Local source impacts on primary and secondary aerosols in the Midwestern United States

Local source impacts on primary and secondary aerosols in the Midwestern United States

Accepted Manuscript Local source impacts on primary and secondary aerosols in the Midwestern United States Thilina Jayarathne, Chathurika M. Rathnayak...

3MB Sizes 0 Downloads 6 Views

Accepted Manuscript Local source impacts on primary and secondary aerosols in the Midwestern United States Thilina Jayarathne, Chathurika M. Rathnayake, Elizabeth A. Stone PII:

S1352-2310(15)30400-3

DOI:

10.1016/j.atmosenv.2015.09.058

Reference:

AEA 14139

To appear in:

Atmospheric Environment

Received Date: 2 June 2015 Revised Date:

22 September 2015

Accepted Date: 23 September 2015

Please cite this article as: Jayarathne, T., Rathnayake, C.M., Stone, E.A., Local source impacts on primary and secondary aerosols in the Midwestern United States, Atmospheric Environment (2015), doi: 10.1016/j.atmosenv.2015.09.058. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

1 2 3 4 5 6 7 8

Local source impacts on primary and secondary aerosols in the Midwestern United States Thilina Jayarathneφ, Chathurika M. Rathnayakeφ, Elizabeth A. Stone* Department of Chemistry, University of Iowa, Iowa City, IA 52242, United States

9

to poor representation of outdoor air pollutants in human exposure assessments. To examine

φ

RI PT

Co-first authors *Corresponding author: +1-319-384-1863; fax: +1-319-335-1270; e-mail: [email protected] Atmospheric particulate matter (PM) exhibits heterogeneity in composition across urban areas, leading

heterogeneity in PM composition and sources across an urban area, fine particulate matter samples

11

(PM2.5) were chemically profiled in Iowa City, IA from 25 August to 10 November 2011 at two monitoring

12

stations. The urban site is the federal reference monitoring (FRM) station in the city center and the peri-

13

urban site is located 8.0 km to the west on the city edge. Measurements of PM2.5 carbonaceous aerosol,

14

inorganic ions, molecular markers for primary sources, and secondary organic aerosol (SOA) tracers

15

were used to assess statistical differences in composition and sources across the two sites. PM2.5 mass

16

ranged from 3 – 26 µg m-3 during this period, averaging 11.2 ± 4.9 µg m-3 (n=71). Major components of

17

PM2.5 at the urban site included organic carbon (OC; 22%), ammonium (14%), sulfate (13%), nitrate (7%),

18

calcium (2.9%), and elemental carbon (EC; 2.2%). Periods of elevated PM were driven by increases in

19

ammonium, sulfate, and SOA tracers that coincided with hot and dry conditions and southerly winds.

20

Chemical mass balance (CMB) modeling was used to apportion OC to primary sources; biomass burning,

21

vegetative detritus, diesel engines, and gasoline engines accounted for 28% of OC at the urban site and

22

24% of OC at the peri-urban site. Secondary organic carbon from isoprene and monoterpene SOA

23

accounted for an additional 13% and 6% of OC at the urban and peri-urban sites, respectfully.

24

Differences in biogenic SOA across the two sites were associated with enhanced combustion activities in

25

the urban area and higher aerosol acidity at the urban site. Major PM constituents (e.g., OC, ammonium,

26

sulfate) were generally well-represented by a single monitoring station, indicating a regional source

27

influence. Meanwhile, nitrate, biomass burning, food cooking, suspended dust, and biogenic SOA were

AC C

EP

TE D

M AN U

SC

10

1

ACCEPTED MANUSCRIPT

not well-represented by a single site and demonstrated local influences. For isoprene SOA, product

29

distributions indicated a larger role for the high-NOx pathway at the urban site. These local sources are

30

largely responsible for differences in population exposures to outdoor PM in the study domain located

31

within the Midwestern US.

32

Keywords: secondary organic aerosol; isoprene; urban air quality; source apportionment

AC C

EP

TE D

M AN U

SC

RI PT

28

2

ACCEPTED MANUSCRIPT

33 34

1. Introduction Ambient particulate matter (PM) contributes substantially to the global burden of disease, via respiratory infections, chronic respiratory disease, cardiovascular disease, and cancer (Lim et al., 2012).

36

PM concentrations and composition varies regionally, locally, and seasonally, driven by differences in

37

sources, climate, meteorology, and topography (Bell et al., 2008; Jerrett et al., 2013). In population

38

exposure studies, PM is typically assumed to be uniformly distributed across an urban area by

39

extrapolating data from a single long-term monitoring station. This assumption may introduce

40

significant biases, confounding associations between PM constituents and health outcomes (Bell et al.,

41

2011). For this reason, it is necessary to identify and assess the impacts of local sources that drive

42

differential exposures to outdoor air pollutants in an urban environment in order to improve their

43

representation in human exposures.

SC

M AN U

44

RI PT

35

Primary PM sources emit particles to the atmosphere and each aerosol source category has a unique chemical fingerprint defined by its elemental and molecular composition (Hildemann et al.,

46

1991; Schauer and Cass, 2000; Simoneit, 1985, 1999). The chemical fingerprint of the aerosol source is

47

embedded in the resultant particulate matter and can be used to identify and quantify source

48

contributions at a receptor location. For the apportionment of carbonaceous aerosol, organic molecular

49

markers that are unique to specific categories are the most useful tracers. To be considered molecular

50

markers, organic species must also be stable in the atmosphere during transport from source to

51

receptor and present in the aerosol phase at ambient temperatures and pressures (Schauer et al., 1996).

52

Some important molecular markers include levoglucosan for biomass burning (Schauer et al., 2001;

53

Simoneit et al., 1999b); sterols that can be used to differentiate between biomass sources (Sheesley et

54

al., 2003); n-alkanes with an odd-carbon preference for vegetative detritus (Rogge et al., 1993b); and

55

cholesterol and fatty acids for food cooking (McDonald et al., 2003; Rogge et al., 1991; Schauer et al.,

56

1999b). Furthermore, hopanes, steranes, and PAH may be used to differentiate between fossil fuel types

AC C

EP

TE D

45

3

ACCEPTED MANUSCRIPT

57

and vehicular emissions (Lough et al., 2007; Rogge et al., 1993c, 1997; Schauer et al., 1999c). Molecular-

58

marker driven source apportionment modeling provides the link between atmospheric PM and its

59

sources that are critical for air resources management. Secondary aerosols formed in the atmosphere can similarly be traced by individual chemical

RI PT

60

species. With the combination of ambient measurements and smog chamber experiments, SOA tracers

62

have been identified from globally-important biogenic gases, such as isoprene (Claeys et al., 2004; Edney

63

et al., 2005; Surratt et al., 2008), monoterpenes (Claeys et al., 2007; Jaoui et al., 2005), sesquiterpenes

64

(Jaoui et al., 2007), and green leaf volatiles (Shalamzari et al., 2014). SOA tracers provide direct insight to

65

the chemical speciation of SOA precursors and products in diverse ambient environments ranging from

66

boreal forests in Finland (Kourtchev et al., 2008), to the Southeastern United States (Lin et al., 2013),

67

Amazonia (Claeys et al., 2010), and Nepal (Stone et al., 2012). The distribution of isoprene SOA products,

68

for example, reveals the influence of NOx; 2-methylglyceric acid is the predominant product under high-

69

NOx conditions, while 2-methyltetrols and five-carbon alkene triols predominate under low-NOx

70

conditions (Surratt et al., 2006). In addition to oxidants, aerosol acidity and relative humidity have

71

proven to be important determinants in the amount and composition of SOA (Chan et al., 2011; Surratt

72

et al., 2007; Zhang et al., 2012). Incorporation of SOA mechanisms and evaluation of models with

73

specific SOA products has markedly improved model predictions of biogenic SOA (Pye et al., 2013).

74

Moreover, the development of the SOA-tracer method to apportion SOA to its precursors using

75

chamber-derived SOA tracer-to-OC ratios (Kleindienst et al., 2007) has helped to close the gap in

76

understanding the sources of carbonaceous aerosol (Lewandowski et al., 2008).

M AN U

TE D

EP

AC C

77

SC

61

The Midwestern US has been shown to be strongly impacted by biogenic and anthropogenic

78

SOA across urban and background sites (Lewandowski et al., 2008; Rutter et al., 2014; Stone et al.,

79

2009). The focus of the present study is to examine the extent of heterogeneity in PM2.5 composition

80

and sources within a single urban area. Paired daily measurements of OC, inorganic ions, and organic

4

ACCEPTED MANUSCRIPT

species at two locations in Iowa City, IA are used to identify the sources of PM that vary within the urban

82

area. The combination of molecular markers for primary sources and SOA tracers allows for a deeper

83

understanding of the impacts of local activities on SOA formation during the late-summer and fall

84

seasons.

RI PT

81

85 2. Materials and methods

87

2.1. PM2.5 sample collection

88

SC

86

PM2.5 was collected daily at two sites in Iowa City from 25 August to 10 November 2011. The periurban site (+41.6647, -91.5845) was located in a peri-urban area, surrounded by woods, agricultural

90

fields, and a parking lot and was generally considered to be a background site. The urban site (+41.6572,

91

-91.5035) was located in an urban area adjacent to an elementary and high school and residences, and

92

was generally considered to have more anthropogenic influences. PM2.5 was collected simultaneously

93

at both locations using medium volume PM2.5 samplers (3000B, URG Corp., Chapel Hill, NC) affixed to

94

platforms 3-4 m above ground level. Particles with aerodynamic diameter less than 2.5 µm were

95

selected by a cyclone (URG) operating at a flow rate of 90 L min-1. Air flow rates were measured before

96

and after sample collection with a calibrated rotameter (Gilmont Inst., Barrington, IL). PM2.5 was

97

collected on 90 mm quartz fiber filters (Pallflex® Tissuquartz™ Filters, Pall Life Sciences, NY) that were

98

pre-cleaned by baking at 550 °C for 18 h to remove any organic species. QFF were stored in aluminum

99

foil-lined petri dishes sealed with Teflon tape, and stored frozen before and after sample collection.

100

Seventy eight filter samples were collected at each site, and field blanks were collected on each fifth

101

day.

