Contribution of Wildland-Fire Smoke to US PM2.5 and Its Influence on

Subscriber access provided by UNIV OF LOUISIANA

Characterization of Natural and Affected Environments

The contribution of wildland-fire smoke to US PM and its influence on recent trends 2.5

Katelyn O'Dell, Bonne Ford, Emily V. Fischer, and Jeffrey R. Pierce Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.8b05430 • Publication Date (Web): 25 Jan 2019 Downloaded from http://pubs.acs.org on January 27, 2019

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 28

Environmental Science & Technology

1

The contribution of wildland-fire smoke to US PM2.5

2

and its influence on recent trends.

3

Katelyn O’Dell*, Bonne Ford, Emily V. Fischer, and Jeffrey R. Pierce

4

Department of Atmospheric Science, Colorado State University, 200 West Lake Street, 1371

5

Campus Delivery, Fort Collins, Colorado 80523, United States

6

7

Abstract

8

Seasonal-mean concentrations of particulate matter with diameters smaller than 2.5 µm (PM2.5)

9

have been decreasing across the United States (US) for several decades, with large reductions in

10

spring and summer in the eastern US. In contrast, summertime-mean PM2.5 in the western US has

11

not significantly decreased. Wildfires, a large source of summertime PM2.5 in the western US,

12

have been increasing in frequency and burned area in recent decades. Increases in extreme PM2.5

13

events attributable to wildland fires have been observed in wildfire-prone regions, but it is

14

unclear how these increases impact trends in seasonal-mean PM2.5. Using two distinct methods,

15

(1) interpolated surface observations combined with satellite-based smoke plume estimates and

16

(2) the GEOS-Chem chemical transport model (CTM), we identify recent trends (2006-2016) in

17

summer smoke, non-smoke, and total PM2.5 across the US. We observe significant decreases in

18

non-smoke influenced PM2.5 in the western US and find increases in summer-mean smoke PM2.5

ACS Paragon Plus Environment

1


19

in fire-prone regions, although these are not statistically significant due to large interannual

20

variability in the abundance of smoke. These results indicate that without the influence of

21

wildland fires, we would expect to have observed improvements in summer fine particle

22

pollution in the western US but likely weaker improvements than those observed in the eastern

23

US.

24

Introduction and Background

25

Wildland fires are a major source of particulate matter with diameters smaller than 2.5 µm

26

(PM2.5) in the United States (US). According to the Environmental Protection Agency (EPA)

27

2011 National Emissions Inventory (NEI), wildland fires emit 25% of primary PM2.5 in the US1.

28

In the southeastern US, wildland fire burn area is dominated by human-ignited prescribed burns

29

or agricultural burns, while in the western US ~70% of wildland fire burned area is from

30

lightning-ignited fires2. In both regions, burned area is heavily tied to environmental conditions

31

and could be similarly impacted by changes in climate2. In the western US, most wildland fire

32

burned area and emissions occur in the summer, and the occurrence of wildland fires varies

33

greatly from year to year. Annual area burned by wildland fires in the US can differ by a factor

34

of 3 or more from year to year3. In the western US, the interannual variability in wildland fire

35

occurrence drives interannual variability in summer aerosol concentrations4,5. Despite their

36

episodic nature and high interannual variability, wildland fires can have significant negative

37

impacts on air quality and public health in the western US, especially in high fire years6,7.

38 39

Over the past several decades, aerosol concentrations have been decreasing across the eastern

40

US8–11. Using the Interagency Monitoring of Protected Visual Environments (IMPROVE)

41

monitor network, statistically significant decreases in average fine particle mass and elemental


2

Page 2 of 28

Page 3 of 28


42

carbon concentrations have been observed since the late 1980s at many US monitoring sites,

43

predominantly in the eastern US8,9. These decreases, including largely unexpected reductions in

44

organic aerosol, have been linked to large reductions in anthropogenic combustion emissions

45

using both observations and models9,1012,13. In contrast to the eastern US, changes to western US

46

PM2.5 concentrations have been small and largely insignificant8,9. Hand et al. (2011) found a

47

significant decline in summer PM2.5 at a few western US sites, however at most western US sites

48

there was an insignificant change in summer PM2.5 over the period 1989 to 2008. There were

49

positive trends observed at a few sites in several western US states over this time period8,9.

50

Previous work has proposed these positive trends in summer-mean PM2.5 in the western US may

51

be due to the influence of wildland fires. Extreme PM2.5 events (the 98th quantile of daily PM2.5

52

concentrations) attributable to wildland fires have been increasing over the early 21st century in

53

wildfire-prone regions of the western US14.

54 55

On decadal timescales, large wildfires in the western US have been increasing in frequency and

56

burned area since the mid-1980s15,16. These increases are largely the result of changing climatic

57

factors, in part driven by anthropogenic climate change, including earlier spring snowmelt,

58

warmer temperatures, and reduced winter precipitation17–20. Increases in human-ignited fires and

59

a build-up of fuels from fire suppression have also had a significant contribution17,21,22 . Climate,

60

ignition changes, and fuel build-up have hence led to approximately 20 additional large wildfires

61

(a 140% increase) per decade and an additional 123,000 ha in burned area (a 390% increase) per

62

decade since the 1973-1982 decade17.. Climate-related increases in wildfire burned area and the

63

number of large fires are projected to continue to increase23–25. Spracklen et al. (2009) project a

64

~50% increase in burned area across the western US between 2009 and 2050.


3


65

As wildfires in the western US increase in frequency and burned area over the next century,

66

emissions of PM2.5 from fires are also expected to increase26–28. Liu et al. (2016) project ~60%

67

and ~30% increases in the frequency and intensity, respectively, of smoke events (≥ 2

68

consecutive days of smoke-elevated PM2.5) in the western US by the mid-21st century27. Ford et

69

al. (2018) estimated that by the end of the century, fire-related PM2.5 will increase by ~50%26,

70

and several studies have estimated that smoke will dominate the summertime PM2.5 and OA

71

concentrations in the West24,26,28,29. Yue et al. (2013) estimate a ~50-70% increase in

72

summertime organic carbon and ~20-30% increase in summertime black carbon in the US West

73

by the mid-21st century due to increased wildfire emissions24.

