Science of the Total Environment 550 (2016) 940–949

Contents lists available at ScienceDirect

Science of the Total Environment

journal homepage: www.elsevier.com/locate/scitoten v

Influence of quarry mining dust on PM2.5 in a city adjacent to a limestone

quarry: Seasonal characteristics and source contributions
Xing Peng a, Guo-Liang Shi a,⁎, Jun Zheng b, Jia-Yuan Liu a, Xu-Rong Shi a,c, Jiao Xu a, Yin-Chang Feng a

a
State Environmental Protection Key Laboratory of Urban Ambient Air Particulate Matter Pollution Prevention and Control, College of Environmental Science and Engineering, Nankai University,
Tianjin 300071, China
b
Huzhou Environmental Monitoring Center, Huzhou 313000, China
c
College of Environmental & Resource Sciences, Shanxi University, Taiyuan 030006, China

H I G H L I G H T S G R AP HI C A L A B S T R AC T

• As a first effort, the influence of mineral

dust on PM2.5 was investigated by
employing the combined receptor
model.
• Reconstructed source contribution
(RCSC) method was firstly established
to estimate the performance of the
model results.
• Air mass residence time (AMRT) analy-
sis was explored to investigate the re-
gional influence on PM2.5.

a r t i c l e i n f o a b s t r a c t

Article history: To understand the influence of quarry mining dust on particulate matter, ambient PM2.5 and quarry mining dust
Received 24 October 2015 source samples were collected in a city near quarry facilities during 2013–2014. Samples were subject to chem-
Received in revised form 27 January 2016 ical analysis for dust-related species (Al, Si, Ca, Fe, Ti), tracer metals, carbon components and water-soluble ions.

Accepted 28 January 2016

Seasonal variations of PM2.5 and its main chemical components were investigated. Distinctive seasonal variations
Available online xxxx

of PM2.5 were observed, with the highest PM2.5 concentrations (112.42 μg m− 3) in fall and lowest concentrations
Editor: D. Barcelo in summer (45.64 μg m− 3). For dust-related species, mass fractions of Si and Al did not show obvious seasonal
variations, whereas Ca presented higher fractions in spring and summer and lower fractions in fall and winter.
Keywords: A combined receptor model (PMF-CMB) was applied to quantify the quarry mining dust contribution to PM2.5.
PM2.5 Seven sources were identified, including quarry mining dust, soil dust, cement dust, coal combustion vehicles,
Quarry mining dust secondary sulfate and secondary nitrate. On a yearly average basis, the contribution of quarry mining dust to
Dust-related species PM2.5 was 6%. The contribution of soil dust to PM2.5 was comparable with cement dust (13% and 13%, respec-
PMF-CMB tively). Other identified sources included vehicle, secondary sulfate, secondary nitrate and coal combustion,
Source apportionment
which contributed 23, 15, 9 and 18% of the total mass, respectively. Air mass residence time (AMRT) analysis

⁎ Corresponding author at: State Environmental Protection Key Laboratory of Urban Ambient Air Particulate Matter Pollution Prevention and Control, College of Environmental Science
and Engineering, Nankai University, Tianjin 300071, China.
E-mail address: nksgl@hotmail.com (G.-L. Shi).