AC C

EP

TE D

M AN U

89

102 103

2.2. Acquisition of PM2.5 mass, back trajectories, and meteorology data

5

ACCEPTED MANUSCRIPT

Hourly measurements PM2.5 mass, measured by a beta attenuation monitor at the urban site,

105

were downloaded from United States Environmental Protection Agency Air Quality System (AQS) Data

106

Mart (USEPA) and averaged to correspond to the filter sampling time. PM2.5 mass measurements were

107

not available for 13-17 October and 28 October – 3 November due to instrument malfunctions. The

108

backward trajectories were analyzed using National Ocean and Atmospheric Administration (NOAA)

109

Hybrid Single-Particle Integrated Trajectory (HYSPLIT) model (NOAA, 2014) with a trajectory time of 48

110

hours at a height of 500 m above ground level (Draxler and Hess, 1997, 1998; Draxler et al., 1999).

111

Meteorological data was accessed from the Iowa City Municipal Airport weather station located within 2

112

km of the two study sites.

113

2.3. Chemical speciation

M AN U

114

SC

RI PT

104

Chemical speciation followed standard methods of analysis, with detailed methodology provided in the Supporting Information. Briefly, water soluble cations (Na+, K+, NH4+, Mg2+, Ca2+) and

116

anions (Cl-, NO3-, SO42-) were quantified in aqueous extracts of filter samples using ion exchange

117

chromatography with suppressed conductivity detection. EC and OC were measured by thermal-optical

118

analysis (Sunset Laboratory Inc.) following the ACE-Asia base case protocol (Schauer et al., 2003). The

119

measurement of organic species required, in most cases, the combination of 2-4 filter samples from

120

consecutive days to achieve instrumental detection limits. Samples collected on Saturdays during

121

University of Iowa home football games (including 3, 17, 24 September; 15 and 22 October; and 5

122

November) were analyzed individually. Filters were solvent-extracted into methanol and

123

dichloromethane and analyzed by gas chromatography mass spectrometry, before and after silylation

124

derivatization. All measurements were field-blank subtracted and were subject to rigorous quality

125

control.

AC C

EP

TE D

115

126 127

2.4. Source apportionment by chemical mass balance modeling

6

ACCEPTED MANUSCRIPT

128

Source apportionment of OC was conducted using the EPA chemical mass balance (CMB) model (version 8.2) (EPA, 2004), which determines the least-squares solution to the linear combination of

130

source profiles contributing to receptor-based PM2.5 measurements by solving for source contributions.

131

Input source profiles included vegetative detritus (Rogge et al., 1993b), natural gas (Rogge et al., 1993c),

132

diesel engines (Lough et al., 2007), non-catalyzed gasoline engines (Lough et al., 2007), biomass burning

133

using an averaged profile for EPA region 5 (Fine et al., 2004a; Sheesley et al., 2007), and coal combustion

134

(Zhang et al., 2008). Fitting species included EC, n-alkanes, levoglucosan, 17β(H)-21α(H)-norhopane,

135

17α(H)-21β(H)-hopane, benzo[k]fluoranthene, indeno(1,2,3-cd)pyrene, benzo(ghi)perylene, and picene.

136

The catalyzed and non-catalyzed gasoline engine profiles from Lough et al. (2007) were co-linear and

137

thus these two source categories could not be resolved. The non-catalyzed gasoline engine was selected

138

for use because prior source apportionment studies in the Midwestern US have shown that non-

139

catalyzed gasoline engines contributions to OC significantly outweigh those from catalyzed gasoline

140

engines (Snyder et al., 2010; Stone et al., 2009). Typical model performance included R2 values > 0.9 and

141

χ2 < 7. Lower model performance (i.e. lower R2 values and higher χ2) were observed for 15 and 22

142

October indicative of a less robust fit of profiles to ambient data and missing OC sources; consequently,

143

CMB results on these days are not reported.

146

SC

M AN U

TE D

EP

145

2.5. Statistical analysis

AC C

144

RI PT

129

Spatial differences in ambient aerosol composition across urban and peri-urban sites were assessed

147

using Minitab software. First, data were log-transformed to achieve normal distributions (Anderson-

148

Darling, p > 0.05). Differences between the two sites were examined by paired t-test (by date) and were

149

considered significant if p < 0.05, corresponding to the 95% confidence interval.

150 151

3. Results and discussion 7

ACCEPTED MANUSCRIPT

152

3.1. PM2.5 composition From 25 August to 10 November 2011, 24 h average PM2.5 mass concentrations at the urban (FRM)

154

site ranged from 3 - 26 µg m-3 (24 h). The maximum PM2.5 concentration occurred on 31 August at 25.6

155

µg m-3 and just exceeded the World Health Organization guideline of 25 µg m-3 designed to protect

156

human health, but not the United States EPA National Ambient Air Quality Standard of 35 µg m-3. The

157

absolute concentrations of PM2.5 mass, OC, EC, and inorganic ions at the urban site are shown in Figure

158

1a. Absolute concentrations of these PM components are reported in Figure S1 for the peri-urban site.

159

In general, days with the highest PM2.5 concentrations in August through October were caused by

160

elevated levels of sulfate, ammonium, and OC, which coincided with high temperatures (Figure 2f), high

161

relative humidity, and southerly winds.

SC

M AN U

162

RI PT

153

The largest component PM2.5 was carbonaceous aerosol as shown in Figure 1b. OC accounted for an average (± standard error) of 22 ± 6 % of PM2.5 (Figure 1b). Organic matter (which includes auxiliary

164

elements such as hydrogen, carbon, and oxygen atoms) was estimated as 1.8 times the OC

165

concentration, approximately 39 ± 11 % of PM2.5, and was the largest contributor to PM2.5 mass. The

166

selected OC-to-OM conversion factor is an intermediate value between those recommended for urban

167

(1.6 ± 0.2) and remote (2.1 ± 0.2) locations (Turpin and Lim, 2001). On average, PM2.5 was comprised of

168

14 ± 6 % ammonium, 13 ± 7 % sulfate, and 7 ± 6 % nitrate. These secondary ions were dominated by

169

ammonium sulfate in August through October with increasing contributions from ammonium nitrate

170

towards November. Water-soluble ions with minor contributions to PM2.5 mass included calcium (2.9%),

171

magnesium (0.3%), potassium (0.4%), and sodium (0.2%). PM mass that was not attributed to OM, EC,

172

or the measured inorganic ions is termed “other mass” and averaged 31% and corresponded to aerosol

173

components that were not measured. Suspended soil and dust (comprised largely of silica, alumina, and

174

iron oxide) were estimated to account for approximately 15% of PM2.5 based on calcium concentrations

175

and other source apportionment studies in Iowa (Kundu and Stone, 2014). Particle-phase liquid water is

AC C

EP

TE D

163

8

ACCEPTED MANUSCRIPT

176

likewise expected to be an important contributor to PM2.5 mass due to the relatively large proportion of

177

hygroscopic salts (e.g. ammonium sulfate and ammonium nitrate) (Carlton and Turpin, 2013; Hodas et

178

al., 2014). Differences in PM composition across the urban scale were assessed by comparing measurements

RI PT

179

across the two sites in two ways. First, time series measurements were juxtaposed (Figure 2) and

181

changes to PM greater than the analytical uncertainty (shown with error bars) across the two sites

182

indicated local influence that caused PM composition to shift at only one site. Second, sustained

183

enhancements of a species was examined using a t-test in which ambient measurements were log-

184

normalized and paired by date to examine statistical differences at the 95% confidence interval (Table

185

1).

M AN U

186

SC

180

Regional transport has been established as the dominant contributor to ambient PM2.5 burdens in the Midwestern US, particularly with regards OC, ammonium, and sulfate, and is responsible for region-

188

wide PM nucleation events (Allen and Turner, 2008).Bulk measurements of OC were not significantly

189

different across the two sites, but its sources varied as discussed in the context of organic molecules in

190

sections 3.2 and 3.4. Similarly, sulfate (Figure 2b) and ammonium (Figure 2c) were not significantly

191

different across the two sites, indicating that ammonium sulfate formation is primarily influenced by

192

regional atmospheric processes, rather than by local activities. Source regions impacting Iowa City were

193

identified using air mass backward trajectories with the NOAA HYSPLIT model. Five general trajectories

194

explained the majority of sampling days and were categorized as northwesterly (18%), northerly (15%),

195

northeasterly (17%), southeasterly (13%), and southwesterly (13%), with representative trajectories are

196

shown in Figure S2 and summarized in Table S1. Air masses transported to Iowa City from the

197

southeasterly and southwesterly regions, in particular, were responsible for heightened levels of OC,

198

ammonium, and sulfate (Table S2), indicating long-range transport from these regions leads to higher

199

PM levels in the Upper Midwest.

AC C

EP

TE D

187

9

ACCEPTED MANUSCRIPT

Nitrate concentrations (Figure 2d) increased into November, when colder temperatures prevailed

201

that promoted the partitioning of nitric acid and ammonium nitrate to the particle phase. Winter nitrate

202

episodes have been previously documented in the Midwest, sometimes causing exceedances of PM2.5

203

NAAQS, and coincided with meteorological stagnation, low pressure systems, and increased

204

temperature and relative humidity relative to typical winter conditions (Katzman et al., 2010; Stanier et

205

al., 2012). Local source influences on nitrate concentrations in Iowa City are apparent in the end of

206

October and November, when nitrate concentrations significantly increased at the urban site relative to

207

the peri-urban site (Figure 2d, Table 1, p = 0.044). The urban excess in nitrate is driven by local

208

combustion activities, particularly biomass burning and food cooking (as discussed in section 3.2).

SC

M AN U

209

RI PT

200

Ambient calcium concentrations ranged from 0.01-1.50 µg m-3 across the two study sites (Figure 2e) with local maxima occurring during dry periods. Maximum calcium concentrations occurred during the

211

month of October when harvesting of row crops occurs in Iowa and conditions were dry for consecutive

212

days (Figure 2f). Calcium concentrations at the peri-urban site generally exceeded those at the urban

213

site and were an average of 19.4 % higher. The peri-urban enhancement in calcium is likely due to the

214

close proximity of agricultural lands and unpaved roadways. The spatial difference in calcium is notable,

215

because calcium-bearing minerals (e.g. calcium carbonate) will add alkalinity to aerosol and neutralize

216

strong acids (Brahney et al., 2013). Thus, it is expected that aerosol alkalinity is greater at the peri-urban

217

site, while aerosol acidity is greater at the urban site.

