Ambient Air Quality Assessment of Al-Mansoriah Residential Area in the State of Kuwait

Air quality data (ground level ozone (O3), non-methane hydrocarbons (NMHC), methane (CH4), nitrogen oxide (NO), nitrogen dioxide (NO2), carbon monoxide (CO), hydrogen sulfide (H2S) and sulfur dioxide (SO2) of a Kuwaiti residential area, Al-Mansoriah, were analyzed to evaluate and determine: (a) the exceedances of air quality from permissible limits set by the Kuwait Environment Public Authority (KUEPA), (b) the diurnal patterns of air pollutants, (c) the predominant sources of airborne pollutants in the surrounding area, and (d) the ''weekend effect'' on ozone levels. The dataset covered the period of five consecutive years, from January 2000 to December 2004. High levels of ozone were witnessed; recording a number of exceedances. Inhabitants of Al-Mansoriah were exposed, during the period of study, to acute and chronic levels of SO2. Concerning NMHC, the permissible limit (0.24 ppmrolling average between 6-9 am), was violated in each of the monitoring years. NO diurnal pattern showed two distinct strong peaks during the months from October to March, in which the low solar radiation does not cause intense photochemical reactions, which lead to NO destruction. A clear H2S source (considered as a primary one) was witnessed from Kuwait city (Sharq district). The backed up sewage lines and maintenance work coinciding with study period clearly affected AlMansoriah. A clear "weekend effect", in terms of difference in levels occurring during midweek and weekends, was ascertained. The analysis suggests that Al-Mansoriah residential area should be considered a NOx sensitive region of Kuwait.


Introduction
Since the autonomy of Kuwait Environment Public Authority (KUEPA) from the jurisdiction of the Ministry of Health (MOH) and the recognition of its legal holding power, pollution prevention in the state of Kuwait has _____________________________________________ *Corresponding author's e-mail: ssalem@ns1.prsc.kisr.edu.kw been of a growing concern over urban and suburban areas. Decision makers in the state express widespread interest in controlling greenhouse gases (GHGs), volatile organic compounds (VOCs) and other airborne pollutants. Increased levels of these pollutants, besides eroding materials, are known to have deleterious effects on human health (Brunekreef andHolgate, 2002, Al-Salem andBouhamrah, 2006), to cause injury to plants (Saitanis et al. 200;Saitanis and Karandinos, 2002), to reduce crops yield (Nali et al. 2002) and to negatively affect ecosystems (Bytnerowicz et al. 2002). Natural ecosystems adjacent to urban areas are exposed increasingly to air pollutants of urban origin. Air pollutants, particularly SO 2 , NO 2 and the major secondary photochemical oxidant O 3 are important threats to plants. Their effects comprise many physiological and biochemical changes in plants, which may result in growth reduction and yield loss, even at low levels chronic exposure. Concerning human health, tropospheric O 3 measured in urban regions and across regional airsheds, are known to affect human health (Riga-Karandinos and Saitanis, 2005).
The objective of this work was to investigate the air pollution of a typical residential area in Kuwait, namely Al-Mansoriah. To this purpose we explored the air pollution levels in this area, with emphasis to ground level ozone (O 3 ), non-methane hydrocarbons (NMHC), methane (CH 4 ), nitrogen oxide (NO), nitrogen dioxide (NO 2 ), carbon monoxide (CO), hydrogen sulfide (H 2 S) and sulfur dioxide (SO 2 ). The exceedances of the air quality threshold values were calculated. The diurnal profiles and the seasonal variations of these gases are studied and the relationships between each other were explored. In addition, pollution sources in the surroundings of the area under investigation were identified using a series of concentration and wind roses analyses. The so called ''weekend effect'', concerning the variations of ozone levels between weekdays and weekends, was also investigated.

Investigated Area Description
Al-Mansoriah residential area is a luxurious suburb of the city which hosts mainly Kuwaiti residents of mid/high class (39,444 residents). Figure 1 shows the area from satellite imagery, relative to Kuwait city. Private and public transport vehicles reach a maximum of 45,000, mostly around the areas main streets.
