Ambient Ozone Control in a Photochemically ... - ACS Publications

Subscriber access provided by The Libraries of the | University of North Dakota

Article

Ambient ozone control in a photochemically active region: short-term despiking or long-term attainment? Jiamin Ou, Zibing Yuan, Junyu Zheng, Zhijiong Huang, Min Shao, Zekun Li, Xiaobo Huang, Hai Guo, and Peter Louie Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.6b00345 • Publication Date (Web): 02 May 2016 Downloaded from http://pubs.acs.org on May 9, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 30

Environmental Science & Technology

Ambient ozone control in a photochemically active region: short-term despiking or long-term attainment? Jiamin Ou,† Zibing Yuan,*,† Junyu Zheng,*,† Zhijiong Huang,† Min Shao,‡ Zekun Li,† Xiaobo Huang,† Hai Guo,§ and Peter K.K. Louie∥ †

School of Environment and Energy, South China University of Technology, Guangzhou 510006, China

‡

State Key Joint Laboratory of Environmental Simulation and Pollution Control, College of Environmental Sciences and Engineering, Peking University, Beijing 100871, China §

Department of Civil and Environmental Engineering, Hong Kong Polytechnic University, Hung Hom, Hong Kong, China ∥ Hong

Kong Environmental Protection Department, Revenue Tower, 5 Gloucester Road, Wanchai, Hong Kong, China

1 ACS Paragon Plus Environment


1

ABSTRACT

2

China has made significant progress decreasing the ambient concentrations of most air

3

pollutants, but ozone (O3) is an exception. O3 mixing ratios during pollution episodes are

4

far higher than the national standard in the Pearl River Delta (PRD), thus greater

5

evidence-based control efforts are needed for O3 attainment. By using a validated O3

6

modeling system and the latest regional emission inventory, this study illustrates that

7

control strategies for short-term O3 despiking and long-term attainment in the PRD may

8

be contradictory. VOC-focused controls are more efficient for O3 despiking in urban and

9

industrial areas, but significant NOx emission reductions and a subsequent transition to a

10

NOx-limited regime are required for O3 attainment. By tracking O3 changes along the

11

entire path towards long-term attainment, this study recommends to put a greater focus

12

on NOx emission controls region-wide. Parallel VOC reductions around the Nansha port

13

are necessary in summertime and should be extended to the urban and industrial areas

14

in fall with a flexibility to be strengthened on days forecasted to have elevated O3.

15

Contingent VOC-focused controls on top of regular NOx-focused controls would lay the

16

groundwork for striking a balance between short-term despiking and long-term

17

attainment of O3 concentrations in the PRD.


Page 2 of 30

Page 3 of 30

18


TOC/Abstract Art

19



20

1. INTRODUCTION

21

Ambient fine particulate matter (PM2.5) and ozone (O3) are criteria air pollutants

22

both associated with significant adverse health effects, but PM2.5 is the primary focus of

23

current air quality management in China. Due to the implementation of a series of

24

stringent control measures, the nearly unabated increase in PM2.5 has recently been

25

curbed and even reversed. Ambient PM2.5 levels were reduced by more than 20% from

26

2013 to 2015 in China’s three major city clusters: Beijing-Tianjin-Hebei, Yangtze River

27

Delta and Pearl River Delta (PRD).1 However, ambient O3 levels have been increasing

28

even with the reduction of PM2.5 levels. In the PRD, the annual average ground-level O3

29

mixing ratio increased from 24 ppbv in 2006 to 29 ppbv in 2014.2 The maximum 1-hour

30

O3 mixing ratio in the PRD reached 150~220 ppbv in summer and fall O3 episodes, and

31

the daily maximum 8-hour standard of ~82 ppbv (160 µg/m3) was exceeded on 11% of

32

the days.2 Thus, the increase in O3 partially negates the public health benefit attained

33

from the PM2.5 reduction.

34

O3 is a secondary pollutant that forms in the troposphere when its mix of precursors,

35

mainly nitrogen oxides (NOx) and volatile organic compounds (VOCs), react in the

36

presence of sunlight. Atmospheric photochemical processes governing O3 formation and

37

loss have been investigated for decades.3-5 The complexity of the O3 formation

38

mechanism is partly reflected by its nonlinear response to changes of precursors under

39

different chemical regimes. For a VOC-limited regime, VOC reduction decreases the O3

40

level, while NOx control measures result in an initial increase of O3, the so-called ‘NOx

41

disbenefits’, due to the competition between NO2 and VOC for OH radicals.6 The O3-


Page 4 of 30

Page 5 of 30


42

VOC-NOx relationship also shows spatiotemporal variabilities due to the impact of

43

meteorological conditions and emission source characteristics.4,7-9

44

Despite the well-established chemistry, the abatement of tropospheric O3 through

45

precursor controls still faces difficulties in many areas. For example, earlier approaches

46

to reduce O3 levels in California’s South Coast Air Basin (SoCAB) generally focused on

47

VOC control, given the fact that O3 formation in most areas was recognized as VOC-

48

limited. O3 levels decreased sharply with VOC-focused control before 2005, but leveled

49

off afterwards even with further VOC emission reductions. SoCAB’s South Coast Air

50

Quality Management District is now focusing primarily on NOx reductions to achieve

51

long-term multipollutant (O3 and PM2.5) attainment objectives.10-12 Such a transition

52

implies a discrepancy between short-term and long-term O3 control measures in areas

53

with significant O3 pollution.

