STATISTICAL ANALYSIS OF INHALABLE
PARTICLES (PM10) AND FINE PARTICLES (PM2.5)
IN URBAN REGION OF KOLKATA, INDIA
Control # 1340
Scheduled in session : AB-1a (1)
Subhankar Naga
A. K. Guptab
U. K. Mukhopadhyayc
Department of Civil Engineering ,Indian Institute of Technology,Kharagpur
Pin-721302, West Bengal,, India.
C West Bengal Pollution Control Board, LA-10A,Sector-III,Salt Lake City,Kolkata,Pin700098, West Bengal India
Corresponding Author. Email address : a - snag@iitian.iitkgp.ernet.in
a & b
ABSTRACT
In this study, the relationship between inhalable particulate (PM10), fine particulate
(PM2.5), coarse particles (PM2.5-10) and meteorological parameters such as temperature,
relative humidity, solar radiation, wind speed were statistically analyzed and modelled
for urban region of Kolkata during winter of 2003-2004. Ambient air quality was
monitored with a sampling frequency of twenty-four hours at three monitoring sites
covering a period of three months from December 2003 to February 2004. The
monitoring sites were located 3-5 m above ground near highly trafficked and congested
areas. The 24-h average PM10 and PM2.5 samples were collected using Thermo-Andersen
high volume samplers. Meteorological parameters such as temperature, relative humidity,
solar radiation, and wind speed were also recorded simultaneously during the sampling
period. The ratios between PM2.5 and PM10 were found to be in the range of 0.6 to 0.9
and the highest ratio was found in the most polluted urban site. Concentrations of PM10
and PM2.5 show the monthly, weekly and daily variations during winter. Statistical
analyses have shown a strong positive correlation between PM10 and PM2.5 and strong
correlation was observed between particulate matter (PM10 and PM2.5) and wind speed.
Regression equations for PM10 and PM2.5 and meteorological parameters were developed.
KEYWORDS: PM10, PM2.5, PM2.5-10, statistical analyses, meteorological parameters,
urban air pollution.
INTRODUCTION
Respiratory health effects are biologically expected to be associated with particles less
than 10 µm, passing the nose and entering lung alveoli (Monn Ch. et. al., 1995, Dockery
et al., 1992,1993, Pope et al., 2002). The impact of ambient particles on human health has
been known since the early 1980s (Lippmann, 1989). Over the last two decades a large
number of epidemiological studies have been conducted around the world that observed
associations between ambient particle concentrations and excesses in daily mortality and
morbidity (Katsouyanni et al., 1997). The health effects range from increased incidences
of pneumonia and asthma, exacerbation of chronic obstructive pulmonary diseases,
increased respiratory symptoms and decreased lung function, to an increased mortality
rate (USEPA, 1995). It has been estimated that for every increase of 10 µg/m 3 of PM10
there will be a rise in mortality of 3.4% from respiratory diseases and 1.4% from
cardiovascular diseases (WBPCB, 2002). The findings on epidemiological studies around
the world have underlined the importance of ambient particles and the need for
monitoring PM10 (inhalable particles) and PM2.5 (fine particles). In addition to negative
health effects, particulate matter reduces visibility and accelerates the deterioration of
buildings. It is assumed that PM10 represents anthropogenic pollutants, whilst the origin
of coarser (>10µm) is mainly natural.
The major objectives of the present study were as follows:
 To collect baseline data of PM10 and PM2.5 from a selected area of study;
 to assess the fraction of PM2.5 within PM10 and its variation over sites and time;
and
 to analyze correlation between air pollutants and meteorological parameters in the
urban region of Kolkata city.
MONITORING AND ANALYSIS
Description of Study Area
Kolkata city, the second-largest metropolis in South Asia is centred around latitude 22°
34’ N and longitude 88° 24’ E. The city is bounded to the west and north-west by the
Hooghly River. The core area of the city is generally flat with elevations ranging from 5
to 6 m above mean sea level. Kolkata Municipal Corporation (KMC) covers an area of
187 km2. However, the wider urban agglomeration surrounding the KMC, which covers
an area of 1,380 km2. Kolkata City has a tropical Savannah climate with a marked
monsoon season. Monthly mean temperatures range from 20 to 31°C, and the maximum
temperatures often exceed 40°C. Summers are hot, the temperatures fluctuating between
maximum 30oC to 35oC and a minimum 14oC to 25oC. Relative humidity is 66% and
69% in winter and summer, respectively. The pre-monsoon and monsoon season is
dominated by strong south- westerly winds with greatest air ventilation potential
(UNEP/WHO, 1992). Moderate northwesterly winds prevail for most of the year with a
high frequency of calms. Kolkata city is located in coastal area and influenced by seabased disturbances. The average wind speed blowing throughout the year is 7 km/hour.