74 75

The most rapid increases in wildfire frequency and burned area over the past two decades have

76

been observed in Pacific Northwest (PNW) forests17. In this region, fire is already a significant

77

source of PM2.5, contributing approximately 1/3 of the annual average PM2.5 in 200026. Across

78

the coming century, the rapid increase of wildland fires is expected to continue in this region

79

with an expected ~80% increase in burned area in the PNW alone23. These increases in fire, in

80

conjunction with decreases with anthropogenic PM2.5, are expected to cause fire PM2.5 to

81

dominate PM2.5 in the region by the end of the 21st century26.

82 83

While there have been previous studies that have examined how aerosol concentrations have

84

changed from the mid-1980s up to 2012 and these studies have noted increases in wildland fire-

85

linked extreme PM2.5 events, the specific impact from fires on trends in seasonal-mean PM2.5 has

86

not yet been quantified. There is growing interest in how changes in fire emissions might impact

87

future air quality and offset projected decreases in anthropogenic emissions (e.g., Ford et al.,

88

2018; Liu et al., 2016; Yue et al., 2013)24,26,27. In this study, we are able to determine how these


4

Page 4 of 28

Page 5 of 28


89

competing drivers are already impacting observed trends in PM2.5. Knowledge of how these

90

sources are changing in the present day is needed to understand future air pollution and its

91

impacts, which can inform decision making.

92

93

In this study, we provide updated estimates of trends in seasonal-mean PM2.5 at US monitoring

94

sites over the most recent decade (2006-2016), and we provide the first estimates of trends in

95

wildland-fire specific PM2.5. Focusing specifically on the PNW, we assess the contribution of

96

smoke PM2.5 to summer-mean PM2.5 and how this impacts trends in regional summer PM2.5. We

97

estimate these trends using two distinct methods: 1) surface monitor observations combined with

98

satellite estimates of smoke influence, and 2) GEOS-Chem simulations using Global Fire

99

Emissions Database version 4 (GFED4) biomass burning emissions.

100 101

Methods

102

Time Period and Domain

103

We investigate changes in continuous US seasonal-mean PM2.5 over an 11-year period (2006-

104

2016). This time period is constrained on both ends by data availability, discussed in the

105

following sections. For calculations of seasonal-mean values, we define seasons as follows;

106

winter: January, February, March (JFM); spring: April, May, June (AMJ); summer: July, August,

107

September (JAS); and fall: October, November, December (OND). This choice of months groups

108

the three most active wildfire months in the northwestern US (July, August, September) together.

109

The number of fires and the frequency of smoke in the atmospheric column peak between July

110

and September in this region30.


5


111

Surface Observations

112

We use in situ 24-hour average PM2.5 observations from the EPAs Air Quality System (AQS)

113

monitoring network (https://aqs.epa.gov/aqsweb/airdata.html). The AQS network contains air

114

pollution data collected by air quality monitors across the US maintained by the EPA, state,

115

local, and tribal agencies. The EPA AQS database contains observations from both urban and

116

rural monitoring sites. These data are collected from sites using both gravimetric and beta-

117

attenuation techniques and both 24-hr and 1-hr sample durations. We use PM2.5 data from two

118

sets of PM2.5 monitors that exist in the AQS database: 1) federal reference method (FRM) sites

119

(EPA parameter code 88101) and 2) acceptable non-FRM sites, which reasonably match the

120

FRM (EPA parameter code 88502). In order for the data from a given monitor to be included in

121

our trend analysis at monitoring sites, data must be available for that monitor for every season

122

between 2006 and 2016 and the monitor must have 80% data availability for the season each

123

year. In total, trends in PM2.5 are calculated for 306 unique PM2.5 monitors across the US. At

124

each monitoring site, we evaluate the trend in PM2.5 using a linear least-squares regression

125

(results in main text Figures 1 and 2) as well as Theil-Sen estimator31 (results in supplementary

126

Figures S1 and S2). These two methods qualitatively agree, and the choice of method does not

127

affect the conclusions of this work. The significance of the resulting best-fit lines is evaluated

128

using a two-sided t-test with a null hypothesis that the slope is zero at a 95% confidence level.

129 130

Interpolation of Surface Observations

131

To estimate trends between surface monitor locations, we calculate a gridded surface of daily

132

observed PM2.5 values by interpolating the surface observations. Ordinary kriging, an

133

interpolation method recently applied to air quality studies32–34, is used to interpolate the PM2.5


6

Page 6 of 28

Page 7 of 28


134

concentration between monitors to a gridded surface. The kriging parameters (sill = 2.6, range =

135

8.5, nugget = 0.1) are determined using a k-fold evaluation with 10 folds, discussed further in the

136

SI. We focus the kriging analysis on the summer season and include all surface sites with data

137

available on a given day for each summer season between 2006 and 2016 in the interpolation.

138

This allows us to retain the maximum number of possible sites for the interpolation on each day.

139

Daily average PM2.5 values measured at these sites are kriged to a 15 x 15 km grid over the

140

contiguous US. We calculate summer area average PM2.5 values over the US PNW and over the

141

contiguous US using the gridded PM2.5 estimates. The gridded estimates, rather than point

142

measurements, are used to calculate the area average seasonal PM2.5 to reduce bias towards

143

urban locations where a larger proportion of the monitors are concentrated.

144 145

Satellite Observations of Smoke Plumes

146

Smoke-specific PM2.5 is estimated using in situ PM2.5 observations concurrently with areal

147

polygons of smoke extent from the Hazard Mapping System (HMS)30,35,36. A detailed description

148

of the HMS fire location and smoke polygon dataset is given in Rolph et al. (2009). HMS is an

149

interactive tool that relies on imagery from several satellites and trained satellite analysts to

150

identify fire locations and smoke plumes across North America. The HMS smoke plume

151

polygons are largely derived from visible satellite imagery from the Geostationary Operational

152

Environmental Satellite (GOES) satellites with an approximate spatial resolution of several

153

kilometers35. The smoke plume polygons indicate locations where there is likely smoke

154

somewhere in the atmospheric column during daylight hours. The HMS smoke plume and fire

155

detection dataset has been operational since 2005. We begin our study period in 2006 because

156

this is the first full year where HMS smoke plumes are available. In general, the number and


7


157

extent of HMS smoke plumes are conservative estimates of smoke occurrence in the atmospheric

158

column30.