http://dx.doi.org/10.1016/j.scitotenv.2016.01.195
0048-9697/© 2016 Elsevier B.V. All rights reserved.
941 X. Peng
X. Peng
et al.
et /al.Science
/ Science
of the
of the
Total
Total
941
showed that northeast and southeast regions might be the major PM2.5 source during the sampling campaign.
The findings of this study can be used to understand the characteristics of quarry mining dust and control strat-
egies for PM2.5.
1. Introduction
PM2.5 is an important pollutant in ambient urban air and one of the
most essential factors for regional air quality, climate impacts, economic
concerns and human health (Ozkaynak and Thurston, 1987; Emmanuel,
2000; Simkhovich et al., 2008; Yan et al., 2009; Tosca et al., 2010; Bressi
et al., 2014). Dust, including natural dust (resulting from natural resus-
pension and sandstorm events, e.g., soil dust and sand storm) and an-
thropogenic dust (deriving from anthropogenic activities, e.g., quarry
mining dust, cement dust), is an important contributor to PM2.5, ac-
counting for more than 25% of PM 2.5 mass concentration in urban
areas (Kleeman et al., 1999; Canepari et al., 2009; Shi et al., 2009;
Wang et al., 2012; Dall'Osto et al., 2013; Tian et al., 2015). Anthropo-
genic dust has high fractions (up to 50%) of the total dust load in the at-
mosphere, increasing the loading of dust (Balkanski et al., 2007),
therefore, researchers have concentrated on the emission and influence
of the anthropogenic dust.
Quarry mining dust, a specific anthropogenic dust, is mainly emitted
from the stockpiling, loading, transportation, and processing of mineral
materials in quarries and from the open air storage of mineral materials
(Elbir and Muezzinoglu, 2004; Silvester et al., 2009; Galindo et al.,
2011). In developing countries (e.g., China, India), quarry mining dust
usually has been taken seriously, as large-scale building construction
and extensive roads associated with urbanization have the potential to
cause the usage of mines. With the rapid development of China, the cur-
rent annual build building area in China is approximately 2 billion m2,
which accounts for half of the total annual building area in the world
(Fernández, 2007; Yang et al., 2014). The consumption of minerals of
China increased from approximately 7000 (1997) to
13,000 million tons (2007) in the construction sector (Wang et al.,
2014). The mineral products sector is one of the most important con-
tributors to quarry mining dust emissions that are important contribu-
tors to the fine particulates. In addition, uncontrolled quarry mining
dust emissions could cause serious human health (Moshammer et al.,
2014). Therefore, it is worth paying attention to the influence of quarry
mining dust.
Researches have been conducted on quarry mining dust (Petavratzi
et al., 2005, 2007; Silvester et al., 2009; Lira et al., 2012; Moshammer
et al., 2014), which were focused on quarry mining dust emissions
(Petavratzi et al., 2005, 2007; Silvester et al., 2009), human health
(Lira et al., 2012; Moshammer et al., 2014) and the influence on coarse
particulate (Bluvshtein et al., 2011) or water-soluble ions (Galindo et al.,
2011). Studies have showed that quarry mining dust could be released
into the atmosphere as a source of coarse particulates as well as fine par-
ticulates (PM2.5) (Bluvshtein et al., 2011). There are extremely few re-
ports in the literature concerning the influence on air quality and
PM2.5, although it is well known that quarry mining dust is potentially
hazardous to PM2.5. However, it is scarcely to quantify the influence of
quarry mining dust on the PM2.5. In this study, the contribution of
quarry mining dust to PM2.5 was estimated, improving the understand-
ing of the quarry mining dust and controlling on the PM2.5.
Receptor models (e.g., Chemical mass balance (CMB), Positive ma-
trix factorization (PMF)) are useful tools to quantify the source contri-
butions to PM and have been widely applied in many studies
(Paterson et al., 1999; Xie et al., 1999; Watson and Chow, 2001; Lee
et al., 2003; Waked et al., 2014). However, for some collinear sources
(sources have similar profiles such as quarry mining dust and soil
dust), single receptor models failed to separate these sources and quan-
tify their contributions from each other due to high correlation coeffi-
cients in time series between these source contributions and/or the
Environment
Environment
550550
(2016)
(2016)
940–949
940–949 941
941
© 2016 Elsevier B.V. All rights reserved.
similarity in their source profiles (Brinkman et al., 2006; Shi et al.,
2009). Therefore, it is a challenge to estimate quarry mining dust contri-
butions to PM2.5 by receptor models, because that the similar chemical
compositions of quarry mining dust and soil dust (Galindo et al.,
2011). A combined model of PMF-CMB was developed in our previous
study to separate collinear sources and quantify their source contribu-
tions (Shi et al., 2009), which has been introduced in the EPA PMF
v5.0 user guild and applied in several studies, playing an important
role in quantifying the collinear sources (Shi et al., 2009; Zeng et al.,
2010; Pant and Harrison, 2013; Taiwo et al., 2014; USEPA, 2014). So,
the combined model can be employed to quantify the source contribu-
tion of quarry mining dust.
To study the influence of quarry mining dust, we would choose an
ideal experimental field close to quarrying facilities. Huzhou is a typical
city characterized by minerals, and it has approximately 100,000 people
engaged in mineral-related careers.
In this work, PM2.5 samples and dust samples (Quarry mining dust,
soil dust) were collected during 2013–2014 at a site in Huzhou. The
dust-related species (e.g., Al, Si, Ca, Fe, and Ti), tracer metals, carbon
compositions and water-soluble ions were analyzed. The current work
aims to: 1) investigate the characteristics of chemical species in PM2.5
and profiles of quarry mining dust and soil dust; 2) quantify the quarry
mining dust and other source contributions to PM2.5 by using a com-
bined model (PMF-CMB), and estimate the performance of the modeled
results by a new method (reconstructed source contribution, RCSC);
and 3) explore the regional influence on PM2.5 by air mass residence
time method (AMRT). To our best knowledge, it is the first effort to
use the RCSC method to estimate the performance of the model results.
The data and analysis will improve the knowledge of chemical compo-
nents of quarry mining dust, and their influence on PM2.5 and human
health.
2. Methods and models
2.1. Study area description
Huzhou (119°14′E–120°29′E, 30°22′N–31°11′N) is located in the
north of Zhejiang province, which is adjacent to Hangzhou (the capital
city of Zhejiang province). Huzhou possesses rich mineral resources
and the total productions of mineral were up to 109.11 megatons
(Mt) in 2009. The major minerals are limestone, building stone and
quartz sandstones. Most of the minerals and their products are supplied
to the surrounding cities, such as Shanghai, Hangzhou, Jiaxing and
Ningbo. Huzhou is a desired candidate for investigating the influence
of quarry mining dust on air quality.
2.2. Sample collection
The 24-h integrated PM2.5 was conducted at four sampling sites in
Huzhou for 15–20 days during every season for one year (March 2013
to February 2014). The details of sampling were presented in Supple-
ment information (SI) Section S1. A total of 258 samples were obtained.
The map of PM2.5 sampling sites is in SI (Fig. S1).
Quarry mining dust and soil dust samples were collected in Huzhou.
Six quarry mining dust samples were collected from two typical quarry-
ing facilities. The map of the quarry plants sites were shown in the
Fig. S1. Soil samples were collected at 15 sites. The mineral and soil sam-
ples were placed in glass jars and kept frozen at − 20 °C after they were
collected. The source samples were suspended in a resuspension system
and the particles were separated by a PM 2.5 cut-off inlet with filters
942 X. Peng
X. Peng
et al.
942
(quartz and polypropylene fiber) for further analysis (Chow et al., 1994;
Watson and Chow, 2001; Bi et al., 2007). More details were reported in
our prior research (Bi et al., 2007). The analysis processes were the same
for the receptor samples (PM2.5).
2.3. Chemical analysis
In this study, dust-related species (e.g., Al, Si, Ca, Fe, Ti) collected on
polypropylene fiber filters were analyzed by inductively coupled
plasma-mass spectrometry (ICP-AES) (IRIS Intrepid II, Thermo Elec-
tron). Simultaneously, other tracer metals, water-soluble ions (NO−
2− +
−
SO4 , Cl and NH4 ) and carbon components (organic carbon (OC)
and element carbon (EC)) were measured, to help the source apportion-
ment. The details of sampling were presented in Supplement informa-
tion Section S2.
In addition, supporting meteorological (temperature, wind speed
and wind direction) and air quality data (PM2.5, sulfur dioxide (SO2), ni-
trogen dioxide (NO2), Carbon monoxide (CO) and ozone (O3)) were ob-
tained from the government monitoring stations located in Huzhou.
2.4. Data analysis
2.4.1. RCFM-evaluating the accuracy of PM2.5 data
The reconstructed fine mass (RCFM) method was used to evaluate
the accuracy of PM2.5 and chemical species associated with PM2.5 mea-
surements (Hand et al., 2011). If the measured PM2.5 data are close to
the RCFM data, this might show the measured data are reasonable.
The reconstructed PM2.5 mass concentrations were calculated based
on the concentrations of major species of PM2.5, as in the USA IMPROVE
study. The details of the reconstructed mass were presented in
Section S3.
2.4.2. RCSC-evaluating the performance of PMF
In this work, a new method was developed to evaluate the rational-
ity of source apportionment results obtained by PMF. This method is
based on the RCFM algorithm: the contributions of each source are re-
constructed using the factor profiles obtained from the PMF.
The reconstructed source contribution (RCSC) of each source can be
calculated as follows:
RC j ¼ ½AS j þ ½AN j þ ½OM j þ ½LAC j þ ½Soil j þ ½SS j ð1Þ
where RCj is the reconstructed contribution of the jth source to the sam-
ples. [AS]j, [AN]j, [OM]j, [LAC]j, [Soil]j and [SS]j are the modeled mass
contribution (μg m− 3) from the jth source, for ammonium sulfate, am-
monium nitrate, organic matter, light absorbing carbon, soil and sea
salts, respectively. They were calculated with the following equations:
h i
2−
½AS j ¼ 1:375 SO4
j
½AN j ¼ 1:29 NO−
3 j
½OM j ¼ 1:8 ½OC j ð4Þ
½LAC j ¼ ½EC j ð5Þ
½Soil j ¼ 2:2 ½Al j þ 2:49 ½Si j þ 1:63 ½Ca j þ 2:42
þ 1:94 ½Ti j
−
½SS j ¼ 1:8 ½Cl j:
The [SO42 −]j, [NO3−]j, [OC]j, [EC]j, [Al]j, [Si]j, [Ca]j, [Fe]j, [Ti]j and [Cl−]j
et /al.Science
/ Science
of the
of the
Total
Total
Environment
Environment
550550
(2016)
(2016)
940–949
940–949 942
942
In this study, the contributions of five sources were reconstructed. If
the calculated contributions from PMF are close to the reconstructed
contributions, this might show the results obtained from PMF are
reliable.
2.5. Combined model (PMF-CMB)
A combined model was employed to quantify the quarry mining
dust and other source contributions to PM2.5. The principle of the com-
3, bined receptor model was described in our previous work in detail (Shi
et al., 2009, 2011). Briefly, the combined model includes two stages: At
stage 1, the original dataset (original receptor) was analyzed by PMF,
and several factors were extracted. Some solo factors could be identified
as one source category (called simplex sources), whereas other solo fac-
tors might contain two or more sources (collinearity sources), which are
called complex factors. At stage 2, the complex factor was considered as
a new receptor (secondary receptor), and the sub-sources in the sec-
ondary receptor were calculated by the CMB model. Finally, the final re-
sult of the combined model was determined by combining the results
obtained in stage 1 and stage 2. The details of the PMF and CMB models
were presented in the SI Sections S4 and S5.
PMF v5.0 was employed to apportion the PM2.5 sources, which was
performed on a matrix made up of 23 columns (number of chemical
species) and 258 rows (number of PM2.5 samples). Hg was removed,
as it had not been detected in most samples.
2.6. AMRT-regional influence analysis
Additionally, we also tried to analyze regional influence on PM2.5. Air
mass residence time was analyzed based on that it over a region is line-
arly related to that region's contribution to pollutants measured at the
receptor site (Xu et al., 2006; Dimitriou and Kassomenos, 2014). There-
fore, the relationship between PM2.5 concentrations and air mass resi-
dence time could be established as follows (Dimitriou and
Kassomenos, 2014):
X
N
PMi ¼ ðB k T ik Þ ð8Þ
k¼1
where PMi is the PM2.5 concentration (μg m− 3) on the ith day, which
was obtained from the government monitor stations in this work. Tik
is the total air mass residence time over each one of the kth region for
the ith day. Bk is the PM2.5 contribution factor that represents the ability
of the kth region to enrich the daily PM2.5 concentrations. N is the num-
ð2Þ ber of the regions. The kth region contribution to PM2.5 during the ith
day is calculated by the multiplication of Bk with Tik.
In this work, 72-h back trajectories, ending at 500 m altitude of the
ð3Þ
sampling site (Dimitriou and Kassomenos, 2014), were calculated
every hour by HYSPLIT4.9 (http://ready.arl.noaa.gov/HYSPLIT.php)
model during the sampling period. 469 trajectories were obtained and
each trajectory was consisted of 73 trajectory points (3 days backward
per trajectory × 24 h per trajectory day + 1 arrival point). Thus, the
sum number of the trajectory points was 469 × 24 × 73 in this study
½Fe j (Dimitriou and Kassomenos, 2014). To ensure that almost all of trajec-
ð6Þ tory points were included, the research area extended across between
the longitude 73°E–140°E and the latitude 17°N–54°N. In order to inves-
ð7Þ tigate the influences regions for research area, five areas were selected
for analyzing: including Zhejiang province (region 4), surrounding re-
gion of Huzhou (region 1–3) and rest region of China (region 5), as
943 X. Peng
X. Peng
et al.
et /al.Science
/ Science
of the
of the
Total
Total
Environment
Environment
550550
(2016)
(2016)
940–949
940–949 943
943 943