The average solar radiation was recorded 1428 W/m2.
Three stations were established (in the busiest urban agglomeration) to measure PM10 and
PM2.5 concentration. The air quality monitoring stations are shown in Figure 1.
Additional justifications for various sites are given in Table 1.
Collection and Analysis of Particulate Matter
PM10 (inhalable particles) and PM2.5 (fine particles) were collected at three monitoring
stations from December 2003 to February 2004 in the study area. Thermo Andersen flow
manager high volume Sampler was used for collecting PM10 and PM2.5 and operated at an
average flow rate of 68.0 m3/h. The PM10 and PM2.5 samples were collected on
preweighed glass fibre filters (GF/A), as per the most widely recommended (IS 5182 PartIV, 1999) technique. The monitoring at each station has been covered at least once in
a week and the sampling period was twenty-four hours throughout the study period.
Meteorological Data
The meteorological station was located in Victoria Memorial, which is very close to the
air quality monitoring stations. The location of meteorological station is shown in Figure
1. The height of the meteorological tower is 10 m. from ground level. On line continuous
meteorological data with an averaging time interval of 1 hour comprising of temperature,
relative humidity, solar radiation, wind speed and wind direction, which were measured
by Environment S.A automatic Analyzers (Envidas,2002).
Figure 1: Location of monitoring sites in the study area
Site
Index
BE
MP
PS
Table 1. Justification of air quality monitoring stations.
Location
Land use
Road
Gradient
Beleghata & Eastern
Metropolitan
Byepass
crossing
Minto Park
Park Street
Type
Industrial/Residential
Flat
Industrial influence
and medium traffic
Commercial
Commercial/
Industrial
Flat
Flat
High traffic density
High traffic density
with
industrial
influence
RESULTS AND DISCUSSION
Precision and data capture
Total of 53 samples were collected during the study period from December 2003 to
February 2004 from three different sites for PM10. Also, a total of 53 samples of PM2.5
have been collected from three different sites during from December 2003 to February
2004.A blank control procedure for particulate sampling was performed, the result of
which showed high accuracy of sampling in most cases. The initial and final weights of
the blank sampler mounted at one of the sites were almost the same and within the
permissible measurement error. Only in few cases final weight was slightly higher, with
maximum difference of 0.1%.
Frequency distribution of PM10 and PM2.5 concentrations
The frequency distributions of PM10 concentrations in intervals of 50 µg/m3 are shown in
Figure 2. A peak in the distribution of PM10 concentrations occurred at 250 to 300 µg/m3.
Approximately 24% of PM10 concentrations (n=53) were below 200 µg/m3,
81% below 400 µg/m3 and 64% below 300 µg/m3. Approximately 90% of PM10
concentrations were exceeding the standard value of 150 µg/m3, as compared with the
standard given in Table 2.
Figure 2: Frequency distribution of PM10 concentration
Number of frequency
14
12
10
8
6
4
2
0
100
200
300
400
500
PM10 concentration (µg/m3)
Table 2. National Ambient Air Quality Standards (NAAQMS, 1997)
Concentration in ambient air
Time weighted average
PM10
PM2.5
3
Annual average
50 µg/m
15 µg/m3
24 h average
150 µg/m3
65 µg/m3
On the other hand, the peak in the PM2.5 concentrations was at in intervals of 50 to 200
µg/m3 are shown in Figure 3. Approximately 74 % of PM10 concentrations (n=53) were
below 200 µg/m3, 98 % below 400 µg/m3. Approximately 96% of PM2.5 concentrations
were exceeding the standard value of 65 µg/m3, as per the standard given in Table 2.
