Measurements of gaseous SO2, NO, NO2, H2S, O3, NH3 and VOCs, in Kuwait city using IVL passive sampler technology, were obtained on monthly basis during the period from March 2011 to February 2012 at 10 locations. The results of this study reveal that those compounds had low concentrations compared to Kuwait Environment Public Authority standards for residential areas. It is found that nitrogen oxides, benzene and xylene were mostly emitted from motor vehicles; in contrast, industrial sources including oil industries were a major contributor to sulfur dioxide and toluene pollution. Higher concentrations of NOx and SO2 were recorded during winter season where the inversion layer is quite pronounced thus trapping more pollutants in ambient air, while higher O3 values were recorded in summer period; the hot season in Kuwait. The spatial distributions of the measured gaseous air pollutants were extremely consistent with the influence of the prevailing wind direction NW. In general and except for PM10, the yearly average air quality indices (AQI) of SO2, NO2, CO and O3 concentrations are coming under the category of “good” and “moderate”.
The State of Kuwait has experienced a very rapid growth rate in the socio-economic and infrastructures over the last four decades. The construction of hundreds of kilometres of urban freeways and arterials has complemented the intense development of the socio-economic sectors. With a population of over three million and a vehicle fleet of more than one million and both growing rapidly, Kuwait City is experiencing increasing volumes of traffic, greater trip frequency and increasing trip length [1] . Thus the outdoor air quality is becoming a major air pollu- tion issue and concern, for the people living in Kuwait City, due to its use for indoor ventalation. It also remains among the key problems of urban environment. Non-methane hydrocarbon concentrations in the ambient air are generally elevated with respect to Kuwait Environment Public Authority (K-EPA) limits [2] ; suggesting oil re- lated activities are the causative agents.
In a survey of more than 100 cities in industrial and non-industrial countries, the World Bank has found that the air in many urban areas remains unhealthy [3] . The study survey indicated that nearly 29% of the surveyed cities experienced sulphur dioxide concentrations (often from power plants) in excess of the maximum levels allowable under WHO guidelines and 71% had nitrogen dioxide concentrations (often from urban traffic) that exceeded WHO maximums [1] . The interaction of primary pollutants with the surrounding environment and the effect of photochemical reactions in the atmosphere results in what is known as secondary pollutants, a valid example of which is ozone (O3). The rate of the conversion of NOx to NO3 affects ozone formation and the fate of the NOx in the atmosphere. NOx is oxidized to Nitric acid (HNO3) in the atmosphere, which in turn forms NO3 particles [4] .
Seven years (1998-2004) of data from three Air Monitoring Stations of Kuwait Environmental Public Author- ity (KEPA) were analyzed by Al-Mutairi and Koushki [1] . The analyses of the data showed that concentrations of the following air pollutants CH4, CO, O3, SO2, NOx and total sulfur (TS), had slightly increased over the sev- en-year study period and only those of the NOx and SO2 exceed the permitted standard levels. Traffic was the main source of air pollution in the district located adjacent to the Kuwait city centre, while oil refineries contri- buted most to ambient air pollution in the rural district. However, the review of literature indicates that no analy- sis of air pollution trend or source identification in Kuwait city has been made in recent years. It is the narrowing of this research gap that this study addresses. This study focused on measurements of selected gaseous air pollu- tants in outdoor ambient air in the vicinity of Kuwait city using passive samplers. The pollutants addressed are sulphur dioxide (SO2), nitrogen monoxide and dioxide (NO, NO2), hydrogen sulfide (H2S), ozone (O3), ammo- nia (NH3), and Total Volatile Organic Compound (TVOC). All the measurements were performed simulta- neously to ensure optimal use of time and effort.
2. Pollution Load in the Study Area
Kuwait City, the capital city of Kuwait, is located at the eastern part of the country on the shore of the Arabian Gulf with an estimated current population of 56,200. It contains most governmental offices, the headquarters of most Kuwaiti corporations and banks. It is the political, cultural and economic center of the state. Automobiles are, by far, the principle mode of travel in Kuwait city. Traffic congestion and rush hours within city limits and along major highways surrounding the city (i.e., 1st to 5th ring road) contribute mainly to the CO, NOx and VOCs le- vels in the air [5] . The average hourly number of vehicles running on the main roads in Kuwait city ranges from 611 to 2792 vehicles/hr (Figure 1). The data, collected from a traffic survey carried by the Ministry of Interior in October 2010 for one week, showed marked day-of-week dependence.
Activities within the city harbor (Mina Shuwaikh port) also contribute to the traffic-related airborne pollutants and pollution load. While traffic is the main source of air pollution in Kuwait City, upstream/downstream Indus- tries such as power stations and oil fields and refineries contribute the least to air pollution emissions in the city (Figure 2). Three major refineries process over 935,000 bbld, situated in the southern part of the country, col- lectively referred to as the refineries belt [6] . In addition, petrochemical processing lines and conversion Indus- tries are all located, somewhat, in the vicinity of this belt (Figure 2). The capacity of the power plant and desa- lination network in the country is estimated at 16,095 MW and 2 × 106 m3/day, respectively [6] . The northern power plants (operating on fossil fuels) are expected to be main sources for SO2 and NOx pollutants in Kuwait city. The emissions of these pollutants from the oil fields are also possible sources in the vicinity of Kuwait city and its surroundings. Kuwait city is also affected by possible emissions of H2S and NH3, which are generated from existing old landfills in the vicinity of the city and agricultural activities, respectively.