54

Similar difficulties for controlling O3 in the SoCAB are also present in the PRD, a

55

photochemically active region with intense precursor emissions. O3 formation regimes in

56

the PRD are spatially intertwined – urban, suburban and industrial areas are typically

57

VOC-limited while rural areas are more NOx-limited.13-18 And as the regime may change

58

under different synoptic conditions, O3 control strategies cannot simply target solely

59

VOCs or NOx throughout the PRD. Furthermore, O3 levels are far higher than the national

60

standard and control efforts to achieve short-term reductions of peak 1-hour O3

61

concentrations (defined as ‘O3 despiking’ in this study) may contradict those for long-

62

term attainment. Formulating an effective O3 control strategy relies on not only scientific

63

evidence but practical feasibility. Initially, an anthropogenic VOC (AVOC)-to-NOx



64

reduction ratio of 3:1 was recommended in urban, suburban and industrial areas of the

65

PRD, as specified in the Guangdong – Hong Kong Joint Emission Reduction Plan (JERP)

66

effective from 1997 to 2010.19 However, such an ambitious AVOC reduction plan proved

67

not successful due to AVOC sources in the PRD being too scattered and diverse for

68

effective control.

69

In this study, we explore the feasibility of disentangling O3 control issues in the PRD

70

by using a WRF/SMOKE-PRD/CMAQ modeling system. A set of control scenarios is

71

designed to investigate the O3-VOC-NOx relationship and their spatial distribution during

72

elevated O3 days. O3 responses to different AVOC and NOx co-control schemes are

73

analyzed, and policy implications on O3 controls for short-term despiking and long-term

74

attainment and practical multipollutant control strategies are discussed.

75

2. DATA AND METHODS

76

2.1 Ozone modeling system

77

The PRD region, composed of 9 administrative cities in Guangdong Province, is the

78

targeted domain of the O3 modeling system (Figure 1). Lambert-Conformal projection

79

centered at 28.5 °N 114 °E, with two true latitudes for the projection at 15 °N and 40 °N,

80

is used as the basic projected coordinate in the modeling system.


Page 6 of 30

Page 7 of 30


81 82 83 84 85

Figure 1. The nested domains used in the WRF/SMOKE-PRD/CMAQ modeling system. D3 covers the entire Pearl River Delta region. Also illustrated in the bottom right figure are the targeted areas for O3 simulation and analysis, i.e. Guangzhou, Dongguan, Nansha and Jiangmen.

86

The Weather Research and Forecast (WRF) model v3.3 is used to provide

87

meteorological data. The 1°1° global reanalysis data obtained from the National

88

Centers for Environmental Protection (NCEP) and the land use data from Moderate

89

Resolution Imaging Spectroradiometer (MODIS) are adopted. The other physical options

90

in WRF include the Rapid Radioactive Transfer Model (RRTM) scheme for long wave

91

radiation, the Dudhia scheme for short wave radiation, the Noah Land Surface Model,

92

the Yonsei Planetary Boundary Lay (PBL) scheme, the WRF Single-Moment 6-class

93

(WSM6) scheme for microphysics, and the Kain-Fritsch scheme for cumulus

94

parameterization. A three-level two-way nested domain is set for WRF, with horizontal 7 ACS Paragon Plus Environment


95

resolutions of 27 km, 9km and 3km, respectively. As illustrated in Figure 1, the coarse

96

domain (D1) covers much of East Asia, Southeast Asia and the northwestern Pacific, the

97

second domain (D2) covers most of Guangdong province, and the fine domain (D3), the

98

target area in this study, includes the PRD, Hong Kong and Macau. WRF runs with 26

99

vertical layers where 18 layers are under 1000 meters above sea surface level.

100

The 2010 bulk emission inventories for Hong Kong and the PRD is adopted20,21 and

101

transformed by the SMOKE-PRD emission processor into hourly gridded model-ready

102

emission data.22 SMOKE-PRD is a successfully localized version of SMOKE with the PRD

103

local source class clarification codes and spatial, temporal and chemical speciation

104

information.22 The emission inventories cover comprehensive anthropogenic sources

105

with the latest local emission factors, detailed local activity data and updated source

106

classification, including but not limited to power plants, residential combustion, on-road

107

and non-road mobile sources, industrial sources, solvent-use sources, and biomass

108

burning, for pollutants of sulfur dioxide (SO2), carbon monoxide (CO), NOx, AVOC and

109

particulates (PM10 and PM2.5). The Model of Emissions of Gases and Aerosols from

110

Nature (MEGAN) model is used to estimate biogenic VOC (BVOC) emissions.23 Emissions

111

from D2 and D3 within Guangdong Province are derived from Guangdong emission

112

inventories.24,25 The model-ready emission data from the Multi-resolution Emission

113

Inventory for China (MEIC) Model with 1°1° resolution (http://www.meicmodel.org/)

114

and Regional Emission inventory in ASia (REAS) are adopted for D1 and D2 but outside

115

Guangdong.

116

CMAQ v4.7.1 with the CB-05 gas-phase chemical mechanism is used to simulate O3


Page 8 of 30

Page 9 of 30


117

mixing ratios and its relationship with VOC and NOx emissions.26 Also incorporated in the

118

CMAQ model are the ISORROPIA v1.7 module27 for inorganic aerosol simulation, SOAP

119

module for organic aerosol simulation, and the Asymmetric Cloud Model 2 (ACM2) PBL

120

scheme28 for cloud treatment and eddy simulation in the PBL. Clean air data provided

121

with CMAQ is used to generate boundary conditions for D1, and the simulated field in

122

outer domains (D1 and D2) is used to generate boundary conditions for inner domains

123

(D2 and D3). To eliminate the impact of initial conditions, a 3-day model spin-up is used

124

in each month’s simulation. Note that the three-nested domains in CMAQ share the

125

same spatial resolution with WRF but their coverage is slightly smaller. CMAQ is one-way

126

nested and runs with 18 vertical layers from the isobaric surface to 50 hPa.

127

Ground-level meteorological data is used to evaluate the performance of the WRF

128

model, while the observed O3 mixing ratios from the PRD regional air quality monitoring

129

network are used to validate the simulated O3 mixing ratios. As shown in Tables S1 and

130

S2, the modeling system can reproduce ambient meteorological conditions and O3

131

mixing ratios fairly well, therefore is reliable for in-depth analysis of the relationship

132

between O3 and its precursors.

