Evidence of Influence of Human Activities and ... - ACS Publications

Subscriber access provided by Kaohsiung Medical University

Environmental Processes

Evidence of Influence of Human Activities and Volcanic Eruptions on Environmental Perchlorate from a 300-Year Greenland Ice Core Record Jihong Cole-Dai, Kari Marie Peterson, Joshua Andrew Kennedy, Thomas S. Cox, and David G. Ferris Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.8b01890 • Publication Date (Web): 26 Jun 2018 Downloaded from http://pubs.acs.org on June 28, 2018

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 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 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.

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

1

Evidence of Influence of Human Activities and Volcanic Eruptions on Environmental

2

Perchlorate from a 300-Year Greenland Ice Core Record

3

Jihong Cole-Dai1, Kari M. Peterson1, Joshua A. Kennedy1*, Thomas S. Cox2, and David G.

4

Ferris3

5 6

1: Department of Chemistry and Biochemistry, South Dakota State University, Box 2202, Avera Health and Science Center, Brookings, SD 57007, USA

7

2: Department of Physical Sciences, Butte College, Oroville, CA 95965, USA

8

3: Department of Earth Sciences, Dartmouth College, Hanover, NH 03755, USA

9

*Corresponding author: [email protected]

ACS Paragon Plus Environment


10 11

Page 2 of 30

ABSTRACT A 300-year (1700-2007) chronological record of environmental perchlorate,

12

reconstructed from high-resolution analysis of a central Greenland ice core, shows that

13

perchlorate levels in the post-1980 atmosphere were two-to-three times those of the pre-1980

14

environment. While this confirms recent reports of increased perchlorate in Arctic snow since

15

1980 compared with the levels for the prior decades (1930-1980), the longer Greenland record

16

demonstrates that the Industrial Revolution and other human activities, which emitted large

17

quantities of pollutants and contaminants, did not significantly impact environmental perchlorate,

18

as perchlorate levels remained stable throughout the eighteenth, nineteenth, and much of the

19

twentieth centuries. The increased levels since 1980 likely result from enhanced atmospheric

20

perchlorate production, rather than from direct release from perchlorate manufacturing and

21

applications. The enhancement is probably influenced by the emission of organic chlorine

22

compounds in the last several decades. Prior to 1980, no significant long-term temporal trends in

23

perchlorate concentration are observed. Brief (a few years) high concentration episodes appear

24

frequently over an apparently stable and low background (~1 ng kg‒1). Several such episodes

25

coincide in time with large explosive volcanic eruptions including the 1912 Novarupta/Katmai

26

eruption in Alaska. It appears that atmospheric perchlorate production is impacted by large

27

eruptions in both high and low latitudes, but not by small eruptions and non-explosive degassing.

28 29 30 31

1 ACS Paragon Plus Environment

Page 3 of 30

32


TOC Art

33 34 35 36

INTRODUCTION Perchlorate (ClO4-) is water soluble, kinetically stable, and ubiquitous in the environment.

37

It has been found to be widespread in both terrestrial soil and water systems.1-8 Environmental

38

perchlorate may be a significant health risk to vulnerable populations, due to its inhibition of

39

iodine uptake in the thyroid, disrupting normal thyroid function.9-10 Ammonium and potassium

40

perchlorate are produced and used in rocket propellant, munitions, fireworks, and road flares,

41

among many applications. During production and use, perchlorate may be disbursed into the

42

environment. In the southwestern United States, for example, manufacture of perchlorate salts

43

and associated waste releases are believed to have elevated perchlorate levels in the lower

44

Colorado River.11 In the United States, however, perchlorate is not designated a pollutant, and

45

no nationwide enforceable limits have been established for perchlorate in drinking water,

46

although several states regulate perchlorate for public health.12

47

Perchlorate is found at trace levels in many areas of the world. Terrestrial areas where

48

perchlorate has been detected include arid regions such as Chile, the southwestern United States,

49

and the Dry Valleys of Antarctica, locations far from known pollution sources.2, 13-15 Perchlorate 2 ACS Paragon Plus Environment


Page 4 of 30

50

has also been detected in subsurface waters with no possible input from pollution or direct

51

atmospheric deposition.6 These findings demonstrate that perchlorate either is or has been