Figure 3: Frequency distribution of PM2.5 concentration
18
Number of frequency
16
14
12
10
8
6
4
2
0
50
100
200
300
400
500
PM2.5 concentration (µg/m3)
Temporal Variation of PM10 and PM2.5
The daily PM10 average concentration of all the sites are presented in Figure 4. Figure 5
presents the same information for PM2.5. The average concentrations of PM10 and PM2.5
all the sites during the study period were 286.1±123.4 µg/m3 and 179.5±91.5 µg/m3,
respectively. The maximum concentration (637.7 µg/m3 in the month of January, while
minimum (66.7 µg/m3) was recorded in the month of December. The maximum and
minimum concentration of PM2.5 was 537.8 µg/m3 and 49.5 µg/m3 in the month of
January and December, respectively.
Spatial Variation of PM10, PM2.5 and PM2.5-10
Figure 6 shows the average concentrations of PM10, PM2.5 and PM2.5-10 at all the three
monitoring stations during the study period. The ratio of highest to lowest concentration
of all sites varies from 5.0 to 9.0 for PM10, 5.3 to 10.9 for PM2.5 and 6.7 to 8.9 for
PM2.5-10, respectively. The ratio of PM2.5/PM10 was highest at the site PS (0.92). This may
be due to heavy traffic and surrounding industrial activities. The ratio of PM2.5/PM10 was
lowest at the site MP (0.60). The concentration of PM2.5-10 was about 39% of PM10
concentration. The ratio of PM2.5-10/PM10 was highest at MP (0.47) in the month of
February, while the lowest was at BE (0.27) during the month of January.
Figure 4: Daily PM10 average concentration of all the sites
700
BE
600
PM10 (µg/m3)
MP
500
PS
400
300
200
100
0
10
Dec
15
Dec
17
Dec
22
Dec
29
Dec
08
Jan
12
Jan
21
Jan
28
Jan
04
Feb
11
Feb
18
Feb
25
Feb
Sampling date
Figure 5: Daily PM2.5 average concentration of all the sites
400
350
PM2.5 (µg/m3)
300
250
200
BE
MP
PS
150
100
50
0
10
15
17
22
29
Dec Dec Dec Dec Dec
08
Jan
12
Jan
21
Jan
Sampling date
28
Jan
04
Feb
11
Feb
18
Feb
25
Feb
Figure 6: Concentrations of PM10, PM2.5 and PM2.5-10 at all the three monitoring stations
Concentration (µg/m3)
400
December
350
January
300
February
250
200
150
100
50
0
PM10
PM2.5 PM2.5- PM10
10
BE
PM2.5 PM2.5- PM10
10
MP
PM2.5 PM2.510
PS
Sampling stations
Relationships between PM10, PM2.5 and PM2.5-10
Results from recent studies have shown that PM10 in urban locations is mainly composed
of fine particles. The particle size analysis shows that concentration of PM2.5 is about
62% of PM10 concentration at all the three sites. Harrison et al., (1997) found that in
Birmingham, approximately 60% of PM10 was PM2.5. Clarke et al., (1999) reported that
60-70% of urban PM10 mass is typically in the PM2.1 fraction and 50% is below
approximately 1.5m.
The PM2.5/PM10 value had shown little variability, and ranged from 0.61 to 0.64. This
suggests that the contributions of PM2.5-10 (coarse particle) and PM2.5 (fine particles) to
PM10 are similar. Finally the relationships between PM10, PM2.5 and PM2.5-10 were
investigated. Annual mean PM2.5/PM10 ratios measured in a large number of urban and
semi-rural US areas varied between 0.3 and 0.7 (USEPA, 2001). Harrison et al. (1997),
analyzed data collected in Birmingham, UK, and found that PM2.5 comprises about 80%
of PM10 during the winter months in UK and a lower fraction was reported for the
summer months with PM2.5 concentration accounting for approximately 50% of PM10
concentrations. High PM10 concentrations were strongly associated with both PM2.5 and
PM2.5-10 concentrations, with Spearman correlation coefficients of 0.93 and 0.76,
respectively. This is expected, since both fine (primary and secondary particles) and
coarse particles (road dust re-suspension, which is enhanced in dry winter climates) are
associated with local traffic.
Correlation between Particulate Data Sets
Figure 7 shows the scatter plots of PM2.5 concentration against that of PM10
concentration. The Figure indicates that these two parameters are highly related to one
another with a linear relationship. The least-square regression line for the daily data has
the following equation:
[PM2.5]= 0.69[PM10] - 17.878
[R2 = 0.87]
[PM10] = Concentration of PM10 in µg/m3
[PM2.5] = Concentration of PM2.5 in µg/m3
(1)
The regression Equation (1) obtained by using the data of all the monitoring stations viz;
BE, MP and PS, respectively. The above equation reveals that the PM10 concentration
increases with increasing PM2.5 concentration.