133

2.2 Scenario design for O3-NOx-VOC isopleth

134

Four areas with different pollutant emission characteristics are selected for scenario

135

analysis, as shown in Figure 1. Guangzhou, the most populated city in the PRD and the

136

regional road network hub, was selected to represent urban conditions. Dongguan, a

137

city with hundreds of toy-making, shoe-making and furniture-manufacturing factories

138

and workshops, was selected as an industrial area. Nansha, a small town close to the



139

Pearl River Estuary is an O3 hot spot and contributes significantly to the air quality non-

140

attainment rate. Nansha is also largely impacted by significant emissions from the

141

nearby container terminal and power plants. Jiangmen, a city in the southwestern PRD

142

with scarce industrial activities, a sparse population and limited road network, was

143

selected as a rural area. These locations are illustrated in Figure 1, and more details for

144

area selection are provided in the Supporting Information.

145

In the PRD, maximum hourly O3 concentrations generally occur in August, while

146

October has the highest monthly average O3 concentration. Therefore, these two

147

months are selected for O3 simulation. Within the two months, elevated O3 days,

148

defined as those with a maximum 1-hour O3 mixing ratio over 102 ppbv (200 µg/m3),

149

Stage II of China’s National Ambient Air Quality Standard (NAAQS), are extracted for

150

analysis, as listed in Table S3. The selected O3 episodes are generally associated with

151

subsidence and a stagnant airmass on the outskirts of a tropical cyclone in the

152

Northwestern Pacific, a typical and common meteorological condition leading to O3

153

episodes in the PRD.29

154

An emission reduction matrix including 1 base case and 39 emission reduction

155

scenarios (% change in gram) is designed to develop O3 isopleth diagrams.30 It is noted

156

that both AVOCs and BVOCs are included in simulating the O3 level, but only AVOCs are

157

subject to the reduction schemes given that BVOCs are uncontrollable. Region-wide

158

emission reductions are conducted for the PRD in D3 while the emissions outside PRD

159

remain unchanged. The 39 scenarios with different combinations of NOx and AVOC

160

emission reductions in the PRD are shown in Figure S1. For example, the “A” point


Page 10 of 30

Page 11 of 30


161

represents the scenario with both NOx and AVOC emissions reduced by 10%. Note that

162

emission reduction scenarios are more concentrated around the base case (the upper

163

right corner of the emission matrix), as they represent more feasible emission reduction

164

scenarios. 18 out of 39 scenarios simulated 30% or less emission reductions while others

165

were evenly distributed from 30% to 100% reductions. Based upon the responses of O3

166

mixing ratios to the 39 reduction scenarios, the O3 isopleth diagrams and response of O3

167

to other changes of AVOC and NOx controls are then developed by interpolation.

168

Our method is different from the widely used response surface model (RSM) in that

169

RSM uses statistical correlation structures to approximate model functions31-33. RSM

170

explores the relationships between parameters with less computational cost. However,

171

it may introduce additional uncertainty from statistical representation compared to

172

direct simulation using CMAQ if we assume the input and CMAQ performance are the

173

same in both methods.34 Therefore, multiple CMAQ runs with this scenario analysis are

174

selected to investigate the nonlinear relationship between O3, VOCs and NOx in this

175

study.

176 177

3. RESULTS

178

3.1 Spatial and seasonal variations of ozone formation regime

179

Figure 2 shows the O3 isopleth diagrams for the selected elevated O3 days from the

180

four areas. In an O3 isopleth, the VOC-limited and NOx-limited regimes are separated by

181

a ridge line, which corresponds to the maximum 1-hour O3 mixing ratio for a given VOC

182

emission to produce. Areas close to the ridge line are defined as a transitional regime.



Page 12 of 30

183

The upper right corner of the plot is the starting point (base case) for any emission

184

reduction scenario and is defined by 100% of emissions for both NOx and AVOCs (BVOCs

185

remain constant). For the modeled elevated O3 days in August, the base case scenarios

186

at Guangzhou and Nansha are above the ridge line, indicating that O3 formation is in a

187

VOC-limited regime in urban and port areas of the PRD. The base case in Dongguan is in

188

a transitional regime, which is sensitive to both AVOC and NOx and has no NOx

189

disbenefits.5 There is no ridge line at Jiangmen, as the O3 formation is always NOx-limited

190

for any possible AVOC-NOx emission combinations. For the modeled elevated O3 days in

191

October, the base case scenarios at Guangzhou, Dongguan and Nansha are VOC-limited

192

and further from the ridge line than in the August modeled days. The ridge line appears

193

in the isopleth for Jiangmen and almost passes through the base case scenario, showing

194

O3 formation in a transitional regime. Table 1 summarizes the O3 formation regimes at

195

the four areas on the elevated O3 days in both months.

Table 1 O3 formation regimes at the four areas on elevated O3 days in both months August October Area a Regime Distance to the ridge line Regime Distance to the ridge line a

196

197 198

Guangzhou VOC-limited

-20% NOx

VOC-limited

-50% NOx

Dongguan Transitional

-8% NOx

VOC-limited

-50% NOx

Nansha VOC-limited

-47% NOx

VOC-limited

-49% NOx

Jiangmen NOx-limited

-

Transitional

-4% NOx

a

The distance along the y-axis from the base case to the intersection of the ridge line with the right boundary of the O3 isopleths plot.


Page 13 of 30


199 200 201 202 203

Figure 2. O3 isopleth profiles (ppbv of 1-hr max) for elevated O3 days. The VOC-limited and NOxlimited regimes are separated by the red ridge lines. Guangzhou and Nansha are in a VOC-limited regime in both August and October, Dongguan is in a transitional regime in August and a VOC-limited in October, while Jiangmen is in a NOx-limited regime in August and in a transitional regime in

204

October. The O3 formation regime is generally more VOC-limited in October than August. 13 ACS Paragon Plus Environment


205 206

The spatial differences in O3 formation regime can be explained by the distribution of

207

total VOC (including BVOC)-to-NOx emission ratios, as shown in Figure 3. The ratio is the

208

lowest at Nansha, followed by Guangzhou and Dongguan, and the highest at Jiangmen.

