Skip to main content
NCBI home page
EPA Author Manuscripts logoLink to EPA Author Manuscripts
. Author manuscript; available in PMC: 2019 Nov 25.
Published in final edited form as: Sci Total Environ. 2019 Jan 12;662:385–392. doi: 10.1016/j.scitotenv.2019.01.116

Source contributions of surface ozone in China using an adjoint sensitivity analysis

MY Wang a, Steve HL Yim a,b,c, DC Wong d, KF Ho e,b
PMCID: PMC6875754  NIHMSID: NIHMS1048013  PMID: 30690372

Abstract

Air pollution has become an adverse environmental problem in China, resulting in serious public health impacts. This study advanced and applied the CMAQ adjoint model to quantitatively assess the source-receptor relationships between surface ozone (O3) changes over different receptor regions and precursor emissions across all locations in China. Five receptor regions were defined based on the administrative division, including northern China (NC), southern China (SC), Pearl River Delta region (PRD), Yangtz River Delta region (YRD), and Beijing-Tianjin-Hebei region (BTH). Our results identified the different influential pathways of atmospheric processes and emissions to O3 pollution. We found that the atmospheric processes such as horizontal and vertical advection could offset the O3 removal through chemical reactions in VOC-limited areas inside the receptor regions. In addition, O3pollution can be induced by transport of O3 directly or its precursors. Our results of relative source contributions to O3 show that transboundary O3 pollution was significant in SC, NC and YRD, while the O3 pollution in PRD and BTH were more contributed by local sources. Anhui, Hubei and Jiangsu provinces were the three largest source areas of NOx and VOC emissions to O3 in SC (>52%) and YRD (>69%). NOx and VOC emissions from Tianjin and Beijing were the largest contributors to O3 in NC (>34%) and BTH (>51%). PRD was the dominant source areas of NOx (>89%) and VOC emissions (~98%) to its own regional O3.

Keywords: Surface ozone, Source-receptor relationships, Ozone-sensitive regime

Graphical Abstract

graphic file with name nihms-1048013-f0004.jpg

1. Introduction

Ambient air pollution has become one of the most severe environmental issues in China due to its rapid industrialization and urbanization accompanied by increasing anthropogenic emissions (Ning et al., 2018; Xie et al., 2017; Yin et al., 2017). Ozone (O3), as a strong oxidant, is the third most important greenhouse gas and is detrimental to human health, crop productivity and natural ecosystem (Feng et al., 2015; Gu et al., 2018; Karlsson et al., 2017). According to the national monitoring results, the mean daily 8-hour maximum O3concentrations in China increased from 69.5 ppbv in 2013 to 75.0 ppbv in 2015 (Wang et al., 2017). The three most developed regions, such as Beijing-Tianjin-Hebei region (BTH), Yangtz River Delta region (YRD) and Pearl River Delta region (PRD), have the highest anthropogenic emissions which have led to severe regional O3 pollution (Ma et al., 2016; Shen et al., 2015; Wei et al., 2015). Quantitative estimation of O3 source-receptor (S–R) relationships has become a critical issue in scientific understanding of O3 pollution abatement in China.

The O3-NOx-VOC sensitivity regimes, which describe the relationship between O3 and its local precursors (NOx and VOC), vary in urban and rural areas in China (Jin and Holloway, 2015). Classifying the photochemical regime of O3 formation (i.e. VOC-limited and NOx-limited) in an area helps determine which emissions should be targeted more to abate O3pollution. The major sources contributing to surface O3 above background levels include transport of O3 and its precursors (Sudo and Akimoto, 2007; Xue et al., 2014), photochemical production (Wang et al., 2010), or vertical transport (Wang et al., 2015).We note that the photochemical regime of O3 formation only describes local ozone-precursor relationships, which does not account for the nonlocal contributions of atmospheric processes to local O3. To fully understand the O3 generation and removal, it is necessary to identify the major influential atmospheric processes and emissions from local and nonlocal sources to O3 in a receptor region, in particular to China which has significant domestic transboundary air pollution (Luo et al., 2018; Hou et al., 2018; Yim et al., 2019; Gu and Yim, 2016).

The complexity in nonlinear chemical response makes it a challenge to quantify the local and transboundary source contributions of surface O3 pollution (Fiore et al., 2012). Solving the nonlinear optimal problem of O3 concentration-emission responses requires calculation of local sensitivities of O3 concentrations (Hakami et al., 2007). It is noted that the chemical transport models (CTM) can determine the source sensitivities and apportionments of modeled chemical species to input variables. Previous studies apportioned source contributions of O3 pollution in China based on ozone source apportionment technology (Gao et al., 2016; Li et al., 2012b), integrated process analysis (Fan et al., 2014; Li et al., 2012a), response surface modeling (Xing et al., 2011), or forward sensitivity analysis based on CTM (Itahashi et al., 2013; Wang et al., 2011). However, these methods are source-oriented and limited in cases of few specific input variables (Zhai et al., 2018). By contrast, adjoint method is receptor-oriented, and presents a unique advantage that it outputs the spatiotemporal variations of the sensitivities of the cost function to location-specific emission sources in one simulation. Therefore, adjoint method is computationally efficient than others and is used for source apportionment (Zhang et al., 2015).

This study advanced and applied the adjoint version of an air quality model to simulate O3sensitivity to NOx and VOC emissions. Specifically, we investigated the surface O3 response over a receptor region to emission changes at each grid cell. Our main objectives were to identify the local and transboundary influential pathways of emissions as well as atmospheric processes, and to quantify the source contributions of emissions to surface O3at intra-national as well as provincial levels in China. This study is expected to provide a fully scientific understanding of the S–R relationships of O3 in China. The model configurations are described in Section 2. The details of adjoint sensitivity methods, model evaluation and validation are elaborated from TextS1 to TextS3 in the Supporting Information (SI). In Section 3.1, the patterns of O3 concentration from baseline case in the study month are discussed. The S–R relationship and source contributions of O3 are respectively provided in 3.2 Surface ozone source-receptor relationships, 3.3 Relative source contributions to O. The discussions and conclusions are provided in Section 4.