3. Materials and methods
Passive (diffusive) samplers were used at 10 locations in Kuwait city to measure the following components: SO2, NO, NO2, H2S, O3, NH3, and VOCs (Benzene, Toluene, n-Octane, Ethyl Benzene, m&p-Xylene, o-Xylene and Nonane). Each of these components was sampled on a monthly basis. Measurements began in March 2011 and were conducted for 12 months. The 10 sampling locations (Figure 3) were systematically selected over Kuwait
Average hourly number of vehicles running on the main roads in Kuwait City during the day time
city with a density of approximately one sample per half-square kilometer. Sites were chosen, wherever possible, to be away from direct local sources, including roads. This study utilized diffusive samplers developed at the IVL (Swedish Environmental Research Institute Ltd—Gothenburg, Sweden) and described in the literature [7] . The IVL samplers are of badge type, 10 mm long and 20 mm internal diameter. A membrane is mounted at the inlet to prevent them from wind-induced turbulent diffusion. The membrane is protected from mechanical dam- age by a stainless steel mesh.
The samplers were mounted under a metal disc in order to protect them from rain and direct sunshine (Figure 4), and fixed at a height 3 m above the ground. Even in the absence of rain, the shelter is required to minimize dust contamination and the effects of advection on the diffusive samplers. Based on the passive sampler manufactur- er’s advice, the samplers were left in the field for about three weeks. Description of the sampling protocol is presented in Table 1. The samplers were prepared at IVL and mailed together with instructions on a monthly basis to the author. The samplers were transported to and from the field in air-tight containers, and minimally handled in the field using latex gloves. The samplers were then sent to IVL for analysis. At IVL, the analytical procedures for the samplers have been accredited. Furthermore, the IVL is participating in the standardization work for testing diffusive samplers within the CEN (European Committee for Standardization). This implies that certified standards are used every time a batch of samples is analyzed. Additional information of the use of IVL-type samplers and their comparison with active sampling results can be found in Ayers et al. [8] [9] and Gillett et al. [10] .
The sampler sites were then geo-referenced and registered to a base-map and the data imported into a surfer software version 10. Statistical data processing was performed using SPSS software package version 19. The data set, consisting of n elements (n = 120), was treated as a 3-dimensional statistical sample defined as:
where i is the air pollutant; j the site ID; k the month ID; x the concentration of the air pollutant i at site j during
Possible local sources of pollutants to Kuwait City
the kth month.
Furthermore, to correlate the concentrations of the measured pollutants with the meteorological conditions, the meteorological data for the period of the sampling were also obtained from Kuwait International Airport.
In the present work, the air quality was also assessed through the use of Air Quality index developed by Al-Shayji et al. [11] for the State of Kuwait, based on the guidelines proposed by the US Environmental Protec- tion Agency [12] . Table 2 shows the AQI ranges, corresponding health effects and associated category codes suggested by the US-EPA. For this purpose continuous measurements of SO2, PM10, NO2, O3 and CO, corres- ponding to a period between March 2011 and February 2012, were used. Specifically, the pollution level time- series were obtained from the Mansouria monitoring station. Mansouria is a district adjacent to the Kuwait city center; i.e., 3 km from the centre. At the station, PM10, CO, NO, NO2, O3 and SO2 concentrations are routinely recorded.
In the station, SO2 and CO concentrations are measured through the UV-fluorescence principle and the infra- red correlation photometry method respectively, while the Beta-gauge method is used to measure PM10 concen- trations. On the other hand, the ozone concentrations are continuously measured through the UV-absorption
Location of the measurement sites used in this study
Passive samplers mounted under a metal disc.
method, whereas NO and NO2 measurements are performed through the chemilumine science principle. It is noted that all analyzers are calibrated regularly through calibration units (gas cylinders, permeation tubes, refer- ence gauges, pre-weighed calibration filters and mass flow controllers).
4. Results and discussion4.1. Basic analysis
Table 3 lists the average results for individual pollutant, detected by the passive samplers at 10 sites for a period of 12 months, including the mean concentrations, coefficients of variation and maximum and minimum values. Most pollutants show a consistency between their median and mean concentrations. The largest disagreements occurred for NO; this may be due to indirect measurement of NO; i.e., NOx and NO2 concentrations were consi- dered to calculate NO value as difference, thus the experimental errors of both NO2 and NOx measurements may be included.