159 160

Estimation of Smoke PM2.5 with Satellite and Surface Observations

161

To estimate the smoke contribution to local PM2.5 concentrations, we first use HMS smoke

162

plumes re-gridded to the 15 x 15 km kriging grid to flag potentially smoke-influenced days in the

163

atmospheric column (referred to as smoke days). We then subset the data from each surface

164

monitor into 1) days with no overlapping HMS smoke plume, and 2) days with an overlapping

165

HMS smoke plume (smoke days). The first subset of data for a given monitor is used to estimate

166

the seasonal-median PM2.5 concentrations on non-smoke-influenced days. We acknowledge that

167

there are limitations to this method identifying the magnitude of smoke-influence on PM2.5. 1)

168

Dilute smoke not identified by HMS may still influence the PM2.5 on days that do not overlap

169

HMS plumes. 2) Smoke plumes identified by HMS could be located above the surface and have

170

little or no impact on surface PM2.5. 3) HMS smoke plumes are identified from observations

171

during daylight hours, while 24-hour PM2.5 observations include overnight observations when

172

smoke plumes can shift and go unnoticed by HMS.

173 174

We calculate the expected PM2.5 concentration from non-smoke sources at each monitor as the

175

seasonal median of days with no overlapping HMS smoke plume. We use the seasonal median

176

value rather than the mean as it is less sensitive to the effect of smoke-influenced days with

177

higher PM2.5 concentrations that were missed by the HMS smoke plumes. On average, the

178

seasonal mean non-smoke PM2.5 is 0.6 µg m-3 higher than the seasonal median non-smoke PM2.5

179

with the difference between the two methods (mean – median) ranging from -1.8 to 4.4 µg m-3


8

Page 8 of 28

Page 9 of 28


180

across locations. We interpolate these non-smoke influenced seasonal PM2.5 concentrations using

181

ordinary kriging to obtain a continuous non-smoke influenced PM2.5 estimate across the domain.

182

A daily smoke PM2.5 (PM2.5 from wildland fires) concentration is then estimated as the

183

difference between the daily-mean PM2.5 concentration and the seasonal non-smoke influenced

184

PM2.5 concentration for locations included in HMS smoke polygons (we set smoke PM2.5

185

concentrations to zero at locations outside of HMS smoke polygons). Smoke PM2.5

186

concentrations are calculated at each interpolated grid cell on each day during the time period.

187

We calculate the seasonal mean of the total PM2.5 concentration and smoke PM2.5 concentration

188

for each summer season over the 11-year period. We will refer to this method of estimating

189

smoke PM2.5 and trends as the “monitor-HMS method”.

190 191

Chemical Transport Model Simulations

192

We also use the GEOS-Chem CTM (geos-chem.org) version 11-01 to estimate the impact of

193

wildland-fire smoke on US trends in PM2.5. The model is driven by assimilated reanalysis

194

meteorology from the Modern Era Retrospective Analysis for Research and Applications,

195

Version 2 (MERRA2) produced by the NASA Global Modeling Assimilation Office (GMAO)37.

196

Globally, we use the Emissions Database for Global Atmospheric Research (EDGAR) v4.2

197

anthropogenic

198

(http://edgar.jrc.ec.europa.eu/). In the US, we follow the convention in GEOS-Chem version 12-

199

01 where the EDGAR emissions are overwritten by the EPA NEI 2011 scaled for each year from

200

2006-2016 (https://www.epa.gov/chief). We use biomass burning emissions of aerosol and gas-

201

phase species from GFED4 at a 0.25° x 0.25° resolution39,40.

emissions

inventory

globally

202


9

with

regional

overwrites38


Page 10 of 28

203

We run global GEOS-Chem simulations from 2006-2016 with 1-month spin up at 2° x 2.5°

204

resolution with 47 vertical layers between the surface and 0.01 hPa. The model includes tracers

205

for hydrophobic and hydrophilic black carbon (BCPO, BCPI), hydrophobic and hydrophilic

206

organic carbon (OCPO, OCPI), dust (DST), sea-salt aerosol (SALA), inorganic sulfur nitrates

207

(NIT), ammonium (NH4), and sulfate (SO4). Monthly mean PM2.5 concentrations were

208

calculated according to,

209 210 211

PM2.5 = 1.33 × (NH4 + NIT + SO4) + BCPI + BCPO + 2.1 × (OCPO + 1.16 × OCPI)

(1)

+ DST1 + 0.38 × DST2 + 1.86 × SALA,

212 213

where DST1 and DST2 are dust aerosol with an effective radius of 0.7 µm and 1.4 µm,

214

respectively40. Scaling factors for NH4, NIT, SO4, OCPI, and SALA account for aerosol water

215

under the assumption of 35% relative humidity. The DST2 bin contains particle sizes both

216

smaller and larger than PM2.5; 38% of the mass in the bin is assumed to be from PM2.5. The 2.1

217

scaling factor for OCPO and OCPI is derived from the recommended OM/OC ratio from the

218

GEOS-Chem Aerosol Working Group. To determine the impact of fires on PM2.5, we use a pair

219

of simulations. One simulation includes all emissions and the second simulation includes all

220

emissions except the biomass burning emissions from GFED4. Smoke PM2.5 was identified as

221

the difference between monthly PM2.5 simulated with biomass burning emissions on and off.

222 223

In general, the GEOS-Chem model has shown to correlate well with surface observations of

224

PM2.542–44. However, we acknowledge that there are limitations to our modeling approach that

225

could impact estimates of smoke PM2.5. The version of the model we use does not account for

226

plume rise and emits all biomass burning emission at the surface which could lead to artificially


10

Page 11 of 28


227

high surface PM2.5 concentrations and impact transport45. By using GFED4 for biomass burning

228

emissions, as opposed to GFED4s, we are likely underestimating emissions from small fires46.

229

However, most of the burn area in the western US during the summer wildfire season is the

230

result of large fires2. As well, there is considerable variability between various smoke emissions

231

inventories47, and we do not test these differences here. In the smoke PM2.5 estimates, we ignore

232

changes in smoke organic aerosol due to evaporation and secondary aerosol formation48. Finally,

233

the coarse resolution of the model used here may be unable to resolve narrow smoke plumes

234

(however, we attempt to mitigate the impact of this limitation by focusing on seasonal and

235

regional means in the analysis). The limitations of this model-based approach are generally

236

orthogonal to the limitations of the monitor-HMS approach described above, and hence these

237

two methods provide a means of testing which results are robust.

