Heavy haze pollution during the COVID-19 lockdown in the Beijing-Tianjin-Hebei region, China

Abstract

To investigate the characteristics of particulate matter with an aerodynamic diameter less than 2.5 μm (PM_2.5) and its chemical compositions in the Beijing-Tianjin-Hebei (BTH) region of China during the novel coronavirus disease (COVID-19) lockdown, the ground-based data of PM_2.5, trace gases, water-soluble inorganic ions, and organic and elemental carbon were analyzed in three typical cities (Beijing, Tianjin, and Baoding) in the BTH region of China from 5-15 February 2020. The PM_2.5 source apportionment was established by combining the weather research and forecasting model and comprehensive air quality model with extensions (WRF-CAMx). The results showed that the maximum daily PM_2.5 concentration reached the heavy pollution level (>150 μg/m³) in the above three cities. The sum concentration of SO₄²⁻, NO₃⁻ and NH₄⁺ played a dominant position in PM_2.5 chemical compositions of Beijing, Tianjin, and Baoding; secondary transformation of gaseous pollutants contributed significantly to PM_2.5 generation, and the secondary transformation was enhanced as the increased PM_2.5 concentrations. The results of WRF-CAMx showed obviously inter-transport of PM_2.5 in the BTH region; the contribution of transportation source decreased significantly than previous reports in Beijing, Tianjin, and Baoding during the COVID-19 lockdown; but the contribution of industrial and residential emission sources increased significantly with the increase of PM_2.5 concentration, and industry emission sources contributed the most to PM_2.5 concentrations. Therefore, control policies should be devoted to reducing industrial emissions and regional joint control strategies to mitigate haze pollution.

Graphical abstract

Introduction

Fine particulate matter (particulate matter with an aerodynamic diameter less than 2.5 μm, PM_2.5) is an important air pollutant in human health, ecological effect, and climate change (Ding et al., 2016; Gao et al., 2017; Zhang et al., 2015). Since 2013, with the implementation of the toughest-ever clean air policy in China, significant declines in PM_2.5 concentrations nationwide, the air quality has improved significantly. From 2013 to 2017, the estimated national population-weighted annual mean PM_2.5 concentrations decreased from 61.8 to 42.0 μg/m³; the highest decrease (38%) in PM_2.5 was observed in the Beijing-Tianjin-Hebei (BTH) region, and the dominant contributions from anthropogenic emission abatements (Chen et al., 2019; Wang et al., 2019a; Zhang et al., 2019b). The annual mean PM_2.5 concentration of the BTH region and its surrounding areas (2+26 cities) was 57 μg/m³ in 2019 (http://www.mee.gov.cn/hjzl/sthjzk/zghjzkgb/), but this value still exceeds the World Health Organization (WHO) standard level for the good health of 25 μg/m³ (WHO, 2006). Moreover, heavy haze pollution events still occur frequently under adverse meteorological conditions.

Since December 2019, a novel coronavirus disease (COVID-19) outbreak was first identified in Wuhan and quickly expanded across China. The Chinese government has gradually taken a series of control measures to contain the epidemic, such as closed communities and restrict transportation. The control measures have significantly reduced the activity level of residents; coal, oil refining, steel output far below the same period levels in the past few years, and over 70% reductions in vehicle transportation and domestic flights (Chang et al., 2020; He et al., 2020). Thus, the atmospheric pollutant emissions from various sources have been reduced. Nevertheless, large-scale and heavy haze pollution with a high PM_2.5 mass concentration was observed in the BTH region after the Lantern Festival (8 February 2020). During this period, the major cities in the BTH region suffered at least one day of heavy haze pollution (PM_2.5 > 150 μg/m³, NAAQS Ⅳ for daily PM_2.5 and heavy pollution limit for haze); especially Beijing, which suffered the worst air quality day since the 2019-2020 heating season (November 2019 to March 2020), and the daily PM_2.5 concentrations reached 207 μg/m³ (http://www.cnemc.cn/).

