Environ Monit Assess (2014) 186:7505–7512 DOI 10.1007/s10661-014-3943-y
Elemental analysis of aerosols in Tehran’s atmosphere using PIXE and identification of pollution sources N. Esmaili & S. Khashman & M. Lamehi-Rachti & D. Agha Aligol & F. Shokouhi & P. Oliaiy & M. Farmahini Farahani
Received: 9 February 2014 / Accepted: 8 July 2014 / Published online: 17 July 2014 # Springer International Publishing Switzerland 2014
Abstract In this study, the proton-induced X-ray emission (PIXE) technique has been applied to measure the elemental composition and concentrations of particulate matter of 220 samples of aerosols in Tehran’s atmosphere within a 450-day time interval starting from March 2009 and ending in June 2010, covering all four seasons. PIXE analysis shows the samples are comprised of various elements including Al, Si, S, Cl, K,
Electronic supplementary material The online version of this article (doi:10.1007/s10661-014-3943-y) contains supplementary material, which is available to authorized users. N. Esmaili : M. Lamehi-Rachti (*) : D. Agha Aligol : F. Shokouhi : P. Oliaiy : M. Farmahini Farahani Nuclear Science Research School, Nuclear Science & Technology Research Institute (NSTRI), Tehran, Iran e-mail: [email protected] N. Esmaili e-mail: [email protected] D. Agha Aligol e-mail: [email protected] F. Shokouhi e-mail: [email protected] P. Oliaiy e-mail: [email protected] M. Farmahini Farahani e-mail: [email protected] S. Khashman Department of Physics, Faculty of Sciences, Central Tehran Branch, Islamic Azad Universit, Tehran, Iran e-mail: [email protected]
Ca, Ti, V, Cr, Mn, Fe, Ni, Cu, Zn, Br, Rb, Sr, and Pb. Also, to obtain more information about the sources of pollution and to identify the major sources of urban particulate matter, principal component analysis (PCA) was used. Furthermore, micro-PIXE was performed to study individual aerosols in some samples. Results revealed that the concentration of elements originating from vehicle emissions increases three times in winter; whereas the concentration of elements with soil origin remains constant. Based on wind rose maps, it is inferred that the high concentrations of the elements Al, Si, K, Ca, Ti, Mn, and Fe are associated with natural dust brought by winds into Tehran from the west. Keywords Aerosol . Air pollution . PIXE . Elemental composition . Principal component analysis
Introduction The city of Tehran, with a population of over seven million, covers an area of approximately 730 km2, containing about three million vehicles, two million motorcycles (Tajrishy and Abrishamchi 2005), and many factories. Air pollution is a major environmental problem in Tehran. Studies show that the main sources of air pollution in megacities are traffic, energy production using fossil fuels and biomass, industrial sources, and re-suspension of soil (Moreno et al. 2007; Sinha and Nag 2011). The Government of Iran and Tehran Municipality have actively participated in an effort for the reduction of local and global air pollution. Atash
7506
(2007) reviewed and evaluated the implementation of the 10-year master plan to control air pollution in Tehran. Air pollutants are emitted into the atmosphere from stationary sources (utility, industrial, institutional, and commercial facilities) and mobile sources (on-road vehicular traffic and transportation). Motor vehicles constitute the most dominant source of pollutants in Tehran, and it is estimated that emissions from mobile sources account for almost 85 % of all the air pollution in Tehran. The types and amounts of pollutants are directly related to the type and quality of fuel used for transportation or in industry and the level of technology involved (Halek et al. 2004; Malakooti 2010). The pollutants directly affect the health of people (Lim et al. 2005; Wang and Mullahy 2006), and their identification and measurement is especially important for any future action to reduce the harmful effects (Wildhaber 2006; Ramanakumar et al. 2007; Kampa and Castanas 2008). In order to assess the problems of urban air pollution in a global context, the World Health Organization (WHO) and the United Nations Environment Program (UNEP) have initiated a detailed study of air quality in many megacities of the world. This study is focused on the monitoring and measurement of sulfur dioxide (SO2), suspended particulate matter (SPM), and lead (Pb). However, Tehran was not included in this study because of lack of data (Mage et al. 1996). During the last decade, the elemental composition of atmospheric aerosol has been studied by different techniques (Yatin et al. 2000; Al-Momani et al. 2005; Basha Fig. 1 Seasonal variation of particulate mass (PM) in microgram per cubic meter during sampling days
Environ Monit Assess (2014) 186:7505–7512
et al. 2010; Li et al. 2010; Shah and Shaheen 2010). Ion beam analysis is known to be suitable for elemental analysis of aerosol and air pollutants (Ghermandi et al. 2005; Chiari et al. 2006; Dallarosa et al. 2008; Angyal et al. 2010; Nava et al. 2012). Recently, a few studies have been done about criteria pollutants in Tehran such as CO, NOX, SOX, and PM10 (Daryabeigi et al. 2007; Malakooti 2010; Karatzas et al. 2012), and few studies have been carried out about trace elements. Sekhavatjou et al. (2011) measured and compared concentration of trace elements in industrial and residential areas by neutron activation analysis. Also, Leili et al. (2008) have done an important study on the concentrations of TPM, PM10, and heavy metals in one station in the center of Tehran by atomic absorption spectrometry. Investigation of aerosols in the air of Tehran was defined as a research project in the annual program of this laboratory. In the present work, sampling of aerosols was performed on 220 working days from March 2009 to June 2010. Map of 22 districts of the Municipality of Tehran, the geographic location of the sampling station and the surface topography (Malakooti 2010) of Tehran and some surrounding areas (55 km×45 km, resolution of 1 km) are shown in Online Resource 1. The location of sampling was near a major highway in western Tehran. PIXE and micro-PIXE were used to determine the concentration of various elements in the samples. Principal component analysis was applied to obtain the sources of the aerosols and their contribution to air pollution in Tehran.
