Review of Photochemical Smog Pollution in Jakarta Metropolitan, Indonesia

An analysis of photochemical smog pollution in Jakarta was attempted using data from the existing air quality monitoring stations. Ground-level ozone potential is high in Jakarta due to the high traffic emissions of ozone precursors and the favorable surface meteorological conditions. Despite the frequent missing data during the 1996-1999 monitoring, which resulted in lower ozone values, ozone episodes were significantly recorded in 1997-1998. The number of hours on which ozone concentrations exceed the 1-hour standard (100 ppb) at an ambient station located in Kelapa Gading (10 km northeast of city center) was 186 hrs in 1997 and 571 hrs in 1998. El Niño phenomenon in 1997-1998 had affected the local meteorology leading to more favorable conditions for photochemical production of ozone. The annual ozone averages in ambient stations located off the city center have exceeded the 1-year standard limit (15 ppb). Although the annual average and 95-percentile values indicated an increasing trend from 1996 to 1998, the trend remains to be seen in the future as more complete data could be expected from the new monitoring system. The number of hours on which ozone exceed the 1-hour standard and the annual average tend to be increasing since 2001 to 2002 in all 3 newly operated stations. The seasonal variations of ozone indicate that ozone level is highest in the dry season (September-November) and is lowest in the wet season (December-March). Correlation between ozone level and meteorological attributes (solar radiation, relative humidity and temperature) was significant at 0.01 confidence level. The diurnal cycle of ozone and its precursors is clearly shown and is typical for polluted urban areas. Improvement of the database of air quality monitoring is very critical for Jakarta. Through better database management, the development and monitoring of cost-effective air pollution control strategy can be made.


INTRODUCTION
Photochemical smog is a condition that develops when primary pollutants (oxides of nitrogen-NO x and volatile organic compounds-VOC created from burning of fossil fuels and biomass) interact in the presence of sunlight to produce a mixture of hazardous secondary pollutants [1]. The major constituent of photochemical smog is ground-level ozone (O 3 ), which is not emitted directly into the atmosphere but formed as the product of photochemical reactions of precursors, NO x and VOC [2]. Basic photochemical cycle of NO, NO 2 and O 3 involves the following reactions: when NO and NO 2 are present in sunlight, ozone formation occurs as a result of the photolysis of NO 2 Development of photochemical smog is typically associated with specific climatic conditions. Local meteorological factors that influence the formation of photochemical smog include solar radiation, cloud cover, temperature and precipitation. Other factors such as wind speed, mixing height and topography determine the dispersion of O 3 in the boundary layer. High temperatures are frequently associated with high pressure, stagnant conditions that lead to suppressed vertical mixing and increased O 3 levels. The database of ground-level ozone observations for urban and suburban areas indicates that at most rural sites, ozone concentrations have been found to vary over a diurnal cycle with a minimum in the early morning hours before dawn and a maximum in the late afternoon [2]. This pattern results from daytime photochemical production or intrusion of ozone from upper layer to the boundary layer, combined with ozone loss by dry deposition and reaction with nitric acid (NO) at night, when photochemical production ceases and vertical transport is inhibited by a temperature inversion. In urban areas the nocturnal minimum in ozone can be quite evident because of the rapid reaction between O 3 and NO. This study reviews the status of photochemical smog pollution in Jakarta Metropolitan during a period of 7 years, 1996-2002 with focus on ozone and NO x pollutants. Average monthly and diurnal concentrations of ozone and its precursor are also evaluated in relation to variations in meteorological conditions.