The receptor (monitoring) point chosen for the area was above (at) the polyclinic of Al-Mansoriah situated near the area's Co-Op, which is associated with restaurants and other air polluting anthropogenic activities. These pollutants include n-CH 4 (resulting from restaurants), NO x and CO (emissions of automobile and other burning sources) and VOCs from the gas and gasoline dispensing station. On the other hand, Al-Arabi sports club is considered one of the largest sports and family functions facility in Kuwait, in terms of children entertainment lounges and other function spaces.
Cairo St. is situated to the north side of the receptor point. It is one of the busiest streets of Kuwait especially during rush hour, linking three main residential areas together in the governorate, ie. Al-Mansoriah, Al-Qadisiah and Al-Daeiah. Automobile vehicles operating with gasoline are the main source of pollution associated with traffic jam in the street. Rush hours are usually between 7:30 am to 9:00 am (weekdays); and 8:00 pm to 9:30 pm (weekends).

Data Acquirement and Methods
The dataset of Al-Mansoriah monitoring station, used in this study, was provided form the Kuwait Environment Public Authority (KUEPA). The data cover the period from Jan 1 st 2000 to Dec 31 st 2004. The station is equipped with a number of air sampling devices and analyzers (ThermoTecg 08, SatCal, Al-Issa Co. ltd.) with accuracy of order 1%, with a measuring frequency of 5 minutes (original data). Air probe was approximately 15 m above sea level. All data were stored and manipulated with EnviDas data acquisition software, which did store up to three months worth of data values. Data collected by the station concerned the following pollutants: CH 4 (ppm), NMHC (ppm), CO (ppm), NO (ppb), NO 2 (ppb), H2S (ppb), SO 2 (ppb), and O 3 (ppb). Metrological conditions data were collected through a weather station recording the following parameters: wind speed (ms -1 ) and direction ( o ), relative humidity (%) and ambient air temperature ( o C). Data points were treated and filtered before performing any analysis. Filtration procedure was performed as indicated by (Al-Salem and Khan, 2006). NO 2 and NO values exceeding 200 ppb were excluded, in order to eliminate any automobile point source effect on data collection. Values of CH 4 levels below 1.3 ppm were also removed to avoid insertion of instrumentation signal noise. Calibration and span check points were also deleted.
Many researchers have been occupied with the great number of reactions and types of urban air pollutants (Wen and Lian, 2009;Fenger, 2009;Ramanathan and Feng, 2009;Gildemeister et al. 2007). One of the most important characteristics of urban air ph otochemistry is the oxidation of SO 2 and NO 2 and their conversion to particulate sulfate 2 4 ( SO ) − and (gaseous and particulate) nitrates (NO 3 ). Moreover, the rate of the conversion of NO x to NO 3 affects ozone formation and the fate of t he NO x in the atmosphere. NO x is oxidized to nitric acid (HNO 3 ) in the atmosphere, which in turn forms NO 3 particles (Matsumoto and Tanaka, 1996). Acidification of precipitation, visibility reduction and deleterious effects on human health and plants, ar e all effects associated with these secondary pollutants. Such assessment of urban air quality will surely aid in the planning step of agencies concerned with green industries and keeping up with threshold and permissible limits in populated areas. Althou gh air quality was investigated in a number of urban areas in the state of Kuwait (Al -Salem and Khan, 2006;Al -Salem et al. 2008a;Al-Salem et al. 2008b;Al-Salem et al. 2008c), there is insufficient or lack of information about the air pollution levels in many areas, which host a high percentage of the population of the country.

