Heliyon 5 (2019) e02236

Contents lists available at ScienceDirect

Heliyon
journal homepage: www.heliyon.com

A meta-analysis of selected near-road air pollutants based on concentration

decay rates
Shi V. Liu *, Fu-lin Chen, Jianping Xue **
National Exposure Research Laboratory, Office of Research and Development, U. S. Environmental Protection Agency, RTP, NC, 27711, USA

A R T I C L E I N F O A B S T R A C T

Keywords: Traffic-related air pollution has been associated with various health risks for human populations living near
Environmental sciences roadways. Understanding the relationship between traffic density and dispersion of vehicle-released air pollutants
Atmospheric science is important for assessing human exposure to near-road air pollutants. We performed a literature survey targeting
Near-road
publications containing measurement data of traffic-related air pollutants near roads with distance information on
Concentration decay
Meta-data analysis
their concentration distribution. Concentration decay rates over down-wind distance away from major roads were
Traffic calculated for black carbon (BC), carbon monoxide (CO) and nitrogen oxides (NO2 or NOx) and meta-data analysis
Air pollution on these rates was performed. These analyses showed metadata-based exponential decay rates of 0.0026, 0.0019,
Distance 0.0004, and 0.0027 m
1 for BC, CO, NO2 and NOx, respectively. Using these measurement data-based decay rates,
Vehicle emission concentrations for BC, CO, NO2 and NOx over various near-road distances were predicted. These results are useful
for enhancing exposure modeling and thus more reliably assessing the health risk of exposure to near road air
pollution.

1. Introduction on common spatial patterns of vehicle-emitted pollutants [19, 20, 21].

These general understandings for near-road air pollution patterns are
Petroleum-fueled motor vehicles release regulated air pollutants such useful in many ways [22, 23]. For example, they can be used for assisting
as black carbon (BC), carbon monoxide (CO), nitric oxides (NOx), and policy-making especially in areas lacking monitoring data [24]. They can
particulate matter (PM) [1, 2, 3, 4, 5]. The health impact of these mobile also be used for evaluating development-planning so that traffic pollution
source air pollutants on human populations has been reflected with can be minimized if not eliminated [25].
increasing observations on elevated adverse health effects among Such generalization can be achieved through meta-data analysis on
near-road residents [6, 7, 8, 9, 10, 11, 12]. With an increasing trend of large data sets consolidated from multiple studies. A meta-analysis has
urbanization in human societies and a growing dependency on motor tried to evaluate factors influencing the spatial extent of mobile source
vehicles for economic functions, traffic-related air pollution is becoming air pollution [26]. It showed a complex interplay of spatial extent defi-
more and more significant [13, 14]. nitions, emission rates, background concentrations, and meteorological
Compared with point-source air pollution associated with a fixed conditions on spatial extent estimates even for non-reactive pollutants. It
location, mobile source air pollution is more difficult to evaluate. Many calls for “high resolution” modeling near the source and developing a
studies have been carried out to characterize the spatial distribution of “zone of influence” for mobile sources.
mobile source air pollutants [15, 16, 17]. Most of these studies report We performed a new meta-data analysis in an aim to address the
monitoring data for selected air pollutants at a specific location [18]. needs identified in the previous meta-data analysis and, with a more
These studies are valuable for site-specific discussion of potential expo- comprehensive coverage of different studies and a more detailed data
sure of local residents to the studied air pollutants. Unfortunately, such analysis, intending to reach some refined understanding of spatial dis-
extensive studies are simply too expensive to be carried out in many tribution of traffic-related air pollutants. We found that, by using
communities where near-road air pollution is prominent. pollutant concentration decay rate as a common parameter for evalua-
There is an urgent need for developing some generalized knowledge tion, some spatial profiles can be seen more clearly for traffic-related air

* Corresponding author.
** Corresponding author.
E-mail addresses: [email protected] (S.V. Liu), [email protected] (J. Xue).

https://doi.org/10.1016/j.heliyon.2019.e02236
Received 31 October 2017; Received in revised form 31 August 2018; Accepted 2 August 2019
2405-8440/Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
S.V. Liu et al. Heliyon 5 (2019) e02236

Table 1 Table 2
Literature search criteria and results. Near-road distance-concentration profiles of selected air pollutants.
Step Search criteria and procedures Database searched and library Results Pollutant Study Distance Concentration Note Ref.
created Range, m Range