52

formed naturally. Available evidence suggests that natural sources of perchlorate are likely to be

53

atmospheric in origin.16 The most likely initial precursor is chloride, the most common form of

54

chlorine in the surface environment. In several proposed perchlorate formation mechanisms1, 17-

55

21

56

phenomena (e.g., lightning and volcanic eruptions) are believed to play a significant role.17, 22 It

57

has been suggested that at least some perchlorate may be formed in the stratosphere.23-24

, important atmospheric oxidants (e.g., ozone) are suggested to be involved and natural

58

Knowledge of the natural background and variability of perchlorate in the environment

59

not only is valuable to the effort to establish regulatory limits and thresholds for environmental

60

perchlorate to protect public health, it can also provide insight into important aspects (e.g.,

61

oxidants and oxidation kinetics) of atmospheric chemistry and of large-scale environmental and

62

natural processes. Such knowledge can be obtained from records of perchlorate in the

63

environment. Snow carries chemical substances from the atmosphere and accumulates

64

continuously in polar regions such as Greenland and Antarctica, and on high-elevation mountain

65

glaciers. Thus, chronological records of chemical substances in the environment can be obtained

66

from ice core chemical measurements. Polar ice cores have yielded valuable records that

67

provide insight into the climate system25 and chemical characteristics26-28 of the atmosphere.

68

Extended and detailed records can be used to assess the relative contributions of natural and

69

anthropogenic sources of a pollutant, and to investigate atmospheric processes and conditions

70

impacting variations of the chemical species over time. Several ice core studies have

71

investigated perchlorate specifically.23, 29-32 However, the brief and/or discontinuous records in

72

those studies make it difficult to assess natural variability, long-term trends, and relative source


Page 5 of 30


73

contributions. We have constructed a 300-year, high-resolution record of perchlorate from a

74

2007 central Greenland ice core, and shorter records from other Greenland ice cores. We use the

75

records to (1) discern major trends of environmental perchlorate in the recent past and over the

76

last three centuries, (2) assess possible anthropogenic impact on environmental perchlorate, and

77

(3) identify important factors affecting variability of perchlorate in the environment.

78 79

MATERIALS AND METHODS

80

Ice Core Collection

81

Several shallow ice cores were drilled during June and July, 2007 near Summit Station,

82

Greenland (73.6° N, 38.5° W). The top 97.98 meters of one core (SM07C2, 211 m) was

83

analyzed for perchlorate and used for this study, except for a few short intervals (19.29-20.28,

84

27.405-28.42, 67.45-69.48, 76.49-76.84, 83.35-84.32, and 86.19-87.14 m) where ice had been

85

consumed for other purposes. Perchlorate was also measured in a 1996 ice core from TUNU in

86

northern Greenland (78.1° N, 34.0° W), and a 2002 core from Basin 4 in southern Greenland

87

(62.3° N, 46.3° W). All ice core sections were wrapped in clean plastic lay-flat tubings and kept

88

frozen during transport from the field to the laboratory, where they were maintained at or below

89

‒20° C until chemical analysis.

90

Perchlorate Measurement

91

The procedures to prepare decontaminated Greenland ice core samples for measurement

92

of perchlorate and other species have been described by Peterson et al.33 Samples (2174 for

93

SM07C2, 445 for TUNU, and 410 for Basin 4) and procedural blanks were analyzed for

94

perchlorate using ion chromatography-tandem mass spectrometry with electrospray ionization 4 ACS Paragon Plus Environment


Page 6 of 30

95

(IC-ESI-MS/MS). As described previously,33 perchlorate was eluted from a Dionex IonPac®

96

AS16 (2 x 250 mm) analytical column with 60 mM NaOH at 0.3 mL min‒1. The effluent from

97

the ion chromatograph was mixed with an acetonitrile/water solution (90/10% v/v at 0.3 mL

98

min‒1). This mixture was delivered to the ionization/nebulization inlet of an AB SCIEX QTRAP

99

5500 triple quadrupole mass spectrometer. Negative ion mode was used to detect 35ClO4‒ and

100

37

101

transitions, respectively. Quantification was performed using the 35ClO4‒ peak area with external

