Atmospheric Environment 94 (2014) 402e408

Contents lists available at ScienceDirect

Atmospheric Environment
journal homepage: www.elsevier.com/locate/atmosenv

Estimation of regional background concentration of CO2 at Lin'an

Station in Yangtze River Delta, China
Jing-Jiao Pu a, Hong-Hui Xu a, *, Jun He b, Shuang-Xi Fang c, Ling-Xi Zhou c, *
a

Zhejiang Institute of Meteorological Sciences, Hangzhou, China

University of Nottingham Ningbo China, Ningbo, China
c
Chinese Academy of Meteorological Sciences (CAMS), China Meteorological Administration (CMA), Beijing, China
b

h i g h l i g h t s

A new method is developed to estimate CO2 background concentration.

The method is established on the observations of black carbon concentration and meteorological data.

The method is better than the statistical method using R software in estimating CO2 background concentration.

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 8 January 2014
Received in revised form
19 May 2014
Accepted 20 May 2014
Available online 21 May 2014

A new method of extracting regional background concentration of CO2 in Yangtze River Delta was
established based on the observations of both black carbon concentration and meteorological parameters. The concentrations of CO2 and black carbon were observed at Lin'an regional background station
from 2009 to 2011. The regional background concentration of CO2 in Yangtze River Delta was obtained by
means of this new method, and the impact of human activities on CO2 concentration in this area was also
assessed. The results showed that the regional background concentration of CO2 extracted by this
approach was comparable to the values obtained by R statistical lter method, and moreover this new
method was better at picking up episodes heavily polluted by anthropogenic emissions. The annual
regional average background concentration of CO2 in Yangtze River Delta from 2009 to 2011 was
approximately 404.7 8.2 ppm, 405.6 5.3 ppm and 407.0 5.3 ppm, respectively, much higher than
global average value, indicating the distinct characteristic of this region. The anthropogenic emissions
from Yangtze River Delta had signicant inuence on the concentration of CO2, increasing the value
roughly 9.1 ppm higher than the regional background concentration of this area.
2014 Elsevier Ltd. All rights reserved.

Keywords:
CO2
Regional background concentration
Black carbon
Yangtze River Delta

1. Introduction
CO2 is one of the important long-lived greenhouse gases, whose
concentration has increased by 39% since the industrial revolution.
Until 2011, the global average of CO2 concentration has reached
390.9 ppm (World Meteorological Organization, 2012). The main
sources of CO2 released to the atmosphere are combustion of fossil
fuels, cement production and land use change (Forster et al., 2007),
while the main carbon sink is the absorption of CO2 by plants and
ocean. Owing to the human activities playing an important role in
the level of atmospheric CO2 concentration, it is necessary to

* Corresponding authors.
E-mail addresses: forsnow@126.com (H.-H. Xu), zhoulx@cams.cma.gov.cn
(L.-X. Zhou).
http://dx.doi.org/10.1016/j.atmosenv.2014.05.060
1352-2310/ 2014 Elsevier Ltd. All rights reserved.

establish an appropriate method to estimate the regional background concentration of CO2, in order to acquire the value of CO2
background concentration in this region and to assess the impact of
human activities on the CO2 concentration.
There are mainly three kinds of methods to estimate the background concentration of greenhouse gases: statistical method,
meteorological method and numerical model method. Statistical
method basically uses statistical analyses of the eld data for estimation. The background concentration can be determined based on
local linear regression of the data by using statistical software (e.g.
R software) or estimated according to the threshold values of
pollutant concentration variation (Reimann et al., 2004; Fang et al.,
2011; Tsutsumi et al., 2006; Giostra et al., 2011). Meteorological
method identies background concentration by meteorological
parameters, such as surface wind direction and air mass backward
o
v et al., 2008).
trajectories (Tsutsumi et al., 2006; Balzani Lo