The correlation analyses of data have been carried out using Spearman rank correlation
coefficient using the commercial package of SPSS package (SPSS, 2003).
Table 3 shows the correlation between PM10 and PM2.5 data sets during the study period
and the results show that PM10 and PM2.5 was strongly correlated at all the site. The
highest correlation 0.98 was obtained between PM10 and PM2.5 at the site MP site, while
0.96 at BE and 0.66 at the site PS, respectively.
Figure 7: Scatter plots of PM2.5 concentration against that of PM10 concentration
600
PM2.5 = 0.6899PM10 - 17.878
3
PM2.5 (µg/m )
500
R2 = 0.87
400
300
200
100
0
0
100
200
300
400
3
PM10 (µg/m )
500
600
700
Table 3. Spearman rank correlation coefficient for particulate data.
PM10BE PM10MP PM10PS PM2.5BE PM2.5MP PM2.5PS
PM10BE 1.00
PM10MP 0.92** 1.00
PM10PS 0.73** 0.72** 1.00
PM2.5BE 0.96** 0.97** 0.71** 1.00
PM2.5MP 0.90** 0.98** 0.71** 0.96** 1.00
PM2.5PS 0.86** 0.96** 0.66** 0.93** 0.96**
1.00
** Correlation is significant at the 0.01 level
* Correlation is significant at the 0.05 level
PM10 refers to inhalable particle data sets, PM2.5 refers to fine particle data sets
and sites BE, MP, PS, represents Beleghata EM Byepass Crossing, Minto Park
and Park Street, respectively.
Values were rounded to two places of decimal.
Influence of Meteorological Parameters on Particulate matter
The air quality of any place varies from time to time, though the emissions may not
undergo many variations, because air quality strongly depends on the dynamics of the
atmosphere and meteorological conditions play a vital role in governing the fate of air
pollutants. In this study, the relationship between ambient particulate matter data and
meteorological factors, such as temperature, relative humidity, solar radiation and wind
speed is statistically analyzed using the SPSS package (SPSS, 2003).
Table 4 shows the descriptive statistics of particulate matter and meteorological factors.
Table 5 depicts the relationship between particle concentrations (PM10 and PM2.5) and
meteorological parameters (temperature, relative humidity, solar radiation and wind
speed). The PM10 and PM2.5 is found to be strongly correlated with each other and
inversely correlated with wind speed with correlation coefficient of -0.63 and –0.66,
respectively. Very poor correlation was observed between particulate matters (PM10,
PM2.5) and meteorological parameters including temperature, relative humidity and solar
radiation.
Table 4. Descriptive statistics of particulate matter and meteorological factors.
Parameters
Minimum
Maximum
Mean
Std.
N
Deviation
PM10 (µg/m3)
66.66
637.70
286.07
123.37
53
PM2.5 (µg/m3)
49.47
537.77
179.49
91.50
53
o
TEMP. ( C)
12.40
23.60
18.96
2.53
53
RH (%)
64.00
98.00
79.07
6.80
53
2
SR (W/m )
535.00
1428.00
1141.36
213.17
53
WS (m/s)
1.60
3.00
2.24
0.31
53
TEMP, RH, SR, WS refers to temperature, relative humidity, solar radiation and wind
speed, respectively
Values were rounded to two places of decimal
Table 5. Spearman rank correlation coefficient of particulate data and meteorological
factors.
PM10
PM2.5 TEMP
RH
SR
WS
PM10
1.000
PM2.5
0.90**
1.00
TEMP
-0.10
-0.11
1.00
RH
-0.07
-0.01 -0.11
1.00
SR
0.04
0.01
-0.23
-0.63*
1.00
WS
-0.63*
-0.66* -0.13
0.30
0.06
1.00
** Correlation is significant at .01 level
* Correlation is significant at .05 level
PM10 refers to PM10 data, PM2.5 to PM2.5 data; TEMP, RH, SR, WS refers to temperature,
relative humidity, solar radiation and wind speed, respectively.
Values were rounded to two places of decimal.
