Relationship between PM2.5 pollution and firms’ emissions in Shaanxi Province, China

The relationship between the high-frequency time series of PM_2.5 in the atmosphere and the air pollutants emitted by industrial firms is not yet fully understood. This study aimed to identify independent PM_2.5 clustering regions in Shaanxi Province and to evaluate the spatio-temporal correlations of PM_2.5 concentrations and pollutant emissions from industrial firms in these regions. To accomplish this, daily data on PM_2.5 concentrations and air pollutants emitted by industrial firms were analyzed using the K-means spatial clustering method and cross-wavelet transformation. The results show that: 1) PM_2.5 concentrations in Shaanxi Province can be divided into three independent clustering regions. 2) The lagged impact of industrial emissions on PM_2.5 concentrations were about 1/4-1/2 period. 3) PM_2.5 was mainly influenced by particulate matter (PM) emissions from industrial plants during the period of 16–32 days, while nitrogen oxides (NO_x) significantly affected PM_2.5 concentrations during the period of 32–64 days. 4) Emissions of PM, NO_x, and sulfur dioxide (SO₂) more significantly affect PM_2.5 concentrations in northern and central Shaanxi, and pollutants emitted by firms in the thermal power generation, utility, and steel industries had more significant effects on PM_2.5 concentrations than those emitted by the cement manufacturing and electric power industries. During the COVID-19 shutdown, the emissions of firms cannot significantly affect PM_2.5 concentrations. These findings suggest that emission reduction initiatives should consider industrial, regional, and periodic differences to reduce PM_2.5 pollution during winter.

1 Introduction

Exposure to high level of fine particulate matter (PM_2.5) in the atmosphere has been associated with severe adverse effects on human health (Barwick et al., 2017; Zhong et al., 2017; Chen et al., 2018; Liu and Salvo, 2018; Cheung et al., 2020; Fan et al., 2020; He et al., 2020), as well as substantial economic losses (Zhang and Mu, 2018; Sun et al., 2019; Chen et al., 2022; Ito and Zhang, 2020). While efforts have been made to manage PM_2.5 concentrations in developing countries like China, with notable progress in recent decades (Greenstone et al., 2021), the annual average PM_2.5 concentrations in China’s 339 cities are still expected to exceed the recommended level by the World Health Organization (MEE, 2022; WHO, 2021). Moreover, severe haze pollution remains a common occurrence in many Chinese cities during autumn and winter (Ma et al., 2020; Zhang et al., 2020b).

To effectively reduce ambient PM_2.5 concentrations, it is crucial to identify the firms that should reduce emissions during periods of heavy pollution as well as during normal times (F. Wang et al., 2019; Zhao et al., 2018; Alari et al., 2021; Frankowski, 2020; Rivera, 2021), for the industrial sector significantly contributes to China’s economic growth but also emits a large amount of air pollutants while consuming fossil energy (Choi et al., 2021; MEE, 2020; Zhang et al., 2014). Various studies have analyzed the correlation between PM_2.5 concentrations and yearly air pollutant emissions using a physicochemical model (Choi et al., 2021; Lu et al., 2021; Wang et al., 2022; Liu et al., 2018b; Zhang et al., 2014; Zhang et al., 2018), likely multiple regression (Zhang et al., 2020), geographically weighted regression model (Wang et al., 2018; Tu et al., 2019) and physicochemical model such as CMAQ (Wang et al., 2022b; Wang et al., 2019b), WRF-Chem (Spiridonov et al., 2019; Azmi et al., 2022; Zhang et al., 2023). Although an increasing body of research begins to concentrate on the interplay between business emissions and air quality (Li et al., 2016; Wu et al., 2023), there has been limited research conducted on the relationship between high-frequency scale (such as a week, day, or even an hour) air pollutant emissions from industrial firms and PM_2.5 concentrations, which can help the government formulate proper strategies for reducing industrial emissions.

In order to better capture the high-frequency variability of PM_2.5, some researchers have employed the wavelet analysis technique to identify the temporal patterns of PM_2.5 changes (Chen et al., 2020; Kapwata et al., 2021; Li et al., 2017; Shi et al., 2014; Sun et al., 2017; Zhao et al., 2009; Zhao et al., 2016). Wavelet analysis has been shown to be effective in studying the periodicity and evolution characteristics of PM_2.5 concentrations, as well as investigating the influence of natural factors on PM_2.5 concentrations. Moreover, PM_2.5 pollution exhibit significant regional variations that often do not align with administrative divisions (Tie et al., 2005; Tie and Cao, 2009; Wang et al., 2011; Zhao et al., 2020), the PM_2.5 pollution areas must be identify before appropriate mitigation strategies can be established.

To fill the knowledge gap mentioned above, this study selects Shaanxi Province in China as a relatively independent study area to examine the correlation between firms’ emissions and PM_2.5. For the unique topography Guanzhong Plain is one of China’s most polluted region for PM_2.5 (Xu et al., 2018; Li et al., 2022); Southern Shaanxi has a good ecology; Northern Shaanxi has many polluting firms) and diverse types of firms (Miao et al., 2019; Wang et al., 2022c) in Shaanxi Province can provide effective lessons for pollutant management in other regions and countries. We collected a unique daily scale emission data from all continuous emission monitoring system (CEMS) installed firms in Shaanxi Province and air quality monitor site data at county level.

This paper are to make the following contributions to the field: 1) Establish a methodology for identifying PM_2.5 pollution control areas using spatial clustering techniques; 2) Identify firms and industries with a significant impact on PM_2.5 pollution areas by utilizing cross-wavelet analysis on high-frequency time scale data; 3) Provide evidence-based support for precisely requiring firms to adopt emission reduction measures during periods of heavy pollution, and effectively reducing socio-economic losses.

The rest of this paper is structured as follows. Part 2 introduces the study area, data, and methodology. Part 3, this study shows cluster and wavelet analysis. In the final part, this paper provides some concluding remarks.