J.-J. Pu et al. / Atmospheric Environment 94 (2014) 402e408

Numerical model method gets the baseline concentration by using

an atmospheric dispersion model to simulate the inuence of
emission sources of local or regional scale on observation data
(Giostra et al., 2011; Manning et al., 2003, 2011; Ryall et al., 2001).
The statistical method use some form of statistical analyses on the
data to lter out polluted peaks without any assumptions about the
origin of the data. It is not clear if this technique can remove all
polluted episodes and it may not necessarily distinguish low levels
of polluted air from measurement noise. The method based on
ground meteorological parameters has some disadvantages. A
backward trajectory only represents one of the many possible trajectories that an air parcel could take route to the targeted area; in
reality, vertical and horizontal mixing will transport those air
masses from a larger scale to affect the local or regional atmosphere. In addition, this method uses wind elds taken from global
scale numerical prediction models to calculate the possible backward trajectories but does not resolve small-scale ow features
which may produce false indications of clean air streams (Ryall
et al., 2001). Numerical model method is set on Lagrange particle
dispersion model Numerical Atmospheric Dispersion Modeling
Environment (NAME), which requires the United Kingdom Meteorological Ofce's (UKMO) numerical weather prediction model to
provide the meteorological eld as the model input. The random
walk technique used in the model NAME needs to release a large
amount of particles in order to obtain stable statistical results of the
pollutant dispersion, which requires lots of computing resource. At
present, this method is mainly applied in Europe (Giostra et al.,
2011; Manning et al., 2003, 2011; Ryall et al., 2001; Cai et al., 2003).
Yangtze River Delta of China is nearly developed and densely
populated. Considerable amount of CO2 is being emitted from
various anthropogenic sources which include fossil fuel combustion, cement production and biomass burning, etc. In addition, CO2
level in this area might be affected by the land use change. CO2
emission from cement production derived from industrial process
takes up 7.8% (Li et al., 2011; Liu et al., 2010; Statistical Bureau of
Jiangsu Province, 2012; Statistical Bureau of Zhejiang Province,
2012; Statistical Bureau of Shanghai, 2012). The rate of forest
coverage in Yangtze River Delta stays at a level of 33% or so (Chinese
Statistical Bureau, 2012). In other words, the emission of CO2 by
cement production and the uctuation of CO2 concentration caused
by land use change might be relatively minor. As the industry in
Yangtze River Delta is still developing fast and the burning of crop
residues in the country side occurs frequently and widely in this
region, the fossil fuel combustion and biomass burning could be the
predominant sources for CO2 emission. Black carbon (BC) is one
type of primary pollutants derived from incomplete combustion of
fossil fuels and biomass. In this area, the anthropogenic emissions
of BC and CO2 have the similar spatial distribution (Qin et al., 2012;
European Commission, Joint Research Centre/Netherlands
Environmental Assessment Agency, 2011). A new integrated
method is set up to lter out those anthropogenically inuenced
polluted air mass, based on the BC concentrations and relevant
meteorological parameters, aiming to obtain the regional background CO2 concentrations in Yangtze River Delta.
2. Experimental
2.1. Sampling
The sampling place is located at the Lin'an Regional Atmospheric
Background Station (119 440 E, 30180 N, 138.6 m asl), which is in the
southern part of Yangtze River Delta, about 60 km away to the west
of Hangzhou and about 150 km away to the southwest of Shanghai.
The station is surrounded by hilly lands, forest and farming areas,
with heavy vegetation coverage. There is no large village around

