Observations of Atmospheric Δ14CO2 at the Global and Regional

Oct 25, 2016 - Six months to more than one year of atmospheric Δ14CO2 were measured in 2014–2015 at one global background site in Waliguan (WLG) an...
0 downloads 4 Views 590KB Size
Subscriber access provided by CORNELL UNIVERSITY LIBRARY

Article

Observations of Atmospheric #14CO2 at the Global and Regional Background Sites in China: Implication for Fossil Fuel CO2 Inputs Zhenchuan Niu, Weijian Zhou, Peng Cheng, Shugang Wu, Xuefeng Lu, Xiaohu Xiong, Hua Du, and Yunchong Fu Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.6b02814 • Publication Date (Web): 25 Oct 2016 Downloaded from http://pubs.acs.org on October 25, 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 29

Environmental Science & Technology

1

Observations of Atmospheric ∆14CO2 at the Global and Regional Background

2

Sites in China: Implication for Fossil Fuel CO2 Inputs

3 4

Zhenchuan Niu†, ‡, Weijian Zhou*, †, ‡, §, Peng Cheng†, ‡, Shugang Wu†, ‡, Xuefeng

5

Lu†,‡, Xiaohu Xiong†, ‡, Hua Du†, ‡, Yunchong Fu†, ‡

6 7



8

Environment, Chinese Academy of Sciences, Xi’an, China

9



State Key Laboratory of Loess and Quaternary Geology, Institute of Earth

Shaanxi Provincial Key Laboratory of Accelerator Mass Spectrometry Technology

10

and Application, Xi’an AMS Center, Xi’an, China

11

§

Beijing Normal University, Beijing, China

12 13

*Phone: +86-29-62336203; fax: +86-29-62336234; e-mail: [email protected]

1

ACS Paragon Plus Environment

Environmental Science & Technology

14

ABSTRACT: Six months to more than one year of atmospheric ∆14CO2 were

15

measured in 2014–2015 at one global background site in Waliguan (WLG) and four

16

regional background sites at Shangdianzi (SDZ), Lin’an (LAN), Longfengshan (LFS)

17

and Luhuitou (LHT), China. The objectives of the study are to document the ∆14CO2

18

background levels at each site and to trace the variations in fossil fuel CO2 (CO2ff)

19

inputs at regional background sites. ∆14CO2 at WLG varied from 7.1±2.9‰ to

20

32.0±3.2‰ (average 17.1±6.8‰) in 2015, with high values generally in

21

autumn/summer and low values in winter/spring. During the same period, ∆14CO2

22

values at the regional background sites were found to be significantly (p < 0.05) lower

23

than those at WLG, indicating different levels of CO2ff inputs at those sites. CO2ff

24

concentrations at LAN (12.7±9.6 ppm) and SDZ (11.5±8.2 ppm) were significantly (p

25

< 0.05) higher than those at LHT (4.6±4.3 ppm) in 2015. There were no significant (p >

26

0.05) seasonal differences in CO2ff concentrations for the regional sites. Regional

27

sources contributed in part to the CO2ff inputs at LAN and SDZ, while local sources

28

dominated the trend observed at LHT. These data provide a preliminary understanding

29

of atmospheric ∆14CO2 and CO2ff inputs for a range of Chinese background sites.

30

2

ACS Paragon Plus Environment

Page 2 of 29

Environmental Science & Technology

2015 Regional background sites, China

Shangdianzi

Luhuitou

Lin'an

TOC Art

33 34

3

ACS Paragon Plus Environment

DEC

NOV

OCT

SEP

AUG

JUL

JUN

MAY

APR

FEB

32

30 25 20 15 10 5 0

Waliguan global background site, China

MAR

31

40 35 30 25 20 15 10 5 0

JAN

CO2ff (ppm)

∆14CO2(‰)

Page 3 of 29

Environmental Science & Technology

35

Page 4 of 29

INTRODUCTION

36

Radiocarbon (14C) has a radioactive half-life of 5730 years,1 and it is naturally