2 Study region, data resource, and methodology

2.1 Study region and data resource

Shaanxi Province, in northwestern China, spans between 105°29′E-111°15′E longitude and 31°42′N-39°35′N latitude, characterized by a complex and diverse terrain. Shaanxi Province, the most affluent among the five northwestern provinces of China, recorded a GDP of 2,980.1 billion RMB in 2021 (National Bureau of Statistics, 2023). The industrial sector played a substantial role, contributing 46.3% to the Shaanxi’s GDP. With a diverse industrial landscape encompassing high-end energy and chemical sectors, equipment manufacturing, aerospace, electronic information, and automobile production, Shaanxi Province’s industrial firms collectively consumed 92.8 million tonnes of standard coal in 2021. This consumption constituted 68.1% of the total standard coal usage within the province. Such considerable energy consumption consequently led to substantial pollutant emissions. Notably, industrial sources in Shaanxi Province contributed to over 60% of total emissions for both Total Suspended Particulates (TSP) and Sulfur Dioxide (SO₂) pollutants in 2021 (Ministry of Ecology and Environment, 2021).

Shaanxi can be geographically divided into three parts: southern Shaanxi, Guanzhong and northern Shaanxi. Northern Shaanxi, which encompasses the cities of Yulin and Yan’an, is abundant in coal, oil, and natural gas resources and houses several large-scale energy chemical firms, including thermal power, coking, coal chemical, and petrochemical. The Guanzhong plain, comprising cities such as Xi’an, Xianyang, Weinan, Baoji, and Tongchuan, is recognized as one of China’s worst air pollution areas in China due to a significant number of thermal power, cement, and steel sectors (Bai et al., 2019). The Southern Shaanxi region, including the cities of Hanzhong, Shangluo, and Ankang, is rich in rare mineral resources, such as molybdenum, rhenium, and mercury, as well as abundant natural resources.

Figure 1 demonstrates the distribution of 169 air quality monitoring stations and 361 firms in Shaanxi Province. Specifically, Guanzhong has 60 air quality monitoring stations and 150 industrial firms, Northern Shaanxi has 59 air quality monitoring stations and 177 industrial firms, while Southern Shaanxi has 50 air quality monitoring stations and 34 industrial firms. Shaanxi’s average PM_2.5 concentrations has decreased from 57 to 43 μg/m³ between 2017 and 2020. However, the current concentration remains above the national standard limit of 30 μg/m³.

FIGURE 1

FIGURE 1. Study area of Shaanxi Province.

The monitoring data of PM_2.5 concentrations are from the ambient air quality monitoring stations in Shannxi Province (http://113.140.66.226:8024/sxAQIWeb/pagecity.aspx?cityCode=NjEwMTAw), and meteorological data including wind speed, wind direction, air temperature, air pressure, and humidity are from Shaanxi air quality real-time release system. The research period is from 1 January 2017 to 31 December 2020. The firms’ emissions data come from key pollutant discharge firms’ monitoring information release platforms (http://113.140.66.227:9777/envinfo_ps/zdyjbxxpublicity/list), including the three most important pollutants, TSP (total suspended particles), SO₂ (sulfur dioxide) and NOx (nitrogen oxide). The pollutant emissions of the firms installing CEMS exceed 65% of the total pollutant emissions in the region (The State Council, 2007), which basically reflects the regional industrial pollutant emissions.

For air quality monitoring data, we remove outliers in the data and ensure the accuracy of the data by using data from neighbouring air quality monitoring points. For firm emission data, we firstly removed extreme outliers in the firms CEMS data; secondly, we determined whether the firms was in a shutdown state through the firms’ flue gas oxygen content, and for missing values in a non-shutdown state, we filled in the pollutant emissions of the firms neighboring time through the missing values to ensure the validity of the data.

2.2 Methodology

2.2.1 Spatiotemporal cluster analysis

PM_2.5 concentrations exhibits spatial clustering on a daily basis. To identify regions that require targeted control during the study period, daily partitioning results of the clustering process are subjected to correlation clustering analysis. If the external transfer of pollutants is not taken into account, PM_2.5 concentrations in the identified areas can be reduced by reducing all types of pollutant emissions, such as those from industrial, mobile and non-organized sources. Thiessen polygons are generated from the monitoring stations to ensure the continuity of the identified pollution area and to avoid the limitations imposed by administrative divisions. The resulting clustering categories are consistent with the geographical division, and further refine the control area for PM_2.5 while smoothing out the boundary of the region.

This paper adopts the K-means clustering method, a fast clustering approach that abstracts clustering units into m-dimensional points and evaluates the similarity between units based on the distance between clustering units (Liu et al., 2018a), to conduct spatial clustering of PM_2.5:

1) Optional K initial clustering centers: Z₁(1), Z₂(1), …, Z_k(1), 1 is the subsequence of iteration.

2) The rest of the samples are allocated to one of the K cluster centers according to the principle of minimum distance:

$$

X∈S_i(k),k is the subsequence of iteration; K is the number of cluster centers.

3) Compute the new vector value of each cluster center: Z_j (k+1) j=1, 2, … , K

$Z_j\left(k+1\right)=\frac{1}{N_j}{\displaystyle \sum _{X\in S_j\left(k\right)}}X,j=1,2,\dots ,K(2)$

4) If Z_j(k+1) $\ne$ Z_j(k), return formula (1), classify the pattern samples center by center, and repeat the iterative calculation. If Z_j(k+1)=Z_j(k), algorithm convergence, calculation completes (Wegner et al., 2012).

2.2.2 Wavelet and cross-wavelet analysis

The wavelet transform is a powerful method for the simultaneous analysis of time series in both the time and frequency domains. It involves the use of a variable window function and is generated from mother and child waves through time translation and scaling, as noted by Gao and Zhang (2016) and Mu et al. (2021). This decomposition results in a set of essential functions where increasing the scale is equivalent to decreasing the frequency and sacrificing the time resolution, while reducing the scale and increasing the frequency is equivalent to sacrificing the frequency resolution. By stretching and translational transform, the wavelet function is obtained from the wavelet generator function. The continuous wavelet function is expressed as:

$W_n^X\left(s\right)=\sqrt{\frac{\delta t}{s}}{\displaystyle \sum _{n^{\prime }=0}^{N-1}}{x}_{n^{\prime }}{\psi }^{*}\left[\frac{\left(n^{\prime }-n\right)\delta t}{s}\right](3)$

where^* represents a conjugate complex, N is the total number of the time series, (δ_t/s)/s is a factor used for the standardization of wavelet function.