2. Material and methods

The Community Multiscale Air Quality Modeling System (CMAQ v4.7.1) model developed by U.S.EPA (Byun and Schere, 2006) was applied in this study. The input meteorological fields for CMAQ simulations were driven by the Weather Research and Forecasting Model (WRF v3.7) model (Skamarock et al., 2008). Both WRF and CMAQ models were constructed with the spatial resolution of 27 km, centered at 37°N, 108.1°E on the Lambert conformal projection and covering the entire China region (Wang et al., 2019) (Fig. S1). Fig. S2 refers to the information of each province and municipal in China mentioned in this study. The boundary condition for CMAQ was provided by GEOS-Chem v8.3.2 model. CMAQ was configured with 10-day spin-up to minimize the influence of initial conditions. The Multi-resolution Emission Inventory for China (MEIC) in 2010 was used as the emission datasets (Li et al., 2017b) (http://www.meicmodel.org/). Monthly emissions with a horizontal resolution at 0.25° × 0.25° was firstly generated from MEIC and hourly emission data was further generated at a spatial resolution of 27 km (Gu and Yim, 2016). The model configurations and evaluations of CMAQ are respectively described in TextS3 and TextS4 in SI. We integrated the adjoint calculation core of CMAQ ADJ v4.5.4 model developed for gas-phase process (Hakami et al., 2007) into the more updated version of CMAQ (v4.7.1). The details of integration and validations are fully presented in Text S3 in SI.

The adjoint model calculated the gradient of a pre-defined cost function with respect to model inputs at each location within model domain. The cost function was a concentration-related function which was defined as hourly O3 concentration in this study. The details of adjoint method of CMAQ model are presented in Text S1. The gradients of cost function with respect to NOx(∂O3∂ENOx) and VOC (∂O3∂Evoc) emissions, were calculated by adjoint model during the integration. The species of VOC (VOC = OLE + ALD2 + ETH + FORM+ISOP+PAR + TOL + XYL) in CB4 chemical mechanism was considered in this study (Hogrefe et al., 2003).The values of gradients (∂O3∂ENOx) and (∂O3∂Evoc) were unitless, which would not reflect the actual influence of NOx and VOC emissions on O3concentrations. If an emission source with a relatively larger gradient had a small value of emission intensity, its actual influence would be negligible. Therefore, the semi-normalized sensitivity coefficients of O3 with respect to NOx emissions (ϕO3−ENOx) and VOC emissions (ϕO3−Evoc) were defined (An et al., 2016) as follows:

ϕO3−ENOx=∂O3∂ENOx×ENOx (Eq. 1)
ϕO3−Evoc=∂O3∂Evoc×Evoc (Eq. 2)

where, ENOxandEvoc referred to NOx and VOC emissions (μg/m3), respectively.

In this way, both ϕO3−ENOxand ϕO3−Evoc had physical meanings with the same unit as emissions and could directly quantify how emissions from sources in different locations influence the overall O3 concentrations in a receptor region.

This study focused on surface O3 in five target regions: the southern China (SC), the northern China (NC), and three major metropolitan regions which have substantial anthropogenic emissions: PRD, YRD and BTH regions (Fig. S1). The SC and NC regions were separated as different receptor regions due to their different climate condition. The Huai River-Qin Mountains, a geographical dividing line acted as a barrier which leaded to differences in climate, air pollution and agricultural diversity, was used to geographically divide SC and NC (Li et al., 2017a; Lin, 2017; Wu et al., 2013; Xue et al., 2008). It was known that mega cities and city clusters in China had severe O3 pollution problems due to larger amounts of anthropogenic emission sources especially in BTH (Qu et al., 2014; Tang et al., 2012), YRD (Li et al., 2012a; Wang et al., 2006), and PRD (Liu et al., 2013; Shao et al., 2009). Therefore, the regions of SC, NC, PRD, YRD and BTH were designed as the receptor regions of O3 concentration. Furthermore, in China, June is a typical month with relatively high overall O3 concentration (Liu et al., 2018). June 2010, a typical summer O3episode without tropical cyclones, was thus selected in this study due to the relatively high O3 levels in China.

3. Results

3.1. Baseline case

Our results show that the monthly mean O3 concentrations in June were 74.5 ppbv in China, 67.9 ppbv in SC, 79.1 ppbv in NC, 74.6 ppbv in PRD, 60.5 ppbv in YRD and 92.1 ppbv in BTH. The spatial distributions of monthly mean surface O3 show an explicit distinction between SC and NC (Fig. S8). In the NC plain, BTH, Taiyuan and Qinghai provinces had the highest O3 concentration levels (>90 ppbv). Most provinces in SC had relatively lower O3concentration levels (<40 ppbv). The relatively low O3 observed in Sichuan, Hubei, Hunan and Jiangxi provinces were related to their substantial NOx emissions (Fig. S9(a)), indicating O3 titration in these areas.

From 4 June in 2010, O3 concentration in BTH was the highest among regions in China (Fig. S10). The O3 concentration in NC was significantly higher than that in SC in June (Fig. S10). Previous studies have reported that a hot and dry environment provides favorable conditions for ozone production (Tong et al., 2018a, Tong et al., 2018b; Vautard et al., 2007; Zhao et al., 2015) as temperature is a critical factor to determine chemical reaction rates and thus largely affects O3 variations (Pusede et al., 2015). O3 concentrations in NC and BTH were thus relatively higher than the other regions in China due to the relatively high temperature (Fig. S11).

The relatively low O3 concentration in SC was probably due to the persistent rainfalls, especially in Guangxi, Guangdong and Fujian provinces (Guo et al., 2016). Based on climate records from NCC, CMA (http://cmdp.ncc-cma.net/cn/index.htm), the monthly average precipitation was 95.0 mm (close to normal) and temperature was 20.5 °C (1.0 °C above the normal) over entire China in June 2010. The cold air in the mid-latitude in East Asia was weakened, and the westerly jet moved southward (Fig. S12), which became a favorable circulation background for persistent rainfalls in SC (Fang et al., 2012). The monthly mean precipitation in NC especially in Northern China Plain (NCP), northeastern China, inner Mongolia and Shaanxi province, was lower than the long-term average. Barriopedro et al. (2012) reported that the northern China experienced severe drought in June of 2010.

The monthly mean O3 concentration in eastern China was 34.02 ppbv, which was within the range of 31.9 ± 17.1 ppbv as reported by Zhao et al. (2015). A positive correlation between O3 over YRD and air temperature in June was found (R = 0.75, p = 0.00000), whereas O3was negatively correlated with air pressure (R = −0.31, p < 0.00001). On the other hand, O3concentration was negatively correlated with water vapor ratio (R = −0.73, p < 0.00001) in PRD as well as in SC (R = −0.61, p = 0.00000). The diurnal variation in O3 over PRD in this study was closely correlated with regional mean daily precipitation from Fang et al. (2012).

