Impacts of Sandstorms on Chemistries of Ambient PAHs in a Small City in North China

Received: January 3, 2023
Revised: March 10, 2023
Accepted: March 14, 2023

 Copyright The Author(s). This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are cited.

Cite this article:

Li, Z., Zhai, Z., Liu, J., Chen, L., Ren, Z., Liu, C., Zhan, Z., Yue, Z., Zhu, W., Wei, J., Gao, H., Guo, S. (2023). Impacts of Sandstorms on Chemistries of Ambient PAHs in a Small City in North China. Aerosol Air Qual. Res. 23, 220460. https://doi.org/10.4209/aaqr.220460

HIGHLIGHTS

• Evaluated the sandstorm (SS) impacts on PM_2.5-bound PAH chemistries.
• No significant change in total PAH concentrations was found in the SS period (SSP).
• Oil combustion/leakage contribution increased markedly due to the sandstorm.
• More toxic PAH congeners and health risks by PAH exposure increased in the SSP.

Keywords: PM2.5, PAHs, Sandstorm, Source apportionment, Exposure risk

1 INTRODUCTION

With the rapid urbanization and economic development, air pollution incidents occur frequently in China, which has aroused public concern (Wang et al., 2022; Li et al., 2023). PAHs are semi-volatile organic compounds derived from both the anthropogenic sources (incomplete incineration of wood, tobacco, coal tar, petroleum or chemical products, and oil leakage) and the natural emissions (volcanic eruption, forest and grassland fire, etc.) (Wang et al., 2021a; Wang et al., 2021b; Li et al., 2022a). Furthermore, PAHs are ubiquitous organic pollutants in different environmental media, which have drawn increasing public attention given their toxicity, carcinogenicity, and adverse health effects (Pheiffer et al., 2019; Vega et al., 2021; Wang et al., 2021a; Li et al., 2022b). With these physicochemical and mutagenic properties, PAHs has been categorized as one of persistent organic pollutants by the United Nations Economic Commission (UNCEC, http://www.unece.org/env/lrtap/pops_h1.htm)(Li et al., 2022a). As the largest global emitter of air pollutants, China has been confronting severe air pollution crises (Li et al., 2023). Among these pollutants, the ambient PAH levels in many Chinese regions were significantly higher than those of the other developing countries, posing great threats to human health by their exposure, especially by the inhalation (Shen et al., 2013; Xie et al., 2017; Kong et al., 2018). In previous studies that mainly focused on urban areas, in particular, the PM_2.5-bound PAHs (PB-PAHs) have been recognized as main contributors to people’s health risks posed by PM_2.5 (Xie et al., 2017). The incremental lifetime lung cancer risk (ILCR) derived from ambient PAH exposure was 3.1 × 10^–5 in 2007 (Shen et al., 2014).

Over the past decades, the rapid industrialization and urbanization in China has led to the obvious increments in PAH emissions (Li et al., 2022b, 2023). Sandstorm is a natural meteorological disaster that occurs repeatedly in many arid and semi-arid regions of the world, which is the sand dust raised by strong winds across dry deserts around the earth (Ma et al., 2022; Wang et al., 2022). Sandstorm can not only deteriorate the air quality from source regions to downstream regions, but also invade the highly urbanized regions near the desert (Ma et al., 2022). At the same time, the sandstorm originated from some industrial agglomeration regions can export local air pollutants including PM_2.5, PM₁₀, and PAHs to remote receptor areas, resulting in serious air pollution and adverse health risks (Li et al., 2022; Wei et al., 2022). In China, sandstorm mainly stems from places with low vegetation coverage and strong surface winds, such as northwest China and Inner Mongolia with a lot of heavy industry (Alghamdi et al., 2015; Chen et al., 2018; Hussein et al., 2020; Wei et al., 2022). Sandstorms are often composed of a complex mixture of inorganic elements, water-soluble anions and cations, elemental carbon (EC), organic carbon (OC), polycyclic aromatic hydrocarbons (PAHs) and other organic compounds, and pathogenic microorganisms, which can get into the human bodies and thus cause various diseases (Goudarzi et al., 2019; Karimi et al., 2020; Ma et al., 2022). Sandstorm can travel thousands of kilometers from the original place, passing towns and oceans, and deposit as dust clouds in areas far from their original source (Goudie and Middleton, 2001; Meo et al., 2021). PAHs attached to sand dust can lead to the adverse human health effects (Wang et al., 2022). Therefore, PB-PAHs derived health risks and risk sources were urgently needed to be evaluated (Ma et al., 2022). At the same time, however, from my perspective, the present studies mainly focused on its impacts on fine particles, gas pollutants, and adverse health effects by exposure to PM_2.5-bound inorganic pollutants (Meo et al., 2021; Ma et al., 2022), water conservation and vegetation restoration in sandstorm source regions of Beijing-Tianjin in China (Wang et al., 2021b), and identification of sandstorm (Ouyang et al., 2022; Wang et al., 2022). However, few studies were deployed on the impacts of sandstorm originated from industrial area on PB-PAHs of receptor area.

