McCarthyetal.2021.AssessmentofmobilesourceairtoxicsinanEnvironmentalJusticeDenvercommunityadjacenttoafreeway.pdf

Full Terms & Conditions of access and use can be found athttps://www.tandfonline.com/action/journalInformation?journalCode=uawm20

Journal of the Air & Waste Management Association

ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/uawm20

Assessment of mobile source air toxics in anEnvironmental Justice Denver community adjacentto a freeway

Michael C. McCarthy, Anondo D. Mukherjee, Michael Ogletree, JonathanFurst, Marie I. Gosselin, Mark Tigges, Gregg Thomas & Steven G. Brown

To cite this article: Michael C. McCarthy, Anondo D. Mukherjee, Michael Ogletree, JonathanFurst, Marie I. Gosselin, Mark Tigges, Gregg Thomas & Steven G. Brown (2021) Assessmentof mobile source air toxics in an Environmental Justice Denver community adjacent toa freeway, Journal of the Air & Waste Management Association, 71:2, 231-246, DOI:10.1080/10962247.2020.1734113

To link to this article: https://doi.org/10.1080/10962247.2020.1734113

View supplementary material Published online: 04 Feb 2021.

Submit your article to this journal Article views: 289

View related articles View Crossmark data

TECHNICAL PAPER

Assessment of mobile source air toxics in an Environmental Justice Denvercommunity adjacent to a freewayMichael C. McCarthya, Anondo D. Mukherjeea, Michael Ogletreeb, Jonathan Furstc, Marie I. Gosselinc,Mark Tiggesc, Gregg Thomasb, and Steven G. Browna

aSonoma Technology, Inc., Petaluma, CA, USA; bColorado Department of Public Health & Environment, Denver, CO, USA; cAir ResourceSpecialists, Inc., Fort Collins, CO, USA

ABSTRACTAir pollutant concentrations are often higher near major roadways than in the surroundingenvironments owing to emissions from on-road mobile sources. In this study, we quantified thegradient in black carbon (BC) and air toxics concentrations from the I-70 freeway in the Elyria-Swansea environmental justice neighborhood in Denver, Colorado, during three measurementcampaigns in 2017–2018. The average hourly upwind-downwind gradient of BC concentrationsfrom the roadway was 500–800 ng/m3, equal to an increment of approximately 30-80% abovelocal background levels within 180 m of the freeway. When integrated over all wind directions,the gradients were smaller, approximately 150–300 ng/m3 (~11-18%) over the course of nearlyfour months of measurements. No statistically significant gradient in air toxics (e.g., benzene,formaldehyde, etc.) was found, likely because the uncertainties in the mean concentrations werelarger than the magnitude of the gradient (<25%). This finding is in contrast to some earlierstudies in which small gradients of benzene and other VOCs were found. We estimate that samplesizes of at least 100 individual measurements would have been required to estimate meanconcentrations with sufficient certainty to quantify gradients on the order of ±10% uncertainty.These gradient estimates are smaller than those found in previous studies over the past twodecades; more stringent emissions standards, the local fleet age distribution, and/or the steadyturnover of the vehicle fleet may be reducing the overall impact of roadway emissions on near-road communities.

Implications: Gradients of near-road pollution may be declining in the near-road environmentas tailpipe emissions from the vehicle fleet continue to decrease. Near-road concentrationgradients of mobile source air toxics, including benzene, 1,3-butadiene, and ethylbenzene, willrequire higher sample sizes to quantify as emissions continue to decline.

PAPER HISTORYReceived October 10, 2019Revised February 10, 2020Accepted February 11, 2020

Introduction

In recent years, near-road air quality has been the focusof many monitoring and modeling studies (Barzyk et al.2015; Bates et al. 2018; Ginzburg et al. 2015; Parvez andWagstrom 2019; Saha, Khlystov, and Grieshop 2018a)due to the potential negative health impacts of airpollution from major roadways and the large portionof the population residing close to major roadways(Brugge, Durant, and Rioux 2007; Ghosh et al. 2017,2016; Health Effects Institute 2010; Health EffectsInstitute Air Toxics Review Panel 2007; Wilhelm et al.2011). Several studies have shown that daytime pollu-tant concentrations can be many times greater thanbackground concentrations within 150 m of a majorroadway, while decreasing rapidly with increasing dis-tance from the roadway. (Baldauf et al. 2008; Hagler

et al. 2009; Jeong et al. 2019; Oakes et al. 2016; Parvezand Wagstrom 2019; Riley et al. 2014; Saha et al. 2018b;Zhu et al. 2002b, 2004). Observational studies haveshown that there is a significant gradient of blackcarbon (BC), ultrafine particles, and nitrogen dioxide(NO2) concentrations next to roadways, and a smallergradient of fine particulate matter (PM2.5) (Karner,Eisinger, and Niemeier 2010; Saha et al. 2018b).Furthermore, concentrations of certain pollutants mea-sured in the near-road environment are heavily influ-enced by the fleet mix of vehicles traveling on theadjacent roadway; for example, ultrafine particlesand BC concentrations are positively correlated to die-sel-fueled vehicles (Brown et al. 2014; Dallmann et al.2012; Dallmann and Harley 2010; Kleeman et al. 2009;Patterson and Harley 2019; Riddle et al. 2008). Several

CONTACT Steven G. Brown [email protected] Sonoma Technology, Inc., 1450 N. McDowell Blvd., Suite 200 Petaluma, CA 94954, USA.The supplemental data for this article can be accessed on the publisher’s website.

JOURNAL OF THE AIR & WASTE MANAGEMENT ASSOCIATION2021, VOL. 71, NO. 2, 231–246https://doi.org/10.1080/10962247.2020.1734113

© 2021 A&WMA

studies have also found significant negative healthimpacts, such as reduced lung function, low birthweights, and increased asthma and risk of heart failure,for individuals living within 300 m of major roadways(Brandt et al. 2014; Ghosh et al. 2016; Health EffectsInstitute 2010; Urman et al. 2014).

Hazardous air pollutants (HAPs), also called air toxics,are ambient air pollutants that pose a wide range of threatsto human health. Section 112 of the Clean Air Act (CAA),as amended in 1990, lists 189 HAPs and mandates regula-tions to control their emissions. Scientific studies haveshown that a number of these air toxics are carcinogenic,and many are associated with a wide variety of negativehealth impacts, including adverse effects on reproductive,developmental, and neurological health (Clark-Reyna,Grineski, and Collins 2016; Loh et al. 2007; Rosenbaumet al. 1999; Wilhelm et al. 2011; Woodruff et al. 1998, 2000).Mobile source air toxics are air toxics that are emittedthrough the combustion cycle of motor vehicles. The 2014National Air Toxics Assessment provides an overview ofwhich HAPs provide the greatest carcinogenic exposurerisk, and which emissions sources contribute to that risk –showing that mobile source emissions from light dutyvehicles, heavy duty vehicles, and diesel vehicles remaina significant contributor to cancer risk (U.S.Environmental Protection Agency 2014b, 2015a). Benzeneis a key hazardous air pollutant and mobile source air toxic(MSAT) that is the focus of the measurements and model-ing here, as it classified as a known human carcinogen forall routes of exposure under the proposed revisedCarcinogen Risk Assessment Guidelines (Fruin et al.2001; Health Effects Institute Air Toxics Review Panel2007; Skov et al. 2001; Smith 2010; U.S. EnvironmentalProtection Agency 1996, 2018; Woodruff et al. 1998).Other potentially important sources of benzene and otherHAPs in the area include emissions from manufacturing,petrochemical and chemical facilities, waste incinerators,non-road mobile sources, and stationary engines.

Mobile source emissions have been a regulatory targetof successive rulemaking procedures following the statu-tory mandate of the 1990 amendments to the CAA(https://www.epa.gov/mobile-source-pollution/regulations-reduce-mobile-source-pollution). Between 2000and 2014, the U.S. Environmental Protection Agency(EPA) promulgated three major regulations: Tier 2Motor Vehicle Emissions Standards and Gasoline SulfurControl Requirements; Control of Hazardous AirPollution from Mobile Sources; and Tier 3 MotorVehicle Emission and Fuel standards (U.S.Environmental Protection Agency 2000, 2014a, 2007).Implementation of Tier 3 standards for motor vehicleemissions and fuel composition are ongoing, and tailpipeemissions standards “generally phase in between model

years 2017 and 2025.” This regulatory framework has ledto significantly lower emissions from the traffic sectoroverall (McDonald et al. 2013; Reid et al. 2016).