In this study, we present results from analyses of PM_2.5 pollution characteristics in Beijing, Tianjin, and Baoding, the typical cities of the BTH region, and characteristics of chemical components of aerosol at urban sites in the above three cities. The weather research and forecasting model (version 3.5.1) and comprehensive air quality model with extensions (version 6.3) (WRF-CAMx) were used to perform numerical simulation analysis for PM_2.5 episodes; the inter-cities and regions transport to PM_2.5 concentration was calculated, and the contributions of different pollution sources to PM_2.5 were discussed. The present study aims to gain detailed clues of the haze event in the BTH region.

1. Materials and methods

1.1. Sites description and analysis datasets

The location of Beijing, Tianjin, and Baoding is an equilateral triangle, and the distance between cities is approximately 150 km (Fig. 1 ). The real-time data of PM_2.5, particulate matter with an aerodynamic diameter less than 10 μm (PM₁₀), nitrogen dioxide (NO₂), sulfur dioxide (SO₂), carbon monoxide (CO), and ozone (O₃) were released to the public by China National Environmental Monitoring Centre (CNEMC) and obtained from the Air Quality Monitoring and Analysis Platform (https://quotsoft.net/air/). Mass concentrations of chemical compositions in PM_2.5 were obtained from BTH regional particulate matter chemical compositions monitoring network, including water-soluble inorganic ions (SO₄ ²⁻, NO₃ ⁻, NH₄ ⁺, Cl⁻, Ca²⁺, K⁺, Na⁺, Mg²⁺, F⁻), organic carbon (OC) and elemental carbon (EC). The chemical compositions in PM_2.5 were observed at urban sites: CNEMC in Beijing (116.42°E, 40.04°N), Zhongshan north road in Tianjin (117.21°E, 39.17°N) and Yangguang north street in Baoding (115.48°E, 38.93°N). Meteorological data were obtained from the China Meteorological Data Service Centre (http://data.cma.cn/), including temperature, relative humidity, wind, and pressure. The height of planetary boundary layer (PBL) was simulated by WRF-CAMx. The online ambient observation was conducted from 5-15 February 2020, and the time resolutions of all the above data were 1 hour.

Fig. 1.

Study area and observation sites.

1.2. Model configuration

The WRF-CAMx model has been widely used for pollution simulation research, especially with Particulate Source Apportionment Technology (PSAT) (Li et al., 2013; Sun et al., 2018; Wagstrom et al., 2008). This study employed the WRF-CAMx model to simulate meteorological conditions, PM_2.5 variations, and source apportionment. The CAMx model was applied and configured using two-level nested modeling domains. A coarse-grid domain with a resolution of 36 km × 36 km was employed to cover the whole East Asia region. At a more local level, a fine-grid domain with a resolution of 12 km × 12 km was used, which covered most parts of northeastern China, including the BTH region and some surrounding provinces. The coarse-grid simulations were used to provide boundary conditions for the fine-grid simulations. All of the following analyses are based on the results from the fine-grid simulations (Zhang et al., 2017).

In order to distinguish the sources of pollutant transportation in the target cities, we reclassified the fine-grid domain into ten areas, including three target cities (1. Beijing; 2. Tianjin; 3. Baoding), four areas in other cities of Hebei Province (4. Zhangjiakou, Chengde, Qinhuangdao (ZCQ); 5. Tangshan (TS); 6. Langfang, Cangzhou (LC); 7. Shijiazhuang, Hengshui, Xingtai, Handan (SHXH)), three areas in other provinces (8. Shandong Province (SD); 9. Shanxi Province (SX); 10. Other regions). For clarity, we displayed a diagram of regional redistricting in Appendix A Fig. S1.

Tropospheric Analysis datasets were provided by the Final Operational Global Analysis of the USA National Center for Environmental Prediction (NCEP FNL data). Emissions data of nature were calculated by the Model of Emissions of Gases and Aerosols from Nature Version 3 (MEGAN3) for estimating biogenic emissions (Guenther et al., 2020). Emissions data of anthropogenic activity were based on the Multi-resolution Emission Inventory for China (MEIC emission inventory Version 1.3 (http://www.meicmodel.org/)), which include PM_2.5, PM₁₀, CO, SO₂, nitrogen oxides (NO_x), black carbon (BC), OC, total suspended particulate (TSP), ammonia (NH₃), volatile organic compounds (VOCs), and carbon dioxide (CO₂). Furthermore, these pollutants were divided into six sources: industry, residential, transportation, power, agriculture, and other sources. The emissions in the 2+26 cities were updated to include the latest survey results. In particular, we have revised the emission data again during the COVID-19 lockdown to ensure the agreement between the simulated and observed values of pollutants (Appendix A Fig. S2) (Huang et al., 2020; Le et al., 2020; Lv et al., 2020; Meng et al., 2021).