1 (air pollution measurement) ProQuest Environmental 6021 BC 26 20–300 1.52–0.78 All direction [17]
AND (major road OR highway) Science Collection BC 26 20–300 1.68–0.82 Downwind [17]
AND (distance OR spatial Was searched using criteria BC 35 5–450 3.5–1.8 upwind [30]
distribution) in anywhere specified. BC 35 5–500 5–2 Downwind [30]
2 Above criteria plus “NOT *” ProQuest Environmental 3799 BC 48 0–60 9–1.4 Downwind [31]
where * is “model”, “health Science Collection BC 55 15–305 14.9–7.4 Location Delft, [32]
effect”, or “land use” Was searched with different Downwind
exclusion terms and then BC 55 32–260 12.2–8.7 Location [32]
results are combined into an Overschie,
EndNote library. Downwind
3 Endnote library was searched Groups containing publication 1879 BC 65 35–300 5.5–1 405 Freeway, [33]
for specific air pollutants such matching each specific air Downwind
as “black carbon”, “carbon pollutant were created within BC 65 15–300 22–5 710 freeway, [33]
monoxide”, “nitrogen oxides”, Endnote library. Downwind
or “particulate” BC 65 17–300 21.7–5.5 Both freeway, [33]
4 Title reading by three Publications passing 373 Downwind
researchers independently concurrent agreement of three CO 10 10–1450 195–30 SH – 71, Parallel [34]
researchers' screening were CO 10 150–390 250–130 I – 35, Parallel [34]
collected into a separate group CO 10 250–1750 75–60 FM – 973, Parallel, [34]
within the EndNote library. Upwind
5 Independent reading of Publications containing useful 65 CO 10 5–1100 190–60 FM –973, Parallel, [34]
abstracts by three researchers measurement data on selected Downwind
air pollutant were collected CO 10 20–25 150–140 I – 35, [34]
into final concurrence set. Perpendicular,
Downwind
CO 10 75–175 75–45 FM – 973, [34]
pollutants near roads. The concentration decay rate-based analysis will Perpendicular,
Upwind
provide some detailed knowledge for more confident estimates on near-
CO 10 5–220 125–60 FM – 973, [34]
road spatial distribution of traffic-related air pollutants. This knowledge Perpendicular,
will be very important for evaluating exposure levels of human pop- Downwind
ulations living at various distances to the mobile source of air pollutants CO 20 20–150 210–180 S1 – S2, Upwind [16]
[24, 25]. CO 20 40–150 400–320 N1 – N4, [16]
Downwind
CO 26 20–300 340–270 All direction [17]
2. Materials and methods CO 26 20–300 400–300 West [17]
CO 44 25–1000 500–300 SW, Upwind [35]
2.1. Literature search CO 44 25–325 400–300 NE, Downwind [35]
CO 49 9–53 850–500 Upwind [36]
CO 49 0–80 1050–460 Downwind [36]
Using ProQuest we searched its Environmental Science Collection CO 65 35–300 2000–300 405 Freeway, [33]
databases for broad scope terms that might hit publications containing Downwind
information on distance distribution of air pollutants near roadways (see CO 65 15–300 2300–100 710 Freeway, [33]
Table 1 for search criteria). Because our goal was to get publications Downwind
CO 65 17–300 2300–200 Both Freeway, [33]
containing concentration measurement data on near-road air pollutants Downwind
we searched the initial matches again to exclude those reporting models, NO2 10 10–1500 5.5–1.5 SH – 71, Parallel [34]
health effects or land uses. From those matched findings we then focused NO2 10 100–400 10.5–3 I – 35, Parallel [34]
on a selected set of air pollutants which included BC, CO, NO2, NOx, and NO2 10 100–1000 3.5–2 FM – 973, Parallel, [34]
Upwind
PM (PM2.5 and PM10). Publications matching these selections were
NO2 10 10–1100 9–2 FM – 973, Parallel, [34]
exported into an EndNote library. Then titles of these publications were Downwind
manually screened by three researchers independently to identify NO2 10 20–25 13.5–13 I – 35, [34]
genuine matches. The results of this screening were combined and the Perpendicular
abstracts of each paper were again independently read by three re- NO2 10 90–150 2–2 FM – 973, [34]
Perpendicular,
searchers to reach the final set of papers containing measurement data on Upwind
at least one of the selected air pollutants. NO2 10 10–220 8–3 FM – 973, [34]
Perpendicular,
Downwind
2.2. Data extraction and collection NO2 14 50–200 20–13 Upwind [34]
NO2 14 50–400 30–18 Downwind [34]
A master spreadsheet was created to collect pollutant concentrations NO2 15 0–950 32–12 Upwind [20]
over distances from roads. Original numerical data contained in a pub- NO2 15 0–1300 32–15 Downwind [20]
NO2 26 20–300 24.1–18.69 All direction [17]
lication or obtained from authors were used when they were available
NO2 26 20–300 27.31–20.95 West [17]
but most times estimated values obtained from published figures were NO2 29 4–48 30–15 [38]
used. Because initial inspection on the extracted data showed no obvious NO2 35 5–450 30–25 Upwind [30]
concentration decay over distance away from road for PM which did not NO2 35 5–500 32–25 Downwind [30]
include ultrafine PM we limited our continued data compilation only to NO2 36 30–60 5.4–4.5 Location A [39]
NO2 36 10–30 5.8–5.6 Location B [39]
BC, CO, NO2, and NOx. In addition to extracting data from publications NO2 36 10–30 6.9–6.6 Location C [39]
found in the literature search we also obtained original data from large NO2 48 0–60 70.5–59.2 Downwind [31]
EPA-sponsored measurement studies such as the Las Vegas study [27] (continued on next page)
and the near-road exposures and effects of urban air pollutants study