37

produced in the atmosphere by the cosmic-ray neutron interactions with nitrogen

38

nuclei:

39

distributed around the globe. The levels of 14C in CO2 are reported as ∆14C, that is, the

40

per mil (‰) deviation from the absolute radiocarbon reference standard corrected for

41

fractionation and decay.2

14

N(n, p)14C. Once produced,

14

C is rapidly oxidized to

൫ భర஼ ൗ భమ஼ ൯

Δ ଵସ‫ = ܥ‬൤൫ భర஼ൗ భమ஼ ൯ ೄಿ − 1൨ × 1000‰

42

ಲಳೄ

43

14

In this equation, (14C/12C)SN is the

CO2 and is rapidly

(1)

14

C/12C ratio of the sample normalized to a

44

conventional δ13C value of –25‰, and (14C/12C)ABS is the absolute radiocarbon

45

reference standard.

46

The level of ∆14CO2 in the atmosphere has been disturbed by the ongoing input of

47

fossil fuel CO2 (CO2ff) at least since 1890, and by a series of atmospheric nuclear

48

weapons tests in the 1950–60s. To study the disturbance, atmospheric ∆14CO2 has

49

been measured at some background sites.3–8 The long-term measurements at these

50

sites indicate that atmospheric ∆14CO2 values were depressed to about –25‰ at the

51

beginning of 1950 due to CO2ff emissions, and subsequently increased drastically due

52

to the addition of 14C produced by above-ground nuclear weapons tests in the 1950–

53

60s. After the cease of nuclear weapons tests, atmospheric ∆14CO2 decreased rapidly

54

between 1963 and 1990, mainly driven by exchanges between the atmosphere, and

55

biosphere and oceans. From 1990 onward, atmospheric ∆14CO2 has decreased slowly

56

mainly due to the CO2ff emissions.3,5

57

The recent decrease of ∆14CO2 resulted principally from the CO2ff emissions, and 1

58

ppm CO2ff emitted to the atmospheric CO2 level of 380 ppm will result in a decrease

59

of ~2.8‰ for ∆14C.9 Thus, the measurement of ∆14CO2 in the atmosphere can provide 4

ACS Paragon Plus Environment

Page 5 of 29

Environmental Science & Technology

60

a quantitative record of atmospheric CO2ff concentration, which is important for the

61

understanding of the increase of atmospheric CO2 concentration, and for the

62

formulation of CO2ff reduction strategies to mitigate this increase.

63

With the rapid economic growth in recent decades, several economic regions have

64

been established in China, and atmospheric CO2 concentrations at some regional

65

background sites were reported to be influenced by fossil fuel emissions in adjacent

66

economic regions.10–12 So, what are the levels of atmospheric CO2ff inputs at those

67

regional background sites in China? To answer this question it is first necessary to

68

ascertain background atmospheric ∆14CO2 levels across China. Currently the available

69

data are limited to some recent measurements of ∆14C in plant material and

70

atmospheric ∆14CO2 in several cities.13–16 This sparse background ∆14CO2 dataset

71

hinders our broader understanding of atmospheric CO2ff values in China. In addition,,

72

background ∆14CO2 values are also important for carbonaceous aerosol apportionment

73

studies using the

74

atmospheric ∆14CO2 measurements were carried out at one global background site and

75

four regional background sites in China from 2014–2015. The objectives of the work

76

include: (1) to clarify the levels and temporal variations in atmospheric ∆14CO2 at

77

global and regional background sites; (2) to trace the variations in CO2ff inputs at

78

regional background sites; (3) to determine the influences of local and regional

79

emissions on these variations.