Figure 8 shows the temporal variation of PM10 and wind speed, while Figure 9 shows the
temporal variation of PM2.5 and wind speed at all the monitoring stations. It was found
that an inverse relationship exists between wind speed and particulate data, and therefore
the predominance of local sources, in this case strong winds flush pollution out of the
system whereas low winds allow pollution level to rise. The correlation studies of the
daily average of PM10 and PM2.5 concentration of all the sites with different
meteorological factors have been conducted to establish their relationship.
The multiple regression equation obtained for PM10 is expressed as:
[PM10]
= 1098.0 - 253.6 [WS] –12.9 [TEMP]
[R2 = 0.45]
(2)
According to this equation the level of PM10 decreases with increasing wind speed and
temperature. The regression equation obtained for PM2.5 is expressed as:
[PM2.5]
= 793.8 –171.3 [WS] - 12.17 [TEMP]
[R2 = 0.42]
(3)
The above equation reveals that the PM2.5 concentration is also decreasing with
increasing concentration of wind speed and temperature.
[PM10] = Concentration of PM10 in µg/m3,
[PM2.5] = Concentration of PM2.5 in µg/m3
[WS] = Wind speed in m/s,
[TEMP] = Temperature in oC.
The above equation reveals that the PM10 (inhalable particle) concentration increases
with increasing PM2.5 (fine particle) concentration and PM10 (inhalable particle)
concentration decreases with increasing wind speed. The equation (2) and (3) is obtained
by using 24-hr average values of PM10, PM2.5 and wind speed.
Figure 8: Temporal variation of PM10 and wind speed
700
y= 1/x
500
3
PM10 (µg/m )
600
400
300
200
100
0
1.5
1.7
1.9
2.1
2.3
2.5
2.7
2.9
3.1
Wind speed (m/s)
Figure 9: Temporal variation of PM2.5 and wind speed
600
PM2.5 (µg/m3)
500
y=1/x
400
300
200
100
0
1.5
1.8
2.1
2.4
Wind speed (m/s)
2.7
3
3.3
CONCLUSIONS
The study provides a valuable baseline data on PM10 and PM2.5 levels and first of its kind
in which such data has been collected in any urban region of India. Data capture rate was
high and the precision of the results was good. The total data were analysed to investigate
temporal variation, meteorological effects and correlation using the code SPSS in order to
gain more understanding on their variability and interrelations. The following
conclusions were reached:
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
The maximum PM10 concentrations at Belaghata EM Byepass crossing, Minto
Park and Park Street stations were 602.63µg/m3, 508.67µg/m3 and 637.70µg/m3
respectively. The minimum PM10 concentrations at Belaghata EM Byepass
crossing, Minto Park and Park Street stations were 66.66 µg/m3, 83.02 µg/m3 and
127.02 µg/m3 respectively.
The maximum PM2.5 concentrations at Belaghata EM Byepass crossing, Minto
Park and Park Street stations were 537.77µg/m3, 330.85µg/m3 and 363.63µg/m3
respectively. The minimum PM2.5 concentrations at Belaghata EM Byepass
crossing, Minto Park and Park Street stations were 49.47µg/m3, 53.94 µg/m3 and
69.14 µg/m3 respectively.
The maximum PM10-2.5 (coarse particles) concentrations at Belaghata EM
Byepass crossing, Minto Park and Park Street stations were 153.53µg/m3,
224.06µg/m3 and 274.06µg/m3 respectively; the minimum PM10-2.5 concentrations
at Belaghata EM Byepass crossing, Minto Park and Park Street stations were
17.19 µg/m3, 27.43 µg/m3 and 40.81 µg/m3 respectively.
The average PM10 concentrations at Belaghata EM Byepass crossing, Minto Park
and Park Street stations were 287.02 µg/m3, 278.24 µg/m3 and 177.63 µg/m3
respectively. The average PM2.5 concentrations at Belaghata EM Byepass
crossing, Minto Park and Park Street stations were 201.48 µg/m3 , 166.58 µg/m3
and 163.64 µg/m3 respectively . The average PM10-2.5 (coarse particles)
concentrations at Belaghata EM Byepass crossing, Minto Park and Park Street
stations were 85.54 µg/m3, 111.67 µg/m3 and 130.27 µg/m3 respectively.
It was found that in the winter (worst season) the PM10 quality was in noncompliance (in 96% cases) with the NAAQM, 1997 and NAAQS, 2003 standards.