238 239

Results and Discussion

240

Seasonal Trends in PM2.5 Observed at Surface Sites

241

Observed trends in seasonal-mean PM2.5 at contiguous US sites are shown in Figure 1 for the

242

period 2006 to 2016. In line with several previous studies of aerosol surface

243

concentrations8,9,11,49, we find a predominantly negative trend in seasonal-mean PM2.5 in all

244

seasons. The largest reductions in absolute PM2.5 concentrations occurred in spring and summer

245

in the eastern US during this time period. In these regions over our 11-year study period, the rate

246

of decline in PM2.5 was as large as -1.66 µg m-3 yr-1, which occurred in McKeesport,

247

Pennsylvania in the summer season. Significant declines in PM2.5 are ubiquitous across the

248

eastern US in all seasons: the PM2.5 trends are significantly negative at 45%, 58%, 47%, and

249

56% of sites east of 100° W in winter, spring, summer, and fall, respectively.


11


250

Figure 1. Slopes of linear fits on seasonal-mean PM2.5 at EPA AQS sites from 2006-2016. Sites

251

with slopes that are significantly different from 0 at the 95% confidence level are outlined in

252

black, and sites with insignificant slopes are outlined in gray. Dashed line drawn at 100° W is

253

used in the analysis to distinguish between the eastern and western US.

254 255

In contrast to the eastern US, fewer sites in the western US exhibit significant declines in total

256

PM2.5. The percent of sites in the western US (west of 100° W) that exhibit a significant decrease

257

in PM2.5 is 19%, 23%, and 37% over winter, spring, and fall. In winter (Figure 1a), we observe a

258

positive, but insignificant, trend at a few sites in the PNW (represented by the gray box in Figure

259

2), possibly due to residential fuel burning for heat. In summer (Figure 1c), 27 of the 48 sites

260

studied in the PNW report positive but insignificant trends in summer-mean PM2.5 and, only 3

261

sites (less than 5% of sites studied in the western US) exhibit a significant decline in PM2.5 over

262

this period. This is in agreement with prior work on summer PM2.5 trends observed at surface


12

Page 12 of 28

Page 13 of 28


263

monitors and aerosol optical depth50 over previous years46,47. We investigate whether the positive

264

trends in summer total PM2.5 can be attributed to positive trends in smoke PM2.5.

265 266

Trends in Smoke and Non-smoke PM2.5

267

Figure 2 presents trends in total, non-smoke, and smoke-influenced PM2.5 using the monitor-

268

HMS method (panels a, b, and c) with 15 km resolution and GEOS-Chem with 2° x 2.5°

269

resolution (panels d, e, and f). total PM2.5

non-smoke PM2.5

smoke PM2.5

monitor-HMS

GEOS-Chem

270 271

Figure 2. Slopes of linear fits on summer-mean (JAS) total PM2.5, non-smoke PM2.5 (summer

272

median for monitor-HMS method), and smoke PM2.5 from 2006-2016 for the monitor-HMS

273

(panels a, b, and c) and the GEOS-Chem (panels d, e, and f) methods. Locations with slopes

274

significantly different from zero at the 95% confidence level are dotted. Grey boxes define the

275

Pacific Northwest region for both the 15 x 15 km kriging grid (panels a, b, and c) and the 2° x

276

2.5° GEOS-Chem grid (panels d, e, and f).

277


13


278

By including CTM estimates in our analysis, we are able to identify discrepancies between

279

observed and simulated trends in PM2.5 over the past decade. In general, the model and monitor-

280

HMS approaches agree in the location and sign of trends in summer-mean total PM2.5. GEOS-

281

Chem predicts a larger spatial extent in significant negative trends in PM2.5 compared to the

282

monitor-HMS method. In the PNW and northern California, the two approaches disagree about

283

the location, sign, and magnitude of the trends in total PM2.5. The inconsistent patterns in PM2.5

284

trends in the PNW and northern California shown in both estimation methods are reflected in the

285

observations from monitoring sites in the region (see Figure 1c). There are PM2.5 data records

286

that have both increased and decreased at different monitoring sites in the PNW and northern

287

California over this time period.

288 289

Trends in non-smoke PM2.5 are shown in Figures 2b and 2e. In the absence of smoke influence,

290

the spatial extent of significant negative changes in PM2.5 expands for both the model and the

291

monitor-HMS method. The monitor-HMS method shows significant declines in PM2.5 across

292

Idaho, Colorado, and Utah while GEOS-Chem shows significant declines in PM2.5 across most

293

of the country. Both GEOS-Chem and the monitor-HMS method produce small negative, or

294

insignificant, changes in non-smoke PM2.5 across most of the PNW. Despite several differences

295

between the two methods of attributing smoke PM2.5, both estimation methods suggest that

296

decreases in summer PM2.5 would be expected over more of the western US in the absence of

297

wildland-fire smoke.

298 299

Figures 2c and 2f show trends in summer-mean smoke PM2.5. Both the monitor-HMS-based and

300

GEOS-Chem-based estimates produce an insignificant increase in smoke PM2.5 in Oregon and


14

Page 14 of 28

Page 15 of 28


301

Washington over this period and an insignificant decrease over portions of Montana. Both

302

methods show small positive smoke PM2.5 trends over the northern Great Plains and New

303

England, but the two methods disagree on the sign of the changes in the Southeast. Simulated

304

surface PM2.5 concentrations in GEOS-Chem are likely more sensitive to changes in fire

305

emissions than reality because all biomass burning emissions in the version of the model we used

306

here are emitted at the surface. However, on average, a large fraction of the smoke from fires in

307

the PNW and Canada are emitted above the boundary layer30,45. A possible origin for the

308

disagreement between the monitor-HMS-based and GEOS-Chem-based estimates in the

309

Southeast is our inability to completely isolate smoke PM2.5 from total PM2.5 in the observations

310

due to the limitations of satellite-smoke plume estimates discussed in the methods.

311 312

Wildland-Fire Smoke Contributions to Summer-Mean PM2.5 in the PNW

313

In Figure 3, we focus on area-averaged summer-mean PM2.5 in the PNW (the grey boxes

314

outlined in Figure 2), where Figure 2 demonstrated most of the increases in summer-mean PM2.5.

(a)

(b)

315 316

Figure 3. Estimated smoke contribution to area-averaged summer (JAS) PM2.5 across the Pacific

317

Northwest, defined by the gray box in Figure 2, using the monitor-HMS method (panel a) and


15


318

GEOS-Chem (panel b). The full height of each bar represents the summer mean of total PM2.5.

319

Estimated summer non-smoke PM2.5 is shown in blue, and the difference (the estimated smoke

320

contribution) is shown in orange.

