Response of aerosol chemistry to clean air action in Beijing, China: Insights from two-year ACSM measurements and model simulations

Journal Pre-proof Response of aerosol chemistry to clean air action in Beijing, China: Insights from twoyear ACSM measurements and model simulations Wei Zhou, Meng Gao, Yao He, Qingqing Wang, Conghui Xie, Weiqi Xu, Jian Zhao, Wei Du, Yanmei Qiu, Lu Lei, Pingqing Fu, Zifa Wang, Douglas R. Worsnop, Qiang Zhang, Yele Sun PII:

S0269-7491(19)33526-2

DOI:

https://doi.org/10.1016/j.envpol.2019.113345

Reference:

ENPO 113345

To appear in:

Environmental Pollution

Received Date: 1 July 2019 Revised Date:

3 October 2019

Accepted Date: 3 October 2019

Please cite this article as: Zhou, W., Gao, M., He, Y., Wang, Q., Xie, C., Xu, W., Zhao, J., Du, W., Qiu, Y., Lei, L., Fu, P., Wang, Z., Worsnop, D.R., Zhang, Q., Sun, Y., Response of aerosol chemistry to clean air action in Beijing, China: Insights from two-year ACSM measurements and model simulations, Environmental Pollution (2019), doi: https://doi.org/10.1016/j.envpol.2019.113345. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier Ltd.

Meteorology

? Emission Reductions

1

Response of aerosol chemistry to clean air action in Beijing, China:

2

insights from two-year ACSM measurements and model simulations

3

Wei Zhoua,b, Meng Gaoc, Yao Hea,b, Qingqing Wanga, Conghui Xiea,b, Weiqi Xua,b, Jian Zhaoa,b, Wei Dua,

4

Yanmei Qiua,b, Lu Leia,b, Pingqing Fud, Zifa Wanga,b,e, Douglas R. Worsnopf, Qiang Zhangg, and Yele

5

Suna,b,e,* a

6 7

State Key Laboratory of Atmospheric Boundary Layer Physics and Atmospheric Chemistry, Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing 100029, China

8

b

9

c

10

d

11 12

e

13

f

14 15

g

16 17

College of Earth and Planetary Sciences, University of Chinese Academy of Sciences, Beijing 100049, China

Department of Geography, Hong Kong Baptist University, Hong Kong SAR, China Institute of Surface-Earth System Science, Tianjin University, Tianjin 300072, China

Center for Excellence in Regional Atmospheric Environment, Institute of Urban Environment, Chinese Academy of Sciences, Xiamen 361021, China Aerodyne Research Inc., Billerica, Massachusetts 01821, USA

Ministry of Education Key Laboratory for Earth System Modeling, Department of Earth System Science, Tsinghua University, Beijing 100084, China *Corresponding author: Yele Sun ([email protected]), 40 Huayanli, Chaoyang District, Beijing 100029, China.

18 19

Abstract. Despite substantial mitigation of particulate matter (PM) pollution during the past decade in Beijing, the response

20

of aerosol chemistry to clean air action and meteorology remains less understood. Here we characterized the changes in

21

aerosol composition as responses to emission reductions by using two-year long-term measurements in 2011/2012 and

22

2017/2018, and WRF-Chem model. Our results showed substantial decreases for all aerosol species except nitrate from

23

2011/2012 to 2017/2018. Chloride exhibited the largest decrease by 65–89% followed by organics (37–70%), mainly due to

24

reductions in coal combustion emissions in winter and agriculture burning in June. Primary and secondary organic aerosol

25

(SOA) showed comparable decreases by 61–70% in fall and winter, and 34–63% in spring and summer, suggesting that

26

reductions in primary emissions might also suppress SOA formation. The changes in nitrate were negligible and even

27

showed increases due to less reductions in NOx emissions and increased formation potential from N2O5 heterogeneous

28

reactions. As a result, nitrate exceeded sulfate and became the major secondary inorganic aerosol species in PM with the

29

contribution increasing from 14–21% to 22–32%. Further analysis indicated that the reductions in aerosol species from

30

2011/2012 to 2017/2018 were mainly caused by the decreases of severely polluted events (PM1 >100 µg m-3). WRF-Chem

31

simulations suggested that the decreases in OA and sulfate in fall and winter were mainly resulted from emission reductions

32

(27–36% and 25–43%) and favorable meteorology (4–10% and 19–30%), while they were dominantly contributed by

33

emission changes in spring and summer. Comparatively, the changes in nitrate were mainly associated with meteorological

34

variations while the contributions of emissions changes were relatively small. Our results highlight different chemical

35

responses of aerosol species to emission changes and meteorology, suggesting that future mitigation of air pollution in China

36

needs species-targeted control policy.

37

Keywords: Clean air action; Meteorology; Emission reductions; WRF-Chem; Aerosol species

1

38

1.

Introduction

39

China has endeavored to tackle severe air pollution in the past decade, particularly in the Beijing-Tianjin-Hebei region

40

where high PM2.5 (particulate matter with an aerodynamic diameter less than 2.5 µm) concentrations have been reported

41

(Huang et al., 2014; Zhang et al., 2012; Zhang et al., 2015). PM2.5 exerts vital effects on environment, climate and human

42

health over a regional and even global scale (Lelieveld et al., 2015). With the successful implementation of “Action Plan on

43

Prevention and Control of Air Pollution” (hereafter 2013 clean air action) ( China State Council, 2013), the annual mean

44

PM2.5 concentrations in Beijing decreased by 43% from 89.5 µg m-3 in 2013 to 51 µg m-3 in 2018 ( Beijing Municipal

45

Ecological Environment Bureau, 2019; China, 2018) as a result of large reductions in anthropogenic emissions (Zheng et al.,

46

2018). However, the annual average PM2.5 concentration in 2018 is still much higher than the Chinese National Ambient Air

47

Quality Standard (an annual average for 35 µg m−3) and far exceeds the World Health Organization’s guideline of 10 µg m−3,

48

implying that air pollution remains yet to be mitigated in Beijing despite continuous improvement of air quality (Sun et al.,

49

2014; Xie et al., 2019). Therefore, it is of importance to evaluate the changes in aerosol composition and sources as

50

responses to emission reductions during the past five years.

51

Extensive ground-/vertical-based field campaigns and models have investigated the formation mechanisms, evolution

52

processes and sources of haze episodes (Guo et al., 2014; Hu et al., 2015; Li et al., 2017; Li et al., 2015; Sun et al., 2016a;

53

Sun et al., 2015a; Wang et al., 2018; Zhong et al., 2018). Aerosol composition and sources in Beijing were found to have

54

strong seasonal variations. Much higher contributions of secondary inorganic aerosol (SIA = sulfate + nitrate + ammonium)

55

to PM were observed in summer than winter, while organics and chloride showed opposite seasonal trends due to enhanced

56

primary emissions from coal combustion in winter (Sun et al., 2015b; Zhang et al., 2013; Zhao et al., 2013). Another

57

long-term measurement reported that neither SIA nor organics showed pronounced contribution differences between summer

58

and winter (Hu et al., 2017). These findings highlight complex seasonal variations of aerosol composition due to their

59

different sources and formation mechanisms, especially when coupled with fickle meteorology (Hu et al., 2015). Therefore, a

60

further investigation on seasonal variations of aerosol particles in a multi-year perspective is essential for a comprehensive

61

understanding of the roles of aerosol chemistry and meteorology.

62

Previous studies characterizing aerosol changes under different emission control scenarios usually lasted one or several

63

weeks during specific events (Du et al., 2017; Huang et al., 2010; Sun et al., 2016c; Xu et al., 2015; Zhang et al., 2016b;

64

Zhang et al., 2016c; Zhao et al., 2017), for example, the 2008 Beijing Olympic Games, the Asia Pacific Economic

65

Cooperation (APEC) summit and the Spring Festival. A recent study by Xu et al. (2019a) evaluated the changes of aerosol

66

chemistry in two different winter seasons of 2014 and 2016. The results showed ubiquitous increases in submicron species

67

by 10−130% from 2014 to 2016. These findings were limited to a certain time, which can be largely influenced by

68

meteorological conditions. For example, Wang et al. (2009) found that meteorology accounted for a larger portion in

69

reducing PM concentration than source control measures during the 2008 Beijing Olympics, while Sun et al. (2016c)

70

reported comparable contributions between emission control and meteorological conditions to the 2014 “APEC blue”. Zhou

71

et al. (2018b) found 30−44% average mass differences in submicron species under similar emission sources yet different

72

meteorological conditions. These results indicate large uncertainties in assessing the impacts of emission control on PM in a

73

relatively short time, highlighting a need for more long-term studies. However, previous studies on the response of aerosol

74

chemistry to long-term emission control, e.g., the 2013 clean air action, are very limited. Zheng et al. (2018) showed

75

substantial decreases of PM2.5 and gaseous precursors from 2010 to 2017 as a consequence of clean air actions by using

76

emission inventory method. Satellite observations also confirmed such large reductions in gaseous species (Lachatre et al.,

77

2018) and particulate matter (Cheng et al., 2019; Fu et al., 2017; Lachatre et al., 2018; Wang et al., 2013). Recently, Li et al.