In the Park Street area the concentration shows higher value due to higher traffic
congestion along with industrial influence from adjacent Park Circus area.
PM2.5 data appears to be a constant fraction of the PM10 at all the sites, indicating
common influences of meteorology and sources.
Statistical analysis of air quality data shows PM10 and PM2.5 are showing poor
correlation with temperature, relative humidity and solar radiation, but very
significant but inverse correlation with wind speed.
There are clear associations between PM10 and PM2.5 data sets at all the measured
sites. The highest correlation (0.98) was obtained between PM10 and PM2.5 at
Minto Park, followed by 0.96 at Beleghata EM Byepass crossing and 0.66 at Park
Street station.
(10)
Considering the simplicity of the stepwise regression models, their performance
was quite satisfactory, in predicting the observed values. Predictive models
explain 87% and 86% of the PM10 and PM2.5 concentration variance respectively.
ACKNOWLEDGEMENT
The authors are thankful to the West Bengal Pollution Control Board for providing their
support for this study.
REFERENCES
1. Clarke, A. G., Azadi-Boogar, G. A., Andrews, G. E., 1999. Particle size and chemical
composition of urban aerosols. The Science of the Total Environment 235, 15-24.
2. Dockery, D. W., Schwartz, J., Spengler, J.D., 1992. Air pollution and daily mortality:
association with particulates and acid aerosols. Environmental Research 59, 362-373.
3. Dockery, D. W., Pope, C.A., Xu X., Spengler, J.D., Ware, J.H., Fay, M.E., Ferris,
B.G., Spiezer, F.E., 1993. An association between air pollution and mortality in six
US cities.New England Journal of Medicine 329, 1753-1759.
4. Envidas, 2002.Environmental Data Acquisition System Manual. Envitech Ltd., Israel.
5. Harrison, R. M., Deacon, A. R., Jones, M. R., Appleby, R. S., 1997. Sources and
processes affecting concentrations of PM10 and PM2.5 particulate matter in
Birmingham (UK). Atmospheric Environment 31, 4103-4107.
6. IS 5182 - PartIV, 1999. Indian Standard Methods of Air Sampling and Analysis Part
IV, Bureau of Indian Standards, Manak Bhavan, 9, Bahadur Shah Zafar Marg, New
Delhi-110002, India.
7. Katsouyanni, K., Touloumi, G., Spix, C., Schwartz, J., Bakducci, F., Me4dina, S.,
Rossi, G., Wojtyniak B., Suniyer, J., Bacharova, L., Schouten J. P., Ponka A.,
Anderson, H. R., 1997. Short term effects of ambient sulphur dioxide and particulate
matter on mortality in 12 European cities: results from time series data from the
APHEA project. BMJ 314, 1658-1663.
8. Lippmann, M., 1989. Background on health effects of acid aerosols. Environmental
Health Perpectives 79, 3-7.
9. Monn, Ch., Braendli O., Schaeppi G., Schindler Ch.,1995. PM10 & TSP in Urban,
Rural and Alpine air in Switzerland. Atmospheric Environment 29(19), pp 2565-2573
10. NAAQMS, 1997. National Ambient Air Quality Monitoring Series, Central Pollution
Control Board, Government of India.
11. Pope, C. A., Burnett, R. T., Thun, M. J., Calle, E. E., Krewski, D., Ito, K., Thurtson,
G. D., 2002. Lung cancer, cardio-pulmonary mortality and long term exposure to fine
particulate air pollution. Journal of the American Medical Association 287 (9), 11321141.
12. SPSS, 2003. “Advance Statistics Manual”, SPSS/PC+V6.0,SPSS Inc., 444 North
Michigan Avenue, Chicago IL, 60611, U.S.A.
13. USEPA, United States Environmental Protection Agency,1995.Air quality criteria for
particulate matter. EPA Report No. EPA/600/AP-95/001c. 3v.Office of Health and
Environmental Assessment, Environmental Criteria and Assessment Office, Research
Triangle Park, NC.
14. USEPA, United States Environmental Protection Agency,2001.Air quality criteria
document on particulate matter. Draft Report
15. WBPCB, 2002. Health effects of Air Pollution-A study on Kolkata. West Bengal
Pollution Control Board.
16. WHO/UNEP, 1992. Urban Air Pollution in the Megacities of the World, World Heath
Organization/United Nations Environment Programme Oxford, Blackwell.