Science of the Total Environment 350 (2005) 225 – 237

www.elsevier.com/locate/scitotenv

Dependence of urban air pollutants on meteorology

Hamdy K. ElminirT
National Research Institute of Astronomy and Geophysics, El-Marsad Street, Helwan, Cairo, Egypt
Received 10 October 2004; accepted 31 January 2005
Available online 11 March 2005

Abstract

Dependence of air pollutants on meteorology is presented with the aim of understanding the governing processes pollutants
phase interaction. Intensive measurements of particulate matter (PM10) and gaseous materials (e.g., CO, NO2, SO2, and O3) are
carried out regularly in 2002 at 14 measurement sites distributed over the whole territory of Great Cairo by the Egyptian
Environmental Affairs Agency to assess the characteristics of air pollutants. The discussions in this work are based upon
measurements performed at Abbassiya site as a case study. The nature of the contributing sources has been investigated and
some attempts have been made to indicate the role played by neighboring regions in determining the air quality at the site
mentioned. The results hint that, wind direction was found to have an influence not only on pollutant concentrations but also on
the correlation between pollutants. As expected, the pollutants associated with traffic were at highest ambient concentration
levels when wind speed was low. At higher wind speeds, dust and sand from the surrounding desert was entrained by the wind,
thus contributing to ambient particulate matter levels. We also found that, the highest average concentration for NO2 and O3
occurred at humidity V 40% indicative for strong vertical mixing. For CO, SO2 and PM10 the highest average concentrations
occurred at humidity above 80%.
D 2005 Elsevier B.V. All rights reserved.

Keywords: Aerosol; Sulfur dioxide, SO2; Nitrogen dioxide, NO2; Particulate matter with diameter less than 10 Am, PM10; Carbon monoxide;
CO and ozone, O3

1. Introduction compounds (Fraser et al., 2003). However, only for

about 200 of the pollutants have the impacts been
To date nearly 3000 different anthropogenic air investigated, and the ambient concentrations are
pollutants have been identified most of them organic determined for an even smaller number (Dzubay and
(including organometals). Combustion sources, espe- Mamane, 1989; Van Borm et al., 1990; Paoletti et al.,
cially motor vehicles, emit about 500 different 1991, 1999; Al-Rajhi et al., 1996; Esteve et al., 1997;
Kasparian et al., 1998; Chan et al., 1999, etc.). To
better understand the effects that exposure to acid
T Tel.: +20 02 5560645; fax: +20 02 5548020. aerosols has on humans, epidemiologic air pollution
E-mail address: [email protected]. studies have begun to characterize the spatial and
0048-9697/$ - see front matter D 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.scitotenv.2005.01.043
226 H.K. Elminir / Science of the Total Environment 350 (2005) 225–237

temporal distribution of acid aerosols (Suh et al., monitoring ambient air quality at 14 sites distributed
1992; Purdue et al., 1992; Samet et al., 2000; Pope et over the whole territory of Great Cairo area. In
al., 2002). But to fully understand the processes general, sites characteristics are different representing
responsible for the spatial and temporal distribution of industrial, traffic, urban, residential and background.
acid aerosols requires analysis of local and regional Not all parameters are being measured by the EEAA
meteorology; especially wind direction; wind speed; program at all sites. This depends on site specifica-
turbulence; and atmospheric stability (Michael, 1997; tion and typical dominating sources in the specific
Hien et al., 2002; Laakso et al., 2003). Chemical area. Therefore, the discussions in this work are
reactions also depend on ambient weather conditions based upon measurements performed during 2002 of
because they are influenced by short-wave radiation; the different indicators measured at Abbassiya station
air temperature; and air humidity. Along with chem- as a case study. The variables monitored were SO2,
ical reactions, dispersion and dilution processes result NO2, PM10, CO and O3. Monthly averages, medians,
in ambient air pollution which shows concentrations standard deviations, minimum and maximum con-
of different substances varying with regard to time centrations for Abbassiya monitoring station are
and space. summarized in Fig. 1. According to the national
Although the relationships between synoptic mete- ambient air quality standards it is clear from Fig. 1
orology and air pollution have been investigated for that atmospheric concentrations of all pollutants have
pollutants such as ozone (Comrie, 1990, 1992, 1994; been found to be consistently much higher than
Eder et al., 1994), SO2 (Kalkstein and Corrigan, permissible limits.
1986), NO2 (Davis and Kalkstein, 1990), greenhouse Fig. 2 presents the corresponding meteorology
gas induced global warming (Kalkstein et al., 1990), taken at Abbassiya station. Ambient air temperature,
and even visibility (Davis, 1991), we still have very (T in 8C); relative humidity (RH in %); wind speed
little information about the dependence of urban (WS in ms
1) and wind direction, (WD in degree) are
aerosol on the city’s geographical, geological and only listed here as the parameters—especially the
meteorological conditions. Several works have been wind patterns—which are the most relevant for
done studying air pollution in Great Cairo area, where interpretation of the results. Temperature is measured
the sources of air pollutants have been identified using dry bulb, wet bulb and maximum–minimum
(Hagazy, 1961; Salam, 1967; Nassr-Allah, 1968; thermometers installed in a Stevenson screen. Relative
Salam and Sowelim, 1976; El-Taieb, 1981; Mosalam, humidity is derived from the dry and wet bulb
1986; Abdel-Rahman et al., 1988; El-dahab, 1990; El- temperatures from a hygrometric chart. Wind speed
Hussainy and Sharobiem, 2002). In view of the above is measured by an automatic anemometer and wind
discussion, it is necessary to define the status of direction by a standard wind vane. Fig. 2 shows
ambient air quality due to the presence of different surface average temperatures between 14 and 31 8C
pollutants in the environment of Cairo. This paper and relative humidity 50–75% with annual average
examines the temporal relationship between meteoro- wind velocity of 3 ms
1. According to Hansen (1974)
logical parameters and urban air pollutants, helping to the seasonal variation observed may be caused by
fill a void in research into the relationships between changes in air mass type and not by any local source
the atmospheric circulation, local meteorology, and of air pollutants.
concentrations of tropospheric air pollutants. The concentration of dust in the atmosphere was
monitored by means of a portable air sampler. To
measure the concentration air was drawn into the
2. Apparatus and measurements sampler, and by virtue of their inertia the particles
were deposited on membrane filters, which are most
Different organizations (e.g., Centre for Environ- commonly equipped with inlets that eliminate par-
mental Hazard Mitigation, CEHM at Cairo Univer- ticles above a specified size cut. Filters are weighed
sity; the Institute of Graduate Studies and Research, under controlled temperature and relative humidity
IGSR at Alexandria University; and the Egyptian conditions before and after sampling, and mass
Environmental Affairs Agency, EEAA) have been concentrations are determined from the increase in
 H.K. Elminir / Science of the Total Environment 350 (2005) 225–237 227