219 220

EP

AC C

218

TE D

210

3.2. Molecular markers of primary sources 3.2.1. Polycyclic aromatic hydrocarbons. Up to 17 PAHs, which are general markers of the

221

combustion of carbonaceous material, were quantified in individual and composite filter samples (Figure

222

S3). The quantified compounds have four to six aromatic rings and their concentrations at the urban

223

site ranged from 0.07-5.0 ng m-3 and averaged 0.76 ± 0.96. PAH levels were generally lower at the peri-

10

ACCEPTED MANUSCRIPT

urban site, where their concentrations ranged from 0.082-0.90 ng m-3 and averaged 0.36 ± 0.18 ng m-3.

225

The observed background PAH levels were comparable to prior observations in Iowa City (Downard et

226

al., 2015) and background locations in the Great Lakes region (1.4 ng m-3), but are well below levels

227

observed in urban and industrial locations in the Midwest (Sun et al., 2006).

228

RI PT

224

Maximum PAH concentrations occurred from 29 September - 3 October and 20-22 October at both sites, with significantly higher concentrations observed in the urban area. These time periods

230

corresponded to elevated levels of both biomass burning PM (see section 3.2.3) and a simultaneous

231

increase in motor vehicle PM (see section 3.2.2) brought on by lower temperatures and meteorological

232

stagnation. In addition, the concentrations of benzo[e]pyrene and benzo[a]pyrene were significantly

233

greater at the urban site relative to the peri-urban site (p < 0.001 for both compounds, Table 1, n = 29),

234

indicating an increase in urban population exposures to carcinogenic PM constituents. 3.2.2.

M AN U

235

SC

229

Hopanes. Ambient concentrations of a homologous series of hopanes, which are

ubiquitous in fossil fuels, are shown in Figure S4. The sum of 17α(H)- 21β(H)-hopane, 17β(H)-21α(H)-30-

237

norhopane, and 17α(H)-22,29,30-trisnorhopane concentrations were consistently < 0.2 ng m-3, save for

238

local maxima at each site. The absolute hopane levels are comparable in magnitude to prior

239

observations in urban areas of the Midwestern US (Stone et al., 2009). The relative abundances of

240

17α(H)- 21β(H)-hopane, 17β(H)-21α(H)-30-norhopane, and 17α(H)-22,29,30-trisnorhopane were 1: 0.75:

241

0.30 at the urban site and 1:0.75:0.30 at the peri-urban site, indicating a consistent source type across

242

the two sites. The observed relative ratios are similar in magnitude to those for diesel engines at 1:

243

0.62: 0.35 (Rogge et al., 1993a) and significantly different from coal combustion at 1: 1.9: 1.5 (Oros and

244

Simoneit, 2000a), supporting their origin from motor vehicles. In particular, hopanes are components of

245

lubricant oil used by both diesel and gasoline vehicles and in the United States are markers of engine

246

emissions rather than tail pipe emissions (Schauer et al., 1999a). Spatial differences across the urban

247

and peri-urban sites were assessed using the two most abundant hopanes and both showed significant

AC C

EP

TE D

236

11

ACCEPTED MANUSCRIPT

248

enhancements in the urban area (Table 1), suggesting an overall greater fossil fuel influence on PM2.5 OC

249

in the urban area, which is discussed in the context of natural gas, coal combustion, diesel and gasoline

250

engines in section 3.3.

251

3.2.3.

RI PT

Levoglucosan. Levoglucosan, a marker of biomass burning emissions (Simoneit et al.,

1999a), was significantly higher in the urban site (Figure S5, Table 1). The average levoglucosan

253

concentrations at the urban and peri-urban sites were 73 and 41 ng m-3, respectively, falling on the

254

upper end of summer and winter concentrations reported for 10 sites spanning urban, suburban, and

255

rural locations in the Midwestern US (Sullivan et al., 2011). With levels above previously reported values

256

for the region, wood burning is identified as an important local source of PM in Iowa City.

M AN U

257

SC

252

The urban excess in biomass burning becomes particularly apparent in October and November, when temperatures decreased (Figure 1f) and spurred the use of wood fireplaces for home heating.

259

Local maxima in levoglucosan concentrations are observed on days with local festivities (17, 24

260

September, 15, 22 October, and 5 November) that involved more than 80,000 people, slow-moving

261

traffic, and outdoor food cooking, particularly grilling. On October 22, levoglucosan concentrations

262

reached 359 ng m-3 at the urban site and 88 ng m-3 at the peri-urban site, coinciding with an urban

263

excess in OC of 2.64 µg m-3 and an overall PM2.5 concentration of 16.0 µg m-3. With biomass burning

264

within the local area having significant influences on ambient PM, this source presents an opportunity

265

for PM reductions at the local level.

EP

AC C

266

TE D

258

3.2.4.

Cholesterol. Cholesterol is an indicator of food cooking and indicates the presence of

267

this source at both sites (Figure S6). Daily concentrations of cholesterol were less than 2.0 ng m-3 and

268

averaged < 0.5 ng m-3, indicating a presence of food cooking as a source of ambient PM, but to a lesser

269

extent than heavily urbanized areas such as Los Angeles (Fine et al., 2004b). Cholesterol exhibited a

270

significant urban excess (Table 1) and its concentrations were not correlated (r = 0.06) across the two

271

sites, indicating that local food cooking impacted the two sites to different extents. Similar to

12

ACCEPTED MANUSCRIPT

levoglucosan, local maxima at the urban site corresponded to dates with heightened local activities

273

including outdoor food cooking. In the absence of other tracers (e.g. trigycerides or fatty acids),

274

cholesterol cannot be used to identify the food or style of cooking; hence, cholesterol levels cannot be

275

used quantitatively to assess the extent of this source on ambient organic carbon levels because the

276

style of cooking and foodstuffs were not known. Nonetheless, cholesterol levels indicate the presence

277

of food cooking emissions in PM2.5 with significant contributions from local sources.

RI PT

272

280

3.3. Source apportionment of organic carbon to primary sources.

Chemical mass balance modeling resolved contributions of from vegetative detritus, biomass

M AN U

279

SC

278

burning, diesel engines, gasoline engines, and coal combustion to PM2.5 organic carbon. Daily

282

contributions of these sources to organic carbon are shown in Figure 3 and mean values are

283

summarized in Table 2. Biomass burning contributed 0.06 – 1.25 µgC m-3 at the urban site and 0.08 –

284

0.52 µgC m-3 at the peri-urban site, with maximum concentrations occurring 29 September – 3 October

285

(along with levoglucosan and PAHs). The drop in temperatures during this time period triggered wood

286

burning for home heating purposes, and impacted the urban area to a larger extent. Even with elevated

287

local wood burning, 24h PM2.5 concentrations at the urban site remained below 13 µg m-3. Nonetheless,

288

control of biomass burning within the urban area would lead to reductions in primary PM2.5 and co-

289

pollutants (i.e. NOx, CO, VOC, etc.).

EP

AC C

290

TE D

281

Gasoline engines were estimated to account for an average of 0.14 ± 0.04 µgC m-3 at the urban

291

site and 0.08 ± 0.03 µgC m-3 at the peri-urban site. The urban increment is likely associated with poorly-

292

functioning gasoline-powered motor vehicles and off-road engine use (e.g. lawnmowers). Diesel engines

293

accounted for an average of 0.08 ± 0.01 µgC m-3 and 0.10 ± 0.02 µgC m-3 at the urban and peri-urban

294

sites, respectively. On average, the combined gasoline and diesel engine contributions to OC were

295

nearly equivalent across the two sites. 13

ACCEPTED MANUSCRIPT

The coal combustion source was driven by the tracer picene (Oros and Simoneit, 2000b) and

297

was statistically significant in 85% of sample periods at the urban site and 59% of sampling periods at

298

the peri-urban site. The maximum contribution at the east site reached 0.14 µgC m-3 on one occasion

299

and otherwise did not exceed 0.10 µgC m-3. These results demonstrate that coal combustion

300

contributes negligibly to ambient PM2.5 OC within the urban area. Natural gas is also used as a fuel

301

source in local power plants, but its contributions to PM2.5 were not statistically significant and were

302

thus deemed negligible.

SC

303

RI PT

296

The presence of vegetative detritus was indicated by an odd-carbon preference of n-alkanes. The contribution of this source to OC was relatively minor at 0.1 µgC m-3, averaging approximately 5% of

305

OC at the two sites. n-Alkanes could not be quantified in all samples due to solvent interferences, and

306

thus vegetative detritus contributions to OC were estimated from a limited number of samples.

307

Consequently, spatial variability in this source could not be assessed.

With the aforementioned primary OC sources contributing to an average of 28% of OC at the

TE D

308

M AN U

304

urban site and 34% at the peri-urban site, there remains a large role for other sources. Additional

310

primary sources of OC are known to be important to the region, but were not included in the model.

311

Apportionment of organic carbon to dust and soil requires measurement of Al and/or Si to constrain this

312

source in the CMB model (Rutter et al., 2011). Likewise, food cooking contributions to OC could not be

313

apportioned, which results from the large degree of variability with different foods, oils, and cooking

314

techniques (as discussed in section 3.2.4), such that the cooking profiles available in the literature

315

(McDonald et al., 2003; Rogge et al., 1991; Schauer et al., 1999b) gave unrealistic model results.

316

Additionally, SOA is known to be an important source of OC in urban and background sites in the

317

Midwestern US (Lewandowski et al., 2008; Stone et al., 2009) that is discussed in section 3.4.

318 319

AC C

EP

309

Comparison to prior source apportionment studies in the Midwestern US reveals several similarities with respect to sources of OC in Iowa City. First, vegetative detritus makes a minor

14

ACCEPTED MANUSCRIPT

contribution (< 0.1 ugC m-3), but one that is consistent throughout the annual cycle (Lewandowski et al.,

321

2008; Snyder et al., 2010). Second, mobile sources represent a significant source of OC that is relatively

322

consistent throughout the year (Lewandowski et al., 2008; Snyder et al., 2010). Importantly, the diesel

323

engine contributions in Iowa City are similar in absolute and relative contributions to prior studies in

324

Cleveland, OH and Detroit, MI, while gasoline engine contributions are significantly lower in this regard

325

(Stone et al., 2009). Overall, the mobile source contributions to OC in Iowa City are similar to a rural site

326

in the Midwest (Bondville, IL) site and are lower than those reported for larger urban areas in the

327

Midwest (e.g. Cincinnati, OH, East St. Louis, IL, Detroit, MI) (Lewandowski et al., 2008; Snyder et al.,

328

2010; Stone et al., 2009). Third, biomass burning is more spatially and temporally variable, while average

329

percent contributions to OC are near the middle of the range of 5-35% previously observed across the

330

Midwest by Sullivan et al. (2011). Finally, primary sources show substantial variation across the urban

331

and peri-urban sites (following molecular markers discussed in section 3.1), which has also been

332

reported in Detroit and Cleveland (Snyder et al., 2010; Stone et al., 2009) and is attributed to the

333

influences of local sources.