Statistically, Table 3 indicates that the variation in the concentrations among each measured pollutant ranges between 40 to 46%. Also, the mean values are found to be greater than the median values leading to positively
Description of the sampling protocol
Month
Sampling duration
Temp ˚C
Humidity %
WD ˚N
WS m/s
Pressure m Bars
Start
End
March
03/03/2011
26/3/2011
19.75
46.17
213.14
3.81
1013.60
April
04/08/2011
23/4/2011
25.49
43.19
155.69
3.77
1007.25
May
05/07/2011
22/5/2012
32.31
37.58
134.14
3.47
996.13
June
06/11/2011
30/6/2011
36.81
30.80
234.38
3.82
994.43
July
16/7/2011
31/7/2011
38.04
32.68
204.67
3.38
994.89
August
08/01/2011
16/8/2011
38.08
33.61
195.70
3.34
1000.00
September
09/05/2011
20/9/2011
35.06
38.10
130.77
3.42
1008.62
October
14/10/2011
31/10/2011
28.55
43.37
171.01
3.51
1014.27
November
16/11/2011
30/11/2011
20.35
51.12
211.51
4.26
1019.65
December
12/02/2011
21/12/2011
15.07
55.16
253.84
3.96
1019.65
January
01/01/2012
28/1/2012
15.83
51.12
187.43
4.07
1016.74
February
02/10/2012
03/02/2012
15.76
55.96
195.84
3.66
1012.89
AQI range, description and corresponding category codes suggested by Al-Shayji et al. [<xref ref-type="bibr" rid="scirp.44735-ref11">11</xref>] and the US Environ- mental Protection Agency
aBreak points for Kuwait AQI, bBreak points for US AQI, cNO2 has no short term US NAAQS & can generate an AQI only above a value of 200.
skewness values (except for NH3); i.e., values approaching to zero from positive direction. Thus, we can consid- er those pollutants are stabled under these circumstances.
The air pollutant concentrations for each pollutant i broken down by sampling site j and sampling month k; i.e., x = xijk (Equation (1)), are presented in Figure 5. It is important to keep in mind the following point when discussing the experimental data. The diffusive samplers are suitable to describe the spatial distribution of air pollutants simply because they can only integrate exposures to give a cumulative value of all the hourly fluctua- tions in pollutant gas concentrations.
A comparison between the measured concentrations and the applicable air quality standards promulgated by Kuwait EPA, as well as US EPA, reveals the following:
1) The maximum measured concentration (8.0 μg/m3) of NH3 is low value compared to the annual KEPA residential standard (148 μg/m3) for NH3.
2) The maximum measured concentration (35.6 μg/m3) of NO2 is lower than the KEPA annual standard (67 μg/m3) for residential areas for NO2.
3) The maximum measured concentration (20.2 μg/m3) of SO2 is about one quarter of the yearly KEPA resi- dential standard (80 μg/m3) for SO2.
Descriptive statistics of average measured pollutant concentration (µg/m<sup>3</sup>) at 10 sites during March 2011 to Feb- ruary 2012
Mean
95% Confidence Interval for Mean
5% Trimmed Mean
Median
Variance
Std. Deviation
Minimum
Maximum
Max/Min
Range
Interquartile Range
Skewness
Kurtosis
25th percentile
75th percentile
Lower Bound
Upper Bound
H2S
5.5
2.2
5.6
3.8
3.8
8.0
2.4
3.9
7.5
2.0
6.3
4.0
0.3
−0.6
4.6
6.3
SO2
13.4
6.5
15.4
10.8
11.4
62.5
6.3
8.8
20.2
14.4
16.0
9.4
0.3
0.1
10.7
15.4
O3
42.8
21.7
46.8
34.2
36.0
339.9
16.3
27.1
65.0
8.7
41.2
28.2
0.1
−0.3
32.4
51.0
NH3
4.8
2.1
5.0
3.5
3.7
4.3
1.9
3.0
8.0
3.8
4.9
3.2
0.0
−0.6
3.7
5.8
NO2
28.9
12.8
29.0
20.8
20.6
163.7
11.4
21.1
35.6
2.3
28.3
20.0
0.1
−0.7
24.3
33.5
NO
42.4
16.0
42.6
28.5
24.3
409.1
18.6
25.7
62.8
4.3
52.9
29.3
0.8
0.7
32.1
53.0
Benzene
2.5
1.4
2.8
2.0
1.9
1.2
1.0
1.8
3.7
2.2
2.7
1.6
0.4
−0.6
2.0
3.1
Toluene
15.7
7.5
18.3
12.6
11.8
64.1
7.5
9.1
26.9
3.2
22.2
11.4
0.6
0.1
10.7
20.4
N-Octane
1.3
0.7
1.4
1.0
1.0
0.3
0.5
0.8
1.8
3.6
1.3
0.8
0.2
−0.8
1.0
1.5
Ethyle-Benzene
4.6
2.2
5.2
3.7
3.5
5.4
2.1
2.4
7.4
10.9
6.0
3.5
0.4
−0.6
3.2
6.0
m&p Xylene
18.1
9.0
20.7
14.6
13.8
79.7
8.2
10.5
29.1
2.9
23.6
12.8
0.4
−0.4
12.4
23.6
O-Xylene
7.3
3.7
8.2
5.9
5.6
12.1
3.2
4.3
11.4
2.8
9.2
5.1
0.3
−0.5
5.1
9.4
Nonane
1.1
0.3
1.3
0.8
0.7
0.6
0.7
0.6
2.6
4.3
2.3
0.6
2.0
4.7
0.7
1.2
The two percentile points define the upper and lower tails of data. The range indicates the difference between the maximum and minimum measured values. Interquartile represents the difference between the 75th and 25th percentile. Skewness number measures to what extent the distribution of values deviates from symmetry around the mean. Kurtosis number measures the “peakedness” or the “flatness” of a distribution. A positive kurtosis indicates a shape flatter than normal.