102

calibration. The limit of detection and lower limit of quantification of the method were 0.1 and

103

0.3 ng kg‒1, respectively.33

104

Ice Core Dating

105

ClO4‒ using multiple reaction monitoring of the m/z 99.0 to 83.0 and m/z 101.0 to 85.0

Ice cores can be dated with the technique of annual layer counting (ALC), in which

106

annual peak-valley-peak oscillations in concentrations of certain chemical species are counted to

107

develop a depth-age relationship. Concentrations of major ions (Na+, K+, Mg2+, Ca2+, Cl‒, NO3‒,

108

SO42‒) in the top 97.98 m (except for 19.27-20.28 m, which was consumed in a previous study)

109

of the SM07C2 core (3259 samples) were measured with ion chromatography.34 The maximum

110

concentration of calcium (Ca2+) during a year in Greenland snow occurs in early spring, and

111

minimum concentration in autumn.34 At Summit, the chloride-to-sodium (Cl‒/Na+) mass ratio in

112

snow fluctuates seasonally and reaches an annual minimum during winter and a maximum

113

during summer (see data in Supporting Information). Oscillations in Ca2+ concentration and

114

Cl‒/Na+ ratio (SI Figure S1) were identified and used as primary annual layer markers to yield

115

307 years (1700-2007) in the SM07C2 core (0-97.98 m). Concentrations of Na+ and SO42‒,

116

which tend to reach maxima in summer and early spring, respectively,35 were used for auxiliary

117

confirmation of annual layers. For the depth interval of 19.27-20.28 m where no major ion data 5 ACS Paragon Plus Environment

Page 7 of 30


118

are available, two years (mid-1964 to mid-1966) were assigned, based on the average annual

119

accumulation rate for the depth interval of 15-25 m. The depth where the annual Cl‒/Na+ ratio

120

minimum occurs is assigned the month of January of each year. Therefore, an ice core year

121

corresponds to a calendar year in this 300-year chronology. No major ions were measured for the

122

TUNU and Basin 4 cores; these were dated using the average annual accumulation rate for each

123

site36-37 with the assumption of constant accumulation over the time covered by the length of

124

those core sections.

125

The dating uncertainty for the SM07C2 core at 100 years (~43 m) is less than 1 year and

126

is ±3 years at 300 years (~100 m). This is consistent with the uncertainty reported by Cole-Dai

127

et al. in an 800-year chronology of the 2007 Summit cores.34 The uncertainty in the dating of the

128

TUNU and Basin 4 cores was not determined but is expected to be slightly higher than the ALC-

129

dated Summit cores as a result of variations in accumulation rate.

130

Annual layer depth interval (thickness) was measured and presented as annual snow

131

accumulation (cm) in water equivalent. The amount of perchlorate (or other ions) in snow is

132

expressed in units of concentration (ng kg‒1 or µg kg‒1) when the analytical measurement is

133

presented, or in units of flux (µg m‒2 yr‒1 or mg m‒2 yr‒1) when discussing air-to-ground

134

deposition. For each year, annual flux was calculated by summing mass deposition of all samples

135

in the year. Since snow accumulation rates at Summit, Greenland have remained relatively stable

136

in recent centuries34, trends in concentration are similar to those in deposition flux.30

137 138

RESULTS AND DISCUSSION

139

Long-term Trend of Environmental Perchlorate 6 ACS Paragon Plus Environment


140

Perchlorate concentrations in the Summit core (Figure 1) are extremely low, with an

141

average concentration (±standard deviation) of 1.6±2.8 ng kg‒1. Prior to 1980, the relatively

142

stable concentration is punctuated by brief spikes of up to a few years in duration. The average

143

perchlorate concentration (Table 1) in the period from mid-nineteenth century to late twentieth

144

century (1850-1979, 1.2±1.2 ng kg‒1, excluding samples in 1912 and 1913 when perchlorate

145

concentration was drastically impacted by an explosive volcanic eruption, to be discussed later)

146

shows no significant change from that in the pre-industrial time (1700-1849, 1.2±1.0 ng kg‒1).