are modeled mass contribution (μg m− 3) from the jth source, for shown in Fig. S2. Table S1 summarizes the coordinates of regions. Back

SO42 −, NO3−, OC, EC, Al, Si, Ca, Fe, Ti and Cl−, respectively, which obtained trajectories cluster analysis for annually and seasonally was conducted.

from PMF. The corresponding results were presented in Figs. S3 and S4.
944 X. Peng
X. Peng
et al.
et /al.Science
/ Science
of the
of the
Total
Total
Environment
Environment
550550
(2016)
(2016)
940–949
940–949 944
944 944

3. Results and discussion

3.1. PM2.5 mass concentration

The measured concentrations of PM 2.5 and the chemical species

were presented in Table 1. An annual average PM2.5 concentration of
78.28 μg m− 3 was found for this study, which was two times higher
than the annual average concentration standard of PM2.5 (35 μg m− 3)
established by the National Ambient Air Quality Standards (NAAQS) of
China (2012).
The average PM2.5 concentrations for each season were listed in
Table 1. A typical seasonal variation in PM2.5 concentrations was re-
vealed, with higher concentrations in fall (September–November,
112.42 μg m− 3), intermediate concentrations in spring (March–May,
72.68 μg m− 3) and winter (December–February, 74.26 μg m− 3), and
Fig. 1. Chemical mass fractions (%) of PM2.5 mass concentration during the whole
lower concentrations in summer (June–August, 45.64 μg m− 3). The sampling period and four seasons. The major chemical components include inorganic
lowest concentrations displayed in summer occur because precipitation ions (SO4 , NO3 , NH4 and Cl ), carbon components (OC, EC) and crustal markers (Si,
2− − + −

in Huzhou is usually concentrated during this period (Fig. S5), which is
similar to Chengdu (Tao et al., 2014). To further investigate the reasons
for PM2.5 seasonal variation, gas pollutants (SO2, NO2, CO and O3) were
measured and their seasonal variations were analyzed. Gas pollutants
(SO2, NO2 and CO) also exhibited the lowest concentrations in summer
(Fig. S6), which could further confirm that lighter pollution in summer
might be affected by good meteorological conditions.
3.2. The characteristics of the dust-related species in PM2.5
To better explore the characteristics of dust sources, the properties
of dust-related chemical compounds in PM2.5 were discussed. According
to literature, Si, Al, Ca, Ti and Fe are usually treated as compounds asso-
ciated with dust sources (Canepari et al., 2009; Duan et al., 2012). In this
work, Si, Al and Ca were the major elements in PM2.5. Their contribu-
tions were all more than 2% of the PM2.5 mass concentrations (Fig. 1).
As shown in Table 1, three species displayed harmonious trends of sea-
sonal variation. The highest concentrations (3.93, 6.03 and 2.65 μg m− 3
for Al, Si and Ca, respectively) were found during fall and the lowest
Al, Ca, Fe and Ti).
(1.39, 2.93 and 1.55 μg m− 3 for Al, Si and Ca, respectively) were ob-
tained during summer. The difference of concentration in two seasons
might be associated with meteorological conditions, similar to the situ-
ation with PM2.5. However, the seasonal variations of the mass fractions
(%) were not consistent with the mass concentrations (μg m− 3) (Fig. 1).
Si was the most abundant element, accounting for 4.8, 6.3, 5.4 and 5.3%
in spring, summer, fall and winter, respectively, with an annual average
fraction of 5.5%. Al presented unremarkable seasonal variation, account-
ing for 2.4% in spring, 3.1% in summer, 3.3% in fall and 3.1% in winter. Ca
showed distinct seasonal variations, with higher fractions in spring
(6.2%) and summer (3.4%) and lower fractions in fall (2.5%) and winter
(2.3%). In summary, mass fractions of Si and Al did not show obvious
seasonal variations, whereas Ca presented higher fractions in spring
and summer and lower fractions in fall and winter. The distinction
might due to the different emission patterns of the origins of the three
species. Ca was the marker of cement dust, which was related to

Species Spring (μg m− 3) Summer (μg m− 3) Fall (μg m− 3) Winter (μg m− 3) Annual (μg m− 3)
Table
PM2.51 72.68 ± 23.73 45.64 ± 18.02 112.42 ± 48.12 74.26 ± 40.25 78.28 ± 43.61
The
Almeasured concentrations of PM
1.80 2.5 and chemical species (mean
± 0.86 ±0.55
1.39 ± standard deviation). 3.93 ± 5.63 2.57 ± 2.00 2.54 ± 3.38
Si 3.55 ± 1.40 2.93 ± 1.35 6.03 ± 3.39 4.41 ± 3.31 4.37 ± 2.93
Ca 4.10 ± 0.94 1.55 ± 0.75 2.65 ± 1.19 1.74 ± 1.05 2.42 ± 1.37
NH+ 4 3.91 ± 2.16 2.64 ± 1.26 6.67 ± 3.02 4.33 ± 2.10 4.53 ± 2.71
−
Cl 1.32 ± 1.43 0.77 ± 0.91 2.34 ± 2.08 2.55 ± 1.3 1.83 ± 1.68
−
6.10 ± 3.84 4.55 ± 2.11 17.22 ± 11.93 12.35 ± 8.99 10.8 ± 9.63
2−
10.95 ± 6.00 8.67 ± 4.26 15.51 ± 8.09 9.89 ± 7.97 11.58 ± 7.43
OC 10.64 ± 4.02 7.46 ± 2.89 16.51 ± 11.6 10.65 ± 5.02 11.60 ± 7.89
EC 3.98 ± 1.88 2.35 ± 1.03 5.84 ± 3.31 4.34 ± 1.99 4.24 ± 2.60
NO3
−3
SO4 Spring (ng m ) Summer (ng m− 3) Fall (ng m− 3) Winter (ng m− 3) Annual (ng m− 3)