321 322

We separate the summer-mean PM2.5 for this region into a non-smoke PM2.5 concentration (what

323

we would expect the summer PM2.5 to be in the absence of wildland-fire smoke), and the mean

324

smoke contribution to the summer-mean PM2.5. The smoke contribution, shown in orange, is the

325

difference between the total summer-mean PM2.5 (total height of each bar) and the estimated

326

non-smoke PM2.5 (blue section of each bar). During the 11-year study period, wildland-fire

327

smoke contributes between 2.2 and 5.8 µg m-3 (monitor-HMS, Figure 3a) or 1.2 and 9.2 µg m-3

328

(GEOS-Chem, Figure 3b) of the summer-mean PM2.5 in the PNW, depending on the year. The

329

year with the largest contribution in this region according to both methods is 2015 with smoke

330

contributing 64% of the summer-mean PM2.5 using the monitor-HMS-based approach and 76%

331

using the GEOS-Chem-based approach. There were many large fires in this region during 2015,

332

and smoke impacted wide swaths of the US51–53. Despite the fact that wildland fires are typically

333

sporadic and transient events, both methods indicate that wildland-fire smoke can greatly

334

increase the summer PM2.5, especially in years with many large fires.

335 336

In the absence of wildland-fire smoke, there is much less inter-annual variability in summer-

337

mean PM2.5 (Figure 3). Both of our approaches indicate that without the influence of smoke, we

338

would expect modest improvements in summer PM2.5 in the PNW in line with improvements

339

elsewhere in the US. Both approaches produce a ~ 1% decrease in summer PM2.5 per year across

340

this region in the absence of smoke PM2.5. A linear fit on the non-smoke mean PM2.5 from the


16

Page 16 of 28

Page 17 of 28


341

GEOS-Chem approach yields a slope of -0.12 µg m-3 yr-1 (R2 = 0.85, p-value = 5.15 x 10-5); a

342

linear fit on the monitor-HMS estimates yields a similar slope of -0.10 µg m-3 yr-1 (R2 = 0.39, p-

343

value = 0.04). We do not find statistically significant increases in the area-average summer-mean

344

PM2.5 attributed to smoke over the 11-year study period using either the monitor-HMS (slope =

345

0.05 µg m-3 yr-1, R2 = 0.02, p-value = 0.65) or GEOS-Chem approaches (slope = 0.10 µg m-3 yr-1,

346

R2 = 0.02, p-value = 0.71). We note that linear trends are highly dependent on the start and end

347

years of the fit (here ending on an exceptionally low fire year), and there was a large interannual

348

variability in the contribution of smoke to PM2.5 during this period (e.g. Figure 3).

349 350

This work adds to the current literature on US trends in PM2.58,9 by attributing observed increases

351

(and lack of significant trends) in PNW summer-average PM2.5 to wildland-fire smoke. Our

352

updated record of US PM2.5 trends shows continuing improvements in air quality in the eastern

353

US. We show that without the impact of wildland-fire smoke, these improvements would also be

354

expected in the summer over much of the western US. We quantify the contribution of wildland-

355

fire smoke to summer-mean PM2.5 in the PNW, a region heavily impacted by wildland fires, over

356

the past 11 years. We show that wildland-fire smoke contributes more than 50% of the summer-

357

mean PM2.5 in large fire years, but the large interannual variability in this source of PM2.5 makes

358

detecting significant trends in annual mean PM2.5 over short time-spans challenging. We

359

hypothesize that smoke PM2.5 trends may be significant in the western US if our method (which

360

involves HMS that does not go back further than 2006) could be applied over a longer time

361

period. As wildland fires are expected to continue to increase in frequency and severity in the

362

western US, we also hypothesize that our methods may be used to determine significant changes

363

in smoke PM2.5 in the future if monitor and HMS data sources operate continuously.


17


Page 18 of 28

364 365

Acknowledgements

366

This work was supported by NASA project number NNX15AG35G.

367 368

369

Supporting Information Available The following files are available free of charge via the Internet at

370

http://pubs.acs.org.

371

Supplemental Information (pdf): Alternative statistical approach to calculate

372

PM2.5 slopes presented in Figures 1 and 2, a description of the process used to

373

determine the kriging parameters, Figure 2 trends for all other seasons using

374

the observation-based approach, and a discussion of interannual variability in

375

two driving factors of smoke PM2.5 concentrations in the PNW: smoke

376

concentration intensity and smoke day frequency.

377

Corresponding Author

378

*[email protected]

379

References

380

(1)

US

EPA,

O.

2011

National

Emissions

Inventory

(NEI)

Documentation

381

https://www.epa.gov/air-emissions-inventories/2011-national-emissions-inventory-nei-

382

documentation.


18

Page 19 of 28

383


(2)

Brey, S. J.; Barnes, E. A.; Pierce, J. R.; Wiedinmyer, C.; Fischer, E. V. Environmental

384

Conditions, Ignition Type, and Air Quality Impacts of Wildfires in the Southeastern and Western

385

United States. Earths Future 2018, 6 (10), 1442–1456. https://doi.org/10.1029/2018EF000972.

386

(3)

National Interagency Fire Center https://www.nifc.gov/fireInfo/fireInfo_statistics.html.

387

(4)

Jaffe, D.; Hafner, W.; Chand, D.; Westerling, A.; Spracklen, D. Interannual Variations in

388

PM2.5 Due to Wildfires in the Western United States. Environ. Sci. Technol. 2008, 42 (8), 2812–

389

2818. https://doi.org/10.1021/es702755v.

390

(5)

Spracklen, D. V.; Logan, J. A.; Mickley, L. J.; Park, R. J.; Yevich, R.; Westerling, A. L.;

391

Jaffe, D. A. Wildfires Drive Interannual Variability of Organic Carbon Aerosol in the Western

392

U.S.

393

https://doi.org/10.1029/2007GL030037.

394

(6)

in

Summer.

Geophys.

Res.

Lett.

2007,

34,

L16816,

(16).

Gan, R. W.; Ford, B.; Lassman, W.; Pfister, G.; Vaidyanathan, A.; Fischer, E.; Volckens,

395

J.; Pierce, J. R.; Magzamen, S. Comparison of Wildfire Smoke Estimation Methods and

396

Associations with Cardiopulmonary-Related Hospital Admissions. GeoHealth 2017, 1 (3),

397

2017GH000073. https://doi.org/10.1002/2017GH000073.