Exceedances of Air Quality Permissible Limits/Threshold Values
In order to assess the air quality in Al-Mansoriah area, measured levels of air pollutants were compared with threshold limits specified in the rules and regulations of Kuwait Environment Public Authority (KUEPA, 2001) and the European Directive standards (EEA, 2003a;EEA, 2003b). Concerning O 3 , for the protection of both, human health and ecosystem, KUEPA rules and regulations set a permissible limit of 80 ppb as an hourly rolling (continu-ous) average for O 3 , not to be exceeded twice in a 30 days period. Inhabitants of Al-Mansoriah were exposed to O 3 levels above the limit, but only a few days (Table 1). In 2000, six exceednaces were recorded based on 8-hrs rolling averages, and one in the year 2004 (Table 1). Judging from the current dataset of Al-Mansoriah, O 3 levels should be considered potentially phytotoxic in this residential area. Investigations in other countries have shown that the nowadays ambient ozone levels can cause injury to plants. For example, (Saitanis et al. 2004) used plants of the ozone bioindicator Bel-W3 tobacco variety, at 28 sites around a town in Greece. In all those sites the used bioindicator plants were highly injured even after one week of exposure. Similar studies, incorporating instrumental monitoring and phytodetection with ozonebioindicator plants, have demonstrated the occurrence of ozone at potentially phytotoxic levels in other rural and forest areas in Greece Saitanis, 2003;Saitanis et al. 2003). KUEPA has not yet established any long-term objectives for ozone levels in the near future (5-15 years). Long-term objectives have been established for Europe after considering the results of the scientific experiments conducted in the framework of the UN-ECE project over several years in many European countries (Riga-Karandinos and Saitanis, 2005).
Current permissible limit in Europe, has been set to an AOT40 value of 3000 ppb*h above which direct adverse effects on trees, crops and natural vegetation may occur (Benton et al. 2000). AOT40 is the sum of the exceedances of the hourly ozone concentrations above the threshold of 40 ppb, during the daylight hours over a period of three months. The level of 40 ppb (80 µg/m 3 ) of O 3 should not be considered as a lower concentration limit in Europe for biological effects, since some biological responses may occur below that value (EEA, 2003a;EEA, 2003b). By comparison with other major European cities, Al-Manosriah exceeded the number of permissible limit exceedances (8 hr rolling average) in the same duration of monitoring in Paris (Vardoulakis et al. 2002) and Aberdeen (Marr et al. 2007).
SO 2 levels in Al-Mansoriah were generally high, which on the other hand, risks inhabitants of acute and chronic exposure to the pollutant. The KUEPA permissible limit (30 ppb) was violated in 2003 (annual threshold) and 5 exceedances were recorded in 2003 based on daily rolling averages. This also was typical of a Greek metropolis (Volos), were levels were monitored on the basis of a rolling average as well (Riga-Karandinos and Saitanis, 2005). It was also noticed that SO 2 levels exceeded the European threshold limit for human protection reported previously (EAA, 2003a). Concerning NO 2 , 279 average exceedances per year were observed during the five years of monitoring for the short term (1 hr) limit. The corresponding annual permissible limit (30 ppb) was exceeded during the monitoring period (Table 1). For carbon monoxide (CO) the limit value (8000 ppb) of a daily mean was exceeded 6 times in 2004. Methane (CH 4 ) and non-methane hydrocarbons (NMHC) levels were monitored as well. KUEPA has not yet established any regulation for CH 4 in Kuwait. It has, however, set a daily permissible limit (0.24 ppm) for the period 6-9 am for NMHC. The analysis of data showed that NMHC recorded 153 exceedances (average) during the five years of monitoring. The number of exceedances increased gradually on an annual basis (Table 1).