3.2. Surface ozone source-receptor relationships

3.2.1. O3 sensitivity in southern and northern China

Figure 1 shows that the values ofϕO3−ENOx andϕO3−EVOC in most of locations were positive within SC and NC. The positive values indicate that NOx and VOC emissions in these locations had the positive relationship with the overall O3 concentration in SC and NC. Even though negative values of ϕO3−ENOxand ϕO3−EVOC were observed, their magnitudes were insignificant. The ϕO3−ENOx values in the capitals of each province such as in Wuhan and Xi’an, and in highly urbanized regions and cities such as in Shanghai and Nanjing, where substantial NOxemissions exist (Gao et al., 2017; Xie et al., 2016), were larger than 0.01 μg/m3, implying that a 10% reduction in NOx emissions in any grid of these areas would lead to >0.01 × 10% μg/m3 decrease in overall O3 concentration in SC. It was noticeable that these areas also had relatively higher ϕO3−EVOCvalues due to substantial VOC emissions. The magnitude ofϕO3−EVOC was relatively smaller than that of ϕO3−ENOx. his result indicates that reducing the same amount of NOx emissions in these areas would lead to a larger decrease in O3 in SC than VOC emissions did. The most influential source areas outside SC were mainly located at its adjacent areas with substantial emissions, i.e. the northern part of Anhui, Xi’an and Zhengzhou, whereas influences of other areas outside to O3 in all receptor regions were negligible. This pinpoints the significant local nature of the overall O3 in SC.

Figure 1.

Figure 1.

Open in a new tab

Spatial distributions of monthly mean sensitivity coefficients of O3 in (a) NC and (b) SC with respect to NOx emissions (ϕO3−ENOx) (μg/m3), and those in (c) NC and (d) SC with respect to VOC emissions (ϕO3−EVOC) (μg/m3). Red boundaries refer to NC in (a) and (c), and SC in (b) and (d).

To further explain the influential pathways of emissions from different source areas that affected O3 in receptor regions, the relative contributions from transport and chemical reactions as well as O3 chemical regimes were discussed. In this study, photochemical indicators such as H2O2/HNO3 and O3/NOy ratios were used to identify O3 photochemical production regimes (Sillman and He, 2002). Liu et al. (2010) suggested that the transition value of H2O2/HNO3 was 1.6 in July in China. Ye et al. (2016) estimated the threshold value of O3/NOy which should be 10.2 in PRD. Fig. S13 suggests that the O3 production regime was NOx-limited in most of areas in China while mega-cities and urban areas were in a VOC-limited regime.

To diagnose the relative contributions of chemical reactions and physical processes, a process analysis was conducted in CMAQ, see Fig. S14. The process analysis implemented in CMAQ estimated the contribution of each process that affects the overall simulated O3 in each synchronization time step (Li et al., 2012a). The contribution of each process was referred to the O3 changes before and after the respective process and was compared with the overall O3 changes in each synchronization time step. The typical VOC-limited areas within SC were PRD, YRD, and capitals, which had negative contributions from chemical reactions and positive contributions from transport process to local O3 formation, and had high positive ϕO3−ENOx and ϕO3−EVOC values. It indicates that the transport process can offset the reduction in local O3 concentration by chemistry.

Adjacent areas outside SC had relatively high positive ϕO3−ENOxand ϕO3−EVOCvalues in source areas such as Xi’an and Lanzhou where substantial anthropogenic emissions exist. Emissions from source areas outside SC might affect the overall O3 in SC through different influential pathways. In VOC-limited source areas outside SC, there were two possible major influential pathways: one was that O3 precursors were transported to SC to increase the overall O3 in SC through local chemical reactions; another pathway was that, during the transport of O3 precursors, O3 might be formed before reaching to receptor regions. For the NOx-limited source areas outside SC, there were three possible major influential pathways: the first two pathways were same as those as discussed above; the third one might be that the locally formed O3 in NOx-limited source areas was directly transported to SC. In the VOC-limited source areas, O3 decreased when increasing NOx due to the reaction of hydroxyl radical and NO2 (Yarwood et al., 2003). Thus, the third pathway might not be important in VOC-limited source areas outside SC.

The high positive ϕO3−ENOxand ϕO3−EVOCvalues in NC also corresponded to source areas with substantial NOx and VOC emissions such as in Beijing, Tianjin, Shijiazhuang and Jinan. The major influential pathways of the overall O3 in NC were similar to those in SC that were thoroughly discussed above. Large amounts of NOx emissions from industry, transportation and power plants coexisted in the southern Beijing, northern Tianjin, and southwestern Hebei, which leaded to relatively high ϕO3−EVOCand ϕO3−EVOCvalues. Although biogenic emissions were high in Yan Mountains that located in southern BTH and Taihang Mountains that located in western BTH, the impacts on surface O3 were low due to the lack of NOxemissions there (Qu et al., 2014). It was noticeable that northern part of Anhui province was the major influential source areas for the overall O3 in both SC and NC. The probable reasons were due to the rapid industrial growth in northern part of Anhui province (Qiu et al., 2017), and the conducive geographical condition that pollutants could be accumulated in the northern part of Anhui since it is the terminal of North China Plain (Yan and Wu, 2017). The most influential source areas outside NC were mainly located at its adjacent areas with substantial emissions, i.e. Nanjing and Wuhan, showing the similar significant local nature of the overall O3 as that in SC.

3.2.2. O3 sensitivity in three developed regions in China

In the other three receptor regions (PRD, YRD and BTH), the high positiveϕO3−EVOCandϕO3−EVOCvalues were related to the source areas with substantial emissions (Fig. 2). The source contributions from emissions to O3 in PRD and BTH were more concentrated in local areas within the receptor regions. The most influential source areas in PRD were the urban areas in Guangzhou, Dongguan and Shenzhen. In BTH, the major source contributions of emissions were from southern Beijing, northern Tianjin and southwestern Hebei. The most influential source areas to the overall O3 in YRD were not only from local areas inside YRD but also from large areas outside YRD such as the northern part of Anhui province, Wuhan and Nanchang. Emissions from outside YRD made comparable contributions as local emissions did, indicating the transboundary nature of O3 pollution in YRD. The significant transboundary O3 may be due to the prevailing southwesterly (~19.5% in frequency) and southerly (~17.9% in frequency) surface wind in June (Fig. S11). For example, the non-negligible contributions from emissions in PRD to overall O3 in YRD are obviously showed in Fig. 2. The prevailing wind direction was conducive to the pollutant transport from PRD to YRD.