14

12

C method.17–20 Thus, six months to more than one year of

80 81

MATERIALS AND METHODS

82

Site Description. Waliguan (WLG) Global Atmosphere Watch (GAW) station (36.28°

83

N, 100.9° E, 3816 m a.s.l.) is located in the northeast part of Qinghai-Tibet Plateau in

84

Qinghai Province, western China (Figure 1). This site is about 100 km southwest of 5

ACS Paragon Plus Environment

Environmental Science & Technology

85

Xining, capital of Qinghai Province, far from industrial and populated centers. There

86

are no inhabitants within 10 km of this station. This region has a continental plateau

87

climate, covered by arid and semiarid desert meadow. This station provides

88

background atmospheric data for the Eurasian continent.21

89

Shangdianzi (SDZ) regional background station (40.65° N, 117.12° E, 287 m a.s.l.)

90

is situated on a mountainside about 100 km northeast of Beijing, North China (Figure

91

1). This mountainous area has a semi-humid continental monsoon climate, covered by

92

woodlands and crops. A small village is located to the south (about 0.8 km) of this site,

93

and a railway runs from the southwest to northwest direction (about 0.6 km). The

94

observation at this site was used to delegate the atmospheric background information

95

of the Beijing-Tianjin-Hebei Economic Region.10,11

96

Lin’an (LAN) regional background station (30.30° N, 119.73° E, 139 m a.s.l.) is

97

located on the top of a small hill about 6 km northeast of Lin’an county, Zhejiang

98

Province, East China. This site is about 40 km from the center of Hangzhou, capital of

99

Zhejiang Province, and 190 km from the center of Shanghai, the largest economic

100

center in China (Figure 1). This mountainous area is covered by woodlands and rice

101

paddies, with a subtropical monsoon climate. It represents the atmospheric

102

background information of Yangtze River Delta Economic Region. 10,11

103

Longfengshan (LFS) regional background station (44.73° N, 127.60° E, 331 m

104

a.s.l.) is located on the top of a hill about 40 km southeast of Wuchang county, 140

105

km the southeast of Harbin, capital of Heilongjiang Province, Northeast China (Figure

106

1). The area has a temperate continental monsoon climate, covered by woodlands and

107

rice paddies. There is a reservoir with an area of about 20 km2 on the northeast side of

108

this station and small villages within several kilometers of it. The station represents

109

the atmospheric background information of the Northeast Old Industrial Bases. 10,11 6

ACS Paragon Plus Environment

Page 6 of 29

Page 7 of 29

Environmental Science & Technology

110

Luhuitou (LHT) (18.22° N, 109.48° E, 10 m a.s.l.) is a coastal station located at the

111

southwest tip of the Luhuitou peninsula in Sanya, Hainan Province, South China

112

(Figure 1). This station is on the southernmost perimeter of Hainan Island, about 200

113

km from the capital city of Haikou, more than 600 km southwest of the Pearl River

114

Delta Economic Region. This area has a tropical oceanic monsoon climate, with

115

luxuriant tropical vegetation.

116 117

Sample Collection. The air samplings were carried out from September 2014 to

118

December 2015 at SDZ and LAN, from September 2014 to February 2015 at LFS,

119

and from January 2015 to December 2015 at LHT and WLG. The air samplings at

120

each site were generally arranged at about 10:00 AM (local time) on the 10th and 25th

121

of each month, and sometimes the samplings were postponed or advanced about one

122

or two days when rainy or snowy days were encountered. Additionally, to study the

123

CO2ff variations at different times of a day at the regional sites, air samples at SDZ,

124

LFS, LHT and LAN were collected at about 10:00, 18:00 and 22:00 on two

125

continuous days during both the summer and winter.

126

Ambient air was collected in two 10-L aluminum foil sampling bags (Delin Gas

127

Packing Co., Ltd, Dalian, China) at WLG, using pumps for approximately 10 min,

128

while 5-L sampling bags were used at the regional sites. Meteorological parameters

129

were recorded during sampling. The bag sampling method has been proven in former

130

studies,16,22 and has little influence on CO2ff calculations.16 Before samples were

131

collected, the bags were flushed out with ambient air three times. The operators held

132

their breath when turning on and off the switch and maintained a distance from the

133

apparatus during the collection. After the collection, the bags were sent to the

134

laboratory immediately. A total of 144 air samples were obtained. The bagged air 7

ACS Paragon Plus Environment

Environmental Science & Technology

135

samples were first measured for CO2 concentration, and then for 14C analysis.