2. Results and discussion

2.1. Overview of PM_2.5 concentrations

The temporal and spatial distributions of daily PM_2.5 concentrations at various sites in the BTH region during haze episodes are shown in Fig. 2 . The daily PM_2.5 concentrations at all sites in the BTH region reached Class Ⅱ of the National Ambient Air Quality Standard (NAAQS Ⅱ, 75 μg/m³) on 5-6 February 2020; air quality levels were good or moderate at different sites (MEE, 2012). After that, slight PM_2.5 pollution began to appear in the southeastern part of the BTH region on 7 February, and large-scaled heavy PM_2.5 pollution gradually appeared in the following days. From 9 to 11 February, the maximum daily PM_2.5 concentration exceeded 150 μg/m³ at all sites of Beijing, Tianjin, and Baoding. Until 14 February, the daily PM_2.5 concentration in the BTH region dropped rapidly; then, the regional daily PM_2.5 concentrations were less than 35 μg/m³ (NAAQS Ⅰ) on 15 February, and the haze episode ended.

Fig. 2.

Characteristics of PM_2.5 at various sites in the Beijing-Tianjin-Hebei region.

During heavy haze pollution episodes, the mean daily PM_2.5 concentrations of Baoding exceeded 150 μg/m³ for five consecutive days and three, two days of Beijing and Tianjin (Table S1). The maximum daily PM_2.5 concentration of Beijing reached 207 μg/m³ (12 February), and this was the most polluted day during the 2019-2020 heating season. The highest hourly PM_2.5 level at all sites of three cities was 330 μg/m³, which appeared in the center of the cities’ triangle (Yufa site, Fig. 2), and this value was four times higher than the NAAQS Ⅱ.

The ratios of PM_2.5/PM₁₀ were 1.20±0.36, 1.12±0.37, 0.81±0.11, and the correlation coefficients (R ²) were 0.96, 0.95, 0.99 in Beijing, Tianjin, and Baoding during the observed period, respectively. There was a reversal phenomenon between PM_2.5 and PM₁₀ under high humidity conditions in Beijing and Tianjin. This could be caused by the different monitoring methods of PM_2.5 and PM₁₀, and another critical factor is that the influence of water was not absolutely eliminated under high humidity (Jiang et al., 2016; Pan et al., 2014).

2.2. Characteristics of PM_2.5 chemical compositions

Fig. 3 shows the time series of meteorological variables, pollutants concentrations, and PM_2.5 chemical compositions in Beijing, Tianjin, and Baoding from 5 to 15 February 2020. For PM_2.5 concentration, the site data closest to the chemical composition monitoring was used: Aoti site in Beijing, Huaihedao site in Tianjin, and Jiedaizhongxin site in Baoding. The converted factor of OC to organic matter (OM) has been considered from 1.4 to 2.2, and it is related to atmospheric oxidation capacity and meteorology (Huang et al., 2017, 2020; Pang et al., 2020; Xu et al., 2019); this study adopted a factor of 1.6 because of high O₃ concentration and low temperature during the observed period.

Fig. 3.

Time series of meteorological variables, pollutants concentrations, and PM_2.5 chemical compositions in (a) Beijing, (b) Tianjin, and (c) Baoding. TEMP: temperature; PBLH: the height of planetary boundary layer; RH: relative humidity; OM: organic matter; OC: organic carbon; EC: elemental carbon.

Beijing, Tianjin, and Baoding have similar characteristics of PM_2.5 concentration and chemical composition, and the mass concentration ratios of chemical compositions to PM_2.5 were 0.80, 0.75, and 1.03 in the above three cities, respectively (Table 1 ). The concentrations and chemical compositions of PM_2.5 in Baoding were consistent; however, Beijing and Tianjin have a large missing part. Trace elements in PM_2.5 were not detected, which can partly explain this phenomenon. Nevertheless, as shown in Section 2.1, there is a reversal between PM_2.5 and PM₁₀, and the uncertainty of PM_2.5 measurement could be another important reason.