Environ Monit Assess (2014) 186:7505–7512
7507
Table 1 Averaged concentrations and standard deviation in nanogram per cubic meter of the detected elements in different seasons Elements
Spring (2009)
Summer (2009)
Autumn (2009)
Winter (2009)
Spring (2010)
Al
500±190
780±390
590±360
660±280
820±320
Si
2,500±1,000
3,480±1,580
2,420±920
3,410±1,580
3,460±1,350
S
830±230
1,030±430
1,160±650
1,420±820
750±230
Cl
430±140
780±540
610±320
810±540
490±270
K
980±440
1,530±600
970±390
1,210±530
1,240±480
Ca
7,700±3,550
11,370±4,810
7,800±2,990
9,980±4,370
9,950±3,940
Ti
240±100
310±150
190±80
260±110
340±160
V
10±10
20±10
10±10
30±30
20±10
Cr
10±10
20±10
10±10
10±10
10±10
Mn
90±30
120±50
90±40
110±50
110±50
Fe
3,250±1,420
4,460±1,980
2,930±1,110
3,990±1,710
4,760±2,440
Ni
20±10
20±10
10±10
20±30
20±20
Cu
40±10
50±30
60±30
80±50
40±20
Zn
180±70
270±150
380±260
530±390
270±190
Br
60±60
50±30
50±30
50±30
40±30
Rb
40±30
60±50
50±30
60±60
40±40
Sr
100±50
130±90
100±50
120±80
100±50
Pb
190±200
310±190
370±320
500±520
200±150
Materials and methods Atmospheric aerosol particles were collected on cellulose-fiber filters. Each filter has a diameter of 47 mm. The sampling was performed at a height of 1.5 m above ground. The filter holder was connected to a vacuum pump [GAST 0522-V3], which was operated at a pumping speed of approximately 3 m3/h. The sampling time was about 8 h from 8 A.M. to 4 P.M. (on working days). The location of sampling was near a major highway in western Tehran. Total particulate matter (TPM) was obtained by weighing each filter before and after sampling with a digital balance (model: Precisa 125A with 0.0001 g accuracy). The samples were bombarded with a 2-MeV proton beam and all measurements were carried out in vacuum of around 10−5 Torr. The beam current was about 6 nA. The X-rays emitted from the samples were detected by a Canberra Si (Li) detector at an angle of 135°. The energy resolution of detectors was 165 eV for the Fe (Kα) Xray. The quantitative analysis was performed by WinQXAS software (Edition IAEA. 1997–2001) and the detector efficiency, cross-sections, and numbers of incident protons were taken into account by RBS technique. To confirm the accuracy of the measurement, two
Fig. 2 Concentration of Zn and Pb in nanogram per cubic meter during sampling days
7508
reference materials (IAEA-SOIL7 and IAEA-SL1, lake sediment) were analyzed under the same conditions, and the results were in good agreement with the recommended values within an accuracy of 5 %. In order to gain further insight into the detailed characteristics of atmospheric aerosols, some samples were analyzed by micro-PIXE method (Agha-Aligol et al. 2007). To identify aerosol source, elemental concentrations were processed using principal component analysis (PCA). The statistical analysis was performed by the Statistics Package for Social Science (SPSS 12.0 Inc, USA). All elemental concentrations extracted from PIXE analysis were transformed to base-10 logarithms and used for PCA. These transformations compensate for the differences in magnitude between major elements and trace elements. Also, the wind rose map was created using the wind speed and wind direction as recorded in the station of meteorological organization (GEO) nearest to the sampling location. The WRPLOT
Environ Monit Assess (2014) 186:7505–7512
view package was used to plot wind rose (WRPLOT view, 1998–2011).
Results and discussion Figure 1 shows the seasonal variation of the total particulate matter (TPM) of each filter in microgram per cubic meter during the sampling days. It can be seen that the total mass concentration gradually increases from spring to winter. The average mass concentration had the lowest value of about 100 μg/m3during the spring of 2009. In the summer and autumn, the concentration increased to around 200 and 250 μg/m3, respectively. The amount of TPM has its highest value of about 300 μg/m3 in winter. Our sampling period started from the spring of 2009 and continued until the spring of 2010. Our measurements show that amount of average pollution in spring 2010 is about 250 μg/m3 which is
Fig. 3 Average concentrations of CO, NO2, SO2, PM10, total particulate matter (TPM), and Pb pollutants over total period of sampling data recorded at the GEO station and our measurements
Environ Monit Assess (2014) 186:7505–7512
7509
twice to that of the spring of 2009. Also, the observed values for average pollution in the spring of 2010 are quite similar to the winter of 2009. By using PIXE, eighteen elements including Al, Si, S, Cl, K, Ca, Ti, V, Cr, Mn, Fe, Ni, Cu, Zn, Br, Rb, Sr, and Pb were identified and their absolute concentrations measured. The average concentrations and standard deviations for the detected elements in different seasons are presented in Table 1 in nanogram per cubic meter. The quantitative results show that elements associated with traffic and vehicle emissions like Cu, Zn, and Pb almost increased threefolds in the winter. The main reason was temperature inversion during most winter days in Tehran. The concentrations of elements associated with soil dust remain at the same levels. Table 1 also shows an increase in elements associated with soil dust in the spring of 2010 compared to the spring of 2009, in agreement with the variation in total particulate matter. Wind rose maps for each season, and for the total period of sampling shows in Online Resource 2. During spring, the prevailing direction of the wind Table 2 Percentage of initial and rotated variance and cumulative variance for all and four main components Components
1 2
Initial eigenvalues
Rotated eigenvalues
Variance (%)
Cumulative (%)
Variance (%)
Cumulative (%)
51.39
51.39
37.32
37.31
13.35
64.75
22.68
is from the west. The wind rose maps for the spring of 2010 and the spring of 2009 show that the wind speed (in % unit) is higher in the spring of 2010 (Online Resource 2). In the spring of 2010, huge amounts of atmospheric dusts arrived in Tehran as well as many other cities in Iran. Thus, one can infer that the high concentrations of Al, Si, K, Ca, Ti, Mn, and Fe are due to the natural dust brought into Tehran with the winds from the west (possibly Iraq). Meteorological data also show an increase in atmospheric dust on certain days in the autumn and winter of 2009. These days correspond to the two highest peaks in Fig. 1 in autumn and winter. Zinc and lead concentrations are higher in autumn and winter and go through similar variations (Fig. 2). Figure 3 shows average concentrations of CO, NO2, SO2, and PM10 pollutants over the total period of sampling. The averages were obtained for our sampling period of about 8 h from the hourly data recorded at the GEO station of Iran’s Meteorological Organization. Also, Fig. 3 shows the average concentrations for the total particulate matter (TPM) and
Table 3 VARIMAX rotated matrix in particulate aerosol composition Components
Component factor 1
2
3
4
Log Al
0.836
0.232
0.295
0.037
60.00
Log Si
0.884
0.285
0.211
0.150
0.169
0.849
0.194
0.083
3
5.59
70.40
9.24
69.25
Log S
4 5
4.37 4.22
74.71 78.94
5.46
74.71
Log Cl
0.323
0.755
0.336
−0.024
Log K
0.845
0.365
0.246
0.085
6
3.91
82.85
Log Ca
0.857
0.330
0.274
0.107
7
3.55
86.40
Log Ti
0.897
0.053
0.252
0.131
8
3.20
89.59
Log V
0.295
0.354
0.043
0.600
9
2.63
92.22
Log Cr
0.627
0.246
−0.300
−0.222
10
1.83
94.05
Log Mn
0.813
0.439
0.164
0.082
11
1.40
95.45
Log Fe
0.926
0.197
0.214
0.0963
12
0.95
97.67
Log Ni
0.480
0.118
0.072
0.536
13
0.88
98.55
Log Cu
0.120
0.671
0.077
0.123
14
0.55
99.07
Log Zn
0.163
0.880
0.036
0.122
15
0.38
99.45
Log Br
0.216
0.350
0.667
−0.017
16
0.20
99.65
Log Rb
0.350
0.304
0.312
−0.450
17
0.20
99.84
Log Sr
0.337
0.126
0.722
0.018
18
0.16
100.00
Log Pb
0.146
0.866
0.099
−0.070
7510
Environ Monit Assess (2014) 186:7505–7512
& & &
& Fig. 4 Simple scatter plot of Ca and Si according to base-10 log values
Pb as obtained from our measurements. It may be worth to note the following features: &
The sharp peak for SO2 at 40 ppb in Jan 2010 coincided with the inversion condition in winter in Tehran.