334

3.4. Secondary organic aerosol tracers

SC

M AN U

TE D

Tracers for biogenic SOA derived from isoprene and monoterpenes were detected in the majority

EP

335

RI PT

320

of PM2.5 samples in this study, while those for 2-methyl-3-butene-2-ol (MBO) and toluene/aromatics

337

were below our method detection limits. This finding is consistent with prior studies of biogenic SOA

338

that have documented higher atmospheric abundances of isoprene and monoterpene SOA tracers in

339

relation to these other precursor gases (Lewandowski et al., 2013; Rutter et al., 2014). Consequently,

340

the following discussion is limited to isoprene and monoterpene-derived SOA and does not provide a

341

complete analysis of all known SOA sources in the Midwest region.

342 343

AC C

336

3.4.1.

Isoprene. Ambient concentrations of three isoprene SOA tracers—2-methylglyceric acid,

2-methyltetrol, and 2-methylerythritol—are shown in Figure 4a-b. Average concentrations of isoprene

15

ACCEPTED MANUSCRIPT

tracers in September and October were comparable to prior studies in the Upper Midwest

345

(Lewandowski et al., 2008; Stone et al., 2009), but are significantly lower than those observed in the

346

Southeastern US (Kleindienst et al., 2007), which is known to be a greater isoprene source region.

347

Generally, the isoprene SOA tracer concentrations were highest in August and decreased as

348

temperatures became cooler. The observed seasonal pattern in Iowa City is consistent with prior

349

studies in the Midwestern US, when isoprene SOA tracers are elevated June through September, decline

350

into October, and reach a minimum in winter (Kleindienst et al., 2007).

SC

351

RI PT

344

Local maxima in isoprene SOA tracers on 29 August – 3 September and 6 – 9 October corresponded to periods with elevated temperatures, southerly winds and high concentrations of

353

ammonium sulfate and OC. The concentrations of these three isoprene tracers were significantly greater

354

at the urban site relative to the peri-urban site (p < 0.003), indicating local influences on SOA tracer

355

levels. As isoprene SOA formation is enhanced by aerosol acidity (Surratt et al., 2007), greater acidity at

356

the urban site (due to significantly lower calcium levels relative to the peri-urban site as discussed

357

section 3.1) may be a driving cause of these spatial differences in isoprene SOA tracer levels. Notably,

358

the ratio of the high-NOx isoprene tracer (2-methylglyceric acid) to low-NOx isoprene tracers (2-

359

methyltetrols) was higher at the urban site compared to the peri-urban site (Figure 4c). During the

360

period of 25 August through 23 September the urban excess was significant (p = 0.009), indicating the

361

greater influence of NOx on SOA formed at the urban site. Higher ratios of 2-methylglyceric acid to 2-

362

methyltetrols have been previously reported for urban areas relative to rural locations (Lewandowski et

363

al., 2013), while this study indicates that these ratios also vary significantly across an urban area.

TE D

EP

AC C

364

M AN U

352

The SOA tracer method developed by Kleindienst et al. (2007) was used to estimating isoprene

365

contributions to secondary OC. To this end, ambient concentrations of the three isoprene tracers were

366

summed and converted to secondary OC using the tracer-to-OC ratio of 0.155 ± 0.039 (Kleindienst et al.,

367

2007). The following estimates of isoprene contributions to OC are subject to substantial uncertainty

16

ACCEPTED MANUSCRIPT

that derives from differences between the chamber experiments used to develop the tracer-to-OC ratios

369

and the ambient environment, and differences in the quantitation method for SOA tracers between this

370

study and Kleindienst et al. (2007). Because the urban and peri-urban samples were collected and

371

analyzed by identical methods, the spatial comparison between the two sites remains valid. Estimated

372

isoprene contributions to OC at the urban site ranged from < 20 – 720 ngC m-3 and averaged 125 ngC m-

373

3

374

ngC m-3 at the peri-urban site. On average, isoprene SOA contributed 4.8% of OC at the urban site and

375

3.8% at the peri-urban site. Isoprene SOA is an important source of carbonaceous aerosol within the

376

study domain, with local influences impacting the urban site to a greater extent than the peri-urban site.

SC

. At the peri-urban site, isoprene contributions to OC ranged from < 20 – 240 ngC m-3 and averaged 54

3.4.2.

M AN U

377

RI PT

368

Monoterpenes. A series of monoterpene SOA tracers were quantified, with ambient

concentrations shown in Figure 5. Unlike isoprene SOA tracers that peaked in August and decreased in

379

cooler months, monoterpene SOA tracers remained elevated into the autumn. This temporal trend is

380

consistent with prior studies of monoterpene SOA tracers that have demonstrated maximum

381

concentrations occur in June and July and levels remain elevated through into November (Lewandowski

382

et al., 2008; Rutter et al., 2014). Consequently, monoterpene SOA remains a significant source of OC into

383

autumn, while isoprene SOA contributions sharply decline after summer.

EP

384

TE D

378

The spatial variation of the two most frequently detected monoterpene SOA tracers—3hydroxyglutaric acid and 2-hydroxy-4,4-dimethylglutaric acid—indicated that monoterpene SOA tracer

386

levels were significantly enhanced at the urban site relative to the peri-urban site. The urban

387

enhancement in monoterpene SOA may result from a combination of higher SOA precursors in the

388

urban area (e.g. NOx) or the greater prevalence of biomass burning in the urban area. The spatial

389

difference in monoterpene SOA is unlikely to be affected by differences in aerosol acidity, as α-pinene

390

SOA formation is less affected by aerosol acidity compared to isoprene (Offenberg et al., 2009). The sum

391

of the observed monoterpene tracers was converted to an estimate of secondary OC using the tracer-to-

AC C

385

17

ACCEPTED MANUSCRIPT

392

OC ratio of 0.168 ± 0.081. At the urban site, monoterpene contributions to OC mass ranged from 0.82-

393

30.1 % while averaging 8.5%. At the peri-urban site, the percent contribution of monoterpene SOA to

394

OC mass ranged 0.9-4.5 %.

396 397

RI PT

395 4. Conclusions

Source apportionment by chemical mass balance (CMB) indicated important roles for biomass burning accounting for 13.0-18.4% of OC across the two sites, with decreasing roles of vegetative

399

detritus (4.5-5.2%), diesel engines (3.8-5.8%), gasoline engines (3.5-4.8%), coal combustion (0.4-1.1%),

400

and natural gas combustion (negligible). The SOA tracer method assigned an average of 4.8% and 8.5%

401

of OC to isoprene and monoterpene SOA at the urban site and 2.4% and 3.8% at the peri-urban site,

402

respectively. Combining primary source biogenic SOA contributions, the majority of the observed OC (>

403

60%) was not assigned, pointing towards additional sources (e.g., food cooking, SOA from

404

toluene/aromatics and MBO). Molecular markers of biomass burning and food cooking reveal a large

405

degree of heterogeneity across the urban area. Urban combustion sources are also responsible for the

406

urban excess in nitrate and a shift in the isoprene SOA product distribution towards the high-NOx

407

isoprene SOA tracer 2-methylglyceric acid. Accordingly, there is a high degree of variability in

408

population exposures to PM from these local sources in outdoor environments, particularly during

409

periods of heightened local activities.

M AN U

TE D

EP

AC C

410

SC

398

411

Acknowledgements

412

We thank undergraduate researchers Sean Staudt and Kevin Frey for sample collection, Lucas Saunders

413

compiling meteorological data, and Tony Nguyen for assistance with sample preparation. We also thank

414

Prof. Keri Hornbuckle at the University of Iowa for establishing the University of Iowa Air Monitoring Site 18

ACCEPTED MANUSCRIPT

(peri-urban) and the Iowa DNR for access to Hoover Elementary Site (urban). We acknowledge the

416

Center for Global and Regional Environmental Research (CGRER) and the University of Iowa for funding.

417

Supporting Information: Ambient concentrations of bulk PM constituents at the peri-urban site (Figure

418

S1). Representative air mass back trajectories (Figure S2), summary of trajectory categories by date, and

419

average PM2.5 OC, sulfate, and ammonium concentrations for each trajectory category. Ambient

420

concentrations of molecular markers of primary aerosol sources measured at peri-urban and urban

421

sites: PAHs (Figure S3), hopanes (Figure S4), levoglucosan (Figure S5), cholesterol (Figure S6).

AC C

EP

TE D

M AN U

SC

RI PT

415

19

ACCEPTED MANUSCRIPT

Figure Captions

423 424 425

Figure 1: (a) Daily PM2.5 mass concentrations and (b) average PM2.5 mass composition at the urban site in Iowa City, IA from 25 August to 10 November 2011. Other mass includes components not measured in this study (e.g. soil/dust and liquid water).

426 427 428 429

Figure 2: Ambient concentrations of select PM2.5 components at the urban and peri-urban sites: organic carbon (a), sulfate (b), ammonium (c), nitrate (d), and calcium (e) juxtaposed with daily minimum and maximum temperatures and cumulative rainfall (f). Error bars on concentration data represent propagated analytical uncertainties.

430 431 432

Figure 3: Source contributions to organic carbon estimated by CMB modeling. Other sources correspond to primary sources not included in the model (e.g. suspended soil organic matter, food cooking, unidentified sources) and SOA.

433 434 435 436 437 438

Figure 4: Ambient concentrations of isoprene SOA tracers (left axis) and isoprene contributions to organic carbon estimated by the SOA-tracer method (right axis) at the peri-urban (a) and urban (b) sites. Samples in which isoprene SOA tracers were below detection limits are marked as non-detects (ND). Ratios of 2-methylglyceric acid (the high-NOx tracer) to 2-methyltetrols (low-NOx tracer) shown in (c) demonstrate the greater influence of the high-NOx isoprene oxidation pathway at the urban site in August and September.

439 440 441 442

Figure 5: Ambient concentrations of monoterpene SOA tracers (left axis) and monoterpene contributions to organic carbon estimated by the SOA-tracer method (right axis) at the peri-urban (a) and urban (b) sites. Samples in which monoterpene SOA tracers were below detection limits are marked as non-detects (ND).