78

(2019) found that the anthropogenic emission reductions are a major driver for the large decreases of aerosol species from

79

winter 2014 to 2017. Gao et al. (2019) underscored the effectiveness of China’s clean air action, and claimed that the strict

80

emission control measures have suppressed the unfavorable influences of climate on wintertime PM2.5 concentrations in 2

81

Beijing since 2002. Despite this, the impacts of clear air action and meteorology on aerosol composition changes over a

82

seasonal scale are yet to be illustrated through field measurements in Beijing.

83

In this study, we conducted two-year real-time submicron aerosol measurements in 2011/2012 and 2017/2018 by using

84

an aerosol chemical speciation monitor. The changes in seasonal average mass concentrations and compositions in the two

85

years are characterized to elucidate aerosol chemistry changes before and after the 2013 clean air action. The diurnal

86

variations of aerosol composition are also analyzed to gain insights into the changes in formation mechanisms from

87

2011/2012 to 2017/2018. Finally, the WRF-Chem model is used to quantitatively estimate the roles of emission reductions

88

and meteorology in changing aerosol species from 2011/2012 to 2017/2018.

89

2.

90

2.1 Sampling site and measurements

Experimental methods

91

The two-year field campaign was performed at the Institute of Atmospheric Physics, Chinese Academy of Sciences

92

(39°58′28″N, 116°22′16″E, ASL: 49 m) from October 2017 to August 2018 (2017/2018), and from October 2011 to August

93

2012 (2011/2012), respectively. An Aerodyne Aerosol Chemical Speciation Monitor (ACSM) (Ng et al., 2011) equipped with

94

standard vaporizer was deployed on the roof of ~8 m building for real-time measurements of non-refractory submicron

95

aerosol (NR-PM1) species with ~15 min time resolution, including organics (Org), sulfate (SO4), nitrate (NO3), ammonium

96

(NH4), and chloride (Chl). The sampling site is a typical urban site surrounded by residential areas and traffic roads, which

97

has been detailed in Sun et al. (2015b) and Zhou et al. (2018a). The detection limits of individual species for 30 min average

98

ACSM data have been reported in Sun et al. (2012), which are 0.54, 0.06, 0.07, 0.25, and 0.03 µg m-3 for organics, sulfate,

99

nitrate, ammonium and chloride respectively. The reproducibility uncertainties of ACSM was estimated to be 9−36% for

100

different species (Crenn et al., 2015). Hourly gaseous species such as sulfur dioxide (SO2; TEI 43i TL SO2 analyzer), carbon

101

monoxide (CO; TEI 48i TL CO analyzer) and ozone (O3; TEI 49C UV absorption analyzer) were acquired from the air

102

quality monitoring station in the Olympic Park nearby (~4 km). The 5 min average detection limits were 0.5 ppb for SO2, 4.0

103

ppm for CO and 1.0 ppb for O3, respectively, and the accuracy was approximate 5%. Meteorological parameters of wind

104

speed (WS), wind direction (WD), relative humidity (RH) and temperature (T) at 102 m were obtained from the Beijing 325

105

m Meteorological Tower, which is approximately 20 m from the sampling site. The annual anthropogenic emissions of SO2,

106

nitrogen oxides (NOx), nonmethane volatile organic compound (NMVOC) and ammonia (NH3) in China during 2011 and

107

2017 from Zheng et al., (2018) were summarized in Table S1, which shows 64% decrease in SO2 from 2011 to 2017 whereas

108

a small reduction in NOx (21%) and stable emissions in NMVOC and NH3.

109

2.2 ACSM data analysis

110

The detailed ACSM data analysis and calibration of NR-PM1 species in 2011/2012 have been given in Sun et al.

111

(2015b). Briefly, the ACSM was calibrated with pure ammonium nitrate to determine its response factor and relative

112

ionization efficiency (RIE) of NH4 (Ng et al., 2011), while the default RIEs were used for Org and Chl (Xu et al., 2015).

113

Because the RIE of sulfate was not calibrated in 2011/2012, a RIE of 1.0 was used which was derived from a two-week

114

inter-comparison between ACSM and High-Resolution Aerosol Mass Spectrometer (AMS hereafter) (Sun et al., 2015a). No

115

calibration was performed on ACSM in 2017/2018, thus the NR-PM1 species of ACSM were corrected by using the

116

regression coefficients between ACSM and AMS from two-month synchronous measurements. The detailed calibration of

117

AMS can be found in Xu et al. (2019b). As shown in Fig. S1, all NR-PM1 species measured by ACSM and AMS were well

118

correlated (R2=0.83−0.97), and the regression slopes varied from 0.61 to 1.54. The different ratios for different aerosol

119

species were likely due to the combined uncertainties for the two instruments, which were estimated to be ~20−40%

120

(Bahreini et al., 2009; Crenn et al., 2015). 3

121

Five organic aerosol (OA) factors were determined in 2011/2012 by using multi-linear engine 2 (ME-2) receptor model

122

(Canonaco et al., 2013), including three primary OA factors (POA), i.e., fossil fuel-related OA (FFOA), cooking OA (COA),

123

and biomass burning OA (BBOA), and two secondary OA factors (SOA), i.e., less oxidized oxygenated OA (LO-OOA) and

124

more oxidized oxygenated OA (MO-OOA) (Sun et al. (2018). Comparatively, positive matrix factorization (Paatero and

125

Tappert, 1994; Ulbrich et al., 2009) was first performed on seasonal OA matrix in 2017/2018 considering different emission

126

sources in different seasons. Because of the limited sensitivity and resolution of ACSM, only two factors were chosen in

127

each season, i.e., POA and SOA. We also utilized ME-2 for a better separation of OA factors using COA and FFOA as

128

constraints. The optimal three-factor ME-2 results still led to the mixture of POA and SOA (Fig. S2). For a better parallel

129

comparison, FFOA, COA and BBOA in 2011/2012 were combined as one POA factor while LO-OOA and MO-OOA were

130

combined as SOA. Fig. 1 shows the monthly average NR-PM1 species and meteorological parameters in 2011/2012 and

131

2017/2018, which covered four seasons including fall (October, November), winter (December, January, February), spring

132

(March, April, May), and summer (June, July, August).

133

2.1 Model simulations

134

We utilized the Weather Research and Forecasting model coupled with Chemistry (WRF-Chem; version 3.8.1) (Grell et

135

al., 2005) to quantify the roles of emissions and meteorological conditions in changing aerosol composition from 2011/2012

136

to 2017/2018. The capability of WRF-Chem model in simulating aerosol pollution in North China has been widely evaluated

137

in previous studies (Gao et al., 2016; Gao et al., 2018). In this study, the two nested domains of East Asia and North China

138

with 81 km and 27 km horizontal resolution, respectively were setup. Simulations of gas-phase reactions and aerosol

139

reactions were made by using the Carbon Bond Mechanism version Z (CBM-Z) (Zaveri and Peters, 1999) gas-phase

140

chemical mechanism coupled with the aerosol module MOSAIC (Model for Simulating Aerosol Interactions and Chemistry)

141

(Zaveri et al., 2008), and the same schemes of microphysics, radiative transfer, boundary layer as those in Gao et al. (2016)

142

were used. The WRF-Chem used the NCEP FNL (Final) Operational Global Analysis as the initial meteorological and

143

boundary conditions, while the simulations of MOZART-4 global model simulations of trace gases and aerosols for the

144

initial chemical and boundary conditions (Emmons et al., 2010).

145

The emission inventories of anthropogenic emissions of SO2, NOx, CO, non-methane volatile organic compounds

146

(NMVOCs), NH3, primary inorganic and organic PM in China were from the Multi-resolution emission inventory (MEIC, Li

147

et al., 2017b; Zheng et al., 2018), while those of outside China were from the MIX inventory (Li et al., 2017c). The biogenic

148

and open biomass burning emissions were derived using the online Model of Emissions of Gases and Aerosols from Nature

149

(MEGAN) (Guenther et al., 2006), and the Global Fire Emissions Database version 4 (GFEDv4) (Giglio et al., 2013).

150 151 152

To quantify the roles of emissions and meteorological conditions in changes of aerosol species, we conducted three sets of simulations. (1) M11E11: using meteorological data of year 2011/2012, and emission data of year 2011; the simulations cover

153

October, November, and December of 2011, and January to September of 2012; simulations were initialized monthly, with

154

five more days discarded as spin-up.

155

(2) M17E11: using meteorological data of year 2017/2018 and emission data of year 2011; the simulations cover

156

October, November, and December of 2017, and January to September of 2018; other settings follow the M11E11 case.

157 158

(3) M17E17: using meteorological data of year 2017/2018, and emission data of year 2017; other settings follow the M17E11 case.

159

The difference between M11E11 and M17E17 represents the simulated chemistry changes from 2011/2012 to

160

2017/2018, which is attributed to the contribution of emission reductions (difference between M17E11 and M17E17) and

161

meteorology (difference between M11E11 and M17E11).

4

162

3.

Results and discussion

163

3.1 Changes in aerosol mass concentrations from 2011/2012 to 2017/2018

164

The seasonal mean concentrations and composition of NR-PM1 species in 2011/2012 and 2017/2018 are shown in Fig.