STUDY AREA AND DATA COLLECTION
Jakarta is the capital city of Indonesia located in the northern part of Java Island, at 106° East and 6° South. The area of Jakarta is approximated at 664 square kilometers with flat topography, close to the shore and has an average elevation of 7 meter above sea level. Jakarta is part of the greater Jabodetabek (Jakarta, Bogor, Depok, Tangerang and Bekasi); bordering with Tangerang in the west, with Bogor and Depok in the south and Bekasi in the east. The total population of Jakarta was estimated at 9 million in 2002, whereas total population of Jabodetabek exceeded 22 million people in 2002. In the meantime, total commuters from Bodetabek to Jakarta today exceed 3 million everyday. Climate in Indonesia is largely influenced by monsoon systems, both Asia and Australian winter monsoons. During wet months (October-March), wind direction extends from west to southwest transporting moist air from Asia continent; while during dry months (April-September) wind direction is from north up to northeast which transports dry air from southern hemisphere. Besides the monsoon systems, the climate is also affected by global circulation occurring on a time scale of several years known as El Niño/Southern Oscillation (ENSO). In 1997, extreme meteorological values were recorded in Indonesia and the region as a result of ENSO phenomenon. The climate in Jakarta is very hot and humid. Because Jakarta lies so close to the equator, the solar heating during the day and the earth cooling during the night may produce local land-sea breeze. The winds are generally weak. Calm conditions often prevail at night. In the afternoon, winds are quite moderate because of the effects of local winds that modify the monsoon winds. Jakarta has experienced serious air pollution problem, which is largely contributed from traffic sources accounting for approximately 70% of the total emissions. The vehicular emissions in Jakarta contributed 69% of total NOx emissions, whereas CO and HC emissions from vehicular sources amount to 564,000 ton and 98,000 ton per year respectively [3]. The data used for this study were collected from the Local Environmental Management Agency (BPLHD) of DKI Jakarta and the National Environmental Management Center (EMC). During a period of 1996-1999, data was supplied from an air quality monitoring network which comprises of 6 stations that monitor NO x , CO, PM 10 , NMHC, THC, O 3 and SO 2 parameters using designated absorption methods. In 2000, the prevailing stations were overhauled and Meteorological data was compiled from 2 major stations, namely Cengkareng International Airport and Curug Airstrip. In addition, In Situ meteorological data is also recorded at air quality monitoring  (1) Serpong, southwest of Jakarta. No nearby specific emission sources. Pulogadung (2) East Jakarta, industrial park. Pluit (3) North Jakarta, residential area. Thamrin (4) Central Jakarta, at curbside near Thamrin roundabout. Casablanca (5) Central Jakarta, commercial and residential area. Gambir (6) Central Jakarta, at curbside. Kelapa Gading (7) North Jakarta, residential area. Pulogadung* (8) East Jakarta, industrial park. Kebon Jeruk* (9) West Jakarta, commercial and residential area. Senayan* (10) South Jakarta, commercial area Kemayoran* (11) Central Jakarta, commercial area Pondok Indah* (12) South Jakarta, commercial and residential area. BPLHD* (13) South/central Jakarta, commercial and residential area *New monitoring stations (2000-present) Source: BPLHD DKI Jakarta, 2001 stations, which are equipped with basic meteorological sensors at 10 m height above the ground. Table 3 through 6 present a summary of ozone and NO 2 pollution levels for Jakarta during the four-year monitoring period from 1996 to 1999 and the two-year period from 2001 to 2002. In general, the highest ozone was found at ambient station number 7 (Kelapa Gading) located approximately 10 km northeast of the city center. The highest hourly maximum level was recorded at 964 ppb at this station. Although this figure looks highly suspicious as if someone was doing a range check on the ozone instrument, it was noted that the total number of hours on which ozone exceed the standard were quite high reaching 186 hrs and 570 hrs in 1997 and 1998 respectively. Station 7 also held the highest average hours per year of ozone exceeding the standard, followed with station 2 and 3. Where ozone data are available in 1996-1999, 1998 was characterized by the highest ozone values (average and 95 percentile) at station 2. At stations 3 and 7, the highest ozone values were recorded in 1997. The discussion on the relation between ozone episodes and the El Niño phenomenon during 1997-1998 is discussed later in the next section. Lower ozone was detected at curbside station 4 and ambient station 5 of which these stations are located closest to the city center. The major factor that caused low ozone values at these stations was due to destruction of ozone by NO. Since most of the NO x is emitted in the form of NO, it can be expected that with high emission of NO from traffic in curbside station, ozone level becomes lower (reaction 3). The processes of destruction and formation of ozone in a large urban area such as Jakarta should be competing at any location. Over the city, ozone level is found to be relatively high except at the very heavy traffic city center and curbside where the ozone destruction by NO is very apparent. However, it is very difficult to make further assessment as to the transport of ozone and its precursors over the larger area of Jakarta (beyond the 15 km perimeter of the city center) because all stations are located within. Frequent missing data and limitation of the current two-year monitoring (2001-2002) make the assessment of ozone annual trend very premature. The initial ozone trend averaged for 3 ambient stations (station 2, 3 and 7) indicated an increase of ozone values from 1996 to 1999; but a substantial drop was noted from 1998 to 1999 (Fig. 3). The substantial drop of ozone level was significantly contributed from station 7. Thus, besides the low percentage of data completeness, other factors such as the change in meteorological conditions and in emission of O 3 precursors in the upwind area of station 7 may also play a role in causing the shift of ozone annual trend. In addition, the annual ozone averages in ambient stations located off the city center (station 2, 3 and 7) indicated exceeding the 1-year standard limit (15 ppb). Ozone fluctuation in terms of standard deviation was quite high showing variations in daily and monthly time scales.  Station  1996  1997  1998  1999  2001  2002  1996  1997  1998  1999  2001  2002  Pulogadung (2) a  11  2  268 18 (7)     The assessment of ozone level in the background area (station 1) was not done due to unavailability of ozone data. It can be expected that O 3 concentration in this background area is fairly high. Rural environments tend to be characterized by relatively high VOC-to-NO x ratios because of rapid removal of NO x from distant sources compared to that of VOC [2]. In general, increasing VOC concentrations means more ozone. Transport of ozone from a very photochemically active source region (Jakarta urban center) is also a possible candidate of potential source of ozone enhancement in the background area. Ozone behavior near the surface may be attributed to local meteorological conditions [4]. During the period of 1997-1999, daily mean temperature was increasing   (2)   Year  of high ozone levels Fig. 4 with warm temperatures, low humidity and high solar radiation. The local land-sea breeze may potentially influence ozone behavior in the area. Analysis of correlation using the two-year database at station 8 indicated that the correlation between ozone and solar radiation, between ozone and temperature and between ozone and humidity was all significant at the 0.01 level (two-tailed student t test) with the correlation coefficient (r) of 0.64, 0.67 and 0.63, respectively. Surface meteorological observation indicates that the southerly to NE wind predominates over the period of May-October (dry season) transporting the ozoneenriched air masses towards neighboring Southeast Asian countries. Zhang et al. related the high ozone in the dry season in Bangkok to, among others, the regional transport of O 3 and precursors from the Asian continent associated with the NE monsoon.