A review study by Karner et al. examined the spatialgradient of pollutant concentrations over a 500-meterdistance from the edge of the roadway, relative to thebackground and relative to the edge of the roadway,synthesizing results from numerous studies prior to2010 (Karner, Eisinger, and Niemeier 2010). They foundthat some volatile organic compounds (VOCs), includingethane, propane, n-butane and n-hexane, showed a rapiddecline in concentration (>50%) starting 150 meters fromthe roadway. They also present a black carbon gradient of~50%, 150 meters from the roadway, and show that totalPM2.5 has a much weaker gradient, 15% over 150 meters.Other studies have focused on the modeled and measuredgradient of vehicle emissions versus distance to road,showing gradients of benzene, BC, and ultrafine particles,generally focused on downwind conditions only (Baldaufet al. 2008; Chang et al. 2015; Parvez and Wagstrom 2019;Riley et al. 2014; Saha et al. 2018b; Zhu et al. 2002b, 2004).

As the health effects from environmental exposure tocriteria pollutants in the United States have been substan-tially reduced over the past four decades (U.S.Environmental Protection Agency 2019), an increasedfocus has been placed on the equity of exposure and tothe environmental and health-related outcomes related tothe location of emissions sources as reflected in environ-mental justice concerns. Nationally, 19% of the popula-tion lives within 500 meters of high traffic volume roads,which are defined as having greater than 25,000 annualaverage daily traffic (AADT) based on 2010 census data(Rowangould 2013). Studies have found inequities ofexposure to air toxics in the near-road population withrespect to race, community segregation, and socioeco-nomic status (Houston, Krudysz, and Winer 2008;Morello-Frosch and Shenassa 2006) as well as race andincome (Apelberg, Buckley, and White 2005; Mirandaet al. 2011; Pope and Dockery 2006; Rowangould 2013).

The Elyria-Swansea neighborhood is an environmentaljustice community located in central Denver near theconvergence of Interstate Highways 70 and 25. I-70 runseast-west through the neighborhood, and a major rail lineruns southwest-northeast through the neighborhood. Thecommunity in the vicinity of Swansea Elementary schoolis the domain of this study and ranks in the 86th percen-tile for PM2.5 exposure, 96th percentile for diesel PMexposure, and in the 97th percentile for traffic proximityand volume, according to EPA’s EJScreen EJ Index data(available interactively through https://ejscreen.epa.gov/mapper/). Higher percentiles mean that the census blocksof the Elyria-Swansea neighborhood are among the worstin the U.S. for these pollution categories (U.S.

232 M.C. MCCARTHY ET AL.

Environmental Protection Agency 2017). Total AADT onI-70 is 153,000. A major redevelopment project was pro-posed to change the roadway configuration of I-70 as ittransits this neighborhood (see https://www.codot.gov/projects/i70east). Currently, I-70 is an elevated viaductapproximately 30 feet higher than the surrounding com-munities. As part of the I-70 redevelopment projectbeginning in 2019, the I-70 viaduct will be replaced witha 30 foot below-grade highway. A 990 foot long stretchadjacent to Swansea Elementary School will be covered bya 4-acre cap and ground-level parkland. One of the majorgoals of this study is to provide baseline concentrationmeasurements of pollution gradients around the elevatedroadway configuration prior to redevelopment. It isanticipated that a second monitoring study will be per-formed after completion of the I-70 redevelopment pro-ject to attempt to quantify the changes in pollutionconcentrations.

In this work, we present results from measurements ofblack carbon and air toxics at multiple temporary com-munity sites perpendicular to I-70 in the Elyria-Swanseaneighborhood, focused on the area at and aroundSwansea Elementary School. We use these measurementsto determine concentrations of BC and air toxics, and theextent to which concentrations are higher with increasingproximity to the freeway. Measured results are comparedto modeling of benzene with AERMOD. As part of workto understand the impact of a future expansion of I-70,the Denver Department of Public Health andEnvironment (DDPHE) conducted modeling of PM2.5,NO2, and benzene for the Elyria-Swansea neighborhood.This was presented in The Going One Step Beyond inNorth Denver study (Thomas, Williams, and Bain 2007),which is a combined neighborhood and project scalemodeling assessment, to compare a 2011 baseline yearwith 2035 predicted conditions for the expanded and fullybuilt out I-70 (Central 70) project. Benzene in Denver ismostly emitted from motor vehicle exhaust in addition toother sources such as gasoline service stations, oil and gasactivities and refining, and wood burning.

Experiments

Spatial domain

Air pollution monitoring of air toxics (carbonyls, VOCs,and BC) was conducted over three campaigns on a north-south axis along Elizabeth Street and Thompson Courtperpendicular to I-70, focused on the area at and aroundSwansea Elementary School, which employed a playgroundadjacent to I-70. Monitoring stations were sited on or nearthe Swansea Elementary School campus north of the high-way and at sites on rooftops and in parks south of the

highway. Figure 1 shows the study domain along withmonitoring site locations for Campaigns I, II, and III.Monitoring site locations for Campaigns I and II areshown in blue, and those for Campaign III are shownin red.

Monitoring campaigns

The study involved three monitoring periods. CampaignsI and II had identical monitoring methods and site loca-tions. Campaign III deployed measurements at differentsite locations as a result of logistical constraints. In all cases,the goal was to identify concentrations in the near-roadenvironment that were representative of concentrationexposures experienced in the community and elementaryschool during school periods.

● Campaign I was run from October 25, 2017,through November 25, 2017.

● Campaign II was run from February 2, 2018,through March 2, 2018.

● Campaign III was run from September 14, 2018,through October 29, 2018.

Table 1 lists the site locations, elevations, names, andmeasurement methods deployed in Campaigns I and II.Table 2 lists the site locations, elevations, names, andmeasurement methods deployed in Campaign III. Notethat in Campaign III, site 2N_a only operated fromSeptember 14 through 25, 2018. That site was thenrelocated to the site labeled 2N_b, which operatedfrom September 26 through October 29, 2018. BC wasmeasured by microAeths at five near-roadway sites inCampaigns I and II. Sites 1S and 1N were closest to theroadway (within 20 meters of the nearest lane of trafficon I-70). Sites 2S, 3N, and 5N were all more than100 meters from the nearest lane on I-70. Site 2S waslocated on top of a building and is elevated relative toother sites in the measurement study.

Monitoring methodology

Four unique measurement methods were employedduring the campaigns.

(1) BC measurements were made in all three cam-paigns using microAethalometers. BC concen-trations are corrected for optical attenuationusing the approach documented in Kirchstetterand Novakov (2007).

(2) Meteorological measurements were made in allthree campaigns, most importantly includingwind speed and wind direction.

JOURNAL OF THE AIR & WASTE MANAGEMENT ASSOCIATION 233

(3) Carbonyl cartridge samples were collected inCampaigns I and II using EPA method TO-11.

(4) Summa canisters were used to collect VOCs inCampaigns I–III and analyzed using GC-MSEPA method TO-15 (U.S. EnvironmentalProtection Agency 1999b).

VOC results from Campaign I, identified midwaythrough campaign II (Feb 2018), were suspected to be

of questionable accuracy due to a laboratory issue.Three split samples to different laboratories late incampaign II confirmed this. This was a primary driverfor conducting Campaign III, since VOC concentra-tions are important to the surrounding communities.

Black carbon – BC concentrations were determinedusing the Aethlabs AE51 personal microAethAethalometers. The microAeth samplers are fully self-con-tained instruments with a built-in pump, flow control, and

Table 1. Site locations and measurements deployed in Campaigns I and II.Site Name Distance from I-70 (m) Latitude Longitude Elevation (m) MicroAeth Carbonyls Met

1N 17 39.78034 104.9564 1582 x x1S 16 39.77983 104.95488 1583 x x2N 146 39.78066 104.9561 1582 x x2S 127 39.77877 104.9547 1586 x x3N 102 39.78111 104.95611 1582 x x4N 177 39.78137 104.9554 1582 x x5N 177 39.78178 104.9561 1581 x x

Table 2. Site locations and measurements deployed in Campaign III.Site Name Distance from I-70 (m) Latitude Longitude Elevation (m) MicroAeth VOCs Met

1S 16 39.779834 104.95488 1583 x x1N* 91 39.781 104.9554 1582 x x x2N_a 146 39.781 104.9559 1582 x x3N 102 39.7811 104.9561 1582 x x2N_b 148 39.7815 104.956 1582 x x4N 177 39.7818 104.9561 1582 x x

Notes. *Site 1N was not at the same location as in Campaigns I and II due to inaccessibility of the location on the school playgroundduring Campaign III.