403

within 3 km. The prevailing wind direction is northeast and

southwest. Special weather phenomenon rarely occurs and the
frequency of thermal inversion is low, with distinct meteorological
characteristics of subtropical monsoon climate and atmospheric
circulation.
The equipment for in-situ observation of CO2 concentration is
G1301 Picarro analyzer. The measuring technique is based on cavity
ring-down spectroscopy. The device comprises three main parts; a
laser, a high nesse optical cavity with two or more mirrors, and a
photo-detector. The laser light is injected into the cavity through
one partially reecting mirror. This light builds up over time in the
cavity and is monitored by a photo-detector through a second
partially reecting mirror. After the laser is rapidly turned off, most
of the light remains trapped in the cavity, and the decaying light
intensity due to cavity mirror loss as well as sample absorption and
scattering is measured. The accuracy of CO2 measured with this
technique is lower than 0.05 ppm per 5 min, and the value of
maximum drift peak reaches 0.5 ppm per month. The time interval
of measurement is 4 s. The more detailed descriptions of the whole
analytical system can be found in Fang et al. (2013). Black carbon
(BC) concentration was measured by aethalometer AE-31 (American Magee Scientic). There are seven wavelengths that could be
used for observations, 370 nm, 470 nm, 520 nm, 590 nm, 660 nm,
880 nm and 950 nm, respectively. The sampling ow rate was 5 L/
min, with 2.5 mm cutting head in the air inlet. The time interval of
measurement was 5 min. The instrument aspirates ambient air
using its inlet tube and black carbon mass concentration is estimated by measuring the change in the transmittance of a quartz
lter tape, on to which particles impinge. These data are automatically recorded in the ash card of the instrument and displayed
on the screen. The data analyzed in this study was observed from
the 880 nm channel which is widely used for the standard BC
measurement of the aethalometer (Hansen, 2005).
The above two co-located measurements were conducted
simultaneously from 1 January 2009 to 31 December 2011.
The meteorological data, including precipitation and surface
wind speed, is collected from the observations of meteorological
station which is located within the Lin'an Regional Atmospheric
Background Station.
2.2. Quality control
For the CO2 measurement, the calibration of the observation
system was conducted regularly every 12 h by measuring a set of
standard working gases which are traceable to WMO/GAW standard gas of level one, and concentrations were corrected slightly
from these calibrations (Fang et al., 2013). Moreover, the outliers
were picked out according to the equipment performance log when
there were instrument troubles or anthropogenic inuences, such
as instrument maintenance or on-site visit.
For the BC measurement, the calibration of the ow rate was
done biweekly at normal condition. However, when it became
noisy during continuous operation, the ow rate was calibrated
accordingly.
For the meteorological measurement, routine maintenance of
the instruments was carried out regularly so that the quality of
observations would not deteriorate signicantly. Routine quality
control checks were also carried out to detect the equipment faults.
2.3. Method of selecting regional background concentration of CO2
During the sampling period from 1 January 2009 to 31
December 2011, the concentrations of both CO2 and BC at Lin'an
Atmospheric Background Station had been observed for 24.092 h,
taking up 91.7% of the total physical hours. The hourly mean

404

J.-J. Pu et al. / Atmospheric Environment 94 (2014) 402e408

concentrations of CO2 and BC are shown in Fig. 1. The range of CO2

concentration was from 369.7 ppm to 488.6 ppm, and the BC
concentrations were ranging between 15.2 ng/m3 and 29640.0 ng/
m3. The two major sources of BC are thought to be fossil fuel
combustion and biomass burning (Cooke et al., 1996; Sharma et al.,
2012). Solid bio-fuel combustion is claimed to be the largest source
of black carbon emission in India (Venkataraman et al., 2005). Fossil
fuels are said to be responsible for about 55%e65% of BC emissions
in China (Qin and Xie, 2012). In this study, the correlation coefcient of the concentrations of these two pollutants arrives at 0.53
with 99% condence intervals and it shows both pollutants were
moderately positively correlated. It could be understandable that
both pollutants have some common main sources as mentioned
above; in addition, the advection and convection of these pollutants occurs under the same meteorological conditions in this highly
developed area. Since BC is thought to be mainly affected by
anthropogenic polluted air masses (Balakrishnaiah et al., 2011;
Dumka et al., 2010; Cao et al., 2009; Zhao et al., 2012), these episodes can be eliminated as the rst step to estimate the CO2
background concentration. Secondly, when wet precipitation occurs, the BC concentration might sharply decline because of rain
scavenging process that may not directly reect if the CO2 level is
affected by any anthropogenic sources as the scavenging ratio of
gaseous CO2 is quite limited (Jurado et al., 2008); hence, under the
circumstances of wet precipitation, Hysplit 4.9 was used to calculate 48 h air mass backward trajectory along with NCEP (National
Centers for Environmental Prediction) reanalyzed meteorological
data to lter out episodes strongly affected by anthropogenic
emissions. Thirdly, calm wind implies relatively stable atmospheric
condition, so the cases under calm wind were also excluded for the
selection of CO2 background concentration.
3. Results and discussion
3.1. CO2 regional background concentration
Fig. 2 exhibits the hourly mean concentration of BC at Lin'an
station from 1 January 2009 to 31 December 2011. It can be seen
from the gure that almost all the hourly average or median BC
concentrations were below 5000 ng/m3. Hence, in this study
5000 ng/m3 is considered as the threshold of BC concentration. The
period when the BC concentration was over 5000 ng/m3 is thought
to be polluted. In total there are 8120 h excluded for background

Fig. 1. Hourly mean concentrations of CO2 and BC at Lin'an station from 2009 to 2011.