Figure 2.

Figure 2.

Open in a new tab

Spatial distributions of the sensitivity of the sensitivity coefficients of O3 in (a) PRD, (b) YRD and (c) BTH with respect to NOx emissions (ϕO3−ENOx) (μg/m3), and those in (d) PRD, (e) YRD and (f) BTH with respect to VOC emissions (ϕO3−EVOC) (μg/m3). Red boundaries refer to PRD in (a) and (d), YRD in (b) and (e), and BTH in (c) and (f). PRD, YRD and BTH are zoomed and displayed in the left corner of each sub-figure.

3.3. Relative source contributions to O3 from different source areas and provinces

This section assesses the relative source contributions to the overall O3 concentration in five receptor regions from their NOx and VOC emissions (Fig. 3). A source contribution was referred to semi-normalized adjoint sensitivity coefficients ϕO3−EVOCandϕO3−EVOC(adjoint sensitivity × emissions). To calculate a relative source contribution, we first calculated the contribution of each source area by summing up the values of ϕO3−EVOCand ϕO3−EVOC over all the grid points in the receptive source area. The relative source contribution was then calculated based on the relative ratio of all source areas including SC, NC, PRD, YRD and BTH. We note that the source contributions of O3 in PRD, YRD and BTH regions were assessed individually so that YRD and PRD were excluded from SC, and BTH was excluded from NC. Each column in Fig. 3 refers to the apportionment of relative contributions by NOxor VOC emissions to their total contributions to the overall O3 concentration in each receptor region. For example, when summing up all relative source contributions of NOx and VOC emissions to O3 in NC, the total percentage values were 100%.

Fig 3.

Fig 3.

Open in a new tab

Relative source contributions (%) of NOx and VOC emissions from different source regions to the overall O3 in receptor regions. YRD and PRD were excluded from SC, and BTH was excluded from NC.

The transboundary nature of O3 pollution was found in SC, NC and YRD. Fig. 3 clearly shows the significant contributions of nonlocal emissions to O3 over SC, NC and YRD. In addition, the total relative source contributions from nonlocal emissions cannot be neglected. For example, nonlocal NOx emissions contributed to around 20% O3 in both NC and SC, and around 55% O3 in YRD. In addition, nonlocal VOC emissions contributed to around 11% O3 in YRD.

The contributions of different source areas to O3 in PRD and BTH were more localized. For example, local NOx and VOC emissions in PRD contributed 61.3% and 32.1% O3 in PRD, respectively. 71.9% and 13.5% O3 in BTH were from local NOx and VOC emissions.

When comparing the source contributions from different emission species, NOx emissions were the key contributor to O3 concentrations in all receptor regions. The NOx emissions from all source areas contributed around 80% O3 in receptor regions expect YRD in which NOx emissions contributed around 67% O3.

The relative source contributions of NOx and VOC emissions from each province in China to the overall O3 in the five receptor regions were quantified. Table S3 shows the relative contributions of provinces from NOx or VOC emissions that accounted for >95% of total contributions in each receptor region. For NOx emissions, Anhui, Hubei, and Jiangsu provinces were the three largest contributors (>16%) to O3 in SC and YRD. For VOC emissions, the major source areas of O3 in SC (>20%) were Guangdong and Hubei provinces, whereas those of O3 in YRD (>20%) were Hubei, Jiangsu provinces and Shanghai municipality. The two major source areas of O3 in BTH and NC were Beijing (>30%) and Tianjin (>50%) municipalities. NOx (~90%) and VOC (~99%) emissions from Guangdong province were dominant contributors to O3 in PRD. Most of the largest contributors of both NOx and VOC emissions were from the same province to O3 in receptor regions except for receptor YRD. The largest source contributions of NOx emissions were from Anhui (~36%) and of VOC emissions were from Jiangsu (~25%) to O3 in YRD. The NOx emissions in June in Anhui provinces were significantly high due to the biomass burningin the month (Ding et al., 2017). Jiangsu province was one of the three largest contributors of VOC emission in China (Wu et al., 2015).

4. Discussion

Emission control efficiency largely depends on the photochemical regimes of O3 production (i.e. VOC-limited or NOx-limited regimes). Indicators such as H2O2/HNO3 and O3/NOy ratios have been widely used to identify the O3 photochemical production regimes in China based on measurements (Xue et al., 2014), OMI satellite products (Jin and Holloway, 2015), and modeling simulations (Xie et al., 2014). Our results were consistent with previous studies that the megacities such as Beijing and Shanghai, and city clusters such as PRD and YRD regions were typical VOC-limited regimes in June. In our study, the indicators were used as the proxy for chemical regimes of O3 to help investigate the most probable influential pathways of emissions outside receptor regions to O3 within different receptors.

The main S-R relationships of O3 were investigated in previous studies which mostly were conducted based on site observations (Liu et al., 2010; Xu et al., 2008) and city clusters such as PRD (Wang et al., 2010). Our study focused on the S-R relationships over entire China including the PRD, YRD and BTH regions. Although we found that reducing both NOxand VOC emissions were effective for O3 control in China, reducing NOx emissions was more important.

While previous studies identified the source contributions of emissions to regional O3 in BTH (Wang et al., 2009), YRD (Gao et al., 2016) and PRD (Li et al., 2012b) during an O3episode, our study further investigated the monthly relative source contributions of NOx and VOC emissions from different source areas to O3 in receptor regions in China. June, a typical summer O3 episode, was selected as the study period so that our study can investigate more information of O3 pollution in China based on prior knowledge from previous studies. Our results show that the major local source areas were those that had substantial anthropogenic emissions such as capitals in each province, PRD and YRD. The nonlocal source areas had different influential pathways to O3 which had been discussed in the results section.

The transboundary O3 pollution in China had been found by many studies. For example, the transport of O3 from city regions to surrounding rural areas was found by Li et al. (2012a) in August in YRD; the transport of O3 precursors from central PRD to downwind rural areas was found by Wang et al. (2010) in October in PRD; the transport of O3 and its precursors from Tianjin and the south of Hebei to Beijing urban areas was found by Wang et al. (2009)during 26 June to 2 July. In our study, the relative contributions from individual chemistry and transport process were assessed as the supplementary information to understand the influential pathways of emissions to O3 in receptor regions. In addition, the relative source contributions of emissions from source areas outside receptor to O3 in receptor to highlight the transboundary nature of O3 pollution were also quantified. In SC, NC and YRD receptors, the transboundary nature of O3 was found. The relative contributions of NOx and VOC emissions from each province were quantified separately to provide provincial information for O3 pollution control.