Figure 1. Monitoring site locations for Campaigns I and II (left) and Campaign III (right).

234 M.C. MCCARTHY ET AL.

data-storage; see https://aethlabs.com/microaeth/ae51/overview. Data were collected at five-minute intervals in allcampaigns but have been aggregated up to hourly meansfor hours with at least 75% completeness of valid measure-ments (status indicator = 0, concentration ≥ 0 ng/m3). Caiet al. found high reproducibility among six microAethunits (relative standard deviation of 8%) and good agree-ment with the rack-mounted Aethalometer AE33 unit(R = 0.92, slope = 1.01) (Cai et al. 2014) Cheng and Linfound differences between the AE51 and an AE31 modelAethalometer of up to 14% but overall good agreement(R2 = 0.97 without AE31 data treatment) (Cheng and Lin2013). Viana et al. found similar results comparing theAE51 to elemental carbon measurements from a ThermoMulti-Angle Absorption Photometer (MAAP), with R2

values between five AE51s and the MAAP ranging between0.75 and 0.85 with slopes between 0.75 and 1.15 (Vianaet al. 2015). The AE51 has been used in multiple studies fordetermining personal exposure as well as BC concentra-tions inside residences and in ambient environments(Lovinsky-Desir et al. 2018; Rivas et al. 2016).

Meteorology – A standard suite of meteorologicalmeasurements were collected at one monitoring siteduring each campaign. Meteorological data collectedincluded resultant wind speed, resultant wind direction,relative humidity, barometric pressure, and tempera-ture. We note that the siting for the meteorologicalstation did not necessarily meet EPA requirements forsiting due to a limited fetch and obstructions fromnearby trees and buildings.

Carbonyls – Carbonyl compounds were analyzed fromsamples collected on an every-other-day and weekly sche-dule using dinitrophenylhydrazine (DNPH)-coated sor-bent cartridges following EPA’s Method TO-11a (U.S.Environmental Protection Agency 1999a). These sampleswere sent to an analytical laboratory for analysis usinghigh-performance liquid chromatography (HPLC).

VOCs – VOCs were collected for 24 hours on an every-other-day schedule using SUMMA stainless steel canistersfollowing EPA method TO-15 (U.S. EnvironmentalProtection Agency 1999b). Collected samples were sentto a local university laboratory for analysis using gaschromatograph mass spectrometry (GC-MS).

Modeling methodology

The AERMOD dispersion model was used to estimatebenzene concentrations from both polygon (area andnonroad) and link-based (highway) sources. Mostassumptions and model settings were kept the same asin the 2016 North Denver study (Thomas, Ogletree,and Clay 2016). Polygon emissions were calculated byallocating the 2011 NEI county benzene emissions to

the polygon transportation analysis zones (TAZs). Theallocation surrogates, obtained from the DenverRegional Council of Governments (DRCOG) data,included vehicle miles traveled (VMT), populationand/or population density, number of oil and gaswells, and miles of railroad tracks. Area sources wereassumed not to change between 2011 and 2018 in themodel; this assumption is likely to result in a negligibleimpact on total emissions in each polygon (~1–2%difference in total emissions if NEI 2014 rates wereused). The 2004 meteorological data used in thisstudy were collected at the nearby Rocky MountainArsenal (surface) and the former Denver StapletonAirport (upper air) and were again used in this model.

In the 2016 North Denver study, the Motor VehicleEmission Simulator (MOVES2014) model was used toestimate county level annual emissions of benzene (aswell as PM2.5 and NOx) for 2011, 2015, 2020, 2025,2030, and 2035 (U.S. Environmental ProtectionAgency 2015b, 2016). Since the year 2018 was notexplicitly modeled, the 2018 county annual emissionsof benzene were calculated by interpolation between2015 and 2020. The MOVES results showed that the2011 on-road emissions of benzene were reduced by57% in 2018. This reduction factor was used to estimatethe polygon 2018 benzene emissions. Benzene emis-sions from non-road, railroad and upstream oil andgas sources were assumed to be reduced from 2011 to2018. Reduction factors based on professional judg-ment were used: 10% reduction in non-road, 10%reduction in railroad emissions, and 25% reduction inupstream oil and gas emissions. There are relatively fewemissions from upstream oil and gas in Denver Countyas compared to counties to the north (20–90 milesaway). The AERMOD-modeled impacts in Denverfrom upstream oil and gas are negligible. There is anoil refinery approximately 1.5 miles to the north of themonitoring study area. Given that prevailing winds aremore frequently from the south, the influence of the oilrefinery is expected to be intermittent; chemical finger-print analysis is used to attempt to identify possiblesource impacts from this the refinery. Given both2035 and 2018 (interpolated) annual benzene emissionsfrom MOVES, as well as link emissions for 2035, thelink emissions for 2018 were estimated using a linearrelationship. For onroad emissions, the project linkswere modeled explicitly based on forecast VMT.However, there are also emissions from local roadsand other sectors in the city not captured in forecasts.For transportation analysis zones (TAZs) bordering thearea around Swansea Elementary school, all emissionswere included (onroad, nonroad, area, and point). ForTAZs where the I-70 on-network highway links fell

JOURNAL OF THE AIR & WASTE MANAGEMENT ASSOCIATION 235

within that TAZ, 10% of previously processed onroademissions were retained to account for estimated con-tributions from traffic on local roads and connectors,mainly in the surrounding neighborhood. The link-based on-road emissions were then modeled on top ofthe remaining emissions.

The study estimated emissions and concentrations ofPM2.5, NO2, and benzene for two scenarios: winter(January) and summer (July) concentrations. Benzeneresults are presented here.

It is important when interpreting the results fromthis project to understand the local meteorology. TheSouth Platte River is about one mile west of the projectboundary shown in Figure 1. Typically, winds are fromthe south during the overnight and early morninghours, following the river drainage. Therefore, locationsnorth of the highway are expected to see higher impactsduring those times. As the mountains to the west ofDenver absorb sunlight and the temperature rises, a lowpressure zone is created (i.e., the Denver Cyclone) andwinds start originating from the northeast by earlyafternoon, especially during the warmer months. Onaverage, it is more likely that a site north is downwindof the highway in the early morning hours and thenupwind of the highway in the afternoon and early

evening hours. This will mask concentration gradientsfrom the highway unless concentrations are binned bywind direction.

Results

Black carbon in Campaigns I, II, and III

Near-road gradients can be examined as either acutephenomena that occur within the context of a localwind field, or as a cumulative chronic gradient thatintegrates over all possible wind directions. Both typesof exposures can be important, as people living, work-ing, and going to school next to roadways have differ-ent activity patterns that may be affected as a result oftheir movements. We examined both the chronic inte-grated exposures and the acute maximum gradientswhen winds are from the roadway.

Diurnal patterns of BC concentrations segregated intothree wind bins are shown in Figures 2 and 3. Wind binsare 120 degree arcs―winds from the north, winds in two60-degree arcs from the east or west parallel to the free-way, and winds from the south. Hours are binned in4 hour aggregates to increase the sample size, reducenotch sizes, and better illustrate central tendencies in

Figure 2. Campaign I and II gradients in hourly average binned BC concentrations (ng/m3) measured by microAeths when winds arefrom the north (top), parallel (middle), and south (bottom).

236 M.C. MCCARTHY ET AL.

patterns. The concentration gradients were characterizedacross all hours when segregated by wind direction andare shown in Tables 3 and 4 for Campaigns I and II, andCampaign III, respectively. Boxplots were generated usingthe R tidyverse ggplot2 library (https://ggplot2.tidyverse.org/reference/geom_boxplot.html); boxes correspond tothe interquartile range (IQR), whiskers extend to 1.5*IQR,and individual outliers are outside the 1.5*IQR. Notchesare 1.58*IQR/sqrt(n), which gives a roughly 95% confi-dence interval for comparing medians.

Mann-Whitney U-Tests were run on wind bin aggre-gated datasets from Campaigns I/II and III by site (i.e., notsegregated by hour-range as shown in Figures 2 and 3) todetermine whether the distributions of concentrationswere statistically significant at the 95% confidence level.Note that the Mann-Whitney U-Test is non-parametric sodoes not assume anything about the shape of the concen-tration distribution. Concentration distributions at Site 1Swere significantly higher than those at other sites fartherfrom I-70 in both Campaigns I/II and Campaign III.