209

The lowest ratio at Nansha, mostly below 0.8, is the result of intensive NOx emissions

210

from the port and the nearby coal-fired power plants. In contrast, most areas in

211

Jiangmen have total VOC-to-NOx emission ratios over 5 or even 20 as a result of

212

significant BVOC emissions in rural areas.

213 214 215 216 217 218

Figure 3. Spatial distribution of the total VOC-to-NOx emission ratios in (a) August and (b) October. Among the four areas of analysis, Jiangmen has the highest ratio, followed by Dongguan and Guangzhou. Nansha shows the lowest ratio due to significant NOx emissions from the nearby port and power plants.

219

The seasonal difference in the O3 formation regime can be partly explained by the

220

seasonal variation of BVOC emissions. BVOC accounts for 40% of total VOC emissions in

221

August and 22% in October, and for 53% of total O3 forming potential in the PRD in


Page 14 of 30

Page 15 of 30


222

August and 32% in October, as a result of higher temperatures and stronger solar

223

radiation in August.20 As BVOC supplies shrink in fall while NOx emissions change only

224

marginally, O3 formation shifts towards VOC-limited conditions, especially in rural areas

225

where BVOC emissions are more significant. Other factors such as a decreased OH

226

production rate in October, due to lower UV radiation and humidity, may also contribute

227

to the transition.7

228

3.2 Response of O3 to different precursor reduction schemes

229

In this section, the response of peak O3 levels to five precursor reduction schemes

230

are examined, including two AVOC control-focused schemes (AVOC control only, and

231

AVOC/NOx reduction ratio=3:1, the ratio is % change in gram / % change in gram), a

232

balanced scheme (AVOC/NOx reduction ratio=1:1), and two NOx control-focused

233

schemes (AVOC/NOx reduction ratio=1:2, NOx control only). Figure 4 provides a

234

conceptual illustration of the reduction paths for all five schemes.

235



236 237 238 239 240

Figure 4. An Illustration overlaid on ozone isopleths of the five precursor reduction schemes in August in Guangzhou. Δ marks the transition from a VOC-limited to NOx-limited regime. Transition occurs for schemes of NOx control only, AVOC/NOx=1:2, and AVOC/NOx=1:1.

241

Figure 5 shows the changes of peak O3 mixing ratios in response to different

242

AVOC/NOx emission reduction schemes. The horizontal axis represents the combined

243

reduction percentage of AVOC and NOx. For example, reduction percentage of 100% in

244

the horizontal axis suggests that the combined reduction percentage of AVOC and NOx is

245

100%, which corresponds to the reduction of AVOC and NOx emissions by 33% and 67%,

246

50% and 50%, and 75% and 25% in the reduction schemes of AVOC/NOx=1:2, 1:1 and

247

3:1, respectively. The maximum reduction percentage of AVOC and NOx individually is

248

assumed to be 90% (180% as combined percentage).


Page 16 of 30

Page 17 of 30


249 250 251 252 253

Figure 5. Response of the O3 mixing ratio to different precursor reduction schemes. Δ marks the transition from the VOC-limited to NOx-limited regime. In select plots, percent O3 increases are given in comparison with their original values, and percent NOx and AVOC reductions are given as that required to reach attainment, which is marked by the red dotted line (1-hr O3=102ppbv). 17 ACS Paragon Plus Environment


254 255

3.2.1 Guangzhou/Dongguan. As shown in Figure 5(a-d), Guangzhou and Dongguan

256

share similar characteristics for the peak O3 mixing ratio response to precursor

257

reductions. In August (Figure 5a-c), peak O3 mixing ratios decrease most rapidly for the

258

‘AVOC control only’ reduction scheme (red line), followed by AVOC/NOx=3:1 (orange

259

line). AVOC-focused controls are therefore more efficient at short-term O3 despiking.

260

However, O3 attainment cannot be reached until 85% of the AVOCs are reduced at

261

Guangzhou, which is not practically feasible considering their diverse and scattered

262

sources. At Dongguan, even if 90% of the AVOCs are reduced in both AVOC-focused

263

control schemes, the peak O3 level still exceeds 102 ppbv, Stage II of NAAQS. Therefore,

264

although the AVOC-focused controls can initially lower the peak O3 level, it is likely that

265

the control schemes would fail to bring the O3 level into attainment.

266

Peak O3 mixing ratios increase slightly (1%) at Guangzhou under the ‘NOx control

267

only’ reduction scheme (cyan line) due to NOx disbenefits. After reducing NOx emissions

268

by 20% (as marked by Δ in Figure 5), the VOC-limited regime transfers to NOx-limited

269

conditions and the O3 mixing ratio decreases sharply in response to NOx reduction. O3

270

attainment at Guangzhou can be reached when NOx emissions are reduced by 73%,

271

which is more practically feasible than reducing 90% of AVOCs. With the transitional

272

regime at Dongguan, O3 mixing ratios do not show any increase under the ‘NOx control

273

only’ scheme, and the conditions become NOx-limited when 8% of NOx emissions are

274

reduced. O3 attainment is achieved when 63% of NOx emissions are reduced.

275

Parallel AVOC and NOx reductions can avoid the slight increase of peak O3 mixing


Page 18 of 30

Page 19 of 30


276

ratios at Guangzhou in August. As shown in Fig 5(a), the O3 mixing ratio steadily

277

decreases in the reduction schemes of AVOC/NOx=1:1 (green line) and 1:2 (blue line),

278

and decreases faster after crossing the inflection points (Δ marks). It is also noted that

279

the reduction schemes of NOx only, AVOC/NOx=1:2 and 1:1 achieve a similar O3 level of

280

about 65 ppbv. In other words, peak O3 mixing ratios at Guangzhou in August are

281

approximately constant when 90% of the NOx emissions are reduced, no matter how

282

much AVOC emissions are reduced in parallel (0%, 45% and 90% AVOC reductions for the

283

NOx only, AVOC/NOx=1:2 and 1:1 schemes, respectively). Therefore, NOx-focused

284

controls have a higher potential of achieving long-term O3 attainment at Guangzhou.