Mg 575 ± 122 728 ± 354 761 ± 318 614 ± 265 676 ± 294
K 1282 ± 601 566 ± 194 1716 ± 910 1004 ± 553 1162 ± 764
Ti 26.7 ± 20.8 12.1 ± 14.3 26.8 ± 18.2 4.2 ± 3.8 16.1 ± 17.5
V 16.8 ± 10.4 17.0 ± 10.1 17.1 ± 17.8 5.0 ± 4.8 13 ± 13.4
Cr 17.8 ± 5.2 21.3 ± 9.2 42.8 ± 24.9 30.2 ± 12.1 29.4 ± 18.4
Mn 22.1 ± 9.7 24.8 ± 22.8 50.9 ± 24.7 24.4 ± 10.8 31.7 ± 22.3
Fe 612 ± 790 427 ± 225 967 ± 588 525 ± 197 647 ± 535
Co 0.7 ± 0.3 0.6 ± 0.3 1.8 ± 4.7 0.8 ± 0.3 1.0 ± 2.6
Ni 11.7 ± 3.2 12.1 ± 6.0 25.1 ± 13.4 20.0 ± 6.9 18.0 ± 10.3
Cu 22.1 ± 7.6 21.6 ± 8.9 33.5 ± 20.7 29.1 ± 20.2 27.3 ± 17
Zn 250 ± 110 169 ± 86 492 ± 313 206 ± 115 289 ± 231
As 0.8 ± 0.5 1.1 ± 0.9 2.8 ± 4.8 1.1 ± 0.8 1.5 ± 2.7
Cd 1.2 ± 0.6 1.0 ± 0.6 2.7 ± 1.5 1.2 ± 1.0 1.6 ± 1.3
Hg nda nd nd nd nd
Pb 48.8 ± 23.5 29.5 ± 17.3 113.5 ± 75.5 66.4 ± 40.1 67.9 ± 57.4
a
nd: not detected.
945 X. Peng
X. Peng
et al.
et /al.Science
/ Science
of the
of the
Total
945
building construction. Around the sampling sites, a large number of con-
struction activities were performed during the spring or summer. In fall
and winter, construction activities were reduced due to the low temper-
atures and the completion of the construction projects. These factors
caused lower fractions of Ca in PM2.5 in fall and winter. Fe and Ti have
not been discussed due to their smaller fractions in PM2.5 (b 1%). Previ-
ous reports showed that Fe and Ti were present in lower fractions of
PM2.5 relative to Al, Si and Ca (Shi et al., 2011).
3.3. The characteristics of water soluble ions and carbon components
Comprehending the characteristics of water-soluble ions and carbon
components can add to knowledge about PM2.5 formation mechanisms.
Hence, the characteristics of ions and carbon components were investi-
gated. As illustrated in Fig. 1, the water-soluble ion (SO42 −, NO3−, NH4+
and Cl−) fractions ranged from 29.9% to 37.9% in four seasons, with an
annual average value of 35.8%, becoming an important contributor to
PM2.5. Sulfate had the highest fraction (with an annual mean of 14.9%)
among the four analyzed ions, followed by NO3− (12.5%), NH4+ (5.9%)
and Cl− (2.4%). A typical seasonal variation of sulfate followed the
order of summer (18.6%), spring (14.7%), fall (14.6%) and winter
(12.3%), which was similar to results reported by Zhang et al. (2013).
In contrast to sulfate, nitrate displayed a seasonal variation with higher
fractions in fall (14.7%) and winter (15.6%) and lower fractions in spring
(8.1%) and summer (9.9%). Temperature might influence the ammo-
nium nitrate evaporation and particulate nitrate formation (Zhang
et al., 2013; Xu et al., 2014).
For EC (Fig. 1), relatively constant fractions were seen for the contri-
butions to PM2.5, with the values of 5.4 (spring), 5.2 (summer), 5.5 (fall)
and 6.3% (winter). OC presented weak seasonal variation, accounting
for 15.0% in spring, 17.0% in summer, 14.1% in fall and 15.2% in winter.
On a yearly basis, the average fractions of OC and EC were 15.3% and
5.6%, respectively. EC is caused by primary emissions (e.g., vehicle emis-
sions, coal combustion) that might exhibit few changes of source emis-
sion strength during different seasons, resulting in relatively constant
fractions of EC. OC is composed of primary emissions (primary organic
carbon, POC) and secondary formations (secondary organic carbon,
SOC). SOC concentrations calculated according to the previous reports
and the method displayed in Section S6 (Castro et al., 1999; Hou et al.,
2011). The SOC concentrations ranged from 3.64 (winter) to
5.09 μg m− 3 (summer), with an annual average concentration of
4.24 μg m− 3. In this study, the SOC presented a pattern with higher con-
centrations in the spring and summer, consistent with the report by Heo
et al. (2013). The results suggested the high OC concentration in
Total
Environment
Environment
550550
(2016)
(2016)
940–949
940–949 945
945
summer was partly ascribed to the suitable formation conditions of
SOC in summer.
3.4. Using RCFM analyze ambient PM2.5 data
PM 2.5 mass in Huzhou was reconstructed by employing the ap-
proach in IMPROVE, and the RCFM results were displayed in the Fig. 2.
To further examine the accuracy of PM2.5 data, a correlation analysis be-
tween the reconstructed and measured mass concentrations for annu-
ally and seasonally was conducted. Fig. 3 illuminates the overall
results. The correlations (r) of the reconstructed and measured mass
concentrations were up to 0.99 and the slopes ranged from 0.92 to
1.03, indicating a good performance of the mass concentrations mea-
sured. In addition, the slopes near 1 supported that 1) the parameters
provided by IMPROVE are suitable for this study, and 2) the measured
PM2.5 and chemical components data might be reasonable. As shown
in Fig. 2, soil dust was an important component, accounting for 27.6%
(22.22 μg m− 3) of RCFM.
3.5. Source apportionment of PM2.5
The combined model (PMF-CMB) was employed to apportion the
sources (especially quarry mining dust) of PM2.5. A brief description of
the model results is included in the following paragraph. Firstly, the
cluster analysis was applied to the species to identify the source of the
PM 2.5 in Huzhou. The results were displayed in Fig. S7. Group 1
4 3
contained SO2 − , NO− and OC that were associated with secondary
source (Viana et al., 2008). Group 2 including Si, Al and EC were associ-
ated with vehicle and dust (Tan et al., 2014; Tao et al., 2014). Group 3
including Ca, Cl− were linked to coal combustion (Watson et al.,
2001). The cluster results indicate that secondary source, vehicle, coal
combustion and dust might be the potential sources.
3.5.1. PMF stage to extract crustal dust and other sources
For the PMF model, the first step was to determine the factor num-
ber. In this study, an optimal factor number of five was found and the
detail process for selecting of factor number could be found in
Section S7.
3.5.1.1. Vehicle exhaust. The profiles of each identified factor were shown
in Fig. 4. According to Fig. 4, factor 1 was dominated by high loadings of
OC and EC that might represent vehicle sources (Kim and Hopke, 2004;
Li et al., 2014; Tao et al., 2014). The values of OC/EC can be used to eval-
uate the PM2.5 sources because the ratio might change from source to
source (Gray, 1986; Watson et al., 2001) reported that the ratio of OC/
EC of vehicle sources and coal combustion were 0.67–2.45 and 0.28–
8.09, respectively. In this study, the OC/EC ratio in factor 1 was 4.0,
which was higher than the ratios reported by Watson et al. (2001),
which might because SOC was also mixed in factor 1, which resulted
in overestimating the OC fraction in this factor. Fig. 5 exhibited the
PMF results of the relative source contributions to the PM2.5 on both a
seasonal and annual basis in Huzhou. This factor represented a contribu-
tion of 18.1 μg m− 3 (23%) to PM2.5, which included parts of the SOC.
3.5.1.2. Secondary sources. The second factor was identified by high load-
2−
ing of NH+ 4 , SO4 and NO− 3 , which might be associated with secondary
sources (Viana et al., 2008; Zhang et al., 2013; Waked et al., 2014). Sec-
ondary sources were the predominant source in China, accounting for
more than 20% of the total PM 2.5 mass in many cities (Zhang et al.,
2013; Shi et al., 2014; Tao et al., 2014; Xiao et al., 2014). In this study,
secondary sources were the main contributor to PM2.5, with a contribu-
2−
tion of 21.0 μg m− 3 (27%). Due to high loading fractions of SO4 and
−
946 X. Peng
X. Peng
et al.
et /al.Science
/ Science
of the
of the
Total
Total
Environment
Environment
550550
(2016)
(2016)
940–949
940–949 946
946 946