Figure 3. Campaign III gradients in hourly average binned BC concentrations (ng/m3) measured by microAeths when winds are fromthe north (top), parallel (middle), and south (bottom).

Table 3. Summary average BC concentrations for CampaignsI and II, segregated by wind direction.

Average BC Concentration (ng/m3)

SiteDistance (m)from I-70

Winds fromthe South

ParallelWinds

Winds fromthe North

All WindDirections

1S 16 1946 1842a 1944 1916a

2S 127 1646 1271 1496 14901N* 17 2133c 1581 1340 1694b

3N 102 2191c 1481 1229 16445N 177 2031c 1292 1190 1519

Notes. aThe distribution of concentrations for Site 1 S was significantlyhigher at the 95% confidence level than for all other sites.

bThe distribution of concentrations for Site 1N was significantly higher atthe 95% confidence level than for Site 2S and Site 5N.

cThe distribution of concentrations for Sites 1N, 3N, and 5N were all higherthan for Sites 1S and 2S at the 95% confidence level.

Table 4. Summary average BC concentrations for Campaign III,segregated by wind direction.

Average BC Concentration (ng/m3)

SiteDistance (m)from I-70

Winds fromthe South

ParallelWinds

Winds fromthe North

All WindDirections

1S 16 1390a 1586a 1798a 1578a

1N 91 1970 1197 1006 14832N 146 2065 1216 1004 14363N 102 1610 1051 903 1254b

4N 177 1727 1072 949 1334

Notes. aThe distribution of concentrations for Site 1S was significantlyhigher at the 95% confidence level than for all other sites when windsblew from the north, were parallel, and under All Wind Directions. It wasstatistically significantly lower when winds blew from the south.

bThe distribution of concentrations for Site 3N was significantly lower at the95% confidence level than for Site 1S and 1N.

JOURNAL OF THE AIR & WASTE MANAGEMENT ASSOCIATION 237

Gradients stratified by wind bins in the 120-degreearcs relative to the roadway were statistically significantacross many of the comparisons in both the CampaignI/II and Campaign III datasets. When winds were per-pendicular to I-70, statistically significant gradients of500–800 ng/m3 were found at all downwind sites. Thehigher downwind concentrations cause higher expo-sures for people downwind of I-70 during thoseperiods; BC was significantly higher than at upwindsites during these conditions.

Thus, we find that proximity to the highway is the mostimportant contributor to exposure over long time periods,but wind direction relative to I-70 is the most importantcontributor for shorter-term exposure periods. Theseshorter-term exposure periods can be grouped into activityexposure periods to determine whether children attendingelementary schools are downwind of I-70 during schoolhours, for example, and to assess whether their relativeexposures will be higher than those of children attendingnearby schools located farther from major roadways.

VOC sampling in Campaign III

In Campaign III, VOC samples were collected on22 days. Individual samples from each site were com-pared to daily mean concentrations to determine therelative increments at each site. The average incrementsat individual sites were never statistically significantlylarger than at other sites, largely as a result of largestandard deviations in the population means.Supplemental Table 1 shows the summary statisticsfor all pollutants collected during Campaign III thathad valid above-detection-limit samples with morethan 75% of samples collected. Given that the samplesize was only 22 days, the uncertainty in mean concen-trations was larger than any concentration differencesacross sites. For example, the percentages of the stan-dard deviations for most of the hydrocarbons (not thechlorofluorocarbons) were usually larger than 40%,resulting in 95% confidence intervals of around 20%for most site-pollutant means. When compared to 95%confidence intervals, all pollutants’ mean incrementswere statistically significantly indistinguishable froma mean value of 100%; i.e., none of the results areindicative of a population mean that is higher thanthe overall average with statistical significance. Todetect gradients of less than 15% (i.e., the magnitudecalculated for aggregated BC gradients) across siteswith the temporal variability in concentrations thatwere observed day-to-day, at least 80 daily samples ateach site would be needed to get confidence intervalsdown to the 10% level. Note: this does not imply thatthere is no gradient or that the gradient is ~15% across

sites; we are simply stating that a gradient of less than30% relative magnitude could not be detected withstatistical confidence based on the limited number ofsamples, the uncertainty in the measurement method,and the shifting winds over a 24-hr period of samplecollection.

Two independent methods were used to identifypotential emissions sources influencing the near-roadsites for VOCs. In the first, we used enrichment ratioplots. Enrichment ratio plots normalize concentrationsof two pollutants on the x-axis and y-axis by dividingthem by the concentration of a third pollutant; thismethod helps to remove some of the meteorologicalvariability from the standard scatter plot of two pollu-tants. In the second method, positive matrix factoriza-tion (PMF) was run on the combined data from all sitesfor VOCs to identify “factors” that correspond to emis-sions sources with covariant pollutant concentrations.

Enrichment ratio plots segregated by wind directionbin provide some evidence that multiple sources areaffecting VOC concentrations at the near-road sites.Figure 4 shows the enrichment ratio plots for benzeneand n-pentane divided by propane; benzene and2,2,4-trimethlypentane divided by propane; and ben-zene and i-butane divided by propane. In the top leftpanel (benzene, n-pentane), the wind direction bins areclearly separated into a cluster of blue points indicatingthat higher ratios of benzene/propane occur when thewind blows from the south, whereas lower ratios ofbenzene/propane occur when the wind blows fromthe north. In contrast, in the top right panel, the rela-tive ratios remain the same but the wind direction binsare aligned linearly. Finally, the bottom left panel showsbenzene/propane and i-butane/propane; similar ton-pentane propane, the samples are clustered by winddirection with clear bifurcation of the patterns betweensoutherly and northerly winds. The contrast betweenthe three sets of figures clearly indicates a secondhydrocarbon source in addition to mobile source emis-sions influencing the near-road sites.

Source apportionment using PMF on VOCs hasbeen performed on data collected in areas such asLos Angeles, California; Houston, Texas; andEdmonton, Alberta (Brown, Frankel, and Hafner2007; Buzcu and Fraser 2008; Field et al. 2015;McCarthy et al. 2013; Miller et al. 2002; Rappenglücket al. 2013; Watson, Chow, and Fujita 2001). PMFrequires only ambient data, and assumptions regard-ing the numbers or types of sources or specific sourceprofiles are not explicitly needed because PMF gener-ates a set of factor profiles and contributions based onthe input data and selected factor number (Brownet al. 2015; Paatero 1999). Multiple PMF scenarios

238 M.C. MCCARTHY ET AL.

were run to examine the potential number of emis-sions sources that could be identified using PMF.Scenarios were run with 3-, 4-, and 5-factor solutions.Initial runs that included isoprene always includeda factor with isoprene as the dominant VOC, repre-senting biogenic emissions. Isoprene was thenexcluded from the analysis because biogenic emissionssources were not of interest. In subsequent 3- and4-factor runs, 3-factor solutions were as stable as4-factor solutions with regard to bootstrapping (88%reproducibility for 3-factor solutions; 86% reproduci-bility for 4-factor solutions). These bootstrappingresults are consistent with a relatively small matrix oftotal records (~120 samples and 50 species). The3-factor solution results are classified in Table 5 andshown in Figure 5. Selecting a 4-factor solution splitthe mobile source factor into two separate categorieswith a long-chain alkane component with key tracersof n-decane and n-nonane; this may be representativeof diesel exhaust emissions.

Aldehyde sampling in Campaigns I and II

Daily carbonyl samples were collected over 36 days atsix sites during Campaigns I and II. Analysis of theaggregate data showed that no gradients were statisti-cally significant, with mean normalized site concentra-tions of formaldehyde and acetaldehyde varying by lessthan 15% across sites, significantly lower than theapproximately 18-20% 95% confidence interval in thenormalized mean. In other words, any gradients wouldnot be statistically distinguishable unless they weremuch larger in magnitude than the small differences

Figure 4. Enrichment ratio scatter plots of (top, left) benzene and n-pentane divided by propane, (top, right) benzene and2,2,4-trimethylpentane divided by propane, and (bottom, left) benzene and i-butane divided by propane. All figures are coloredby wind direction bins, and sites are indicated by different shapes.