165

2. The average (±1σ) mass loading of total NR-PM1 was 28.1±25.8, 22.0±25.6, 38.7±29.8 and 31.0±21.0 µg m-3 in fall,

166

winter, spring and summer of 2017/2018, respectively, which were decreased by 24–63% compared with those in 2011/2012

167

(Fig. 3), indicating substantial PM reductions in 2017/2018 after the implementation of clean air action in 2013. The most

168

dramatic decrease occurred in winter (63%) due to the influences of both emission controls and favorable meteorological

169

conditions as indicated by much higher frequency of northerly winds in 2017 than 2011 (Fig. S3, 75% vs. 39% of the time).

170

Comparatively, the smallest decrease (24%) in NR-PM1 was observed in spring. One reason was due to the high frequency of

171

southerly winds in 2017/2018 that facilitated the regional transport of air pollution from the south where air pollution was

172

more severe. As a result, the NR-PM1 in spring presented the highest mass concentration among the four seasons in

173

2017/2018. The NR-PM1 in 2017/2018 winter was also much lower than those previously reported in Beijing, for example

174

77.8 µg m-3 in 2012 (Hu et al., 2017), 60.5−92.9 µg m-3 from 2013 to 2016 (Xu et al., 2019a), suggesting a continuous

175

mitigation of heavy pollution in winter in recent years. Indeed, the frequency of high aerosol loadings (NR-PM1>100 µg m-3)

176

was substantially reduced from 20% in winter 2011/2012 to 2% in 2017/2018 (Fig. S4). We also noticed that the heavily

177

polluted events (NR-PM1 >160 µg m-3) due to biomass burning impacts in summer of 2011/2012 (Cheng et al., 2014; Hu et

178

al., 2016; Sun et al., 2016b; Zhang et al., 2016a) were significantly reduced in 2017/2018 because of the ban of burning crop

179

residual in recent years.

180

The changes in NR-PM1 species from 2011/2012 to 2017/2018 were significantly different among different aerosol

181

species (Fig. 3). Chloride showed the largest decrease by 65−89% in 2017/2018, followed by organics (37−70%) during all

182

seasons. Previous studies showed that chloride was mainly from coal combustion in winter (Wang et al., 2015) and biomass

183

burning emissions in summer (Sun et al., 2016b).Therefore, the results suggested an effective control of these two important

184

primary source emissions in recent years. For example, all coal-fired boilers in urban Beijing have been replaced by natural

185

gas, and most rural areas also switched from coal to natural gas or electricity for residential heating in winter 2017.

186

Consistently, organics, a large fraction from coal combustion emissions (Sun et al., 2013; Wang et al., 2015), presented the

187

largest decrease in winter (70%) than the other three seasons (37–62%). It is interesting to note that both primary and

188

secondary OA showed similar decreases as the total OA with the highest reduction in fall and winter seasons (61–70%),

189

while the reduction in SOA was more significant than POA in summer (63% vs. 45%). These results suggest that controlling

190

emissions of POA and volatile organic compounds might have helped reduce SOA formation substantially, and such an effect

191

was more important in summer when photochemical production was more significant (Sun et al., 2015b).

192

Secondary inorganic sulfate and nitrate showed very different changes from 2011/2012 to 2017/2018. Sulfate showed

193

considerable decreases during all seasons, and the decreases in fall and winter (60–67%) were nearly twice of those (33–44%)

194

in spring and summer. These results suggest a complex effect of emission control and meteorological conditions on

195

secondary sulfate formation in different seasons. As listed in Table S2, the gaseous SO2 showed large reductions by 56–82%

196

in 2017/2018, while O3 was increased approximately by a factor of 2. We found that the decreases in sulfate were not as

197

much as its precursor SO2, for example 66% vs. 82% in winter and 33% vs. 56% in spring. One possible explanation is the

198

increased oxidation capacity together with the decreased PM facilitated more efficient oxidation of SO2 to sulfate, consistent

199

with the observations over the eastern United States (Shah et al., 2018). Compared with sulfate, the seasonal changes in

200

nitrate were more complex. Nitrate showed the largest decrease in summer (48%) followed by winter (29%), yet the

201

reduction percentages were generally smaller than other species. While the nitrate difference in fall season was negligible, it

202

even showed an increase by 16% in spring from 2011/2012 to 2017/2018. One explanation is the consistently high NO2

203

concentration in Beijing during the last decade, and the decrease was much smaller than that of SO2 (Cheng et al., 2019). For

5

204

instance, the annual average concentration of NO2 was decreased approximately by 20% from 52–55 µg m-3 in 2011/2012 to

205

42–46 µg m-3 in 2017/2018 according to the Beijing Environmental Statement. Another possibility is the increasing nitrate

206

formation potential in PM pollution during fall and spring seasons, consistent with our previous measurements (Xu et al.,

207

2019a; Zhou et al., 2018b). Our results highlight the complex chemical response of secondary formation to the changes in

208

precursors and meteorological parameters.

209

3.2 Changes in aerosol composition from 2011/2012 to 2017/2018

210

The different variations in NR-PM1 species caused significant changes in aerosol composition from 2011/2012 to

211

2017/2018. As shown in Fig. 2, the contribution of OA to the total NR-PM1 mass was decreased significantly from 43–55%

212

in 2011/2012 to 36–41% in 2017/2018 although it was the major fraction of NR-PM1 in both years. Similarly, the chloride

213

contribution showed decreases from 2–8% to 0.5–5% as a response to efficient control of primary emissions. Comparatively,

214

the contribution of sulfate was comparable during all seasons in the two years despite the significant decreases in mass

215

concentrations of sulfate and SO2. These results suggest that sulfate continuously plays an important role in PM pollution

216

during the last decade despite its reduction. We also noticed remarkably enhanced sulfate contribution in summer, e.g., from

217

9–15% in other seasons to 26% in summer 2011/2012, and from 12–17% to 22% in summer 2017/2018, respectively,

218

attributing to more photochemical processing and aqueous-phase production of sulfate in summer (Sun et al., 2012).

219

Compared with organics and sulfate, nitrate displayed substantially higher contributions in 2017/2018 than 2011/2012,

220

and they were nearly twice in fall and winter (26–32% vs. 14–16%). In fact, the mass concentrations of nitrate far exceeded

221

those of sulfate in fall, winter and spring seasons of 2017/2018, and the mass ratios of nitrate to sulfate ranged from 1.8 to

222

3.1, which were 1.6–2.0 times of those in 2011/2012. These results indicate that nitrate has become the most important

223

secondary inorganic aerosol species and plays an increasing role in PM pollution in recent years (Li et al., 2019; Xu et al.,

224

2019a). One reason was likely due to much less reduction in NOx compared to SO2, e.g., 21% for NOx and 64% for SO2

225

(Lachatre et al., 2018; Zheng et al., 2018). Enhanced nitrate formation pathway from N2O5 heterogeneous hydrolysis could

226

also be an explanation (Wang et al., 2017; Xing et al., 2018). Note that the nitrate contributions in fall and winter of

227

2017/2018 (26–32%) were much higher than those previously reported in Beijing during 2010–2016, e.g., 21–24% in fall

228

and 11–20% in winter (Hu et al., 2017; Hu et al., 2016; Sun et al., 2013; Xu et al., 2019a; Xu et al., 2015; Zhang et al., 2014;

229

Zhao et al., 2017), further supporting the increasing role of nitrate in PM pollution. In comparison, the contribution of nitrate

230

to NR-PM1 was comparable to that of sulfate in summer, and the NO3/SO4 ratios were also close between 2011/2012 and

231

2017/2018 (1.01 vs. 1.08). These results suggest very different chemical responses of sulfate and nitrate formation to

232

emission changes between summer and the other three seasons. For example, nitrate showed a similar decrease as sulfate in

233

summer (48% vs. 44%) while it even showed increases during the other seasons in spite of large decreases in sulfate.

234

Fig. 4 shows the mean mass concentrations of NR-PM1 species as a function of total NR-PM1 mass loadings. Although

235

the seasonal average concentrations in nitrate showed overall decreases in 2017/2018 except spring (Fig. 3), we observed

236

consistently much higher nitrate in 2017/2018 than 2011/2012 during periods with the same PM levels (Fig. 4). Our results

237

demonstrate significantly enhanced nitrate formation from 2011/2012 to 2017/2018, while the decreases in total nitrate

238

concentrations were mainly caused by the reduced severely polluted events with NR-PM1 mass larger than 100 µg m-3.

239

Similarly, the changes in sulfate from 2011/2012 to 2017/2018 during periods with the same PM levels (NR-PM1 < 90 µg m-3;

240

Fig. 4) were small and even showed slight increases despite the significant reductions in their total average concentrations.