The Morlet wavelet function is commonly used in wavelet analysis as it effectively separates and reconstructs waves of different frequency bands without losing time resolution. This function maintains its shape through frequency shift and has excellent temporal aggregation and high-frequency resolution (Morlet et al., 1982). Therefore, Morlet wavelet function is employed to analyze the periodicity of PM_2.5 concentrations:

${\psi }_0\left(t\right)={\pi }^{-1/4}{e}^{i{\omega }_{0}t}{e}^{-t^2/2}(4)$

Where t is time, ${\omega }_0$ is dimensionless frequency. If ${\omega }_0$ =6, scale s equals Fourier period. Wavelet power spectrum ${\left|W_n^X\left(s\right)\right|}^2$ shows the variation characteristics of the time series on a specific scale and its variation with time. This unbiased and consistent estimation of the real power spectrum of time series is achieved through the use of the full wavelet spectrum (Torrence and Compo, 1998; Torrence and Webster, 1999).

${\overline{W}}^2\left(s\right)=\frac{1}{N}{\displaystyle \sum _{n=o}^{N-1}}{\left|W_n\left(s\right)\right|}^2(5)$

The boundary effects and errors encountered when handling finite time series are addressed by testing the statistical significance of the wavelet power spectrum using an appropriate equation (Cazelles et al., 2008; Furon et al., 2008; Grinsted et al., 2004; Liu et al., 2018a). Therefore, the test measure display as follows:

$P_k=\frac{1-{\alpha }^2}{{\left|1-\alpha e^{-2i\pi k}\right|}^2}(6)$

Cross-wavelet analysis can compare the frequencies of two-time series X_t and Y_t, and obtain the resonance period and phase of the two sequences in some periods. Cross wavelet spectra of two-time series XWT is $W^{XY}=W^X{W}^{Y*}$, cross wavelet power is $\left|W^{XY}\right|$. The period changes with time, reflected by the cross wavelet power can be visualized by the full spectrum (Veleda et al., 2012). The confidence degree of cross wavelet power can be calculated by the square root of the product of two chi-square distributions. The confidence and correlation coefficient of the two-time series are as follows:

$D\left(\frac{\left|W_n^X\left(s\right)W_n^{Y*}\left(s\right)\right|}{{\sigma }_X{\sigma }_Y}<P\right)=\frac{Z{}_v\left(p\right)}{v}\sqrt{P_k^X{P}_k^Y}(7)$

$R_n^2\left(s\right)=\frac{{\left|S\left(W_n^{XY}\left(s\right)\right)\right|}^2}{S\left({\left|W_n^X\left(s\right)\right|}^2\right)•S\left({\left|W_n^Y\left(s\right)\right|}^2\right)}(8)$

The phase difference can be transferred to the difference of [-π, π], and its value indicates the lag characteristics of the two sequences (Aguiar-Conraria and Joana Soares, 2011; Addesso et al., 2022). The phase difference formula is as follows:

${\phi }_{xy}=\mathit{arctan}\left(\frac{\Im \left(W_n^{XY}\left(s\right)\right)}{\Re \left(W_n^{XY}\left(s\right)\right)}\right)(9)$

3 Results analysis

3.1 Spatiotemporal distribution of PM_2.5 in study region

The seasonal distribution of PM_2.5 concentrations in each region is relatively even, with the highest levels in winter and the lowest in summer (Figure 2). During the winter and spring seasons, the PM_2.5 concentrations in the three regions display significant variations, ranging from 65 to 95 μg/m³ and 29.5–47 μg/m³, respectively. This indicates that the concentrations during winter are approximately twice as high as those during spring. Conversely, PM_2.5 concentrations slightly vary among the three regions during the summer and autumn seasons. Moreover, meteorological factors such as temperature, wind speed, and relative humidity significantly promote the diffusion of atmospheric pollutants. In mid-March, the cessation of heating results in a reduction of pollution sources, leading to lower PM_2.5 concentrations during the spring season than during the winter season.

FIGURE 2

FIGURE 2. Spatiotemporal distribution of PM_2.5 concentrations.

Annual PM_2.5 concentrations in the three regions follow the order of Guanzhong > Northern Shaanxi > Southern Shaanxi. Guanzhong is the most economically developed region in Shaanxi Province and has several large coal-fired power plants, cement manufacturers, chemical plants, and metal smelters, which contribute to high PM_2.5 concentrations. The Guanzhong area’s proximity to the Qinling Mountains in the South and the Loess Plateau in the North further hinders airflow movement and pollutant diffusion, thereby intensifying the pollutant concentration (Wang et al., 2015). In Northern Shaanxi, several coal-fired power plants and factories discharge pollutants, and the unique linear canyon terrain in Yan’an City tends to converge local pollutants in the area. Southern Shaanxi primarily relies on natural resources to develop the primary industry, has low urbanization, and has a less prominent contribution of firm emission sources to PM_2.5 concentrations.

3.2 Clustering results of PM_2.5 in study region

From Figure 2, it is obvious shows that PM_2.5 pollution in Shaanxi Province has a significant regional feature, in order to accurately identify each PM_2.5 pollution region in Shaanxi Province, so as to more accurate identification of the link between pollutant emissions from firms and PM_2.5 in the air, this part use the PM_2.5 data of the air monitoring points in the study time period, and the method of spatial clustering to perform spatial clustering. Figure 3 depicts the spatial clustering of air pollution areas in Shaanxi Province. By overlaying the zoning results onto the elevation map of Shaanxi Province, it becomes apparent that delimited regions of PM_2.5 overlap with the locations of the Guanzhong Plain, Northern Shaanxi, and Southern Shaanxi, suggesting that PM_2.5 concentrations are influenced by the topography. Three regions were identified based on their coverage area and PM_2.5 concentration levels. Region 1 covers most of the Guanzhong area, region 2 includes the northern Shaanxi area and a portion of the Guanzhong area, while region 3 comprises the Southern Shaanxi area and a portion of the Guanzhong area. Region 1 had a mean PM_2.5 of 57.4 μg/m³, and 335 days per year exceeded 24-h PM_2.5 thresholds (<75 μg/m³), representing 20.41% of total days. In Region 2, the mean PM_2.5 was 40.3 μg/m³, with 151 days exceeding the 24-h PM_2.5 benchmark, representing 9.2% of the total number of days. Finally, the mean PM_2.5 concentration in region 3 was 33.2 μg/m³, with 79 days exceeding the 24-h PM_2.5 benchmark, representing 4.81% of the total number of days.

FIGURE 3

FIGURE 3. Spatial clustering PM_2.5 in Shaanxi.