Table 5. Factors identified and key pollutants used to classifyemissions sources in PMF.Factor Factor Name Key Pollutants in Factor

Factor 1 MobileSource

Acetylene, ethene, toluene,2,2,4-trimethylpentane, xylenes

Factor 2 Background Carbon tetrachloride, FreonsFactor 3 Short-Chain

AlkanesEthane, propane, n-butane, i-butane,n-pentane

JOURNAL OF THE AIR & WASTE MANAGEMENT ASSOCIATION 239

seen. To detect gradients of less than 15% across siteswith the temporal variability in concentrationsobserved day-to-day, at least 120 daily samples at eachsite would be needed to get confidence intervals downto the 10% level. Daily concentration statistics for thecarbonyls are provided in Supplemental Table 2.

AERMOD modeling results

The AERMOD model was run separately for TAZ andlink sources. The benzene concentrations from bothmodels were then summed to calculated the predictedambient benzene concentrations. Figure 6 (top) dis-plays the spatial pattern of modeled benzene concen-trations for the month of January in the study area. Theexact model and monitor values are shown on thefigure (shown as model(monitor)) and model to moni-tor ratios range from 0.83 to 1.38. Figure 6 (bottom)shows the model estimates of average benzene concen-trations during the month of July as a representative ofsummer season. Model-to-monitor ratios range from0.40 to 0.7 with most ratios less than a factor of 2. It isimportant to note that even though the benzene mea-surements were collected during September andOctober, comparison of these measurements with themodel estimates of January and July showed goodmodel performance. When applying what we knowabout meteorology and climatology by season/monthin Denver, the modeled concentration for the Sep-Oct2018 monitoring study would fall somewhere betweenJanuary and July; midway would be a reasonableassumption. If that assumption is correct, then model-to-monitor ratios would range from 0.61 to 1.04, which

would be considered excellent model performance.Overall, the model gradient (>0.5 ppb) in concentrationis much larger than that observed in the measurements.

Discussion

For this study, multiple temporary monitoring stationswere deployed in the environmental justice communityof Elyria-Swansea in the vicinity of Swansea ElementarySchool. Monitoring stations were within 200 m of I-70,a large freeway with AADT of 153,000 vehicles.

MicroAethalometer measurements of BC providedstatistically significant concentration gradients for thenear-road environment in the community. Over thethree campaigns, under downwind conditions, hourlyaverage BC concentrations at downwind sites were onthe order of 500–800 ng/m3 higher than those upwindof I-70 during the same hours. This increment was 30-80% higher than average concentrations of BC upwindof I-70. However, aggregation of the concentrationsover the entire campaign timeframe with winds alter-nating between upwind and downwind significantlyreduced the total expected exposures of persons livingor working next to I-70. Maximum incrementsobserved at sites within 20 m of I-70 were150–300 ng/m3 higher than those observed at sites150 m from I-70 during the same time period.Aggregate increments were reduced relative to thewind-segregated bins due to the shifting winds relativeto I-70’s east-west orientation. These total incrementswere only 11-18% higher than concentrations observed100–150 m from I-70 under the same wind conditions.This apparent reduction may be an important

Figure 5. Percent of pollutants identified in each factor with a 3-factor PMF solution excluding isoprene.

240 M.C. MCCARTHY ET AL.

consideration when modeling exposures and expectedhealth impacts for persons living or working in thenear-road environment.

In contrast to previous work in this area, we found thatintegrated gradients in concentrations in the near-roadenvironment were relatively small. The classic work byZhu et al. (2002a) showed an exponential decrease inroadway BC concentrations in Los Angeles, and manyfollow-up studies in the 2000s showed that BC gradientswere large relative to background, i.e., > 80% (Woodruffet al. 1998). Saha et al. (2018b) show a BC gradient around500–800 ng/m3 above background (400 m upwind site) in

the wintertime, using measurements that are primarilydownwind. This study found a comparable BC downwindgradient on the order of 500–800 ng/m3, relative to theupwind site (16 meters upwind). In contrast to the study bySaha et al. (2018b) this study found a small BC gradient of150–300 ng/m3, over 150 meters from the roadway. Thiscould be due to differences in dispersion related to windspeed in Denver, or because the sampling time of themeasurements was integrated over longer periods for thisstudy. While total upwind versus downwind gradientswere as large as a 40% increment on the downwind sideof the highway, the integrated gradients over all wind

Figure 6. Predicted 24-hour averages of benzene concentrations during the month of January 2018 (top) or July 2018 (bottom) andthe observed benzene measurements (24-hour averages) collected from September 15 through October 28, 2018. The values on themap are shown as: Model (Monitor).

JOURNAL OF THE AIR & WASTE MANAGEMENT ASSOCIATION 241

directions and samples were small, only 15%. This meansthat hours when receptors are downwind of a freeway BCconcentrations are well above background, and that whenintegrated over multiple months, BC concentrations are15% higher next to the freeway than at greater distancesfrom the freeway. The smaller gradients observed here maybe due to cleaner diesel vehicles resulting from tighteremissions standards as the vehicle fleet becomes cleanerover time. This is consistent with an overall observeddecreasing trend in black carbon in near-road and urbansettings (Dallmann and Harley 2010; McDonald et al. 2013;Milando, Huang, and Batterman 2016; Rattigan et al. 2013)Alternately, the smaller gradient may be due to the longertime period and integrated measurement relative to mostshorter-term studies (Saha et al. 2018b), or it could bea local fleet with a younger age distribution than in otherstudy locations. It is also possible that the emissions area result of a shifting emissions profile, with lower blackcarbon emissions and higher amounts of other pollutantsthat were not measured in this study (e.g., ultrafines).Finally, the smaller gradient may be partly attributable tothe elevated roadway configuration (Heist, Perry, andBrixey 2009; Steffens et al. 2014).

Following up on the BC comparison, the VOCs andcarbonyls provided no quantitative gradients that couldbe significantly distinguished from the concentrations atother sites. While this may appear to be a null result, wecan constrain the possible size of gradient that could bepresent given our confidence in our ability to measure theconcentrations of the VOCs and carbonyls. Our typical95% confidence intervals for the individual site-parametercombinations were on the order of ±20%. Propagatinguncertainty across multiple sites, we can estimate thatconcentration gradients on the order of 29% would betheoretically detectable with a 95% confidence level. Notethat the 29% value is derived from the comparison of twomean concentrations at two separate sites that each havean individual uncertainty of ±20%; the combined uncer-tainty is larger because the two means have independentuncertainties that are combined in quadrature. None ofthe individual sites or parameters displayed a 30% gradi-ent in concentrations, as shown in the supplementaltables. Therefore, we can quantitatively state that if thereare gradients in concentrations of these species in thenear-road environment, they are smaller than 20% inmagnitude, i.e., essentially as small as or smaller thanthe gradients in BC. In addition, we can state that withtypical day-to-day variance in meteorology (wind speedand wind direction), we would need at least four times asmany samples to theoretically quantify gradients on theorder of 10% (similar to the size of the aggregated gradientobserved for BC). This may be a useful guideline forfuture studies considering near-road gradient analyses.

Model performance was excellent in terms of allmonitors predicting benzene concentrations at levelswithin a factor of two of observed measured concentra-tions. We note that the winds observed at the RockyMountain Arsenal site used for the modeling are sig-nificantly higher in speed than those observed at theSwansea monitoring site, and this may bias the modelconcentration estimates lower than what would beexpected with wind speeds more characteristic of theurban area.

Conclusion

Air pollutant concentrations are often higher nearmajor roadways than in the surrounding environ-ment owing to the proximity of emissions from on-road mobile sources. In this study, we quantified thegradient in black carbon concentrations in the near-road environment of the Elyria-Swansea environ-mental justice neighborhood in Denver, Colorado.The gradient in concentrations segregated by winddirection from the roadway was 500–800 ng/m3,equal to an increment of approximately 30-80%above local background levels. When integratedover all wind directions, the gradients were smaller,approximately 150–300 ng/m3 (~11-18%) over thecourse of about four months of measurements.These gradient estimates are smaller than thosefound in previous studies and may be due to cleanerdiesel vehicles complying with Tier 2 and 3 emis-sions standards.

Measurements of VOCs and carbonyls were unableto statistically quantify gradients because the gradientswere too small in magnitude (<30%) compared to thenumber of samples and uncertainty in measurements.This finding is also in contrast to studies reported inKarner, Eisinger, and Niemeier (2010), but may bedue to significant reductions in emissions over thepast 10 to 20 years that result in a smaller near-roadgradient over an urban background. We estimate thatsample sizes of at least 100 individual measurementswould have been required to estimate mean concen-trations with sufficient certainty to quantify gradientson the order of 15%. These results suggest that recentdecades of improved emissions standards and thesteady turnover of the vehicle fleet may be reducingthe impact of the roadway environment on near-roadcommunities.