Figure 8-13 presents the diurnal variation of pollutants during 1-year monitoring in 1996. The diurnal cycle of pollutants is typically shown for polluted urban areas. Indeed, this pattern is almost similar to the diurnal pattern of photochemical pollutants in Bangkok [5].
When the morning rush hours begin at around 7.00 a.m. to 8.00 a.m.; CO, NO, NMHC and CH 4 also reaches its peak at around 8.00 a.m. due to emissions from traffic. NO peaks at around 7.00 a.m., while NO 2 peaks at around 9.00 a.m. (about two hours after NO). NO concentration drops fast after 8.00 a.m. (after sunrise) as O 3 increases reaching its highest at around 11.00-13.00. CO, NMHC, CH 4 and NO 2 also drop when O 3 reaches its maximum during the afternoon hours. The reduction of these pollutant concentrations should be caused by the photochemical reactions that consume the pollutants such as reactions of pollutants with OH and HO 2 radicals and the pollutant dispersion factors such as mixing height and wind speed. As the boundary layer increases gradually due to convective heating during noon hours, the mixing of air in the lower heights with air in the higher heights causes pollutants to be ventilated upwards [5]. The local wind may also contribute to the pollutant dilution.
In the afternoon towards evening, the pollutant concentrations increase steadily as the traffic peaks again but the photochemical production of O 3 ceases due to a minimum-to-zero solar radiation. As the traffic subsides, the concentrations of NO, NO 2 , NMHC and CO decrease slowly during night hours until before dawn. Interestingly, CH 4 continues to increase slightly in the early hours, which may be contributed from other sources than road transport, i.e. natural, fossil-fuel related and biospheric carbon sources. In urban centers, when NO x predominates over VOC, the OH-NO 2 reaction will remove OH radicals retarding the further O 3 production. Because OH reacts more rapidly with NO 2 than with VOCs, NO x tends to be removed faster than VOCs. In the absence of fresh NO x emissions, NO x is depleted more rapidly leading to increasing ratio of VOC to NO x . As a result, OH reacts preferentially with VOCs to keep the ozone-forming  This shall explain why higher pollutant concentrations are found in station 2 as compared to station 3, besides the fact that station 3 is located near to the coastal area. Daily average value of pollutants is generally higher in dry months. Figure 14 exhibits

CONCLUSION
Photochemical smog pollution potential in Jakarta metropolitan area is high due to the high traffic emissions of ozone precursors and the favorable surface meteorological conditions, i.e. warm temperature, high solar radiation and calm wind conditions. Over the two-year period from 2001 to 2002, the ground-level ozone indicates an upward trend. The trend remains to be seen in the future as more data is made available. Despite the frequent missing database, which produces uncertainty in analysis, ozone episodes are potentially detected at downwind area off the city center. Further investigation is needed with regard to the transport of ozone and its precursors over the larger area of Jakarta and Bodetabek. The seasonal variations of ozone illustrate the effect of local meteorological conditions and regional monsoons on ozone formation and accumulation. Ozone level is highest in the dry season (September-November) and is lowest in the wet season (December-March). Correlation between ozone level and meteorological attributes (solar radiation, relative humidity and temperature) is significant. The lower ozone level at curbside stations reflects the destruction of ozone by NO, which is emitted from traffic sources. The diurnal variations of ozone and its precursors are clearly demonstrated and are typical for polluted urban areas. Improvement of the database of air quality monitoring is very critical for Jakarta. Areas of improvement shall include the development of national guidelines on the standard procedures of air quality monitoring including site location, instrumentation, quality assurance and statistical procedures for data analysis.
The local government implementation plans on collection and analysis of data and inventorying emissions from pollution sources is a key to the identification and selection of emissions control measures. This will lead to development of efficient and cost-effective control strategy.