Table 1.

Mass concentrations of PM_2.5 chemical compositions (μg/m³) in Beijing, Tianjin, and Baoding.

	Beijing	Tianjin	Baoding
OC	6.7±3.5	14.0±9.7	17.1±12.1
EC	2.3±1.8	3.1±2.5	4.3±2.9
SO42−	16.7±9.6	11.9±8.6	33.6±24.9
NO3−	33.4±24.0	14.5±12.1	31.5±26.9
NH4+	15.9±8.8	14.7±9.0	22.3±16.0
Cl−	2.4±1.6	2.3±2.0	3.7±2.7
Ca2+	0.0±0.0	0.1±0.1	0.3±0.4
K+	1.4±1.5	0.5±0.6	1.2±0.9
Na+	0.5±0.1	0.8±0.6	0.6±0.3
Mg2+	0.1±0.1	0.1±0.1	0.1±0.2
F−	0.1±0.1	0.1±0.1	/
SNA/PM2.5	63%±16%	41%±28%	73%±21%
CC/PM2.5	80%±17%	75%±69%	103%±23%

SNA: SO₄²⁻, NO₃⁻ and NH₄⁺; CC: chemical compositions.

2.2.1. OC and EC

The correlation coefficients (R ²) of OC and EC concentration are all greater than 0.87 among the three cities, which means that OC and EC had the same source. The concentration of total carbon (TC, the sum of OC and EC) was the lowest in Beijing, followed by Tianjin and the highest in Baoding. The TC of Beijing, Tianjin, and Baoding accounted for 9%, 17%, and 18% of PM_2.5, respectively, which was lower than that of similar cities (Li et al., 2019; Pang et al., 2020). This is closely related to the industrial layout in the BTH region (Xu et al., 2019). A lower proportion means a reduction in primary emissions intensity, and this phenomenon is closely related to the decrease of human activity level during the COVID-19 lockdown (Geng et al., 2019; Lv et al., 2020; Wang et al., 2019b).

The EC tracer method was widely used to estimate the photochemical contribution of secondary organic carbon (SOC) (Lim and Turpin, 2002; Zhang et al., 2021; Zheng et al., 2015). We estimated SOC by using the EC-tracer method, formulae as follows:

$$\text{POC}={\left(\text{OC}/\text{EC}\right)}_{\text{primary}}\times \text{EC}$$ (1)

$$\text{SOC}=\text{TOC}-\text{POC}$$ (2)

where POC is primary organic carbon, and TOC is the estimated total organic carbon. Data in the lowest 10% by OC/EC ratios were used to estimate the values of ${(\text{OC}/\text{EC})}_{\text{primary}}$ (Appendix A Fig. S3). The analyzes show that SOC constituted 29%, 40%, 36% of TOC in Beijing, Tianjin, and Baoding. These values were slightly higher than the winter average, which is consistent with the haze episodes in the BTH region (Pang et al., 2020; Shao et al., 2018), indicating that POC was the main contributor to OC in all three cities, but the accumulation of SOC to OC increased during the heavy PM_2.5 pollution period.

2.2.2. Water-soluble ions

The average fractions of SNA (SO₄ ²⁻, NO₃ ⁻, NH₄ ⁺) in PM_2.5 were 41%-73% in these three cities (Table 1). These values were higher than the winter average. Pang et al. (2020) has reported approximately 30%-40% in Beijing, Tianjin, and another city (Langfang) in the BTH region during the heating season. Previous studies have shown that the mass ratio of SNA to PM_2.5 was highest in the summer, about 45%–66%, due to strong photochemical reactions (Huang et al., 2017; Xu et al., 2019). In this study, the values of Beijing and Baoding were even higher than the seasonal maximum, indicating that efficient secondary conversion ability during the COVID-19 lockdown.