147

This is strong evidence that the large-scale human activities of the Industrial Revolution

148

beginning in the early-to-mid-nineteenth century did not increase the level of perchlorate in

149

Greenland snow and in the atmospheric environment. It has been speculated38-39 that the

150

widespread use of Chilean nitrate throughout North America and Europe in the nineteenth

151

century and the first half of the twentieth century may have introduced significant amounts of

152

perchlorate, an impurity in Chilean nitrate, to the environment; the Summit ice core record

153

suggests that the impact of Chilean nitrate use on environmental perchlorate was minimal.

154

Perchlorate Increase since 1980

155

Page 8 of 30

A remarkable change in perchlorate concentration is observed in Summit snow since

156

1980 (Figure 1). The average perchlorate concentration during 1980-2007 (2.7±2.1 ng kg‒1) is

157

more than double the pre-1980 average (1.2±1.2 ng kg‒1) and more than three times that (0.8±0.6

158

ng kg‒1) during the immediately preceding 30 years (1950-1979, Table 1), when rockets in early

159

space exploration could have introduced perchlorate into the environment. This observation was

160

initially reported in Peterson et al.30 using a shorter and discontinuous part of the SM07C2 core

161

dataset. The increase of environmental perchlorate since 1980 was first discovered by Rao et al.31

162

who found that perchlorate concentration in post-1980 snow samples from Eclipse Icefield in 7 ACS Paragon Plus Environment

Page 9 of 30


163

Canada increased to over 2.2 ng kg‒1 from the pre-1980 level (0.6 ng kg‒1 during 1973-1976).

164

Those findings are unambiguously confirmed in this much longer (300-year) and continuous

165

record with preindustrial background. Moreover, similar post-1980 increases are found in ice

166

cores from the other Greenland locations, TUNU and Basin 4 (Table 2, SI Figures S2 and S3).

167

The change in perchlorate concentrations in North America and Greenland snow after 1980 is

168

also similar to increases observed in snow samples and ice cores (Table 2) from several other

169

Arctic locations (Nunavut, Canada; Yukon Territory, Canada). For example, Furdui et al.32

170

recently reported that annual perchlorate deposition flux on the Agassiz Ice Cap in Nunavut

171

increased from 0.42 µg m‒2 during the 1936-1979 period to 1.14 µg m‒2 during 1980-2007.

172

Possible or probable causes of the apparent increase of environmental perchlorate since

173

1980 in the Arctic, and probably North America, as documented in the recent studies (Table 2),

174

need to be investigated. The additional perchlorate may be released during industrial perchlorate

175

production or large-scale applications. Available data suggest that industrial production of

176

perchlorate salts in the United States began to increase in the late 1950s and a substantial

177

production increase may have occurred during the 1980s.38 Perchlorate pollution stemming from

178

release during production or disposal of unused perchlorate salts is likely to be confined to only

179

local or regional surface environments and groundwater. Perchlorate in localized pollution may

180

become airborne and possibly be transported to remote areas via atmospheric circulation. In the

181

atmosphere, however, the non-volatile perchlorate is likely associated with dust,40 which is

182

generally not transported efficiently over long distances,41 and therefore unlikely to be recorded

183

in Arctic snow. In addition, industrial production of perchlorate in the United States is estimated

184

to have peaked during the late 1980s,38 whereas the perchlorate concentrations in Greenland

185

snow are highest in the early and mid-1990s (Figure 2). The difference in timing provides



Page 10 of 30

186

further evidence that perchlorate release during production or disposal is unlikely to be

187

responsible for the observed increase of perchlorate in the Arctic and North America since 1980.

188

It was speculated30 that perchlorate residues in solid fuels during frequent rocket launches

189

(e.g., the United States Space Shuttle program, 1981-2005) may be dispersed over the Northern

190

Hemisphere and the High Arctic. However, analyses of space shuttle launch plumes and

191

highway road flares have shown that essentially no perchlorate residue remains when

192

perchlorate-containing explosives are used as intended.38 Therefore, no significant amount of

193

perchlorate would be expected to be released into the environment by major applications; this is

194

supported by the fact that the average perchlorate concentration during the early rocket-age

195

period of 1950-1979 (Table 1) is no higher than during the previous 250 years.

