Figure 2) reveals that the level of NO₂ has tremendously reduced in major cities such as Guangzhou, Beijing, Anshan, Zhengzhou, and Chengdu after the pandemic. The air quality has started improving from the 2nd week of January just after the lockdown throughout the country.

Similarly, the 60-day time-average map (Figure 3) of NO₂ tropospheric column also displayed that the lowest range of NO₂ concentration was 2^e+14 and the maximum concentration was 3.5^e+16 before imposition of lockdown whereas the lowest range of NO₂ concentration came down to 1.4^e+15 and the maximum to 1.8^e+16 after the lockdown.

The absolute concentration of NO₂ tropospheric column (Figure 4) also revealed that all 15 major cities of China experienced lower emissions of NO₂ just after lockdown. The maximum concentration of NO₂ tropospheric column was 2.8^e+16 before lockdown (24 November–22 January) and the minimum concentration was 1^e+16 after lockdown (23 January–22 March) (Figure 4a). While we look at the proportion of NO₂ concentration reduction, Guangzhou received only 37% of NO₂ in the atmosphere after the pandemic. Similarly, the air quality of other cities also improved because of low emissions of NO₂ after lockdown. The graph (Figure 4b) revealed that Wuhan received only 35% of NO₂ concentration; whereas, Chengdu received 39%, Beijing 41%, Nanjing 48%, Xiamen 50%, Chongqing 53%, and Harbin 57% after the pandemic outbreak (23 January–22 March). On the other hand, the minimum reduction in NO₂ concentration was found in cities of Anshan (67%), Lanzhou (67%), Jinan (66%), Xinjiang (65%) where NO₂ concentration was comparatively low even before the pandemic in these same regions (Figure 4b).

The time-series of NO₂ tropospheric column graph (30% cloud screened) displays that the total mass column of NO₂ in China was highest during the first two weeks of December 2019 and remained as high as 4.3^e+15 till the second week of January. After that, the NO₂ mass column suddenly dropped till March 22 (Figure 5).

3.2. Total NO₂ Load over the Area Using air Quality Zonal Modeling

A total 60-day average NO₂ tropospheric column was demonstrated through the air quality zonal modeling (Figure 6). Here, the whole country was divided into 145 grids (255 × 255 km²) where the spatial distribution of NO₂ (Figure 6a) was displayed for the micro-level understanding of the air quality changes in different regions of the country. The area was also calculated under each grid category. The NO₂ concentration ranges between (1.7^e+15–3.2^e+15) had 131 grids and the area was 5.5 million sq.km before the lockdown but the grids have increased to 191 and the area has also increased to 80 million sq.km. Conversely, the areas under the high NO₂ concentration category (5.5^e+15–9.8^e+15) have significantly decreased from 1.1 to 0.2 million sq.km. Moreover, there was a significant decrease in NO₂ concentration in the areas that come under a very high concentration category where the NO₂ tropospheric column ranges between 1.6^e+16 and 2.6^e+16; 13 grids were accounting for 0.55 million km² before the lockdown but none of the areas were under this very high category after lockdown (Figure 6a).

The total column of NO₂ decline rate in percentage was grouped into three categories, i.e., less than 50%, 51–70%, and more than 70%. The maximum decline rate (more than 70%) of the NO₂ total column was measured in Guangzhou, Wuhan, and Chengdu city of China before and after the pandemic COVID-19 event (Figure 6b). Figure 6c reveals the trend of the total NO₂ tropospheric column of 145 random sites before and after the pandemic where it is clearly visible that almost all the sample sites have experienced a low concentration of NO₂ in the atmosphere.

In the graphs (Figure 7), steepness of the histogram revealed the degree of reduction in the NO₂ concentration. Guangzhou city showed high steepness of NO₂ reduction after the pandemic. The maximum (almost 60%) declining rate of NO₂ was observed in Guangzhou, Wuhan, Chengdu, and Beijing; whereas the minimum (almost 40%) declining rate of NO₂ total column in the atmosphere was found in Jinan, Shanghai, Anshan, Zhengzhou, Xinjiang, and Lanzhou.

3.3. Trends of Air Pollutants in China before and after the Pandemic

The overall air quality change in 50 days (

Table 1. Temporal change in air pollutants of 14 cities of China before (4 December 2019 to 22 January 2020) lockdown and after lockdown (23 January to 12 March 2020). The 100-day daily average concentration in total and change in percentage. Total 6 pollutants NO₂, O₃, PM₁₀, PM_2.5, SO₂ (µg/m³), CO (mg/m³) including AQI (change in percentage).