16
CO(mg/m3)

12
Max.
8 Ave. Med.
Min.
4 S. D.
0
120
O3(µg/m3)

80 Max.
Ave.
40 Med.
S. D.
Min.
0
160
NO2(µg/m3)

120

80 Max.
40 Ave. Med .
Min. S. D.
0
J F M AM J J A S ON D J F MAM J J A S ON D J F MAM J J A S ON D J F MAM J J A S O N D J F MAM J J A S ON D
100 Months
80 Max.
SO2(µg/m3)

60
40 Ave. Med.
Min. S. D.
20
0
500
PM10(µg/m3)

400 Max.
300
200 Ave. Med.
Min. S. D.
100
0
J F MAM J J A S ON D J F MAM J J A S ON D J F MAM J J A S ON D J F MAM J J A S O N D J F MAM J J A S ON D
Months

Fig. 1. Monthly averages (Ave.); median (Med.); minimum (Min); maximum (Max) values; and standard deviation (S.D.) of the air pollutant
concentrations observed at Abbassiya station during 2002.

filter mass and the volume of air sampled. The weight and autumn (October and November). The average
was divided by the surface area from which they were summer temperature is 29 8C, while that of winter is 16
collected to give the dust deposition density in Ag/m2. 8C with an average difference between day and night
Mineralogical analyses of atmospheric particulates of 10 8C. This great temperature difference promotes
collected on filter media have been performed by X- the formation of dew at dawn as the relative humidity
ray diffraction. The methodology of X-ray diffraction of the air becomes generally high, especially during
is described by Ignasi et al., (2001). the winter season. The prevailing wind velocity with
an average of 3 ms
1, but gusts of up to 4–5 ms
1 may
be experienced in early morning and late afternoon.
3. Site and climate dependence of tropospheric Dust storms also occur during April and May when the
aerosol collection Khamsin winds blow over the Egyptian western desert
with a wind speed of the order of 10 ms
1. Relative
Like most subtropical regions along the North humidity fluctuates between 59% in June and 71% in
African desert, Cairo city has mainly two seasons: December, with visibility of about 5 km. This
summer (May–mid-October) and winter (December– restricted visibility is the result of the presence of
February), with a very short spring (March and April) solid particles in the atmosphere.
228 H.K. Elminir / Science of the Total Environment 350 (2005) 225–237

N
NNW NNE

NW NE

WNW ENE

W E

WSW ESE

<=2
SW SE
>2 -4
>4 -5
SSW SSE >5
S
35 90

30 80
Ambient air temperature (oC)

Relative humidity (%)

R.H (%)
25 70

20 60

T (oC)
15 50

10 40

Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec.

Fig. 2. Meteorological parameters recorded at Abbassiya monitoring station during 2002.