2
S.V. Liu et al. Heliyon 5 (2019) e02236

Table 2 (continued ) Table 3

Pollutant Study Distance Concentration Note Ref.
Decay rates of air pollutants in different ranges of distance from road.
Range, m Range Pollutant Study* Decay Rate (m
1)
NO2 55 15–305 47.8–30.6 Location Delft, [32] 20~100 m 100~300 m 20~300 m
Downwind
NO2 55 32–260 44.8–32.1 Location [32] BC LV 0.0028 0.0015 0.0018
Overschie, DX 0.0049 0.0013 0.0024
Downwind O35 0.0023 0.0018 0.0024
NOx 10 10–1500 17–3 SH – 71, Parallel [34] O48 0.0123 0.0061 0.0073
NOx 10 100–400 25–5 I – 35, Parallel [34] O55 0.0048 0.0011 0.0024
NOx 10 100–1000 7–3 FM – 973, Parallel, [34] O65 0.0130 0.0017 0.0045
Upwind CO LV 0.0012 0.0004 0.0009
NOx 10 10–1100 32–3 FM – 973, Parallel, [34] DX 0.0048 0.0002 0.0011
Downwind O10-FM-973 0.0054 0.0002 -0.0034
NOx 10 20–25 38–33 I -35, [34] O10-SH-71 0.0069 0.0004 0.0050
Perpendicular, O44 0.0014 0.0001 0.0007
Downwind O49 0.0120 NA NA
NOx 10 90–150 4–3 FM – 973, [34] O65 0.0198 0.0044 0.0082
Perpendicular, NO2 LV 0.0018 0.0008 0.0011
Upwind DX 0.0020 -0.0004 0.0004
NOx 10 10–220 40–5 FM – 973, [34] O10 0.0064 -0.0013 0.0029
Perpendicular, O15 0.0041 0.0014 0.0021
Downwind O26 0.0017 0.0006 0.0009
NOx 14 50–200 35–25 I – 93, Upwind [37] O35 0.0014 -0.0003 0.0002
NOx 14 50–400 65–30 I -93, Downwind [37] O55 0.0039 0.0005 0.0011
NOx 26 20–300 24.1–18.69 All direction, first [17] NOx LV 0.0035 0.0010 0.0018
period DX 0.0087 0.0002 0.0028
NOx 26 20–300 27.31–20.95 West, first period [17] O10 0.0225 0.0031 0.0064
NOx 26 20–300 46.89–30.64 All direction, [17] O26 0.0033 0.0027 0.0015
Second period O44 0.0029 0.0008 0.0023
NOx 26 20–300 56.08–34.86 West, Second [17] O49 0.0072 0.0037 NA
period *
Studies identified as “LV” and “DX” are for Las Vegas study [28] and the
NOx 36 30–60 8.7–6.6 Location A [39]
near-road exposures and effects of urban air pollutants study (NEXUS) [29],
NOx 36 10–30 11.5–9.7 Location B [39]
respectively. Other studies identified as “O##“, plus additional suffix to identify
NOx 36 10–30 10.4–10.1 Location C [39]
NOx 44 25–550 50–28 SouthWest [35] sub-sets of data within the same study (see Table 2 for details).
NOx 44 25–325 35–20 Northeast [35]
NOx 48 0–60 189.9–93.6 [31] 3. Results
NOx 49 0–53 200–80 Upwind [36]
NOx 49 0–80 200–100 Downwind [36]
3.1. Scientific interest and research depth on near road air pollution
Concentration as ug/m3 for BC and ppb for CO, NO2 and NOx, respectively.
With our initial search criteria we found 6021 publications in the
(NEXUS) [28]. broad search performed in July 2013 (Table 1). However, as our focus in
this study was to better understand the spatial extent or dynamics of
2.3. Data analysis near-road air pollutants, publications dealing with models, health effects,
or land use were excluded from the initial matches and only 3799 pub-
Pollutant concentration decay rate (R) was calculated according to lications were retained. With a further limitation of our focus on BC, CO,
the following formula: NO2, NOx, and PM we narrowed our collection to 1879 publications
R ¼ Ln (Ca/Ci)/D, where R is the decay rate, D is the distance be- which were imported into an Endnote library. Three researchers inde-
tween pollutant's concentrations at the initial measurement site (Ci) and pendently reviewed the titles of this collection to make their respective
after a distance (Ca). Ln represents natural logarithm. selections. Then a total of 373 publications were collected after discus-
The above two EPA-sponsored near-road air pollutant measurement sion. The abstracts of this collected set were independently read by three
studies were analyzed first using SAS program to see concentration dis- researchers and a final set of 65 publications were found to contain
tribution pattern of BC and to obtain means and standard deviations of measurement data on at least one of the selected air pollutants with a
air pollutant concentrations over monitoring distances. Pollutants’ decay kind of distance profile from major road.
rates were calculated for distance range 20–100 and 100–300 m as well
as 20–300 m.
3.2. Distance and concentration ranges of selected air pollutant in various
Meta-data analysis on calculated concentration decay rates was per-
studies
formed according to random model described [29]. Specifically, a
random effects model was used for the meta-analyses. In the Random
Initial evaluation of those 65 publications indicated that PMs showed
effects model, the true effect could vary from study to study while one
no apparent concentration decay over within near-road distances. Thus
true effect size was assumed in the fixed-effect model. Mean, standard
we focused our spatial profile analysis of concentration decay only on BC,
error and sample size were used for meta-analyses to calculate the overall
CO, NO2 and NOx. We found 12 unique studies contained sufficient
decay rate across all studies. Missing values of standard deviation for
concentration and distance measurement information for our concen-
decay rates in literature-extracted studies were filled in with equivalent
tration decay analysis (Table 2).
values derived from analysis on two large data sets of original mea-
Most of the collected studies showed measurement distance ranges up
surement data from the above two EPA-sponsored studies.
to 300 m while a few studies showed measurement distance up to 1500
m. In general, larger distance ranges were covered by traffic-related air
pollutants in down-wind direction than those in up-wind direction.
However, the actual concentration ranges varied greatly among different
studies, indicating complex effects of local landscape, road construction,