City		Jinan	Guangzhou	Shanghai	Nanjing	Xiamen	Beijing	Zhengzhou	Wuhan	Harbin	Chongqing	Chengdu	Xian	Lanzhou	Anshan
CO	Before	1.46	0.98	0.83	0.97	0.61	0.86	1.28	1.09	1.22	1.02	0.98	1.33	1.87	1.60
	After	0.92	0.73	0.65	0.69	0.44	0.86	0.89	0.91	0.85	0.79	0.67	0.95	1.12	1.37
	Change	−37	−25	−21	−29	−27	0	−31	−17	−30	−22	−31	−28	−40	−14
NO2	Before	56.57	60.86	57.60	53.73	25.75	45.17	54.83	52.73	56.73	46.05	48.70	61.48	65.80	43.11
	After	27.02	27.68	27.68	25.91	13.01	26.31	25.83	19.54	28.89	24.51	21.94	32.29	40.40	30.42
	Change	−52	−55	−52	−52	−49	−42	−53	−63	−49	−47	−55	−47	−39	−29
O3	Before	25.74	43.01	39.73	30.62	64.35	25.44	25.95	23.69	32.37	14.45	23.85	17.08	34.74	33.95
	After	64.18	48.69	71.77	68.07	75.40	49.59	66.24	58.47	70.48	35.36	52.35	51.20	62.74	53.45
	Change	149	13	81	122	17	95	155	147	118	145	120	200	81	57
PM10	Before	141.31	74.48	49.40	80.86	43.07	57.45	121.43	86.54	141.98	83.93	102.22	148.41	107.27	100.39
	After	89.01	33.03	34.84	50.75	30.66	59.60	89.67	47.61	77.55	58.11	65.05	107.01	88.11	94.80
	Change	−37	−56	−29	−37	−29	4	−26	−45	−45	−31	−36	−28	−18	−6
PM2.5	Before	98.66	38.18	53.15	56.17	30.95	44.69	106.48	63.58	129.74	59.86	75.56	116.60	61.27	69.40
	After	62.39	21.92	35.13	36.18	21.22	63.89	72.21	39.08	68.51	44.44	47.27	81.10	40.58	68.14
	Change	−37	−43	−34	−36	−31	43	−32	−39	−47	−26	−37	−30	−34	−2
SO2	Before	21.52	8.64	7.80	7.80	5.31	4.87	10.24	9.46	36.40	8.96	8.22	14.16	29.77	21.01
	After	13.27	5.49	5.66	5.38	4.40	4.53	8.37	7.50	27.25	6.70	6.92	10.51	18.23	18.89
	Change	−38	−36	−27	−31	−17	−7	−18	−21	−25	−25	−16	−26	−39	−10
AQI	Before	131.38	164.22	73.80	79.16	48.00	65.71	141.37	88.17	61.70	82.16	102.16	153.66	88.42	96.03
	After	88.44	92.98	51.59	54.42	36.95	88.12	101.06	56.58	35.91	63.37	68.98	114.82	73.79	92.22
	Change	−33	−43	−30	−31	−23	34	−29	−36	−42	−23	−32	−25	−17	−4

Source: Arranged by author, Ministry of Environmental Protection of China (MEPC), China Ministry of Ecology and Environment, 2020.

) reveals that Harbin recorded the highest reduction (43%) in air pollutants followed by Guangzhou (42%), Wuhan (36%), Jinan (33%), Chengdu (32%), Nanjing (31%), Shanghai (30%), Zhengzhou (29%), Xian (25%), Xiamen (23%), Chongqing (23%), Lanzhou (17%), and Anshan (4%). The most interesting fact in this analysis is that when all the cities were experiencing improved air quality, Beijing recorded deteriorating air quality; it recorded a 34% increase in air pollutants during the 50-day study period after lockdown. In this context, it is important to look into the individual pollutants in detail for an insight into understanding their roles in determining air quality.

During the 50 days of study after lockdown, the concentration of carbon monoxide (mg/m³) reduced by 40% in Lanzhou, 37% in Jinan, 31% in Zhengzhou and Chengdu, 30% in Harbin, 29% in Nanjing, 28% in Xian, 27% in Xiamen, and 25% in Guangzhou (Figure 8b). The temporal changes in NO₂ (Figure 8c) reveal that the more than 50% concentration of NO₂ has reduced in Wuhan (63%), Chengdu and Guangzhou (55%), Zhengzhou (53%), Jinan, Nanjing and Shanghai (52%) while, other cities received below 50% NO₂ concentration (ranging between 20% and 49%).