3.3 Air pollutant emissions from industrial firms in study region

The analysis results presented in Figure 4 shows the variations in pollutant emissions among industrial firms in Shaanxi Province. Figure 4 shows that region 1 and region 2 emit about the same amount of TSP pollutants, while region 2 emits a much larger amount of SO₂ and NOx than region 1 and region 3. This emission situation is closely related to the distribution of industries in different regions. This difference was attributed to the lots of power plants located in region 2, which accounts for 50% of the power plants in Shaanxi Province and contributes to over 70% of total pollutant emissions from power plants. Region 1 is distributed with many processing and manufacturing firms, which mainly emit TSP pollutants, making the emissions of TSP pollutants in region 1 and region 2 similar, although there are significant differences in SO₂ and NOx emissions. Simultaneously, it can be observed that TSP emissions in region 3 saw a notable increase during the COVID-19 period in 2020. In contrast, emissions in regions 1 and 2 experienced a certain degree of decline. There’s a reasonable suspicion that regions 1 and 2 redirected a portion of their production capacity to region 3, which may account for this dynamic change.

FIGURE 4

FIGURE 4. Daily variation curves of air pollutants, including TSP, SO_2, and NO_x, emitted by industrial firms from 2017 to 2020.

3.4 Analysis of cross-wavelet in different regions

This part we employ the Morlet wavelet function to conduct a wavelet transform of daily data on TSP, SO₂, and NO_x emissions from industrial firms in three regions of Shaanxi Province from January 2017 to December 2020. The wavelet spectrums of TSP, SO₂, and NO_x are analyzed, with time d as the horizontal axis and time scales as the vertical axis. A wavelet coefficient of 0 signifies a mutation point wherein the correlation coefficient shifts from higher to lower values. The white shadow area in the wavelet spectrum represents the influence cone of the wavelet boundary, while the area enclosed by thick white solid lines indicates that the noise test has passed at a 95% significance level.

Figure 5 shows that the pollutants emitted by industrial firms during the winter prevention period (November to February) in the three regions demonstrate substantial periodicities of 16–64 and 64–128 days, with the impact of contaminants on PM_2.5 concentrations lagging 1/4-3/8 cycles. The TSP and PM_2.5 manifest a strong correlation in the 16–32 days range, followed by NO_x and PM_2.5, while SO₂ and PM_2.5 exhibit the weakest correlation. These results coincide with the fact that TSP directly influences PM_2.5 concentrations in the atmosphere, while NOx and SO2 necessitate a chemical reaction to generate nitrate and sulfate and form secondary particles. Simultaneously, the figure shows a notable correlation between pollutant emissions from firm and PM_2.5 concentrations within short timeframes in both region 1 and region 2. This strong correlation strongly suggests that firms’ pollutant emissions play a pivotal role in driving fluctuations in PM_2.5 concentrations, particularly during the winter months. However, this short-term influence on PM_2.5 concentrations as a result of corporate emissions is not evident in region 3.

FIGURE 5

FIGURE 5. Spatiotemporal correlation between PM_2.5 concentrations and industrial firms’ emissions in different regions of PM_2.5 pollution. Note: region 1 represents Guanzhong, region 2 represents Northern Shaanxi, and region 3 represents Southern Shaanxi. (A,D,G) Region 1 TSP, Region 2 TSP, Region 3 TSP; (B,E,H) Region 1 S0₂, Region 2 S0₂, Region 3 S0₂; (C,F I) Region 1 NOx, Region 2 NOx, Region 3 NOx.

Figure 5 also shows that the emissions from industrial firms in region 1 strongly resonate with PM_2.5 from November to February 2017 and 2018. However, during a short period of 32–64 days, the emissions negatively related with PM_2.5 concentrations, implying that changes in PM_2.5 concentrations lag behind alterations in TSP emissions from firms. In the winter of 2019, there is a significant correlation between firm emissions and PM_2.5 concentrations, but no significant correlation is found in the spring of 2020 since industrial emissions reduced during the COVID-19 lockdown, which caused an unprecedented cessation of human activities that affected China’s industrial production and pollutant emissions. Therefore, controlling industrial emissions can alleviate PM_2.5 pollution in Shaanxi.

3.5 Analysis of cross-wavelet in different industries

Using region 1 as an example, the study found a strong correlation between PM_2.5 concentrations and pollutants emitted by thermal production, supply, and steel industries within a short timeframe of 8–32 days. However, the correlation coefficient displays irregular fluctuations (Figure 6). Conversely, there was a weak correlation between PM_2.5 and emissions from cement and power sectors. During a medium to long timeframe of 32–64 days, TSP emissions from heat production and supply firms in 2018 and 2019 exhibited correlations with PM_2.5 concentrations. Moreover, SO₂ positively related PM_2.5 concentrations in 2018 and 2019, while NO_x and PM_2.5 displayed positive correlations in 2019 and 2020. Between 2017 and 2019, TSP, SO₂, and NO_x emissions from cement manufacturing firms demonstrated strong positive correlations with PM_2.5. In contrast, there were weak correlations between PM_2.5 concentrations and pollutants emitted by firms in the power and iron-steel industries. Within a long timeframe of 64–128 days, the correlation between emissions in the above four industries and PM_2.5 concentrations gradually weakened, except for significant relationships between PM_2.5 and NO_x emitted by thermal production and supply firms.

FIGURE 6

FIGURE 6. Spatiotemporal correlation between PM_2.5 concentration and the air pollutants emitted by different industrial firms in region1 (Guanzhong). (A,D,G,J) TSP steel industry, TSP thermal power industry, TSP cement manufacturing, PM heat production and supply; (B,E,H,K) S0₂ steel industry, S0₂ thermal power industry, S0₂ cement manufacturing, S0₂ heat production and supply; (C,F,I,L) NO_x steel industry, NO_x thermal power industry, NO_x cement manufacturing, NO_x heat production and supply.

3.6 Spatiotemporal correlation between PM_2.5 concentrations and meteorological factors

In addition to firms’ emissions, meteorological factors are also an important driver of PM_2.5 concentrations (Jia et al., 2020). This section takes region 1 as an example to further analyse whether meteorology further influences the changes in PM_2.5 concentrations on basis of emissions from firms. Specifically, the study examines the periodic effects of atmospheric pressure, temperature, humidity, wind speed, and sunshine on PM_2.5 using time-series data and Morlet cross-wavelet transform analysis. Figure 7 shows that the PM_2.5 concentration has a negative correlation with atmospheric pressure, humidity, wind speed, and sunshine on an 8–32 days time scale, while no significant correlations are observed with air temperature. In addition, the degree of correlation between PM_2.5 concentration and wind direction, air pressure, and humidity during the winter defense period in 2017–2018 and 2018–2019 is significantly higher than that in other years. In a certain extent, it can be inferred from Figures 5A–C that a connection exists between the heightened responsiveness of firms’ emissions to ambient PM_2.5 and alterations in meteorological conditions. This dynamic suggests that firms emit comparable levels of pollutants but contribute to more PM_2.5 pollution, possibly due to changing meteorological factors.