241

These results imply that sulfate was not synchronously mitigated as SO2 under relatively low polluted conditions, especially

242

when NR-PM1 was less than 50 µg m-3. Therefore, a stricter control of SO2 over a wider regional scale is critical to reduce

243

sulfate in PM in the future. Overall, our results illustrate very different chemical responses of nitrate and sulfate to emission

244

changes since the clear air action in 2013. Nitrate showed ubiquitous increases in both mass concentrations and contributions

245

under similar polluted conditions during all seasons from 2011/2012 to 2017/2018, and the contributions in NR-PM1 6

246

increased substantially from ~10% to 30–40% as a function of PM loadings in 2017/2018 (Fig. 4). Comparatively, the

247

changes in sulfate were comparably small and the contributions in NR-PM1 as a function of PM loading were also relatively

248

stable from 2011/2012 to 2017/2018 (Fig. 4). Due to the increasing importance of particulate nitrate in PM pollution, strict

249

emission control measures should be taken on NOx sources such as transportation and coal-fired plants sectors. Although

250

previous studies suggest the vital role of ammonia in nitrate formation as neutralizer of nitric acid (Guo et al., 2018),

251

effective reduction in ammonia might be unpractical as a result of its dominant emission from agriculture productions in

252

North China Plain (Li et al., 2019).

253

3.3 Changes in diurnal variations of aerosol species from 2011/2012 to 2017/2018

254

Fig. 5 compares the diurnal cycles of seasonal mass concentrations of NR-PM1 species between 2011/2012 and

255

2017/2018. Nearly all NR-PM1 species showed ubiquitously lower concentrations in 2017/2018 than 2011/2012 with the

256

ratios (Ratio2017/2011) being 0.08–1.18. The large variations in Ratio2017/2011 highlight different responses of aerosol chemistry

257

to emission control due to their different sources and formation mechanisms, e.g., primary emissions and secondary

258

production. Although organics showed unique diurnal characteristics with two peaks at meal times from cooking emissions

259

(He et al., 2010), the Ratios2017/2011 were quite flat with diurnal differences being less than 0.1 during four seasons. Similarly,

260

chloride showed the largest reductions by 60–92% associated with relatively constant diurnal ratios. These results suggest

261

that the patterns of anthropogenic activities for primary emissions did not change significantly during the past decade

262

although the total emissions were reduced substantially. Nitrate showed 12–33% higher mass loadings at nighttime in fall

263

and spring of 2017/2018 than 2011/2012, while they were comparable at daytime, likely due to more nocturnal nitrate

264

formation from N2O5 heterogeneous reaction in 2017/2018 (Wang et al., 2017; Zhou et al., 2018c). In winter, the nitrate

265

concentrations at daytime in 2017/2018 was much lower than those in 2011/2012 whereas the differences were small at

266

nighttime, further supporting stronger nocturnal nitrate formation in 2017/2018 winter. In contrast, the Ratios2017/2011 of

267

nitrate reached the lowest values in summer, e.g., 0.33–0.59, which was likely due to enhanced evaporative loss associated

268

with higher T in summer in 2017/2018 than 2011/2012.

269

3.4 Responses of aerosol changes to emission reductions and meteorology

270

3.4.1 Model evaluation

271

Fig. 6 shows the comparisons of model simulated seasonal mean concentrations of organics, sulfate, nitrate and

272

ammonium with the observations, and Fig. S5 presents the comparisons of meteorological parameters (R2 = 0.82-0.99).

273

Consistent with previous studies (Gao et al., 2016), WRF-Chem model underestimated OA substantially during all seasons

274

except winter. The largest underestimation in summer was likely due to the significant underestimation of SOA in models

275

(Lin et al., 2016). Comparatively, model simulations of OA agreed relatively well with the observations in winter when

276

primary OA was dominant. Although model underestimated OA in both 2011/2012 and 2017/2018, the decreases calculated

277

from model simulations overall agreed with those from the observations (Fig. 7). WRF-Chem simulated sulfate reasonably

278

well during most seasons except winter. In fact, previous studies also showed that including heterogeneous reactions of SO2

279

into models can improve the simulations substantially in winter haze episodes (Shao et al., 2019; Zheng et al., 2015). The

280

model ubiquitously overestimated nitrate during all seasons, particularly in the summer of 2011. Because the overestimation

281

percentages of nitrate by WRF-Chem model in summer were significantly different between 2011/2012 and 2017/2018, the

282

calculated changes from 2011/2012 to 2017/2018 through simulations and observations were also different (Fig. 7). For

283

example, the observed nitrate showed considerable reductions by ~40% in summer from 2011/2012 to 2017/2018, while

284

model simulations showed increases in June and July, and much smaller decreases in August. Such differences could cause

285

some uncertainties in quantification of relative contributions of emission reductions and meteorology in summer.

7

286

3.4.2 Contributions of emission reductions and meteorology

287

Fig. 7 shows the changes of aerosol species from 2011/2012 to 2017/2018 in Beijing caused by emission reductions and

288

meteorology, respectively. As shown in Fig. 7e, emission reductions played a dominant role in reducing OA, which caused

289

27–36% decreases in fall and winter from 2011/2012 to 2017/2018. In comparison, meteorology appeared not to play an

290

important role which only caused 4–10% decreases in OA except January (26%). Considering that the changes of OA

291

calculated from model simulations were 17–33% lower than those from the observations (Fig. 2a), the model might

292

underestimate the role of meteorology in OA decreases from 2011/2012 to 2017/2018. Compared with OA, emission

293

reductions and meteorology played comparable roles in sulfate changes in fall and winter, of which 25–43% reduction of

294

sulfate was caused by emission changes, and 19–30% by meteorology (53% in January). These results indicate that the

295

decreases in OA and sulfate in fall and winter in 2017/2018 were caused by both emission reductions and favorable

296

meteorological conditions. In fact, meteorology played the most important role in changing nitrate concentrations by

297

decreasing 20–28% and 30–65% in fall and winter, respectively from 2011/2012 to 2017/2018. Different from OA and

298

sulfate, the emission changes even caused increases in nitrate concentrations by 10–16% and 3–15% in fall and winter,

299

respectively. One reason is due to the small decreases in NOx emissions associated with increased oxidation capacity from

300

2011/2012 to 2017/2018 (Cheng et al., 2019). Therefore, our results highlight that air quality improvement in fall and winter

301

in Beijing in the future can benefit substantially from strict control of NOx emissions by decreasing nitrate concentration in

302

PM.

303

The unfavorable meteorology in spring 2018 caused ubiquitously increases in all aerosol species compared with 2012,

304

especially in March and April. Although the emission changes caused 42–56% and 52–88% reductions for OA and sulfate,

305

respectively in spring 2018, the total reductions of 30–48% and 18–42% were the smallest among the four seasons because

306

of negative impacts of meteorology. Similar to fall and winter, the emission changes led to increases in nitrate concentrations

307

by 3–31% in spring. Note that OA and sulfate overall showed the smallest decreases in March (30% and 37%, respectively),

308

while nitrate presented the largest increase (35%). In addition to the unfavorable meteorological conditions, the increased

309

emissions in March might be another explanation. During the fall and winter in 2017/2018, the government implemented

310

more stringent emission control measures than previous years in Beijing, Tianjin and 26 other cities in the provinces of

311

Hebei, Shandong, Henan and Shanxi to meet the goal of clean air action. After March 2018, the less strict emission controls

312

could lead to the increase in emissions to some extent. In summer except June, we found that meteorology was relatively

313

similar between the two years, and thus, the changes due to meteorological influences were small (< 15%) for all aerosol

314

species. Comparatively, the reductions in OA and sulfate due to emissions changes were significant in summer, which were

315

41–57% and 75–90%, respectively. These results suggest that the significant reductions in VOCs and SO2 from 2011/2012 to

316

2017/2018 (Cheng et al., 2019) can reduce the formation of SOA and sulfate in summer. The changes in nitrate due to

317

emission changes however were small (7–24%). In fact, the observations showed substantial decreases by 41–52% in

318

summer 2018. One reason was due to the significant overestimation of nitrate by models in summer 2018 (Fig. 6).

319

Figs. S6 and S7 further display the chemical responses of aerosol changes to meteorology and emission reductions

320

during four seasons in north China. Consistent with the results above, meteorological conditions played very different roles

321

in changing aerosol species between fall/winter and spring/summer seasons. The favorable meteorology caused ~20–60%

322

reductions in OA and sulfate in most regions in north China during fall and winter, and it dominated the nitrate changes by

323

contributing ~80–100%. Compared with meteorology, emission changes caused approximately 40–80% reductions in OA

324

and sulfate. These results indicate that emission controls over a regional scale under favorable meteorological conditions

325

were the major reason for air quality improvement in fall and winter in 2017/2018. Nitrate showed very different responses

326

to emission changes compared with OA and sulfate. As shown in Fig. 7c, the emission changes even caused increases in

327

nitrate concentrations in the regions to the north and west of Beijing which were likely due to the insignificant reduction of

328

NOx emissions in these areas. In contrast, nitrate in the regions to the south and southeast of Beijing showed clear decreases 8

329

due to emission reductions in winter. During spring and summer, the decreases in OA and sulfate in north China were

330

dominantly contributed by emission changes, while meteorological conditions played a minor and even negative impacts on

331

the decreases of these two species, especially in the regions to the north and northwest of Beijing. Differently, emission

332

changes and meteorology played comparable roles in changing nitrate concentration in these two seasons.

333

4.