Diurnal Patterns of the Pollutants
A good way of unraveling the dynamics and patterns of air pollutants is to examine their diurnal pattern. The average diurnal patterns of O 3 , NMHC, CH 4 , NO, NO 2 , CO, H 2 S and SO 2 levels for each month of the year were examined (Fig. 2). Significant variations were observed between the pollutants studied in both, the recorded levels and diurnal patterns. It was observed that, in Al-Mansoriah residential area, CO, NO and NO 2 , exhibit the typical urban daily pattern of primary air pollutants characterized by two peaks, one in the morning and one late in the evening (Riga-Karandinos and Saitanis, 2005). This pattern coincides with the human activities (traffic, open market hours and central heating). SO 2 exhibited an unstable diurnal fluctuation. SO 2 levels have been commented to be driven not only for the emissions but also from the levels of O 3 and OH (Harrison and Perry, 1986) and the solar radiation (Riga-Karandinos and Saitanis, 2005). Khoder's (2002) results show that the oxidation processes and conversion of SO 2 to sulfate and of NO 2 to nitrate (NO 3 ) depend on the photochemical oxidation and that the sulfur conversion ratio increases with increasing ozone concentration.
The NO levels were higher during the months from October to March. During these months, the lower solar radiation intensity does not allow intense photochemical reactions (Riga-Karandinos and Saitanis, 2005), which would lead to NO destruction. The concentrations of NO, from April to September, were lower than during October to March months. However, during all, high and low NO months, the two typical peaks -one early in the morning and another in the evening -were present The patterns of O 3 show a clear seasonal and diurnal variation (increased ozone levels during months from April to September and a strong daytime photochemical ozone buildup due to photooxidation of precursor gasses) showing more intense photochemical processes during the luminous hours of the days and during the high sunlight season (Fig. 2). According to (Simpson et al. 1995), annual cycle of ozone over mid-latitudes shows mainly two kinds of seasonal behavior, either with a broad summer maximum, generally attributed to local photochemical productions and typical of populated and industrialized areas, or with a spring maximum, of more controversial origin, typical of remote regions and associated to background conditions. The high rush-hour traffic during early morning between 7:30 am to 9:00 am (weekdays); and 8:00 pm to 9:30 pm (weekends), well before sun radiation is strong enough to trigger photochemical reactions of ozone formation leads to ozone destruction shown in Fig.  2 as ozone decrease.
The witnessed patterns of CO were, somewhat, inverse to those of ozone. It was clear that seasonal variations of ozone affected indirectly the seasonal pattern of CO. Ozone is the photochemical precursor of hydroxyl radicals (OH) in the lower troposphere. Reaction with OH removes most gases of natural and anthropogenic origin from the atmosphere (Jockel et al. 2003). It is also a fact that the main mechanism for the removal of CO from the atmosphere is the reaction of OH. On the other hand, tropospheric oxidation reactions of CO in the presence of sufficient NO also lead to the production of substantial amounts of ozone. The comparison of the diurnal patterns of O 3 and CO, suggests that the depletion of O 3 and the enhancement of CO, during the non-photochemistry hours (early morning and late in the evening), come from the O 3 titration when mixed with urban emissions and simultaneously the anthropogenic release of CO. This pattern of CO may be the typical signature of urban influence (Fig.  2).
Regarding CH 4 patterns, a clear increase was witnessed in the early morning hours (5-7 am) associated with restaurants and other gas operated businesses opening hours. The same is clear for non-methane hydrocarbons (NMHC). The levels increase for both pollutants and reach their peak in the three summer break months, ie. July, August and September. This could be attributed to the increasing activities in luxurious areas like Al-Mansoriah in the summer time.

Wind and Pollutants' Concentrations Roses
In order to investigate the predominant sources affecting Al-Mansoriah residential area, five years mean wind rose and concentration roses for the air pollutants (O 3 , NMHC, CH 4 , NO, NO 2 , CO, H 2 S and SO 2 ), around the receptor point, (Al-Mansoriah polyclinic) were analyzed. Kuwait is characterized by north/northwestern winds (primary direction) contributing to over 70% of the winds around the year (Khan and Al-Salem, 2007). Figure 3 shows the 15 0 wind speed rose executed for the period of study around Al-Mansoriah. West/west-northern and north/northwestern winds covering the corridor of wind direction between 285-3300, establish the primary wind direction with average winds exceeding 6 ms -1 . All wind and pollutants' concentrations' roses in this study are in a blowing form manner.