4) The maximum measured concentration (7.5 μg/m3) of H2S is almost within the yearly KEPA standard for residential areas (8 μg/m3) for H2S.
5) The maximum measured concentration (3.6 μg/m3) of Benzene is less than the US-EPA annual threshold (5 μg/m3).
4.2. Influence of seasons on the concentration of pollutants
Time-series of monthly average concentration values of measured pollutants are presented in Figure 6. The seasonal variation in ambient levels varies from component to component. For example, SO2 during the winter season, reflects an increase in emissions (16.3 µg/m3) associated with domestic heating in Kuwait city and a de- crease in the rate of the gas-phase loss of SO2 via chemical conversion to sulfate or in part a result of short-range transport of SO2 from oil fields., while a dramatic decrease during the summer is mainly associated with pre- vailing wind direction NW that transports the SO2 from oil fields away from Kuwait city [6] . The change in “normality” behavior between winter and summer months, an increase of concentrations up to 6.8 µg/m3 on average, may also be explained considering that SO2 emissions may be likely associated to the nearby Power Plant located upwind of the city.
Similar, high concentrations of NO and NO2 were recorded for the winter period (43.3 and 35 µg/m3, respec- tively); i.e., dry season for the city. On average, increase of NO and NO2 concentrations between winter and summer months reached up to 34 and 25.6 µg/m3, respectively. Furthermore, the inversion layer, which is quite pronounced during the winter season, traps more pollutants in ambient air and the relatively calmer conditions in dry months due to the preponderance of windless days that favour the accumulation of pollutants. Thus, the seasonality in emissions from the nearby power plants should also be considered. The increased fuel consump- tion in buildings with increased or decreased temperature is likely to partially contribute to high concentrations of NOx and SO2. Their plumes may be emitted below the boundary layer, during early morning, causing them to reach the ground. One other factor, which could account for the elevated amounts of NO and NO2 in winter season, is the higher traffic volume due to the seasonal influx of visitors to the city during this period. On the other hand, a factor contributing to the lower levels of air pollution emissions over the summer season could be associated with the fact that the majority of the population in Kuwait, especially the non-Kuwaiti professionals, leaves the State for summer vacation. This will considerably reduce the traffic contribution to air pollution
Concentrations of measured pollutants.
emissions over this month (9.5, 9.4 and 9.3 µg/m3 for SO2, NO2 and NO, respectively).
The minimum average values of NH3 and H2S occur in fall. The seasonal change seems to be limited and does not reflect the impact of meteorological factors such as humidity and intensity of winds. H2S and NH3 concen- trations could be a result of old-closed landfill sites in the vicinity of the surrounding areas (personnel commu- nication with K-EPA). However, agricultural activities, in particular street plantation, could be the dominant source of NH3 emissions to the atmosphere.
High concentrations of O3 (45.8 µg/m3, on average) were recorded for the summer period, the hot season in Kuwait. This is a major concern with high ambient temperatures that induce ozone formation. Generally, in- crease in power consumption in the summer, reaching up to 10,890 MW (personnel communication with Minis- try of Electricity and Water), due to the increased use of indoor air conditioning, results in increase in emission rate of O3.
For Volatile Organic Compounds (VOC), the seasonal changes of n-Octane and Nonane pollutants seem to be limited and do not reflect the impact of meteorological factors such as humidity and intensity of wind or traffic factor. The concentrations of Toluene ranged from 9.1 to 26.9 µg/m3, with slight increase of concentrations dur- ing winter season (15.9 µg/m3) and decrease in fall season (8.9 µg/m3). Similarly, the concentrations of Benzene, Ethyl-Benzene, o-Xylene and m&p-Xylene showed slight increase of concentrations during winter season (2.96, 4.8, 7.5 and 19.5 µg/m3, respectively) compared to the lower measured concentrations in fall season (1.4, 2.6, 4.3 and 10.83 µg/m3, respectively). In winter season the prevailing southeast wind direction seems to be more effec- tive in diffusing VOCs plume from the downwind emission sources such as the refineries and oil fields to the Kuwait city. Also higher traffic emissions contribution is particularly evident in winter season. Generally speaking, the measured volatile organic compounds seem to behave in a way similar to each other, and the main source of VOC pollutants in Kuwait city comes from motor vehicles, although industrial sources are also rele- vant [6] . However, the low measured concentrations of VOCs indicate that they mostly decay before reaching the city, thus confirming that the contribution of the oil industry sources in Kuwait is minimal. For example, the Xylene value being much higher than Benzene may be interesting since it indicates either Xylene emission sources are different from those of Benzene, or the same source emitting the various compounds at different rates.