The strong emissions of trace gases and aerosol one can see that air pollution level is high during
particles by vehicles traveling on the city’s narrow winter months and create winter syndrome due to low
roads, industry and resuspended soil dust, together temperature, low mixing depth, pollution inversion,
with secondary aerosol, coupled with the unfavorable traffic density, driving habits or due to ratio of
natural conditions of dispersion, are responsible for automobiles to trucks (Kolehmainen et al., 2000). In
the high concentrations of pollutants observed in the several monitoring stations distributed over the
Great Cairo metropolitan area. A brownish cloud of metropolitan region of Great Cairo, PM10 concen-
air pollution forms over Cairo can be seen by the trations consistently exceed the national air quality
naked eye. This cloud denotes a high level of both PM10 standard of 50 Ag
3. Concentrations of CO,
NO2 pollutant and of man-made aerosols. In general, SO2, and NO2 are also high due to high motor vehicle
 H.K. Elminir / Science of the Total Environment 350 (2005) 225–237 229

emissions. The high level of pollutants occurs, general, remarkable similarities are present in urban
indicating that air pollution issue is actually a regional and rural concentration values in and around monitor-
issue, not limited anymore to the city borders. ing sites. While rural values are as a rule lower, the
general features remain unchanged indicating that the
conditions are mainly governed by the same air
4. Chemical and mineralogical composition masses both in urban and rural sites. According to
elemental and mineralogical analysis of the dust
The question of the existence and trace power of collected from all traps in the city of Cairo, the dust
regional elemental characteristics reflecting the struc- particles were composed mostly of quartz and calcite,
ture of emission sources at a given location has been with smaller amounts of dolomite and clay minerals.
treated in a number of publications. As summarized in The major constituents were Si and Ca; minor
some reference papers (Pacyna et al., 1984; Tsoar and constituents were Al, Fe, Mg, K, and Na. The
Pye, 1987; Pye, 1992; Church et al., 1990; Koltay, characteristic features of the dataset presented in
1990) single element traces or ratios of elemental Fig. 3 can be summarized as follows:
concentrations can be used for studying the nature of Silicon, Si, and calcium, Ca, are regarded as crustal
main emission sources in the region as well as for element. They are important components of PM10
pinpointing the sources areas of transported pollutants measured in Cairo. The two elements hold about 23%
to the site of observation. Among others, sulfur is an of PM10 mass. The presence of coarse silicon can be
indicator for fossil fuel burning, fine fraction vana- ascribed to erosion from desert land surrounding
dium and nickel are excellent indicators for residual monitoring station. Ongoing building activities and
fuel burning, copper and zinc are often observed to be degradation of building elements can be considered as
linked with traffic (together with lead), arsenic is a additional sources for increasing silicon level. The
tracer for coal combustion, selenium has a variety of same effect appears in soil-derived elements, Ca and
sources, but mostly fossil fuel burning, leaded titanium, Ti. Calcium, Ca, concentrations are much
gasoline; smelters and waste incinerators are all similar in all locations except for increased coarse
possible sources of lead (Borbély-Kiss et al., 1999). fraction in Helwan site which are to be ascribed to
The availability of coarse dust particles in Cairo partly degradation of buildings, partly to emanating
city is explained by the effects of human activities from industries such as cement plants. Sulfur, S,
resulting in disturbances to the soil, notably con- concentration indicate the determining contribution of
struction work, and also by motorized traffic. In long-range transport processes during which gaseous

20
March
18 April
May
16
Relative concentration (%)

June
14 July
August
12

10

0
Na Ca Zn Si Cl Al K Ti Fe S Ce Mg
Mineralogical composition

Fig. 3. X-ray diffraction analysis of dust collected from all traps in the city of Cairo.
230 H.K. Elminir / Science of the Total Environment 350 (2005) 225–237

sulfur emission is subject to a continuous conversion of NOx. Being an urban site, NO concentrations at
to particulate secondary sulfur. The increased values Cairo are generally found to be well above the
in sulfur, S, concentration can be considered as an threshold level for O3 production. For example, the
indication for possible influence of Shoubra El- well-known photo-oxidation cycle of CO can be
Kheima on Cairo air. Potassium was ascribed to represented as:
mixed natural (e.g., soil derived contribution as in the
CO þ OHYO2 þ H;
silicon case) and anthropogenic sources.
Low chlorine, Cl and sodium Na concentrations H þ O2 þ MYHO2 þ M;
can be understood as a strong sea-derived contribu-
tion. No attempt has been made to use seawater- HO2 þ NOYNO2 þ OH;
related enrichment factors to search for a sea-spray
contribution to the local chlorine in Cairo city. Small NO2 þ hmðk b 420nmÞYOð3 PÞ þ NO;
concentration elements of aluminum Al, zinc, Zn,
manganese, Mg and iron, Fe indicating the importance
Oð3 PÞ þ O2 þ MYO3 þ M;
of soil derived dust of traffic origin. Finally, the
composition of dust deposited in Cairo city during
Net: CO þ 2O2 þ hmYCO2 þ O3 :
Khamasin storm was similar to that deposited during
normal weather conditions, indicating that the com- In this process NOx acts as a catalyst and continues
position of dust deposited during the storm over a to do so until it is permanently removed by physical
wide area of the desert around Cairo was uniform. processes (e.g. surface deposition) or transformed to
other NOx compounds. Irrespective of a nearly
consistent diurnal pattern, O3 concentration can also
5. Trends of air pollutant concentrations vary significantly from one day to the next as depicted
in Fig. 5. The datasets illustrated in Fig. 5 were
The diurnal pattern of O3 (see Fig. 4) is charac- constituted from 365 days, but for a clearer reading no
terized by maximum concentration in the afternoon symbols per month are reported. Note also that,
and minimum concentration during early morning missing data were replaced with the expected values
hours. The increase in O3 during day is basically due provided from a nearby Gomhoria station and
to photo-oxidation of precursor gases, like CO, CH4, suspected erroneous values were removed after care-
and hydrocarbons in the presence of sufficient amount ful examination. The variation of monthly surface