We carried out an intensive on-site observation during March 16 to April 9, 2021, in a small city covering two sandstorm events to examine the sandstorm impacts on PAH characteristics. Dingxing County was selected, locating within the Beijing-Tianjin-Hebei (BTH) region, which was frequently subjected to the sandstorm pollution derived from Inner Mongolia with lots of industries. This study mainly aimed to: (1) analyze the variations in PB-PAH levels and associated source contributions between the sandstorm period and non-sandstorm period; and (2) assess the potential health risks of PB-PAHs in the sandstorm period.

 
2 METHODOLOGY

 
2.1 Sampling Area Description

Fig. 1 shows the specific location of PM_2.5 sampling sites. PM_2.5 collection was completed on the roof of a main building of Dingxing County government. Dingxing County is located in northern Hebei Province and surrounded by the cities of Beijing, Tianjin, and Baoding. Dingxing County is 89 km away from downtown Beijing City in the north, 122 km away from downtown Tianjin City in the east, and 54 km away from downtown Baoding City in the west. The total area is 714 km², and the cultivated area is 49,300 hectares. It has jurisdiction over 1 provincial industrial cluster, 1 urban area, and 274 administrative villages, which is one of the 35 counties (cities) around Beijing Tianjin metropolitan area in Hebei Province. The sampling point is situated at urban area of Dingxing County, surrounding by residential buildings within a radius of 3 km, which is identified as a typical residential area.

 Fig. 1. Location of the sampling site (marks in green color).

 
2.2 PM_2.5 Sampling

The sampling processes can be found in Tao et al. (2014) and Wei et al. (2022). Two air samplers (TH-150C, Wuhan Tianhong Instrument Co., Ltd) were employed to collect PM_2.5 at the same time at 100 L min^–1. Two different types of filters were used in this study, including a quartz fiber (QF) filter (Pall, USA) and a polytetrafluoroethylene (PTFE) filter (Whatman UK), and both of them were 90 mm in diameter. Sampling began on March 16, 2021, and ended on April 9, 2021. The sample duration for each collection was 23 hours (from 8:00 AM to 7:00 AM). Totally 50 PM_2.5 samples and 5 blank samples were acquired. The QF and PTFE filters were roasted at 450°C and 60°C, respectively and then sampled prior to sampling. All the filters were put into a room with a constant temperature and humidity for 48 hours before and after sampling for weighing using a Sartorius ME-5F microbalance (Sartorius, Göttingen, Germany). Deviations among three duplicate weights were all lower than 15 µg for each filter (Yang et al., 2022). Net mass was calculated as the difference between the pre-weight and post-weight. All the filters were placed in a freezer at –20°C before analysis.

 
2.3 PAH Analysis

In the aggregate, eighteen PAH congeners were involved in this study including naphthalene (NA), acenaphthylene (ACL), acenaphthylene (AC), fluorene (Fl), phenanthrene (PHE), anthracene (AN), fluoranthene (FA), pyrene (PY), benzo(a)anthracene (BaA), chrysene (CHR), benzo(b)fluoranthene (BbF), benzo(k)fluoranthene (BkF), benzo(a)pyrene (BaP), indeno(1,2,3-cd)pyrene (IP), dibebz(a.h)anthracene (DBA), benzo(g,h,i)perylene (BgP), benzo(e)pyrene (BeP), and coronene (COR). They were categorized into 2-ring PAHs (NA), 3-ring PAHs (ACL, AC, FL, PHE and AN), 4-ring PAHs (FA, PY, BaA and CHR), 5-ring PAHs (BbF, BkF, BaP and BeP), 6-ring PAHs (IP, DBA and BgP), and 7-ring PAHs (COR) in term of their number of benzene rings. Moreover, they can be further divided into low molecular weight PAHs (LMW-PAHs; 2- and 3-ring), medium molecular weight PAHs (MMW-PAHs, 4-ring), and high molecular weight PAHs (HMW-PAHs, 5-, 6-, and 7-ring). The samples were analyzed by a GC/MS system (GC6890/MSD5973i; Hewlett-Packard, USA) (Li et al., 2021, 2022b).