196

Perchlorate Formation in the Atmosphere

197

Since perchlorate is also formed in the atmosphere, the post-1980 increase of perchlorate

198

may be caused by enhanced atmospheric production. Among several proposed chemical

199

mechanisms of atmospheric perchlorate formation, two appear to be the leading or principal

200

pathways. In one process, as has been demonstrated17, 22, 38 in laboratory studies, perchlorate is

201

produced from chloride in aerosols subjected to electric discharge simulating lightning. In

202

another process, further oxidation of chlorine radicals and oxy-chlorine species generated in

203

reactions involving ozone and other oxidants leads to perchlorate formation17, 21, 38, among many

204

other reaction products (e.g., chlorate).

205

Furdui et.al. proposed32 that, prior to 1960, a significant fraction of perchlorate in Arctic

206

snow results from chloride aerosols, based on a strong correlation between chloride and

207

perchlorate in an Agassiz (Nunavut, Canada) ice core during 1940-1959. Despite this proposal,


Page 11 of 30


208

Furdui et al. did not attribute the post-1980 perchlorate increase in Arctic snow to the source of

209

aerosol chloride, for no correlation between chloride and perchlorate is observed in Agassiz

210

snow samples after 1980. No significant correlation between chloride and perchlorate is found

211

in the Summit, Greenland core for the period of 1940-1959, or during any previous time periods

212

in the 300-year record (SI Figures S4-S6). This suggests that, at Summit, perchlorate is not

213

dominated by formation from aerosol chloride either before or after 1980. Therefore, the increase

214

since 1980 cannot be attributed to elevated levels of either aerosol chloride from sea-salt or HCl

215

emitted by volcanic eruptions and degassing. A significantly increased frequency or intensity of

216

lightning could enhance perchlorate formation rate; however, there is no evidence that lightning

217

frequency has increased since 1980.42-43

218

A major source of chlorine radicals and oxy-chlorine species in the atmosphere are

219

organic chlorine compounds from both natural and anthropogenic emissions.44 In a recent study,

220

Jiang et al.23 found higher perchlorate levels in Antarctic snow since the 1970s than those in

221

older snow and suggested that this increase is correlated with stratospheric chlorine, which has

222

increased significantly since the 1970s due to anthropogenic emission of organic chlorine

223

compounds.45 Because of possible post-depositional change to perchlorate in Antarctic snow23,

224

the impact of anthropogenic chlorine on perchlorate was difficult to assess. In exploring possible

225

causes of the post-1980 increase in the Arctic, Furdui et al. examined29, 32 the emission records of

226

several organic chlorine compounds and suggested that the increased emission of methyl

227

chloroform (CH3CCl3) since 1970 and the drastic decrease since the mid-1990s are likely the

228

main contributors to the perchlorate trend since 1980. This suggestion is similar to that by Jiang

229

et al.23: a substantial increase of perchlorate in polar snow beginning around 1980, followed by a



230

leveling-off in the mid-1990s and a slight decrease since that time, likely reflects the trend of

231

anthropogenic emissions of organic chlorine compounds.

232

Correlation between Atmospheric Chlorine and Perchlorate

233

Page 12 of 30

Evidence of the trend of organic chlorine and impact of emissions of organic chlorine

234

compounds can be found in atmospheric measurements for the recent decades, and in

235

atmospheric concentrations calculated for time periods prior to the measurements. For example,

236

tropospheric equivalent chlorine (EC), 80% of which is contributed by organic chlorine,

237

increased from approximately 2 ppbv in the 1970s to about 5 ppbv during the 1990s, and has

238

decreased slightly (to about 4.5 ppbv) since the mid-1990s.44 Jiang et al. found23 an apparent

239

correlation between perchlorate in Antarctic snow and equivalent effective stratospheric chlorine

240

(EESC). EESC is calculated from measurements of organic chlorine compounds in the

241

troposphere or surface air and represents the ozone-depleting potential of halogens in the

242

stratosphere. The trend of EESC lags that of EC by approximately 3 years at the equator and up

243

to 7 years at the poles.45 The increase of both EC and EESC in the second half of the twentieth

244

century was a result of emissions of chlorofluorocarbons (CFCs) and other long-lived organic

245

chlorine compounds (very short-lived chlorine compounds account for a small fraction of EC).45-