120
Average diurnal variation of O3 (µg/m3)

100

80

60

40

20

0
1 3 5 7 9 11 13 15 17 19 21 23
Time of day

Fig. 4. Diurnal variation of ozone (O3) concentration at Abbassiya monitoring station.

 H.K. Elminir / Science of the Total Environment 350 (2005) 225–237 231
PM10 (ug/m3)
450
300
150
0

160
NO2 (ug/m3)

120
80
40

100
SO2 (ug/m3)

75
50
25
0

16
CO (mg/m3)

12
8
4
0

120
O3 (ug/m3)

80
40
0
Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec.

Fig. 5. Daily variations of air pollutant concentrations recorded at Abbassiya monitoring station during 2002.

ozone concentrations are depicted in Fig. 1. Fig. 1 extensive observations of O3 and its precursors by
shows that O3 measurements have average higher UV photometry at different sites are needed to
concentrations (79 Ag/m3) in June and lower concen- understand O3 formation processes. The yearly dis-
trations (6 Ag/m3) in December. The reason for this tributions of the monthly average concentration values
high O3 in June (low O3 in December) is mainly due for NO2 show the same regularity, (e.g., NO2 exhibit
to active (inactive) photo-chemical production mech- well expressed summer and spring maxima). Increase
anism along with favorable (unfavorable) meteoro- in the average levels of NO2 are observed during July
logical conditions and the related increasing (103 Ag/m3), from August to December, the monthly
(decreasing) precursors concentration. The increase average concentration of NO2 decreased from about
in O3 is attributed to the increase in NOx and other O3 82 to 62 Ag/m3.
precursor emissions by different sources in the The diurnal pattern of SO2 is characterized by
proximity of this site. On seasonal basis, O3 has maximum concentration in the morning. Higher levels
small concentration in wintertime. This is due to low of SO2 during morning hours at Cairo are due to the
solar radiation intensity and high emissions of combinations of anthropogenic emissions, boundary
reductive substances. In spring, the pattern for Cairo layer processes, chemistry as well as local surface
goes up to higher concentrations. In summer and wind patterns. During night hours, the boundary layer
autumn the patterns of surface ozone concentration descends and remains low till early morning, thereby
seem to be the same. resisting the mixing of the anthropogenic emissions
In fact O3 concentration observed at Cairo is low with the upper layer. Hence, pollutants get trapped in
(about 46 Ag/m3) probably due to low NOx concen- the shallow surface layer and show higher levels.
tration as compared to O3 concentrations observed at During these hours of maximum concentrations of
many sites in US, Mexico, Europe, China and Taiwan SO2, anthropogenic emissions are also prominent due
(Oltmans and Levy, 1994; Aneja et al., 1999; Luo et to rush hours (Suh et al., 1992; El-Hussainy and
al., 2000; Wang et al., 2000). However, more Sharobiem, 2002). It is important to note that the
232 H.K. Elminir / Science of the Total Environment 350 (2005) 225–237

major anthropogenic source for CO, SO2 and NO2 is 6. Dependence of tropospheric pollutants on
fossil fuel burning (combustion in motor vehicles). meteorological parameters
Whereas during noontime the higher height of the
boundary layer provides a larger mixing region and 6.1. Contribution of wind aspects to concentration of
hence the pollutants get diluted. The monthly varia- air pollutants
tion in average SO2 concentrations ranges from a
minimum value of about 20 Ag/m3 during January to a The prevailing winds, which may transport mois-
high value of about 55 Ag/m3 during December. On ture or aerosol particles from distant sources, play a
seasonal basis, SO2 exhibit well expressed autumn major role on the seasonal variation of turbidity. A
maxima. On the other hand, high levels of CO correlation analysis indicates significant negative
generally occur in areas with heavy traffic congestion. correlation between total urban concentration and
In Cairo 85–95% of CO emissions may come from wind speed data. Such behavior is plausible for non-
motor vehicle exhaust. Other sources of CO emissions crustal elements where transport effects increased by
include industrial processes and residential waste higher wind speed, give an explanation for the clear-
burning. The daily 8-h average CO concentration up of the local air. An investigation into the effect of
was exceeded 10 mg/m3 in 33% of the time. wind direction aspects on airborne particulate data and
Inspecting the results in Fig. 5 it is apparent that, the subsequent analysis of results obtained in this
there is a very clear seasonal variation in CO and study may show possible source direction. In this
PM10 concentrations. Higher concentrations have section, the possible source direction that affects the
been found during autumn/winter months and lowest PM10, CO, NO2, SO2, and O3 data is identified
concentrations are found during summer. These roughly by comparing average contributions of wind
revealed the effect of weather condition on the aspect to the concentrations of mass with the average
distribution pattern. These cycles, which are more frequencies of wind direction.
pronounced for PM10 than for gases, appear in the According to statistical results of the analysis of
monthly averages as well as in median concentrations meteorological factors measured at the Abbassiya
(see Fig. 1). Data for the months November, monitoring station, the major prevailing wind direc-
December and January (cold season) and June, July tions are easterly (E, SSE, ES, and ESE) and westerly
and August (warm season) were combined and (W, WSW, WS, and SSW) and least frequent from the
averaged. Winter to summer concentration ratios for north and northwest. The yearly average frequencies
pollutant components PM10, CO and SO2 were 2.141, for easterly and westerly wind were 28.2% and
1.676 and 1.656, respectively. From the previous 35.9%, respectively. Table 1, demonstrates the sig-
analysis one concludes that, the high atmospheric nificance of wind direction for the resulting air
loading for air pollutants was frequently observed in pollution. This behavior closely reflects the geological
wintertime. structure of the region. The concentration levels of O3