The method of EPA TO-13A provided the detailed pretreatment and analysis procedures in this study. Briefly, a quarter of QF filter was ultrasonically extracted with dichloromethane, concentrated using a rotary evaporator, and then put into a silica gel for cleanup and purification. The elute was reduced to ~1 mL by a nitrogen blowing instrument, and then spiked with internal standards before analysis. The adopted chromatographic conditions were consistent with those in Li et al. (2021, 2022b) and shown as follows: held at 70°C for 2 mins, heated at 10°C min^–1 to 260°C for 8 mins, and then rose to 300°C at 5°C min^–1 and held for 5 mins. More details about quality assurance and quality control (QA/QC) were documented in Kong et al. (2018) and Li et al. (2021). The same treatment procedures in regular samples were adopted to analyze field and laboratory blank filters. PAHs were quantified by both the intention time and peak areas of the calibration standards. Internal standard solutions 14-deuterium tribiphenyl and 4-bromo-2-fluorobiphenyl were used in this study. Method detection limits (MDLs) of 18 PAHs ranged between 0.010 and 1.20 ng, with a mean value of 0.22 ± 0.18 ng. The recoveries of 18 PAHs in all the collected samples fell within the range of 80% to 120%. The average recoveries of two internal standards were 90 ± 15% and 95 ± 20%, respectively. The relative standard deviations (RSD) of the 6 repeated samples were all less than 10%. The reported PAH results were subtracted from blanks, but not corrected by recoveries.

 
2.4 Backward Trajectory Calculation

The 72 h backward trajectories of air masses arrived at Dingxing County with the starting height of 500 m were calculated by the NOAA Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model 4.8, which is a specialized model developed by NOAA's Air Resources Laboratory, ARL. It is widely used to calculate and analyze the transport and diffusion trajectories of air pollutants. Then these trajectories were clustered via the TrajStat software. Clustering procedure was presented in Hysplit User's Guide-Version 4. We got 112 trajectories and then divided them into five paths by minimizing the angle distances (Xu et al., 2019).

 
2.5 Source Apportionment of PAHs

2.5.1 Diagnostic Ratios (DRs)

In this study, diagnostic ratios were used to determine the changes in PAH sources between the sandstorm and non-sandstorm periods. The DRs including AN/(AN + PHE), FA/(FA + PY), BaA/(BaA + CHR), and IP/(IP + BgP) were commonly used as indicators to qualitative identification of PAH emission sources such as petroleum products, petroleum combustion, and coal or biomass burning (Suman et al., 2016). We examined these DRs and compared them with those of typical emission sources reported in Kong et al. (2018).

 
2.5.2 Positive Matrix Factorization (PMF) Model

PMF, based on the weighted least square fit approach, is a multivariable factor analysis tool for the quantitative identification of main sources of PAHs (Li et al., 2023). Its advantages include no need to measure the source fingerprint spectrums, non-negative decomposition matrix, and optimization using the data standard deviations (Li et al., 2021). PMF decomposes a matrix of specific sample data into factor contribution matrix (G) and factor profile matrix (F), which is presented as Eq. (1):

where x_ij is the concentrations of jth PAH species in the ith sample; g_ik is the contribution of the kth source to the ith sample; f_kj is the mass fraction of the jth compound from the kth source; and e_ij is the residual for each sample and species. The factor profile matrix was matched to the documented source profiles and then coupled with the local emissions, which can recognize the adequate sources (Wang et al., 2020a; Li et al., 2023). U.S. Environmental Protection Agency PMF version 5.0, is applied in source apportionment of PAHs in this study (Taghvaee et al., 2018; Wang et al., 2020a). Model inputs included the measured volume concentrations and uncertainties of 18 PAH congeners. In addition, Li et al. (2022b) provided the calculations of uncertainties and presented by Eqs. (2) and (3):

An extra modeling uncertainty of 10% was used in this study. PMF was run 20 times for each PAH congener.