Fig. 2. Hourly mean concentration of BC at Lin'an station from 2009 to 2011 (The
upper and lower quartiles (25% and 75%) are represented by the box, the central red
line of each box is the median, and the black is the average. The dash lines extended
from both sides of the box show those extreme data points but not considered as
outliers. Red cross points are outliers which are larger than q3 1.5*(q3  q1) or
smaller than q1  1.5*(q3  q1), where q1 and q3 are the 25th and 75th percentiles).
(For interpretation of the references to color in this gure legend, the reader is referred
to the web version of this article.)

concentration measurements, accounting for 33.7% of the total

observation hours. Afterward, among the events with wet precipitations, those episodes with strong contributions from
anthropogenic emissions were also eliminated based on the air
mass backward trajectory analysis as shown in Fig. 3. In total 2019 h
were ltered out. After that, the events with calm wind atmospheric conditions appeared for 266 h, which have also been
excluded for selection of CO2 background events. Finally, 7108 h
when the wind speed was less than 2 m/s are excluded for they
were thought to be local events (Tsutsumi et al., 2006). After all the
above mentioned ltering processes, there are in total 6579 h left,
accounting for 27.3% of all sampling hours in this study. The temporal distribution of CO2 regional background concentrations could
be obtained as shown in Fig. 4.
3.2. Characteristics of regional CO2 background concentration in
Yangtze River Delta
Fig. 5a describes the diurnal variation of regional CO2 background concentration observed at Lin'an station. It can be seen from
the gure that the diurnal trend of CO2 background concentration
at Lin'an station varied in the range between 397.9 9.2 ppm
(16:00 PM) and 415.9 12.1 ppm (4:00 AM). The diurnal amplitude
was quite signicant with the peak of 31.1 ppm in summer and the
minimum of 6.7 ppm in winter. This result is quite similar to the
measurement at Fraserdale site, which is in the northern Ontario
boreal forest in Canada (Higuchi et al., 2003). During the growing
season, regional biospheric inuence is quite evident, resulting in
higher daily variation of CO2 concentration in summer. Lin'an station is located in a subtropical monsoon climate region with heavy
vegetation coverage, which may affect the CO2 diurnal trend predominantly through the plant photosynthesis and respiration.
Meanwhile, the daily evolution of meteorological conditions could
take great effect on the CO2 concentration. At night time, stable
conditions would lead to higher concentration, especially when
thermal inversion occurs, and in the day time, the expansion of
mixing height would be benecial to decrease CO2 concentration.

J.-J. Pu et al. / Atmospheric Environment 94 (2014) 402e408

405

Fig. 4. Background and non-background concentrations of CO2 at Lin'an station.

concentration in Yangtze River Delta was almost consistent with

that at global atmospheric background stations, with the peak and
bottom values in May and August, respectively. However, there was
some difference between the CO2 background concentrations in
Yangtze River Delta and those values from global background stations. At Lin'an station, the sub-peak concentration of CO2 showed
up during the period from November to February, which was absent at both Waliguan station and Mauna Loa station. It might be
because during the winter time more CO2 could be emitted as the

Fig. 3. (a) Anthropogenic emissions of CO2 surrounding the sampling site; (b) Air mass
backward trajectories affecting the sampling site.

Fig. 5b shows the monthly mean background and nonbackground concentration of CO2. The annual amplitude of CO2
background concentration was 7.7 ppm, with the lowest value
401.9 ppm in September, the second lowest value of 402.6 ppm in
August and the highest concentration 409.5 ppm in January and
February, the second highest value of approximately 408.8 ppm in
May. From 2009 to 2011, the global CO2 monthly mean concentration trend showed as a monomodal curve (CMDL/NOAA, 2012a;
CMDL/NOAA, 2012b). At Waliguan station in Qinghai Province,
China, the peak concentration of CO2 appears in April or May,
meanwhile the lowest value appears in July or August, with the
annual amplitude of 8.9 ppm; at Mauna Loa station in Hawaii, the
highest value of CO2 concentration appears in May, and the lowest
value appears in October, with the annual amplitude of 5.6 ppm.
Therefore, the monthly temporal trend of CO2 background

Fig. 5. (a) Diurnal trend of CO2 background concentration at Lin'an station; (b)
Monthly average of CO2 background and non-background concentration at Lin'an
station and CO2 concentrations at Mauna Loa station and Waliguan station.