246

46

247

and adjustments to reduce ozone-depleting substances in the atmosphere, tropospheric chlorine

248

reached the highest level in the mid-1990s and has been declining gradually, while stratospheric

249

chlorine has followed a similar trend. Between 1950 and 1991, tropospheric/surface chlorine

250

concentration, estimated from the emission of long-lived organic chlorine compounds45 (CCl4,

251

CH3Cl, CH3CCl3, CFC-12, and CFC-11) summed and scaled for the number of chlorine atoms in

252

each compound, shows an increase that accelerated significantly in the 1970s and 1980s (Figure

Because of the implementation of the 1987 Montreal Protocol and subsequent amendments


Page 13 of 30


253

3, details in SI Table S1 and Figure S7). Measurements of major organic chlorine species since

254

the early 1990s shows a gradual decrease in tropospheric chlorine (Figure 3).47 The increase in

255

perchlorate concentration during the 1980s and the slight decrease since the mid-1990s in

256

Summit snow appears to largely follow the trend of tropospheric chlorine (Figure 3), indicating

257

that environmental perchlorate may be sensitive to changes in the atmospheric burden of organic

258

chlorine. Note that in Figure 3, several brief high-concentration episodes appear to be

259

superimposed on the broad trend paralleling that of tropospheric chlorine. The perchlorate

260

maxima in the mid-1960s, early 1980s and early 1990s coincide in time with the Agung (1963),

261

El Chichón (1982) and Pinatubo (1991) volcanic eruptions, which may have briefly enhanced

262

perchlorate production in the atmosphere (discussed next). Because of the influence of the

263

Pinatubo eruption, it is not possible to determine the exact timing of maximum perchlorate

264

deposition in the 1990s and to use the perchlorate data in Figure 3 to determine if perchlorate in

265

the snow is dominated by tropospheric or stratospheric chlorine.

266

The records from North America, Greenland, the Canadian Arctic, and Antarctica

267

indicate that the perchlorate increase since 1980 is likely global, rather than regional, and is

268

consistent with the global impact of anthropogenic emissions of organic chlorine compounds.

269

Volcanic Influence on Episodic Perchlorate Increase

270

Numerous relatively short (less than 3 years) episodes of elevated perchlorate

271

concentration and deposition flux are observed in the 300-year record (Figure 1). The largest of

272

these events occurs in the depth range of 38.3 to 39.2 m (Figure 4) during 1912-1914. The

273

highest perchlorate concentrations during this period approach 50 ng kg‒1, compared to the

274

average of approximately 0.9 ng kg‒1 during the periods (about 10 years) immediately before and



275

after 1912-1914. The perchlorate flux for 1912 through 1913 (7.7 and 4.9 µg m‒2 yr‒1,

276

respectively) represents the largest deposition in the 300-year record.

Page 14 of 30

277

A strong and explosive (VEI 6) eruption of the Novarupta (Katmai) volcano (58.27° N,

278

155.16° W) in Alaska occurred on June 6, 1912. It is well established that sulfate in polar snow

279

increases significantly following large explosive volcanic eruptions.34-35 The Novarupta eruption

280

is marked in the ice core by substantially increased sulfate concentrations (Figure 4) during 1912

281

and 1913, which coincides with the drastically elevated perchlorate concentration and deposition

282

flux (Figure 4). This appears to support previous suggestions that eruptions may increase the

283

amount of perchlorate in the environment.29, 31 The extraordinarily high perchlorate

284

concentrations and deposition, along with simultaneous increase of sulfate following the

285

Novarupta eruption, were also found in a replicate Summit core (SM07C4, SI Figure S8),

286

demonstrating that the remarkable perchlorate episode is not an artifact from ice core sampling

287

or chemical analysis.