136 137

CO2 Concentration Measurement. The CO2 concentrations of air samples were

138

measured by a Picarro G2131-I CO2 Isotopic Analyzer (Picarro Inc.). This type of

139

equipment employs a cavity ring-down spectroscopy (CRDS) technology, with high

140

linearity, precision and stability for CO2 measurement. A cavity ring-down

141

spectrometer is made up of a laser, a high finesse optical cavity consisting of two or

142

more mirrors, and a photodetector. The “ring-down” measurement is made by rapidly

143

turning off the laser and measuring the time constant of the light intensity as it

144

exponentially decays.23,24

145

Briefly, the ambient air in the bag was filtered, dried in an ethanol-liquid nitrogen

146

cold trap (–90 ºC), and then introduced into a high-finesse optical cavity. The optical

147

absorbance of the sample, a function of CO2 concentration, was determined by the

148

light dissipation rate in the optical cavity. Each sample was measured for 6 minutes.

149

Because of the dead volumes when switching to a new sample, only the data in the

150

last 4 minutes was averaged for a sample. The data of 12 CO2_dry and 13CO2_dry were

151

summed to get the total CO2 concentration of an air sample.

152

The instrument was calibrated by a standard gas (395.49±0.02 ppm) obtained from

153

the Chinese Academy of Meteorological Sciences. This standard gas is pressurized in

154

a 29.5 L treated aluminum alloy cylinder (Scott-Marrin Inc., California) fitted with

155

high-purity, two-stage gas regulator, and calibrated with cylinders assigned by the

156

WMO/GAW CO2 Central Calibration Laboratory operated by NOAA/ESRL. The

157

precision of CO2 measurements in this study was below 0.1 ppm.

158 159

Purification, Graphitization, and

14

C Measurement. In order to get pure CO2 8

ACS Paragon Plus Environment

Page 8 of 29

Page 9 of 29

Environmental Science & Technology

160

samples, the air in the bag was first passed through a liquid nitrogen trap (–196 ºC) in

161

a vacuum system at a flow rate of about 200 mL min–1 to trap CO2 and water, and then

162

the trapped water was removed with an ethanol-liquid nitrogen trap (–90 ºC).8

163

A zinc-iron method was used for the graphitization of CO2, with zinc particles and

164

iron powder as reductant and catalyst, respectively.25,26 Then the obtained graphite

165

(1.0–1.2 mg) from ambient air samples was pressed into aluminum target holders for

166

14

167

produced from standards and anthracite coal blanks were processed using the same

168

procedure as the ambient air samples. A vacuum system blank (–998.4±0.1‰) was

169

obtained.

C measurement. Additionally, after combustion with excess CuO powder, the gases

170

An aliquot of standard air with ∆14C value of 6.7±2.3‰ obtained from the Chinese

171

Academy of Meteorological Sciences was periodically (about two months) used to

172

assess the uncertainty for chemical processing, with the same treatment processing as

173

the bag air samples, and an average uncertainty of 2.3‰ was obtained. This value will

174

result in an uncertainty of about 1.3 ppm in the CO2ff calculations. The

175

14

C levels in the air samples were measured using a 3 MV accelerator mass

176

spectrometer (AMS) in Xi’an, China. Each batch contains forty-eight targets,

177

including thirty-eight air samples, six OX-II samples as primary standards, two

178

Chinese sugar carbon (CSC) samples as secondary standards, and two anthracite coal

179

samples as blanks. They were arranged in order into a sample-holding wheel, and then

180

placed into the AMS ion source for 14C measurement. Each sample recorded 300 000–

181

400 000

182

fractionation corrections. The precision of a typical 14C measurement was 3‰. 27 The

183

14

14

C counts, and on-line δ13C measurements were used for isotopic

C levels in the air samples are expressed as ∆14C values at each site in Table S1.