5. Conclusions

This study advanced and applied the CMAQ adjoint model to calculate the O3 sensitivity with respect to emissions in China. The aim of this study was to investigate the source-receptor relationship between O3 changes over different receptor regions and its emission precursors at each grid cell, with a special focus on relative source contributions to O3 at different geographical levels such as at different source areas and at provincial level. This effort should provide a fully scientific understanding of O3 generation and removal, and its most influential pathways of emission and atmospheric processes in China. We found that the atmospheric processes such as horizontal and vertical transport can offset the O3 removal through chemical reactions in VOC-limited areas inside the receptor regions. In addition, O3pollution can be induced by transport of O3 directly or its precursors. Our results found that the reducing NOx and VOC emissions from local source areas would be effective for O3pollution control in all receptor regions especially in PRD and BTH. The transboundary O3pollution in SC, NC and YRD should also be paid attention since the large source contributions from nonlocal emissions were found. Therefore, reducing local emissions and nonlocal emissions were both important for O3 pollution control in SC, NC and YRD. Anhui, Hubei, and Jiangsu provinces were the three largest source areas of NOx emissions to O3 in SC and YRD. The largest contributors of VOC emission sources to O3 in SC were Guangdong and Hubei provinces, whereas those of O3 in YRD were Hubei and Jiangsu provinces, and Shanghai. The top two largest contributors of O3 in BTH and NC were emissions from Tianjin and Beijing. NOx and VOC emissions from Guangdong province were dominant contributors (>89%) to O3 in PRD. The largest source areas of NOx emission were Anhui and those of VOC emissions were Jiangsu to O3 in YRD. We should pay more attention to emission reductions in these provinces and municipalities to develop the most expeditious and cost-effective emission control strategies in China.

Supplementary Material

Suppl Data

Acknowledgments

This work was jointly funded by The Vice-Chancellor’s Discretionary Fund of The Chinese University of Hong Kong (grant no. 4930744) and the Environment and Conservation Fund(ECF project 07/2014). We would like to thank the Hong Kong Environmental ProtectionDepartment and the Hong Kong Observatory for providing air quality and meteorological data, respectively. We acknowledge the support of the CUHK Central High Performance Computing Cluster, on which computation in this work have been performed. The authors declare no competing financial interest.