AC C

EP

TE D

M AN U

SC

RI PT

422

20

ACCEPTED MANUSCRIPT

Table 1: Summary of ambient concentrations of PM2.5 organic and elemental carbon, inorganic ions, and organic species at the peri-urban and urban sites. Paired t-tests were used to examine spatial differences between the two sites using log-transformed ambient concentrations. Statistically significant differences have a p-value < 0.05, corresponding to the 95% confidence interval, and are marked with a star (*). Statistical analysis was not performed (NP) when concentration data were not normally distributed prior to or post log transformation. Peri-urban site Range Mean ± St Dev

Range

Urban site Mean ± St Dev

Statistic p-value

Organic carbon

(µg m-3)

0.58 - 5.33

2.12 ± 1.04

0.66 - 6.18

2.31 ± 1.20

0.354

Elemental carbon

(µg m-3)

0.05 - 1.46

0.29 ± 0.22

0.01 - 0.54

0.22 ± 0.10

0.076

Potassium

(µg m-3)

0.01 - 0.11

0.04 ± 0.02

SC

PM2.5 Component

RI PT

443 444 445 446 447 448

0.04 ± 0.03

0.162

Ammonium

(µg m-3)

0.08 - 3.63

Sulfate

(µg m-3)

0.23 - 6.22

Nitrate

(µg m-3)

0.05 - 3.90

Calcium

(µg m-3)

0.01 - 1.50

-3

Pyrene

(pg m ) -3

Benzo[e]pyrene

(pg m ) -3

(pg m )

17β(H)-21α(H)-30-Norhopane 17α(H)- 21β(H)-Hopane

(pg m )

0.838

1.57 ± 1.31

0.15 - 6.32

1.58 ± 1.39

0.893

0.57 ± 0.57

0.10 - 4.47

0.72 ± 0.66

0.044*

0.37 ± 0.35

0.01 - 1.41

0.29 ± 0.28

NP

24 ± 10

6 - 192

38 ± 36

0.005*

7 - 62

31 ± 15

9 - 553

73 ± 107

< 0.001*

< 7 - 44

17 ± 9

< 7 - 340

43 ± 67

< 0.001*

< 7 - 296

28 ± 53

11 - 97

33 ± 19

0.040*

< 7 - 222

29 ± 39

18 - 99

40 ± 20

0.022*

14 - 88

41 ± 18

73 ± 69

0.004*

-3

(pg m ) (ng m )

AC C

SOA tracers

0.72 ± 0.63

-3

Levoglucosan Cholesterol

-3

0.08 - 3.20

6 - 47

EP

Benzo[a]pyrene

0.75 ± 0.66

TE D

Molecular markers of primary sources

0.01 - 0.14

M AN U

Inorganic ions

8 - 359

-3

(ng m )

< 0.07 - 1.78

0.31 ± 0.41

0.05 - 1.97

0.47 ± 0.39

0.010*

-3

< 1 - 10.3

3.2 ± 2.6

< 0.1 - 26.3

8.6 ± 7.0

< 0.001*

-3

< 1 - 10.5

1.6 ± 2.5

< 0.1 - 26.5

4.0 ± 6.5

0.003*

-3

< 1 - 18.0

3.1 ± 4.2

< 0.1 - 59.4

7.4 ± 11.9

< 0.001*

-3

< 2 - 19.7

7.2 ± 5.5

4.22 - 60.2

17.3 ± 15.3

< 0.001*

-3

< 0.2 - 11.7

4.0 ± 3.4

< 2 - 27.1

8.6 ± 7.1

< 0.001*

2-Methylglyceric acid

(ng m )

2-Methylthreitol

(ng m )

2-Methylerythritol

(ng m )

3-Hydroxyglutaric acid

(ng m )

2-Hydroxy-4,4-dimethylglutaric acid

(ng m )

21

ACCEPTED MANUSCRIPT

Table 2: Absolute and percent contributions of primary source categories to organic carbon at the periurban and urban sites determined by CMB modeling. Other sources include food cooking, soil organic matter, unidentified primary sources, and SOA. Reported errors represent 95% confidence intervals (n = 27). Peri-urban site (µ µgC m-3) 0.09 ± 0.03 0.27 ± 0.05 0.10 ± 0.02 0.08 ± 0.03 0.007 ± 0.003 negligible 1.77 ± 0.38

Contribution (%) 4.5 ± 1.7 13.8 ± 2.9 4.8 ± 0.7 3.8 ± 1.7 0.4 ± 0.2 negligible 75.8 ± 4.1

(µ µgC m-3)

0.11 ± 0.01 0.38 ± 0.11 0.08 ± 0.01 0.14 ± 0.04 0.023 ± 0.012 negligible 1.79 ± 0.42

AC C

EP

TE D

M AN U

Vegetative detritus Biomass burning Diesel engines Gasoline engines Coal combustion Natural gas Other 453

Urban site

RI PT

Source Category

SC

449 450 451 452

22

Contribution (%) 5.2 ± 0.9 18.4 ± 5.3 3.5 ± 0.4 5.8 ± 1.5 1.1 ± 0.6 negligible 70.6 ± 5.6

ACCEPTED MANUSCRIPT

Works Cited

455 456 457

Allen, D.T., Turner, J.R., 2008. Transport of atmospheric fine particulate matter: Part 1 - Findings from recent field programs on the extent of regional transport within North America. Journal of the Air & Waste Management Association 58, 254-264.

458 459 460

Bell, M.L., Ebisu, K., Peng, R.D., 2011. Community-level spatial heterogeneity of chemical constituent levels of fine particulates and implications for epidemiological research. Journal of Exposure Science and Environmental Epidemiology 21, 372-384.

461 462 463

Bell, M.L., Ebisu, K., Peng, R.D., Walker, J., Samet, J.M., Zeger, S.L., Dominici, F., 2008. Seasonal and Regional Short-term Effects of Fine Particles on Hospital Admissions in 202 US Counties, 1999-2005. American Journal of Epidemiology 168, 1301-1310.

464 465

Brahney, J., Ballantyne, A.P., Sievers, C., Neff, J.C., 2013. Increasing Ca2+ deposition in the western US: The role of mineral aerosols. Aeolian Research 10, 77-87.

466 467

Carlton, A.G., Turpin, B.J., 2013. Particle partitioning potential of organic compounds is highest in the Eastern US and driven by anthropogenic water. Atmospheric Chemistry and Physics 13, 10203-10214.

468 469

Chan, M.N., et al., 2011. Influence of aerosol acidity on the chemical composition of secondary organic aerosol from beta-caryophyllene. Atmospheric Chemistry and Physics 11, 1735-1751.

470 471

Claeys, M., et al., 2004. Formation of secondary organic aerosols through photooxidation of isoprene. Science 303, 1173-1176.

472 473 474

Claeys, M., et al., 2010. Polar organic marker compounds in atmospheric aerosols during the LBASMOCC 2002 biomass burning experiment in Rondonia, Brazil: sources and source processes, time series, diel variations and size distributions. Atmospheric Chemistry and Physics 10, 9319-9331.

475 476

Claeys, M., et al., 2007. Hydroxydicarboxylic acids: Markers for secondary organic aerosol from the photooxidation of alpha-pinene. Environmental Science & Technology 41, 1628-1634.

477 478

Downard, J., et al., 2015. Uncontrolled combustion of shredded tires in a landfill - Part 1: Characterization of gaseous and particulate emissions. Atmospheric Environment 104, 195-204.

479 480

Draxler, R.R., Hess, G., 1997. Description of the HYSPLIT4 modeling system. NOAA Tech. Memo. ERL ARL224, NOAA Air Resources Laboratory, Silver Spring, MD., 24.

481 482

Draxler, R.R., Hess, G., 1998. An overview of the HYSPLIT_4 modelling system for trajectories. Australian Meteorological Magazine 47, 295-308.

483 484

Draxler, R.R., Stunder, B., Rolph, G., Taylor, A., 1999. HYSPLIT4 User’s Guide. NOAA Technical Memorandum ERL ARL 230, 35.

485 486 487 488

Edney, E.O., Kleindienst, T.E., Jaoui, M., Lewandowski, M., Offenberg, J.H., Wang, W., Claeys, M., 2005. Formation of 2-methyl tetrols and 2-methylglyceric acid in secondary organic aerosol from laboratory irradiated isoprene/NOX/SO2/air mixtures and their detection in ambient PM2.5 samples collected in the eastern United States. Atmospheric Environment 39, 5281-5289.

489

EPA, U.S., 2004. EPA-CMB8.2 Users Manual.

490 491 492

Fine, P.M., Cass, G.R., Simoneit, B.R.T., 2004a. Chemical characterization of fine particle emissions from the wood stove combustion of prevalent United States tree species. Environmental Engineering Science 21, 705-721.

AC C

EP

TE D

M AN U

SC

RI PT

454

23

ACCEPTED MANUSCRIPT

Fine, P.M., Chakrabarti, B., Krudysz, M., Schauer, J.J., Sioutas, C., 2004b. Diurnal variations of individual organic compound constituents of ultrafine and accumulation mode particulate matter in the Los Angeles basin. Environmental Science & Technology 38, 1296-1304.

496 497

Hildemann, L.M., Markowski, G.R., Cass, G.R., 1991. Chemical-Composition Of Emissions From Urban Sources Of Fine Organic Aerosol. Environmental Science & Technology 25, 744-759.

498 499 500

Hodas, N., et al., 2014. Aerosol Liquid Water Driven by Anthropogenic Nitrate: Implications for Lifetimes of Water-Soluble Organic Gases and Potential for Secondary Organic Aerosol Formation. Environmental Science & Technology 48, 11127-11136.

501 502 503

Jaoui, M., Kleindienst, T.E., Lewandowski, M., Offenberg, J.H., Edney, E.O., 2005. Identification and quantification of aerosol polar oxygenated compounds bearing carboxylic or hydroxyl groups. 2. Organic tracer compounds from monoterpenes. Environmental Science & Technology 39, 5661-5673.

504 505 506

Jaoui, M., Lewandowski, M., Kleindienst, T.E., Offenberg, J.H., Edney, E.O., 2007. beta-caryophyllinic acid: An atmospheric tracer for beta-caryophyllene secondary organic aerosol. Geophysical Research Letters 34.

507 508

Jerrett, M., et al., 2013. Spatial Analysis of Air Pollution and Mortality in California. American Journal of Respiratory and Critical Care Medicine 188, 593-599.

509 510 511

Katzman, T.L., Rutter, A.P., Schauer, J.J., Lough, G.C., Kolb, C.J., Van Klooster, S., 2010. PM2.5 and PM102.5 Compositions during Wintertime Episodes of Elevated PM Concentrations across the Midwestern USA. Aerosol Air Quality Research 10, 140-U113.