Table 1
Average concentration of air pollutants as a function of wind direction
Atmospheric air pollutants Atmospheric air pollutants
PM10 CO NO2 SO2 O3 PM10 CO NO2 SO2 O3
(Ag/m3) (mg/m3) (Ag/m3) (Ag/m3) (Ag/m3) (Ag/m3) (mg/m3) (Ag/m3) (Ag/m3) (Ag/m3)
N NON NON NON NON NON S 181 6 94 23 76
NNE 171 6 62 18 24 SSW 182 6 80 25 64
NE 195 5 78 22 22 SW 179 6 71 26 49
ENE 188 6 78 28 34 WSW 174 7 69 39 25
E 189 6 80 26 44 W 156 5 80 54 24
ESE 178 5 80 32 66 WNW 140 5 60 46 22
SE 167 6 79 24 75 NW 185 5 65 66 14
SSE 163 6 70 21 80 NNW 122 5 68 66 16
 H.K. Elminir / Science of the Total Environment 350 (2005) 225–237 233

Table 2 are slightly higher at a lower relative humidity since

Average concentration of air pollutants determined within selected the reaction of NO2 with OH is decreased. For all
ranges of relative humidity
other components the highest average concentrations
PM10 CO SO2 NO2 O3
occurred at humidity above 80%. An increase in
R.H. V 40% 175 5 28 83 63 humidity reflects an updraft of boundary layer air
R.H. 41–60 148 5 28 77 61
masses to the 3 km level, leading to a higher level of
R.H. 61–80 180 6 31 76 44
R.H. N 80% 367 7 41 69 25 air pollution.
In order to study the relationship between the trace
gas concentrations and relative humidity a correlation
and NO2 were highest when transport was from the analysis was performed. Table 3, shows the correla-
easterly wind. These observables also have high tion matrix between relative humidity and the
values for the south where an iron and steel factory corresponding data for CO, NO2, PM10, SO2, and
is located. CO, PM10 and SO2 concentrations did not O3 concentrations. It can be noticed the barely
show the same directionality, with average concen- positive relation between SO2, PM10 and CO concen-
trations of 5.67 mg/m3, 171.3 and 34.4 Ag/m3, from trations with relative humidity, and significant neg-
all directions except the westerly wind for SO2. This is ative trend for O3 and NO2. The correlation of low
further evidenced that the markers for primary humidity with low concentrations can be attributed to
emissions are of urban origin. the influence of clean free tropospheric air masses
which is in agreement with the findings by Seinfeld
6.2. Tropospheric pollutants dependence on relative and Spyros (1998). On the other hand, high humidities
humidity may also indicate precipitation events accompanied
by in-cloud scavenging, which results in low concen-
Analysis of the surface wind speed and direction trations of gas and aerosol concentrations. NO2 and
alone will not adequately explain the variability in the O3 should be more affected by in-cloud gas scaveng-
concentrations of air pollutants. Hence, analyses of ing than SO2 or PM10. For this reason the highest
the meteorological parameters affecting ambient con- average concentration for surface ozone occurred at
centrations of air pollutants should include an humidity V 40. This finding is in good agreement with
indicator of atmospheric moisture (e.g. water vapor/ the seasonal cycle of NO2 component.
humidity content). The statistical analysis of daily air From the previous discussion one concludes that,
pollutant concentrations and daily values of the the correlation coefficient between air pollutant
relative humidity for the whole period at Abbassiya concentrations and relative humidity is not very high.
station was summarized in Table 2. The lower correlation coefficient for all pollutants
Inspecting the results we found that, the highest illustrates the competition of two mechanisms: atmos-
average concentration for O3 and NO2 occurred at pheric dispersion (particles are removed from con-
humidity less than or equal to 40%. This can be taminated surface air by dry deposition and by wet
mainly attributed to the enhanced oxidation of hydro- deposition in precipitation), and aerosolization from
carbons in the afternoon, which support the ozone surfaces (emissions of aerosol particles by vehicles
production. This is in agreement with the findings of traveling on the city’s narrow roads, industry and
Ma and van Weele, (2000). The NO2 concentrations resuspended soil dust).