Footnotes

Appendix A. Supplementary data

References

  1. An XQ, Zhai SX, Jin M, Gong S, & Wang Y. (2016). Development of an adjoint model of GRAPES–CUACE and its application in tracking influential haze source areas in north China. Geoscientific Model Development, 9(6), 2153–2165. 10.5194/gmd-9-2153-2016 [DOI] [Google Scholar]
  2. Barriopedro D, Gouveia CM, Trigo RM, & Wang L. (2012). The 2009/10 Drought in China: Possible Causes and Impacts on Vegetation. Journal of Hydrometeorology, 13(4), 1251–1267. 10.1175/JHM-D-11-074.1 [DOI] [Google Scholar]
  3. Byun D, Schere KL, 2006. Review of the governing equations, computational algorithms, and other components of the Models-3 community multiscale air quality (CMAQ) modeling system. Appl. Mech. Rev 59 (2), 51 10.1175/1.2128636 [DOI] [Google Scholar]
  4. Ding J, van der ARJ, Mijling B, Levelt PF, 2017. Space-based NOx emission estimates over remote regions improved in DECSO. Atmos. Meas. Tech 10 (3), 925–938. 10.5194/amt-10-925-2017 [DOI] [Google Scholar]
  5. Fan Q, Yu W, Fan S, Wang X, Lan J, Zou D, … Chan P. (2014). Process analysis of a regional air pollution episode over Pearl River Delta Region, China, using the MM5-CMAQ model. Journal of the Air & Waste Management Association, 64(4), 406–418. 10.1080/10962247.2013.816387 [DOI] [PubMed] [Google Scholar]
  6. Fang Y, Chen W, & Zhou W. (2012). Analysis of the role played by circulation in the persistent precipitation over South China in June 2010. Advances in Atmospheric Sciences, 29(4), 769–781. 10.1007/s00376-012-2018-7 [DOI] [Google Scholar]
  7. Feng Z, Hu E, Wang X, Jiang L, Liu X, 2015. Ground-level O3 pollution and its impacts on food crops in China: a review. Environ. Pollut 199, 42–48 10.1016/j.envpol.2015.01.016 [DOI] [PubMed] [Google Scholar]
  8. Fiore Arlene M., Naik, Spracklen DV, Steiner A, Unger N, Prather M, … Zeng G. (2012). Global air quality and climate. Chemical Society Reviews, 41(19), 6663 10.1039/c2cs35095e [DOI] [PubMed] [Google Scholar]
  9. Gao J, Zhu B, Xiao H, Kang H, Hou X, & Shao P. (2016). A case study of surface ozone source apportionment during a high concentration episode, under frequent shifting wind conditions over the Yangtze River Delta, China. Science of The Total Environment, 544, 853–863. 10.1016/j.scitotenv.2015.12.039 [DOI] [PubMed] [Google Scholar]
  10. Gao W, Tie X, Xu J, Huang R, Mao X, Zhou G, Chang L, 2017. Long-term trend of O3 in a mega City (Shanghai), China: characteristics, causes, and interactions with precursors. Sci. Total Environ 603–604, 425–433. 10.1016/j.scitotenv.2017.06.099 [DOI] [PubMed] [Google Scholar]
  11. Gu Y, & Yim SHL (2016). The air quality and health impacts of domestic trans-boundary pollution in various regions of China. Environment International, 97, 117–124. 10.1016/j.envint.2016.08.004 [DOI] [PubMed] [Google Scholar]
  12. Gu Y, Wong TW, Law CK, Dong GH, Ho KF, Yang Y, Yim SHL, 2018. Impacts of sectoral emissions in China and the implications: air quality, public health, crop production, and economic costs. Environ. Res. Lett 13 (8), 084008. 10.1088/1748-9326/aad138 [DOI] [Google Scholar]
  13. Guo J, Deng M, Lee SS, Wang F, Li Z, Zhai P, ... Li X, 2016. Delaying precipitation and lightning by air pollution over the Pearl River Delta. Part I: observational analyses: aerosol delays diurnal cycle in rainfall. J. Geophys. Res.-Atmos. 121 (11), 6472–6488. 10.1002/2015JD023257 [DOI] [Google Scholar]
  14. Hakami A, Henze DK, Seinfeld JH, Singh K, Sandu A, Kim S, … Li Q. (2007). The Adjoint of CMAQ. Environmental Science & Technology, 41(22), 7807–7817. 10.1021/es070944p [DOI] [PubMed] [Google Scholar]
  15. Hogrefe C, Sistla G, Zalewsky E, Hao W, Ku J-Y, 2003. An assessment of the emissions inventory processing systems EMS-2001 and SMOKE in grid-based air quality models. J. Air Waste Manage. Assoc 53 (9), 1121–1129. 10.1080/10473289.2003.10466261 [DOI] [PubMed] [Google Scholar]
  16. Hou X, Chan CK, Dong GH, Yim SHL, 2018. Impacts of transboundary air pollution and local emissions on PM2.5 pollution in the Pearl River Delta region of China and the public health, and the policy implications. Environ. Res. Lett [Google Scholar]
  17. Itahashi S, Uno I, & Kim S. (2013). Seasonal source contributions of tropospheric ozone over East Asia based on CMAQ–HDDM. Atmospheric Environment, 70, 204–217. 10.1016/j.atmosenv.2013.01.026 [DOI] [Google Scholar]
  18. Jin X, & Holloway T. (2015). Spatial and temporal variability of ozone sensitivity over China observed from the Ozone Monitoring Instrument: OZONE SENSITIVITY OVER CHINA. Journal of Geophysical Research: Atmospheres, 120(14), 7229–7246. 10.1002/2015JD023250 [DOI] [Google Scholar]
  19. Karlsson PE, Klingberg J, Engardt M, Andersson C, Langner J, Karlsson GP, Pleijel H, 2017. Past, present and future concentrations of ground-level ozone and potential impacts on ecosystems and human health in northern Europe. Sci. Total Environ 576, 22–35. 10.1016/j.scitotenv.2016.10.061 [DOI] [PubMed] [Google Scholar]
  20. Li J, Yang W, Wang Z, Chen H, Hu B, Li J, … Zhang Y. (2016). Modeling study of surface ozone source-receptor relationships in East Asia. Atmospheric Research, 167, 77–88. 10.1016/j.atmosres.2015.07.010 [DOI] [Google Scholar]
  21. Li L, Chen CH, Huang C, Huang HY, Zhang GF, Wang YJ, … Jang CJ (2012). Process analysis of regional ozone formation over the Yangtze River Delta, China using the Community Multi-scale Air Quality modeling system. Atmospheric Chemistry and Physics, 12(22), 10971–10987. 10.5194/acp-12-10971-2012 [DOI] [Google Scholar]
  22. Li S, Wang T, Huang X, Pu X, Li M, Chen P, … Wang M. (2018). Impact of East Asian Summer Monsoon on Surface Ozone Pattern in China. Journal of Geophysical Research: Atmospheres, 123(2), 1401–1411. 10.1002/2017JD027190 [DOI] [Google Scholar]
  23. Li X, Liu J, Mauzerall DL, Emmons LK, Walters S, Horowitz LW, & Tao S. (2014). Effects of trans-Eurasian transport of air pollutants on surface ozone concentrations over Western China: Trans-Eurasian Transport of O 3 to China. Journal of Geophysical Research: Atmospheres, 119(21), 12,338–12,354. 10.1002/2014JD021936 [DOI] [Google Scholar]
  24. Li Y, Lau AK-H, Fung JC-H, Zheng JY, Zhong LJ, & Louie PKK (2012). Ozone source apportionment (OSAT) to differentiate local regional and super-regional source contributions in the Pearl River Delta region, China: OZONE SOURCE APPORTIONMENT STUDY. Journal of Geophysical Research: Atmospheres, 117(D15), n/a-n/a. 10.1029/2011JD017340 [DOI] [Google Scholar]