512 513 514

Kleindienst, T.E., Jaoui, M., Lewandowski, M., Offenberg, J.H., Lewis, C.W., Bhave, P.V., Edney, E.O., 2007. Estimates of the contributions of biogenic and anthropogenic hydrocarbons to secondary organic aerosol at a southeastern US location. Atmospheric Environment 41, 8288-8300.

515 516 517

Kourtchev, I., et al., 2008. Determination of isoprene and alpha-/beta-pinene oxidation products in boreal forest aerosols from Hyytiala, Finland: diel variations and possible link with particle formation events. Plant Biology 10, 138-149.

518 519

Kundu, S., Stone, E.A., 2014. Composition and sources of fine particulate matter across urban and rural sites in the Midwestern United States. Environmental Science: Processes & Impacts 16, 1360-1370.

520 521 522

Lewandowski, M., Jaoui, M., Offenberg, J.H., Edney, E.O., Sheesley, R.J., Schauer, J.J., 2008. Primary and Secondary Contributions to Ambient PM in the Midwestern United States. Environmental Science & Technology 42, 3303-3309.

523 524 525 526

Lewandowski, M., Piletic, I.R., Kleindienst, T.E., Offenberg, J.H., Beaver, M.R., Jaoui, M., Docherty, K.S., Edney, E.O., 2013. Secondary organic aerosol characterisation at field sites across the United States during the spring-summer period. International Journal of Environmental Analytical Chemistry 93, 10841103.

527 528 529

Lim, S.S., et al., 2012. A comparative risk assessment of burden of disease and injury attributable to 67 risk factors and risk factor clusters in 21 regions, 1990-2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet 380, 2224-2260.

530 531 532

Lin, Y.H., Knipping, E.M., Edgerton, E.S., Shaw, S.L., Surratt, J.D., 2013. Investigating the influences of SO2 and NH3 levels on isoprene-derived secondary organic aerosol formation using conditional sampling approaches. Atmospheric Chemistry and Physics 13, 8457-8470.

AC C

EP

TE D

M AN U

SC

RI PT

493 494 495

24

ACCEPTED MANUSCRIPT

Lough, G.C., Christenson, C.C., Schauer, J.J., Tortorelli, J., Bean, E., Lawson, D., Clark, N.N., Gabele, P.A., 2007. Development of Molecular Marker Source Profiles for Emissions from On-Road Gasoline and Diesel Vehicle Fleets. Journal of the Air & Waste Management Association 57, 1190-1199.

536 537 538

McDonald, J.D., Zielinska, B., Fujita, E.M., Sagebiel, J.C., Chow, J.C., Watson, J.G., 2003. Emissions from charbroiling and grilling of chicken and beef. Journal of the Air & Waste Management Association 53, 185-194.

539

NOAA, 2014. HYSPLIT - Hybrid Single Particle Lagrangian Integrated Trajectory Model.

540 541 542

Offenberg, J.H., Lewandowski, M., Edney, E.O., Kleindienst, T.E., Jaoui, M., 2009. Influence of aerosol acidity on the formation of secondary organic aerosol from biogenic precursor hydrocarbons. Environmental Science & Technology 43, 7742-7747.

543 544

Oros, D., Simoneit, B., 2000a. Identification and emission rates of molecular tracers in coal smoke particulate matter. Fuel 79, 515-536.

545 546

Oros, D.R., Simoneit, B.R.T., 2000b. Identification and emission rates of molecular tracers in coal smoke particulate matter. Fuel 79, 515-536.

547 548

Pye, H.O.T., et al., 2013. Epoxide Pathways Improve Model Predictions of Isoprene Markers and Reveal Key Role of Acidity in Aerosol Formation. Environmental Science & Technology 47, 11056-11064.

549 550 551

Rogge, W.F., Hildemann, L.M., Mazurek, M.A., Cass, G.R., Simoneit, B.R., 1993a. Sources of fine organic aerosol. 2. Noncatalyst and catalyst-equipped automobiles and heavy-duty diesel trucks. Environmental Science & Technology 27, 636-651.

552 553 554

Rogge, W.F., Hildemann, L.M., Mazurek, M.A., Cass, G.R., Simoneit, B.R.T., 1993b. Sources Of Fine Organic Aerosol .4. Particulate Abrasion Products From Leaf Surfaces Of Urban Plants. Environmental Science & Technology 27, 2700-2711.

555 556

Rogge, W.F., Hildemann, L.M., Mazurek, M.A., Cass, G.R., Simoneit, B.R.T., 1993c. Sources Of Fine Organic Aerosol .5. Natural-Gas Home Appliances. Environmental Science & Technology 27, 2736-2744.

557 558

Rogge, W.F., Hildemann, L.M., Mazurek, M.A., Cass, G.R., Simoneit, B.R.T., 1997. Sources of fine organic aerosol .8. Boilers burning No. 2 distillate fuel oil. Environmental Science & Technology 31, 2731-2737.

559 560 561

Rogge, W.F., Hildemann, L.M., Mazurek, M.A., Cass, G.R., Simonelt, B.R.T., 1991. Sources Of Fine Organic Aerosol .1. Charbroilers And Meat Cooking Operations. Environmental Science & Technology 25, 11121125.

562 563 564

Rutter, A.P., Snyder, D.C., Schauer, J.J., Sheesley, R.J., Olson, M.R., DeMinter, J., 2011. Contributions of resuspended soil and road dust to organic carbon in fine particulate matter in the Midwestern US. Atmospheric Environment 45, 514-518.

565 566 567

Rutter, A.P., Snyder, D.C., Stone, E.A., Shelton, B., DeMinter, J., Schauer, J.J., 2014. Preliminary assessment of the anthropogenic and biogenic contributions to secondary organic aerosols at two industrial cities in the upper Midwest. Atmospheric Environment 84, 307-313.

568 569

Schauer, J.J., Cass, G.R., 2000. Source apportionment of wintertime gas-phase and particle-phase air pollutants using organic compounds as tracers. Environmental Science & Technology 34, 1821-1832.

570 571 572

Schauer, J.J., Kleeman, M.J., Cass, G.R., Simoneit, B.R., 1999a. Measurement of emissions from air pollution sources. 2. C1 through C30 organic compounds from medium duty diesel trucks. Environmental Science & Technology 33, 1578-1587.

AC C

EP

TE D

M AN U

SC

RI PT

533 534 535

25

ACCEPTED MANUSCRIPT

Schauer, J.J., Kleeman, M.J., Cass, G.R., Simoneit, B.R.T., 1999b. Measurement of emissions from air pollution sources. 1. C-1 through C-29 organic compounds from meat charbroiling. Environmental Science & Technology 33, 1566-1577.

576 577 578

Schauer, J.J., Kleeman, M.J., Cass, G.R., Simoneit, B.R.T., 1999c. Measurement of emissions from air pollution sources. 2. C-1 through C-30 organic compounds from medium duty diesel trucks. Environmental Science & Technology 33, 1578-1587.

579 580 581

Schauer, J.J., Kleeman, M.J., Cass, G.R., Simoneit, B.R.T., 2001. Measurement of emissions from air pollution sources. 3. C-1-C-29 organic compounds from fireplace combustion of wood. Environmental Science & Technology 35, 1716-1728.

582 583

Schauer, J.J., et al., 2003. ACE-Asia intercomparison of a thermal-optical method for the determination of particle-phase organic and elemental carbon. Environmental Science & Technology 37, 993-1001.

584 585 586

Schauer, J.J., Rogge, W.F., Hildemann, L.M., Mazurek, M.A., Cass, G.R., 1996. Source apportionment of airborne particulate matter using organic compounds as tracers. Atmospheric Environment 30, 38373855.

587 588 589

Shalamzari, M.S., Kahnt, A., Vermeylen, R., Kleindienst, T.E., Lewandowski, M., Cuyckens, F., Maenhaut, W., Claeys, M., 2014. Characterization of Polar Organosulfates in Secondary Organic Aerosol from the Green Leaf Volatile 3-Z-Hexenal. Environmental Science & Technology 48, 12671-12678.

590 591 592

Sheesley, R.J., Schauer, J.J., Chowdhury, Z., Cass, G.R., Simoneit, B.R.T., 2003. Characterization of organic aerosols emitted from the combustion of biomass indigenous to South Asia. Journal of Geophysical Research-Atmospheres 108.

593 594 595

Sheesley, R.J., Schauer, J.J., Zheng, M., Wang, B., 2007. Sensitivity of molecular marker-based CMB models to biomass burning source profiles. Atmospheric Environment, doi:10.1016/j.atmosenv.2007.08.011.

596 597 598

Simoneit, B.R., Schauer, J.J., Nolte, C., Oros, D.R., Elias, V.O., Fraser, M., Rogge, W., Cass, G.R., 1999a. Levoglucosan, a tracer for cellulose in biomass burning and atmospheric particles. Atmospheric Environment 33, 173-182.

599 600

Simoneit, B.R.T., 1985. Application of Molecular Marker Analysis to Vehicular Exhaust for Source Reconciliations. International Journal of Environmental Analytical Chemistry 22, 203-233.

601 602

Simoneit, B.R.T., 1999. A review of biomarker compounds as source indicators and tracers for air pollution. Environmental Science and Pollution Research 6, 159-169.

603 604 605

Simoneit, B.R.T., Schauer, J.J., Nolte, C.G., Oros, D.R., Elias, V.O., Fraser, M.P., Rogge, W.F., Cass, G.R., 1999b. Levoglucosan, a tracer for cellulose in biomass burning and atmospheric particles. Atmospheric Environment 33, 173-182.

606 607 608

Snyder, D.C., Rutter, A.P., Worley, C., Olson, M., Plourde, A., Bader, R.C., Dallmann, T., Schauer, J.J., 2010. Spatial variability of carbonaceous aerosols and associated source tracers in two cites in the Midwestern United States. Atmospheric Environment 44, 1597-1608.

609 610 611

Stanier, C., et al., 2012. Overview of the LADCO winter nitrate study: hourly ammonia, nitric acid and PM2.5 composition at an urban and rural site pair during PM2.5 episodes in the US Great Lakes region. Atmospheric Chemistry and Physics 12, 11037-11056.

612 613

Stone, E.A., Nguyen, T.T., Pradhan, B.B., Dangol, P.M., 2012. Assessment of biogenic secondary organic aerosol in the Himalayas. Environmental Chemistry 9, 263-272.