406

J.-J. Pu et al. / Atmospheric Environment 94 (2014) 402e408

increment of energy consumption that could lead to the increased

CO2 background concentrations in this area. In addition, centralized
heating systems start to work during this period in North China
which could input a lot more CO2 emissions; the prevailing wind
direction in winter for this region is northwest and continental air
mass may bring more CO2 into the Yangtze River Delta. From 2009
to 2011, the annual regional CO2 background concentration of
Yangtze River Delta was about 404.7 8.2 ppm, 405.6 5.3 ppm
and 407.0 5.3 ppm, respectively, signicantly higher than the
global average value 386.8 ppm (2009), 389.0 ppm (2010) and
390.9 ppm (2011) (World Meteorological Organization, 2010, 2011,
2012).
3.3. Comparison with the CO2 background concentrations selected
out by R statistical method
The CO2 background concentration can be obtained by statistical
software, e.g. R software. Fig. 6a presents the temporal variation of
CO2 background concentration at Lin'an station from 2009 to 2011
by R statistical method. This method is based on robust local
regression estimation, and the function rfbaseline from R statistical package is used here with a parameter NoXP 1440, which
corresponds to a local time window of about 60 days. The more
detailed description of this method can be seen in Fang et al. (2011).
By this statistical ltering process, 11,034 h were excluded accounting for 45.8% of all the sampling hours. The CO2 background
concentration in Yangtze River Delta was 406.9 ppm in the period
from 2009 to 2011, about 1.1 ppm higher than the value obtained by

Fig. 7. Diurnal distribution of CO2 background and non-background concentrations at

Lin'an station.

the new method based on BC concentrations, air mass backward

trajectories and meteorological parameters, which indicated that
the results from these two approaches of estimating CO2 background concentrations are comparable. Therefore, it is believable
that the regional CO2 background concentration of Yangtze River
Delta is about 406 ppm.
At the same time, the statistical method, e.g. by using R software, is more disadvantageous to be used in the area with heavy
vegetation coverage, for the diurnal uctuation of CO2 concentration in these regions is signicant due to the plant photosynthesis
and respiration, but the statistical method would eliminate the
situation when there is obvious uctuation of CO2 concentration.
Thus, the diurnal distribution of CO2 concentration remains relatively stable by statistical method (shown in Fig. 6b). In comparison,
the new method used in this study might exclude those nonbackground events (i.e. sampling periods inuenced by anthropogenic emissions) in a more complete manner by considering the
correlation between BC and CO2 concentrations, air mass backward
trajectories and atmospheric stable conditions. Then, the diurnal
trends of CO2 background concentrations owing to photosynthesis
and respiration of green plants could be seen clearly (shown in
Fig. 5a). Moreover, statistical method is not suitable for use when
CO2 concentration has less variation due to the effects of both
carbon source and sink. For example, it is difcult to distinguish
whether or not the CO2 concentration is affected by anthropogenic
pollution in the afternoon because the carbon sink is strengthened
in that period.
3.4. Inuence of anthropogenic emissions on CO2 concentration in
Yangtze River Delta

Fig. 6. (a) Background and non-background CO2 concentrations by R statistical method

at Lin'an station; (b) Diurnal distribution of CO2 background concentration by R statistical method at Lin'an station.