On the other hand, the results displayed a drastic rise in ozone in the atmosphere in almost all cities after the lockdown. The detailed analysis of O₃ (Figure 8d) explained that Xian experienced a 200% rise in O₃. Moreover, most of cities experienced more than 100% escalation of O₃, for instance, Zhengzhou (155%), Jinan (149%), Wuhan (147%), Chongqing (145%), Nanjing (122%), Chengdu (120%), and Harbin (118%). Interestingly Beijing, the city where the overall quality of air has declined, also experienced a high rise in O₃ (95%). In addition, the concentration of ozone has also increased in Shanghai (81), Lanzhou (81%), and Anshan (57%). Cities such as Xiamen and Guangzhou experienced only 17% and 13% increases in O₃, respectively.

The temporal change in PM₁₀ (μg/m³) (Figure 8e) reveals its sudden decrease in all cities except Beijing after the lockdown (Table 1). The maximum decrease in PM₁₀ concentration was observed in Guangzhou (56%) followed by Wuhan and Harbin (45%). The other cities that observed a substantial decrease in PM₁₀ are Nanjing and Jinan (37%), Chengdu (36%), Shanghai and Xiamen (29%), Xian (28%), and Zhengzhou (26%). Conversely, PM₁₀ concentration increased by 4% in Beijing after the lockdown. Like PM₁₀, the concentration of PM_2.5 also considerably decreased after the pandemic lockdown (Figure 8f). The cities which observed low concentrations of PM_2.5 are Harbin (47%), Guangzhou (43%), Wuhan (39%), Jinan and Chengdu (37%), Nanjing (36%), Shanghai and Lanzhou (34%), Zhengzhou (32%), Xiamen (31%), and Xian (30%). Simultaneously, Beijing experienced a very high concentration of PM_2.5 (43%) after the lockdown.

Along with other pollutants, the temporal change in SO₂ concentration (Figure 8g) displays a substantial decrease in all cities after the lockdown. The cities that experienced a high proportion decrease in SO₂ concentration are Lanzhou (39%), Jinan (38%), Guangzhou (36%), Nanjing (31%), Shanghai (27%), Xian (26%), and Chongqing and Harbin (25%) (Table 1).

The daily average concentration of air pollutants for 100 days (50 days before lockdown and 50 days after lockdown) of 14 cities (Figure 9) reveals that the air quality has remarkably improved in Guangzhou where 43% of pollutants have declined during this study period. The concentration of PM₁₀ and NO₂ has reduced by 56% and 55%, respectively. Similarly, in Shanghai, the overall pollutants in the atmosphere have decreased by 30% (Table 1). In Nanjing, the overall air quality (AQI) has improved by 31%. Moreover, the air quality of Xiamen has improved by 23% and the concentration of NO₂ and PM_2.5 has declined by 49% and 31%, respectively. In Jinan, the overall air quality has improved by 33% and the NO₂ concentration has decreased by 52%. However, the concentration of O₃ has amplified by 149%. In Zhengzhou, the concentration of NO₂ has reduced by 53% and the concentration of O₃ has grown by 155%. In Wuhan, the major pandemic affected area, the concentration of NO₂ has reduced by 63% and PM₁₀ by 45% whereas the concentration of O₃ has increased by 147%. The air quality has also improved in Harbin where the concentration of NO₂ has reduced by 49% and the concentration of O₃ has increased by 118%. In Chongqing, the concentration of NO₂ has reduced by 47% and the concentration of O₃ has improved by 145%. Similarly, in Chengdu, there has been a 55% reduction in NO₂ concentration and a 120% increase in O₃ concentration in the atmosphere. In Xian, there is a 47% change in NO₂ and a 200% increase in O₃ in the atmosphere. In Anshan, the concentration of NO₂ has reduced by 29% whereas the concentration of O₃ has increased by 57%.