  25. Lin M, Fiore AM, Horowitz LW, Cooper OR, Naik V, Holloway J, … Wyman B. (2012). Transport of Asian ozone pollution into surface air over the western United States in spring: ASIAN INFLUENCE ON U.S. SURFACE OZONE. Journal of Geophysical Research: Atmospheres, 117(D21), n/a-n/a. 10.1029/2011JD016961 [DOI] [Google Scholar]
  26. Lin M, Fiore AM, Horowitz LW, Langford AO, Oltmans SJ, Tarasick D, & Rieder HE (2015). Climate variability modulates western US ozone air quality in spring via deep stratospheric intrusions. Nature Communications, 6(1). 10.1038/ncomms8105 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Liu XH, Zhang Y, Xing J, Zhang Q, Wang K, Streets DG, ... & Hao JM (2010). Understanding of regional air pollution over China using CMAQ, part II. Process analysis and sensitivity of ozone and particulate matter to precursor emissions. Atmospheric Environment, 44(30), 3719–3727. [Google Scholar]
  28. Liu H, Wang XM, Pang JM, & He KB (2013). Feasibility and difficulties of China’s new air quality standard compliance: PRD case of PM2.5 and ozone from 2010 to 2025. Atmospheric Chemistry and Physics, 13(23), 12013–12027. 10.5194/acp-13-12013-2013 [DOI] [Google Scholar]
  29. Liu Hongyu, Jacob DJ, Chan LY, Oltmans SJ, Bey I, Yantosca RM, … Martin RV (2002). Sources of tropospheric ozone along the Asian Pacific Rim: An analysis of ozonesonde observations: ASIAN OUTFLOW OF OZONE. Journal of Geophysical Research: Atmospheres, 107(D21), ACH 3–1-ACH 3–19. 10.1029/2001JD002005 [DOI] [Google Scholar]
  30. Liu H, Wang XM, Pang JM, & He KB (2013). Feasibility and difficulties of China’s new air quality standard compliance: PRD case of PM 2.5 and ozone from 2010 to 2025. Atmospheric Chemistry and Physics, 13(23), 12013–12027. [Google Scholar]
  31. Liu X-H, Zhang Y, Xing J, Zhang Q, Wang K, Streets DG, … Hao J-M (2010). Understanding of regional air pollution over China using CMAQ, part II. Process analysis and sensitivity of ozone and particulate matter to precursor emissions. Atmospheric Environment, 44(30), 3719–3727. 10.1016/j.atmosenv.2010.03.036 [DOI] [Google Scholar]
  32. Liu Z, Yim SH, Wang C, & Lau NC (2018). The impact of the aerosol direct radiative forcing on deep convection and air quality in the Pearl River Delta region. Geophysical Research Letters, 45(9), 4410–4418. [Google Scholar]
  33. Luo M, Hou X, Gu Y, Lau NC, & Yim SHL (2018). Trans-boundary air pollution in a city under various atmospheric conditions. Science of The Total Environment, 618, 132–141. [DOI] [PubMed] [Google Scholar]
  34. Ma Z, Xu J, Quan W, Zhang Z, Lin W, & Xu X. (2016). Significant increase of surface ozone at a rural site, north of eastern China. Atmospheric Chemistry and Physics, 16(6), 3969–3977. 10.5194/acp-16-3969-2016 [DOI] [Google Scholar]
  35. Ning G, Wang S, Ma M, Ni C, Shang Z, Wang J, & Li J. (2018). Characteristics of air pollution in different zones of Sichuan Basin, China. Science of The Total Environment, 612, 975–984. [DOI] [PubMed] [Google Scholar]
  36. Pusede SE, Steiner AL, & Cohen RC (2015). Temperature and Recent Trends in the Chemistry of Continental Surface Ozone. Chemical Reviews, 115(10), 3898–3918. 10.1021/cr5006815 [DOI] [PubMed] [Google Scholar]
  37. Qiu F, Yuan H, Bai L, & Li F. (2017). Spatial-temporal heterogeneity of industrial structure transformation and carbon emission effect in Xuzhou Metropolitan Area. Chinese Geographical Science, 27(6), 904–917. [Google Scholar]
  38. Qiu F, Yuan H, Bai L, & Li F. (2017). Spatial-temporal heterogeneity of industrial structure transformation and carbon emission effect in Xuzhou Metropolitan Area. Chinese Geographical Science, 27(6), 904–917. [Google Scholar]
  39. Qu Y, An J, Li J, Chen Y, Li Y, Liu X, & Hu M. (2014). Effects of NO x and VOCs from five emission sources on summer surface O3 over the Beijing-Tianjin-Hebei region. Advances in Atmospheric Sciences, 31(4), 787–800. 10.1007/s00376-013-3132-x [DOI] [Google Scholar]
  40. Shao M, Zhang Y, Zeng L, Tang X, Zhang J, Zhong L, & Wang B. (2009). Ground-level ozone in the Pearl River Delta and the roles of VOC and NOx in its production. Journal of Environmental Management, 90(1), 512–518. [DOI] [PubMed] [Google Scholar]
  41. Shen J, Zhang Y, Wang X, Li J, Chen H, Liu R, … Lv W. (2015). An ozone episode over the Pearl River Delta in October 2008. Atmospheric Environment, 122, 852–863. 10.1016/j.atmosenv.2015.03.036 [DOI] [Google Scholar]
  42. Sillman S, & He D. (n.d.). Some theoretical results concerning O3-NOx-VOC chemistry and NOx-VOC indicators. Journal of Geophysical Research: Atmospheres, 107(D22), ACH 26–1-ACH 26–15. 10.1029/2001JD001123 [DOI] [Google Scholar]
  43. Skamarock WC, Klemp JB, Dudhi J, Gill DO, Barker DM, Duda MG, … Powers JG (2008). A Description of the Advanced Research WRF Version 3. Technical Report, (June), 113 10.5065/D6DZ069T [DOI] [Google Scholar]
  44. Sudo K, & Akimoto H. (2007). Global source attribution of tropospheric ozone: Long-range transport from various source regions. Journal of Geophysical Research, 112(D12). 10.1029/2006JD007992 [DOI] [Google Scholar]
  45. Tang G, Wang Y, Li X, Ji D, Hsu S, & Gao X. (2012). Spatial-temporal variations in surface ozone in Northern China as observed during 2009–2010 and possible implications for future air quality control strategies. Atmospheric Chemistry and Physics, 12(5), 2757–2776. 10.5194/acp-12-2757-2012 [DOI] [Google Scholar]
  46. Tong CHM, Yim SHL, Rothenberg D, Wang C, Lin CY, Chen YD, & Lau NC (2018). Projecting the impacts of atmospheric conditions under climate change on air quality over the Pearl River Delta region. Atmospheric Environment, 193, 79–87. [Google Scholar]
  47. Tong CHM, Yim SHL, Rothenberg D, Wang C, Lin CY, Chen YD, & Lau NC (2018). Projecting the impacts of atmospheric conditions under climate change on air quality over the Pearl River Delta region. Atmospheric Environment, 193, 79–87. [Google Scholar]
  48. Vautard R, Beekmann M, Desplat J, Hodzic A, & Morel S. (2007). Air quality in Europe during the summer of 2003 as a prototype of air quality in a warmer climate. Comptes Rendus Geoscience, 339(11–12), 747–763. 10.1016/j.crte.2007.08.003 [DOI] [Google Scholar]
  49. Wang TJ, Lam KS, Xie M, Wang XM, Carmichael G, & Li YS (2006). Integrated studies of a photochemical smog episode in Hong Kong and regional transport in the Pearl River Delta of China. Tellus B: Chemical and Physical Meteorology, 58(1), 31–40. 10.1111/j.1600-0889.2005.00172.x [DOI] [Google Scholar]