Average monthly variation of measured pollutants
4.3. Analysis of correlations between the different pollutants
Table 4 presents the seasonal analysis of correlations of the measured pollutants. The correlations analysis showed that in winter season SO2 and O3 are strongly correlated (r > 0.7) with NO2 and NH3, respectively, while in summer season SO2 and NH3 are strongly correlated with O3 and NO, respectively. In fall season O3 and NO2 are strongly correlated with NO2 and NO, respectively. However, VOC pollutants showed strongest correlation among themselves, especially Toluene with Xylene, indicating sources in common (traffic emissions). As stated by Costabilea et al. [13] the lowest correlation values are associated to residential emissions and the highest to traffic emissions. The comparison between VOCs correlation (r) with NO2 (they are stronger in summer season) may be a further evidence of the shared traffic sources. On the other hand, several pollutants show significant negative correlations with meteorological factors such as temperature, humidity, wind speed, and wind direction. For example O3 showed strong negative correlation with ambient temperature in winter season and strong posi- tive correlation in fall season.
Correlation coefficient, r
Winter
H2S
SO2
O3
NH3
NO2
NO
Benzene
Ethyle Benzene
Nonane
oxylene
mpXylene
Toluene
nOctane
Temp
Humidity
Pressure
WD
WS
H2S
1.00
SO2
0.02
1.00
O3
0.35
−0.10
1.00
NH3
−0.07
0.34
−0.755**
1.00
NO2
0.13
0.768**
−0.01
0.16
1.00
NO
−0.45
−0.17
0.07
−0.26
−0.15
1.00
Benzene
−0.02
0.06
0.15
0.04
0.32
0.38
1.00
Ethyle Benzene
0.27
0.29
−0.12
0.00
0.38
0.14
0.578**
1.00
Nonane
0.709**
−0.26
−0.01
−0.06
0.18
−0.32
0.29
0.18
1.00
oxylene
0.18
0.28
−0.13
0.02
0.35
0.21
0.620**
0.992**
0.14
1.00
mpXylene
0.23
0.29
−0.13
0.02
0.37
0.16
0.590**
0.991**
0.18
0.990**
1.00
Toluene
0.14
−0.26
−0.07
0.00
0.15
0.27
0.730**
0.33
0.747**
0.34
0.35
1.00
nOctane
0.26
−0.449*
−0.06
−0.03
−0.01
0.10
0.557**
0.19
0.858**
0.18
0.20
0.929**
1.00
Temp
−0.40
−0.763**
−0.07
−0.25
−0.742**
0.577**
−0.03
−0.25
−0.01
−0.19
−0.20
0.22
0.25
1.00
Humidity
−0.40
−0.724**
−0.05
−0.22
−0.728**
0.555**
−0.04
−0.31
0.03
−0.24
−0.24
0.25
0.28
0.978**
1.00
Pressure
0.36
0.513**
−0.01
0.11
0.575**
−0.409*
0.06
0.38
−0.12
0.33
0.30
−0.29
−0.31
−0.773**
−0.889**
1.00
WD
0.40
0.766**
0.08
0.25
0.742**
−0.579**
0.03
0.24
0.02
0.18
0.19
−0.21
−0.25
−0.999**
−0.970**
0.752**
1.00
WS
0.15
−0.08
−0.10
−0.09
0.04
0.10
0.07
0.397*
−0.21
0.37
0.31
−0.20
−0.20
−0.18
−0.379*
0.761**
0.14
1
Spring
H2S
SO2
O3
NH3
NO2
NO
Benzene
Ethyle Benzene
Nonane
oxylene
mpXylene
Toluene
nOctane
Temp
Humidity
Pressure
WD
WS
H2S
1
SO2
0.327
1
O3
−0.128
0.260
1
NH3
−0.181
0.587**
0.550**
1
NO2
−0.401
−0.534*
−0.610**
−0.176
1
NO
−0.308
0.476
0.548
0.629*
0.347
1
Benzene
−0.146
0.241
−0.354
0.248
0.365
0.638*
1
Ethyle Benzene
0.476*
−0.115
−0.153
−0.191
−0.167
−0.051
0.398*
1
Nonane
0.569**
0.053
−0.422*
−0.273
0.048
−0.212
0.298
0.738**
1
oxylene
0.443*
−0.105
−0.225
−0.190
−0.126
0.018
0.472*
0.991**
0.769**
1
mpXylene
0.436*
−0.119
−0.206
−0.197
−0.135
−0.023
0.442*
0.994**
0.757**
0.997**
1
Toluene
0.130
0.118
−0.139
0.103
−0.108
0.055
0.730**
0.741**
0.580**
0.766**
0.746**
1
nOctane
0.472*
0.382*
−0.327
0.030
−0.193
0.030
0.643**
0.656**
0.753**
0.711**
0.680**
0.744**
1
Temp
0.804**
0.402*
−0.009
−0.169
−0.633**
−0.423
−0.147
0.326
0.277
0.273
0.270
0.100
0.400*
1
Humidity
−0.834**
−0.441*
0.096
0.174
0.598**
0.423
0.083
−0.323
−0.323
−0.282
−0.276
−0.109
−0.463**
−0.992**
1
Pressure
−0.734**
−0.332
−0.119
0.155
0.673**
0.423
0.236
−0.320
−0.200
−0.252
−0.254
−0.083
−0.296
−0.983**
0.952**
1
WD
−0.677**
−0.281
−0.201
0.144
0.693**
0.423
0.290
−0.310
−0.146
−0.234
−0.238
−0.070
−0.222
−0.953**