Table 3
Correlation matrix for air pollutants and relative humidity
Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec. Yearly value
SO2
0.34
0.04
0.03 0.01 0.39 0.39 0.25
0.39 0.01 0.54 0.55 0.18 0.196
PM10 0.23 0.10
0.01
0.02 0.57 0.64 0.02
0.29
0.06 0.36 0.70 0.15 0.252
CO 0.50
0.10
0.15 0.56 0.55 0.63 0.18
0.66 0.28 0.26
0.20
0.53 0.204
O3
0.49
0.07
0.07
0.51
0.75
0.64
0.02 0.41
0.13
0.80
0.76
0.35
0.422
NO2
0.36
0.23
0.19
0.15
0.14 0.47
0.04 0.23 0.00
0.75
0.24
0.42
0.138
234 H.K. Elminir / Science of the Total Environment 350 (2005) 225–237

6.3. Air pollutants dependence on ambient air observed. The association between primary pollu-
temperature tants and temperature was found to be weak or
insignificant. The temperature changes did not
The temperature variation experienced during the significantly influence SO2 and NO2 concentrations
present study is sufficient to examine the relationship (see Fig. 6). Although considerable positive correla-
of ambient temperature with the trace gases tion was found between ambient temperature and

(A) (B)
120 500
R = +0.695 450 R = - 0.412
Surface ozone (µg/m3)

100
400

PM10 (µg/m3)
80 350
300
60 250
200
40
150
20 100
50
0 0
5 10 15 20 25 30 35 40 5 10 15 20 25 30 35 40
Ambient air temperature (oC) Ambient air temperature (oC)

(C) (D)
100 160
R = - 0.204 140 R = + 0.187
80
120
NO2 (µg/m3)
SO2 (µg/m3)

60 100

40 80

60
20
40

0 20
5 10 15 20 25 30 35 40 5 10 15 20 25 30 35 40
Ambient air temperature (oC) Ambient air temperature (oC)

(E)
20
R= - 0.497
16
CO (mg/m3)

12

0
5 10 15 20 25 30 35 40
Ambient air temperature (oC)

Fig. 6. Air pollutant concentrations as a function of ambient air temperature.

 H.K. Elminir / Science of the Total Environment 350 (2005) 225–237 235

Table 4
Correlation matrix for air pollutants and ambient air temperature
Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec.
SO2 0.74 0.38 0.21 0.15
0.52
0.48
0.77
0.30
0.35
0.42
0.66
0.94
PM10
0.24
0.19
0.12 0.02
0.62
0.84
0.76
0.51
0.58
0.56
0.83
0.53
CO
0.93
0.44
0.48
0.21
0.45
0.83
0.89
0.79 0.30
0.01
0.03 0.75
O3 0.96 0.83 0.61 0.24 0.66 0.89 0.80 0.89 0.35 0.39 0.86 0.97
NO2 0.30 0.12
0.29 0.12 0.00
0.66 0.65 0.41
0.35 0.38 0.10 0.80

concentration of NO2 in July and December (Table 7. Conclusion

4). The same trend was observed for SO2 but with
negative correlation. The increase in NO2 concen- The aim of the present study was to establish
tration during summer months may be attributed to which meteorological parameter influences the ele-
an enhanced formation of a secondary aerosol. In the mental concentrations to characterize the short range
case of the secondary pollutants the situation is transport of traffic-related airborne particles and to
different. Surface ozone O3, which is an indicator of standardize the concentrations with respect to the
photoxidation processes, show a clear trend of meteorological parameters. On the basis of the
increasing with temperature (Table 4). Higher O3 present study, it can be concluded that wind and
concentrations have been found during spring/sum- relative humidity are shown to be the most important
mer months and lowest concentrations are found meteorological parameters influencing the behavior
during winter. Non-availability of sufficient solar of air pollutants, which are further affected by the air
radiation and washout of pollutants result in near mass history. Measurements show that wind direction
absence of photochemical ozone production during is not directly responsible for daytime higher O3
this period. Further, rainfall and sky cover, which are concentrations, but these directions have some con-
higher during winter shows negative correlation with tributions in the diurnal variations of CO, PM10, SO2
temperature and surface O3 concentration (Michael, and NO2. The pollutants associated with traffic (e.g.,
1997). CO, NO2, SO2, and PM10) were at highest ambient
On seasonal basis, PM10 exhibit well expressed concentration levels when wind speed was low. At
winter and spring maxima. The yearly distribution of higher wind speeds, dust and sand from the
the monthly average concentration values for CO surrounding desert was entrained by the wind, thus
show the same regularity. Table 5 summarizes contributing to ambient particulate matter levels.
average concentration values determined within Measurements also show that, no significant differ-
ambient air temperature ranges 10–15, 16–20, 21– ences are found between the chemical composition
30 and N30 8C. Inspecting the results we found that, deposited during Khamasin storm and that deposited
the highest average concentration for NO2 and O3 during normal weather conditions, indicating that the
occurred at ambient temperature greater than 30 8C. composition of dust deposited during the storm over
For PM 10 and CO the highest concentrations a wide area of the desert around Cairo was uniform.
occurred at ambient air temperature between 10 However, analysis of the surface wind speed and
and 20 8C. direction alone will not adequately explain the
variability in the concentrations of air pollutants.
Table 5 Analyses of the meteorological parameters affecting
Average concentration values determined within selected ranges of concentrations of air pollutants should include the
temperature temperature and an indicator of atmospheric moisture
PM10 CO SO2 NO2 O3 (e.g. the relative humidity), and mixing heights that
T = 10–15 8C 197 8 12 56 16 facilitates the regional transport of ambient air
T = 16–20 8C 237 6 20 67 28 pollutants.
T = 21–30 8C 154 5 33 80 61 On the basis of relative humidity we found that, the
N30 8C 110 5 58 107 85 highest average concentration for NO2 and O3
236 H.K. Elminir / Science of the Total Environment 350 (2005) 225–237