  50. Wang X, Li J, Zhang Y, Xie S, & Tang X. (2009). Ozone source attribution during a severe photochemical smog episode in Beijing, China. Science in China Series B: Chemistry, 52(8), 1270–1280. 10.1007/s11426-009-0137-5 [DOI] [Google Scholar]
  51. Wang X, Zhang Y, Hu Y, Zhou W, Lu K, Zhong L, ... & Russell AG (2010). Process analysis and sensitivity study of regional ozone formation over the Pearl River Delta, China, during the PRIDE-PRD2004 campaign using the Community Multiscale Air Quality modeling system. Atmospheric Chemistry and Physics, 10(9), 4423–4437. 10.5194/acp-10-4423-2010 [DOI] [Google Scholar]
  52. Wang X, Zhang Y, Hu Y, Zhou W, Zeng L, Hu M, ... & Russell AG (2011). Decoupled direct sensitivity analysis of regional ozone pollution over the Pearl River Delta during the PRIDE-PRD2004 campaign. Atmospheric environment, 45(28), 4941–4949. 10.1016/j.atmosenv.2011.06.006 [DOI] [Google Scholar]
  53. Wang N, Guo H, Jiang F, Ling ZH, & Wang T. (2015). Simulation of ozone formation at different elevations in mountainous area of Hong Kong using WRF-CMAQ model. Science of the total environment, 505, 939–951. 10.1016/j.scitotenv.2014.10.070 [DOI] [PubMed] [Google Scholar]
  54. Wang XueSong, Li J, Zhang Y, Xie S, & Tang X. (2009). Ozone source attribution during a severe photochemical smog episode in Beijing, China. Science in China Series B: Chemistry, 52(8), 1270–1280. 10.1007/s11426-009-0137-5 [DOI] [Google Scholar]
  55. Wang Y, Chan A, Lau GNC, Li Q, Yang Y, & Hung‐LamYim S. Effects of urbanization and global climate change on regional climate in the Pearl River Delta and thermal comfort implications. International Journal of Climatology. [Google Scholar]
  56. Wei W, lv Z, Cheng S, Wang L, Ji D, Zhou Y, … Wang L. (2015). Characterizing ozone pollution in a petrochemical industrial area in Beijing, China: a case study using a chemical reaction model. Environmental Monitoring and Assessment, 187(6). 10.1007/s10661-015-4620-5 [DOI] [PubMed] [Google Scholar]
  57. Wu B, Tian J, Bai C, Xiang M, Sun J, & Liu X. (2013). The biogeography of fungal communities in wetland sediments along the Changjiang River and other sites in China. The ISME journal, 7(7), 1299 10.1038/ismej.2013.29 [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Wu X, Huang W, Zhang Y, Zheng C, Jiang X, Gao X, & Cen K. (2015). Characteristics and uncertainty of industrial VOCs emissions in China. Aerosol Air Qual. Res, 15, 1045–1058. [Google Scholar]
  59. Xie M, Zhu K, Wang T, Yang H, Zhuang B, Li S, ... & Ouyang Y. (2014). Application of photochemical indicators to evaluate ozone nonlinear chemistry and pollution control countermeasure in China. Atmospheric environment, 99, 466–473. 10.1016/j.atmosenv.2014.10.013 [DOI] [Google Scholar]
  60. Xie M, Zhu K, Wang T, Chen P, Han Y, Li S, ... & Shu L. (2016). Temporal characterization and regional contribution to O3 and NOx at an urban and a suburban site in Nanjing, China. Science of the Total Environment, 551, 533–545. 10.1016/j.scitotenv.2016.02.047 [DOI] [PubMed] [Google Scholar]
  61. Xie Y, Dai H, Zhang Y, Hanaoka T, & Masui T. (2017). Health and Economic Impacts of Ozone Pollution in China: a provincial level analysis. Atmospheric Chemistry and Physics Discussions, 1–63. 10.5194/acp-2017-849 [DOI] [Google Scholar]
  62. Xing J, Wang SX, Jang C, Zhu Y, & Hao JM (2011). Nonlinear response of ozone to precursor emission changes in China: a modeling study using response surface methodology. Atmospheric Chemistry and Physics, 11(10), 5027–5044. 10.5194/acp-11-5027-2011 [DOI] [Google Scholar]
  63. Xu J, Zhang Y, Fu JS, Zheng S, & Wang W. (2008). Process analysis of typical summertime ozone episodes over the Beijing area. Science of The Total Environment, 399(1–3), 147–157. 10.1016/j.scitotenv.2008.02.013 [DOI] [PubMed] [Google Scholar]
  64. Xue F, Wang Y, Xu S, Zhang F, Wen B, Wu X, ... & Jin L. (2008). A spatial analysis of genetic structure of human populations in China reveals distinct difference between maternal and paternal lineages. European Journal of Human Genetics, 16(6), 705 10.1038/sj.ejhg.5201998 [DOI] [PubMed] [Google Scholar]
  65. Xue LK, Wang T, Gao J, Ding AJ, Zhou XH, Blake DR, … Wang WX (2014). Ground-level ozone in four Chinese cities: precursors, regional transport and heterogeneous processes. Atmospheric Chemistry and Physics, 14(23), 13175–13188. 10.5194/acp-14-13175-2014 [DOI] [Google Scholar]
  66. Yan S, & Wu G. (2017). SO2 Emissions in China–Their Network and Hierarchical Structures, Scient. Rep, 7, 46216 10.1038/srep46216 [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Yarwood G, Stoeckenius TE, Heiken JG, & Dunker AM (2003). Modeling weekday/weekend ozone differences in the Los Angeles region for 1997. Journal of the Air & Waste Management Association, 53(7), 864–875. 10.1080/10473289.2003.10466232 [DOI] [PubMed] [Google Scholar]
  68. Ye L, Wang X, Fan S, Chen W, Chang M, Zhou S, ... & Fan Q. (2016). Photochemical indicators of ozone sensitivity: application in the Pearl River Delta, China. Frontiers of Environmental Science & Engineering, 10(6), 15 10.1007/s11783-016-0887-1 [DOI] [Google Scholar]
  69. Yim SHL, Hou X, Guo J, & Yang Y. (2019). Contribution of local emissions and transboundary air pollution to air quality in Hong Kong during El Niño-Southern Oscillation and heatwaves. Atmospheric Research, 218, 50–58. 10.1016/j.atmosres.2018.10.021 [DOI] [Google Scholar]
  70. Yin P, Chen R, Wang L, Meng X, Liu C, Niu Y, … Kan H. (2017). Ambient Ozone Pollution and Daily Mortality: A Nationwide Study in 272 Chinese Cities. Environmental Health Perspectives, 125(11). 10.1289/EHP1849 [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Zhai S, An X, Zhao T, Sun Z, Wang W, Hou Q, ... & Wang C. (2018). Detection of critical PM 2.5 emission sources and their contributions to a heavy haze episode in Beijing, China, using an adjoint model. Atmospheric Chemistry and Physics, 18(9), 6241–6258. 10.5194/acp-18-6241-2018 [DOI] [Google Scholar]
  72. Zhang Q, Yuan B, Shao M, Wang X, Lu S, Lu K, … Liu SC (2014). Variations of ground-level O3 and its precursors in Beijing in summertime between 2005 and 2011. Atmospheric Chemistry and Physics, 14(12), 6089–6101. 10.5194/acp-14-6089-2014 [DOI] [Google Scholar]
  73. Zhao H, Wang S, Wang W, Liu R, & Zhou B. (2015). Investigation of Ground-Level Ozone and High-Pollution Episodes in a Megacity of Eastern China. PLOS ONE, 10(6), e0131878. 10.1371/journal.pone.0131878 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Suppl Data
NIHMS1048013-supplement-Suppl_Data.pdf (2.6MB, pdf)

RESOURCES