0.908**
0.993**
1
WS
−0.845**
−0.496**
0.257
0.178
0.512*
0.423
−0.041
−0.297
−0.39*
−0.282
−0.271
−0.123
−0.563**
−0.934**
0.971**
0.852**
0.782**
1
Summer
H2S
SO2
O3
NH3
NO2
NO
Benzene
Ethyle Benzene
Nonane
oxylene
mpXylene
Toluene
nOctane
Temp
Humidity
Pressure
WD
WS
H2S
1
SO2
0.126
1
O3
0.145
0.873**
1
NH3
0.118
0.925**
0.837**
1
NO2
0.045
0.406
0.121
0.484*
1
NO
−0.214
0.611
0.690
0.848**
0.682*
1
Benzene
−0.015
0.437*
0.266
0.383
0.583**
0.616
1
Ethyle Benzene
−0.101
0.040
0.005
0.136
0.560*
0.552
0.494*
1
Nonane
0.608**
0.316
0.097
0.271
0.432
−0.031
0.472*
0.261
1
oxylene
−0.167
0.184
0.103
0.219
0.621**
0.584
0.673**
0.873**
0.320
1
mpXylene
−0.124
0.058
0.021
0.172
0.574*
0.565
0.500*
0.996**
0.257
0.890**
1
Toluene
0.132
0.186
0.224
0.104
0.018
−0.094
0.667**
−0.141
0.500*
0.042
−0.130
1
nOctane
0.294
0.493*
0.275
0.537*
0.747**
0.656
0.665**
0.538**
0.711**
0.616**
0.543**
0.432
1
Temp
0.072
0.401
0.203
0.551**
0.585**
0.190
0.276
0.321
0.332
0.236
0.349
−0.004
0.383
1
Humidity
0.090
0.524*
0.258
0.601**
0.641**
0.190
0.403
0.214
0.408
0.223
0.249
0.101
0.456*
0.954**
1
Pressure
−0.047
−0.246
−0.131
−0.467*
−0.488*
−0.190
−0.124
−0.396*
−0.230
−0.227
−0.416*
0.116
−0.282
−0.958**
−0.828**
1
WD
−0.084
−0.484*
−0.240
−0.587**
−0.625**
−0.190
−0.361
−0.255
−0.384
−0.230
−0.288
−0.065
−0.433*
−0.981**
−0.994**
0.883**
1
WS
−0.075
−0.423*
−0.213
−0.561**
−0.596**
−0.190
−0.298
−0.306
−0.345
−0.235
−0.336
−0.013
−0.396
−0.999**
−0.967**
0.943**
0.989**
1
Fall
H2S
SO2
O3
NH3
NO2
NO
Benzene
Ethyle Benzene
Nonane
oxylene
mpXylene
Toluene
nOctane
Temp
Humidity
Pressure
WD
WS
H2S
1
SO2
−0.293
1
O3
0.584
0.378
1
NH3
−0.144
−0.487*
−0.560*
1
NO2
−0.643
−0.254
−0.829**
0.447*
1
NO
−0.720*
−0.084
−0.694**
0.450*
0.849**
1
Benzene
−0.574
0.185
−0.554*
0.200
0.657**
0.694**
1
Ethyle Benzene
0.010
0.489
−0.334
−0.244
0.439
0.239
0.531
1
Nonane
0.282
0.041
−0.232
−0.231
−0.025
−0.022
0.320
0.612*
1
oxylene
0.016
0.577
−0.293
−0.349
0.470
0.328
0.637*
0.905**
0.717*
1
mpXylene
0.060
0.572
−0.183
−0.400
0.371
0.220
0.532
0.911**
0.689*
0.992**
1
Toluene
−0.170
0.296
−0.228
−0.113
0.266
0.281
0.733**
0.894**
0.643**
0.952**
0.926**
1
nOctane
−0.183
0.047
−0.437
0.051
0.387
0.415
0.831**
0.662*
0.736**
0.774**
0.703*
0.877**
1
Temp
0.842**
0.623**
0.680**
−0.515*
−0.780**
−0.626**
−0.354
−0.114
0.051
−0.087
−0.078
−0.149
−0.265
1
Humidity
−0.839**
−0.651**
−0.680**
0.545*
0.763**
0.604**
0.332
0.088
−0.057
0.052
0.045
0.130
0.252
−0.999**
1
Pressure
−0.835**
−0.484*
−0.659**
0.370
0.828**
0.702**
0.435
0.224
−0.023
0.238
0.221
0.221
0.311
−0.983**
0.974**
1
WD
−0.844**
−0.578**
−0.677**
0.467*
0.802**
0.656**
0.384
0.153
−0.042
0.140
0.128
0.175
0.283
−0.998**
0.994**
0.993**
1
WS
−0.766**
−0.793**
−0.625**
0.711**
0.586**
0.403
0.157
−0.095
−0.095
−0.191
−0.185
−0.004
0.144
−0.936**
0.950**
0.855**
0.911**
1
**Correlation is significant at the 0.01 level (2-tailed). *Correlation is significant at the 0.05 level (2-tailed).
A cluster analysis was also applied to the obtained concentration data for the measured pollutants using Sta- tistics Toolbox function, with average linkages as the criterion for forming clusters of elements. Figure 7 dis- plays three clusters: 1) H2S, NH3, Benzene, N-Octane, Ethyl-Benzene, o-Xylene and Nonane; 2) SO2, Toluene, and m & pXylene; 3) O3, NO, and NO2. It is observed, however, that clusters 1 and 2 join together at a relatively lower level implying perhaps an uncommon source, while the long distance of gases in cluster 3 may create an impression that NOx may have a different source from other gases; mainly from traffic movement. The cluster analysis also confirms the interaction of NOx pollutants with the surrounding environment resulting in ozone formation.