FIGURE 7

FIGURE 7. Spatiotemporal correlation between PM_2.5 concentrations and meteorological factors in region 1 (Guanzhong). (A) Region 1 pressure, (B) Region 1 temperature, (C) Region 1 humidity, (D) Region 1 sunshine hours, (E) Region 1 wind speed.

4 Conclusion

This study utilizes data of PM_2.5 concentrations collected from 169 air quality monitoring stations and 361 industrial firms situated in all county-level cities within Shaanxi Province. The spatial cluster analysis method is used to determine the clustering region of PM_2.5, while the spatiotemporal correlation between PM_2.5 concentrations and industrial firm emissions in different smog-contaminated areas is evaluated using the cross-wavelet analysis method.

The findings of this study indicate that the mean PM_2.5 concentrations in Shaanxi Province decreased from 50.2 to 38.1 μg/m³ between 2017 and 2020, which can be attributed to the 3-year plan to control haze and improve air quality. The mean PM_2.5 concentrations in summer and winter were 38.9 μg/m³ and 77.6 μg/m³, respectively, with higher concentrations observed during winter. Additionally, the mean PM_2.5 concentrations in Guanzhong, Northern Shaanxi, and Southern Shaanxi were 57.4, 40.3, and 33.2 μg/m³, respectively, with Guanzhong having a higher PM_2.5 concentration than Southern Shaanxi.

The K-means clustering method is employed to cluster daily PM_2.5 concentrations in winter, and three pollution areas are identified based on ground elevation information and geographical subregions of Shaanxi Province. PM_2.5 concentrations in the three regions of Shaanxi Province during winter show periodic changes, with pollutants emitted by firms lagging about 1/4-1/2 of the period. In the 16–32 days period, PM_2.5 concentrations are significantly affected by PM, followed by NO_x and SO₂. NO_x has significantly affected PM_2.5 concentrations in the 32–64 days period, while in the 64–128 days period, the impact of NO_x and SO₂ on PM_2.5 concentrations is similar. Industrial firms in regions 1 and 2 significantly affect local PM_2.5 concentrations, while the impact of industrial firms in region 3 is minor.

We also found during the COVID-19 lockdown, industrial firm emissions, excluding those in the heat production and supply industry, had no significant impact on PM_2.5 concentrations, indicating that controlling air pollutants emitted from firms can alleviate PM_2.5 in Shaanxi. Furthermore, pollutants emitted by firms in the thermal production, supply, and steel industries have a greater impact on PM_2.5 concentrations than those from the cement manufacturing and power industries.

The study findings suggest that the government should focus on reducing and eliminating backward production capacity, strengthening discharge control of thermal power production, supply, and steel-iron industries, and promoting industrial structure optimization and upgrading in Shaanxi Province to reduce PM_2.5 concentrations during winter. Specifically, for the Guanzhong and northern Shaanxi regions, there is a particular need to control pollutant emissions from firms under unfavourable meteorological conditions in order to effectively mitigate the severe PM2.5 pollution caused by firms’ emissions. For the southern Shaanxi region, there is a need to prevent firms in the northern Shaanxi and Guanzhong regions from relocating or relocating some of their production orders to southern Shaanxi.

Data availability statement

The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author.

Author contributions

JZ: Data curation, Formal Analysis, Methodology, Software, Writing–original draft. LY: Formal Analysis, Writing–original draft. CJ: Visualization, Writing–review and editing. PG: Visualization, Writing–review and editing.

Acknowledgments

We would like to acknowledge the reviewers for their valuable comments.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

References

Addesso, R., Bellino, A., and Baldantoni, D. (2022). Underground ecosystem conservation through high-resolution air monitoring. Environ. Manage. 69, 982–993. doi:10.1007/s00267-022-01603-0

PubMed Abstract | CrossRef Full Text | Google Scholar

Aguiar-Conraria, L., and Joana Soares, M. (2011). Business cycle synchronization and the euro: A wavelet analysis. J. Macroecon. 33, 477–489. doi:10.1016/j.jmacro.2011.02.005

CrossRef Full Text | Google Scholar

Alari, A., Schwarz, L., Zabrocki, L., Le Nir, G., Chaix, B., Benmarhnia, T., et al. (2021). The effects of an air quality alert program on premature mortality: A difference-in-differences evaluation in the region of paris. Environ. Int. 156, 106583. doi:10.1016/j.envint.2021.106583

PubMed Abstract | CrossRef Full Text | Google Scholar

Azmi, S., Sharma, M., and Nagar, P. K. (2022). NMVOC emissions and their formation into secondary organic aerosols over India using WRF-Chem model. Atmos. Environ. 287, 119254. doi:10.1016/j.atmosenv.2022.119254

CrossRef Full Text | Google Scholar

Bai, K., Ma, M., Chang, N.-B., and Gao, W. (2019). Spatiotemporal trend analysis for fine particulate matter concentrations in China using high-resolution satellite-derived and ground-measured PM2.5 data. J. Environ. Manage. 233, 530–542. doi:10.1016/j.jenvman.2018.12.071

PubMed Abstract | CrossRef Full Text | Google Scholar

Barwick, P. J., Li, S., Rao, D., and Zahur, N. (2017). Air pollution, health spending and willingness to pay for clean air in China. SSRN Electron. J. doi:10.2139/ssrn.2999068

CrossRef Full Text | Google Scholar

Cazelles, B., Chavez, M., Berteaux, D., Ménard, F., Vik, J. O., Jenouvrier, S., et al. (2008). Wavelet analysis of ecological time series. Oecologia 156, 287–304. doi:10.1007/s00442-008-0993-2

PubMed Abstract | CrossRef Full Text | Google Scholar

Chen, S., Guo, C., and Huang, X. (2018). Air pollution, student health, and school absences: evidence from China. J. Environ. Econ. Manage. 92, 465–497. doi:10.1016/j.jeem.2018.10.002