&
The variation of NO2 (in the range of 37–62 ppb) follows the same pattern as that of SO2 except for June 2009. CO shows a smooth variation between 2 and 8 ppm except for a sharp peak in November 2009. The lowest Pb concentration occurs in April 2009 and gradually increases, reaching its maximum in January 2010 (similar to SO2 and NO2) when the inversion condition prevails in Tehran. The concentration of TPM reaches a maximum in the spring of 2010 when a huge amount of dust was carried with the wind into Tehran. The largest fluctuation is seen in the concentration of PM10 which drops from a maximum value of about 110 μg/m3 in September 2009 to a minimum of about 35 μg/m3 due to the rains in November of 2009.
To identify and classify different sources of pollution in Tehran, principal component analysis (PCA) was
LOG Al
LOG K
LOG Mn
LOG Fe
LOG Zn
LOG Pb Fig. 5 Scatter plot of correlation matrix of some elements of soil dust and vehicle sources
Environ Monit Assess (2014) 186:7505–7512
performed on all the elements appearing in our data (Härdle and Simar 2007). In Table 2, the initial eigenvalues, percentage of variances, and cumulative variances for all components are listed alongside the rotated eigenvalues for the four main principal components which account for 75 % of the variance observed in the eighteen original components. The first and second components are the most significant factors with variances of 37.31 and 22.68 %, respectively. Table 3 shows the VARIMAX rotated component matrix of all elements. The first factor has high factor loadings for Al, Si, K, Ca, Ti, Mn, and Fe; therefore, it represents resuspended soil dust in agreement with the results obtained by Artaxo et al. (1999). The second factor associated with S, Cl, Cu, Zn, and Pb represents automobile and vehicular sources. In general, Pb is derived from gasoline; Zn and Cu are due to the wear and tear of vulcanized rubber tires, lubricating oil, etc.; Cl probably comes from residual oil combustion and winter salting of the streets (Angyal et al. 2010). Thus, the second factor can be classified as a common urban anthropogenic source of pollution (Kar et al. 2010). The third factor, associated with Br and Sr, represents residual fuel oil combustion, but the presence of Sr makes it difficult to determine the exact sources. The fourth factor contains V and Ni and represents fuel oil (Miranda 1996). In general, our results show that the main components of pollution in Tehran are aerosols from traffic, vehicular emissions, and dust re-suspension. The simple scatter plot of the base-10 logarithm of the concentrations of Ca and Si, which are the elements present in soil dust (Fig. 4), shows a strong correlation between these two elements. Figure 5 shows a matrix of similar scatter plots for all possible pairs of elements associated with soil dust and automobile and vehicular emissions, in particular Al, K, Mn, and Fe associated with soil dust and Zn and Pb associated with vehicular sources. This matrix clearly shows a positive linear correlations within elements in soil dust group and within elements associated with vehicular sources (marked with brown squares in Fig. 5), as well as anticorrelation between elements in the soil dust group and elements associated with vehicular sources. The distribution of chemical elements in some samples was mapped using micro-PIXE. Online Resource 3 shows a typical map of distribution of several elements within a 300×300 μm2 scanned area. In this particular sample, various elements were not distributed
7511
homogeneously over the filter area, indicating they are associated with particulate matter (in this case, soil dust). Moreover, it can be seen in Online Resource 3, Ca and Si are the dominant elements in this sample, and correlations can be seen between the distributions between various pairs of elements such as Ca–S, Ca–K, Si–K, Si–Ca, and Cl–Si in agreement with a previous study (Ghermandi et al. 2005).
Conclusion Based on our quantitative analysis, the dominant elements of air pollution in Tehran are Ca, Fe, and Si whose origin is soil dust. PCA analysis shows that the main sources of pollution to be soil dust and traffic. The results obtained from the PIXE analysis shows that Tehran is highly polluted in winter. As expected, due to vehicular emission and slow speed of wind in winter and temperature inversion during the cold season, the pollution levels are almost three times the levels in spring. The concentration of elements in the aerosols originating from the soil dust remains almost constant through all the seasons. By using micro-PIXE and statistical analysis, different correlations were identified. In general, this investigation provides baseline information which can be used by air quality control and environmental organizations.
References Agha-Aligol, D., Khosravi, F., Lamehi-Rachti, M., Baghizadeh, A., Oliaiy, P., & Shokouhi, F. (2007). Analysis of 18th–19th century’s historical samples of Iranian ink and paper belonging to the Qajar dynasty. Applied Physics A: Materials, 89(3), 799–805. Al-Momani, I. F., Daradkeh, A. S., Haj-Hussein, A. T., Yousef, Y. A., Jaradat, Q. M., & Momani, K. A. (2005). Trace elements in daily collected aerosols in Al-Hashimya, central Jordan. Atmospheric Environment, 73(1–2), 87–100. Angyal, A., Kertész, Z., Szikszai, Z., & Szoboszlai, Z. (2010). Study of Cl-containing urban aerosol particles by ion beam analytical methods. Nuclear Instruments and Methods, B268, 2211–2215. Artaxo, P., Oyola, P., & Martinez, R. (1999). Aerosol composition and sources apportionment in Santiago de Chile. Nuclear Instruments and Methods, B150, 409–416. Atash, F. (2007). The deterioration of urban environments in developing countries: mitigating the air pollution crisis in Tehran, Iran. Cities, 24(6), 399–409.