occurred at humidity V 40%. This can be mainly Eder B, Davis J, Bloomfield P. An automated classification scheme
attributed to the enhanced oxidation of hydrocarbons designed to better elucidate the dependence of ozone on
meteorology. J Appl Meteorol 1994;33.
in the afternoon, which support the ozone production. El-dahab, A. Study of the limit of carbon monoxide in different
For other components the highest average concen- units in the iron and steel company. Faculty of science. Helwan
trations occurred at humidity above 80%. An increase University Meeting, 25 March; 1990.
in humidity reflects an updraft of boundary layer air El-Taieb, N. Some studies on air pollution and solar radiation in
Helwan industrial area. M.Sc. thesis, Al-Azhar University; 1981.
masses to the 3 km level, leading to a higher level of
El-Hussainy F, Sharobiem W. Studies on atmospheric sulfur dioxide
air pollution. We also found that, the association values and trends. Al-Azhar Bull Sci 2002;13(1).
between PM10 and temperature was found to be Esteve V, Rius J, Ochando E. Quantitative X-ray diffraction phase
insignificant. In the case of the secondary pollutants analysis of coarse airborne particulate collected by cascade
the situation is different. Surface ozone O3 and NO2 impactor sampling. Atmos Environ 31;1997.
show clear trend of increasing with temperature. In Fraser P, Yue W, Buzcu B. Source apportionment of fine particulate
matter in Houston, TX, using organic molecular markers. Atmos
future work, we will study the seasonal variability of Environ 2003;37:2117 – 23.
the global UV radiation and its dependence on urban Hagazy, A. Physical studies on polluted air in Egypt. M.Sc. Thesis,
air pollutants. Cairo University; 1961.
Hansen V. Determination of atmospheric turbidity parameters from
spectral solar radiation measurements. Arch Meteorol Geophys
Bioklimatol 1974;22.
References Hien D, Bac T, Tham C, Nhan D, Vinh D. Influence of
meteorological conditions on PM2.5 concentrations during the
Abdel-Rahman M, Said S, Shuaib N. Comparison between monsoon season in Hanoi. Atmos Environ 2002;36.
atmospheric turbidity coefficients of desert and temperate Kalkstein L, Corrigan P. A synoptic climatological approach for
climate. Sol Energy 1988;40(3). geographical analysis: assessment of sulfur dioxide concen-
Al-Rajhi M, Al-Shayeb S, Seaward D, Edwards M. Particle size trations. Ann Assoc Am Geogr 1986;76.
effect for metal pollution analysis of atmospherically deposited Kalkstein L, Dunne C, Vose S. Detection of climatic change in the
dust. Atmos Environ 1996;30. Arctic using a synoptic climatological approach. J Climatol
Aneja VP, Oommen RG, Riordan AJ, Arya SP, Wayland RJ, Murray 1990;3:1153 – 67.
GC. Ozone patterns for three metropolitan statistical areas in Kasparian J, Frejafon E, Rambaldi P, Yu J, Vezin B, Weolf P, et al.
North Carolina, USA. Atmos Environ 1999;33:5081 – 93. Characterization of urban aerosols using SEM-microscopy, X-
Borbély-Kiss I, Koltay E, Szabó GY, Bozó L, Tar K. Composition ray analysis and Lidar measurements. Atmos Environ
and sources of urban and rural atmospheric aerosol in eastern 1998;32.
Hungary. J Aerosol Sci 1999;30(3):369 – 91. Kolehmainen M, Martikainen H, Hiltunen T, Ruuskanen J.
Chan C, Simpson W, Mctainsh H, Vowles D, Cohen D. Source Forecasting air quality using hybrid neural network modeling.
apportionment of PM2.5 and PM10 aerosols in Brisbane by Environ Monit Assess 2000;65.
receptor modeling. Atmos Environ 1999;33. Koltay E. Elemental analysis of atmospheric aerosols: results and
Church T, Arimoto R, Barrie L, Dehairs F, Dulac F, Zoller H. The perspectives of the PIXE technique. Int J PIXE 1990;1.
long-range atmospheric transport of trace elements A critical Laakso L, Hussein T, Aarnio P, Komppula M, Hiltunen V, Viisanen
evaluation In the long-range atmospheric transport of natural Y, et al. Diurnal and annual characteristics of particle mass and
and contaminant substances, NATO ASI Series, vol. 297. number concentrations in urban, rural and arctic environments
Dordrech7 Kluwer; 1990. in Finland. Atmos Environ 2003;37.
Comrie A. The climatology of surface ozone in rural areas: a Luo C, St John JC, Xiuji Z, Lam KS, Wang T, Chameides WL. A
conceptual model. Phys Geogr 1990;14:295 – 316. nonurban ozone air pollution episode over eastern China:
Comrie A. An enhanced synoptic climatology of ozone using a observations and model simulations. J Geophys Res 2000;105:
sequencing technique. Phys Geogr 1992;13:53 – 65. 1889 – 908.
Comrie A. A synoptic climatology of rural ozone pollution at three Michael P. An analysis of the meteorological parameters affecting
forest sites in Pennsylvania. Atmos Environ 1994;28:1601 – 14. ambient concentrations of acid aerosols in Uniontown, Penn-
Davis R. A synoptic climatological analysis of winter visibility sylvania. Atmos Environ 1997;31(6).
trends in the Mideastern United States. Atmos Environ 1991; Mosalam A. Atmospheric transmission and diffuse fraction of solar
25B:165 – 75. energy over Egypt. Proc Int Symp on Applications of Solar and
Davis R, Kalkstein L. Development of an automated spatial Renewable Energy ASRE 86 1986;1.
synoptic climatological classification. Int J Climatol 1990;10: Nassr-Allah, M. Air contaminants in the industrial community of
769 – 94. Helwan. M.Sc. Thesis, Ain Shims University; 1968.
Dzubay T, Mamane Y. Use of electron microscopy data in receptor Oltmans SJ, Levy II H. Surface ozone measurements from a global
models. Atmos Environ 1989;23. network. Atmos Environ 1994;28:9 – 24.
 H.K. Elminir / Science of the Total Environment 350 (2005) 225–237 237