Fig. 2. The average (annual) and seasonal average mass fractions of each components. The
mass fractions of components were estimated by using IMPROVE method. AS: ammonium
sulfate, AN: ammonium nitrate, OM: organic matter, LAC: light absorbing carbon, Soil: soil,
SS: sea salts.
NO3 , this factor was a complex factor (called complex source, complex
factor 1) that included secondary sulfate and nitrate sources. The pre-
cursor gases of secondary sulfate and nitrate were mainly emitted
from coal combustion, vehicle exhaust and power plants.
947 X. Peng
X. Peng
et al.
et /al.Science
/ Science
of the
of the
Total
Total
Environment
Environment
550550
(2016)
(2016)
940–949
940–949 947
947 947

Fig. 3. Correlation between reconstructed and measured PM2.5 concentrations during the whole sampling period and four seasons.
3.5.1.3. Coal combustion. Factor 3 was distinguished by the relatively Watson et al. (2001). However, the Ca (2.6%) abundance for factor 3 was
high loading of OC and EC as well as the moderately high loading of higher than the corresponding values (0.7 ± 0.7%) from Sheffield and
Al, Si, Ca and Cl−, and might be identified as coal combustion (Watson Gordon (1986). These differences might because the profiles of coal
et al., 2001). This source accounted for 18% of the PM2.5 mass concentra- combustion vary from region to region (Watson et al., 2001).
tion, with an absolute contribution of 14.1 μg m− 3. The OC/EC ratio in
this factor fell into the range of coal combustion (0.28–8.09) (Watson 3.5.1.4. Cement dust. The fourth factor was relevant to cement dust,
et al., 2001). Al and Si accounted for 3.7% and 6.1% of this factor profile, which was typically characterized by high loadings of Ca and accounted
respectively, which was similar to the corresponding values reported by for 9.6 μg m− 3 (13% to PM2.5) (Zhang et al., 2005). With the

Fig. 4. Factor profiles obtained by PMF model. Complex factor 1: secondary sulfate source and secondary nitrate source. Complex factor 2: soil dust and quarry mining dust.
948 X. Peng
X. Peng
et al.
et /al.Science
/ Science
of the
of the
Total
948
Fig. 5. Source contributions obtained by PMF model. (a) Annual; (b) Spring; (c) Summer; (d) Fall; (e) Winter. Complex factor 1: secondary sulfate source and secondary nitrate source.
Complex factor 2: soil dust and quarry mining dust.
development of the city, massive construction works – spanning ap-
proximately 11.70 million m2 – increased the usage of cement and
might product dust. A lack of control measures also resulted in cement
dust pollution.
3.5.1.5. Crustal dust. Factor 5 was characterized by high loadings of Al, Si
and Ca, which pointed to crustal dust (Duan et al., 2012; Tan et al., 2014;
Tao et al., 2014). In addition, for the crustal elements, such as Fe, factor 5
exhibited higher fractions than the other factors, further suggesting that
factor 5 might have a crustal dust source (Canepari et al., 2009; Duan
et al., 2012). As mentioned above, Huzhou is a typical mineral city
with a total mineral production of 109.11 Mt. The exploiting, processing
and transportation of mines might account for the dust. Therefore,
quarry mining dust is a special dust source in Huzhou. Factor 5 might
be attributed to the mixture of soil dust and quarry mining dust (called
complex factor 2), which accounted for 19%, with an absolute contribu-
tion of 14.6 μg m− 3. The complex factors (1 and 2) would be introduced
into the CMB model to apportion the sub-sources.
The total contribution of three types of dust in this study was 32%,
which was consisted with the soil fraction (27.6%) of RCFM. This value
was similar to the reported value of 30% for the dust (including soil
dust, crustal dust, brake dust) in Barcelona (Dall'Osto et al., 2013),
which was in agreement with the reported value in India (26%)
(Gummeneni et al., 2011). Therefore, according to the PMF results,
dust might be the dominant source, and effective measures should be
taken to control the dust in Huzhou.
Total
Environment
Environment
550550
(2016)
(2016)
940–949
940–949 948
948
Variations in the seasonal averaged contributions of dust to PM2.5
were shown in Fig. 5. Each dust exhibited diverse seasonal variations
owing to its emission variations and meteorological conditions, depend-
ing on the season. As expected, cement dust presented a seasonal pat-
tern with higher contributions during spring and summer than fall
and winter (17%, 23%, 9% and 8%, respectively), consistent with seasonal
variation of Ca. In terms of contributions in absolute concentrations,
contributions of cement dust were in the range from 5.6 to
11.5 μg m− 3. The highest contributions were observed in spring and
summer. This could be explained because cement dust is a typical an-
thropogenic source, mainly influenced by construction activities
(e.g., building and road traffic). As mentioned in Section 3.2, the con-
struction activities happened in hot seasons (spring and summer) and
were reduced in cold seasons (fall and winter), which caused the higher
contributions in spring and summer. For complex source 2 (contained
soil and quarry mining dust), contributions from this source to PM2.5
were higher in fall (25.0 μg m− 3, 22%) and winter (11.2 μg m− 3, 16%)
than in summer (8.5 μg m− 3, 17%) and spring (11.6 μg m− 3, 17%). It
could be interpreted that the dominant part of complex source 2 was
soil dust (accounted for 68% of complex source 2, see Section 3.5.3).
Soil dust is a nature source largely affected by meteorological factors,
such as wind speed and rainfall. Therefore, complex source 2 was
mainly influenced by meteorological conditions. The seasonal average
precipitations were higher during spring and summer, relative to the
other two seasons (Fig. S5), resulting in this factor presenting lower
contributions to PM2.5. The higher contributions of complex factor 2,
949 X. Peng
X. Peng
et al.
et /al.Science
/ Science
of the
of the
Total
Total
Environment
Environment
550550
(2016)
(2016)
940–949
940–949 949
949 949

Table 2 Table 3
The results of reconstructed and calculated source contributions. On a yearly basis, the final source apportionment results of PM2.5 obtained by the com-
bined receptor model.
Source RCSC (μg m− 3)a CC (μg m− 3)a RCSC (%) CC (%)

Source Contribution (%) Contribution (μg m− 3)