The ratio of NO₃ ⁻/PM_2.5 was the highest (32%) among the water-soluble ions in Beijing, followed by SO₄ ²⁻ (16%) and NH₄ ⁺ (15%). In Tianjin, the ratios of NO₃ ⁻, SO₄ ²⁻, NH₄ ⁺ to PM_2.5 were similar, 15%, 15%, and 12%, respectively. In Baoding, the above three ratios were 28%, 26%, and 19%. Generally, the primary sources of NO₃ ⁻, SO₄ ²⁻, NH₄ ⁺ are NO_x, SO₂, and NH₃, respectively (Seinfeld and Pandis, 2016). The BTH region is densely populated with heavy industries, such as steel, cement, and other heavy industry enterprises, with huge emissions of NO_x and SO₂ (Geng et al., 2019; Zheng et al., 2018). Additionally, as a main grain-producing region in China, the southern part of the BTH region produces a large amount of NH₃ emission (Kang et al., 2016; Liu et al., 2019).

The ratio of NO₃ ⁻/SO₄ ²⁻ was generally used to judge the relative contributions of transportation and stationary sources (Arimoto et al., 1996; Pang et al., 2020). Beijing has the highest ratio of 2, and Baoding has the lowest value of 0.94. Comparatively, transportation sources in Beijing and stationary sources in Baoding contributed to PM_2.5 were more significant (Wang et al., 2014). The concentrations of SO₄ ²⁻ and TC in Baoding were more than twice that of Beijing, but NO₃ ⁻ concentration was similar. These phenomena reflected the transportation emissions intensity of the above three cities is similar, and the stationary sources emission of Baoding is more significant than Beijing and Tianjin.

2.2.3. Variations in different haze polluted conditions

We reclassified PM_2.5 polluted conditions into eight categories and analyzed the change characteristics of chemical compositions (Fig. 4 a). There are various characteristics in different cities. For Beijing, with the increase of PM_2.5 concentration, the proportion of NO₃ ⁻ raised rapidly. When the concentration of PM_2.5 > 200 μg/m³, the proportion of NO₃ ⁻ could reach 46%. Pang et al. (2020) and Shao et al. (2018) reported that the proportion of NO₃ ⁻ in PM_2.5 accounted for 20% and 16% on haze episodes during the 2016-2017 heating season. The result in this study is higher than that of previous studies, and NO₃ ⁻ played a dominant position in promoting PM_2.5 concentration increase. In addition, the proportion of OM decreased with the increase of PM_2.5 concentration, and when PM_2.5 pollution was the most serious (>200 μg/m³), the proportion of organic matter was the lowest, only 11%. The results show that during the PM_2.5 episodes in the COVID-19 lockdown, the direct contribution of primary source emissions to PM_2.5 was irrelevant in Beijing. For Tianjin, OM was in a major position, about 30%-40% of all chemical compositions. Furthermore, there is no significant change in the proportion of any major chemical compositions when the PM_2.5 concentration increased. The absolute concentrations of major chemical compositions have similarly increased rates with the increase of PM_2.5 concentration. For Baoding, NO₃ ⁻ proportion raised gradually when PM_2.5 concentration increased, and the maximum ratio of NO₃ ⁻ was 29%. This phenomenon is similar to the change in Beijing but not as evident. Besides, the trend of other major chemical compositions (OM, NH₄ ⁺) was proportional to PM_2.5 concentration (similar to Tianjin).

Fig. 4.

Variations of (a) chemical compositions and (b) nitrogen oxidation ratio (NOR), sulfur oxidation ratio (SOR), RH, temperature as a function of PM_2.5 concentration in Beijing (up), Tianjin (middle), and Baoding (down).