7512 Basha, S., Jhala, J., Thorat, R., Goel, S., Trivedi, R., Shah, K., et al. (2010). Assessment of heavy metal content in suspended particulate matter of coastal industrial town, Mithapur, Gujarat, India. Atmospheric Research, 97(1–2), 257–265. Chiari, M., Lucarelli, F., Migliori, A., Nava, S., Valli, G., Vecchi, R., et al. (2006). PIXE analysis of p M10–2.5 and PM2.5 with hourly resolution from Michelozzo’s Courtyard in Palazzo Vecchio (Florence, Italy). Nuclear Instruments and Methods, B249, 552–555. Dallarosa, J., Calesso Teixeira, E., Meira, L., & Wiegand, F. (2008). Study of the chemical elements and polycyclic aromatic hydrocarbons in atmospheric particles of PM10 and PM2.5 in the urban and rural areas of South Brazil. Atmospheric Research, 89(1–2), 76–92. Daryabeigi Zand, A., Nabi bidhendi, G., Mikaeili, T. A., & Pezeshk, H. (2007). The influence of deposit control additives on exhaust CO and HC emissions from gasoline engines (case study: Tehran)’. Transportation Research Part D: Transport and Environment, 12(3), 189–194. Ghermandi, G., Cecchi, R., Lusvarghi, L., Laj, P., Zappoli, S., & Ceccato, D. (2005). Internal/external mixing of aerosol particles elemental composition retrieved from micro-PIXE and PIXE. Nuclear Instruments and Methods, B240, 313–320. Halek, F., Kavouci, A., & Montehaie, H. (2004). Role of motorvehicles and trend of air borne particulate in the Great Tehran area, Iran. International Journal of Environmental Health Research, 14(4), 307–313. Härdle, W. K., & Simar, L. (2007). Applied multivariate statistical analysis. London: Springer. Limited. Kampa, M., & Castanas, E. (2008). Human health effects of air pollution. Environmental Pollution, 151, 362–367. Kar, S., Maity, J. P., Chandra Samal, A., & Chandra Santra, S. (2010). Metallic components of traffic-induced urban aerosol, their spatial variation, and source apportionment. Environmental Monitoring and Assessment, 168, 561–574. Karatzas, K., Nejadkoorki, F., Riga, M. (2012). Preliminary analysis of Tehran air quality data : Findings and future research protection and restoration of the environment XI, 986-992. Leili, M., Naddafi, K., Nabizadeh, R., Yunesian, M., & Mesdaghinia, A. (2008). The study of TSP and PM10 concentration and their heavy metal content in central area of Tehran, Iran. Air Quality, Atmosphere and Health, 1, 159–166. Li, Z., Zhao, S., Edwards, R., Wang, W., & Zhou, P. (2010). Characteristics of individual aerosol particles over Urumqi Glacier No. 1 in eastern Tianshan, central Asia, China. Atmospheric Research, 99(1), 57–66. Lim, M. C. H., Ayoko, G. A., & Morawska, L. (2005). Characterization of elemental and polycyclic aromatic hydrocarbon compositions of urban air in Brisbane. Atmospheric Environment, 39, 463–476.
Environ Monit Assess (2014) 186:7505–7512 Mage, D., Ozolins, G., Peterson, P., Webster, A., Orthoferi, R., Vandeweerds, V., & Gwynnet, M. (1996). Urban air pollution in megacities of the world. Atmospheric Environment, 30(5), 681–686. Malakooti, H. (2010). Meteorology and air-quality in a mega-city: application to Tehran, Iran. Miranda, J. (1996). Studies of atmospheric aerosol in large urban area using PIXE an overview. Nuclear Instruments and Methods, B109(110), 439–444. Moreno, T., Alastuey, A., Querol, X., Font, O., & Gibbons, W. (2007). The identification of metallic elements in airborne particulate matter derived from fossil fuels at Puertollano, Spain. International Journal of Coal Geology, 71(2–3), 122–128. Nava, S., Becagli, S., Calzolai, G., Chiari, M., Lucarelli, F., Prati, P., et al. (2012). Saharan dust impact in central Italy: an overview on three years elemental data records. Atmospheric Environment, 60, 444–452. Ramanakumar, A. V., Parent, M.-E., & Siemiatycki, J. (2007). Risk of lung cancer from residential heating and cooking fuels in Montreal, Canada. American Journal of Epidemiology, 165, 634–642. Sekhavatjou, M. S., Hosseini Alhashemi, A., & Rostam, A. (2011). Comparison of trace element concentrations in ambient air of industrial and residential areas in Tehran city. Biological Trace Element Research, 143, 1413–1423. Shah, M. H., & Shaheen, N. (2010). Seasonal behaviours in elemental composition of atmospheric aerosols collected in Islamabad, Pakistan. Atmospheric Research, 95(2–3), 210– 223. Sinha, S.N., Nag, P.K. (2011). Air pollution from solid fuels, Encyclopedia of Environmental Health, Elsevier Science, pp. 46-52. Tajrishy, M., & Abrishamchi, A. (2005). Integrated approach to water and wastewater management for Tehran, Iran. Water conservation, reuse, and recycling: Proceedings of an Iranian-American Workshop (pp. 217–230). Washington, D.C: National Academies Press. Wang, H., & Mullahy, J. (2006). Willingness to pay for reducing fatal risk by improving air quality: a contingent valuation study in Chongqing, China. Science of Total Environment, 367, 50–57. Wildhaber, J. H. (2006). Aerosols: the environmental harmful effect. Paediatric Respiratory Reviews, 7(Suppl 1), S86–S87. WRPLOT view, Wind rose plots for metrological data, 1998-2011 Lakes Environmental software, version 7.0.0. Yatin, M., Tuncel, S., Aras, N. K., Olmez, I., Aygun, S., et al. (2000). Atmospheric trace elements in Ankara, Turkey: 1. Factors affecting chemical composition of fine particles. Atmospheric Environment, 34(8), 1305–1318.