Conclusion

334

We presented a detailed analysis of the changes in aerosol chemistry due to the 2013 clean air action in urban Beijing

335

with two-year (2011/2012 and 2017/2018) measurements of aerosol particle composition. Our results showed substantial

336

decreases in mass concentrations of all aerosol species except nitrate from 2011/2012 to 2017/2018. Chloride showed the

337

largest decrease by 65–89%, followed by organics (37–70%), mainly due to the reductions in coal combustion emissions in

338

winter and agriculture burning in June. Correspondingly, the contributions of OA and chloride in NR-PM1 were decreased

339

from 43–55% to 36–41%, and from 2–8% to 0.5%–5%, respectively. Further analysis of the changes of aerosol species as a

340

function of PM levels indicated that the reductions from 2011/2012 to 2017/2018 were mainly caused by the decreases of

341

severely polluted events (PM1 > 100 µg m-3), while the reductions under similar PM levels were much small.

342

Secondary inorganic sulfate and nitrate presented different chemical response to emission changes. While sulfate

343

showed significant reductions by 33%–67% owing to large decreases in SO2, the changes in nitrate were small and even

344

showed slight increases due to less reductions in NOx emissions and increased formation potential from N2O5 heterogeneous

345

reaction in 2017/2018. In fact, nitrate exceeded sulfate during all seasons except summer, and became the major secondary

346

inorganic aerosol species in PM. The nitrate contribution in PM also showed large increases from 14–21% in 2011/2012 to

347

22–32% in 2017/2018, indicating an increasingly important role of nitrate in PM pollution in Beijing. WRF-Chem

348

simulations showed that the decreases in OA and sulfate in fall and winter were mainly resulted from emission reductions

349

(27–36% and 25–43%) and favorable meteorology (4–10% and 19–30%), while emission changes played more important

350

roles in spring and summer. Comparatively, the changes in nitrate were mainly associated with those of meteorological

351

parameters while the contributions of emissions changes were relatively small. Our results highlight very different chemical

352

responses of aerosol species to emission changes and meteorology, which help mitigate air pollution in China in the future.

353

Author contributions. YS designed the research. MG provided WRF-Chem data. WZ, YH, QW, CX, WX, JZ, WD, YQ, and

354

LL conducted the measurements. WZ, MG, YH, and YS analyzed the data. MG, PF, ZW, DW, and YS reviewed and

355

commented on the paper. WZ and YS wrote the paper.

356

Competing interests. The authors declare that they have no conflict of interest.

357

Acknowledgements. This work was supported by the National Natural Science Foundation of China (91744207, 41575120,

358

41571130034), and the National Key Research and Development Program of China (2017YFC0209601, 2017YFC0212704).

359

References

360

Bahreini, R., Ervens, B., Middlebrook, A.M., Warneke, C., de Gouw, J.A., DeCarlo, P.F., Jimenez, J.L., Brock, C.A.,

361

Neuman, J.A., Ryerson, T.B., Stark, H., Atlas, E., Brioude, J., Fried, A., Holloway, J.S., Peischl, J., Richter, D., Walega,

362

J., Weibring, P., Wollny, A.G., Fehsenfeld, F.C., 2009. Organic aerosol formation in urban and industrial plumes near

363

Houston and Dallas, Texas. J. Geophys. Res. 114, D00F16.

364

Beijing Municipal Ecological Environment Bureau, 2019. The annual average concentration of PM2.5 in Beijing was 51

365

micrograms

per

cubic

meter

during

2018,

366

http://www.bjepb.gov.cn/bjhrb/xxgk/jgzn/jgsz/jjgjgszjzz/xcjyc/xwfb/844164/index.html. 9

and

the

air

quality

continued

to

improve.

367

Canonaco, F., Crippa, M., Slowik, J.G., Baltensperger, U., Prévôt, A.S.H., 2013. SoFi, an IGOR-based interface for the

368

efficient use of the generalized multilinear engine (ME-2) for the source apportionment: ME-2 application to aerosol

369

mass spectrometer data. Atmos. Meas. Tech. 6, 3649-3661.

370

Cheng, J., Su, J., Cui, T., Li, X., Dong, X., Sun, F., Yang, Y., Tong, D., Zheng, Y., Li, Y., Li, J., Zhang, Q., He, K., 2019.

371

Dominant role of emission reduction in PM2.5 air quality improvement in Beijing during 2013–2017: a model-based

372

decomposition analysis. Atmos. Chem. Phys. 19, 6125-6146.

373

Cheng, Y., Engling, G., He, K.B., Duan, F.K., Du, Z.Y., Ma, Y.L., Liang, L.L., Lu, Z.F., Liu, J.M., Zheng, M., Weber, R.J.,

374

2014. The characteristics of Beijing aerosol during two distinct episodes: impacts of biomass burning and fireworks.

375

Environ. Pollut. 185, 149-157.

376 377 378 379

China,

2018.

Air

quality

targets

set

by

the

Action

Plan

have

been

fully

realized.

http://www.gov.cn/xinwen/2018-02/01/content_5262720.htm. China State Council, 2013. Action Plan on Prevention and Control of Air Pollution, China State Council, Beijing, China. http: //www.gov.cn/zwgk/2013-09/12/content_2486773.htm.

380

Crenn, V., Sciare, J., Croteau, P.L., Verlhac, S., Fröhlich, R., Belis, C.A., Aas, W., Äijälä, M., Alastuey, A., Artiñano, B.,

381

Baisnée, D., Bonnaire, N., Bressi, M., Canagaratna, M., Canonaco, F., Carbone, C., Cavalli, F., Coz, E., Cubison, M.J.,

382

Esser-Gietl, J.K., Green, D.C., Gros, V., Heikkinen, L., Herrmann, H., Lunder, C., Minguillón, M.C., Močnik, G.,

383

O'Dowd, C.D., Ovadnevaite, J., Petit, J.E., Petralia, E., Poulain, L., Priestman, M., Riffault, V., Ripoll, A., Sarda-Estève,

384

R., Slowik, J.G., Setyan, A., Wiedensohler, A., Baltensperger, U., Prévôt, A.S.H., Jayne, J.T., Favez, O., 2015. ACTRIS

385

ACSM intercomparison – Part 1: Reproducibility of concentration and fragment results from 13 individual Quadrupole

386

Aerosol Chemical Speciation Monitors (Q-ACSM) and consistency with co-located instruments. Atmos. Meas. Tech. 8,

387

5063-5087.Du, W., Zhao, J., Wang, Y., Zhang, Y., Wang, Q., Xu, W., Chen, C., Han, T., Zhang, F., Li, Z., Fu, P., Li, J.,

388

Wang, Z., Sun, Y., 2017. Simultaneous measurements of particle number size distributions at ground level and 260 m on

389

a meteorological tower in urban Beijing, China. Atmos. Chem. Phys. 17, 6797-6811.

390

Emmons, L.K., Walters, S., Hess, P.G., Lamarque, J.F., Pfister, G.G., Fillmore, D., Granier, C., Guenther, A., Kinnison, D.,

391

Laepple, T., Orlando, J., Tie, X., Tyndall, G., Wiedinmyer, C., Baughcum, S.L., Kloster, S., 2010. Description and

392

evaluation of the Model for Ozone and Related chemical Tracers, version 4 (MOZART-4). Geosci. Model Dev. 3,

393

43-67.

394

Fu, X., Wang, S., Xing, J., Zhang, X., Wang, T., Hao, J., 2017. Increasing Ammonia Concentrations Reduce the Effectiveness

395

of Particle Pollution Control Achieved via SO2 and NOX Emissions Reduction in East China. Environ. Sci. Technol.

396

Lett. 4, 221-227.

397 398

Gao, M., Carmichael, G.R., Wang, Y., Saide, P.E., Yu, M., Xin, J., Liu, Z., Wang, Z., 2016. Modeling study of the 2010 regional haze event in the North China Plain. Atmos. Chem. Phys. 16, 1673-1691.

399

Gao, M., Han, Z., Liu, Z., Li, M., Xin, J., Tao, Z., Li, J., Kang, J.E., Huang, K., Dong, X., Zhuang, B., Li, S., Ge, B., Wu, Q.,

400

Cheng, Y., Wang, Y., Lee, H.J., Kim, C.H., Fu, J.S., Wang, T., Chin, M., Woo, J.H., Zhang, Q., Wang, Z., Carmichael,

401

G.R., 2018. Air quality and climate change, Topic 3 of the Model Inter-Comparison Study for Asia Phase III

402

(MICS-Asia III) – Part 1: Overview and model evaluation. Atmos. Chem. Phys. 18, 4859-4884.

403

Gao, M., Liu, Z., Zheng, B., Ji, D., Sherman, P., Song, S., Xin, J., Liu, C., Wang, Y., Zhang, Q., Wang, Z., Carmichael, G.,

404

McElroy, M., 2019. China's Clean Air Action has suppressed unfavorable influences of climate on wintertime

405

PM2.5 concentrations in Beijing since 2002. Atmos. Chem. Phys. Discuss. 1-19.

406 407 408 409

Giglio, L., Randerson, J.T., van der Werf, G.R., 2013. Analysis of daily, monthly, and annual burned area using the fourth-generation global fire emissions database (GFED4). J. Geophys. Res. - Biogeosci. 118, 317-328. Grell, G.A., Peckham, S.E., Schmitz, R., McKeen, S.A., Frost, G., Skamarock, W.C., Eder, B., 2005. Fully coupled “online” chemistry within the WRF model. Atmos. Environ. 39, 6957-6975. 10

410

Guenther, A., Karl, T., Harley, P., Wiedinmyer, C., Palmer, P.I., Geron, C., 2006. Estimates of global terrestrial isoprene

411

emissions using MEGAN (Model of Emissions of Gases and Aerosols from Nature). Atmos. Chem. Phys. 6, 3181-3210.