4. Discussion

The coronavirus first outbreak started in December 2019 and human-to-human transmission was confirmed on 20 January 2020 [14]. Wuhan city was locked from other regions of the country to stop the spread of the SARS-CoV virus since 23 January 2020. After a few days, the administrative authorities of Guangzhou and Beijing declared the lockdown. The highest level of public health emergency was announced within a week in a few administrative units [13,15]. The anthropogenic emissions from industrial and manufacturing units were closed after the spreading of coronavirus which caused improvement in overall air quality in China [14].

The present study found that there was a substantial decrease in NO₂ concentration and consequently improvement in air quality. It is noteworthy that NO₂ concentration has tremendously decreased in the cities of Guangzhou, Wuhan, and Chengdu. These cities come under the same category, where NO₂ tropospheric column (molecules/cm²) ranges between 1.6^e+16 and 3^e+16 before the pandemic but after the shutdown none of these cities reported high emissions of NO₂. NASA and European Space Agency (ESA) pollution monitoring satellites have also detected a substantial decrease in NO₂ over China due to a complete shutdown. The concentration of nitrogen dioxide was reduced as the sources of this pollutant such as motor vehicles, power plants, and industrial facilities were closed for two months [58]. In addition, the lockdown has coincided with the Lunar New Year celebration which also facilitated to decrease the concentration of NO₂ in all major cities in China’s mainland. Some of the preceding studies have also established that many cities such as Guangzhou, Beijing, Chengdu, Shenzhen, Nanjing, Shanghai, Chongqing, etc., have experienced low concentrations of NO₂ and particulate matter in the atmosphere due to regulatory effects in China [76,77,78,79].

The study found a substantial decline in CO in all cities except Beijing where it remained static. The maximum declining rate of CO was observed in Lanzhou (40%), Jinan (37%), and Zhengzhou (31%) whereas the minimum declining rate of CO was measured in Anshan (14%), Wuhan (17%), and Shanghai (21%) because of the complete halt of vehicles, wood-burning, and industry. Similar to this line, several studies established that vehicles, wood-burning, and industry are the main sources of CO [80,81,82,83,84]. The reduction in CO in 60 days is a significant indicator for air quality because of its 1–2-month life span in the atmosphere, thereby improved quality of air [85,86].

The particulate pollutants such as PM_2.5 and PM₁₀ have also decreased substantially during the lockdown period as the sources of these pollutants were restricted. Several previous studies have documented that the primary sources of PM_2.5 and PM₁₀ are automobile emissions, incomplete combustion, wind-blown soil and dust, construction dust, and biomass burning [82,87,88]. The result displays that the maximum declining rate of PM₁₀ and PM_2.5 was measured in Guangzhou (56% and 43%), Wuhan (45% and 39%), Jinan (37% and 38%), and Nanjing (37% and 36%). Similarly, some other studies have also found that Beijing, Lanzhou, Tianjin, Nanning, and Chongqing reported lower concentrations of these particulate matters during the regulation effect on certain occasions [89,90,91]. On the other hand, the concentration of SO₂ has also declined in all cities due to strict regulation and the closing of heavy industrial factories such as iron, steel, and cement industries [92,93,94,95]. The major sources of SO₂ are coal combustion of biomass in coal-fired power plants and industry sector [9,96,97,98] that were completely locked during the pandemic. Some other studies have exhibited that SO₂, combined with volatile organic compounds (VOCs), enhances the formation of new particles, particularly sulphate which is one of the main components of PM_2.5 [99,100,101,102].

Consistent with previous studies, this study [103,104,105] also found that due to the decrease in all air pollutants including CO, NO₂, PM_2.5, PM₁₀, and SO₂ the total concentration of O₃ has increased in all 15 cities in China during the study period. The maximum increasing rate of O₃ was found in Xian (200%), Zhengzhou (155%), Jinan (149%), and Chongqing (145%) whereas the minimum increasing trend was found in Guangzhou (13%), Xiamen (17%), and Anshan (57%). Several reasons have been proposed to explain this complex relation: (a) inverse relation between O₃ and its precursors NO₂, PM_2.5, PM₁₀, and SO₂ lower emission of these pollutants results in faster ozone production in nitrogen oxide (NO_x) concentrated areas [106,107,108,109]; (b) low concentration of NO_x also causes less destruction of ozone [88,110]; (c) decreased level of atmospheric pollutants leads to more clear sunshine that also accentuates more ozone production [108,111]. In addition, a pause of vehicles also resulted in a reduction in CO and NO_x including NO and NO₂ facilitating an increase in the level of ozone in the atmosphere [81,112,113,114]. The production of ozone under the influence of anthropogenic activities in the troposphere and involving catalysis by NO₂ and NO should be significant [22]. Its precursor compounds NO_x and VOC have a wide variety of sources and can exhibit a non-linear effect on ozone production, while its accumulation is strongly influenced by meteorological parameters [115]. Although reductions in atmospheric ozone allow more solar radiation to reach ground level, resulting in higher surface temperatures, a decrease in the downward longwave radiation emitted by CO₂, O₃, and H₂0 from a cooler lower stratosphere with less ozone would result in a decrease in surface temperatures [22].