285

Without an O3 increase along its reduction path, NOx-focused controls also favor long-

286

term O3 attainment at Dongguan.

287

In October, however, the NOx disbenefits are more evident when the ‘NOx control

288

only’ reduction scheme is imposed. Peak O3 mixing ratios increase by 23% when 50% of

289

NOx emissions are reduced, and return to the original level only when 78% of NOx

290

emissions are reduced. In comparison, O3 shows a flatter bulge (6% at Guangzhou and

291

8% at Dongguan) along the reduction path of AVOC/NOx=1:2, and reaches the NOx-

292

limited regime by reducing 30% of AVOC and 60% of NOx emissions. O3 attainment can

293

be reached when around 40% of AVOC and 80% of NOx emissions are reduced. Although

294

a steady decrease in the peak O3 level is realized along the reduction path of

295

AVOC/NOx=1:1, the reduction efficiency is too low and O3 attainment is only achieved

296

when more than 70% of AVOC and NOx are reduced, which is not practically feasible.

297

More than 50% of AVOC needs to be reduced in the schemes of ‘AVOC only’ and



298

AVOC/NOx=3:1 which is feasible but challenging. Therefore, AVOC/NOx=1:2 has a higher

299

potential in achieving long-term O3 attainment at both sites considering its practical

300

feasibility and only a slight elevation in the O3 level initially. In contrast, AVOC-focused

301

controls have a higher potential for short-term O3 despiking.

302

3.2.2 Nansha. As discussed previously, Nansha is characterized by strong NOx

303

emissions from the nearby port and coal-fired power plants. The O3 formation regime, as

304

a result, is VOC-limited in both months. Similar as in Guangzhou and Dongguan, the O3

305

level decreases most rapidly with the ‘AVOC control only’ approach, but fails to reach

306

attainment, as shown in Figure 5(e-f). Peak O3 levels increase along the reduction paths

307

of ‘NOx control only’, AVOC/NOx=1:2 and 1:1, with maximum increases in the peak O3

308

mixing ratio of 16%, 11% and 6% in August and 20%, 14% and 6% in October,

309

respectively. Despite the initial O3 increase, these three reduction schemes can lower

310

the O3 mixing ratio to 110 ppbv in August, close to the NAAQS Stage II of 102 ppbv. In

311

these parallel reduction scenarios, NOx plays the key role while simultaneous control of

312

AVOC restrains any increase in O3 along the reduction path. The AVOC/NOx=3:1

313

reduction scenario elevates the peak O3 level relative to the ‘AVOC control only’ scheme.

314

Therefore, AVOC controls should be the focus for short-term O3 despiking while NOx-

315

focused controls are recommended for balancing practical feasibility, moderation of

316

short-term peak O3 increases, and a higher chance long-term to bring the O3 level into

317

attainment.

318

3.2.3 Jiangmen. As shown in Figure 5(g-h), O3 responses to precursor reductions at

319

Jiangmen are the most straightforward. Due to stronger BVOC emissions and weaker


Page 20 of 30

Page 21 of 30


320

influence from human activities, the O3 formation regime is NOx-limited in August and

321

transitional in October. As a result, ‘NOx control only’ is recommended for O3 attainment

322

in both months. Different from the other three areas, ‘NOx control only’ is also the most

323

efficient method in short-term despiking at Jiangmen. Due to the dominant BVOC

324

emissions, control efforts should be primarily directed to NOx emissions.

325

4. DISCUSSION

326

Results show that in the urban and port areas of the PRD, the most efficient way for

327

short-term despiking of the peak O3 level is AVOC control, which is consistent with the

328

VOC-limited O3 formation regimes in these areas. However, the AVOC-focused controls

329

cannot bring the O3 level into attainment in some areas, i.e. the O3 reduction by AVOC

330

control is “short-term efficient but not long-term effective”. NOx-focused controls,

331

instead, have the potential of reaching O3 attainment.

332

A NOx control-focused strategy for long-term attainment is based upon an implicit

333

expectation that, with sufficient NOx reductions, O3 formation would transition from

334

VOC- to a NOx-limited regime throughout the PRD and further AVOC reductions would

335

not be effective at that point. This is illustrated in Figure 5 whereby peak O3 mixing ratios

336

in the three schemes with 90% reduction of NOx (NOx only, AVOC/NOx=1:2 and 1:1) drop

337

almost to the same level. This indicates that the O3 reduction potential is essentially

338

driven by the degree of NOx control and independent of AVOC reductions. Therefore,

339

transition from the current VOC-limited to NOx-limited O3 formation regime is the

340

prerequisite for O3 attainment in the urban, industrial and port areas of the PRD. After

341

transition, the effect of NOx reduction on O3 is enhanced when further NOx emissions



342

have been reduced. The main function of parallel AVOC control is to circumvent or lower

343

O3 elevation in the initial NOx reduction stages.

344

Different approaches for short-term despiking and long-term attainment may have

345

conflicting effects and complicate the O3 control strategy, and inevitably impact the

346

PM2.5 reduction schemes. For example, short-term AVOC reductions lead to greater

347

availability of OH radical in the atmosphere, which reacts with the abundant NO2 to form

348

nitric acid and subsequently ammonia nitrate, a significant addition to PM2.5.35

349

Moreover, AVOC reduction intensifies the limitation of VOC, leading to O3 levels that are

350

more sensitive to the NOx levels in the atmosphere. Peak O3 levels would then increase

351

in response to any reduction of NOx emission, a required step for long-term O3 and PM2.5

352

attainment. This may explain why PM2.5 reduction is often accompanied by O3 elevation

353

in China.