CrossRef Full Text | Google Scholar

Chen, C., Sun, Y., Lan, Q., and Jiang, F. (2020). Impacts of industrial agglomeration on pollution and ecological efficiency-A spatial econometric analysis based on a big panel dataset of China’s 259 cities. J. Clean. Prod. 258, 120721. doi:10.1016/j.jclepro.2020.120721

CrossRef Full Text | Google Scholar

Chen, Z., Wang, F., Liu, B., and Zhang, B. (2022). Short-term and long-term impacts of air pollution control on China’s economy. Environ. Manage. 70, 536–547. doi:10.1007/s00267-022-01664-1

PubMed Abstract | CrossRef Full Text | Google Scholar

Cheung, C. W., He, G., and Pan, Y. (2020). Mitigating the air pollution effect? The remarkable decline in the pollution-mortality relationship in Hong Kong. J. Environ. Econ. Manage. 101, 102316. doi:10.1016/j.jeem.2020.102316

PubMed Abstract | CrossRef Full Text | Google Scholar

Choi, W. J., Jung, B., Lee, D., Kang, H., Kim, H., Hong, H., et al. (2021). An investigation into the effect of emissions from industrial complexes on air quality in the ulsan metropolitan city utilizing trace components in PM2.5. Appl. Sci. 11, 10003. doi:10.3390/app112110003

CrossRef Full Text | Google Scholar

Fan, M., He, G., and Zhou, M. (2020). The winter choke: coal-fired heating, air pollution, and mortality in China. J. Health Econ. 71, 102316. doi:10.1016/j.jhealeco.2020.102316

PubMed Abstract | CrossRef Full Text | Google Scholar

Frankowski, J. (2020). Attention: smog alert! Citizen engagement for clean air and its consequences for fuel poverty in Poland. Energy Build. 207, 109525. doi:10.1016/j.enbuild.2019.109525

CrossRef Full Text | Google Scholar

Furon, A. C., Wagner-Riddle, C., Smith, C. R., and Warland, J. S. (2008). Wavelet analysis of wintertime and spring thaw CO2 and N2O fluxes from agricultural fields. Agric. For. Meteorol. 148, 1305–1317. doi:10.1016/j.agrformet.2008.03.006

CrossRef Full Text | Google Scholar

Greenstone, M., He, G., Li, S., and Zou, E. Y. (2021). China’s war on pollution: evidence from the first 5 years. Rev. Environ. Econ. Policy 15, 281–299. doi:10.1086/715550

CrossRef Full Text | Google Scholar

Grinsted, A., Moore, J. C., and Jevrejeva, S. (2004). Application of the cross wavelet transform and wavelet coherence to geophysical time series. Nonlinear process. geophys. 11, 561–566. doi:10.5194/npg-11-561-2004

CrossRef Full Text | Google Scholar

He, G., Liu, T., and Zhou, M. (2020). Straw burning, PM2.5, and death: evidence from China. J. Dev. Econ. 145, 102468. doi:10.1016/j.jdeveco.2020.102468

CrossRef Full Text | Google Scholar

Ito, K., and Zhang, S. (2020). Willingness to pay for clean air: evidence from air purifier markets in China. J. Polit. Econ. 128, 1627–1672. doi:10.1086/705554

CrossRef Full Text | Google Scholar

Jia, C., Wang, Y., and Yang, X. (2020). Analysis on the periodic variation of PM2.5 concentration and its causes in Guanzhong Plain. Arid. Zo. Res. 37. doi:10.13866/j.azr.2020.06.18

CrossRef Full Text | Google Scholar

Kapwata, T., Wright, C. Y., du Preez, D. J., Kunene, Z., Mathee, A., Ikeda, T., et al. (2021). Exploring rural hospital admissions for diarrhoeal disease, malaria, pneumonia, and asthma in relation to temperature, rainfall and air pollution using wavelet transform analysis. Sci. Total Environ. 791, 148307. doi:10.1016/j.scitotenv.2021.148307

PubMed Abstract | CrossRef Full Text | Google Scholar

Li, G., Deng, S., Li, J., Gao, J., Lu, Z., Yi, X., et al. (2022). Levels, ozone and secondary organic aerosol formation potentials of BTEX and their health risks during autumn and winter in the Guanzhong Plain, China. Air Qual. Atmos. heal. 15, 1941–1952. doi:10.1007/s11869-022-01228-6

CrossRef Full Text | Google Scholar

Li, M., Ren, X., Zhou, L., and Zhang, F. (2016). Spatial mismatch between pollutant emission and environmental quality in China — a case study of NOx. Atmos. Pollut. Res. 7, 294–302. doi:10.1016/j.apr.2015.10.005

CrossRef Full Text | Google Scholar

Li, Z., Sun, Z., Shao, X., Liao, X., Zhang, X., Xiong, Y., et al. (2017). Using Morlet wavelet analysis to analyze multiple time scale periodically in PM2.5 in Beijing. China Environ. Sci. 37, 407–415. doi:10.3969/j.issn.1000-6923.2017.02.002

CrossRef Full Text | Google Scholar

Liu, G., Yang, J., Hao, Y., and Zhang, Y. (2018a). Big data-informed energy efficiency assessment of China industry sectors based on K-means clustering. J. Clean. Prod. 183, 304–314. doi:10.1016/j.jclepro.2018.02.129

CrossRef Full Text | Google Scholar

Liu, H., and Salvo, A. (2018). Severe air pollution and child absences when schools and parents respond. J. Environ. Econ. Manage. 92, 300–330. doi:10.1016/j.jeem.2018.10.003

CrossRef Full Text | Google Scholar

Liu, Y., Wu, J., Yu, D., and Hao, R. (2018b). Understanding the patterns and drivers of air pollution on multiple time scales: the case of northern China. Environ. Manage. 61, 1048–1061. doi:10.1007/s00267-018-1026-5

PubMed Abstract | CrossRef Full Text | Google Scholar

Lu, W., Tam, V. W., Du, L., and Chen, H. (2021). Impact of industrial agglomeration on haze pollution: new evidence from bohai sea economic region in China. J. Clean. Prod. 280, 124414. doi:10.1016/j.jclepro.2020.124414

CrossRef Full Text | Google Scholar

Ma, S., Xiao, Z., Zhang, Y., Wang, L., and Shao, M. (2020). Assessment of meteorological impact and emergency plan for a heavy haze pollution episode in a core city of the North China plain. Aerosol Air Qual. Res. 20, 26–42. doi:10.4209/aaqr.2019.08.0392

CrossRef Full Text | Google Scholar

Mee, (2022). 2021 communiqué on the status of China’s ecology and environment. Avaialble at: http://www.gov.cn/xinwen/2022-05/28/content_5692799.htm (accessed May 28, 2022).