Disclosure statement

No potential conflict of interest was reported by the authors.

242 M.C. MCCARTHY ET AL.

Funding

This work was supported by the Environmental ProtectionAgency [17 FAAA 93541].

About the authors

Michael C. McCarthy and Anondo D. Mukherjee areAtmospheric Scientists at Sonoma Technology, Inc. (STI),located in Petaluma, California

Michael Ogletree is an Air Quality Program Manager at theDenver Department of Public Health & Environment inDenver, Colorado.

Jonathan Furst and Marie I. Gosselin are Field Specialists atAir Resource Specialists

Mark Tigges is a Program Manager at Air ResourceSpecialists, located in Fort Collins, Colorado.

Gregg Thomas is the Environmental Quality DivisionDirector at the Denver Department of Public Health &Environment.

Steven G. Brown is a Vice President of STI and Manager ofthe Data Science Department.

References

Apelberg, B. J., T. J. Buckley, and R. H. White. 2005.Socioeconomic and racial disparities in cancer risk fromair toxics in Maryland. Environ. Health Perspect. 113(6):693–99. doi:10.1289/ehp.7609.

Baldauf, R., E. Thoma, M. Hays, R. Shores, J. S. Kinsey,B. Gullet, S. Kimbrough, V. Isakov, T. Long, R. Snow,et al. 2008. Traffic and meteorological impacts onnear-road air quality: Summary of methods and trendsfrom the Raleigh near-road study. J. Air Waste Manage.Assoc. 58:865–78. doi:10.3155/1047-3289.58.7.865.

Barzyk, T. M., V. Isakov, S. Arunachalam, A. Venkatram,R. Cook, and B. Naess. 2015. A near-road modeling systemfor community-scale assessments of traffic-related air pol-lution in the United States. Environ. Modell. Softw.66:46–56. doi:10.1016/j.envsoft.2014.12.004.

Bates, J. T., A. F. Pennington, X. Zhai, M. D. Friberg, F. Metcalf,L. Darrow, M. Strickland, J. Mulholland, and A. Russell. 2018.Application and evaluation of two model fusion approachesto obtain ambient air pollutant concentrations at a fine spa-tial resolution (250m) in Atlanta. Environ. Modell. Softw.109:182–90. doi:10.1016/j.envsoft.2018.06.008.

Brandt, S., L. Perez, N. Künzli, F. Lurmann, J. Wilson,M. Pastor, and R. McConnell. 2014. Cost of near-roadway and regional air pollution–attributable childhoodasthma in Los Angeles County. J. Allergy Clin. Immunol.134 (5):1028–35. doi:10.1016/j.jaci.2014.09.029.

Brown, S. G., S. Eberly, P. Paatero, and G. A. Norris. 2015.Methods for estimating uncertainty in PMF solutions:Examples with ambient air and water quality data andguidance on reporting PMF results. Sci. Total Environ.518-519:626–35. doi:10.1016/j.scitotenv.2015.01.022.

Brown, S. G., A. Frankel, and H. R. Hafner. 2007. Sourceapportionment of VOCs in the Los Angeles area using

positive matrix factorization. Atmos. Environ. 41:227–237.STI-2725. doi:10.1016/j.atmosenv.2006.08.021.

Brown, S. G., M. C. McCarthy, J. L. DeWinter, D. L. Vaughn,and P. T. Roberts. 2014. Changes in air quality atnear-roadway schools after a major freeway expansion inLas Vegas, Nevada. J. Air Waste Manage. Assoc. 64(9):1002–12. STI-3889. doi:10.1080/10962247.2014.907217.

Brugge, D., J. L. Durant, and C. Rioux. 2007. Near-highwaypollutants in motor vehicle exhaust: A review of epidemio-logic evidence of cardiac and pulmonary health risks.Environ. Health 6 (23). doi:10.1186/1476-069X-6-23.

Buzcu, B., and M. P. Fraser. 2008. Comparison of VOCemissions inventory data with source apportionmentresults for Houston, TX. Atmos. Environ. 42:5032–43.doi:10.1016/j.atmosenv.2008.02.025.

Cai, J., B. Yan, J. Ross, D. Zhang, P. L. Kinney,M. S. Perzanowski, K. Jung, R. Miller, and S. N. Chillrud.2014. Validation of MicroAeth® as a black carbon monitorfor fixed-site measurement and optimization for personalexposure characterization. Aerosol. Air Quality Res. 14(1):1–9. doi:10.4209/aaqr.2013.03.0088.

Chang, S. Y., W. Vizuete, M. Breen, V. Isakov, andS. Arunachalam. 2015. Comparison of highly resolvedmodel-based exposure metrics for traffic-related air pollutantsto support environmental health studies. Int. J. Environ. Res.Public Health 12 (12):15605–25. doi:10.3390/ijerph121215007.

Cheng, Y.-H., and M.-H. Lin. 2013. Real-time performance ofthe microAeth® AE51 and the effects of aerosol loading on itsmeasurement results at a traffic site. Aerosol. Air Quality Res.13 (6):1853–63. doi:10.4209/aaqr.2012.12.0371.

Clark-Reyna, S. E., S. E. Grineski, and T. W. Collins. 2016.Health status and residential exposure to air toxics:What are the effects on children’s academic achieve-ment? Fam. Community Health 39 (3):160–68.doi:10.1097/fch.0000000000000112.

Dallmann, T., S. DeMartini, T. Kirchstetter, S. Herndon, andT. Onasch. 2012. On-road measurement of gas and particlephase pollutant emission factors for individual heavy-dutydiesel trucks. Environ. Sci. Technol. 46 (15):8511–18.doi:10.1021/es301936c.

Dallmann, T. R., and R. A. Harley. 2010. Evaluation ofmobile source emission trends in the United States.J Geophys. Res. Atmos. 115. doi:10.1029/2010jd013862.

Field, R. A., J. Soltis, M. C. McCarthy, S. Murphy, andD. C. Montague. 2015. Influence of oil and gas field opera-tions on spatial and temporal distributions of atmosphericnon-methane hydrocarbons and their effect on ozone for-mation in winter. Atmos. Chem. Phys. 15:3527–42. STI-6271. doi:10.5194/acp-15-3527-2015.

Fruin, S. A., M. J. St. Denis, A. M. Winer, S. D. Colome, andF. W. Lurmann. 2001. Reductions in human benzeneexposure in the California South Coast Air Basin. Atmos.Environ. 35:1069–77. doi:10.1016/S1352-2310(00)00306-X.

Ghosh, R., W. J. Gauderman, H. Minor, H. A. Youn,F. Lurmann, K. R. Cromar, L. Chatzi, B. Belcher,C. R. Fielding, and R. McConnell. 2017. Air pollution,weight loss and metabolic benefits of bariatric surgery:A potential model for study of metabolic effects of envir-onmental exposures. Pediatr. Obes. doi:10.1111/ijpo.12210.

Ghosh, R., F. Lurmann, L. Perez, B. Penfold, S. Brandt,J. Wilson, M. Milet, N. Künzli, and R. McConnell. 2016.Near-roadway air pollution and coronary heart disease:

JOURNAL OF THE AIR & WASTE MANAGEMENT ASSOCIATION 243

Burden of disease and potential impact of greenhouse gasreduction strategy in southern California. Environ. HealthPerspect. 124 (2):193–200. doi:10.1289/ehp.1408865.

Ginzburg, H., X. Liu, M. Baker, R. Shreeve, R. K. M. Jayanty,D. Campbell, and B. Zielinska. 2015. Monitoring study ofthe near-road PM2.5 concentrations in Maryland. J. AirWaste Manage. Assoc. 65 (9):1062–71. doi:10.1080/10962247.2015.1056887.

Hagler, G. S. W., R. W. Baldauf, E. D. Thoma, T. R. Long,R. F. Snow, J. S. Kinsey, L. Oudejans, and B. K. Gullett.2009. Ultrafine particles near a major roadway in Raleigh,North Carolina: Downwind attenuation and correlationwith traffic-related pollutants. Atmos. Environ.43:1229–34. doi:10.1016/j.atmosenv.2008.11.024.

Health Effects Institute. 2010. Traffic-related air pollution:A critical review of the literature on emissions, exposure, andhealth effects. Special Report 17, Report prepared by theHealth Effects Institute, Boston, MA, January. https://www.healtheffects.org/publication/traffic-related-air-pollution-critical-review-literature-emissions-exposure-and-health.