In Fig. 7, the diurnal distribution of CO2 non-background concentration was between 406.6 ppm and 426.0 ppm, with the
maximum appearing at 6:00 AM and the minimum showing up at
15:00 PM. The diurnal amplitude was 19.4 ppm. Therefore, the
diurnal variation trend of CO2 background concentrations was
similar to that of CO2 non-background concentrations. It might be
because both trends were mainly affected by diurnal variation of
meteorological conditions and plant photosynthesis/respiration
processes (Higuchi et al., 2003; Henninger et al., 2010; Miyaoka
et al., 2007; Aikawa et al., 1995). In the early morning, thermal
inversion was present and plant respiration was strong, hence the
CO2 concentrations of both background and non-background were
high. In the afternoon, the depth of atmospheric mixing layer was
higher and the dispersion condition was better, and plant

J.-J. Pu et al. / Atmospheric Environment 94 (2014) 402e408

photosynthesis was stronger, consequently both the background

and non-background CO2 concentrations were lower. During the
period when it was heavily affected by anthropogenic polluted air
masses, non-background CO2 concentrations rose signicantly,
about 9.1 ppm over than the background values.
In Fig. 5b, the seasonal distribution of CO2 non-background concentrations was similar to that of CO2 background concentrations.
The minimum monthly average of non-background concentrations
was in August and September, approximately with a value of
409.2 ppm, and the maximum value was in December, followed by in
May and January, which were above 417.5 ppm. The annual amplitude of monthly mean concentrations was 11.5 ppm, much higher
than that of the background concentrations. Hence it can be
concluded that air masses polluted by anthropogenic emissions
inuenced non-background CO2 concentrations signicantly,
resulting in 9 ppm more than the background baselines in this region.
4. Conclusion
A new method was developed in order to estimate the regional
CO2 background concentration in Yangtze River Delta. The sampling site is in the southern part of Yangtze River Delta, which is
named Lin'an Regional Atmospheric Background Station. By
analyzing the simultaneous measurements of both CO2 and BC
concentrations from 2009 to 2011, the correlation coefcient of the
concentrations of these two pollutants showed that they were
moderately positively correlated, for fossil fuel combustion and
biomass burning are thought to be the predominant sources for
both of them. So a method based on BC concentration combined
with meteorological parameters is used here to estimate the
regional CO2 background concentration in Yangtze River Delta. In
comparison to the results from statistical method by R package, this
method is better at excluding the inuence of air mass polluted by
anthropogenic emissions, and also it could do better in representing the diurnal variation of CO2 concentration caused by daily cycle
of the plant photosynthesis/respiration and the daily evolution of
boundary mixing height.
The annual regional CO2 background concentration in Yangtze
River Delta obtained by the new method was about
404.7 8.2 ppm (2009), 405.6 5.3 ppm (2010) and
407.0 5.3 ppm (2011), respectively, signicantly higher than the
globally averaged concentration. The diurnal and seasonal variation
of regional CO2 background concentration was characterized
distinctly. The diurnal trend of CO2 background concentration
varied with the peak at 4:00 AM and the bottom at 16:00 PM; the
seasonal trend showed as a dualmodal curve, with the lowest value
in September/August and the highest concentration in January/
February/May.
Anthropogenic emissions would make great effect on CO2 concentration in Yangtze River Delta. When the air mass polluted by
anthropogenic emission passed by, CO2 concentration rose up
signicantly, about 9.1 ppm over than the background
concentration.
Acknowledgments
The authors express their great thanks to the staff of Lin'an
Regional Atmospheric Background Station, who have contributed
to observations of greenhouse gases and the staff of key laboratory
for atmospheric chemistry (CMA), who performed calibration and
quality control of the data. The authors also thank Mauna Loa in the
United States (Hawaii) and Waliguan in China for providing
monthly CO2 data from 2009 to 2011 which can be downloaded
from World Data Centre for Greenhouse Gases (WDCGG). This work
was supported by the Strategic Priority Research Program (B) of