412

Guo, H., Otjes, R., Schlag, P., Kiendler-Scharr, A., Nenes, A., and Weber, R. J., 2018. Effectiveness of ammonia reduction on

413

control of fine particle nitrate. Atmos. Chem. Phys. 18, 12241-12256.Guo, S., Hu, M., Zamora, M.L., Peng, J., Shang,

414

D., Zheng, J., Du, Z., Wu, Z., Shao, M., Zeng, L., Molina, M.J., Zhang, R., 2014. Elucidating severe urban haze

415

formation in China. P. Natl. Acad. Sci. USA. 111, 17373-17378.

416

He, L.Y., Lin, Y., Huang, X.F., Guo, S., Xue, L., Su, Q., Hu, M., Luan, S.J., Zhang, Y.H., 2010. Characterization of

417

high-resolution aerosol mass spectra of primary organic aerosol emissions from Chinese cooking and biomass burning.

418

Atmos. Chem. Phys. 10, 11535-11543.

419 420

Hu, M., Guo, S., Peng, J., Wu, Z., 2015. Insight into characteristics and sources of PM2.5 in the Beijing–Tianjin–Hebei region, China. Natl. Sci. Rev. 2, 257-258.

421

Hu, W., Hu, M., Hu, W.-W., Zheng, J., Chen, C., Wu, Y., Guo, S., 2017. Seasonal variations in high time-resolved chemical

422

compositions, sources, and evolution of atmospheric submicron aerosols in the megacity Beijing. Atmos. Chem. Phys.

423

17, 9979-10000.

424

Hu, W., Hu, M., Hu, W., Jimenez, J.L., Yuan, B., Chen, W., Wang, M., Wu, Y., Chen, C., Wang, Z., Peng, J., Zeng, L., Shao,

425

M., 2016. Chemical composition, sources, and aging process of submicron aerosols in Beijing: Contrast between

426

summer and winter. J. Geophys. Res. - Atmos. 121, 1955-1977.

427

Huang, R.J., Zhang, Y., Bozzetti, C., Ho, K.F., Cao, J.J., Han, Y., Daellenbach, K.R., Slowik, J.G., Platt, S.M., Canonaco, F.,

428

Zotter, P., Wolf, R., Pieber, S.M., Bruns, E.A., Crippa, M., Ciarelli, G., Piazzalunga, A., Schwikowski, M., Abbaszade,

429

G., Schnelle-Kreis, J., Zimmermann, R., An, Z., Szidat, S., Baltensperger, U., El Haddad, I., Prevot, A.S., 2014. High

430

secondary aerosol contribution to particulate pollution during haze events in China. Nature. 514, 218-222.

431

Huang, X.F., He, L.Y., Hu, M., Canagaratna, M.R., Sun, Y., Zhang, Q., Zhu, T., Xue, L., Zeng, L.W., Liu, X.G., Zhang, Y.H.,

432

Jayne, J.T., Ng, N.L., Worsnop, D.R., 2010. Highly time-resolved chemical characterization of atmospheric submicron

433

particles during 2008 Beijing Olympic Games using an Aerodyne High-Resolution Aerosol Mass Spectrometer. Atmos.

434

Chem. Phys. 10, 8933-8945.

435 436

Lelieveld, J., Evans, J.S., Fnais, M., Giannadaki, D., Pozzer, A., 2015. The contribution of outdoor air pollution sources to premature mortality on a global scale. Nature 525, 367.

437

Lachatre, M., Fortems-Cheiney, A., Foret, G., Siour, G., Dufour, G., Clarisse, L., Clerbaux, C., Coheur, P.-F., Van Damme,

438

M., Beekmann, M., 2018. The unintended consequence of SO2 and NO2 regulations over China: increase of ammonia

439

levels and impact on PM2.5 concentrations. Atmos. Chem. Phys. Discuss. 1-25.

440 441 442 443 444 445

Li, H., Cheng, J., Zhang, Q., Zheng, B., Zhang, Y., Zheng, G., He, K., 2019. Rapid transition in winter aerosol composition in Beijing from 2014 to 2017: response to clean air actions. Atmos. Chem. Phys. Discuss. 1-26. Li, J., Du, H., Wang, Z., Sun, Y., Yang, W., Li, J., Tang, X., Fu, P., 2017a. Rapid formation of a severe regional winter haze episode over a mega-city cluster on the North China Plain. Environ. Pollut. 223, 605-615. Li, M., H. Liu, G. N. Geng, C. P. Hong, F. Liu, Y. Song, D. Tong, B. Zheng, H. Y. Cui, H. Y. Man, Q. Zhang, K. B. He, 2017b. Anthropogenic emission inventories in China: a review. Natl. Sci Rev. 4, 834-866.

446

Li, M., Zhang, Q., Kurokawa, J. I., Woo, J. H., He, K., Lu, Z., Ohara, T., Song, Y., Streets, D. G., Carmichael, G. R., Cheng,

447

Y., Hong, C., Huo, H., Jiang, X., Kang, S., Liu, F., Su, H., Zheng, B., 2017c. MIX: a mosaic Asian anthropogenic

448

emission inventory under the international collaboration framework of the MICS-Asia and HTAP. Atmos. Chem. Phys.

449

17, 935-963.

450 451 452

Li, P., Yan, R., Yu, S., Wang, S., Liu, W., Bao, H., 2015. Reinstate regional transport of PM2.5 as a major cause of severe haze in Beijing. P. Natl. Acad. Sci. USA. 112, E2739-2740. Lin, J., An, J., Qu, Y., Chen, Y., Li, Y., Tang, Y., Wang, F., Xiang, W., 2016. Local and distant source contributions to 11

453

secondary organic aerosol in the Beijing urban area in summer. Atmos. Environ. 124, 176-185.

454

Ng, N.L., Herndon, S.C., Trimborn, A., Canagaratna, M.R., Croteau, P.L., Onasch, T.B., Sueper, D., Worsnop, D.R., Zhang,

455

Q., Sun, Y.L., Jayne, J.T., 2011. An Aerosol Chemical Speciation Monitor (ACSM) for Routine Monitoring of the

456

Composition and Mass Concentrations of Ambient Aerosol. Aerosol Sci. Technol. 45, 780-794.

457 458

Paatero, P., Tappert, U., 1994. Positive and matrix factorization: A non-negative factor model with optimal utilization of error estimates of data values. Environmetrics 5, 111-126.

459

Shah, V., Jaeglé, L., Thornton, J.A., Lopez-Hilfiker, F.D., Lee, B.H., Schroder, J.C., Campuzano-Jost, P., Jimenez, J.L., Guo,

460

H., Sullivan, A.P., Weber, R.J., Green, J.R., Fiddler, M.N., Bililign, S., Campos, T.L., Stell, M., Weinheimer, A.J.,

461

Montzka, D.D., Brown, S.S., 2018. Chemical feedbacks weaken the wintertime response of particulate sulfate and

462

nitrate to emissions reductions over the eastern United States. P. Natl. Acad. Sci. USA 115, 8110-8115.

463

Shao, J., Chen, Q., Wang, Y., Lu, X., He, P., Sun, Y., Shah, V., Martin, R.V., Philip, S., Song, S., Zhao, Y., Xie, Z., Zhang, L.,

464

Alexander, B., 2019. Heterogeneous sulfate aerosol formation mechanisms during wintertime Chinese haze events: air

465

quality model assessment using observations of sulfate oxygen isotopes in Beijing. Atmos. Chem. Phys. 19, 6107-6123.

466

Sun, Y., Chen, C., Zhang, Y., Xu, W., Zhou, L., Cheng, X., Zheng, H., Ji, D., Li, J., Tang, X., Fu, P., Wang, Z., 2016a. Rapid

467

formation and evolution of an extreme haze episode in Northern China during winter 2015. Sci. Rep. 6, 27151.

468

Sun, Y., Du, W., Wang, Q., Zhang, Q., Chen, C., Chen, Y., Chen, Z., Fu, P., Wang, Z., Gao, Z., Worsnop, D.R., 2015a.

469

Real-Time Characterization of Aerosol Particle Composition above the Urban Canopy in Beijing: Insights into the

470

Interactions between the Atmospheric Boundary Layer and Aerosol Chemistry. Environ. Sci. Technol. 49, 11340-11347.

471

Sun, Y., Jiang, Q., Wang, Z., Fu, P., Li, J., Yang, T., Yin, Y., 2014. Investigation of the sources and evolution processes of

472

severe haze pollution in Beijing in January 2013. J. Geophys. Res. - Atmos. 119, 4380-4398.

473

Sun, Y., Jiang, Q., Xu, Y., Ma, Y., Zhang, Y., Liu, X., Li, W., Wang, F., Li, J., Wang, P., Li, Z., 2016b. Aerosol

474

characterization over the North China Plain: Haze life cycle and biomass burning impacts in summer. J. Geophys. Res.