Health Effects Institute Air Toxics Review Panel. 2007.Mobile-source air toxics: A critical review of the literatureon exposure and health effects. Special Report 16, Preparedby the Health Effects Insitute, Boston, MA, November.doi:10.1094/PDIS-91-4-0467B.

Heist, D. K., S. G. Perry, and L. A. Brixey. 2009. A windtunnel study of the effect of roadway configurations on thedispersion of traffic-related pollution. Atmos. Environ.43:5101–11. doi:10.1016/j.atmosenv.2009.06.034.

Houston, D., M. Krudysz, and A. Winer. 2008. Diesel trucktraffic in port-adjacent low income and minority commu-nities; environmental justice implications of near roadwayland use conflicts. Transp. Res. Rec. 2076:38–46.doi:10.3141/2067-05.

Jeong, C.-H., J. M. Wang, N. Hilker, J. Debosz, U. Sofowote,Y. Su, M. Noble, R. M. Healy, T. Munoz, E. Dabek-Zlotorzynska, et al. 2019. Temporal and spatial variabilityof traffic-related PM2.5 sources: Comparison of exhaustand non-exhaust emissions. Atmos. Environ. 198:55–69.doi:10.1016/j.atmosenv.2018.10.038.

Karner, A., D. S. Eisinger, and D. Niemeier. 2010. Near-roadway air quality: Synthesizing the findings fromreal-world data. Environ. Sci. Technol. 44:5334–5344. STI-3923. doi:10.1021/es100008x.

Kirchstetter, T. W., and T. Novakov. 2007. Controlled gen-eration of black carbon particles from a diffusion flameand applications in evaluating black carbon measurementmethods. Atmos. Environ. 41 (9):1874–88. doi:10.1016/j.atmosenv.2006.10.067.

Kleeman, M. J., S. G. Riddle, M. A. Robert, C. A. Jakober,P. M. Fine, M. D. Hays, J. J. Schauer, and M. P. Hannigan.2009. Source apportionment of fine (PM1.8) and ultrafine(PM0.1) airborne particulate matter during a severe winterpollution episode. Environ. Sci. Technol. 43 (2):272–79.doi:10.1021/es800400m.

Loh, M. M., J. I. Levy, J. D. Spengler, E. A. Houseman, andD. H. Bennett. 2007. Ranking cancer risks of organichazardous air pollutants in the United States. Environ.Health Perspect. 115 (8):1160–68. doi:10.1289/ehp.9884.

Lovinsky-Desir, S., J. Lawrence, K. H. Jung, A. G. Rundle,L. A. Hoepner, B. Yan, F. Perera, M. S. Perzanowski,R. L. Miller, and S. N. Chillrud. 2018. Assessment of

exposure to air pollution in children: Determining whetherwearing a personal monitor affects physical activity.Environ. Res. 166:340–43. doi:10.1016/j.envres.2018.06.017.

McCarthy, M. C., Y.-A. Aklilu, S. G. Brown, and D. A. Lyder.2013. Source apportionment of volatile organic com-pounds measured in Edmonton, Alberta. Atmos. Environ.81:504–516.STI-5652. doi:10.1016/j.atmosenv.2013.09.016.

McDonald, B. C., D. R. Gentner, A. H. Goldstein, andR. A. Harley. 2013. Long-term trends in motor vehicleemissions in U.S. urban areas. Environ. Sci. Technol. 47(17):10022–31. doi:10.1021/es401034z.

Milando, C., L. Huang, and S. Batterman. 2016. Trends inPM2.5 emissions, concentrations and apportionments inDetroit and Chicago. Atmos. Environ. 129:197–209.doi:10.1016/j.atmosenv.2016.01.012.

Miller, S. L., M. J. Anderson, E. P. Daly, and J. B. Milford.2002. Source apportionment of exposures to volatileorganic compounds. I. Evaluation of receptor modelsusing simulated exposure data. Atmos. Environ. 36(22):3628–41. doi:10.1016/S1352-2310(02)00279-0.

Miranda, M. L., S. E. Edwards, M. H. Keating, and C. J. Paul.2011. Making the environmental justice grade: The relativeburden of air pollution exposure in the United States. Int.J. Environ. Res. Public Health 8 (6):1755–71. doi:10.3390/ijerph8061755.

Morello-Frosch, R., and E. D. Shenassa. 2006. The environ-mental “riskscape” and social inequality: Implications forexplaining maternal and child health disparities. Environ.Health Perspect. 114 (8):1150–53. doi:10.1289/ehp.8930.

Oakes, M. M., J. M. Burke, G. A. Norris, K. D. Kovalcik,J. P. Pancras, and M. S. Landis. 2016. Near-road enhance-ment and solubility of fine and coarse particulate matter traceelements near a major interstate in Detroit, Michigan. Atmos.Environ. 145:213–24. doi:10.1016/j.atmosenv.2016.09.034.

Paatero, P. 1999. The multilinear engine – A table-driven,least squares program for solving multilinear problems,including the n-way parallel factor analysis model.J. Graph. Statist. 8:854–88.

Parvez, F., and K. Wagstrom. 2019. A hybrid modeling fra-mework to estimate pollutant concentrations and expo-sures in near road environments. Sci. Total Environ.663:144–53. doi:10.1016/j.scitotenv.2019.01.218.

Patterson, R. F., and R. A. Harley. 2019. Evaluatingnear-roadway concentrations of diesel-related air pollutionusing RLINE. Atmos. Environ. 199:244–51. doi:10.1016/j.atmosenv.2018.11.016.

Pope, C. A., and D. W. Dockery. 2006. Health effects of fineparticulate air pollution: Lines that connect. J. Air WasteManage. Assoc. 56:709–42. doi:10.1080/10473289.2006.10464485.

Rappenglück, B., G. Lubertino, S. Alvarez, J. Golovko,B. Czader, and L. Ackermann. 2013. Radical precursorsand related species from traffic as observed and modeled atan urban highway junction. J. Air Waste Manage. Assoc. 63(11):1270–86. doi:10.1080/10962247.2013.822438.

Rattigan, O. V., K. Civerolo, P. Doraiswamy, H. D. Felton,and P. K. Hopke. 2013. Long term black carbon measure-ments at two urban locations in New York. Aerosol AirQual. Res 13 (4):1181–96. doi:10.4209/aaqr.2013.02.0060.

Reid, S., S. Bai, Y. Du, K. Craig, G. Erdakos, L. Baringer,D. Eisinger, M. McCarthy, and K. Landsberg. 2016.Emissions modeling with MOVES and EMFAC to assess

244 M.C. MCCARTHY ET AL.

the potential for a transportation project to create particu-late matter hot spots. Transp. Res. Rec. 2570:12–20. STI-6330. doi:10.3141/2570-02.

Riddle, S. G., M. A. Robert, C. A. Jakober, M. P. Hannigan, andM. J. Kleeman. 2008. Size-resolved source apportionment ofairborne particle mass in a roadside environment. Environ.Sci. Technol. 42 (17):6580–86. doi:10.1021/es702827h.

Riley, E. A., L. Banks, J. Fintzi, T. R. Gould, K. Hartin,L. Schaal, M. Davey, L. Sheppard, T. Larson, M. G. Yost,et al. 2014. Multi-pollutant mobile platform measurementsof air pollutants adjacent to a major roadway. Atmos.Environ. 98:492–99. doi:10.1016/j.atmosenv.2014.09.018.

Rivas, I., D. Donaire-Gonzalez, L. Bouso, M. Esnaola,M. Pandolfi, M. de Castro, M. Viana, M. Àlvarez-Pedrerol, M. Nieuwenhuijsen, A. Alastuey, et al. 2016.Spatiotemporally resolved black carbon concentration,schoolchildren’s exposure and dose in Barcelona. IndoorAir 26 (3):391–402. doi:10.1111/ina.12214.

Rosenbaum, A. S., D. A. Axelrad, T. J. Woodruff, Y. H. Wei,M. P. Ligocki, and J. P. Cohen. 1999. National estimates ofoutdoor air toxics concentrations. J. Air Waste Manage. Assoc.49 (10):1138–52. doi:10.1080/10473289.1999.10463919.

Rowangould, G. M. 2013. A census of the US near-roadwaypopulation: Public health and environmental justiceconsiderations. Transp. Res. Part D Transport Environ.25:59–67. doi:10.1016/j.trd.2013.08.003.