3
S.V. Liu et al. Heliyon 5 (2019) e02236

Fig. 1. Meta-data analysis of decay rates for selected traffic-related air pollutants. The different types of air pollutants are shown in separate plots for Black carbon
(BC), carbon monoxide (CO), nitrogen dioxide (NO2) and nitrogen oxides (NOx), respectively. A red line with open circle represents the 95% confidence intervals (CI)
and the mean each individual study. Diamond under each set of distance range represents the average of a meta-analysis for that distance rage. The diamond aligned
with “All range mean” at the bottom of each plot refers to meta-analysis on all three ranges for a given pollutant. Study IDs are as explained in Table 3.

traffic density, and meteorology. 3.4. Meta-data analysis on decay rates

3.3. Decay rate of air pollutants in near-road distances Using a random model of meta-data analysis, we showed the means
and their confident interval of air pollutant concentration decay rates
Initial inspection of a selected group of measurement-based studies among different studies and the overall from meta-analyses (Fig. 1). In
revealed that traffic-related air pollutants often had a larger impact range general, smaller variations in the decay rates were seen in most studies on
in the down-wind than in the up-wind area. Thus, we selected only down- BC than the other air pollutants. The variation of average decay rates in
wind measurement data for calculating decay rates for a selected set of individual studies were statistically significant for BC but not for other air
air pollutants. Based on our initial analysis on rich measurement data pollutants (data not shown). However, the decay rate variations were
obtained from EPA-sponsored near-road air pollution study in Las Vegas, significant on the meta-data analysis level for all air pollutants in the
we found that concentration decay of BC was better described with two examination.
different rates for the 20–100- and the 100–300-meter distance ranges, The meta-data analysis-based decay rates for BC are 0.0038, 0.0016,
respectively, rather than with a single rate covering the entire distance and 0.0025 m
1 for distance ranges 20–100, 100–300, and 20–300 m,
range. Thus, we used 20–100- and 100–300-meter distance ranges for respectively. The meta-data analysis-based decay rates for CO are 0.0030,
separate decay rate calculations in addition to a whole-range decay rate 0.0019, and 0.0011 m
1 for distance ranges 20–100, 100–300, and
calculation for all other studies (Table 3). 20–300 m, respectively. The meta-data analysis-based decay rates for
The mean decay rates for BC are always larger in distance range NOx and NO2 are 0.0047 and 0.0019, 0.0008 m
1 and 0.0002, and
20–100 m than in distance range 100–200 m while the whole distance 0.0025 and 0.0005 m
1 for distance ranges 20–100, 100–300, and
range (20–300 m) showed an intermediate value. The mean decay rates 20–300 m, respectively.
for CO are always largest in distance range 20–100 m. But the mean The overall meta-data analysis-based decay rates are 0.0026, 0.0019,
decay rates in distance range 100–200 m are sometimes larger than the 0.0004, and 0.0027 m
1 for BC, CO, NO2 and NOx, respectively (Fig. 1).
values for the whole distance range (20–300 m). The mean decay rates
for NOx and NO2 are always larger in distance range 20–100 m than in 3.5. Decay rate-based concentration prediction over distance from road
distance range 100–200 m while the whole distance range (20–300 m)
showed intermediate values. Having obtained meta-data analysis-derived air pollutant decay rates