Google Scholar

Miao, Z., Baležentis, T., Shao, S., and Chang, D. (2019). Energy use, industrial soot and vehicle exhaust pollution—China’s regional air pollution recognition, performance decomposition and governance. Energy Econ. 83, 501–514. doi:10.1016/j.eneco.2019.07.002

CrossRef Full Text | Google Scholar

Ministry of Ecology and Environment, (2021). China ecological environment Statistics annual report 2021. Avaialble at: https://www.mee.gov.cn/hjzl/sthjzk/sthjtjnb/202301/W020230118392178258531.pdf.

Google Scholar

Morlet, J., Arens, G., Fourgeau, E., and Giard, D. (1982). Wave propagation and sampling theory - Part II. Sampling theory and complex waves. Geophysics 47, 222–236. doi:10.1190/1.1441329

CrossRef Full Text | Google Scholar

National Bureau of Statistics, (2023). China statistical yearbook 2022. Avaialble at: http://www.stats.gov.cn/sj/ndsj/2022/indexch.htm.

Google Scholar

Rivera, N. M. (2021). Air quality warnings and temporary driving bans: evidence from air pollution, car trips, and mass-transit ridership in Santiago. J. Environ. Econ. Manage. 108, 102454. doi:10.1016/j.jeem.2021.102454

CrossRef Full Text | Google Scholar

Shi, M., Wu, H., Zhang, S., Li, H., Yang, T., Liu, W., et al. (2014). Weekly cycle of magnetic characteristics of the daily PM2.5 and PM2.5-10 in Beijing, China. Atmos. Environ. 98, 357–367. doi:10.1016/j.atmosenv.2014.08.079

CrossRef Full Text | Google Scholar

Spiridonov, V., Jakimovski, B., Spiridonova, I., and Pereira, G. (2019). Development of air quality forecasting system in Macedonia, based on WRF-Chem model. Air Qual. Atmos. heal. 12, 825–836. doi:10.1007/s11869-019-00698-5

CrossRef Full Text | Google Scholar

Sun, Z., Lu, W., Wang, Y., Li, Y., Wei, H., Shi, J., et al. (2017). ER model based supervision system analysis using information disclosure. Clust. Comput. 20, 215–224. doi:10.1007/s10586-017-0779-6

CrossRef Full Text | Google Scholar

Sun, C., Zheng, S., Wang, J., and Kahn, M. E. (2019). Does clean air increase the demand for the consumer city? Evidence from beijing. J. Reg. Sci. 59, 409–434. doi:10.1111/jors.12443

CrossRef Full Text | Google Scholar

The State Council, (2007). Environmental Protection Bureau releases list of national key monitoring enterprises. Avaialble at: https://www.gov.cn/zfjg/content_566589.htm.

Google Scholar

Tie, X., and Cao, J. (2009). Aerosol pollution in China: present and future impact on environment. Particuology 7, 426–431. doi:10.1016/j.partic.2009.09.003

CrossRef Full Text | Google Scholar

Tie, X., Madronich, S., Walters, S., Edwards, D. P., Ginoux, P., Mahowald, N., et al. (2005). Assessment of the global impact of aerosols on tropospheric oxidants. J. Geophys. Res. D. Atmos. 110, D03204. doi:10.1029/2004JD005359

CrossRef Full Text | Google Scholar

Torrence, C., and Compo, G. P. (1998). A practical guide to wavelet analysis. Bull. Am. Meteorol. Soc. 79, 61–78. doi:10.1175/1520-0477(1998)079<0061:APGTWA>2.0.CO;2

CrossRef Full Text | Google Scholar

Torrence, C., and Webster, P. J. (1999). Interdecadal changes in the ENSO-monsoon system. J. Clim. 12, 2679–2690. doi:10.1175/1520-0442(1999)012<2679:icitem>2.0.co;2

CrossRef Full Text | Google Scholar

Tu, M., Liu, Z., He, C., Fang, Z., and Lu, W. (2019). The relationships between urban landscape patterns and fine particulate pollution in China: A multiscale investigation using a geographically weighted regression model. J. Clean. Prod. 237, 117744. doi:10.1016/j.jclepro.2019.117744

CrossRef Full Text | Google Scholar

Veleda, D., Montagne, R., and Araujo, M. (2012). Cross-wavelet bias corrected by normalizing scales. J. Atmos. Ocean. Technol. 29, 1401–1408. doi:10.1175/JTECH-D-11-00140.1

CrossRef Full Text | Google Scholar

Wang, F., Zhao, H., Zhang, X., Niu, C., and Ma, J. (2019a). Understanding individual-level protective responses to air pollution warning: A case study of Beijing, China. Hum. Ecol. Risk Assess. Int. J. 25, 1473–1487. doi:10.1080/10807039.2018.1468995

CrossRef Full Text | Google Scholar

Wang, K., Gao, J., Liu, K., Tong, Y., Dan, M., Zhang, X., et al. (2022a). Unit-based emissions and environmental impacts of industrial condensable particulate matter in China in 2020. SSRN Electron. J. doi:10.2139/ssrn.4035367

CrossRef Full Text | Google Scholar

Wang, K., Gao, J., Liu, K., Tong, Y., Dan, M., Zhang, X., et al. (2022b). Unit-based emissions and environmental impacts of industrial condensable particulate matter in China in 2020. Chemosphere 303, 134759. doi:10.1016/j.chemosphere.2022.134759

PubMed Abstract | CrossRef Full Text | Google Scholar

Wang, M., Wang, W., and Wu, L. (2022c). Application of a new grey multivariate forecasting model in the forecasting of energy consumption in 7 regions of China. Energy 243, 123024. doi:10.1016/j.energy.2021.123024

CrossRef Full Text | Google Scholar

Wang, P., Cao, J., Tie, X., Wang, G., Li, G., Hu, T., et al. (2015). Impact of meteorological parameters and gaseous pollutants on PM2.5 and PM10 mass concentrations during 2010 in Xi’an, China. Aerosol Air Qual. Res. 15, 1844–1854. doi:10.4209/aaqr.2015.05.0380