354

In addition, the initial increase of O3 lessens the appeal for adopting NOx-focused

355

controls, especially in fall. Although the magnitude of O3 increase can be alleviated by

356

parallel AVOCs control, it is challenging to eliminate completely due to the poor

357

understanding of the AVOC source variability and distribution. Therefore, in the process

358

of achieving long-term attainment, a balance needs to be reached between the

359

acceptable level of peak O3 and the efforts paid to AVOC reduction. The degree of

360

parallel AVOC reduction in a particular area should be determined based on the degree

361

of NOx reduction, the associated degree of short-term increased peak O3, and the

362

practical feasibility for AVOC control.

363

Although the O3-NOx-VOCs relationship described here for the PRD is sensitive to


Page 22 of 30

Page 23 of 30


364

uncertainties in the emission inventory, physical and chemical parameterization of the

365

air quality modeling, and the limited number of simulation days, a comprehensive view

366

of the findings indicate that pursuing a NOx-limited regime is necessary for O3

367

attainment in the PRD. Greater efforts at NOx control have to be made for expeditious

368

passing through the NOx disbenefits period towards O3 attainment. Given the magnitude

369

of these required emission reductions, it is critical that the PRD continues to progress

370

and work actively towards achieving as many specific emissions reductions as possible,

371

as specified in the JERP (2015-2020).36 As NOx control is also a required step in reducing

372

the ambient PM2.5 level, a NOx-focused control may lead to synergistic reduction of O3

373

and PM2.5. The development of a robust control strategy demands integrated planning to

374

identify, to the extent that is feasible, co-benefit opportunities for achieving multi-

375

pollutant reductions. As such, control measures for attainment of the O3 standard can

376

assist in the attainment of the PM2.5 standard.

377

In the PRD, fossil fuel combustion dominates NOx emissions, with 53% of total NOx

378

emissions derived from coal-fired power plants, industrial boilers and other stationary

379

sources. These sources are typically large- to medium-size enterprises under the

380

surveillance of environmental authorities.21 Their NOx emissions can be reduced to

381

relatively low levels by denitrification technologies. On-road mobile sources account for

382

30% of NOx emissions, which could be reduced by 43% with implementation of Stage V

383

of the Vehicle Emission Standard slated for the upcoming years. Marine vessels

384

contribute 16% of NOx emissions in the PRD, but could be reduced by 80% with fuel

385

improvements and/or setting up emission control areas.37 Therefore, NOx has a



386

significant reduction potential (~80%) in the PRD.

387

It must be pointed out that a NOx-focused control strategy for long-term attainment

388

does not mean AVOC controls are negligible. AVOC emission reductions are essential to

389

short-term despiking and alleviating the O3 increase as a result of NOx controls. In

390

addition, some AVOC species are of higher toxicity and pose a public health risk,

391

therefore should be tackled with health-oriented air quality management. AVOC

392

controls, however, should be conducted in a more targeted manner. According to our

393

results, AVOC emission reductions should be preferentially targeted around the Nansha

394

port area in summertime and extended to the urban and industrial areas in fall.

395

Contingency measures during high O3 pollution periods should allow for further AVOC

396

control. Practical feasibility is another important consideration when formulating AVOC

397

control strategies.

398

In comparison with NOx, AVOC emission controls are much more challenging.

399

Industrial solvent usage and industrial processes contribute 57% of AVOC emissions in

400

the PRD, involving a wide range of industrial sectors and medium- to small-size

401

enterprises.20 Contributions from solvent use may be even higher considering large

402

amounts of fugitive emissions dissipate in open or semi-open workshops and fail to

403

channel through chimneys. On-road mobile sources contribute 28% of AVOCs and these

404

emissions are overlooked by the Stage V of the Vehicle Emission Standard. Therefore, we

405

estimate that at most 40% of AVOC emissions can be effectively reduced, yielding an

406

approximate AVOC-to-NOx reduction ratio of 1:2, which is that adopted by the JERP

407

(2015-2020) for its practical feasibility.36 This study provides support for the AVOC-to-


Page 24 of 30

Page 25 of 30


408

NOx 1:2 reduction ratio and more detailed refinement of the NOx control-focused

409

strategies in the PRD.

410

As shown in this study, the AVOC-to-NOx reduction should be determined based on

411

scientific evidence and be spatially dynamic. It might also be helpful if stage-wise control

412

strategies are adopted, i.e. VOC-focused control strategies (but lower than 3:1) in the

413

initial stage with a transition to NOx-focused control strategies in the later stage. Such

414

stage-wise approaches may reduce the short-term NOx disbenefits while reaching

415

attainment in the long-term. Further investigations on the actual AVOC-to-NOx control

416

ratios in different stages and the AVOC-control to NOx-control transitioning point are

417

needed to reveal an optimal control strategy.

418

Ambient O3 in the PRD is not only formed locally, but also transported from outside

419

the PRD. Studies have shown that super-regional contribution to O3 in the PRD is

420

important in summer and fall and is an increasing trend.18,38 In this study, a separate

421

sensitivity test was conducted by removing all NOx and anthropogenic and biogenic

422

VOCs emissions within the PRD to quantify O3 arriving from outside, as detailed in the

423

Supporting Information. Using this method indicates O3 contributions from outside

424

range from 28-54% in August and 26-46% in October for the four areas. This regional

425

input of O3 helps keep background O3 high, enhancing the challenge of O3 control in the

426

PRD, and requiring a significant reduction of local precursors to bring O3 levels into

427

attainment. Therefore, co-control with other parts of the country and even other

428

countries in East Asia are required to reduce the background level and lower the barrier

429

for O3 attainment in the PRD.