184

9

ACS Paragon Plus Environment

Environmental Science & Technology

Page 10 of 29

185

Calculation of CO2ff. To quantify the inputs of CO2ff to atmospheric CO2 at the

186

regional sites, CO2ff concentrations at SDZ, LAN, LHT and LFS were calculated

187

according to the equations below. CO2 in the air sample (CO2a) is thought to be a

188

mixture of background CO2 (CO2bg), CO2ff and other CO2 (CO2other), and the ∆14C

189

values for CO2a, CO2bg, CO2other and CO2ff are expressed as ∆a, ∆bg, ∆other and ∆ff (–

190

1000‰), respectively. Two following equations were obtained according to the mass

191

balance of CO2 and 14C. 28 CO2a = CO2bg + CO2other + CO2ff

192

CO2a ∆a = CO2bg ∆bg + CO2other ∆other + CO2ff ∆ff

193

(3)

From eq 2 and eq 3, CO2ff can be calculated with the following equation:

194

CO 2 ff =

195

196

(2)

CO 2 a ( ∆bg − ∆a ) CO 2 other ( ∆other − ∆bg ) + ∆bg − ∆ff ∆bg − ∆ff

(4)

The second term on the right-hand-side of eq 4 is a small bias (β) from other small 14

197

sources of

C mainly from the heterotrophic respiration and nuclear industry. CO2ff

198

bias from the heterotrophic respiration will be underestimated by 0.2~0.3 ppm during

199

the winter and 0.4~0.8 ppm during the summer.9, 28, 29 CO2ff bias resulted from nuclear

200

power plants (NPP) is more than –0.25 ppm over large regions, and up to several ppm

201

near to nuclear sites.30,31 A small correction (–0.25 ~ –0.5 ppm) was used for CO2ff

202

calculation at LAN regional site, about 120 km to the west of Qinshan NPP, for the

203

following reasons.

204

Reactors (PWRs) in Qinshan NPP with a low 14CO2 emission factor, and Graven and

205

Gruber (2011) showed that the CO2ff bias around the LAN site was low.30 Additionally,

206

previous study using moss and pine needles around the Qinshan NPP showed that the

207

14

208

background value (223.8 Bq/kg C) at a distance of 6.5 km.32

14

C is mainly released as

14

CH4 from the Pressurized Water

C specific activity decreases with increasing distance from the NPP, and reaches a

209 10

ACS Paragon Plus Environment

Page 11 of 29

Environmental Science & Technology

210

Back-trajectory Analysis. A Hysplit Trajectory Model33 was used to study the

211

influence of air mass transport on ∆14CO2/CO2ff temporal variations at each sampling

212

site. A Global Data Assimilation System (GDAS, 2006–present) meteorological

213

dataset with a spatial resolution of 1° × 1° and temporal resolution of 3 hours, was

214

used in the model, with a total run time of 72 h. The heights of 100 m AGL and 1000

215

m AGL were chosen as the typical heights of the atmospheric surface layer and the

216

planetary boundary layer (PBL), respectively, and the height of 500 m AGL was

217

chosen to provide more information. The start time (UTC) in this model was 8 h later

218

than the local time.

219 220

Data Analysis. Variance analysis of ∆14CO2 or CO2ff concentrations were performed

221

by ANOVA/Duncan’s test using a SPSS statistical software (V. 17),34 and values of p

222

< 0.05 were considered to be statistically significant.

223 224

RESULTS AND DISCUSSION

225

∆14CO2 variations at WLG global background site. Figure 2A shows ∆14CO2

226

variations at the WLG global background site in 2015, and two abnormal values (–

227

0.3±2.8‰ and –61.5±3.2‰) were removed as obvious non-background samples.

228

∆14CO2 at this site varied from 7.1±2.9‰ to 32.0±3.2‰, with an average of

229

17.1±6.8‰.