4
S.V. Liu et al. Heliyon 5 (2019) e02236

Fig. 2. Near-road concentration predictions for selected traffic-related air pollutants. Legends x12, x13, and x23 inside each plot represents prediction based on meta-
data analysis-derived decay rates for 20–100-m, 20–300-m, and 100–300-m, respectively.

based on real measurement data from multiple studies it is convenient Further evaluation of these 65 publications indicated that PMs
and also useful to make some predictions on distance-concentration showed no apparent concentration decay over near-road distances, we
profiles for the selected air pollutants. We used an initial concentration focused our spatial profile analysis of concentration decay only on BC,
of 5000 ug/m3 for BC or 5000 ppb for CO, NO2 and NOx and predicted CO, NO2 and NOx. We found 12 unique studies contained sufficient
their concentration decay patterns when different decay rates were used concentration and distance measurement information for our concen-
(Fig. 2). tration decay analysis (Table 2).
Based on these predictions it is calculated that about 32% BC, 26% A major problem in comparing different studies on near-road air
CO, 17% NO2, and 37% NOx would be decreased at 100 m from the road pollutants is the widely varied initial on-road or road-edge concentra-
and about 59% BC, 58% CO, 21% NO2, and 52% NOx would be reduced tions in different studies. This variation greatly influences the spatial
at 300 m from the road. extent of the traffic emission-released air pollutants. A study [40] used
two different normalization methods (background vs. edge normaliza-
3.6. Concentration prediction with confident interval among different tion) to investigate the near-road air pollutant concentration decay. It
decay rates was found that, for almost all pollutants, the influence of the roadway on
air pollution concentrations decays to background was 115–570 m and
The concentration decay of mobile-source air pollutants is faster in 160–570 m according to the background normalization and the edge
the near road distance range than the further away from road distance normalization, respectively. This study also found that the changes in
range. Thus, using different rates for concentration calculation will have pollutant concentrations with increasing distance from the road fell into
some effect on the scope of the prediction. To illustrate this rate-selection one of three groups: rapid decay (e.g., CO), consistent decay (e.g., NO2)
effect on concentration prediction we present the concentration decay and no apparent decay (e.g., PM).
patterns and the 95% confidence intervals of BC (Fig. 3). Thus, it appeared that using concentration decay rates may be a better
way to describe “spatial dynamics” of air pollutants, regardless of
4. Discussion methods of concentration normalization [41]. Pollutant concentration
decay rates are often inclusive for many confounding factors and thus
Near-road air pollution is a very hot topic in environmental research many variations such as local land use, season and time are absorbed into
which could be reflected with an easy retrieval of more than 6000 pub- decay rates without a need to differentiate their compounding impacts. A
lications matching general terms on this topic. However, when we tried previous modeling study of concentrations of nitrogenous compounds
to find specific information on spatial extent or spatial patterns of traffic- with distance from roads has led to a selection of exponential decay
related air pollutants near roads, we were disappointed to learn that model for meaningful reflection of the measurement data [20]. Another
many publications did not contain measurement data with desired spatial study has found a logarithmic relationship between NO2 concentration
resolution. This limitation could reflect the difficulties in initiating near- and the distance from a highroad [21]. Some more complicated models
road measurement studies which would be cost-expensive at least. In have been used for characterizing air dispersion of CO and diesel par-
comparation with a previous study which showed 33 studies in the ticulate matter (DPM) [22].
literature meeting its selection criteria for examining spatial extent of In this study we used a simple exponential decay model to charac-
mobile source air pollution [26] and an important study which synthe- terize the concentration decay of all selected air pollutants. Because
sized findings on near-roadway air quality collected data from 41 road- many roads are built in such a way with clear buffering zone near the
side monitoring studies in publications from 1978 to 2008 [40], our roads and more building structures are found with increasing distance
study was initially based on 65 publications with some measurement from the roads, we calculated concentration decay rate in three different
data (Table 1). distance ranges as 20–100 m, 100–300 m and 20–300 m. We found that