407

the Chinese Academy of Sciences (No. XDB05020403), the China

Special Fund for Meteorological Research in the Public Interest (No.
GYHY201206011) and the Zhejiang Provincial Natural Science
Foundation (No. LY12D05001, No. LQ13D050001).
References
Aikawa, M., Yoshikawa, K., Tomida, M., et al., 1995. Continuous monitoring of the
carbon dioxide concentration in the urban atmosphere of Nagoya, 1991e1993.
Anal. Sci. 11, 357e362.
Balakrishnaiah, G., Kumar, K.R., Reddy, B.S.K., et al., 2011. Anthropogenic impact on
the temporal variations of black carbon and surface aerosol mass concentrations at a tropical semi-arid station in southeastern region of India. J. Asian
Earth Sci. 42 http://dx.doi.org/10.1016/j.jseaes.2011.07.016.
o
v, J.M., Henne, S., Legreid, G., et al., 2008. Estimation of background
Balzani Lo
concentrations of trace gases at the Swiss Alpine site Jungfraujoch (3580 m asl).
J. Geophys. Res. 113 http://dx.doi.org/10.1029/2007JD009751.
Cai, X.H., Chen, J.Y., Kang, L., 2003. An atmospheric diffusion model for conditions of
nuclear accident. Radiat. Prot. 23, 293e299.
Cao, J.J., Zhu, C.S., Chow, J.C., et al., 2009. Black carbon relationships with emissions
and meteorology in Xi'an, China. Atmos. Res. 94 http://dx.doi.org/10.1016/
j.atmosres.2009.05.009.
Chinese Statistical Bureau, 2012. Statistical Yearbook of China 2012. Chinese Statistical Press, Beijing, p. 400.
CMDL/NOAA, 2012a. Atmospheric CO2 Monthly Mean Concentration, Mt. Waliguan.
In: WMO WDCGG DATA. Japan Meteorological Agency, Tokyo. http://ds.data.
jma.go.jp/gmd/wdcgg/.
CMDL/NOAA, 2012b. Atmospheric CO2 Monthly Mean Concentration, Mauna Loa.
In: WMO WDCGG DATA. Japan Meteorological Agency, Tokyo. http://ds.data.
jma.go.jp/gmd/wdcgg/.
Cooke, W.F., Wilson, J.J.N., 1996. A global black carbon aerosol model. J. Geophys.
Res. 101 http://dx.doi.org/10.1029/96JD00671.
Dumka, U.C., Moorthy, K.K., Kumar, R., et al., 2010. Characteristics of aerosol black
carbon mass concentration over a high altitude location in the Central Himalayas from multi-year measurements. Atmos. Res. 96 http://dx.doi.org/10.1016/
j.atmosres.2009.12.010.
European Commission, Joint Research Centre (JRC)/Netherlands Environmental
Assessment Agency (PBL), 2011. Emission Database for Global Atmospheric
Research (EDGAR), Release Version 4.2. http://edgar.jrc.ec.europa.eu.
Fang, S.X., Zhou, L.X., Masarie, K.A., et al., 2013. Study of atmospheric CH4 mole
fractions at three WMO/GAW stations in China. J. Geophys. Res. 118 http://
dx.doi.org/10.1002/jgrd.50284.
Fang, S.X., Zhou, L.X., Zang, K.P., et al., 2011. Measurement of atmospheric CO2
mixing ratio by cavity ring-down spectroscopy (CRDS) at the 4 background
stations in China. Acta Sci. Circumstantiae 31, 624e629.
Forster, P., Ramaswamy, V., Artaxo, P., et al., 2007. Changes in atmospheric constituents and in radiative forcing. In: Climate Change 2007: the Physical Science
Basis. Contribution of Working Group 1 to the Fourth Assessment Report of the
Intergovernmental Panel on Climate Change. Cambridge University Press,
Cambridge, pp. 511e513.
Giostra, U., Furlani, F., Arduini, J., et al., 2011. The determination of a regional
atmospheric background mixing ratio for anthropogenic greenhouse gases: a
comparison of two independent methods. Atmos. Environ. 45 http://dx.doi.org/
10.1016/j.atmosenv.2011.06.076.
Hansen, A.D.A., 2005. The Aethalometer Operating Manual. http://www.
mageesci.com/images/stories/docs/Aethalometer_book_2005.07.03.pdf.
Henninger, S., Kuttler, W., 2010. Near surface carbon dioxide within the urban area of
Essen, Germany. Phys. Chem. Earth 35. http://dx.doi.org/10.1016/j.pce.2010.03.006.
Higuchi, K., Worthy, D., Chan, D., et al., 2003. Regional source/sink impact on the
diurnal, seasonal and inter-annual variations in atmospheric CO2 at a boreal
forest site in Canada. Tellus 55B, 115e125.
Jurado, E., Dachs, J., Duarte, C.M., et al., 2008. Atmospheric deposition of organic and
black carbon to the global oceans. Atmos. Environ. 42 http://dx.doi.org/10.1016/
j.atmosenv.2008.07.029.
Li, X., Shi, J.P., Lv, S.Z., et al., 2011. CO2 emissions from the Chinese cement industry
and methods to reduce them. Acta Sci. Circumstantiae 31, 1115e1120.
Liu, Z.C., Wang, A.J., Yu, W.J., et al., 2010. Research on regional carbon emissions in
China. Acta Geosci. Sin. 31, 727e732.
Manning, A.J., O'Doherty, S., Jones, A.R., et al., 2011. Estimating UK methane and
nitrous oxide emissions from 1990 to 2007 using an inversion modeling
approach. J. Geophys. Res. 116 http://dx.doi.org/10.1029/2010JD014763.
Manning, A.J., Ryall, D.B., Derwent, R.G., 2003. Estimating European emissions of
ozone-depleting and greenhouse gases using observations and a modeling
back-attribution technique. J. Geophys. Res. 108 http://dx.doi.org/10.1029/
2002JD002312.
Miyaoka, Y., Inoue, H.Y., Sawa, Y., et al., 2007. Diurnal and seasonal variations in
atmospheric CO2 in Sapporo, Japan: anthropogenic sources and biogenic sinks.
Geochem. J. 41, 429e436.
Qin, Y., Xie, S.D., 2012. Spatial and temporal variation of anthropogenic black carbon
emissions in China for the period 1980e2009. Atmos. Chem. Phys. 12 http://
dx.doi.org/10.5194/acp-12-4825-2012.
Reimann, S., Schaub, D., Stemmler, K., et al., 2004. Halogenated greenhouse gases at
the Swiss High Alpine Site of Jungfraujoch (3580 m asl): continuous