 
2.5.3 Potential source contribution function

The potential source contribution factor analysis method is based on the conditional probability function to identify the location of possible pollution sources. PSCF calculation is used to evaluate the probability of a source located at latitude i and longitude j. PSCF values for grid cells were obtained by PSCF_ij = m_ij/n_ij. n_ij refers to the number of segments with endpoints that fell in the ij^th cell, m_ij is the number of trajectories that arrived at a receptor site with pollutant concentrations higher than the 75^th percentile concentration of a source of PAHs.

 
2.6 Health Risk Assessment

BaP equivalent concentration (BaP_eq) proposed by the U.S. Environmental Protection Agency and incremental lifetime cancer risk (ILCR) were applied to assess the health risks of PAHs by inhalation exposure (Kong et al., 2018). BaP_eq values of PAHs can be obtained as the product of mass concentration of PAHs (C_i) and corresponding toxicity equivalent factor (TEF) (Li et al., 2021), which were shown as follows:

What’s more, the sum of BaP_eq value of individual PAH congener was used to examine the total carcinogenicity (Kong et al., 2018). ILCR is an effective evaluation model based on BaP to assess human health risk. Among three exposure pathways, inhalation is a crucial pathway to PB-PAHs. In return, inhalation was selected in this study for risk evaluation (Shen et al., 2014).

Detailed description about meaning and adopted values of aforementioned parameters was listed in Table S1. According to the standards set by the U.S. EPA, the ILCR greater than 1 × 10^–6 and lower than 1 × 10^–4 is acceptable risk, and greater than 1 × 10^–4 may cause harm to people (Wang et al., 2020a).

 
3 RESULT AND DISCUSSION

 
3.1 PAH Concentrations

The statistics about volume concentrations of PAHs and PM_2.5, and mass concentrations of PAHs were shown in Fig. 2. The average PM_2.5 concentrations enhanced significantly by 23.8% in the sandstorm period (SSP) compared to the non-sandstorm period (NSSP). At the same time, the average PAH concentrations exhibited a small enhancement from 9.16 ng m^–3 to 10.3 ng m^–3, with the increase of 12.4%. This suggestion might be related to the inputs of PAHs emitted from the region of sandstorm origin. On the contrary, the mass concentration of PAHs decreased from 147 µg g^–1 in the NSSP to 137 µg g^–1 in the SSP due to high increase in particle levels in the SSP, which was consistent with the results of Mount Heng in 2015 (Yang et al., 2015).

Fig. 2. PM_2.5 concentrations, and PAH contents and concentrations for each sampling day, and the sandstorm and non-sandstorm period.

It should be noted, the average daily PM_2.5 and PAH concentrations after March 31 markedly decreased, which should partly be ascribed to the cease of heating behavior other than the sandstorm events before March 31. In 2021, the end of domestic heating in sampling site extended from March 15 to March 31 due to the COVID-19.

Table 1 showed the comparisons of PAH levels between the sampling site and other Chinese cities. The PAHs in a small city of this study were much lower than those norther Chinese cities before implementation of air pollution control policies in 2013 and 2017, such as Dalian in Jul. to Aug. 2017, Tianjin in Jul. to Aug. 2007, Zhengzhou in 2016, Beijing and Yantai in 2010–2011 and 2015–2016, Taiwan in Jul.–Sept. 2001, Xi’an in 2005–2007, indicating the effectiveness of “Clean Air Action Plan” since 2013 and “Clean Heating” policy since 2017 (Li et al., 2023). Moreover, the PAH levels in this study not only in the NSSP but in the SSP were comparable to that of Hangzhou in 2015–2016.

 
3.2 Levels of Individual PAH

Totally 17 of 18 PAHs congeners except for ACL were detected in this study and their statistic values were listed in Table S2. BbF had the highest levels in both the NSSP and SSP, which increased from 1.25 ng m^–3 in the NSSP to 1.44 ng m^–3 in the SSP. This increase should be related to the long-distance immigration of oil-related industries in Inner Mongolia, which was discussed in detail in Section 3.5 and 3.6 (Lang and Yang, 2014). Furthermore, the increases in levels of BkF, Fl, and BaA should be attributed to the elevated contribution of coal burning (Wang et al., 2020a). The decreases of AC and Fl in the SSP compared to the NSSP were in accordance with the reduced contribution of vehicle exhaust in the SSP. Acting as a widely used indicator for carcinogenic risk assessment of PAHs, Bap increased from 0.687 ng m^–3 in the NSSP to 0.749 ng m^–3 in the SSP. High BaP concentrations in the SSP should be more concerned though they were lower than the recommended threshold of 1 ng m^–3 by the U.S. EPA (Kong et al., 2018). BaP mainly comes from the burning of coal tar and petroleum, as well as cigarette smoke, vehicle exhaust, and cooking smoke (Wang et al., 2020a). When the actual situation in the sampling site was taken into account, the increases in contributions from petroleum combustion and coal burning would be the explanation for increased BaP. Furthermore, petroleum combustion was closely associated with the petrochemical industries within Inner Mongolia.