5
S.V. Liu et al. Heliyon 5 (2019) e02236

Fig 3. Confidence intervals of black carbon concentration predictions based on different decay rates. Meta-data analysis-derived decay rates for distance ranges
20–100-m, 20–300-m, and 100–300-m are used for making predictions shown at the top, middle, and bottom plots, respectively. The mean and the 95% upper and low
limits are represented by solid and dashed lines, respectively.

the fitting of this model to the spatial patterns of the traffic-related air concentrations can be found or estimated.
pollutants was better if the concentration decay was calculated using In comparing the concentration decay patterns of different air
different decay rates for the closer-to-road (20–100 m) and the further pollutant in the selection we found that BC appeared to be the “best in-
away-from-road (100–300 m) distance ranges, rather than using a single dicator” for near-road air pollution modeling because its much “tighter”
rate to cover the entire distance range (20–300 m). This phenomenon dispersion trend in the meta-data analysis than those revealed for the
may be a reflection that traffic emission-released air pollutants in the other air pollutants such as CO, NO2 and NOx. CO showed a “reversed”
near-road clear buffering zone (20–100 m) and more distance from the decay rate for having a faster decay rate for the whole distance range
road (100–300 m) were influenced by different sets of factors which than the far-way distance range. The reason for this anomaly in this study
shaping their concentration dynamics. is unknown but it could be a reflection of limited data with extreme
Based on these model calculations it is predicted that about 32% and outliers.
59% of BC would decreased within 100- and 300-meters, respectively. We should point out that, by applying a very restricted set of selection
About 26% and 58% of CO would disappear within 100- and 300-meters, criteria and limiting our analysis only on down-wind measurement data,
respectively. About 17% and 21% of NO2 and about 37% and 52% of the results of our meta-analysis would certainly not be an inclusive
NOx would disappear within 100- and 300-meters, respectively. Such reflection of some different spatial patterns for near-road air pollutants
information would be valuable in assessing the exposure potential of that may exist in certain locations. The utilization of a simple exponential
these air pollutants once their emission rate and thus initial equation for calculating decay rate is a kind of arbitrary. Thus, provision