Saha, P. K., A. Khlystov, and A. P. Grieshop. 2018a. Downwindevolution of the volatility and mixing state of near-roadaerosols near a US interstate highway. Atmos. Chem. Phys.18 (3):2139–54. doi:10.5194/acp-18-2139-2018.

Saha, P. K., A. Khlystov, M. G. Snyder, and A. P. Grieshop.2018b. Characterization of air pollutant concentrations,fleet emission factors, and dispersion near a NorthCarolina interstate freeway across two seasons. Atmos.Environ. 177:143–53. doi:10.1016/j.atmosenv.2018.01.019.

Skov, H., A. B. Hansen, G. Lorenzen, H. V. Andersen,P. Lofstrom, and C. S. Christensen. 2001. Benzene expo-sure and the effect of traffic pollution in Copenhagen,Denmark. Atmos. Environ. 35 (14):2463–71. doi:10.1016/S1352-2310(00)00460-X.

Smith, M. T. 2010. Advances in understanding benzene healtheffects and susceptibility. Annu. Rev. Public Health 31(1):133–48. doi:10.1146/annurev.publhealth.012809.103646.

Steffens, J., D. Heist, S. Perry, V. Isakov, R. Baldauf, andK. M. Zhang. 2014. Effects of roadway configurations onnear-road air quality and the implications on roadwaydesigns. Atmos. Environ. 94:74–85. doi:10.1016/j.atmosenv.2014.05.015.

Thomas, G. W., M. Ogletree, and L. Clay. 2016. Going onestep beyond in North Denver, a neighborhood scale airpollution modeling assessment, Part II: Predicted ambientconcentrations in 2035. Colorado Department ofTransportation. https://www.codot.gov/programs/research/pdfs/2007/goodneighbor.pdf.

Thomas, G. W., S. M. Williams, and D. L. Bain. 2007. Going onestep beyond: A neighborhood scale air toxics assessment innorth Denver (the Good Neighbor Project). Technical reportprepared for the The Colorado Department ofTransportation, Denver, CO, and the Federal HighwayAdministration, Colorado Division, Lakewood, CO, by theCity and County of Denver Department of EnvironmentalHealth, March 9. doi:10.1094/PDIS-91-4-0467B.

U.S. Environmental Protection Agency. 1996. Guidelines forcarcinogen risk assessment. Washington, DC: U.S. EPA.

U.S. Environmental Protection Agency. 1999a. Compendiumof methods for the determination of toxic organic com-pounds in ambient air: Compendium method TO-11A. 2nd

ed. Cincinnati, OH: prepared by the U.S. EnvironmentalProtection Agency, Office of Research and Development.EPA/625/R-96/010b, January. https://www3.epa.gov/ttnamti1/files/ambient/airtox/to-11ar.pdf.

U.S. Environmental Protection Agency. 1999b. Compendiumof methods for the determination of toxic organic com-pounds in ambient air: Compendium method TO-15. 2nded. Cincinnati, OH: prepared by the U.S. EnvironmentalProtection Agency, Office of Research and Development.EPA/625/R-96/010b, January. https://www3.epa.gov/ttnamti1/files/ambient/airtox/tocomp99.pdf.

U.S. Environmental Protection Agency. 2000. Control of airpollution from new motor vehicles: Tier 2 motor vehicleemissions standards and gasoline sulfur controlrequirements. Fed. Regist. 65 (28):6698–822. Final rule,40 CFR Parts 80, 85, and 86, February 10. http://www.epa.gov/otaq/regs/ld-hwy/tier-2/frm/fr-t2pre.pdf.

U.S. Environmental Protection Agency. 2007. Control ofhazardous air pollutants from mobile sources, Chapter3: Air quality and resulting health and welfare effects ofair pollution from mobile sources. Regulatory impactanalysis prepared by the Assessment and StandardsDivision, Office of Transportation and Air Quality, U.S.Environmental Protection Agency, Research TrianglePark, NC, EPA420-R-07-002, February. doi:10.1094/PDIS-91-4-0467B.

U.S. Environmental Protection Agency. 2014a. Control of airpollution from motor vehicles: Tier 3 motor vehicle emis-sion and fuel standards. 79 Fed. Regist. 81:23414–886.https://www.epa.gov/regulations-emissions-vehicles-and-engines/final-rule-control-air-pollution-motor-vehicles-tier-3#rule-history.

U.S. Environmental Protection Agency. 2014b. National airtoxics program: the second integrated urban air toxicsreport to Congress. August 21. https://www2.epa.gov/sites/production/files/2014-08/documents/082114-urban-air-toxics-report-congress.pdf.

U.S. Environmental Protection Agency. 2015a. 2011 NationalAir Toxics Assessment (NATA). December 17. https://www.epa.gov/national-air-toxics-assessment.

U.S. Environmental Protection Agency. 2015b. MOVES2014and MOVES2014a technical guidance: Using MOVES toprepare emission inventories for state implementationplans and transportation conformity. EPA-420-B-15-093,November. http://www3.epa.gov/otaq/models/moves/documents/420b15093.pdf.

U.S. Environmental Protection Agency. 2016. Air toxic emis-sions from on-road vehicles in MOVES2014. EPA-420-R-16-016, November. https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=P100PUNO.pdf.

U.S. Environmental Protection Agency. 2017. EJSCREENenvironmental justice mapping and screening tool:Technical documentation. August. https://www.epa.gov/ejscreen.

U.S. Environmental Protection Agency. 2018. Technical sup-port document: EPA’s 2014 National Air ToxicsAssessment. 2014 NATA TSD, August.

JOURNAL OF THE AIR & WASTE MANAGEMENT ASSOCIATION 245

U.S. Environmental Protection Agency. 2019. Our nation’sair: Status and trends through 2018. https://gispub.epa.gov/air/trendsreport/2019.

Urman, R., R. McConnell, T. S. Islam, E. L. Avol,F. W. Lurmann, H. Vora, W. S. Lin, E. B. Rappaport,F. D. Gilliland, and W. J. Gauderman. 2014. Associationsof children’s lung function with ambient air pollution:Joint effects of regional and near-roadway pollutants.Thorax 69 (6):540–47. doi:10.1136/thoraxjnl-2012-203159.

Viana, M., I. Rivas, C. Reche, A. S. Fonseca, N. Pérez,X. Querol, A. Alastuey, M. Álvarez-Pedrerol, andJ. Sunyer. 2015. Field comparison of portable and station-ary instruments for outdoor urban air exposureassessments. Atmos. Environ. 123:220–28. doi:10.1016/j.atmosenv.2015.10.076.

Watson, J. G., J. C. Chow, and E. M. Fujita. 2001. Review ofvolatile organic compound source apportionment by che-mical mass balance. Atmos. Environ. 35:1567–84.doi:10.1016/S1352-2310(00)00461-1.

Wilhelm, M., J. K. Ghosh, J. Su, M. Cockburn, M. Jerrett, andB. Ritz. 2011. Traffic-related air toxics and preterm birth:A population-based case-control study in Los Angeles

County, California. Environ. Health 10 (89). doi:10.1186/1476-069X-10-89.

Woodruff, T. J., D. A. Axelrad, J. Caldwell, R. Morello-Frosch,and A. Rosenbaum. 1998. Public health implications of 1990air toxics concentrations across the United States. Environ.Health Perspect. 106 (5):245–51. doi:10.1289/ehp.98106245.

Woodruff, T. J., J. Caldwell, V. J. Cogliano, andD. A. Axelrad. 2000. Estimating cancer risk from outdoorconcentrations of hazardous air pollutants in 1990.Environ. Res. 82 (3):194–206. doi:10.1006/enrs.1999.4021.

Zhu, Y., W. C. Hinds, S. Kim, S. Shen, and C. Sioutas. 2002a.Study of ultrafine particles near a major highway withheavy-duty diesel traffic. Atmos. Environ. 36(27):4323–35. doi:10.1016/s1352-2310(02)00354-0.

Zhu, Y. F., W. C. Hinds, S. Kim, and C. Sioutas. 2002b.Concentration and size distribution of ultrafine particlesnear a major highway. J. Air Waste Manage. Assoc. 52(9):1032–42. doi:10.1080/10473289.2002.10470842.

Zhu, Y. F., W. C. Hinds, S. Shen, and C. Sioutas. 2004. Seasonaltrends of concentration and size distribution of ultrafineparticles near major highways in Los Angeles. Aerosol Sci.Technol. 38:5–13. doi:10.1080/02786820390229156.

246 M.C. MCCARTHY ET AL.