408

J.-J. Pu et al. / Atmospheric Environment 94 (2014) 402e408

measurements and their use for regional European source allocation.

J. Geophys. Res. 109 http://dx.doi.org/10.1029/2003JD003923.
Ryall, D.B., Derwent, R.G., Manning, A.J., et al., 2001. Estimating source regions of
European emissions of trace gases from observations at Mace Head. Atmos.
Environ. 35, 2507e2523.
Sharma, R.K., Bhattarai, B.K., Sapkota, B.K., et al., 2012. Black carbon aerosols variation in Kathmandu valley, Nepal. Atmos. Environ. 63, 282e288.
Statistical Bureau of Jiangsu Province, 2012. Statistical Yearbook of Jiangsu Province.
Chinese Statistical Press, Beijing, pp. 385e386.
Statistical Bureau of Shanghai, 2012. Statistical Yearbook of Shanghai. Chinese
Statistical Press, Beijing, pp. 229e231.
Statistical Bureau of Zhejiang Province, 2012. Statistical Yearbook of Zhejiang
Province. Chinese Statistical Press, Beijing, pp. 265e267.
Tsutsumi, Y., Mori, K., Ikegami, M., et al., 2006. Long-term trends of greenhouse
gases in regional and background events observed during 1998e2004 at

Yonagunijima located to the east of the Asian continent. Atmos. Environ. 40

http://dx.doi.org/10.1016/j.atmosenv.2006.04.036.
Venkataraman, C., Habib, G., Eiguren-Fernandez, A., et al., 2005. Residential biofuels
in south Asia: carbonaceous aerosol emissions and climate impacts. Science
307. http://dx.doi.org/10.1126/science.1104359.
World Meteorological Organization, 2010. Greenhouse Gas Bulletin. http://www.
wmo.int/pages/prog/arep/gaw/ghg/GHGbulletin.html.
World Meteorological Organization, 2011. Greenhouse Gas Bulletin. http://www.
wmo.int/pages/prog/arep/gaw/ghg/GHGbulletin.html.
World Meteorological Organization, 2012. Greenhouse Gas Bulletin. http://www.
wmo.int/pages/prog/arep/gaw/ghg/GHGbulletin.html.
Zhao, S., Ming, J., Xiao, C., et al., 2012. A preliminary study on measurements of
black carbon in the atmosphere of northwest Qilian Shan. J. Environ. Sci. 24
http://dx.doi.org/10.1016/S1001-0742(11)60739-0.