288

Peterson et al. found30 in a portion of the Summit (SM07C2) perchlorate record that high

289

perchlorate episodes in recent snow coincide in time with the eruptions of Pinatubo (15.14° N,

290

120.35° E) in 1991, and El Chichón (17.36° N, 93.23° W) in 1982, two stratospheric eruptions in

291

the tropics (Table 3). A few additional instances of elevated perchlorate levels are found to be

292

associated with large volcanic eruptions during the 300-year period. Deposition flux of

293

perchlorate is significantly elevated (Table 3 and Figure 5) following the eruptions of Tambora

294

(1815), Babuyan Claro (1831), Cosigüina (1835), and Krakatoa (1883). Note that, due to lack of

295

ice, no perchlorate measurement was made in the period of 1808-1816 when Tambora and

296

another large eruption occurred; and, therefore, the perchlorate response to Tambora is only

297

partially seen in Figure 5. These four, as well as El Chichón and Pinatubo, are known to have 13 ACS Paragon Plus Environment

Page 15 of 30


298

been powerful eruptions (VEI 4-7) with probable injection of volcanic ash and gas (e.g., SO2)

299

directly into the stratosphere.34 The Summit record suggests that low-latitude stratospheric

300

eruptions are also capable of elevating perchlorate concentrations in the atmosphere leading to

301

increased deposition in Arctic snow. Powerful eruptions that inject SO2 into the stratosphere are

302

associated with short-term (a few years) increases in aerosol optical depth (AOD). The El

303

Chichón and Pinatubo (and the 1963 eruption of Agung in Indonesia) impact on Northern

304

Hemisphere AOD, for instance, is clearly observed (Figure 2), and the perchlorate response in

305

each case is contemporaneous with the increase in stratospheric aerosol; the perchlorate response

306

to Agung is only partially seen (Figures 2 and 3), due to lack of ice for measurement. In the

307

Agassiz ice core, no perchlorate spike was observed to correspond to elevated AOD by El

308

Chichón or Pinatubo;32 however, unusually high perchlorate flux was detected in 1981 and in

309

1990, just one year prior to the eruptions of El Chichón and Pinatubo, respectively, and the

310

increase in AOD.

311

Furdui et al. investigated32 a possible connection between volcanic eruptions and

312

perchlorate by examining the relationship between perchlorate and chloride in the Agassiz ice

313

core, based on the assumption that chlorine (HCl) emission from recent volcanic eruptions in

314

Alaska, Aleutian Islands, Kuril Islands and Kamchatka Peninsula may enhance atmospheric

315

perchlorate production. No correlation was found during the period of 1970-2007 in that study;

316

in addition, no correlation was found between sulfate and perchlorate. It was therefore

317

concluded that volcanic activity may not significantly increase perchlorate deposition at Agassiz.

318

The Summit record supports the conclusion by Furdui et al. that chlorine emissions from

319

volcanic eruptions and non-explosive degassing in mid- and high latitudes have no significant

320

impact on atmospheric perchlorate production.



321

Page 16 of 30

The 300-year Summit record provides strong evidence of impact on perchlorate by

322

volcanic eruptions that inject a substantial amount of gas and aerosols directly into the

323

stratosphere. It is apparent that this impact is not limited to high latitude volcanic eruptions (e.g.,

324

Novarupta), and that injections to the stratosphere from a low latitude eruption (e.g., Pinatubo

325

and Tambora) could significantly enhance perchlorate production for one or two years

326

immediately following the eruption. Although it is not clear how these large eruptions increase

327

perchlorate levels for brief time periods following the eruptions, it appears that direct injection of

328

volcanic aerosols into the stratosphere is necessary. However, a comprehensive examination of

329

the volcanic impact on atmospheric perchlorate formation is beyond the scope of this study. The

330

specific processes and possible reaction mechanisms by which volcanic eruptions impact

331

perchlorate will be investigated in future work.

332

The Summit data show that many small perchlorate episodes in the 300-year record are

333

not associated with known large volcanic eruptions. This suggests that perchlorate production

334

may be impacted by factors other than large, stratospheric volcanic eruptions, and the emission

335

of organic chlorine compounds.

336 337

ASSOCIATED CONTENT

338

The following content is presented in Supporting Information: (1) Summit ice core

339

dating; (2) Perchlorate in other Greenland ice cores; (3) Correlation analysis between perchlorate

340

and chloride; (4) Estimating tropospheric chlorine concentration; and (5) Perchlorate response to

341

the 1912 Novarupta/Katmai eruption in two central Greenland ice cores.