475

Atmos. 121, 2508-2521.

476 477

Sun, Y., Wang, Z., Dong, H., Yang, T., Li, J., Pan, X., Chen, P., Jayne, J.T., 2012. Characterization of summer organic and inorganic aerosols in Beijing, China with an Aerosol Chemical Speciation Monitor. Atmos. Environ. 51, 250-259.

478

Sun, Y., Wang, Z., Wild, O., Xu, W., Chen, C., Fu, P., Du, W., Zhou, L., Zhang, Q., Han, T., Wang, Q., Pan, X., Zheng, H., Li,

479

J., Guo, X., Liu, J., Worsnop, D.R., 2016c. "APEC Blue": Secondary Aerosol Reductions from Emission Controls in

480

Beijing. Sci. Rep. 6, 20668.

481

Sun, Y., Xu, W., Zhang, Q., Jiang, Q., Canonaco, F., Prévôt, A.S.H., Fu, P., Li, J., Jayne, J., Worsnop, D.R., Wang, Z., 2018.

482

Source apportionment of organic aerosol from 2-year highly time-resolved measurements by an aerosol chemical

483

speciation monitor in Beijing, China. Atmos. Chem. Phys. 18, 8469-8489.

484

Sun, Y.L., Wang, Z.F., Du, W., Zhang, Q., Wang, Q.Q., Fu, P.Q., Pan, X.L., Li, J., Jayne, J., Worsnop, D.R., 2015b.

485

Long-term real-time measurements of aerosol particle composition in Beijing, China: seasonal variations,

486

meteorological effects, and source analysis. Atmos. Chem. Phys. 15, 10149-10165.

487 488 489 490

Sun, Y.L., Wang, Z.F., Fu, P.Q., Yang, T., Jiang, Q., Dong, H.B., Li, J., Jia, J.J., 2013. Aerosol composition, sources and processes during wintertime in Beijing, China. Atmos. Chem. Phys. 13, 4577-4592. Ulbrich, I.M., Canagaratna, M.R., Zhang, Q., Worsnop, D.R., Jimenez, J.L., 2009. Interpretation of organic components from Positive Matrix Factorization of aerosol mass spectrometric data. Atmos. Chem. Phys. 9, 2891-2918.

491

Wang, H., Lu, K., Chen, X., Zhu, Q., Chen, Q., Guo, S., Jiang, M., Li, X., Shang, D., Tan, Z., Wu, Y., Wu, Z., Zou, Q., Zheng,

492

Y., Zeng, L., Zhu, T., Hu, M., Zhang, Y., 2017. High N2O5 Concentrations Observed in Urban Beijing: Implications of a

493

Large Nitrate Formation Pathway. Environ. Sci. Technol. Lett. 4, 416-420.

494

Wang, Q., Sun, Y., Jiang, Q., Du, W., Sun, C., Fu, P., Wang, Z., 2015. Chemical composition of aerosol particles and light

495

extinction apportionment before and during the heating season in Beijing, China. J. Geophys. Res.-Atmos. 120, 12

496

12708-12722.

497

Wang, Q., Sun, Y., Xu, W., Du, W., Zhou, L., Tang, G., Chen, C., Cheng, X., Zhao, X., Ji, D., Han, T., Wang, Z., Li, J., Wang,

498

Z., 2018. Vertically resolved characteristics of air pollution during two severe winter haze episodes in urban Beijing,

499

China. Atmos. Chem. Phys. 18, 2495-2509.

500 501 502 503

Wang, W., Primbs, T., Tao, S., Simonich, S.L.M., 2009. Atmospheric Particulate Matter Pollution during the 2008 Beijing Olympics. Environ. Sci. Technol. 43, 5314-5320. Wang, Y., Zhang, Q.Q., He, K., Zhang, Q., Chai, L., 2013. Sulfate-nitrate-ammonium aerosols over China: response to 2000– 2015 emission changes of sulfur dioxide, nitrogen oxides, and ammonia. Atmos. Chem. Phys. 13, 2635-2652.

504

Xie, C., Xu, W., Wang, J., Wang, Q., Liu, D., Tang, G., Chen, P., Du, W., Zhao, J., Zhang, Y., Zhou, W., Han, T., Bian, Q., Li,

505

J., Fu, P., Wang, Z., Ge, X., Allan, J., Coe, H., Sun, Y., 2019. Vertical characterization of aerosol optical properties and

506

brown carbon in winter in urban Beijing, China. Atmos. Chem. Phys. 19, 165-179.

507

Xing, J., Ding, D., Wang, S., Zhao, B., Jang, C., Wu, W., Zhang, F., Zhu, Y., Hao, J., 2018. Quantification of the enhanced

508

effectiveness of NOx control from simultaneous reductions of VOC and NH3 for reducing air pollution in the Beijing–

509

Tianjin–Hebei region, China. Atmos. Chem. Phys. 18, 7799-7814.

510

Xu, W., Sun, Y., Wang, Q., Zhao, J., Wang, J., Ge, X., Xie, C., Zhou, W., Du, W., Li, J., Fu, P., Wang, Z., Worsnop, D.R., Coe,

511

H., 2019a. Changes in Aerosol Chemistry from 2014 to 2016 in Winter in Beijing: Insights from High Resolution

512

Aerosol Mass Spectrometry. J. Geophys. Res.- Atmos. 124.

513

Xu, W., Xie, C., Karnezi, E., Zhang, Q., Wang, J., Pandis, S.N., Ge, X., Wang, Q., Zhao, J., Du, W., Qiu, Y., Zhou, W., He, Y.,

514

Zhang, J., An, J., Li, Y., Li, J., Fu, P., Wang, Z., Worsnop, D.R., Sun, Y., 2019b. Summertime aerosol volatility

515

measurements in Beijing, China. Atmos. Chem. Phys. Discuss. 2019, 1-19.

516

Xu, W.Q., Sun, Y.L., Chen, C., Du, W., Han, T.T., Wang, Q.Q., Fu, P.Q., Wang, Z.F., Zhao, X.J., Zhou, L.B., Ji, D.S., Wang,

517

P.C., Worsnop, D.R., 2015. Aerosol composition, oxidation properties, and sources in Beijing: results from the 2014

518

Asia-Pacific Economic Cooperation summit study. Atmos. Chem. Phys. 15, 13681-13698.

519 520 521 522 523 524 525 526 527 528

Zaveri, R.A., Easter, R.C., Fast, J.D., Peters, L.K., 2008. Model for Simulating Aerosol Interactions and Chemistry (MOSAIC). J. Geophys. Res. - Atmos. 113. Zaveri, R.A., Peters, L.K., 1999. A new lumped structure photochemical mechanism for large-scale applications. J. Geophys. Res. - Atmos. 104, 30387-30415. Zhang, J.K., Cheng, M.T., Ji, D.S., Liu, Z.R., Hu, B., Sun, Y., Wang, Y.S., 2016a. Characterization of submicron particles during biomass burning and coal combustion periods in Beijing, China. Sci. Total. Environ. 562, 812-821. Zhang, J.K., Sun, Y., Liu, Z.R., Ji, D.S., Hu, B., Liu, Q., Wang, Y.S., 2014. Characterization of submicron aerosols during a month of serious pollution in Beijing, 2013. Atmos. Chem. Phys. 14, 2887-2903. Zhang, J.K., Wang, L.L., Wang, Y.H., Wang, Y.S., 2016b. Submicron aerosols during the Beijing Asia–Pacific Economic Cooperation conference in 2014. Atmos. Environ. 124, 224-231.

529

Zhang, Q., He, K., Huo, H., 2012. Cleaning China,s air. Nature 484, 161.

530

Zhang, R., Jing, J., Tao, J., Hsu, S.C., Wang, G., Cao, J., Lee, C.S.L., Zhu, L., Chen, Z., Zhao, Y., Shen, Z., 2013. Chemical

531

characterization and source apportionment of PM2.5 in Beijing: seasonal perspective. Atmos. Chem. Phys. 13,

532

7053-7074.

533 534

Zhang, R., Wang, G., Guo, S., Zamora, M.L., Ying, Q., Lin, Y., Wang, W., Hu, M., Wang, Y., 2015. Formation of urban fine particulate matter. Chem. Rev. 115, 3803-3855.

535

Zhang, Y., Sun, Y., Du, W., Wang, Q., Chen, C., Han, T., Lin, J., Zhao, J., Xu, W., Gao, J., Li, J., Fu, P., Wang, Z., Han, Y.,

536

2016c. Response of aerosol composition to different emission scenarios in Beijing, China. Sci. Total. Environ. 571,

537

902-908.

538

Zhao, J., Du, W., Zhang, Y., Wang, Q., Chen, C., Xu, W., Han, T., Wang, Y., Fu, P., Wang, Z., Li, Z., Sun, Y., 2017. Insights 13

539

into aerosol chemistry during the 2015 China Victory Day parade: results from simultaneous measurements at ground

540

level and 260 m in Beijing. Atmos. Chem. Phys. 17, 3215-3232.