230

Generally, high values were observed in autumn (Sep.–Nov.) / summer (Jun.–Aug.)

231

months and low values were observed in winter (Jan., Feb. and Dec.) / spring (Mar.–

232

May) months. These seasonal features have also been observed at other global

233

background sites.3,4,8,35 Back-trajectory analyses indicate that the air masses during

234

the winter and spring sampling periods mainly pass over the northern part of China 11

ACS Paragon Plus Environment

Environmental Science & Technology

235

(Figure S1). Thus, fossil fuel consumption for heating in the northern part of China

236

and even in the range of middle to high latitudes in the Northern Hemisphere36 might

237

be related to the relatively low values in winter and spring. Additionally, the seasonal

238

changes in atmospheric vertical mixing height also contributed to the relatively low

239

values in winter and spring.37 As a summer resort in China, many people travel to

240

Qinghai Province by car, and their fossil fuel emissions could potentially reduce the

241

atmospheric ∆14CO2 values recorded at WLG in the summer. From the wind-rose

242

distribution patterns of ∆14CO2 along 16 directions (Figure S2), it can be seen that

243

high values generally occur along the SW-NNW sector and low values along the

244

N-SSE sector. This distribution can be explained as follows: Xining City, capital of

245

Qinghai Province, lies to the east of this site, as well as the more developed and

246

densely populated region with relatively strong CO2 emissions from fuel combustion,

247

while the region to the west of this site is sparsely populated with weak CO2

248

emissions from fuel combustion.38

249

The calculation of CO2ff concentrations is influenced by the choice of background.

250

The high altitude mountain site is an approximation of the background. Because of the

251

influence of local or regional fossil fuel sources on mountain sites, there is about a 2‰

252

difference in ∆14CO2 between the free tropospheric background and high-altitude

253

mountain background, 28 and this difference might bring out a bias of about 0.8 ppm

254

in CO2ff calculations. Although previous measurements at WLG around 2005 showed

255

that the values of ∆14CO2 at this site were similar to those at Ulaan Uul, Mongolia

256

(UUM, 44.45° N, 111.10° E, 914 m a.s.l.) and Niwot Ridge, Colorado (NWR, 40.05°

257

N, 105.58° W, 3526 m a.s.l.),12 ∆14CO2 data at WLG for this study still require

258

filtering according to wind directions and CO values to remove non-background

259

influences. The wind direction of NE–ENE–E in all seasons is regarded as the major 12

ACS Paragon Plus Environment

Page 12 of 29

Page 13 of 29

Environmental Science & Technology

260

non-background section of atmospheric CO2 at WLG,10,36 and our data generally did

261

not occur in that direction. Additionally, CO values at this site39 were used to filter

262

obvious non-background data. After this filtering, seasonal averages were used as the

263

background for CO2ff calculations at the regional sites.

264 265

∆14CO2 variations at the regional background sites. As shown in Figure 2, ∆14CO2

266

values vary highly during the sampling periods at the four regional background sites.

267

∆14CO2 values were in the ranges of –53.0±3.4‰~32.6±3.0‰ (average –6.8±21.1‰)

268

at SDZ, –8.3±3.0‰~24.1±2.8‰ (average 7.2±11.9‰) at LFS, –4.6±2.9‰~31.8±2.8‰

269

(average 10.5±10.3‰) at LHT, and –66.1±3.0‰~27.4±2.8‰ (average –12.1±24.2‰)

270

at LAN. Previous observations at SDZ in around 2010 also showed high variations

271

(about –48~50‰) in ∆14CO2.12

272

Additionally, one abnormal value (173.5±3.3‰) was observed on November 26th,

273

2014 at LAN, about eight times the average background value. It is difficult to deem

274

this single abnormal value to be related to the Qinshan NPP, based on the following

275

several reasons: (1) the 14CO2 emission factor is low for Qinshan NPP,30 because 14C

276

released from PWRs is mainly in the form of 14CH4. (2) 14C specific activity reaches a

277

background level at a distance of 6.5 km to Qinshan NPP.32 (3) That sampling day had

278

a wind direction of NNE and a low wind speed (0.8 m s–1), and LAN is located about

279

120 km west of the Qinshan NPP. Because of the indefinable reasons for that sample,

280

its abnormal value was removed. A similar abnormal value (100.7±2.2‰) was

281

reported by Turnbull et al.