342 15 ACS Paragon Plus Environment

Page 17 of 30

343 344


ACKNOWLEDGEMENTS Funding for this work was provided by U.S. National Science Foundation (Awards

345

1203533, 1443663; Major Research Instrumentation Award 0922816 for the acquisition of a

346

tandem mass spectrometer). The U.S. National Ice Core Laboratory provided the TUNU and

347

Basin 4 ice cores. The South Dakota State University Campus Core Mass Spectrometry Facility

348

and Linhong Jing are acknowledged for providing technical support and maintenance of the mass

349

spectrometer. Erica Manandhar, Scott Splett, Alexandria Kub, and T. Zack Crawford assisted

350

with the laboratory analysis.

351



Page 18 of 30

352 353 354

Figure 1. Concentration of perchlorate (smoothed with a 3-sample running mean) in the Summit (SM07C2) core from 0 to 97.98 meters depth, covering the period from 2007 to 1700.

355

356 357 358

Figure 2. Aerosol optical depth (a), annual perchlorate flux (b), and annual chloride flux (c) in the period 1940-2006 in the Summit (SM07C2) core.


Page 19 of 30


359

360 361 362 363 364 365 366

Figure 3. Measured surface chlorine concentrations (thick line) from Montzka et al.47; the total atmospheric chlorine (dashed line, scaled for number of chlorine atoms) is the sum of estimates of CCl4, CH3Cl, CFC-12, CFC-11, and CH3CCl3 surface concentrations adapted from data in Newman et al.45, and annual perchlorate flux (3-year running mean, histogram) in the Summit core. See Supporting Information for description of estimating surface chlorine concentrations.



Page 20 of 30

367

368 369 370 371

Figure 4. Perchlorate (a) and sulfate (b) concentrations in the Summit core impacted by the 1912 Novarupta/Katmai eruption (dashed vertical line); data smoothed with a 3-sample running mean.


Page 21 of 30


372

373 374 375 376 377 378

Figure 5. Perchlorate (a) and sulfate (b) annual flux in the period of 1800-1900 are impacted by the eruptions of Tambora (1815), Babuyan Claro (1831), Cosigüina (1835), and Krakatoa (1883). Dashed vertical lines indicate the eruption years. Perchlorate data are not available for 18091816, when Tambora and an unidentified volcano erupted (1809), because of lack of ice for perchlorate measurement.

379 380



381 382 383 384

Page 22 of 30

Table 1. Average perchlorate concentrations in the Summit core during several time periods since 1700. Perchlorate concentrations in samples in the 1912-1913 period are excluded from average calculations, due to unusually high concentrations influenced by a volcanic eruption. Averages and sample numbers including the 1912-1913 samples are given in parentheses. Depth (m) 0-97.98 0-13.84 13.84-25.90 25.90-58.49 58.49-97.98

Years 1700-2007 1980-2007 1950-1979 1850-1979 1700-1849

Average Concentration (ng kg-1) 1.4 (1.6) th Late 20 Century 2.7 Early Rocket Age 0.8 Industrial Age 1.2 (1.6) Pre-Industrial 1.2 Period

Number of Samples 2147 (2174) 264 250 1024 (1051) 859

385 386 387 388 389 390 391

Table 2. Annual snow accumulation rates (in water equivalent) and perchlorate concentrations at sites in the Arctic and North America. Ice cores cover the time periods of 1700-2007 (Summit), 1918-1996 (TUNU), 1972-2002 (Basin 4), 1996-2005 (Devon Island), 1936-2007 (Agassiz Ice Cap, estimated median concentration), 1970-1973 and 1982-1986 (Eclipse Icefield), and 1726-1993 (Fremont Glacier, Wyoming, United States). Location

Greenland Summit34 TUNU36 Basin 4 (ref. 27) Canadian Arctic Devon Island48 Agassiz Ice Cap29, 32, 49 Eclipse Icefield31 North America Fremont Glacier, U.S.A.31

Annual Accumulation (cm)

Average Concentration (ng kg-1) Pre-1980

Post-1980

22.6 12.5 41.1

1.2 1.0 0.9

2.7 3.6 2.8

24.1 10.0 130

N/A 4.2 0.6

5.5 11.4 2.3

76

Evidence of Influence of Human Activities and ... - ACS Publications

Recommend Documents