Pacyna N, Semb A, Hassen J. Emission and long-range transport of Salam, A. Study of air pollution in Helwan industrial area. Report
trace elements in Europe Tellus, vol. 36; 1984. submitted to the high committee for planning of Great Cairo;
Paoletti L, Diociauti M, Falchi M, Pisant D, Ziemacki G. 1967.
Quantitative analysis of airborne breathable particles A compar- Salam A, Sowelim M. Dust deposits in the city of Cairo. Atmos
ison between different analytical techniques. Atmos Environ Environ 1976;1.
1991;25. Samet M, Dominici F, Curriero C, Coursac I, Zeger L. Fine
Paoletti L, Diociaiuti M, Beradis B, Santucci S, Lozzi L, Picozzi P. particulate air pollution and mortality in 20 US cities,
Characterisation of aerosol individual particles in a controlled 1987–1994. N Engl J Med 2000;343.
underground area. Atmos Environ 1999;33. Seinfeld J, Spyros N. Atmospheric chemistry and physics from air
Pope A, Burnett R, Thun J, Calle E, Krewskik D, Thurston D. Lung pollution to climate change. New York7 John Wiley & Sons,
cancer, cardiopulmonary mortality, and long term exposure to Inc.; 1998.
fine particulate air pollution. JAMA 2002;287. Suh H, Spengler D, Koutrakis P. Personal exposures to acid aerosols
Purdue L, Pahl A, Wilson E. Overview of the AREAL acid aerosol and ammonia. Environ Sci 1992;26.
research program. In Proc I992 US EPAIA and WMA Int Symp Tsoar H, Pye K. Dust transport and the question of loess formation.
on Measurement of toxic and related air pollutants, VIP-25. Sedimentology J 1987;34.
Pittsburgh7 Air and Waste Management Association; 1992. Van Borm W, Adams F, Maenhaut W. Receptor modeling of the
p. 259 – 63. Antwerp aerosol. Atmos Environ 1990;24b.
Pye K. Aeolian dust transport and deposition over Crete and van Weele Michiel, Ma Jianzhong. Effect of stratospheric ozone
adjacent parts of the Mediterranean Sea. Earth Surf Processes depletion on the net production of ozone in polluted rural areas.
Landf 1992;17. Chemosphere, Glob Chang Sci 2000;2:23 – 37.
Queralt Ignasi, Sanfeliu Teófilo, Gomez Eva, Alvarez Carlos. X-ray Wang Z, Sha W, Ueda H. Numerical modeling of pollutant transport
diffraction analysis of atmospheric dust using low-background and chemistry during a high-ozone event in northern Taiwan.
supports. Aerosol Sci 2001;32:453 – 9. Tellus 2000;52:1189 – 205.