6
S.V. Liu et al. Heliyon 5 (2019) e02236

of decay rates calculated for different distance ranges may be necessary [6] A.J. Badyda, P. Dabrowiecki, W. Lubinski, P.O. Czechowski, G. Majewski, Exposure
to traffic-related air pollutants as a risk of airway obstruction, Adv. Exp. Med. Biol.
and useful for exposure assessment and development planning in
755 (2013) 35–45.
different urban communities. [7] M.S. Brown, S.E. Sarnat, K.A. DeMuth, L.A. Brown, D.R. Whitlock, S.W. Brown,
P.E. Tolbert, A.M. Fitzpatrick, Residential proximity to a major roadway is
associated with features of asthma control in children, PLoS One 7 (5) (2012)
5. Conclusion
e37044.
[8] N.A. Clark, P.A. Demers, C.J. Karr, M. Koehoorn, C. Lencar, L. Tamburic, M. Brauer,
With a focused literature search for near-road studies on air pollutants Effect of early life exposure to air pollution on development of childhood asthma,
Environ. Health Perspect. 118 (2) (2010) 284–290.
with concentration measurements and distance records we performed a
[9] W.Q. Gan, M. Koehoorn, H.W. Davies, P.A. Demers, L. Tamburic, M. Brauer, Long-
concentration decay rate-based metaanalysis for major traffic-related air Term exposure to traffic-related air pollution and the risk of coronary heart disease
pollutants. In general, BC, CO, NO2 and NOx followed the same pattern of hospitalization and mortality, Environ. Health Perspect. 119 (4) (2011) 501–507.
[10] H. Gül, E.O. Gaga, T. D€ ogeroglu, O.€ Ozden,
€ € Ayvaz, S. Ozel,
O. € G. Güng€or,
logarithmic concentration decay. BC appeared to be an ideal candidate
Respiratory health symptoms among students exposed to different levels of air
for near-road air pollution modeling because its consistent dispersion pollution in a Turkish city, Int. J. Environ. Res. Public Health 8 (4) (2011)
trend reflected in the meta-data analysis. This information are useful for 1110–1125.
enhancing exposure modeling studies on near-road traffic air pollutants [11] U. Kr€amer, C. Herder, D. Sugiri, K. Strassburger, T. Schikowski, U. Ranft,
W. Rathmann, Traffic-related air pollution and incident type 2 diabetes: results
and thus more reliably assessing the health risk of near road air pollution. from the SALIA cohort study, Environ. Health Perspect. 118 (9) (2010) 1273–1279.
[12] E.E. Nkwocha, R.O. Egejuru, Effects of industrial air pollution on the respiratory
Declarations health of children, Int. J. Environ. Sci. Technol. 5 (4) (2008) 509–516.
[13] C. Borrego, O. Tchepel, A.M. Costa, H. Martins, J. Ferreira, A.I. Miranda, Traffic-
related particulate air pollution exposure in urban areas, Atmos. Environ. 40 (37)
Author contribution statement (2006) 7205–7214.
[14] Y. Qin, L.Y. Chan, Traffic source emission and street level air pollution in urban
areas of Guangzhou, South China (P.R.C.), Atmos. Environ. Part B – Urban Atmos.
Shi V. Liu, Fu-lin Chen, Jianping Xue: Conceived and designed the 27B (3) (1993) 275–282.
experiments; Performed the experiments; Analyzed and interpreted the [15] R. Baldauf, E. Thoma, M. Hays, et al., Traffic and meteorological impacts on near-
data; Contributed reagents, materials, analysis tools or data; Wrote the road air quality: summary of methods and trends from the raleigh near-road study,
J. Air Waste Manag. Assoc. 58 (7) (2008) 865–878.
paper. [16] G.S.W. Hagler, E.D. Thoma, R.W. Baldauf, High-resolution mobile monitoring of
carbon monoxide and ultrafine particle concentrations in a near-road environment,
J. Air Waste Manag. Assoc. 60 (3) (2010) 328–336.
Funding statement [17] S. Kimbrough, R.W. Baldauf, G.S.W. Hagler, R.C. Shores, W. Mitchell,
D.A. Whitaker, C.W. Croghan, D.A. Vallero, Long-term continuous measurement of
This research did not receive any specific grant from funding agencies near-road air pollution in Las Vegas: seasonal variability in traffic emissions impact
on local air quality, Air Qual. Atmos. Health 6 (1) (2013) 295–305.
in the public, commercial, or not-for-profit sectors.
[18] Y. Zhu, J. Pudota, D. Collins, D., et al., Air pollutant concentrations near three Texas
roadways, Part I: ultrafine particles, Atmos. Environ. 43 (30) (2009) 4513–4522.
[19] R.E. Ducret-Stich, M.-Y. Tsai, M.S. Ragettli, A. Ineichen, N. Kuenzli, H.C. Phuleria,
Competing interest statement
Role of highway traffic on spatial and temporal distributions of air pollutants in a
Swiss Alpine valley, Sci. Total Environ. 456–457 (2013) 50–60.
The authors declare no conflict of interest. [20] T. Faus-Kessler, M. Kirchner, G. Jakobi, Modelling the decay of concentrations of
nitrogenous compounds with distance from roads, Atmos. Environ. 42 (19) (2008)
4589–4600.
Additional information [21] H. Pleijel, G. Pihl Karlsson, E. Binsell Gerdin, On the logarithmic relationship
between NO2 concentration and the distance from a highroad, Sci. Total Environ.
332 (1–3) (2004) 261–264.
No additional information is available for this paper. [22] J. Lin, D. Yu, Traffic-related air quality assessment for open road tolling highway
facility, J. Environ. Manag. 88 (4) (2008) 962–969.
[23] T.M. Barzyk, B.J. George, A.F. Vette, R.W. Williams, C.W. Croghan, C.D. Stevens,
Acknowledgements
Development of a distance-to-roadway proximity metric to compare near-road
pollutant levels to a central site monitor, Atmos. Environ. 43 (4) (2009) 787–797.
We thank Carry Croghan and Daniel Vallero in National Exposure [24] J. Cyrys, M. Hochadel, U. Gehring, G. Hoek, et al., GIS-based estimation of exposure
Research Laboratory (NERL) and Sue Kimbrough in National Risk Man- to particulate matter and NO2 in an urban area: stochastic versus dispersion
modeling, Environ. Health Perspect. 113 (8) (2005) 987–992.
agement Research laboratory (NRMRL) of US EPA Office of Research and [25] X. Han, L.P. Naeher, A review of traffic-related air pollution exposure assessment
Development for providing large data sets on near-road air pollutant studies in the developing world, Environ. Int. 32 (1) (2006) 106–120.
measurements in Las Vegas and Detroit, respectively. We also thank [26] Y. Zhou, J.I. Levy, Factors influencing the spatial extent of mobile source air
pollution impacts: a meta-analysis, BMC Public Health 7 (2007) 89.
Janet Burke and Lisa Baxter of NERL for reviewing this manuscript. [27] E.S. Kimbrough, R.W. Baldauf, N. Walkins, Seasonal and dirrnal analysis of NO2
The United States Environmental Protection Agency (EPA) through its concentrations from a long-duration study conducted in Las Vegas, Nevada, J. Air
Office of Research and Development conducted the research described in Waste Manag. Assoc. 63 (8) (2013) 934–942.
[28] A. Vette, J. Burke, G. Norris, et al., The near-road exposures and effects of urban air
this paper. This manuscript has been reviewed by EPA and approved for pollutants study (NEXUS): study design and methods, Sci. Total Environ. 448
publication. The views expressed in this article are those of the authors (2013) 38–47.
and do not necessarily represent the views or policies of the EPA. Any [29] M. Borenstein, L.V. Hedges, J.P.T. Higgins, H.R. Rothstein, Introduction to Meta-
Analysis, Wiley, West Sussex, United Kingdom, 2009.
mention of trade names should not be construed as an endorsement. [30] P. Massoli, E.C. Fortner, M.R. Canagaratna, et al., Pollution gradients and chemical
characterization of particulate matter from vehicular traffic near major roadways:
References results from the 2009 queens college air quality study in NYC, Aerosol Sci. Technol.
46 (11) (2012) 1201–1218.
[31] L. Pirjola, T. L€ahde, J.V. Niemi, et al., Spatial and temporal characterization of
[1] J. Cohen, R. Cook, C.R. Bailey, E. Carr, Relationship between motor vehicle
traffic emissions in urban microenvironments with a mobile laboratory, Atmos.
emissions of hazardous pollutants, roadway proximity, and ambient concentrations
Environ. 63 (2012) 156–167.
in Portland, Oregon, Environ. Model. Softw 20 (1) (2005) 7–12.
[32] M.C. Roorda-Knape, N.A.H. Janssen, J.J. De Hartog, et al., Air pollution from traffic
[2] H.-D. He, W.-Z. Lu, Spectral analysis of vehicle pollutants at traffic intersection in
in city districts near major motorways, Atmos. Environ. 32 (11) (1998) 1921–1930.
Hong Kong, Stoch. Environ. Res. Risk Assess. 26 (8) (2012) 1053–1061.
[33] Y. Zhu, W.C. Hinds, S. Kim, S. Shen, C. Sioutas, Study of ultrafine particles near a
[3] W. Kam, J.W. Liacos, J.J. Schauer, R.J. Delfino, C. Sioutas, On-road emission factors
major highway with heavy-duty diesel traffic, Atmos. Environ. 36 (27) (2002)
of PM pollutants for light-duty vehicles (LDVs) based on urban street driving
4323–4335.
conditions, Atmos. Environ. 61 (2012) 378–386.
[34] A.L. Clements, Y. Jia, A. Denbleyker, et al., Air pollutant concentrations near three
[4] L. Pirjola, et al., “Sniffer”—a novel tool for chasing vehicles and measuring traffic
Texas roadways, part II: chemical characterization and transformation of pollutants,
pollutants, Atmos. Environ. 38 (22) (2004) 3625–3635.
Atmos. Environ. 43 (2009) 4523–4534.
[5] Q. Zhang, G. Sun, S. Fang, W. Tian, X. Li, H. Wang, Air pollutant emissions from
vehicles in China under various energy scenarios, Sci. Total Environ. 450 (2013)
250–258.