Sulfur oxidation ratio (SOR) and nitrogen oxidation ratio (NOR) were used to evaluate the degree of conversion of SO₂ and NO₂ into secondary aerosols, calculated as a molar fraction using the following equations (Wang et al., 2005).

$$\text{SOR}=\frac{n\left[{\text{NSS}-\text{SO}}_4^{2-}\right]}{n\left[{\text{NSS}-\text{SO}}_4^{2-}\right]+n\left[\mathrm{S}{\mathrm{O}}_2\right]}$$ (3)

$$\text{NOR}=\frac{n\left[{\text{NO}}_3^{-}\right]}{n\left[{\text{NO}}_3^{-}\right]+n\left[\mathrm{N}{\mathrm{O}}_2\right]}$$ (4)

Here, n refers to the molar concentration, and NSS refers to non-sea salt. NSS-SO₄ ²⁻ was estimated with the following equation in Tianjin (Pang et al., 2020).

$${\text{NSS}-\text{SO}}_4^{2-}=\left[{\text{SO}}_4^{2-}\right]-0.2455\left[\mathrm{N}{\mathrm{a}}^{+}\right]$$ (5)

Beijing and Baoding are far away from the sea, and the influence of sea salt is not considered. We analyzed the distributions of SOR, NOR and meteorological factors (temperature, relative humidity) under various PM_2.5 polluted conditions (Fig. 4 b). Overall, the SOR and NOR gradually increased with the increase of PM_2.5 concentration. In Beijing, the lowest SOR and NOR values were 0.07 and 0.02 in the condition of PM_2.5 was 0-15 μg/m³. The highest SOR was 0.73, which was found in the PM_2.5 range was 76-100 μg/m³, and the highest NOR was 0.33 when the PM_2.5 range was >200 μg/m³; the maximum values of SOR and NOR did not occur in the same range of PM_2.5 concentrations. These phenomena show the various trends as the characteristics of chemical compositions of different PM_2.5 polluted conditions. The correlation coefficients (R) of SOR, NOR and PM_2.5 were 0.50 and 0.76. Both the trend and correlation between NOR and PM_2.5 were greater than SOR, indicating the enhanced secondary transformation of NO₂ to NO₃ ⁻ is aggravated by increased PM_2.5 concentrations. In Tianjin, the trends of SOR, NOR, and PM_2.5 were similar to that of Beijing. The highest SOR and NOR values occurred when the PM_2.5 was at the range of 51-75 and >200 μg/m³, respectively. The correlation coefficients of SOR, NOR and PM_2.5 were 0.67 and 0.68. When PM_2.5 concentration was higher than 100 μg/m³, there was little difference for the proportion of SO₄ ²⁻, but the percentage of NO₃ ⁻ shows a decreasing trend with the increase of PM_2.5 concentration (Fig. 4 a). The above phenomena indicate that the increase in PM_2.5 concentration could be affected by other factors. Further analysis results about chemical compositions of PM_2.5 in Tianjin found that OM accounts for the largest proportion in each category of PM_2.5 polluted conditions (Figs. 3 and 4 a). The SOR and NOR of Baoding were different from those of Beijing and Tianjin. The correlation coefficients of SOR, NOR and PM_2.5 were 0.82 and 0.89, which were significantly higher than the other two cities. The maximum values of SOR and NOR occurred when the most serious PM_2.5 polluted condition, with 0.70 and 0.51, respectively. Compared with Beijing and Tianjin, the NOR of Baoding shows a rapid increase trend with PM_2.5 concentration increased.

2.3. Source apportionment of PM_2.5

2.3.1. Regional contribution

We used the WRF-CAMx model to distinguish the sources of pollutant transportation in the target cities, and the results showed that the local contribution to the PM_2.5 concentrations accounted for 28%, 40%, and 49% in Beijing, Tianjin, and Baoding, respectively (Fig. 5 ). This reflects the obvious inter-transport of PM_2.5 in the BTH region. Beijing was mainly affected by its eastern region; the contribution of TS, LC, and Tianjin to Beijing ranked the top three, with 12%, 10%, and 8%, respectively. Tianjin was mainly affected by its southern region; LC and SD contributed significantly to 12% and 11%. Baoding was mainly affected by its southern and western areas; SHXH and SX substantially contributed to Baoding, with 9% and 8%, respectively. The WRF-CAMx results of regional contributions are well in agreement with the ground-based meteorological monitoring data. According to the meteorological variable information, the dominant wind directions were east, south, and south in Beijing, Tianjin, and Baoding (Fig. 3). Additionally, the other regions also significantly impact the target cities, which means an increase in regional background PM_2.5 concentrations during this analysis period.

Fig. 5.