AC C

EP

TE D

M AN U

SC

RI PT

573 574 575

26

ACCEPTED MANUSCRIPT

Stone, E.A., Zhou, J.B., Snyder, D.C., Rutter, A.P., Mieritz, M., Schauer, J.J., 2009. A Comparison of Summertime Secondary Organic Aerosol Source Contributions at Contrasting Urban Locations. Environmental Science & Technology 43, 3448-3454.

617 618 619 620

Sullivan, A.P., Frank, N., Onstad, G., Simpson, C.D., Collett, J.L., Jr., 2011. Application of highperformance anion-exchange chromatography-pulsed amperometric detection for measuring carbohydrates in routine daily filter samples collected by a national network: 1. Determination of the impact of biomass burning in the upper Midwest. Journal of Geophysical Research-Atmospheres 116.

621 622

Sun, P., Blanchard, P., Brice, K.A., Hites, R.A., 2006. Trends in polycyclic aromatic hydrocarbon concentrations in the Great Lakes atmosphere. Environmental Science & Technology 40, 6221-6227.

623 624

Surratt, J.D., et al., 2008. Organosulfate formation in biogenic secondary organic aerosol. Journal of Physical Chemistry A 112, 8345-8378.

625 626 627

Surratt, J.D., Lewandowski, M., Offenberg, J.H., Jaoui, M., Kleindienst, T.E., Edney, E.O., Seinfeld, J.H., 2007. Effect of acidity on secondary organic aerosol formation from isoprene. Environmental Science & Technology 41, 5363-5369.

628 629

Surratt, J.D., et al., 2006. Chemical composition of secondary organic aerosol formed from the photooxidation of isoprene. Journal of Physical Chemistry A 110, 9665-9690.

630 631

Turpin, B.J., Lim, H.J., 2001. Species contributions to PM2.5 mass concentrations: Revisiting common assumptions for estimating organic mass. Aerosol Science and Technology 35, 602-610.

632 633

USEPA, Technology Transfer Network, National Ambient Air Quality Standards (NAAQS), http://www.epa.gov/ttn/naaqs/standards/pm/s_pm_index.html, accessed 2013.

634 635

Zhang, H., et al., 2012. Secondary organic aerosol formation from methacrolein photooxidation: roles of NOx level, relative humidity and aerosol acidity. Environmental Chemistry 9, 247-262.

636 637 638

Zhang, Y.X., Schauer, J.J., Zhang, Y.H., Zeng, L.M., Wei, Y.J., Liu, Y., Shao, M., 2008. Characteristics of particulate carbon emissions from real-world Chinese coal combustion. Environmental Science & Technology 42, 5068-5073.

SC

M AN U

TE D

EP

640

AC C

639

RI PT

614 615 616

27

ACCEPTED MANUSCRIPT

Figure 1

a) Daily PM2.5 mass and composition

RI PT

25 20

SC

15 10 5

M AN U

Concentration (g m-3)

30

TE D

Aug 24 Aug 27 Aug 30 Sep 02 Sep 05 Sep 08 Sep 11 Sep 14 Sep 17 Sep 20 Sep 23 Sep 26 Sep 29 Oct 02 Oct 05 Oct 08 Oct 11 Oct 14 Oct 17 Oct 20 Oct 23 Oct 26 Oct 29 Nov 01 Nov 04 Nov 07 Nov 10

0

AC C

EP

b) Average PM2.5 composition

Elemental carbon Organic matter (1.8*OC) Ammonium Sulfate Nitrate Calcium Magnesium Potassium Sodium Other PM2.5 mass

f) Meteorological data

100

Nov 10

Precipitation Min. temperature Max. temperature

2.5

2.0

80

1.5

60

1.0

40

0.5

20

0.0

Precipitation (in.)

ACCEPTED MANUSCRIPT

Nov 07

a) Organic carbon (OC)

Nov 01 Nov 04

4

Oct 29

RI PT

0

Oct 26

2

Oct 23

SC

4

Oct 17 Oct 20

3

Oct 14

c) Ammonium

Oct 11

d) Nitrate

Oct 08

0

M AN U

1

Oct 02 Oct 05

2

Sep 29

Sep 26

0

TE D

6

Sep 20 Sep 23

EP

2

Sep 17

e) Calcium

Sep 14

1.5

Sep 11

120

AC C

0.0

Sep 05 Sep 08

0.5

Sep 02

1.0

Aug 30

4

Aug 27

Concentration (g m-3) 0 8

Aug 24

o

Temperature ( F)

Figure 2

8

Urban Peri-urban

4

2

b) Sulfate

6

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

Highlights

EP

TE D

M AN U

SC

RI PT

Ammonium and sulfate were well-represented by a single monitoring station Nitrate and high-NOx isoprene SOA products were enhanced at the urban site Biomass burning, cooking, dust and biogenic SOA had significant local influences First reported source apportionment of organic carbon in the state of Iowa

AC C

   

ACCEPTED MANUSCRIPT

Local source impacts on primary and secondary aerosols in the Midwestern United States Thilina Jayarathne, Chathurika M. Rathnayake, Elizabeth A. Stone* Department of Chemistry, University of Iowa, Iowa City, IA 52242, United States 

RI PT

Co-first authors *Corresponding author: +1-319-384-1863; fax: +1-319-335-1270; e-mail: [email protected]

Detailed description of chemical speciation Inorganic ions by ion chromatography

SC

Water soluble cations (Na+, K+, NH4+, Mg2+, Ca2+) and anions (Cl-, NO3-, SO42-) were quantified in aqueous extracts. A sub-sample of each QFF (totaling 2.56 cm2) was extracted into 15.00 mL ultra-pure

M AN U

water (>18.2 MΩ resistivity) by rotary shaking at 125 rpm for 12 h. Extracts were passed through a 0.45 m syringe filter (Whatman, PTFE) to remove insoluble particles and filter fragments. Extractions were conducted in batches that consisted of 10-12 sample filters, a lab blank, 2-3 field blanks, and quality control samples spiked with cation and anion standards.

Inorganic ions were measured by ion chromatography with suppressed conductivity detection (Dionex ICS-5000). The instrument was equipped with an auto-sampler, 25 µL sample loop, and

TE D

conductivity detector. Cation analysis utilized a Dionex IonPacTM CS12A analytical and guard columns, CSRSTM 300 suppressor, and20 mM methanesulfonic acid mobile phase at a flow rate of 0.5 mL min-1. Anion analysis utilized a Dionex IonPacTM AS22 analytical and guard columns, ASRSTM 300 suppressor, and mobile phase of sodium carbonate (4.5 mM) and sodium bicarbonate (1.4 mM) flowing at a rate of

EP

1.2 mL min-1. The instrument was calibrated using Dionex Six Cation-II and Seven Anion standards. Instrument detection limit, method detection limit, and representative spike recovery values are reported by Jayarathne et al. (2014). Uncertainties were propagated from method detection limits and

AC C

10% of the measured value.

Elemental and organic carbon by thermal-optical analysis Elemental carbon (EC) and organic carbon (OC) were measured on a 1.0 cm2 punch of QFF by thermal-optical analysis (Sunset Laboratory Inc.) following the ACE-Asia base case protocol (Schauer et al., 2003). The average OC value in field blanks (0.090 g cm-2) was subtracted from all sample data. The uncertainty in OC was propagated from the standard deviation of field blank values (0.05 g cm-2) and

1

ACCEPTED MANUSCRIPT

10% of the measured value. EC was not detected in field blanks and the uncertainty in EC was propagated from 0.10 g cm-2, 10% of the measured value, and 10% of the pyrolyzed carbon value.

Organic speciation by gas chromatography mass spectrometry

RI PT

All glassware used was washed with soap and rinsed with tap, deionized, and ultra-pure water, then sealed with aluminum foil, baked in an oven for 5.5 h at 500 oC, and solvent rinsed prior to use. Prior to solvent extraction, filters were spiked with deuterated internal standards including pyrene-D10,

benz(a)anthracene-D12, coronene-D12, cholestane-D4, pentadecane-D32, eicosane-D42, tetracosane-D50,

SC

triacontane-D62, dotriacontane-D66, hexatriacontane-D74, cholesterol-D6 and levoglucosan-13C6. Filters were extracted with two 20 mL aliquots of dichloromethane followed by two 20 mL aliquots of methanol for 10 minutes each by ultrasonication (Branson 5510). Extracts were reduced in volume

M AN U

under high-purity nitrogen using a Turbovap LV Evaporator (Caliper Life Sciences) and Reacti-VapTM Evaporator (Thermo Scientific) to a final volume of 100 µL (or 50 µL for extracts of single filters). Samples were stored at - 20 °C until analysis.

Solvent-extractable organic compounds were analyzed by gas chromatography-mass spectrometry (GCMS, Agilent Technologies 7890A and 5975C, respectively) with a DB-5 capillary column (Agilent, 30m x 0.250mm x 0.25μm) with helium carrier gas (Praxair, 99.999%). A 3.0 μL aliquot of each sample and

TE D

standard was injected to inlet held at 300oC. The GC oven temperature was held at 65oC for 10min, and then increased at a rate of 10oC min-1 to 300oC and held for 26.5min. The MS quadrupole detector scanned m/z ranging from 50-450 Da.

For the GCMS analysis of analytes containing hydroxyl or carboxylic acid function groups (i.e. SOA

EP

tracers, levoglucosan, cholesterol) silylation derivatization was performed. A 10 μL aliquot of each sample was transferred to an amber vial that contained a glass conical insert, and was then blown down

AC C

to near dryness under a gentle stream of nitrogen (Praxair, 99.999%). Then, 10 μL of pyridine and 20 μL of derivitization agent (N,O-bis(trimethylsilyl)-trifluoroacetamide with 1% trimethylchlorosilane (Fluka Analytical, 99%) were added. Samples were then capped and allowed to react at 70 oC for three hours. Silylated samples were then directly analyzed by GCMS within 24 h. A 1 μL aliquot of each sample was injected to an inlet held at 270oC. The GC stationary and mobile phases were identical to those described previously. The oven temperature was held at 84oC for 1 min, then increased at a rate of 8 oC min-1 to 200 oC and held for 2 min, then increased at a rate of 10 oC min-1 and held at 300 oC for 15 min. The MS source utilized chemical ionization with methane (Airgas, 99.999%) as the ionizing reagent and the quadrupole scanned from 50-550 Da. 2

ACCEPTED MANUSCRIPT

Analytes were quantified using five to seven point calibration curves normalized to internal standards. Calibration validity was assessed routinely with a check standard (allowing for ±10% variation). Quality control measures included analysis of one laboratory blank, two field blanks, and one QFF spike recovery sample for every 10 PM2.5 samples. The uncertainty in organic species

RI PT

measurements was propagated from the standard deviation of field blank samples (or from the

instrument detection limit when not detected in field blanks) and a 20% of the measurement value. For SOA tracers that are not commercially available, the use of surrogate standards and uncertainty

AC C

EP

TE D

M AN U

SC

estimates followed Stone et al. (2012).