Southeastern winds could be considered as secondary winds ( 6 ms -1 ) in the case of Al-Mansoriah residential area, as well as the state of Kuwait over the past 30 years (Al-Hajraf et al. 2005).
In the case of H 2 S, the concentration rose reveals a spike value (3450-exceeding 5 ppb) in the direction of Sharq district in Kuwait city (Fig. 4). Backed up sewage and municipal work has been commenced in the period of study in which pedestrians in the district use to complain for suffering rotten eggs odor (typical of H 2 S vented or other type of emissions from municipal manholes) in Sharq. Other directions on the concentration rose show similar emission strength from wind corridors, which are typical for urban and suburban areas (Khan and Al-Salem, 2007).
Methane gas followed typical behavior in almost every case (excluding industrial sites), being uniform from all wind directions (Fig. 5). CH 4 averaged at 2 ppm, which is also typical of Kuwaiti urban areas as previously reported in Fahaheel (Al-Salem and Khan, 2006;Al-Salem et al. 2008a) and Al-Riqa (Al-Salem et al. 2008b;Al-Salem et al. 2008c;Al-Salem and Al-Fadhlee, 2008). CO also showed typical urban behavior and concentration rose trend; being uniform around the receptor point and averaging about 1 ppm (Fig. 5). Daily activities in Al-Mansoriah contribute greatly to both CH 4 and CO concentration roses. Automobile emissions, dry cleaning services, gas dispensing networks and restaurant emissions; all contribute to both pollutants. Sources of these emissions  are distributed around the receptor point, which explains the trends left by both CH 4 and CO on the roses. Regarding NMHC, the concentration rose shows a peak in the emission strength in the northwestern direction (Fig.  5). Figure 6 shows the concentration rose for NO and NO 2 gases around the receptor point. Northern winds (covering 345-150) were clearly the richest in both pollutants (exceeding 60 ppb for NO 2 and 70 ppb for NO). This direction corresponds to Kuwait city, which hosts a number of combustion sources that contribute to NO and NO 2 levels in the ambient.
Opposite to NO x , winds blowing from northwestern and eastern directions showed higher ozone levels (Fig.  7). NO x titration and other sets of reactions fade ozone from the lower atmosphere, which also explains the NO primary direction of north. SO 2 was present around the receptor point in an almost equal level of 25 ppb (Fig. 8).

Weekend/ Weekday Variations
During the past decade, weekend/weekday differences in levels of air pollutants (especially of O 3 ) have been a subject of research interest (Marr and Harley, 2002;Vandaele et al. 2002;Fujita et al. 2003, Riga-Karandinos andSaitanis, 2005). Although, weekly changes in the emissions caused by human activities are known to affect the weekly cycle of ambient pollutant concentrations, this emission-concentration relationship, in urban, suburban and rural sites, is not well elucidated as it is reflected in controversial reports (Blier and Winer, 1999;Pun et al. 2003). In the state of Kuwait, during the period of the study, the weekend was on Thursdays and Fridays (nowadays the weekend has shift to Fridays and Saturdays).
In weekends (Thursdays and Fridays), the emissions of anthropogenic pollutants and especially of O 3 precursors, are believed to be lower compared to those occurring during weekdays, because car traffic is lower and several polluting plants and factories may be less active or inactive. Many approaches have been used for the study of the socalled weekend effect, meaning the occurrence of increased O 3 levels and decreased NO x during weekends in comparison to weekdays. Near constant O 3 levels, during weekdays, in the face of strong weekday/weekend variations in NO x , should also be viewed as a weekend effect (Fujita et al. 2003).
After taking into account the shift in the relaxing days and the days of high anthropogenic activities in Kuwait, in comparison to the west countries, in this study, we analyzed the monthly averages of mean hourly O 3 and NO concentrations during Sunday-Tuesday vs. the corresponding concentrations on Wednesday, Thursday, Friday and Saturday, recorded in Al-Mansoriah, (Fig. 9). In Table 2, the relevant linear regression equations and the corresponding correlation coefficients are given.