CrossRef Full Text | Google Scholar

Wang, P., Guo, H., Hu, J., Kota, S. H., Ying, Q., and Zhang, H. (2019b). Responses of PM2.5 and O3 concentrations to changes of meteorology and emissions in China. Sci. Total Environ. 662, 297–306. doi:10.1016/j.scitotenv.2019.01.227

PubMed Abstract | CrossRef Full Text | Google Scholar

Wang, S., Liu, X., Yang, X., Zou, B., and Wang, J. (2018). Spatial variations of PM2.5 in Chinese cities for the joint impacts of human activities and natural conditions: A global and local regression perspective. J. Clean. Prod. 203, 143–152. doi:10.1016/j.jclepro.2018.08.249

CrossRef Full Text | Google Scholar

Wang, Y., Wan, Q., Meng, W., Liao, F., Tan, H., and Zhang, R. (2011). Long-term impacts of aerosols on precipitation and lightning over the Pearl River Delta megacity area in China. Atmos. Chem. Phys. 11, 12421–12436. doi:10.5194/acp-11-12421-2011

CrossRef Full Text | Google Scholar

Wegner, T., Hussein, T., Hämeri, K., Vesala, T., Kulmala, M., and Weber, S. (2012). Properties of aerosol signature size distributions in the urban environment as derived by cluster analysis. Atmos. Environ. 61, 350–360. doi:10.1016/j.atmosenv.2012.07.048

CrossRef Full Text | Google Scholar

WHO (2021). Global air quality guidelines. Avaialble at: https://www.who.int/news-room/questions-and-answers/item/who-global-air-quality-guidelines.

Google Scholar

Wu, Q., Hong, S., Yang, L., Mu, H., Huang, C., Niu, X., et al. (2023). Coupling coordination relationships between air pollutant concentrations and emissions in China. Atmos. Pollut. Res. 14, 101678. doi:10.1016/j.apr.2023.101678

CrossRef Full Text | Google Scholar

Xu, Y., Ying, Q., Hu, J., Gao, Y., Yang, Y., Wang, D., et al. (2018). Spatial and temporal variations in criteria air pollutants in three typical terrain regions in Shaanxi, China, during 2015. Air Qual. Atmos. heal. 11, 95–109. doi:10.1007/s11869-017-0523-7

CrossRef Full Text | Google Scholar

Zhang, R., Jing, J., Tao, J., Hsu, S.-C., Wang, G., Cao, J., et al. (2014). Corrigendum to “Chemical characterization and source apportionment of PM_2.5 in Beijing: seasonal perspective” published in Atmos. Chem. Phys., 13, 7053–7074, 2013. Atmos. Chem. Phys. 14, 175. doi:10.5194/acp-14-175-2014

CrossRef Full Text | Google Scholar

Zhang, H., Song, H., Wang, X., Wang, Y., Min, R., Qi, M., et al. (2023). Effect of agricultural soil wind erosion on urban PM2.5 concentrations simulated by WRF-chem and weps: A case study in kaifeng, China. Chemosphere 323, 138250. doi:10.1016/j.chemosphere.2023.138250

PubMed Abstract | CrossRef Full Text | Google Scholar

Zhang, J., and Mu, Q. (2018). Air pollution and defensive expenditures: evidence from particulate-filtering facemasks. J. Environ. Econ. Manage. 92, 517–536. doi:10.1016/j.jeem.2017.07.006

CrossRef Full Text | Google Scholar

Zhang, L., An, J., Liu, M., Li, Z., Liu, Y., Tao, L., et al. (2020a). Spatiotemporal variations and influencing factors of PM2.5 concentrations in Beijing, China. Environ. Pollut. 262, 114276. doi:10.1016/j.envpol.2020.114276

PubMed Abstract | CrossRef Full Text | Google Scholar

Zhang, W., Wang, H., Zhang, X., Peng, Y., Zhong, J., Wang, Y., et al. (2020b). Evaluating the contributions of changed meteorological conditions and emission to substantial reductions of PM2.5 concentration from winter 2016 to 2017 in Central and Eastern China. Sci. Total Environ. 716, 136892. doi:10.1016/j.scitotenv.2020.136892

PubMed Abstract | CrossRef Full Text | Google Scholar

Zhang, X., Shi, M., Li, Y., Pang, R., and Xiang, N. (2018). Correlating PM2.5 concentrations with air pollutant emissions: A longitudinal study of the beijing-tianjin-hebei region. J. Clean. Prod. 179, 103–113. doi:10.1016/j.jclepro.2018.01.072

CrossRef Full Text | Google Scholar

Zhao, H., Wang, F., Niu, C., Wang, H., and Zhang, X. (2018). Red warning for air pollution in China: exploring residents’ perceptions of the first two red warnings in beijing. Environ. Res. 161, 540–545. doi:10.1016/j.envres.2017.11.058

PubMed Abstract | CrossRef Full Text | Google Scholar

Zhao, S., Yu, Y., Yin, D., He, J., Liu, N., Qu, J., et al. (2016). Annual and diurnal variations of gaseous and particulate pollutants in 31 provincial capital cities based on in situ air quality monitoring data from China National Environmental Monitoring Center. Environ. Int. 86, 92–106. doi:10.1016/j.envint.2015.11.003

PubMed Abstract | CrossRef Full Text | Google Scholar

Zhao, X., Thomas, C. W., and Cai, T. (2020). The evolution of policy instruments for air pollution control in China: A content analysis of policy documents from 1973 to 2016. Environ. Manage. 66, 953–965. doi:10.1007/s00267-020-01353-x

PubMed Abstract | CrossRef Full Text | Google Scholar

Zhao, X., Zhang, X., Xu, X., Xu, J., Meng, W., and Pu, W. (2009). Seasonal and diurnal variations of ambient PM2.5 concentration in urban and rural environments in Beijing. Atmos. Environ. 43, 2893–2900. doi:10.1016/j.atmosenv.2009.03.009

CrossRef Full Text | Google Scholar

Zhong, N., Cao, J., and Wang, Y. (2017). Traffic congestion, ambient air pollution, and health: evidence from driving restrictions in beijing. J. Assoc. Environ. Resour. Econ. 4, 821–856. doi:10.1086/692115

CrossRef Full Text | Google Scholar