3

ACCEPTED MANUSCRIPT

Figure S1: Absolute concentrations of PM2.5 components measured at the peri-urban site in Iowa City in 2011.

15

Elemental carbon

Calcium

Organic matter (1.8*OC)

Magnesium

Ammonium

Potassium

Sulfate

Sodium

RI PT

20

4

Nov 10

Nov 07

Nov 04

Nov 01

Oct 29

Oct 26

Oct 17

Oct 14

Oct 11

Oct 08

Oct 05

Oct 02

Sep 29

Sep 26

Sep 23

Sep 20

Sep 17

AC C

EP

TE D

Sep 14

Sep 11

Sep 08

Sep 05

Sep 02

Aug 30

Aug 27

0

M AN U

5

Oct 23

10

Oct 20

SC

Nitrate

Aug 24

-3 Concentration ( g m )

25

ACCEPTED MANUSCRIPT

Figure S2: Five common NOAA HYSPLIT backward air mass trajectories for the sampling period b)

Northwesterly

October 14 - 18

September 14 - 16 e)

Southeasterly

Southwesterly

c)

Northeasterly

September 6 - 11 Day 1 Day 2 Day 3 Day 4 Day 5 Day 6 Day 7

M AN U

SC

d)

Northerly

RI PT

a)

September 1 - 4

AC C

EP

TE D

October 7 - 13

5

ACCEPTED MANUSCRIPT

Figure S3: Ambient concentrations of PAH measured at the peri-urban (a) and urban (b) sites in Iowa City, IA during 2011. Saturdays with local football events associated with heightened traffic and outdoor cooking activities are marked with a star (*). Pyrene Benzo[ghi]fluoranthene Cyclopenta[cd]pyrene Benz[a]anthracene Chrysene 1-Methylchrysene Retene Benzo[b]fluoranthene Benzo[k]fluoranthene

(a) Peri-urban

4

2

0 5

SC

1

(b) Urban

M AN U

4 3 2 1 0

AC C

EP

TE D

Aug 27-28 Aug 29-31 Sept 01-02 Sept 03* Sept 04-07 Sept 08-09 Sept 10-11 Sept 12-14 Sept 15-16 Sept 17* Sept 18-20 Sept 21-23 Sept 24* Sept 25-28 Sept 29 - Oct 01 Oct 02-03 Oct 04-05 Oct 06-07 Oct 08-09 Oct 10-14 Oct 15* Oct 16-19 Oct 20-21 Oct 22* Oct 23-25 Oct 26-28 Oct 29-31 Nov 01-04 Nov 05* Nov 06-10

Concentration (ng m-3)

3

Benzo[j]fluoranthene Benzo[e]pyrene Benzo[a]pyrene Perylene Indeno[1,2,3-cd]pyrene Benzo[ghi]perylene Dibenz[ah]anthracene Picene

RI PT

5

6

ACCEPTED MANUSCRIPT

(a) Peri-urban

0.4

17(H)-22,29,30-Trisnorhopane 17(H)-21(H)-30-Norhopane 17(H)-21(H)-Hopane

0.2 0.1 --NQ--

(b) Urban

0.4 0.3 0.2

TE D

0.1 0.0

NQ

M AN U

NQ

0.0 0.5

AC C

EP

Aug 27-28 Aug 29-31 Sept 01-02 Sept 03* Sept 04-07 Sept 08-09 Sept 10-11 Sept 12-14 Sept 15-16 Sept 17* Sept 18-20 Sept 21-23 Sept 24* Sept 25-28 Sept 29 - Oct 01 Oct 02-03 Oct 04-05 Oct 06-07 Oct 08-09 Oct 10-14 Oct 15* Oct 16-19 Oct 20-21 Oct 22* Oct 23-25 Oct 26-28 Oct 29-31 Nov 01-04 Nov 05* Nov 06-10

Concentration (ng m-3)

0.3

SC

0.5

RI PT

Figure S4: Ambient concentrations of a homologous series of hopanes, which are molecular markers of fossil fuel combustion, measured at the peri-urban (a) and urban (b) sites in Iowa City, IA during 2011. Saturdays with local football events associated with heightened traffic and outdoor cooking activities are marked with a star (*). Samples in which hopanes could not be measured due to solvent interferences are marked as not quantifiable (NQ).

7

ACCEPTED MANUSCRIPT

Figure S5: Ambient concentrations of levoglucosan, a molecular marker of biomass burning, measured at the peri-urban (a) and urban (b) sites in Iowa City, IA during 2011. Saturdays with local football events and outdoor festivities are marked with a star and corresponding to local maxima in levoglucosan concentrations (*).

RI PT

400

(a) Peri-urban

200

SC

100 0

(b) Urban

M AN U

300 200 100 0

AC C

EP

TE D

Aug 27-28 Aug 29-31 Sept 01-02 Sept 03* Sept 04-07 Sept 08-09 Sept 10-11 Sept 12-14 Sept 15-16 Sept 17* Sept 18-20 Sept 21-23 Sept 24* Sept 25-28 Sept 29 - Oct 01 Oct 02-03 Oct 04-05 Oct 06-07 Oct 08-09 Oct 10-14 Oct 15* Oct 16-19 Oct 20-21 Oct 22* Oct 23-25 Oct 26-28 Oct 29-31 Nov 01-04 Nov 05* Nov 06-10

Concentration (ng m-3)

300

8

ACCEPTED MANUSCRIPT

RI PT

Figure S6: Ambient concentrations of cholesterol, a molecular marker of meat cooking, measured at the peri-urban (a) and urban (b) sites in Iowa City, IA during 2011. Saturdays with local football events and outdoor festivities are marked with a star (*) and corresponding to local maxima in cholesterol concentrations at the urban site. Samples in which cholesterol levels were below the detection limit are marked as non-detects (ND). 3.0 2.5

(a) Peri-urban

1.5

SC

1.0 0.5

2.5

ND

(b) Urban

M AN U

0.0

2.0 1.5 1.0 0.5 0.0

--ND--

ND

AC C

EP

TE D

Aug 27-28 Aug 29-31 Sept 01-02 Sept 03* Sept 04-07 Sept 08-09 Sept 10-11 Sept 12-14 Sept 15-16 Sept 17* Sept 18-20 Sept 21-23 Sept 24* Sept 25-28 Sept 29 - Oct 01 Oct 02-03 Oct 04-05 Oct 06-07 Oct 08-09 Oct 10-14 Oct 15* Oct 16-19 Oct 20-21 Oct 22* Oct 23-25 Oct 26-28 Oct 29-31 Nov 01-04 Nov 05* Nov 06-10

Concentration (ng m-3)

2.0

9

ACCEPTED MANUSCRIPT

Table S1: NOAA HYSPLIT backward air mass trajectory categories for Iowa City in 2011. Direction Other NW SW NE N NW NW NW NW SW Other Other Other Other SW Other Other NW Other NW

AC C

EP

TE D

M AN U

SC

RI PT

Date Direction Date Direction Date Direction Date 25-Aug NW 14-Sep N 4-Oct Other 24-Oct 26-Aug N 15-Sep N 5-Oct SW 25-Oct 27-Aug Other 16-Sep N 6-Oct SW 26-Oct 28-Aug NE 17-Sep Other 7-Oct SE 27-Oct 29-Aug NE 18-Sep SE 8-Oct SE 28-Oct 30-Aug Other 19-Sep SE 9-Oct SE 29-Oct 31-Aug SE 20-Sep NW 10-Oct SE 30-Oct 1-Sep SW 21-Sep Other 11-Oct SE 31-Oct 2-Sep SW 22-Sep N 12-Oct SE 1-Nov 3-Sep SW 23-Sep N 13-Oct SE 2-Nov 4-Sep SW 24-Sep N 14-Oct NW 3-Nov 5-Sep N 25-Sep NE 15-Oct NW 4-Nov 6-Sep NE 26-Sep Other 16-Oct NW 5-Nov 7-Sep NE 27-Sep Other 17-Oct NW 6-Nov 8-Sep NE 28-Sep Other 18-Oct NW 7-Nov 9-Sep NE 29-Sep Other 19-Oct N 8-Nov 10-Sep NE 30-Sep N 20-Oct NE 9-Nov 11-Sep NE 1-Oct N 21-Oct NE 10-Nov 12-Sep Other 2-Oct NE 22-Oct Other 24-Oct 13-Sep SW 3-Oct Other 23-Oct NW 25-Oct NW – Northwesterly; N – Northerly; NE – Northeasterly; SE – Southeasterly; SW – Southwesterly; Other – Not represented by any of the directional categories

10

ACCEPTED MANUSCRIPT

Table S2: Average OC, sulfate and ammonium concentration and standard deviation with respect to the air mass direction OC (gC m-3) 1.9 ± 0.7

Sulfate (g m-3) 1.2 ± 0.9

Ammonium (g m-3) 0.6 ± 0.4

Northerly

1.5 ± 0.6

0.6 ± 0.5

0.3 ± 0.2

Northeasterly

2.2 ± 0.8

1.2 ± 0.6

Southeasterly

2.9 ± 1.2

2.6 ± 1.4

Southwesterly

2.9 ± 1.7

2.3 ± 1.8

RI PT

Air mass back trajectory Northwesterly

0.5 ± 0.3 1.2 ± 0.7

AC C

EP

TE D

M AN U

SC

1.0 ± 0.7

11

ACCEPTED MANUSCRIPT

Works Cited Jayarathne, T., Stockwell, C.E., Yokelson, R.J., Nakao, S., Stone, E.A., 2014. Emissions of Fine Particle Fluoride from Biomass Burning. Environmental Science & Technology 48, 12636-12644.

RI PT

Schauer, J.J., et al., 2003. ACE-Asia intercomparison of a thermal-optical method for the determination of particle-phase organic and elemental carbon. Environmental Science & Technology 37, 993-1001.

AC C

EP

TE D

M AN U

SC

Stone, E.A., Nguyen, T.T., Pradhan, B.B., Dangol, P.M., 2012. Assessment of biogenic secondary organic aerosol in the Himalayas. Environ. Chem. 9, 263-272.

12