4.4. Spatial distribution of pollutants
Average resulting contour maps for measured pollutant concentrations in Kuwait city, during March 2011 to February 2012, are illustrated in Figure 8. The spatial distribution patterns of VOCs concentrations are general- ly similar for all compounds except n-Nonane, with elevated average rates at two spot areas, one at northwest and another at southeast of the city. This may confirm that a significant portion of the VOCs is generated from local sources associated with heavy traffic movement at these areas. The elevated concentration of n-Nonane is found in the southeast sector. This trend is well agreed with the trend of blowing wind in Kuwait; i.e., the pre- vailing wind blows from the northwest; the northwesterly wind represents 60% of the total wind directions in Kuwait (Figure 9). This may confirm that the traffic is not a source for n-Nonane.
SO2 concentration distribution showed similarity in spatial distribution patterns of VOC; i.e., the elevated concentrations are found in northeast and southwest sectors. This trend is well agreed with the trend of blowing
Hierarchical dendogram for measured pollutants (the distances reflect the degree of correlation between different elements)
Contour plots for average annual concentrations of measured pollutants (µg/m<sup>3</sup>) in Kuwait City.
wind in Kuwait; i.e., the wind blows dominantly from the northwest during summer season and, to a lesser ex- tent during the winter season, indicating possible sources for SO2, mainly from the power plants in the northeast of the city and other power plants and refineries in the south.
When the relative magnitudes of air pollutant concentrations are displayed on a map, the local distribution of the pollutants becomes apparent, providing clues to the emission sources. For NO and NO2, the high concentra- tion area is widely spread in the northwest sector of the city, i.e., downwind of emission sources such as seaport and the northern power stations, indicating external contribution sources beside the traffic movement. NH3 con- centration distribution showed similarity in spatial distribution patterns of NOx, with expected external source due to sanitary water treatment plants, located in the northwest and west of the city, beside ongoing plantation activities along the roads. The elevated concentration of O3 is radiating from the center of the city, indicating that traffic is a significant source of city pollutants. The elevated concentration of H2S is found in the southeast sector where construction activities are going on. Kuwait city in the last 10 years has been exposed to H2S emit- ted from exposing the underground water during ongoing constructions; the water has been raised up and pol- luted over old landfills.
4.5. Air quality index (AQI)
Figure 10 shows the results of AQI values for the hourly average SO2, NO2, CO, O3 and PM10 concentrations for the period from March 2011 to February 2012 at Mansouria AQM station. The calculated AQI values for the hourly average SO2 and NO2 concentrations are categorized as “good” and “moderate” during the study period, except in May and September AQI for NO2 showed 3.1% of total hourly data as “unhealthy for sensitive” cate
The seasonal and annual windrose plots for Kuwait City based on metrological data for the period from March 2011 to February 2012.
Summary of total annual AQI values, based on 24 h average, for SO<sub>2</sub>, NO<sub>2</sub>, CO, O<sub>3</sub> and PM<sub>10</sub> concentrations in Kuwait City
gory. Roughly 67 and 29% of the calculated AQI values for PM10 concentrations are categorized as ‘‘good” and “moderate”, respectively. In winter season, 51% - 79% of the average monthly data was categorized as “mod- erate”, while in summer season the percentage reduced to 25% - 46%. The calculated AQI values for the hourly average O3 are categorized mainly as “good” and partially as “moderate”, except in May AQI for O3 showed 6% of total hourly data as “unhealthy for sensitive” category. All calculated AQI values for the hourly average CO concentrations are categorized as “good” during the study period.
The overall assessment of the air quality in Kuwait city, during the months March 2011 to February 2012, in- dicate that less than 2% of the calculated hourly average AQI values are categorized as “unhealthy for sensitive” and “very unhealthy”. The remaining 98% of AQI values are categorized as “good” and “moderate”.
The AQI was defined taking into account each pollutant separately. The idea behind this approach is that dealing with each pollutant separately is easier due to the known health effects resulting from the population’s exposure to it. Nevertheless, it seems that up to now Kuwait city has no adverse health effects caused by human exposure to a mixture of air pollutants, although their synergetic impact may not always be easily measured. PM10 may be the only worrying pollutant in the city and its elevated values on some occasions is directly related to the natural dust fallout which is a natural common phenomenon in Kuwait, especially in the summer season.