Contributions of different cities and regions to the PM_2.5 concentrations in Beijing, Tianjin, and Baoding. The numbers 1, 2, 3, 4, 5, 6, 7, 8, 9 and 10 are Beijing, Tianjin, Baoding, Zhangjiakou Chengde Qinhuangdao (ZCQ), Tangshan (TS), Langfang Cangzhou (LC), Shijiazhuang Hengshui Xingtai Handan (SHXH), Shandong province (SD), Shanxi province (SX) and Other regions, respecitively.

2.3.2. Contribution of pollution sources

The source apportionments of PM_2.5 in Beijing, Tianjin, and Baoding during this analysis period are shown in Fig. 6 . From the entire region, industrial sources have a dominant contribution to the PM_2.5 concentration in Beijing, Tianjin, and Baoding with 39%, 37%, 40%, respectively, indicated that the current situation of high industrial emission intensity has not changed during the COVID-19 lockdown, and haze pollution is likely to occur in the event of unfavorable meteorological conditions (Cai et al., 2020).

Fig. 6.

Source apportionments of PM_2.5 in Beijing, Tianjin, and Baoding.

Transportation is a major source of urban PM_2.5 concentration (http://sthjj.beijing.gov.cn/). Previous studies have shown that the contribution of transportation sources to PM_2.5 in Beijing was approximately 21%-43%, whereas Tianjin and Baoding were 17%-24% and 10%-14% (Gao et al., 2018; Xu et al., 2019; Zhang et al., 2019a). In this study, the contributions of transportation sources to PM_2.5 concentrations were significantly lower than the previously reported results, with 15%, 14%, 10% in Beijing, Tianjin, and Baoding cities, respectively; this is related to the regional control measures during the COVID-19 lockdown (Lv et al., 2020). The result is directly evidenced by road traffic flow data, which showed a 69% decrease in Hebei Province in February 2020 compared to February 2019 (Meng et al., 2021).

To better judge the contribution of different sources to PM_2.5 with varying levels of pollution, the results of daily source apportionments are shown in Appendix A Fig. S4. It was observed that the contribution of industrial and residential sources increased significantly with the increase of PM_2.5 concentration. On the heavy PM_2.5 pollution days of Beijing, Tianjin, and Baoding, the maximum daily contributions from industrial sources were 45%, 46%, 45%, respectively, and 22%, 27%, 34% from residential sources. In contrast, the contribution of mobile sources increased slightly, contributing up to 17% in the above three cities.

3. Conclusions

This study investigated the polluted characterization and evolution of PM_2.5 and its chemical compositions in Beijing, Tianjin, and Baoding cities of the BTH region during the COVID-19 lockdown (5-15 February 2020) and discussed the source apportionment of PM_2.5. The results showed that the maximum daily average PM_2.5 concentration reached the heavy pollution level (>150 μg/m³) in the above three cities; for Beijing, the concentration was the highest value (207 μg/m³) during the 2019-2020 heating season. The peak hourly PM_2.5 concentration at a site of the BTH region reached 330 μg/m³, a value that was four times than the NAAQS Ⅱ. The sum concentration of SO₄ ²⁻, NO₃ ⁻ and NH₄ ⁺ played a dominant position in PM_2.5 chemical compositions of Beijing, Tianjin, and Baoding, which means that secondary transformation of gaseous pollutants contributed significantly to the generation of PM_2.5; the SOR and NOR increased with the increase of PM_2.5 concentration, indicated secondary transformation was enhanced as the increased PM_2.5 concentration. WRF-CAMx simulations showed that the local contribution to the PM_2.5 concentrations did not exceed 50% in the above three cities, and obviously inter-transport of PM_2.5 in the BTH region was found. Beijing, Tianjin, and Baoding were mainly affected by the eastern, southern, and southwest regions. The transportation source emissions were significantly reduced and lower than previous reports because of the regional control measures in the BTH region during the COVID-19 lockdown. However, the contribution of industrial and residential emission sources increased significantly with the increase of PM_2.5 concentration. The contribution of industrial emission sources was the largest, indicating that the status quo of high industrial emission intensity has not changed. The results highlight the importance of regional joint control strategies and industry emission reduction for mitigating haze pollution.