Vehicle 12.2 18.1 16 23
Complex factor 1a 21.5 21.0 27 27 Vehicle 23 18.1
Coal combustion 15.5 14.1 20 18 Secondary sulfate 15 12.0
Cement dust 7.8 9.6 10 13 Secondary nitrate 9 6.9
Complex factor 2a 16.7 14.6 21 19 Soil dust 13 9.9
Total 73.7 77.4 94 100 Quarry mining dust 6 4.4
a Cement dust 13 9.6
RCSC: reconstructe d source conce ntrations; CC: calcula ted source concentrations;

Coal combustion 18 14.1

Complex factor 1: secondary sulfate & secondary nitrate sources; Complex factor 2: soil &
quarry mining dust.

on the other hand, observed during fall and winter might be explained
by air masses coming from northern China. As shown in Fig. S4, in fall
and winter, air masses transported from Shandong province and picked
up crustal materials. For quarry mining dust, further research should be
conducted to investigate the seasonal variation of contribution.
3.5.2. Results of RCSC
To further evaluate the rationality of the results obtained from the
PMF, the source contributions were reconstructed according to the
method described in Section 2.4.2, using the source profile (F matrix)
and contribution (G matrix) obtained by PMF. The reconstructed contri-
butions (RCSC) were compared with the calculated contributions (CC)
by the PMF model. The results of RCSC and CC were listed in Table 2.
The RCSC for five sources (vehicles, complex factor 1, coal combustion,
cement dust and complex factor 2) were 12.2 μg m− 3, 21.5 μg m− 3,
15.5 μg m− 3, 7.8 μg m− 3 and 16.7 μg m− 3, respectively, which were
similar to the calculated source contributions obtained from PMF
model. The sum of reconstructed source concentrations was
73.7 μg m− 3, accounting for approximately 94% of the PM2.5 mass con-
centration. This finding demonstrates that the PMF model provided rea-
sonable source profiles and source contributions in this study.
3.5.3. CMB stage to apportion the quarry mining dust and other sub-sources
One of the major objectives in this work is to study the contribution
of the quarry mining dust to PM2.5. As described in Fig. 4, for the PMF
results, quarry mining dust was included in complex factor 2. To quan-
tify the contribution of quarry mining dust, profiles of quarry mining
dust and soil dust were measured and introduced into the CMB
model. Simultaneously, complex factor 1 as a secondary receptor was
also introduced into the CMB model. The uncertainties of complex factor
1 and 2 were calculated according to our previous study (Shi et al.,
2011).
Two dust profiles were illuminated in Fig. 6. The profiles of quarry
mining dust were characterized by high abundances of the species of
Al (4.1%), Si (7.7%), Ca (6.8%), Fe (1.6%) and OC (2.5%). For soil dust, Al
(3.8%), Si (11.5%) and Ca (8.3%) were also dominant elements, and Mg
(1.8%), Fe (3.0%) and OC (5.4%) exhibited relatively high fractions. By
comparing the two source profiles, it could be found that quarry mining
dust and soil dust were too similar to separate by single model.
The final source apportionment results of PM2.5 were listed in
Table 3 and Fig. 7. According to the calculation results of the CMB
stage, the annual average contributions of quarry mining dust and soil
dust to PM 2.5 mass were 4.4 μg m− 3 (6%) and 9.9 μg m− 3 (13%),
respectively.
Similarly, complex factor 1 was also analyzed by the CMB model to
evaluate secondary sulfate and secondary nitrate contributions. The sec-
ondary sulfate and nitrate source profiles were established according to
the composition of the molecular formulas of (NH4)2SO4 and NH4NO3
(Shi et al., 2011). According to the CMB results, the secondary sulfate
(12.0 μg m− 3, 15%) contribution for PM2.5 was higher than the second-
ary nitrate (6.9 μg m− 3, 9%). This observation was similar to the report
950 X. Peng
X. Peng
et al.
et /al.Science
/ Science
of the
of the
Total
Total
Environment
Environment
550550
(2016)
(2016)
940–949
940–949 950
950 950

Fig. 6. The measured profiles of four sub-sources.