398

(7)

Reid, C. E.; Brauer, M.; Johnston, F. H.; Jarrett, M.; Balmes, J. R.; Elliott, C. T. Critical

399

Review of Health Impacts of Wildfire Smoke Exposure. Environ. Health Perspect. 2016, 124

400

(9), 1334–1343. https://doi.org/DOI:10.1289/ehp.1409277.

401

(8)

Hand, J. L.; Copeland, S. A.; Day, D. E.; Dillner, A. M.; Indresand, H.; Malm, W. C.;

402

McDade, C. E.; Moore, C. T.; Pitchford, M. L.; Schichtel, B. A.; et al. Spatial and Seasonal

403

Patterns and Temporal Variability of Haze and its Constituents in the United States: Report V


19


404

June

405

variability-of-haze-and-its-constituents-in-the-united-states-report-v-june-2011/.

406

(9)

2011

Page 20 of 28

http://vista.cira.colostate.edu/Improve/spatial-and-seasonal-patterns-and-temporal-

Murphy, D. M.; Chow, J. C.; Leibensperger, E. M.; Malm, W. C.; Pitchford, M.;

407

Schichtel, B. A.; Watson, J. G.; White, W. H. Decreases in Elemental Carbon and Fine Particle

408

Mass

409

https://doi.org/10.5194/acp-11-4679-2011.

in

the

United

States.

Atmos

Chem

Phys

2011,

11

(10),

4679–4686.

410

(10) Leibensperger, E. M.; Mickley, L. J.; Jacob, D. J.; Chen, W.-T.; Seinfeld, J. H.; Nenes,

411

A.; Adams, P. J.; Streets, D. G.; Kumar, N.; Rind, D. Climatic Effects of 1950–2050 Changes in

412

US Anthropogenic Aerosols – Part 1: Aerosol Trends and Radiative Forcing. Atmos Chem Phys

413

2012, 12 (7), 3333–3348. https://doi.org/10.5194/acp-12-3333-2012.

414

(11) Schichtel, B. A.; Husar, R. B.; Falke, S. R.; Wilson, W. E. Haze Trends over the United

415

States,

1980–1995.

Atmos.

Environ.

2001,

416

https://doi.org/https://doi.org/10.1016/S1352-2310(01)00317-X.

35

(30),

5205–5210.

417

(12) Ridley, D. A.; Heald, C. L.; Ridley, K. J.; Kroll, J. H. Causes and Consequences of

418

Decreasing Atmospheric Organic Aerosol in the United States. Proc. Natl. Acad. Sci. 2018, 115

419

(2), 290–295. https://doi.org/10.1073/pnas.1700387115.

420

(13) Malm, W. C.; Schichtel, B. A.; Hand, J. L.; Collett, J. L. Concurrent Temporal and

421

Spatial Trends in Sulfate and Organic Mass Concentrations Measured in the IMPROVE

422

Monitoring Program. J. Geophys. Res. Atmospheres 2017, 122 (19), 2017JD026865.

423

https://doi.org/10.1002/2017JD026865.


20

Page 21 of 28

424


(14) McClure, C. D.; Jaffe, D. A. US Particulate Matter Air Quality Improves except in

425

Wildfire-Prone

Areas.

Proc.

Natl.

426

https://doi.org/10.1073/pnas.1804353115.

Acad.

Sci.

2018,

115

(31),

7901-7906.

427

(15) Westerling, A. L.; Hidalgo, H. G.; Cayan, D. R.; Swetnam, T. W. Warming and Earlier

428

Spring Increase Western U.S. Forest Wildfire Activity. Science 2006, 313 (5789), 940–943.

429

https://doi.org/10.1126/science.1128834.

430

(16) Dennison, P. E.; Brewer, S. C.; Arnold, J. D.; Moritz, M. A. Large Wildfire Trends in the

431

Western United States, 1984–2011. Geophys. Res. Lett. 2014, 41 (8), 2014GL059576.

432

https://doi.org/10.1002/2014GL059576.

433

(17) Westerling, A. L. Increasing Western US Forest Wildfire Activity: Sensitivity to Changes

434

in the Timing of Spring. Phil Trans R Soc B 2016, 371 (1696), 20150178.

435

https://doi.org/10.1098/rstb.2015.0178.

436

(18) Pechony, O.; Shindell, D. T. Driving Forces of Global Wildfires over the Past

437

Millennium and the Forthcoming Century. Proc. Natl. Acad. Sci. 2010, 107 (45), 19167–19170.

438


439

(19) Abatzoglou, J. T.; Williams, A. P. Impact of Anthropogenic Climate Change on Wildfire

440

across Western US Forests. Proc. Natl. Acad. Sci. 2016, 113 (42), 11770–11775.

441


442

(20) Barbero, R.; Abatzoglou, J. T.; Steel, E. A.; Larkin, N. K. Modeling Very Large-Fire

443

Occurrences over the Continental United States from Weather and Climate Forcing. Environ.

444

Res. Lett. 2014, 9 (12), 124009. https://doi.org/10.1088/1748-9326/9/12/124009.


21


Page 22 of 28

445

(21) Balch, J. K.; Bradley, B. A.; Abatzoglou, J. T.; Nagy, R. C.; Fusco, E. J.; Mahood, A. L.

446

Human-Started Wildfires Expand the Fire Niche across the United States. Proc. Natl. Acad. Sci.

447

2017, 114 (11), 2946–2951. https://doi.org/10.1073/pnas.1617394114.

448

(22) Steel, Z. L.; Safford, H. D.; Viers, J. H. The Fire Frequency-Severity Relationship and

449

the Legacy of Fire Suppression in California Forests. Ecosphere 2015, 6 (1), art8.

450

https://doi.org/10.1890/ES14-00224.1.

451

(23) Spracklen, D. V.; Mickley, L. J.; Logan, J. A.; Hudman, R. C.; Yevich, R.; Flannigan, M.

452

D.; Westerling, A. L. Impacts of Climate Change from 2000 to 2050 on Wildfire Activity and

453

Carbonaceous Aerosol Concentrations in the Western United States. J. Geophys. Res.

454

Atmospheres 2009, 114 (D20), D20301. https://doi.org/10.1029/2008JD010966.

455

(24) Yue, X.; Mickley, L. J.; Logan, J. A.; Kaplan, J. O. Ensemble Projections of Wildfire

456

Activity and Carbonaceous Aerosol Concentrations over the Western United States in the Mid-

457

21st

458

https://doi.org/10.1016/j.atmosenv.2013.06.003.

Century.

Atmos.