7
S.V. Liu et al. Heliyon 5 (2019) e02236

[35] L.T. Padro-Martinez, A.P. Patton, J.B. Trull, et al., Mobile monitoring of particle [39] K. McAdam, P. Steer, K. Perrotta, Using continuous sampling to examine the
number concentration and other traffic-related air pollutants in a near-highway distribution of traffic related air pollution in proximity to a major road, Atmos.
neighborhood over the course of a year, Atmos. Environ. 61 (2012) 253–264. Environ. 45 (12) (2011) 2080–2086.
[36] L. Pirjola, P. Paasonen, D. Pfeiffer, et al., Dispersion of particles and trace gases [40] A. Karner, et al., Near-roadway air quaity: synthesizing the findings from real-world
nearby a city highway: mobile laboratory measurements in Finland, Atmos. data, Environ. Sci. Technol. 44 (2010) p5334–5344.
Environ. 40 (5) (2006) 867–879. [41] HEI Panel on the Health Effects of Traffic-Related Air Pollution, Traffic-Related Air
[37] J.L. Durant, C.A. Ash, E.C. Wood, et al., Short-term variation in near-highway air Pollution: A Critical Review of the Literature on Emissions, Exposure, and Health
pollutant gradients on a winter morning, Atmos. Chem. Phys. 10 (2) (2010) Effects. HEI Special Report 17, 2010. Available at, https://www.healtheffects.org
5599–5626. /publication/traffic-related-air-pollution-critical-review-literature-emissions-e
[38] S. Lal, R.S. Patil, Monitoring of atmospheric behaviour of NOx from vehicular xposure-and-health.
traffic, Environ. Monit. Assess. 68 (1) (2001) 37–50.