282

and she hypothesized that this data represented an unidentified analytical problem and

283

excluded the data.

284

12

for the observation at Tae–Ahn Peninsula, South Korea,

At SDZ and LHT, the seasonal averages of ∆14CO2 in summer/autumn are not 13

ACS Paragon Plus Environment

Environmental Science & Technology

285

significantly different from those in winter/spring (p > 0.05), while the seasonal

286

average in summer was significantly (p < 0.05) higher than that in winter/spring at

287

LAN. For the LFS site, the seasonal ∆14CO2 average in autumn was significantly (p
70‰).42 For 2015, the average 14

ACS Paragon Plus Environment

Page 14 of 29

Page 15 of 29

Environmental Science & Technology

310

contributions of CO2ff to the annual ∆CO2 offsets are 34.7±26.4% at LHT, 61.7±25.6%

311

at SDZ, and 56.7±38.0% at LAN. A ∆CO2 offset indicates differences in CO2

312

concentrations between the regional sites and the WLG site (401.0±3.5 ppm in

313

2015).43

314

The annual average CO2ff concentrations at LAN (12.7±9.6 ppm) and SDZ

315

(11.5±8.2 ppm) were significantly (p < 0.05) higher than those at LHT (4.6±4.3 ppm)

316

in 2015. LAN and SDZ are located in the Yangtze River Delta Economic Region and

317

Beijing-Tianjin-Hebei Economic Region, respectively, and strong CO2 emissions

318

from fuel combustions in these two most developed regions38 in China could result in

319

the high CO2ff concentrations at LAN and SDZ. The megacities and industrial sources

320

in the two regions were also regarded as the main reasons for the high atmospheric

321

CO2 values at LAN and SDZ.10,11 However, the low CO2ff concentrations at LHT

322

might be related to advantageous atmospheric diffusion (with generally high wind

323

speeds of 3.4–7.9 m s–1) for this coastal site, no fossil fuel consumption for heating

324

during winter in Sanya, and a smaller number of heavy industries in this tourist city.

325

Reduced CO2 emissions from fuel combustions in Hainan Island are supported by the

326

study of Wang et al. (2013).38

327 328

Temporal CO2ff variations. No significant (p > 0.05) differences in CO2ff

329

concentrations were found among different seasons at SDZ, LHT and LAN (Figure

330

S3). Additionally, no significant (p > 0.05) differences in CO2ff concentrations were

331

found according to the time of day (10:00, 18:00, and 22:00) at SDZ, LFS, LHT and

332

LAN (Figure S4).

333 334

Evaluation of local and regional emissions on CO2ff variations. Figure 4 shows 15

ACS Paragon Plus Environment

Environmental Science & Technology

335

CO2ff concentrations along 16 directions at SDZ, LAN and LHT. This analysis was

336

not carried out for LFS due to the short observation period. It clearly shows that CO2ff

337

at SDZ, LAN and LHT occurred with wind directions from the NNE–WSW sector,

338

N–WSW sector and N–SE sector, respectively.

339

To further ascertain whether CO2ff inputs at these sites were influenced primarily by

340

local sources (≤ ~ 10 km) or by regional sources (> ~ 10 km), CO2ff concentrations

341

were divided according to the method used by Fang et al. (2014) based on long time

342

series observations at these regional background sites. 10 They identified the locations

343

and directions of the primary fossil fuel emission sources around these background

344

sites, and defined a local event when wind passed those sources. Additionally, they

345

ascribed CO2 concentrations at low wind speeds (