Environ.

2013,

77,

767–780.

459

(25) Barbero, R.; Abatzoglou, J. T.; Larkin, S.; Kolden, C. A.; Stocks, B. Climate Change

460

Presents Increased Potential for Very Large Fires in the Contiguous United States. Int. J.

461

Wildland Fire 2015, 24 (7), 892–899. https://doi.org/10.1071/WF15083.

462

(26) Ford, B.; Martin, M. V.; Zelasky, S. E.; Fischer, E. V.; Anenberg, S. C.; Heald, C. L.;

463

Pierce, J. R. Future Fire Impacts on Smoke Concentrations, Visibility, and Health in the

464

Contiguous United States. GeoHealth 2018, 2, 229-247. https://doi.org/10.1029/2018GH000144.


22

Page 23 of 28


465

(27) Liu, J. C.; Mickley, L. J.; Sulprizio, M. P.; Dominici, F.; Yue, X.; Ebisu, K.; Anderson,

466

G. B.; Khan, R. F. A.; Bravo, M. A.; Bell, M. L. Particulate Air Pollution from Wildfires in the

467

Western

468

https://doi.org/10.1007/s10584-016-1762-6.

US

under

Climate

Change.

Clim.

Change

2016,

138

(3–4),

655–666.

469

(28) Val Martin, M.; Heald, C. L.; Lamarque, J. F.; Tilmes, S.; Emmons, L. K.; Schichtel, B.

470

A. How Emissions, Climate, and Land Use Change Will Impact Mid-Century Air Quality over

471

the United States: A Focus on Effects at National Parks. Atmospheric Chem. Phys. 2015, 15,

472

2805–2823.

473

(29) Hallar, A. G.; Molotch, N. P.; Hand, J. L.; Livneh, B.; McCubbin, I. B.; Petersen, R.;

474

Joseph Michalsky; Lowenthal, D.; Kunkel, K. E. Impacts of Increasing Aridity and Wildfires on

475

Aerosol Loading in the Intermountain Western US. Environ. Res. Lett. 2017, 12 (1), 014006.

476

https://doi.org/10.1088/1748-9326/aa510a.

477

(30) Brey, S. J.; Ruminski, M.; Atwood, S. A.; Fischer, E. V. Connecting Smoke Plumes to

478

Sources Using Hazard Mapping System (HMS) Smoke and Fire Location Data over North

479

America. Atmos Chem Phys 2018, 18 (3), 1745–1761. https://doi.org/10.5194/acp-18-1745-2018.

480

(31) Theil, H. A Rank-Invariant Method of Linear and Polynomial Regression Analysis. I.

481

Nederl Akad Wetensch Proc 1950, 53, 386–392.

482

(32) Lassman, W.; Ford, B.; Gan, R. W.; Pfister, G.; Magzamen, S.; Fischer, E. V.; Pierce, J.

483

R. Spatial and Temporal Estimates of Population Exposure to Wildfire Smoke during the

484

Washington State 2012 Wildfire Season Using Blended Model, Satellite, and In-Situ Data.

485

GeoHealth 2017, 1, 106-121. 2017GH000049. https://doi.org/10.1002/2017GH000049.


23


Page 24 of 28

486

(33) Janssen, S.; Dumont, G.; Fierens, F.; Mensink, C. Spatial Interpolation of Air Pollution

487

Measurements Using CORINE Land Cover Data. Atmos. Environ. 2008, 42 (20), 4884–4903.

488


489

(34) Jerrett, M.; Burnett, R. T.; Ma, R.; Pope, C. A. I.; Krewski, D.; Newbold, K. B.;

490

Thurston, G.; Shi, Y.; Finkelstein, N.; Calle, E. E.; et al. Spatial Analysis of Air Pollution and

491

Mortality

492

https://doi.org/10.1097/01.ede.0000181630.15826.7d.

493 494

in

Los

Angeles.

Epidemiology

2005,

16

(6),

727.

(35) Ruminski, M.; Kondragunta, S.; Draxler, R. R.; Zeng, J. Recent Changes to the Hazard Mapping System, 2006.

495

(36) Rolph, G. D.; Draxler, R. R.; Stein, A. F.; Taylor, A.; Ruminski, M. G.; Kondragunta, S.;

496

Zeng, J.; Huang, H.-C.; Manikin, G.; McQueen, J. T.; et al. Description and Verification of the

497

NOAA Smoke Forecasting System: The 2007 Fire Season. Weather Forecast. 2009, 24 (2), 361–

498

378. https://doi.org/10.1175/2008WAF2222165.1.

499

(37) Gelaro, R.; McCarty, W.; Suárez, M. J.; Todling, R.; Molod, A.; Takacs, L.; Randles, C.

500

A.; Darmenov, A.; Bosilovich, M. G.; Reichle, R.; Wargan, K.; Coy, L.; Cullather, R.; Draper,

501

C.; Ankella, S.; Buchard, V.; Conaty, A.; da Silva, A.; Gu, W.; Kim, G.; Koster, R.; Lucchesi,

502

R.; Merkova, D.; Nielsen, J. E.; Partyka, G.; Pawson, S.; Putman, W.; Reineker, M.; Schubert, S.

503

D.; Sienkiewicz, M.; Zhao, B. The Modern-Era Retrospective Analysis for Research and

504

Applications,

505

https://doi.org/10.1175/JCLI-D-16-0758.1.

Version

2

(MERRA-2).

J.

Clim.

2017,

30

(14),

5419–5454.

506

(38) Olivier, J. G. J.; Bouwman, A. F.; Van der Maas, C. W. M.; Berdowski, J. J. M. Emission

507

Database for Global Atmospheric Research (EDGAR): Version 2.0. In Studies in Environmental


24

Page 25 of 28


508

Science; Zwerver, S., van Rompaey, R. S. A. R., Kok, M. T. J., Berk, M. M., Eds.; Climate

509

Change Research; Elsevier, 1995; Vol. 65, pp 651–659. https://doi.org/10.1016/S0166-

510

1116(06)80262-1.

511

(39) van der Werf, G. R.; Randerson, J. T.; Giglio, L.; Collatz, G. J.; Mu, M.; Kasibhatla, P.

512

S.; Morton, D. C.; DeFries, R. S.; Jin, Y.; van Leeuwen, T. T. Global Fire Emissions and the

513

Contribution of Deforestation, Savanna, Forest, Agricultural, and Peat Fires (1997–2009). Atmos

514

Chem Phys 2010, 10 (23), 11707–11735. https://doi.org/10.5194/acp-10-11707-2010.