 
3.3 Ring Number Distribution of PAHs

The trigonometric diagram of PAHs composition with different ring numbers was shown in Fig. 3. HMW-PAHs (> 4-ring) and MMW-PAHs (MMW-PAHs; 4-ring) increased in the SSP compared with those in the NSSP, which was attributed to the increases in oil combustion and coal burning (Zhang et al., 2013; Feng and Cao, 2019). HMW- and MMW-PAHs increased from 5.65 to 6.62 ng m^–3 and from 3.04 to 3.30 ng m^–3, respectively. Consequently, the corresponding mass fractions of HMW-PAHs increased from 61.5% to 64.0%, while they decreased from 33.1% to 31.9% and from 5.34% to 4.06%, respectively. Singh et al. (2021) indicated that HMW-PAHs were more toxic than MMW- and LMW-PAHs, therefore the sandstorm derived the modifications of ring distribution of PAHs enhanced the exposure risks.

Fig. 3. Ratios and concentrations of PAHs with different ring numbers in the sandstorm and non-sandstorm periods.

 
3.4 Diagnostic Ratios (DRs)

Four DRs were calculated in this study, which have been widely used as the helpful indicators to qualitatively identify PAH emission sources (Kong et al., 2018; Wang et al., 2020a). The scatter diagram of the four PAH DRs in this study were shown in Fig. 4. The average FA/(FA + PY) ratio was lower in the SSP than that in the NSSP, indicating the relative importance of fuel burning in the SSP. At the same time, the values for BaA/(BaA + CHR), AN/(AN + PHE), and IP/(IP + BgP) fell within the range of 0.35–0.5, 0.01–0.1 and 0.4–0.55 in both the NSSP and SSP, implying the main PAHs sources in the two periods were similar. In a word, the sandstorm episode did not cause a great difference in PAH source types in both two periods, while made fuel combustion more important in the SSP compared to the NSSP.

Fig. 4. Graphic illustration of the diagnostic ratios for PAH source identifications.

 
3.5 Source Contributions

Fig. S1 presented the four PAHs source segments identified by the PMF model over the entire sampling period (ESP). Factor 1 possessed high loadings of NA, Fl, and PHE, and thus was recognized as the industrial emission. NA, Fl, and PHE were mostly emitted from the industrial boilers and factory chimneys (Ravindra et al., 2008). NA was a representative compound released from the iron-steel industry, and PHE mostly came from coke oven emissions. Factor 2 was characterized by high levels and contributions of DBA, IP, BaP, and BbF, indicating the emissions of oil leakage and combustion (OLC). DBA mostly comes from heavy oil combustion, BbF often marks oil leakage, and IP is a marker of diesel burning (Lang and Yang, 2014). Factor 3 possessed high levels of BkF, BbF, BaA, and FA, manifesting the coal burning (Wang et al., 2020a). Factor 4 with high levels of BaA, InP, BbF, BgP, Fl, AC, and BaP was determined as automobile exhaust emission source (Dickhut et al., 2000; He et al., 2008; Miguel et al., 1998).

Fig. 5 showed the source contributions in the SSP, NSSP, and ESP period, respectively. Fig. S2 showed the time series of source contributions for PAHs in whole sampling period. A large difference in source contributions existed between the SSP and NSSP. Oil leakage and combustion (OLC) fraction far increased from 18.0% in the NSSP to 34.4% in the SSP. Such a large increase in OLC fraction might be related to the immigration of emissions from petrochemical industries in Inner Mongolia in terms of no significant changes for emission intensities of local sources. A relative small increase was found in contributions of coal burning (CB), which enhanced from 21.4% to 25.0%. Unlike OLP and CB, vehicle exhaust (VE), and industrial emission (IE) decreased in the SSP compared to the NSSP. VE and IE fractions decreased from 36.2% to 27.6% and from 24.4% to 13.1%, respectively. The high reduction of 46.3% for IE was attributed to the strict emission control measures issued by local government in the SSP. As a result, the largest contributor changed from VE in the NSSP to OLC in the SSP.