5. Conclusions
The investigation, using multipoint passive samplers, has revealed a good degree of uniformity of pollutants dis- tribution over Kuwait city. The study has provided a preliminary assessment of source contributions and fre- quency distributions of major air pollutants in Kuwait city. For the measured compounds, it was found that they had low concentrations with no violation of Kuwait EPA air quality standards. Statistical analysis of the data also confirmed that motor vehicles are the largest source emitting nitrogen oxides. Traffic movement is also a main possible emission source of benzene and other VOC pollutants, influenced, however, by a more variable type of emissions related to oil industries. The concentrations of SO2 were strongly influenced by the wind direction and from the nearby power plants due to the use of high sulphur content in locally available crude oil (4%). The cor- relation analysis showed that O3 is strongly correlated (r > 0.7) with NO2 in hot season and SO2 in cold season. Furthermore, the inversion layer, which is quite pronounced during the winter season, and temperature varia- tions have clear influence on the seasonal variation in pollutant accumulation in the city, particularly on the concentrations of NOx and SO2.
The result shows that air pollution levels in the ambient air quality can be characterized as “good” and “mod- erate” for SO2 and CO for all the time, 99% of the times for NO2 and O3, and 96% for PM10, during twelve months of measurement (March 2011 to February 2012). The remaining percentage of the time for which PM10 is categorized as unhealthy, is partially related to regular dust fallout in the country.
With such appreciably elevated levels of pollutants in Kuwait city, actions should be taken to reduce emission and studies are needed to assess the role of local sources on long-term environmental impacts. Vehicle emission control could be vital management practice to reduce the input of emissions into the atmosphere. However, for pollutants emitting from oil industrial processes, other precautionary measures should be considered.
Acknowledgements
This work was supported by Kuwait University Research Grant No. [SE03/08]. The author gratefully acknowl- edge the assistance, co-operation and logistical support provided by Mr. Ahmed El-Tabei, Mr. Younis Al-Naif and Mr. EmadAl-Khabaz at the Department of Earth and Environmental Sciences.
ReferencesAl-Mutairi, N. and Koushki, P. (2009) Potential Contribution of Traffic to Air Pollution in the State of Kuwait. American Journal of Environmental Sciences, 5, 218-222. http://dx.doi.org/10.3844/ajessp. 2009.218.222Al-Awadhi, F.A. and Al-Awadhi, S.A. (2006) Bayesian Statistical Modeling for Different Pollutants in the State of Kuwait. Environmetrics, 17, 739-752. http://dx.doi.org/10.1002/env.791World Bank (2003) World Development Indicators. World Bank Publications, Washington DC.Matsumoto, K. and Tanaka, H. (1996) Formation and Dissociation of Atmospheric Particulate Nitrate and Chloride: An Approach Based on Phase Equilibrium. Atmospheric Environment, 30, 639-648.
http://dx.doi.org/10.1016/1352-2310(95)00290-1Ettouney, R.S., Abdul-Wahab, S.A. and Elkilani, A.S. (2009) Emissions Inventory, ISCST, and Neural Network Modelling of Air Pollution in Kuwait. International Journal of Environmental Studies, 66, 193-206.
http://dx.doi.org/10.1080/00207230902859929Al-Salem, S.M. and Khan, A.R. (2008) Comparative Assessment of Ambient Air Quality in Two Urban Areas Adjacent to Petroleum Downstream/Upstream Facilities in Kuwait. Brazilian Journal of Chemical Engineering, 25, 683-696. http://dx.doi.org/10.1590/S0104-66322008000400006Ferm, M. and Rodhe, H. (1997) Measurements of Air Concentrations of SO2, NO2 and NH3 at Rural and Remote Sites in Asia. Journal of Atmospheric Chemistry, 27, 17-29. http://dx.doi.org/10.1023/ A:1005816621522Ayers, G.P., Keywood, M.D., Gillet, R.W., Manins, P.C., Malfroy, H. and Bardsley, T. (1998) Validation of Passive Diffusion Samplers for SO2 and NO2. Atmospheric Environment, 32, 3593-3609.
http://dx.doi.org/10.1016/S1352-2310(98)00079-XAyers, G.P., Peng, L.C., Gillet, R.W. and Fook, L. (2002) Rainwater Composition and Acidity at Five Sites in Malaysia in 1996. Water, Air and Soil Pollution, 133, 15-30. http://dx.doi.org/10.1023/ A:1012967614759Gillett, R.W., Ayers, G.P., Mhwe, T., Selleck, P.W. and Harjan, H. (2000) Concentrations of Nitrogen and Sulfur Species in Gas and Rainwater from Several Sites in Indonesia. Water, Air and Soil Pollution, 120, 205-2015.
http://dx.doi.org/10.1023/A:1005223124903Al-Shayji, K., Lababidi, H.M., Al-Rushoud, D. and Al-Adwani, H. (2008) Development of a Fuzzy Air Quality Performance Indicator. Kuwait Journal of Science and Engineering, 35, 101-126.US Environmental Protection Agency (2006) Guideline for Reporting of Daily Air Quality—Air Quality Index (AQI). Office of Air Quality Planning and Standards, Research Triangle Park.Costabilea, F., Bertonia, B., Desantisa, F., Wanga, F., Weiminb, H., Fengleib, L. and Allegrinia, I. (2006) A Preliminary Assessment of Major Air Pollutants in the City of Suzhou, China. Atmospheric Environment, 40, 6380-6395.
http://dx.doi.org/10.1016/j.atmosenv.2006.05.056