515

(40) Louis, G.; Randerson James T.; Werf Guido R. Analysis of Daily, Monthly, and Annual

516

Burned Area Using the Fourth‐generation Global Fire Emissions Database (GFED4). J.

517

Geophys. Res. Biogeosciences 2013, 118 (1), 317–328. https://doi.org/10.1002/jgrg.20042.

518

(41) David, L. M.; Ravishankara, A. R.; Kodros, J. K.; Pierce, J. R.; Venkataraman, C.;

519

Sadavarte, P. Premature Mortality Due to PM2.5 over India: Effect of Atmospheric Transport

520

and Anthropogenic Emissions. GeoHealth 0 (ja). https://doi.org/10.1029/2018GH000169.

521

(42) Ford, B.; Heald, C. L. Aerosol Loading in the Southeastern United States: Reconciling

522

Surface and Satellite Observations. Atmos Chem Phys 2013, 13 (18), 9269–9283.

523

https://doi.org/10.5194/acp-13-9269-2013.

524

(43) Pye, H. O. T.; Liao, H.; Wu, S.; Mickley, L. J.; Jacob, D. J.; Henze, D. K.; Seinfeld, J. H.

525

Effect of Changes in Climate and Emissions on Future Sulfate-Nitrate-Ammonium Aerosol

526

Levels

527

https://doi.org/10.1029/2008JD010701.

in

the

United

States.

J.

Geophys.

Res.

Atmospheres


25

2009,

114

(D1).


Page 26 of 28

528

(44) Shah, V.; Jaeglé, L.; Thornton, J. A.; Lopez-Hilfiker, F. D.; Lee, B. H.; Schroder, J. C.;

529

Campuzano-Jost, P.; Jimenez, J. L.; Guo, H.; Sullivan, A. P.; et al. Chemical Feedbacks Weaken

530

the Wintertime Response of Particulate Sulfate and Nitrate to Emissions Reductions over the

531

Eastern

532


United

States.

Proc.

Natl.

Acad.

Sci.

2018,

115

(32),

8110–8115.

533

(45) Zhu, L.; Val Martin, M.; Gatti, L. V.; Kahn, R.; Hecobian, A.; Fischer, E. V.

534

Development and Implementation of a New Biomass Burning Emissions Injection Height

535

Scheme (BBEIH v1.0) for the GEOS-Chem Model (v9-01-01). Geosci. Model Dev. 2018, 11

536

(10), 4103–4116. https://doi.org/https://doi.org/10.5194/gmd-11-4103-2018.

537

(46) van der Werf, G. R.; Randerson, J. T.; Giglio, L.; van Leeuwen, T. T.; Chen, Y.; Rogers,

538

B. M.; Mu, M.; van Marle, M. J. E.; Morton, D. C.; Collatz, G. J.; et al. Global Fire Emissions

539

Estimates during 1997-2016. 2017, 9 (2), 697–720. https://doi.org/10.5194/essd-9-697-2017.

540

(47) Koplitz, S. N.; Nolte, C. G.; Pouliot, G. A.; Vukovich, J. M.; Beidler, J. Influence of

541

Uncertainties in Burned Area Estimates on Modeled Wildland Fire PM2.5 and Ozone Pollution

542

in

543


the

Contiguous

U.S.

Atmos.

Environ.

2018,

191,

328–339.

544

(48) Shrivastava, M.; Cappa, C. D.; Fan, J.; Goldstein, A. H.; Guenther, A. B.; Jimenez, J. L.;

545

Kuang, C.; Laskin, A.; Martin, S. T.; Ng, N. L.; et al. Recent Advances in Understanding

546

Secondary Organic Aerosol: Implications for Global Climate Forcing. Rev. Geophys. 2017, 55

547

(2), 509–559. https://doi.org/10.1002/2016RG000540.

548

(49) Blanchard, C. L.; Hidy, G. M.; Tanenbaum, S.; Edgerton, E. S.; Hartsell, B. E. The

549

Southeastern Aerosol Research and Characterization (SEARCH) Study: Temporal Trends in Gas


26

Page 27 of 28


550

and PM Concentrations and Composition, 1999–2010. J. Air Waste Manag. Assoc. 2013, 63 (3),

551

247–259. https://doi.org/10.1080/10962247.2012.748523.

552

(50) Attwood, A. R.; Washenfelder, R. A.; Brock, C. A.; Hu, W.; Baumann, K.;

553

Campuzano‐Jost, P.; Day, D. A.; Edgerton, E. S.; Murphy, D. M.; Palm, B. B.; et al. Trends in

554

Sulfate and Organic Aerosol Mass in the Southeast U.S.: Impact on Aerosol Optical Depth and

555

Radiative

556

https://doi.org/10.1002/2014GL061669.

Forcing.

Geophys.

Res.

Lett.

2014,

41

(21),

7701–7709.

557

(51) Lindaas, J.; Farmer, D. K.; Pollack, I. B.; Abeleira, A.; Flocke, F.; Roscioli, R.; Herndon,

558

S.; Fischer, E. V. Changes in Ozone and Precursors during Two Aged Wildfire Smoke Events in

559

the Colorado Front Range in Summer 2015. Atmospheric Chem. Phys. 2017, 17 (17), 10691–

560

10707. https://doi.org/https://doi.org/10.5194/acp-17-10691-2017.

561

(52) Wettstein, Z. S.; Hoshiko, S.; Fahimi, J.; Harrison, R. J.; Cascio, W. E.; Rappold, A. G.

562

Cardiovascular and Cerebrovascular Emergency Department Visits Associated With Wildfire

563

Smoke Exposure in California in 2015. J. Am. Heart Assoc. 2018, 7 (8).

564

(53) Ford, B.; Burke, M.; Lassman, W.; Pfister, G.; Pierce, J. R. Status Update: Is Smoke on

565

Your Mind? Using Social Media to Assess Smoke Exposure. Atmos Chem Phys 2017, 17 (12),

566

7541–7554. https://doi.org/10.5194/acp-17-7541-2017.

567

568

Table of Contents Graphic


27


Non-smoke PM2.5 Trend

569

Smoke PM2.5 Trend

Total PM2.5 Trend

Change in summer PM2.5 per year (2006 – 2016)


28

Page 28 of 28

Contribution of Wildland-Fire Smoke to US PM2.5 and Its Influence on

Recommend Documents