Fig. 5. PAH source contributions for different periods.

 
3.6 Potential Source Contribution Function

Backward trajectory clustering result indicated that the air masses held five origins (Fig. S3). Three of five trajectories originated from Inner Mongolia and they accounted for 42.2% of the total trajectories. Moreover, another two trajectories came from south Hebei Province (36.6%) and east Bohai Sea (21.2%). Fig. 6 showed the PSCF analysis result. The potential source regions were located in local Dingxing County for three sources including VE, IE, and CB. Instead, OLC was partly impacted by air mass from Inner Mongolia, proving the OLC contribution in the sampling site was highly impacted by sandstorm event.

Fig. 6. Potential source contribution function (PSCF) analysis results for industrial emission (IE), coal burning (CB), oil leakage and combustion (OLC), and vehicle exhaust (VE).

 
3.7 Health Risks

Daily BaP_eq concentrations and ILCRs for adults and children in the entire sampling period (ESP) were shown in Fig. S4. Both them increased in the SSP compared with the NSSP. BaP_eq concentrations increased from 1.55 ng m^–3 to 1.89 ng m^–3 ranged from 0.51 ng m^–3 to 3.15 ng m^–3, and ILCR values elevated from 3.90 × 10^–7 to 4.74 × 10^–7 for adults, and from 2.41 × 10^–7 to 2.93 × 10^–7 for children, respectively, which was lower than the 4.27 ×10⁻⁶ in Anshan (Wang et al., 2020b) and 6.25 × 10⁻⁷ to 1.33 × 10⁻⁶ in Harbin in 2014 (Jiang et al., 2022).

PMF model combined with ILCR model was used to apportion the ILCR sources and the result was shown in Fig. 7. Attributing to the impacts of petrochemical industries in Inner Mongolia, OLC has become the largest contributor to ILCRs in the SSP with the contributions of 1.63 × 10^–7 and 1.01 × 10^–7 for adults and children, respectively, accounting for 34.4% of the total ILCRs. At the same time, however, VE contributed most to the ILCRs, with the contributions of 1.41 × 10^–7 and 8.72 × 10^–8 for adults and children, respectively.

Fig. 7. Source contributions to the incremental lifetime cancer risks (ILCRs).

4 CONCLUSIONS

This study examined the impacts of sandstorm episode on PM_2.5-bound PAHs in levels, compositions, sources, and exposure risks in a small city. We conducted an intensive observation in the spring of 2021, covering two sandstorm events including 16 to 17 March and 28 to 31 March.

1. Sandstorm largely elevated PM_2.5 concentration of 75.2 µg m^–3 in the sandstorm period (SSP) compared to 60.7 µg m^–3 in the non-sandstorm period (NSSP). However, ∑18PAHs exhibited a small increase from 8.90 ng m^–3 to 10.3 ng m^–3.
   
   
2. Unlike the total PAHs, sandstorm obviously changed the concentrations of individual PAH congener. More harmful PAH congeners including BaP, BbF, BkF, Fl, and BaA increased in the SSP compared with the NSSP. In consequence, mass percentages of high molecular weight PAHs (ring number > 4) rose from 61.5% to 64.0%.
   
   
3. Sandstorm largely alter the source contributions of PAHs. Immigration of emissions from petrochemical industries in Inner Mongolia far increased the oil leakage and combustion (OLC) contributions in the SSP compared with the NSSP, with the increase of 91.1%. At the same time, emission control on local industries and traffic limitation in the SSP reduced the contributions from vehicle exhaust (VE) and industrial emission (IE). Potential source contribution function (PSCF) model analysis showed that the OLC was partly originated from Inner Mongolia. Accordingly, the largest contributor of health risks changed from VE in the NSSP to OLC in the SSP.

 
ACKNOWLEDGEMENTS

This study was supported by the Beijing Natural Science Foundation (8212034), the Natura Science Foundation of Hebei Province (B2020502006 and B2020502007), the Science and Technology Projects of Baoding (2141ZF321), and the Fundamental Research Funds for the Central Universities (2020MS125).