Elemental analysis of aerosols in Tehran’s atmosphere using PIXE and identification of pollution sources N. Esmaili & S. Khashman & M. Lamehi-Rachti & D. Agha Aligol & F. Shokouhi & P. Oliaiy & M. Farmahini Farahani
Received: 9 February 2014 / Accepted: 8 July 2014 / Published online: 17 July 2014 # Springer International Publishing Switzerland 2014
Abstract In this study, the proton-induced X-ray emission (PIXE) technique has been applied to measure the elemental composition and concentrations of particulate matter of 220 samples of aerosols in Tehran’s atmosphere within a 450-day time interval starting from March 2009 and ending in June 2010, covering all four seasons. PIXE analysis shows the samples are comprised of various elements including Al, Si, S, Cl, K,
Electronic supplementary material The online version of this article (doi:10.1007/s10661-014-3943-y) contains supplementary material, which is available to authorized users. N. Esmaili : M. Lamehi-Rachti (*) : D. Agha Aligol : F. Shokouhi : P. Oliaiy : M. Farmahini Farahani Nuclear Science Research School, Nuclear Science & Technology Research Institute (NSTRI), Tehran, Iran e-mail: [email protected] N. Esmaili e-mail: [email protected] D. Agha Aligol e-mail: [email protected] F. Shokouhi e-mail: [email protected] P. Oliaiy e-mail: [email protected] M. Farmahini Farahani e-mail: [email protected] S. Khashman Department of Physics, Faculty of Sciences, Central Tehran Branch, Islamic Azad Universit, Tehran, Iran e-mail: [email protected]
Ca, Ti, V, Cr, Mn, Fe, Ni, Cu, Zn, Br, Rb, Sr, and Pb. Also, to obtain more information about the sources of pollution and to identify the major sources of urban particulate matter, principal component analysis (PCA) was used. Furthermore, micro-PIXE was performed to study individual aerosols in some samples. Results revealed that the concentration of elements originating from vehicle emissions increases three times in winter; whereas the concentration of elements with soil origin remains constant. Based on wind rose maps, it is inferred that the high concentrations of the elements Al, Si, K, Ca, Ti, Mn, and Fe are associated with natural dust brought by winds into Tehran from the west. Keywords Aerosol . Air pollution . PIXE . Elemental composition . Principal component analysis
Introduction The city of Tehran, with a population of over seven million, covers an area of approximately 730 km2, containing about three million vehicles, two million motorcycles (Tajrishy and Abrishamchi 2005), and many factories. Air pollution is a major environmental problem in Tehran. Studies show that the main sources of air pollution in megacities are traffic, energy production using fossil fuels and biomass, industrial sources, and re-suspension of soil (Moreno et al. 2007; Sinha and Nag 2011). The Government of Iran and Tehran Municipality have actively participated in an effort for the reduction of local and global air pollution. Atash
7506
(2007) reviewed and evaluated the implementation of the 10-year master plan to control air pollution in Tehran. Air pollutants are emitted into the atmosphere from stationary sources (utility, industrial, institutional, and commercial facilities) and mobile sources (on-road vehicular traffic and transportation). Motor vehicles constitute the most dominant source of pollutants in Tehran, and it is estimated that emissions from mobile sources account for almost 85 % of all the air pollution in Tehran. The types and amounts of pollutants are directly related to the type and quality of fuel used for transportation or in industry and the level of technology involved (Halek et al. 2004; Malakooti 2010). The pollutants directly affect the health of people (Lim et al. 2005; Wang and Mullahy 2006), and their identification and measurement is especially important for any future action to reduce the harmful effects (Wildhaber 2006; Ramanakumar et al. 2007; Kampa and Castanas 2008). In order to assess the problems of urban air pollution in a global context, the World Health Organization (WHO) and the United Nations Environment Program (UNEP) have initiated a detailed study of air quality in many megacities of the world. This study is focused on the monitoring and measurement of sulfur dioxide (SO2), suspended particulate matter (SPM), and lead (Pb). However, Tehran was not included in this study because of lack of data (Mage et al. 1996). During the last decade, the elemental composition of atmospheric aerosol has been studied by different techniques (Yatin et al. 2000; Al-Momani et al. 2005; Basha Fig. 1 Seasonal variation of particulate mass (PM) in microgram per cubic meter during sampling days
Environ Monit Assess (2014) 186:7505–7512
et al. 2010; Li et al. 2010; Shah and Shaheen 2010). Ion beam analysis is known to be suitable for elemental analysis of aerosol and air pollutants (Ghermandi et al. 2005; Chiari et al. 2006; Dallarosa et al. 2008; Angyal et al. 2010; Nava et al. 2012). Recently, a few studies have been done about criteria pollutants in Tehran such as CO, NOX, SOX, and PM10 (Daryabeigi et al. 2007; Malakooti 2010; Karatzas et al. 2012), and few studies have been carried out about trace elements. Sekhavatjou et al. (2011) measured and compared concentration of trace elements in industrial and residential areas by neutron activation analysis. Also, Leili et al. (2008) have done an important study on the concentrations of TPM, PM10, and heavy metals in one station in the center of Tehran by atomic absorption spectrometry. Investigation of aerosols in the air of Tehran was defined as a research project in the annual program of this laboratory. In the present work, sampling of aerosols was performed on 220 working days from March 2009 to June 2010. Map of 22 districts of the Municipality of Tehran, the geographic location of the sampling station and the surface topography (Malakooti 2010) of Tehran and some surrounding areas (55 km×45 km, resolution of 1 km) are shown in Online Resource 1. The location of sampling was near a major highway in western Tehran. PIXE and micro-PIXE were used to determine the concentration of various elements in the samples. Principal component analysis was applied to obtain the sources of the aerosols and their contribution to air pollution in Tehran.