430 431

AUTHOR INFORMATION

432

Corresponding Author

433

*Z.Y. Phone: 86-20-39380021; fax: 86-20-39380021; e-mail: [email protected]

434

*J.Z. Phone: 86-20-39380021; fax: 86-20-39380021; e-mail: [email protected]

435

Present Address

436 437

School of Environment and Energy, South China University of Technology, Guangzhou 510006, China

438 439

ACKNOWLEDGMENTS

440

This work was supported by Environmental Protection Department, Government of the

441

Hong Kong Special Administrative Region under the project “Characterisation of VOC

442

Sources and Integrated Photochemical Ozone Analysis in Hong Kong and the Pearl River

443

Delta region (Tender Ref. 12-02909)”, National Distinguished Young Scholar Science Fund

444

of the National Natural Science Foundation of China (No. 41325020), National Key

445

Technology Research and Development Program of the Ministry of Science and

446

Technology of China (No. 2014BAC21B03), and Fundamental Research Funds for the

447

Central Universities of China (No. D2154370 and D2156980). We thank Dr. Stephen

448

Griffith of the Hong Kong University of Science and Technology for editing this

449

manuscript.

450 451

SUPPORTING INFORMATION AVAILABLE

452

Four figures, four tables and additional information on (1) emission reduction matrix of

453

NOx and VOC for O3 sensitivity study; (2) detailed description of area selection for O3 26 ACS Paragon Plus Environment

Page 26 of 30

Page 27 of 30


454

isopleth development; (3) distribution of urban area, industrial factories, and road

455

network in the PRD; (4) observed and simulated O3 time series in August and October

456

2010; (5) WRF and CMAQ performance statistics in August and October 2010; (6)

457

selected O3 elevated days for different areas in August and October 2010; and (7)

458

regional contribution of O3 from outside the PRD. This information is available free of

459

charge via the Internet at http://pubs.acs.org.

460 461

DISCLAIMER

462

The opinions expressed in this paper are those of the authors and do not necessarily

463

reflect the views or policies of the Government of the Hong Kong Special Administrative

464

Region, nor does mention of trade names or commercial products constitute an

465

endorsement or recommendation of their use.



References 1.

Monthly/Quarterly Report of Air Quality of 74 Cities; China National Environmental Monitoring Centre, 2016; http://www.cnemc.cn/publish/106/0536/newList_1.html.

2.

Pearl River Delta Regional Air Quality Monitoring Network – A Report of Monitoring Results in 2014; Guangdong Provincial Environmental Monitoring Centre, Environmental Protection Department, Hong Kong SARG, Environmental Protection Bureau and Meteorological and Geophysical Bureau, Macao SARG, 2015; http://www.epd.gov.hk/epd/sites/default/files/epd/english/resources_pub/publications/files/PRD_ 2014_report_en.pdf.

3.

Cleveland, W.S.; Graedel, T.E. Photochemical air pollution in the northeast United States. Science 1979, 204, 1273–1278.

4.

Chameides, W.L.; Lindsay, R.W.; Richardson, J.; Kiang, C.S. The role of biogenic hydrocarbons in urban photochemical smog: Atlanta as a case study. Science 1988, 241, 1473–1475.

5.

Sillman, S. The relation between ozone, NOx and hydrocarbons in urban and polluted rural environments. Atmos. Environ. 1999, 33, 1821-1845.

6.

Finlayson-Pitts, B.J.; Pitts, J.N.. Atmospheric chemistry of tropospheric ozone formation: Scientific and regulatory implications. J. Air & Waste Manage. Assoc. 1993, 43, 1091–1100.

7.

Jacob, D.J.; Horowitz, L.W.; Munger, W.; Heikes, B.G.; Dickerson, R.R.; Artz, R.S.; Keene, W.C. Seasonal transition from NOx- to hydrocarbon-limited conditions for ozone production over the eastern United States in September. J. Geophys. Res. 1995, 100, 9315–9324.

8.

Marr, L.C.; Harley, R.A. Spectral analysis of weekday-weekend differences in ambient ozone, nitrogen oxide, and non-methane hydrocarbon time series in California. Atmos. Environ. 2002, 36, 2327–2335.

9.

Murphy, J.G.; Day, D.A.; Cleary, P.A.; Wooldridge, P.J.; Millet, D.B.; Goldstein, A.H.; Cohen, R.C.. The weekend effect within and downwind of Sacramento: Part 1. Observations of ozone, nitrogen oxides, and VOC reactivity. Atmos. Chem. Phys. 2007, 7, 5327-5339.

10. Fujita, E.M.; Campbell, D.E.; Stockwell, W.R.; Lawson, D.R. Past and Future Ozone Trends in California’s South Coast Air Basin: Reconciliation of Ambient Measurements with Past and Projected Emission Inventories. J. Air & Waste Manage. Assoc. 2013, 63, 54-69. 11. Final 2012 Air Quality Management Plan; South Coast Air Quality Management District, 2013. 12. Projected Ozone Trends and Changes in the Ozone-precursor Relationship in the South Coast Ari Basin in Responses to Varying Reductions of Precursor Emissions; Final Report [CRC report No. A-91]; Coordinating Research Council, Inc (CRC), 2015. 13. Zhang, Y.H.; Hu, M.; Zhong, L.J.; Wiedensohler, A.; Liu, S.C.; Andreae, M.O.; Wang, W.; Fan, S.J. Regional Integrated Experiments on Air Quality over Pearl River Delta 2004 (PRIDEPRD2004): Overview. Atmos. Environ. 2008, 42, 6157–6173. 14. Zhang, Y.H.; Su, H.; Zhong, L.J.; Cheng, Y.F.; Zeng, L.M.; Wang, X.S.; Xiang, Y.R.; Wang, J.L.; Gao, D.F.; Shao, M. et al. Regional ozone pollution and observation-based approach for analyzing ozone–precursor relationship during the PRIDE-PRD2004 campaign. Atmos. Environ. 2008, 42, 6203–6218. 15. Shao, M.; Zhang, Y.H.; Zeng, L.M.; Tang, X.Y.; Zhang, J.; Zhong, L.J.; Wang, B. Ground-level ozone in the Pearl River Delta and the roles of VOC and NO(x) in its production. J. Environ. 28 ACS Paragon Plus Environment