Concerning O 3 , the slops of linear regressions revealed that Wednesday concentrations were very close to those occurring in during Sun-Tue ( Fig. 9; Table 2). Unexpectedly, Thursdays-Saturdays O 3 concentrations were much lower than Sun-Tue concentrations. Regarding NO, during Wednesdays and Thursday the slops were clearly close to one indicating similar levels occurring during these two days to those occurring during Sun-Tue. On the other hand, NO levels were clearly much lower on Fridays than ones occurring during Sun-Tue ( Fig. 9; Table 2). Elevated NO x levels can always inhibit O 3 production either by NO titration of O 3 or by reaction of NO 2 and OH to form HNO 3 -nitric acid-(Riga-Karandinos and Saitanis, 2005). Synoptically, O 3 decreases more rapidly with reductions in NOx emissions, than with reductions in VOCs (Volatile Organic Compounds) in NO x -sensitive regions (ie. abundant VOCs relative to NO x ). While in VOCs-sensitive regions (ie. abundant NOx relative to VOCs) a decrease in NO x emissions may cause an increase in O 3 because of reduced titration or reduced formation of HNO 3 . Locations where ozone increases during weekends, while NO x emissions are significantly lower should be considered VOCs-limited (Blanchard and Tanenbaum, 2003). Thus, in the case of Al-Mansoriah residential area, signs of being a NO x sensitive region are very clear although NO x levels are not always higher on weekends, suggesting a strongly influential background effect.
The reduction of anthropogenic emissions during weekends is expected to reduce proportionally equally anthropogenic VOCs (AVOCs) and NO x . In addition to the above-discussed role of VOCs in the ''weekend effect'', it is the high early morning rush-hour traffic on weekends that allow ozone concentrations to rise from a significantly lower concentration and therefore reaching lower overall weekend concentrations.
A number of studies demonstrate a wide range in the weekday/weekend O 3 behavior. Marr and Harley (2002) suggest that NO x concentrations are consistently lower on weekends relative to weekdays by approximately 30 -40%, while VOC changes are less pronounced ( approximately 10%). The explanation of weekday/weekend variation, due to NO x reduction, is supported by a wide range of analyses of ambient data and several photochemical studies, while changes in the timing and location of emissions and meteorological factors play sma ller roles in weekend O 3 behavior (Heuss et al. 2003;Riga-Karandinos and Saitanis, 2005). Analyses of ambient air quality data and emissions forecasts for weekdays and weekends may improve considerably our understanding of the effects of control strategi es and of future changes in emissions on future ambient O 3 concentrations.

Conclusions
After analyzing the dataset obtained for Al-Mansoriah residential area, a number of exceedances were recorded, violating KUEPA permissible limits of air quality. Concerning the exceedances of O 3 , there were no exceedances violating annual or daily limits. However, inhabitants of Al-Mansoriah were exposed to acute SO 2 levels. Recorded NO 2 annual, daily and hourly exceedances in the five years of monitoring. The diurnal patterns of the studied pollutants were investigated and revealed a seasonal variation in ozone levels. An inverse pattern was witnessed for CO. On the other hand, methane patterns showed an increase in the early morning hours associated with the anthropogenic activates in the area.
Series of concentration roses were executed around the receptor point (ie. polyclinic) to determine the influential pollution sources around Al-Mansoriah. Methane and CO showed typical urban area patterns, while NO and NO 2 concentration roses revealed a primary source situated in northern wind corridor (covering 345-150). Weekend effect analysis revealed that the area under investigation was a NO x sensitive one. The causes of the weekend O 3 effect are probably the weekend/weekday differences in NO x emissions and the complex non-linear photochemistry of ozone. The data of this study would be useful for future comparisons, after the operation of the new infrastructures at Kuwait city (Sharq district), which certainly will cause changes in pollutants levels. It is obvious that, for future air quality management strategies, the development of accurate, temporally and spatially resolved dayof-week emission inventories, including separate inventories for weekends can promote a better understanding.