Environ Monit Assess (2014) 186:7505–7512
7507
Table 1 Averaged concentrations and standard deviation in nanogram per cubic meter of the detected elements in different seasons Elements
Spring (2009)
Summer (2009)
Autumn (2009)
Winter (2009)
Spring (2010)
Al
500±190
780±390
590±360
660±280
820±320
Si
2,500±1,000
3,480±1,580
2,420±920
3,410±1,580
3,460±1,350
S
830±230
1,030±430
1,160±650
1,420±820
750±230
Cl
430±140
780±540
610±320
810±540
490±270
K
980±440
1,530±600
970±390
1,210±530
1,240±480
Ca
7,700±3,550
11,370±4,810
7,800±2,990
9,980±4,370
9,950±3,940
Ti
240±100
310±150
190±80
260±110
340±160
V
10±10
20±10
10±10
30±30
20±10
Cr
10±10
20±10
10±10
10±10
10±10
Mn
90±30
120±50
90±40
110±50
110±50
Fe
3,250±1,420
4,460±1,980
2,930±1,110
3,990±1,710
4,760±2,440
Ni
20±10
20±10
10±10
20±30
20±20
Cu
40±10
50±30
60±30
80±50
40±20
Zn
180±70
270±150
380±260
530±390
270±190
Br
60±60
50±30
50±30
50±30
40±30
Rb
40±30
60±50
50±30
60±60
40±40
Sr
100±50
130±90
100±50
120±80
100±50
Pb
190±200
310±190
370±320
500±520
200±150
Materials and methods Atmospheric aerosol particles were collected on cellulose-fiber filters. Each filter has a diameter of 47 mm. The sampling was performed at a height of 1.5 m above ground. The filter holder was connected to a vacuum pump [GAST 0522-V3], which was operated at a pumping speed of approximately 3 m3/h. The sampling time was about 8 h from 8 A.M. to 4 P.M. (on working days). The location of sampling was near a major highway in western Tehran. Total particulate matter (TPM) was obtained by weighing each filter before and after sampling with a digital balance (model: Precisa 125A with 0.0001 g accuracy). The samples were bombarded with a 2-MeV proton beam and all measurements were carried out in vacuum of around 10−5 Torr. The beam current was about 6 nA. The X-rays emitted from the samples were detected by a Canberra Si (Li) detector at an angle of 135°. The energy resolution of detectors was 165 eV for the Fe (Kα) Xray. The quantitative analysis was performed by WinQXAS software (Edition IAEA. 1997–2001) and the detector efficiency, cross-sections, and numbers of incident protons were taken into account by RBS technique. To confirm the accuracy of the measurement, two
Fig. 2 Concentration of Zn and Pb in nanogram per cubic meter during sampling days
7508
reference materials (IAEA-SOIL7 and IAEA-SL1, lake sediment) were analyzed under the same conditions, and the results were in good agreement with the recommended values within an accuracy of 5 %. In order to gain further insight into the detailed characteristics of atmospheric aerosols, some samples were analyzed by micro-PIXE method (Agha-Aligol et al. 2007). To identify aerosol source, elemental concentrations were processed using principal component analysis (PCA). The statistical analysis was performed by the Statistics Package for Social Science (SPSS 12.0 Inc, USA). All elemental concentrations extracted from PIXE analysis were transformed to base-10 logarithms and used for PCA. These transformations compensate for the differences in magnitude between major elements and trace elements. Also, the wind rose map was created using the wind speed and wind direction as recorded in the station of meteorological organization (GEO) nearest to the sampling location. The WRPLOT
Environ Monit Assess (2014) 186:7505–7512
view package was used to plot wind rose (WRPLOT view, 1998–2011).
Results and discussion Figure 1 shows the seasonal variation of the total particulate matter (TPM) of each filter in microgram per cubic meter during the sampling days. It can be seen that the total mass concentration gradually increases from spring to winter. The average mass concentration had the lowest value of about 100 μg/m3during the spring of 2009. In the summer and autumn, the concentration increased to around 200 and 250 μg/m3, respectively. The amount of TPM has its highest value of about 300 μg/m3 in winter. Our sampling period started from the spring of 2009 and continued until the spring of 2010. Our measurements show that amount of average pollution in spring 2010 is about 250 μg/m3 which is
Fig. 3 Average concentrations of CO, NO2, SO2, PM10, total particulate matter (TPM), and Pb pollutants over total period of sampling data recorded at the GEO station and our measurements
Environ Monit Assess (2014) 186:7505–7512
7509
twice to that of the spring of 2009. Also, the observed values for average pollution in the spring of 2010 are quite similar to the winter of 2009. By using PIXE, eighteen elements including Al, Si, S, Cl, K, Ca, Ti, V, Cr, Mn, Fe, Ni, Cu, Zn, Br, Rb, Sr, and Pb were identified and their absolute concentrations measured. The average concentrations and standard deviations for the detected elements in different seasons are presented in Table 1 in nanogram per cubic meter. The quantitative results show that elements associated with traffic and vehicle emissions like Cu, Zn, and Pb almost increased threefolds in the winter. The main reason was temperature inversion during most winter days in Tehran. The concentrations of elements associated with soil dust remain at the same levels. Table 1 also shows an increase in elements associated with soil dust in the spring of 2010 compared to the spring of 2009, in agreement with the variation in total particulate matter. Wind rose maps for each season, and for the total period of sampling shows in Online Resource 2. During spring, the prevailing direction of the wind Table 2 Percentage of initial and rotated variance and cumulative variance for all and four main components Components
1 2
Initial eigenvalues
Rotated eigenvalues
Variance (%)
Cumulative (%)
Variance (%)
Cumulative (%)
51.39
51.39
37.32
37.31
13.35
64.75
22.68
is from the west. The wind rose maps for the spring of 2010 and the spring of 2009 show that the wind speed (in % unit) is higher in the spring of 2010 (Online Resource 2). In the spring of 2010, huge amounts of atmospheric dusts arrived in Tehran as well as many other cities in Iran. Thus, one can infer that the high concentrations of Al, Si, K, Ca, Ti, Mn, and Fe are due to the natural dust brought into Tehran with the winds from the west (possibly Iraq). Meteorological data also show an increase in atmospheric dust on certain days in the autumn and winter of 2009. These days correspond to the two highest peaks in Fig. 1 in autumn and winter. Zinc and lead concentrations are higher in autumn and winter and go through similar variations (Fig. 2). Figure 3 shows average concentrations of CO, NO2, SO2, and PM10 pollutants over the total period of sampling. The averages were obtained for our sampling period of about 8 h from the hourly data recorded at the GEO station of Iran’s Meteorological Organization. Also, Fig. 3 shows the average concentrations for the total particulate matter (TPM) and
Table 3 VARIMAX rotated matrix in particulate aerosol composition Components
Component factor 1
2
3
4
Log Al
0.836
0.232
0.295
0.037
60.00
Log Si
0.884
0.285
0.211
0.150
0.169
0.849
0.194
0.083
3
5.59
70.40
9.24
69.25
Log S
4 5
4.37 4.22
74.71 78.94
5.46
74.71
Log Cl
0.323
0.755
0.336
−0.024
Log K
0.845
0.365
0.246
0.085
6
3.91
82.85
Log Ca
0.857
0.330
0.274
0.107
7
3.55
86.40
Log Ti
0.897
0.053
0.252
0.131
8
3.20
89.59
Log V
0.295
0.354
0.043
0.600
9
2.63
92.22
Log Cr
0.627
0.246
−0.300
−0.222
10
1.83
94.05
Log Mn
0.813
0.439
0.164
0.082
11
1.40
95.45
Log Fe
0.926
0.197
0.214
0.0963
12
0.95
97.67
Log Ni
0.480
0.118
0.072
0.536
13
0.88
98.55
Log Cu
0.120
0.671
0.077
0.123
14
0.55
99.07
Log Zn
0.163
0.880
0.036
0.122
15
0.38
99.45
Log Br
0.216
0.350
0.667
−0.017
16
0.20
99.65
Log Rb
0.350
0.304
0.312
−0.450
17
0.20
99.84
Log Sr
0.337
0.126
0.722
0.018
18
0.16
100.00
Log Pb
0.146
0.866
0.099
−0.070
7510
Environ Monit Assess (2014) 186:7505–7512
& & &
& Fig. 4 Simple scatter plot of Ca and Si according to base-10 log values
Pb as obtained from our measurements. It may be worth to note the following features: &
The sharp peak for SO2 at 40 ppb in Jan 2010 coincided with the inversion condition in winter in Tehran.