Page 28 of 30

Page 29 of 30


Manage. 2009, 90, 512–518. 16. Wang, X.S.; Zhang, Y.H.; Hu, Y.T.; Zhou, W.; Zeng, L.M.; Hu, M.; Cohan, D.S.; Russell, A.G. Decoupled direct sensitivity analysis of regional ozone pollution over the Pearl River Delta during the PRIDE-PRD2004 campaign. Atmos. Environ. 2011, 45, 4941–4949. 17. Li, Y.; Lau, A.K.H.; Fung, J.C.H.; Zheng, J.; Liu, S. Importance of NOx control for peak ozone reduction in the Pearl River Delta region. J. Geophys. Res. 2013, 118, 9428–9443. 18. Xue, L.K.; Wang, T.; Louie, P.K.K.; Luk, C.W.Y.; Blake, D.R.; Xu, Z. Increasing external effects negate local efforts to control ozone air pollution : A case study of Hong Kong and implications for other Chinese cities. Environ. Sci. Technol., 2014, 48, 10769-10775. 19. Pearl River Delta Regional Air Quality Management Plan Mid-term Review Report; Hong KongGuangdong Joint Working Group on Sustainable Development and Environmental Protection, 2008. 20. Ou, J.M.; Zheng, J.Y.; Li, R.R.; Huang, X.B.; Zhong, Z.M.; Zhong, L.J.; Lin, H. Speciated OVOC and VOC emission inventories and their implications for reactivity-based ozone control strategy in the Pearl River Delta region, China. Sci. Total Environ. 2015, 530-531, 393–402. 21. Gridded Emission Inventories to Support the Upgraded PATH Modeling System; Final Report; Chinese University of Hong Kong, 2015. 22. Wang, S.S.; Zheng, J.Y.; Fu, F.; Yin, S.S.; Zhong, L.J. Development of an emission processing system for the Pearl River Delta Regional air quality modeling using the SMOKE model: Methodology and evaluation. Atmos. Environ. 2011, 45, 5079–5089. 23. Guenther, A.; Karl, T.; Harley, P.; Wiedinmyer, C.; Palmer, P. I.; Geron, C. Estimates of global terrestrial isoprene emissions using MEGAN (Model of Emissions of Gases and Aerosols from Nature). Atmos. Chem. Phys. 2006, 6, 3181-3210. 24. Shen, X.L.; Yin, S.S.; Zheng, J.Y.; Lu, Q.; Zhong, L.J. Anthropogenic ammonia emission inventory and its mitigation potential in Guangdong Province. Acta Scientiae Circumstantiae 2014, 34, 43-53. 25. Ye. S.Q.; Zheng, J.Y.; Pan, Y.Y.; Wang, S.S.; Lu, Q. Marine emission inventory and its temporal and spatial characteristics in Guangdong Province. Acta Scientiae Circumstantiae, 2014, 34, 537547. 26. Foley, K.M; Roselle, S.J.; Appel, K.W.;, Bhave, P.V.; Pleim, J.E.; Otte, T.L.;, Mathru, R.; Sarwar, G.; Young, J.O.; Gilliam, R.C. et al. Incremental testing of the Community Multiscale Ari Quality (CMAQ) modeling system version 4.7. Geosci. Model Dev. 2010, 3, 205-226. 27. Nenes, A.; Pandis, S.N.;, Pilinis, C. ISORROPIA: A new thermodynamic equilibrium model for multiphase multicomponent inorganic aerosols. Aquat. Geochem. 1998, 4, 123-152. 28. Pleim, J.E. A combined local and nonlocal closure model for the atmospheric boundary layer. Part II: Application and evaluation in a mesoscale meteorological model. J. Appl. Meteor. Climatol. 2007, 46, 1396-1409. 29. Huang, J.P.; Fung, J.C.H.; Lau, A.K.H.; Yu, Q. Numerical simulation and process analysis of typhoon-related ozone episodes in Hong Kong. J. Geophys. Res. 2005, 110, D05301, doi: 10.1029/2004JD004914. 30. Kinosian, J.R. Ozone-Precursor Relationships from EKMA Diagrams. Environ. Sci. Technol. 1982, 16, 880–883. 31. Fann, N.; Fulcher, C.M.; Hubbell, B.J. The influence of location, source, and emission type in 29 ACS Paragon Plus Environment


estimates of the human health benefits of reducing a ton of air pollution. Air Qual. Atmos. Health. 2009, 2, 169-176. 32. Foley, K.M.; Napelenok, S.L.; Jang, C.; Phillips, S.; Hubbell, B.J.; Fulcher, C.M. Two reduced form air quality modelikng techniques for rapidly calculating pollutant mitigation potential across many sources, locations and precursor emission types. Atmos. Environ. 2014, 98, 283-289. 33. Xing, J.; Wang, S.X.; Jang, C.; Zhu, Y.; Hao, J.M. Nonlinear response of ozone to precursor emission changes in China: a modeling study using response surface methodology. Atmos. Chem. Phys. 2011, 11, 5027-5044. 34. Technical Support Document for the Proposed PM NAAQS Rule-Response Surface Modeling. U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards, Research Triangle Park, NC 27711. February 2006. 35. Atkinson, R. Atmospheric chemistry of VOCs and NOx. Atmos. Environ. 2000, 34, 2063-2101. 36. Air Pollution Emisison Reduction Plan up to 2020. Advisory Council on the Environment, Hong Kong. December, 2012. 37. Ng, S.K.W.; Loh, C.; Lin, C.B.; Booth, V.; Chan, J.W.M.; Yip, A.C.K.; Li, Y.; Lau, A.K.H. Policy change driven by an AIS-assisted marine emission inventory in Hong Kong and the Pearl River Delta. Atmos. Environ. 2013, 76, 102-112. 38. Li, Y.; Fung, J.C.H.; Zheng, J.Y.; Zhong, L.J.; Louie, P.K.K. Ozone source apportionment (OSAT) to differentiate local regional and super-regional source contributions in the Pearl River Delta region, China. J. Geophys. Res. 2012, 117, D15305, doi:10.1029/2011JD017340.


Page 30 of 30

Ambient Ozone Control in a Photochemically ... - ACS Publications

Recommend Documents