541

Zhao, P.S., Dong, F., He, D., Zhao, X.J., Zhang, X.L., Zhang, W.Z., Yao, Q., Liu, H.Y., 2013. Characteristics of

542

concentrations and chemical compositions for PM2.5 in the region of Beijing, Tianjin, and Hebei, China. Atmos. Chem.

543

Phys. 13, 4631-4644.

544

Zheng, B., Tong, D., Li, M., Liu, F., Hong, C., Geng, G., Li, H., Li, X., Peng, L., Qi, J., Yan, L., Zhang, Y., Zhao, H., Zheng,

545

Y., He, K., Zhang, Q., 2018. Trends in China’s anthropogenic emissions since 2010 as the consequence of clean air

546

actions. Atmos. Chem. Phys., 14095-14111.

547

Zheng, B., Zhang, Q., Zhang, Y., He, K.B., Wang, K., Zheng, G.J., Duan, F.K., Ma, Y.L., Kimoto, T., 2015. Heterogeneous

548

chemistry: a mechanism missing in current models to explain secondary inorganic aerosol formation during the January

549

2013 haze episode in North China. Atmos. Chem. Phys. 15, 2031-2049.

550

Zhong, J., Zhang, X., Dong, Y., Wang, Y., Liu, C., Wang, J., Zhang, Y., Che, H., 2018. Feedback effects of boundary-layer

551

meteorological factors on cumulative explosive growth of PM2.5 during winter heavy pollution episodes in Beijing from

552

2013 to 2016. Atmos. Chem. Phys. 18, 247-258.

553

Zhou, W., Sun, Y., Xu, W., Zhao, X., Wang, Q., Tang, G., Zhou, L., Chen, C., Du, W., Zhao, J., Xie, C., Fu, P., Wang, Z.,

554

2018a. Vertical Characterization of Aerosol Particle Composition in Beijing, China: Insights From 3-Month

555

Measurements With Two Aerosol Mass Spectrometers. J. Geophys. Res. Atmos. 123, 13016-13029.

556

Zhou, W., Wang, Q., Zhao, X., Xu, W., Chen, C., Du, W., Zhao, J., Canonaco, F., Prévôt, A.S.H., Fu, P., Wang, Z., Worsnop,

557

D.R., Sun, Y., 2018b. Characterization and source apportionment of organic aerosol at 260 m on a meteorological tower

558

in Beijing, China. Atmos. Chem. Phys. 18, 3951-3968.

559

Zhou, W., Zhao, J., Ouyang, B., Mehra, A., Xu, W., Wang, Y., Bannan, T.J., Worrall, S.D., Priestley, M., Bacak, A., Chen, Q.,

560

Xie, C., Wang, Q., Wang, J., Du, W., Zhang, Y., Ge, X., Ye, P., Lee, J.D., Fu, P., Wang, Z., Worsnop, D., Jones, R.,

561

Percival, C.J., Coe, H., Sun, Y., 2018c. Production of N2O5 and ClNO2 in summer in urban Beijing, China. Atmos.

562

Chem. Phys. 18, 11581-11597.

14

563

Figure Captions:

564

Fig. 1. Monthly average NR-PM1 species (a-g; Org, SO4, NO3, NH4, Chl, POA, SOA) and meteorological parameters (h-j;

565

WS, T, RH) in 2011/2012 and 2017/2018.

566

Fig. 2. Seasonal average mass concentrations (left panel) and mass fractions (right panel) of NR-PM1 species (POA, SOA,

567

SO4, NO3, NH4, Chl) in 2011/2012 and 2017/2018.

568

Fig. 3. Average change percentages of NR-PM1 species (Org, SO4, NO3, NH4, Chl, POA, SOA) from 2011/2012 to

569

2017/2018 during four seasons.

570

Fig. 4. Variations of mass concentrations of NR-PM1 species as a function of NR-PM1 mass loadings during four seasons in

571

2011/2012 and 2017/2018. The change percentages from 2011/2012 to 2017/2018 are shown on the right axis.

572

Fig. 5. Diurnal variations of NR-PM1 species during four seasons in 2011/2012 and 2017/2018. Also shown are the ratios

573

between 2017/2018 and 2011/2012 (Ratio2017/2011) on the right axis.

574

Fig. 6. Comparisons of model simulated monthly average concentrations with the observations for NR-PM1 species (Org,

575

SO4, NO3, NH4) in 2011/2012 and 2017/2018.

576

Fig. 7. (a-d) Comparisons of simulated monthly change percentages of NR-PM1 species from 2011/2012 to 2017/2018 (=

577

[2011/2012 – 2017/2018]/[2011/2012]×100) with the obervations; (e-h) average monthly changes caused by emission

578

reductions and meteorology.

15

(b) SO4

20 15 10 5 0

(c) NO3

30

(g) SOA

20 10 0

30 20 10 0

(i) T

80 60 40 20 0

(j) RH

-1

4 3 2 1 0

o

(d) NH4

(e) Chl

4 2 0

(f) POA

WS (m s )

6

16 12 8 4 0

T ( C)

12 8 4 0

-3

20 15 10 5 0

2011/2012 2017/2018

Mass Conc. (µg m )

(a) Org

RH (%)

-3

Mass Conc. (µg m )

40 30 20 10 0

Oct Nov Dec Jan Feb Mar Apr May Jun

Jul Aug

(h) WS

Oct Nov Dec Jan Feb Mar Apr May Jun

Jul Aug

Fig. 1. Monthly average NR-PM1 species (a-g; Org, SO4, NO3, NH4, Chl, POA, SOA) and meteorological parameters (h-j; WS, T, RH) in 2011/2012 and 2017/2018.

Fig. 2. Seasonal average mass concentrations (left panel) and mass fractions (right panel) of NR-PM1 species (POA, SOA, SO4, NO3, NH4, Chl) in 2011/2012 and 2017/2018.

Fig. 3. Average change percentages of NR-PM1 species (Org, SO4, NO3, NH4, Chl, POA, SOA) from 2011/2012 to 2017/2018 during four seasons.

-3

Org (μg m )

-3

SO4 (μg m )

-3

NO3 (μg m )

0 -10

2011 2017 Changes

20

-30 -40 -50

(b)

40

30

20

20

0

10

-20 -40

(c)

40

120

30

80

20

40

10

Change perc. (%)

-3

Summer

-20

0 30 NH4 (μg m )

Spring

40

0 50

0 80

(d)

60

20

40

10

20 0

0 20 -3

Winter

60

0 40

Chl (μg m )

Fall

(a)

80

0

(e)

15

-20

10

-40

5

-60

0

-80 0

40

80

120 160 200

0

40

80

120 160 200

0

40

80

120 160 200

0

40

80

120 160 200

NR-PM1 (µg m-3)

Fig. 4. Variations of mass concentrations of NR-PM1 species as a function of NR-PM1 mass loadings during four seasons in 2011/2012 and 2017/2018. The change percentages from 2011/2012 to 2017/2018 are shown on the right axis.

-3

Mass Conc. (µg m )

50 40 30 20 10 0

Org

10 8 6 4 2 0

SO4

12

NO3

(a) Fall 2011/2012 2017/2018 Ratio

12

8

8

4

4

Chl

0

4

1.0 0.8 0.6 0.4 0.2 0.0

0

12

0

16 12 8 4 0

20 15 10 5 0

10 8 6 4 2 0

10 8 6 4 2 0

12

0

1.0 0.8 0.6 0.4 0.2 0.0

8 6 4 2 0

4 3 2 1 0

4 3 2 1 0

1.0 0.8 0.6 0.4 0.2 0.0

4

8 6 4 2 0

Ratio2017/2011

0

4 NH4

(d) Summer

50 40 30 20 10 0 20 15 10 5 0

8

8 6 4 2 0

(c) Spring

40 30 20 10 0

12

8 0

(b) Winter

50 40 30 20 10 0

8 12 16 20 24

0

4

8 4

8 12 16 20 24 0 4 Hour of Day

8 12 16 20 24

0

4

1.0 0.8 0.6 0.4 0.2 0.0 1.2 0.8 0.4 0.0

8 12 16 20 24

Fig. 5. Diurnal variations of NR-PM1 species during four seasons in 2011/2012 and 2017/2018. Also shown are the ratios between 2017/2018 and 2011/2012 (Ratio2017/2011) on the right axis.

30 -3

Sim (μg m )

25

(b) SO4

20 15

Fall Winter Spring Summer

10 5 0

30 -3

Sim (μg m )

25

0

10 20 -3 Obs (μg m )

30

(c) NO3

20 15 10 5 0

0

10 20 -3 Obs (μg m )

30

Fig. 6. Comparisons of model simulated monthly average concentrations with the observations for NR-PM1 species (Org, SO4, NO3, NH4) in 2011/2012 and 2017/2018.

Fig. 7. (a-d) Comparisons of simulated monthly change percentages of NR-PM1 species from 2011/2012 to 2017/2018 (= [2011/2012 – 2017/2018]/[2011/2012]×100) with the obervations; (e-h) average monthly changes caused by emission reductions and meteorology.

Highlights: •

Substantial decreases for organics and chloride from 2011/2012 to 2017/2018

•

Small changes and even increases for nitrate in Beijing during the past six years

•

Decreases in aerosol species were driven by the reductions of severely haze events

•

Different chemical responses of aerosol species to emission change and meteorology