&
The variation of NO2 (in the range of 37–62 ppb) follows the same pattern as that of SO2 except for June 2009. CO shows a smooth variation between 2 and 8 ppm except for a sharp peak in November 2009. The lowest Pb concentration occurs in April 2009 and gradually increases, reaching its maximum in January 2010 (similar to SO2 and NO2) when the inversion condition prevails in Tehran. The concentration of TPM reaches a maximum in the spring of 2010 when a huge amount of dust was carried with the wind into Tehran. The largest fluctuation is seen in the concentration of PM10 which drops from a maximum value of about 110 μg/m3 in September 2009 to a minimum of about 35 μg/m3 due to the rains in November of 2009.
To identify and classify different sources of pollution in Tehran, principal component analysis (PCA) was
LOG Al
LOG K
LOG Mn
LOG Fe
LOG Zn
LOG Pb Fig. 5 Scatter plot of correlation matrix of some elements of soil dust and vehicle sources
Environ Monit Assess (2014) 186:7505–7512
performed on all the elements appearing in our data (Härdle and Simar 2007). In Table 2, the initial eigenvalues, percentage of variances, and cumulative variances for all components are listed alongside the rotated eigenvalues for the four main principal components which account for 75 % of the variance observed in the eighteen original components. The first and second components are the most significant factors with variances of 37.31 and 22.68 %, respectively. Table 3 shows the VARIMAX rotated component matrix of all elements. The first factor has high factor loadings for Al, Si, K, Ca, Ti, Mn, and Fe; therefore, it represents resuspended soil dust in agreement with the results obtained by Artaxo et al. (1999). The second factor associated with S, Cl, Cu, Zn, and Pb represents automobile and vehicular sources. In general, Pb is derived from gasoline; Zn and Cu are due to the wear and tear of vulcanized rubber tires, lubricating oil, etc.; Cl probably comes from residual oil combustion and winter salting of the streets (Angyal et al. 2010). Thus, the second factor can be classified as a common urban anthropogenic source of pollution (Kar et al. 2010). The third factor, associated with Br and Sr, represents residual fuel oil combustion, but the presence of Sr makes it difficult to determine the exact sources. The fourth factor contains V and Ni and represents fuel oil (Miranda 1996). In general, our results show that the main components of pollution in Tehran are aerosols from traffic, vehicular emissions, and dust re-suspension. The simple scatter plot of the base-10 logarithm of the concentrations of Ca and Si, which are the elements present in soil dust (Fig. 4), shows a strong correlation between these two elements. Figure 5 shows a matrix of similar scatter plots for all possible pairs of elements associated with soil dust and automobile and vehicular emissions, in particular Al, K, Mn, and Fe associated with soil dust and Zn and Pb associated with vehicular sources. This matrix clearly shows a positive linear correlations within elements in soil dust group and within elements associated with vehicular sources (marked with brown squares in Fig. 5), as well as anticorrelation between elements in the soil dust group and elements associated with vehicular sources. The distribution of chemical elements in some samples was mapped using micro-PIXE. Online Resource 3 shows a typical map of distribution of several elements within a 300×300 μm2 scanned area. In this particular sample, various elements were not distributed
7511
homogeneously over the filter area, indicating they are associated with particulate matter (in this case, soil dust). Moreover, it can be seen in Online Resource 3, Ca and Si are the dominant elements in this sample, and correlations can be seen between the distributions between various pairs of elements such as Ca–S, Ca–K, Si–K, Si–Ca, and Cl–Si in agreement with a previous study (Ghermandi et al. 2005).
Conclusion Based on our quantitative analysis, the dominant elements of air pollution in Tehran are Ca, Fe, and Si whose origin is soil dust. PCA analysis shows that the main sources of pollution to be soil dust and traffic. The results obtained from the PIXE analysis shows that Tehran is highly polluted in winter. As expected, due to vehicular emission and slow speed of wind in winter and temperature inversion during the cold season, the pollution levels are almost three times the levels in spring. The concentration of elements in the aerosols originating from the soil dust remains almost constant through all the seasons. By using micro-PIXE and statistical analysis, different correlations were identified. In general, this investigation provides baseline information which can be used by air quality control and environmental organizations.
References Agha-Aligol, D., Khosravi, F., Lamehi-Rachti, M., Baghizadeh, A., Oliaiy, P., & Shokouhi, F. (2007). Analysis of 18th–19th century’s historical samples of Iranian ink and paper belonging to the Qajar dynasty. Applied Physics A: Materials, 89(3), 799–805. Al-Momani, I. F., Daradkeh, A. S., Haj-Hussein, A. T., Yousef, Y. A., Jaradat, Q. M., & Momani, K. A. (2005). Trace elements in daily collected aerosols in Al-Hashimya, central Jordan. Atmospheric Environment, 73(1–2), 87–100. Angyal, A., Kertész, Z., Szikszai, Z., & Szoboszlai, Z. (2010). Study of Cl-containing urban aerosol particles by ion beam analytical methods. Nuclear Instruments and Methods, B268, 2211–2215. Artaxo, P., Oyola, P., & Martinez, R. (1999). Aerosol composition and sources apportionment in Santiago de Chile. Nuclear Instruments and Methods, B150, 409–416. Atash, F. (2007). The deterioration of urban environments in developing countries: mitigating the air pollution crisis in Tehran, Iran. Cities, 24(6), 399–409.