951 X. Peng
X. Peng
et al.
et /al.Science
/ Science
of the
of the
Total
951
Fig. 7. On a yearly basis, the source apportionment results of complex factor 1 (a) and
complex factor 2 (b) obtained by the CMB model.
Table 4
Contribution percent and regression coefficients of four regions for PM2.5.
Region Contribution (%) Regression coefficient
2 61.73 0.17
3 1.91 0.09
4 26.88 0.11
5 9.25 0.02
Total 99.77
conducted by Huang et al. (2014) for the PM2.5 source apportionment in
Shenzhen.
3.6. Result of AMRT
The results of regional contributions were estimated by AMRT, to as-
sess the potential influences from different areas. Stepwise regression
method was applied to analyze the measured daily PM2.5 mass concen-
trations and the air mass residence time over each region. Region 1
(northwest) was excluded in the following analysis, as it was not statis-
tically significant. The sum of the four regional influences was up to
99.77%, indicating that these four regions might contain all the possible
pollution areas. The Pearson's correlation between measured and pre-
dicted daily PM2.5 mass concentrations was about 0.5 (P b 0.01). Regres-
sion coefficients could qualitatively reflect the region influence, which
were listed in the Table 4. As shown in the Table 4, region 2 (northeast)
and 4 (southeast) might have the major influence on PM2.5. The analysis
of regional influence on PM2.5 was based on the air masses of trajectory
information, therefore, air masses have a great impact on the results
during the sampling period. Region 2 was the biggest important con-
tributor might because the air masses mainly came from northeast dur-
ing the sampling periods. According to Figs. S3 and S9, the air mass and
surface wind directions were predominantly from north and east, which
were consisted with the results of regional influence analysis. It could be
concluded that northeast and southeast regions might be the major
PM2.5 source during the sampling campaign.
4. Conclusions
Based on one year of PM2.5 observations in Huzhou, this study inves-
tigated the influences of quarry mining dust and other sources to PM2.5
by analyzing their seasonal characteristics and source contributions.
Some interesting findings were observed in terms of the seasonal char-
acteristics of dust-related species. Mass fractions of Si and Al did not
show obvious seasonal variations. Ca presented higher fractions during
spring and summer and lower fractions in fall and winter, which might
be influenced by construction activities. By employing the combined
Total
Environment
Environment
550550
(2016)
(2016)
940–949
940–949 951
951
model, the contribution of quarry mining dust (6%) was quantified,
and other six sources were discovered, including vehicle sources
(23%), secondary sulfate (15%), secondary nitrate (9%), cement dust
(13%), coal combustion (18%) and soil dust (13%). Dust (soil, mineral
and cement) was the dominant source of PM2.5, accounting for 32%. In
particular, anthropogenic dust (mineral and cement dust) accounted
for approximately 18% of PM2.5 in Huzhou. Due to technical limitations
of the CMB model, quarry mining dust seasonal patterns could not be
solely investigated, which was discussed with soil dust. Seasonal aver-
age contributions of mineral and soil dust to PM2.5 showed an obvious
seasonal variation with higher contributions during fall and winter
when compared to spring and summer. This result might be associated
with the fact that soil dust was primarily influenced by meteorological
conditions in this study. Cement dust displayed higher contributions
in spring and summer, as it was affected by anthropogenic activities
(e.g., building and road construction). AMRT analysis shows that north-
east and southeast regions might be the major PM2.5 source during the
sampling campaign.
In this study, the characteristics of quarry mining dust were investi-
gated, which can benefit understanding the physical and chemical prop-
erties of quarry mining dust. In addition, source contributions to PM2.5
quarry mining dust were understood, which can provide information
for the formation mechanism and controlling strategies for PM2.5.
Acknowledgments
This study is supported by the National Natural Science Foundation
of China (21207070 and 41205089), Special Funds for Research on Pub-
lic Welfares of the Ministry of Environmental Protection of China
(201409003, 201509020) and the Blue Sky Foundation.
Appendix A. Supplementary data
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.scitotenv.2016.01.195.
References
Balkanski, Y., Schulz, M., Claquin, T., Guibert, S., 2007. Reevaluation of mineral aerosol ra-
diative forcings suggests a better agreement with satellite and AERONET data. Atmos.
Chem. Phys. 7, 81–95.
Bi, X.H., Feng, Y.C., Wu, J.H., Wang, Y.Q., Zhu, T., 2007. Source apportionment of PM10 in six
cities of Northern China. Atmos. Environ. 41, 903–912.
Bluvshtein, N., Mahrer, Y., Sandler, A., Rytwo, G., 2011. Evaluating the impact of a lime-
stone quarry on suspended and accumulated dust. Atmos. Environ. 45, 1732–1739.
Bressi, M., Sciare, J., Ghersi, V., Mihalopoulos, N., Petit, J.E., Nicolas, J.B., et al., 2014. Sources
and geographical origins of fine aerosols in Paris (France). Atmos. Chem. Phys. 14,
8813–8839.
Brinkman, G., Vance, G., Hannigan, M.P., Milford, J.B., 2006. Use of synthetic data to eval-
uate positive matrix factorization as a source apportionment tool for PM2.5 exposure
data. Environ. Sci. Technol. 40, 1892–1901.
Canepari, S., Pietrodangelo, A., Perrino, C., Astolfi, M.L., Marzo, M.L., 2009. Enhancement of
source traceability of atmospheric PM by elemental chemical fractionation. Atmos.
Environ. 43, 4754–4765.
Castro, L.M., Pio, C.A., Harrison, R.M., Smith, D.J.T., 1999. Carbonaceous aerosol in urban
and rural European atmospheres: estimation of secondary organic carbon concentra-
tions. Atmos. Environ. 33, 2771–2781.
Chow, J.C., Watson, J.G., Houck, J.E., Pritchett, L.C., Rogers, C.F., Frazier, C.A., et al., 1994. A
laboratory resuspension chamber to measure fugitive dust size distributions and
chemical compositions. Atmos. Environ. 28, 3463–3481.
Dall'Osto, M., Querol, X., Amato, F., Karanasiou, A., Lucarelli, F., Nava, S., et al., 2013. Hourly
elemental concentrations in PM2.5 aerosols sampled simultaneously at urban back-
ground and road site during SAPUSS — diurnal variations and PMF receptor model-
ing. Atmos. Chem. Phys. 13, 4375–4392.
Dimitriou, K., Kassomenos, P., 2014. A study on the reconstitution of daily PM10 and PM2.5
levels in Paris with a multivariate linear regression model. Atmos. Environ. 98,
648–654.
Duan, J.C., Tan, J.H., Wang, S.L., Hao, J.M., Chai, F.H., 2012. Size distributions and sources of
elements in particulate matter at curbside, urban and rural sites in Beijing. J. Environ.
Sci. (China) 24, 87–94.
Elbir, T., Muezzinoglu, A., 2004. Estimation of emission strengths of primary air pollutants
in the city of Izmir, Turkey. Atmos. Environ. 38, 1851–1857.
Emmanuel, S.C., 2000. Impact to lung health of haze from forest fires, the Singapore expe-
rience. Respirology 5, 175–182.
952 X. Peng
X. Peng
et al.
et /al.Science
/ Science
of the
of the
Total
Total
952
Fernández, J.E., 2007. Resource consumption of new urban construction in China. J. Ind.
Ecol. 11, 99–115.
Galindo, N., Yubero, E., Nicolas, J.F., Crespo, J., Pastor, C., Carratala, A., et al., 2011. Water-
soluble ions measured in fine particulate matter next to cement works. Atmos. Envi-
ron. 45, 2043–2049.
Gray, H.A., 1986. Control of Atmospheric Fine Primary Carbon Particle Concentrations.
EQL Report No. 23. Environmental Quality Laboratory, California Institute of Technol-
ogy, Pasadena, California, pp. 103–108.
Gummeneni, S., Bin Yusup, Y., Chavali, M., Samadi, S.Z., 2011. Source apportionment of
particulate matter in the ambient air of Hyderabad city, India. Atmos. Res. 101,
752–764.
Hand J, Copeland S, McDade C, Day D, Moore Jr C, Dillner A, et al. Interagency monitoring
of protected visual environments (IMPROVE). In., Spatial and Seasonal Patterns and
Temporal Variability of Haze and Its Constituents in the United States. Report V
ISSN, 2011;0737-5352-87.
Heo, J.B., Dulger, M., Olson, M.R., McGinnis, J.E., Shelton, B.R., Matsunaga, A., et al., 2013.
Source apportionments of PM2.5 organic carbon using molecular marker positive ma-
trix factorization and comparison of results from different receptor models. Atmos.