One of the major outcomes of reduced atmospheric pollutants is a substantial improvement in overall air quality except for Beijing. Since most of the industrial production, vehicle movements, and other anthropogenic activities in the cities were closed, the reduced level of NO₂, CO, PM_2.5, PM₁₀, and SO₂ improved the air quality in these cities. Similar to this, several previous studies have found that air quality is positively related to the low concentration of atmospheric pollutants [116,117,118]. Conversely, Beijing has reported a decline in the overall air quality during the study period despite the closure of manufacturing units. The reason may be the increasing level of PM_2.5 and PM₁₀ during the same period. The concentration of PM_2.5 has increased manyfold in Beijing because the city experienced a number of severe haze events during the lockdown period. Some of the previous studies [25,31,79] mentioned that complex relationships among changes in relative humidity, near-surface wind speed and direction, planetary boundary layer height, and precipitation have influenced the increasing concentration of PM_2.5 in Beijing. Climatologically, Beijing has dry air during the wintertime, but a larger than normal amount of moisture accumulated near the surface during the lockdown period. This has facilitated multiphase reactions for aerosol formation and growth [25]. Moreover, wind conditions also facilitate to formation of haze in Beijing as the mean wind speed declined. In addition, the wind direction changed to southerly which usually carries polluted air from Hebei Province’s industrial regions. Moreover, the planetary boundary layer height in northern China also declined during the lockdown and this lower height facilitated stagnant air and subsequently resulted in increasing PM_2.5 in Beijing. Further, during the lockdown period, precipitation mainly occurred in southern China and the Northern part did not receive enough rain to wash out the haze that formed over the region [40].

5. Conclusions

The assessment of major atmospheric pollutants during lockdown due to the COVID-19 pandemic has markedly influenced the air quality in China. The cities such as Guangzhou, Wuhan, and Chengdu that have shown a very high concentration of NO₂ tropospheric column (molecules/cm²) before the pandemic and have experienced a sudden decline after the lockdown. The air quality zonal model has also displayed about 390,150 km² areas that were under the high level of air pollutants have come under low concentration areas. In addition, the spatio-temporal evaluation of the NO₂ and other main pollutants in the major 15 cities of China exhibited a remarkable reduction which in turn, facilitated to improve the overall air quality except in Beijing where the air quality has been degraded due to the overwhelming concentration of PM_2.5 and PM₁₀.

The overall air quality of eastern coastal and industrial cities such as Jinan, Zhengzhou, Wuhan, Guangzhou, Chengdu, and Xian have recorded low concentrations of air pollutants and such reduction in atmospheric pollutants has improved the air quality. The overall air quality change in 50 days reveals that Harbin has recorded the highest reduction (43%) in air pollutants followed by Guangzhou (42%), Wuhan (36%), Jinan (33%), Chengdu (32%), Nanjing (31%), Shanghai (30%), Zhengzhou (29%), Xian (25%), Xiamen (23%), Chongqing (23%), Lanzhou (17%), and Anshan (4%). Since most of the industrial productions, vehicle movements, and other anthropogenic activities in the cities were restricted, the study revealed that the levels of NO₂, CO, PM_2.5, PM₁₀, and SO₂ have remarkably reduced. It is noteworthy that the reduction in atmospheric pollutants and consequently improved air quality would have a positive impact on the environment and human health. To address the emission of atmospheric pollutants, particularly NO₂, CO, PM_2.5, PM₁₀, and SO₂, the government needs to establish an environmental policy framework involving central, provincial, and local governments with stringent laws for environmental protection. The restriction and regulation of all anthropogenic sources of pollutants due to the COVID-19 outbreak was an example of such stringent enforcement of the law to protect human health. Similarly, to protect the environment from atmospheric pollutants strong political commitment, technological development, and policy enforcement are essential to make policies, such as 11th and 12th five-year plans of China, successful towards improving and protecting the environment and human health.