7512 Basha, S., Jhala, J., Thorat, R., Goel, S., Trivedi, R., Shah, K., et al. (2010). Assessment of heavy metal content in suspended particulate matter of coastal industrial town, Mithapur, Gujarat, India. Atmospheric Research, 97(1–2), 257–265. Chiari, M., Lucarelli, F., Migliori, A., Nava, S., Valli, G., Vecchi, R., et al. (2006). PIXE analysis of p M10–2.5 and PM2.5 with hourly resolution from Michelozzo’s Courtyard in Palazzo Vecchio (Florence, Italy). Nuclear Instruments and Methods, B249, 552–555. Dallarosa, J., Calesso Teixeira, E., Meira, L., & Wiegand, F. (2008). Study of the chemical elements and polycyclic aromatic hydrocarbons in atmospheric particles of PM10 and PM2.5 in the urban and rural areas of South Brazil. Atmospheric Research, 89(1–2), 76–92. Daryabeigi Zand, A., Nabi bidhendi, G., Mikaeili, T. A., & Pezeshk, H. (2007). The influence of deposit control additives on exhaust CO and HC emissions from gasoline engines (case study: Tehran)’. Transportation Research Part D: Transport and Environment, 12(3), 189–194. Ghermandi, G., Cecchi, R., Lusvarghi, L., Laj, P., Zappoli, S., & Ceccato, D. (2005). Internal/external mixing of aerosol particles elemental composition retrieved from micro-PIXE and PIXE. Nuclear Instruments and Methods, B240, 313–320. Halek, F., Kavouci, A., & Montehaie, H. (2004). Role of motorvehicles and trend of air borne particulate in the Great Tehran area, Iran. International Journal of Environmental Health Research, 14(4), 307–313. Härdle, W. K., & Simar, L. (2007). Applied multivariate statistical analysis. London: Springer. Limited. Kampa, M., & Castanas, E. (2008). Human health effects of air pollution. Environmental Pollution, 151, 362–367. Kar, S., Maity, J. P., Chandra Samal, A., & Chandra Santra, S. (2010). Metallic components of traffic-induced urban aerosol, their spatial variation, and source apportionment. Environmental Monitoring and Assessment, 168, 561–574. Karatzas, K., Nejadkoorki, F., Riga, M. (2012). Preliminary analysis of Tehran air quality data : Findings and future research protection and restoration of the environment XI, 986-992. Leili, M., Naddafi, K., Nabizadeh, R., Yunesian, M., & Mesdaghinia, A. (2008). The study of TSP and PM10 concentration and their heavy metal content in central area of Tehran, Iran. Air Quality, Atmosphere and Health, 1, 159–166. Li, Z., Zhao, S., Edwards, R., Wang, W., & Zhou, P. (2010). Characteristics of individual aerosol particles over Urumqi Glacier No. 1 in eastern Tianshan, central Asia, China. Atmospheric Research, 99(1), 57–66. Lim, M. C. H., Ayoko, G. A., & Morawska, L. (2005). Characterization of elemental and polycyclic aromatic hydrocarbon compositions of urban air in Brisbane. Atmospheric Environment, 39, 463–476.
Environ Monit Assess (2014) 186:7505–7512 Mage, D., Ozolins, G., Peterson, P., Webster, A., Orthoferi, R., Vandeweerds, V., & Gwynnet, M. (1996). Urban air pollution in megacities of the world. Atmospheric Environment, 30(5), 681–686. Malakooti, H. (2010). Meteorology and air-quality in a mega-city: application to Tehran, Iran. Miranda, J. (1996). Studies of atmospheric aerosol in large urban area using PIXE an overview. Nuclear Instruments and Methods, B109(110), 439–444. Moreno, T., Alastuey, A., Querol, X., Font, O., & Gibbons, W. (2007). The identification of metallic elements in airborne particulate matter derived from fossil fuels at Puertollano, Spain. International Journal of Coal Geology, 71(2–3), 122–128. Nava, S., Becagli, S., Calzolai, G., Chiari, M., Lucarelli, F., Prati, P., et al. (2012). Saharan dust impact in central Italy: an overview on three years elemental data records. Atmospheric Environment, 60, 444–452. Ramanakumar, A. V., Parent, M.-E., & Siemiatycki, J. (2007). Risk of lung cancer from residential heating and cooking fuels in Montreal, Canada. American Journal of Epidemiology, 165, 634–642. Sekhavatjou, M. S., Hosseini Alhashemi, A., & Rostam, A. (2011). Comparison of trace element concentrations in ambient air of industrial and residential areas in Tehran city. Biological Trace Element Research, 143, 1413–1423. Shah, M. H., & Shaheen, N. (2010). Seasonal behaviours in elemental composition of atmospheric aerosols collected in Islamabad, Pakistan. Atmospheric Research, 95(2–3), 210– 223. Sinha, S.N., Nag, P.K. (2011). Air pollution from solid fuels, Encyclopedia of Environmental Health, Elsevier Science, pp. 46-52. Tajrishy, M., & Abrishamchi, A. (2005). Integrated approach to water and wastewater management for Tehran, Iran. Water conservation, reuse, and recycling: Proceedings of an Iranian-American Workshop (pp. 217–230). Washington, D.C: National Academies Press. Wang, H., & Mullahy, J. (2006). Willingness to pay for reducing fatal risk by improving air quality: a contingent valuation study in Chongqing, China. Science of Total Environment, 367, 50–57. Wildhaber, J. H. (2006). Aerosols: the environmental harmful effect. Paediatric Respiratory Reviews, 7(Suppl 1), S86–S87. WRPLOT view, Wind rose plots for metrological data, 1998-2011 Lakes Environmental software, version 7.0.0. Yatin, M., Tuncel, S., Aras, N. K., Olmez, I., Aygun, S., et al. (2000). Atmospheric trace elements in Ankara, Turkey: 1. Factors affecting chemical composition of fine particles. Atmospheric Environment, 34(8), 1305–1318.