Environ. 73, 51–61.
Hou, B., Zhuang, G.S., Zhang, R., Liu, T.N., Guo, Z.G., Chen, Y., 2011. The implication of car-
bonaceous aerosol to the formation of haze. Revealed from the characteristics and
sources of OC/EC over a mega-city in China. J. Hazard. Mater. 190, 529–536.
Huang, X.F., Yun, H., Gong, Z.H., Li, X., He, L., Zhang, Y.H., et al., 2014. Source apportion-
ment and secondary organic aerosol estimation of PM2.5 in an urban atmosphere in
China. Sci. China Earth Sci. 57, 1352–1362.
Kim, E., Hopke, P.K., 2004. Source apportionment of fine particles in Washington, DC, uti-
lizing temperature-resolved carbon fractions. J. Air Waste Manage. Assoc. 54,
773–785.
Kleeman, M.J., Hughes, L.S., Allen, J.O., Cass, G.R., 1999. Source contributions to the size
and composition distribution of atmospheric particles: Southern California in Sep-
tember 1996. Environ. Sci. Technol. 33, 4331–4341.
Lee, P.H.K., Brook, J.R., Dabek–Zlotorzynska, E., Mabury, S.A., 2003. Identification of the
major sources contributing to PM2.5 observed in Toronto. Environ. Sci. Technol. 37,
4831–4840.
Li, L., Li, M., Huang, Z., Gao, W., Nian, H., Fu, Z., et al., 2014. Ambient particle characteriza-
tion by single particle aerosol m-ass spectrometry in an urban area of Beijing. Atmos.
Environ. 94, 323–331.
Lira, M., Rabbani, E.K., Barkokebas, B., Lago, E., 2012. Risk evaluation and exposure control
of mineral dust containing free crystalline silica: a study case at a quarry in the Recife
Metropolitan Area. Work-A Journal of Prevention Assessment and Rehabilitation 41,
3109–3116.
Moshammer, H., Hochgatterer, K., Haluza, D., 2014. 0034 Occupational exposure to min-
eral dust. Effects of on lung function in a nine-year study. Occup. Environ. Med. 71,
A62.
Ozkaynak, H., Thurston, G.D., 1987. Associations between 1980 U.S. mortality rates and al-
ternative measures of airborne particle concentration. Risk Anal. 7, 449–460.
Pant, P., Harrison, R.M., 2013. Estimation of the contribution of road traffic emissions to
particulate matter concentrations from field measurements: a review. Atmos. Envi-
ron. 77, 78–97.
Paterson, K.G., Sagady, J.L., Hooper, D.L., 1999. Analysis of air quality data using positive
matrix factorization. Environ. Sci. Technol. 33, 635–641.
Petavratzi, E., Kingman, S.W., Lowndes, I.S., 2005. Particulates from mining operations: a
review of sources, effects and regulations. Miner. Eng. 18, 1183–1199.
Petavratzi, E., Kingman, S.W., Lowndes, I.S., 2007. Assessment of the dustiness and the
dust liberation mechanisms of limestone quarry operations. Chem. Eng. Process. 46,
1412–1423.
Sheffield AE, Gordon GE. Variability of particle composition from ubiquitous sources. Re-
sults from a new source-composition library. In., Pace, T.G. (Ed.), Transactions, Recep-
tor Methods for source Apportionment, Real World Issues and Applications. Air
Pollution Control Association, Pittsburgh, PA, 1986; pp. 9–22.
Shi, G.L., Li, X., Feng, Y.C., Wang, Y.Q., Wu, J.H., Li, J., et al., 2009. Combined source appor-
tionment, using positive matrix factorization – chemical mass balance and principal
component analysis/multiple linear regression – chemical mass balance models.
Atmos. Environ. 43, 2929–2937.
Shi, G.L., Zeng, F., Li, X., Feng, Y.C., Wang, Y.Q., Liu, G.X., et al., 2011. Estimated contribu-
tions and uncertainties of PCA/MLR–CMB results. Source apportionment for synthetic
and ambient datasets. Atmos. Environ. 45, 2811–2819.
Environment
Environment
550550
(2016)
(2016)
940–949
940–949 952
952
Shi, G.L., Tian, Y.Z., Ye, S., Peng, X., Xu, J., Wang, W., et al., 2014. Source apportionment of
synchronously size segregated fine and coarse particulate matter, using an improved
three-way factor analysis model. Sci. Total Environ. 505, 1182–1190.
Silvester, S.A., Lowndes, I.S., Hargreaves, D.M., 2009. A computational study of particulate
emissions from an open pit quarry under neutral atmospheric conditions. Atmos. En-
viron. 43, 6415–6424.
Simkhovich, B.Z., Kleinman, M.T., Kloner, R.A., 2008. Air pollution and cardiovascular in-
jury. J. Am. Coll. Cardiol. 52, 719–726.
Taiwo, A.M., Harrison, R.M., Shi, Z.B.A., 2014. A review of receptor modelling of industri-
ally emitted particulate matter. Atmos. Environ. 97, 109–120.
Tan, J.H., Duan, J.C., Ma, Y.L., Yang, F.M., Cheng, Y., He, K.B., et al., 2014. Source of atmo-
spheric heavy metals in winter in Foshan, China. Sci. Total Environ. 493, 262–270.
Tao, J., Gao, J., Zhang, L., Zhang, R., Che, H., Zhang, Z., et al., 2014. PM2.5 pollution in a mega-
city of southwest China. Source apportionment and implication. Atmos. Chem. Phys.
14, 8679–8699.
Tian, Y.Z., Shi, G.L., Han, B., Wu, J.H., Zhou, X.Y., Zhou, L.D., et al., 2015. Using an improved
Source Directional Apportionment method to quantify the PM2.5 source contributions
from various directions in a megacity in China. Chemosphere 119, 750–756.
Tosca, M.G., Randerson, J.T., Zender, C.S., Flanner, M.G., Rasch, P.J., 2010. Do biomass burn-
ing aerosols intensify drought in equatorial Asia during El Nino. Atmos. Chem. Phys.
10, 3515–3528.
USEPA, 2014. EPA Positive Matrix Factorization (PMF) 5.0 Fundamentals and User Guide.
Viana, M., Kuhlbusch, T.A.J., Querol, X., Alastuey, A., Harrison, R.M., Hopke, P.K., et al., 2008.
Source apportionment of particulate matter in Europe. A review of methods and re-
sults. J. Aerosol Sci. 39, 827–849.
Waked, A., Favez, O., Alleman, L.Y., Piot, C., Petit, J.E., Delaunay, T., et al., 2014. Source ap-
portionment of PM 10 in a north-western Europe regional urban background site
(Lens, France) using positive matrix factorization and including primary biogenic
emissions. Atmos. Chem. Phys. 14, 3325–3346.
Wang, Z.S., Wu, T., Shi, G.L., Fu, X., Tian, Y.Z., Feng, Y.C., et al., 2012. Potential source anal-
ysis for PM10 and PM2.5 in autumn in a Northern City in China. Aerosol Air Qual. Res.
12, 39–48.
Wang, H.M., Tian, X., Tanikawa, H., Chang, M., Hashimoto, S., Moriguchi, Y., et al., 2014. Ex-
ploring China's materialization process with economic transition: analysis of raw ma-
terial consumption and its socioeconomic drivers. Environ. Sci. Technol. 48,
5025–5032.
Watson, J.G., Chow, J.C., 2001. Source characterization of major emission sources in the
Imperial and Mexicali Valleys along the US/Mexico border. Sci. Total Environ. 276,
33–47.
Watson, J.G., Chow, J.C., Houck, J.E., 2001. PM2.5 chemical source profiles for vehicle ex-
haust, vegetative burning, geological material, and coal burning in Northwestern Col-
orado during 1995. Chemosphere 43, 1141–1151.
Xiao, S., Wang, Q.Y., Cao, J.J., Huang, R.J., Chen, W.D., Han, Y.M., et al., 2014. Longterm
trends in visibility and impacts of aerosol composition on visibility impairment in
Baoji, China. Atmos. Res. 149, 88–95.
Xie, Y.L., Hopke, P.K., Paatero, P., Barrie, L.A., Li, S.M., 1999. Identification of source nature
and seasonal variations of arctic aerosol by positive matrix factorization. J. Atmos. Sci.
56, 249–260.
Xu, J., DuBois, D., Pitchford, M., Green, M., Etyemezian, V., 2006. Attribution of sulfate
aerosols in federal class I areas of the western United States based on trajectory re-
gression analysis. Atmos. Environ. 40, 3433–3447.
Xu, J., Zhang, Q., Chen, M., Ge, X., Ren, J., Qin, D., 2014. Chemical composition, sources, and
processes of urban aerosols during summertime in northwest China. Insights from
high-resolution aerosol mass spectrometry. Atmos. Chem. Phys. 14, 12593–12611.
Yan, B., Zheng, M., Hu, Y.T., Ding, X., Sullivan, A.P., Weber, R.J., et al., 2009. Roadside, urban,
and rural comparison of primary and secondary organic molecular markers in ambi-
ent PM2.5. Environ. Sci. Technol. 12, 4287–4293.
Yang, L., He, B.J., Ye, M., 2014. Application research of ECOTECT in residential estate plan-
ning. Enegr Buildings 72, 195–202.
Zeng, F., Shi, G.L., Li, X., Feng, Y.C., Bi, X.H., Wu, J.H., et al., 2010. Application of a combined
model to study the source apportionment of P-M10 in Taiyuan, China. Aerosol Air
Qual. Res. 10, 177–184.
Zhang, D., Shi, G., Iwasaka, Y., Hu, M., Zang, J., 2005. Anthropogenic calcium particles ob-
served in Beijing and Qingdao, China. Water Air Soil Pollut. 5, 261–276.
Zhang, R., Jing, J., Tao, J., Hsu, S.C., Wang, G., Cao, J., et al., 2013. Chemical characterization
and source apportionment of PM2.5 in Beijing., seasonal perspective. Atmos. Chem.
Phys. 13, 7053–7074.