Characteristics of Ambient Particle-bound Polycyclic Aromatic Hydrocarbons in the Angkor Monument Area of Cambodia

The characteristics of ambient particle-bound polycyclic aromatic hydrocarbons (PAHs) in the Angkor monument area of Cambodia were studied to evaluate the status of air pollution and the influence of possible emission sources. Ambient particulates were sampled at four sites during the same seasons: Angkor monument area, including Angkor Wat, in Siem Reap city, along a main road, and on a small hill beside Lake Tonle Sap. Monitoring of ambient particles in the city and Angkor Wat were conducted on a continuous basis. Meteorological data were continuously monitored at two sites to evaluate the role of wind in the transport of pollutants. The concentrations of total suspended particulates (TSP), PAHs and heavy metals were analyzed to examine the influence of anthropogenic emissions, as well as contributions from soil and miscellaneous particles. Profiles of 15 PAH compounds were compared to determine the influence of wind direction on the seasonal characteristics of PAHs. TSP concentration proved to be proportional to the sum of the concentrations of Al, Ca and Fe, which were the major metals present in particles, regardless of location and season. The mass fraction of PAHs with 4-6 rings in Angkor Wat was almost equal to, or larger than, that found in the city. During the rainy season’s southwest monsoon, the PAH profile in Angkor Wat was similar to that found in the city, while it was similar to that found in the rural areas during the dry season’s north, or northeast, monsoon. This indicates that in the rainy season air pollutant transport from the city to the monument is an important factor. PAH concentration in Angkor Wat was similar to that measured in Bangkok, indicating a serious air pollution situation. * Corresponding author. Tel.: +81762344646, Fax: +81762344644 E-mail address: mfuru@t.kanazawa-u.ac.jp


INTRODUCITON
During the past 10 years, tourism in the Angkor monument areas of Cambodia has grown dramatically, while the populations in Siem Reap and the monument areas also have increased significantly (JICA, 2001).Tourist numbers have increased by a factor of about 100, to the point of almost one million a year (Statistics & Tourism Information Dept., Ministry of Tourism, Cambodia, 2005).According to the Authority for the Protection of the Site and the Management of Angkor and the Region of Siem Reap (APSARA), the population in the preserved monument areas (called Zones I and II) has increased from 20,000 (1992) to 100,000 (2002) and continues to grow (APSARA, 2006a).
Increases in the consumption of energy and water, traffic and the demand for waste material treatment associated with rapid tourism growth have also increased, and insufficient pollution control devices and facilities have compounded serious environmental impacts on the monument area.Air pollution is an unexpected factor that has not been recognized by the local population.
Because of increased energy consumption, such as power generation for electricity, burning fuels for cooking and lighting, traffic and waste incineration, the amount of air pollutants from these emission sources presumably have increased significantly.
Air pollution in this area may now be a serious issue that could have a severe influence, not only for human health, but also on biodiversity and the sustainability of the Angkor monuments.
In order to apply effective methodologies for reducing such environmental loads, it is essential to understand the status of air pollution and contributions from emission sources.However, little information is available, although the Department of Water and Forestry of APSARA has continued to monitor gaseous pollutants twice annually since 2004.This paper's authors have begun preliminary investigations, as well (Furuuchi et al., 2005a;Furuuchi et al., 2006a).
For this study, air pollution in the Angkor monument area was evaluated to examine the present status of air pollution and possible emission sources, focusing on particle-bound polycyclic aromatic hydrocarbons (PAHs), which are typical hazardous and carcinogenic pollutants emitted from various anthropogenic sources.Ambient particulates were continuously sampled using high-volume air samplers at different sites in Siem Reap and in the Angkor monument area.Meteorological data were monitored at two different sites to determine the influence of weather conditions, particularly focusing on the transport of pollutants by wind., 2006a).Zones I and II, which are important as areas needing preservation, make up more than 350 km 2 .Lake Tonle Sap, south of Siem Reap, is the largest lake in Southeast Asia and is famous for its unique characteristic of changing remarkably with the seasons (Tsukawaki et al., 2006).Four sampling sites were established: Angkor Wat (Site 1); Siem Reap (Site 2); Phnom Krom, a small hill (elevation ~140 m) beside Lake Tonle Sap (Site 3); and Route 6 (Site 4).Site 1 was located in an open space between the western corridor of Angkor Wat and the moat surrounding it, which features terrain typical for the area, such as forest, open space and small roads for tourism.Site 2 was located on the balcony of a room on the second floor of a two-story building ("Tap Rohm Hotel"), facing a partly paved city road.The distance to the road was about 2.5 m, and the wall height and width of the balcony were 0.6 and 1 m, respectively.Daytime traffic on the road was rather busy.This site was in a typical busy area of the central city near the "Old Market," prone to possible emission sources, such as traffic, cooking, heating and electricity generation.

Sampling and analysis Locations
Site 3 was located first on a hilltop (May, 2005), then moved to the east side of the hill (June and November, 2006).Site 3 characteristics were suitably representative of areas with fewer residents.These areas were covered by bush and low trees and were partly submerged under the lake from August through December.Depending on the availability of electricity, the different locations were used for Site 3 in 2005 and 2006, but both locations were very similar with few emission sources in the surrounding vicinity.Site 4 was located at the northern side of a temple "Preah Ang Chaik Preah Ang Chorm," at a distance of 5 m from Route 6, the busiest road in Siem Reap.Each sampling site was chosen as representative of the surrounding environment.Sites A and B were APSARA gas-sampling sites.

Sampling equipment and procedures
Total suspended particulates (TSP) were collected using high-volume air samplers (SHIBATA KAGAKU, HV-500F) on binder-less quartz-fiber filters (110 mm, ADVANTEC, QR-100) prepared in a desiccator at a constant temperature and humidity (room temperature and ~50% RH) for 72 hours.The present work is not only for the evaluation of the status of air pollution, but also as a first step in building a monitoring network for air pollution while transferring the monitoring methodology to APSARA coworkers.Therefore, the high-volume air samplers were used because of their simplicity of handling, which is practical and very important for the transfer of technology in developing countries (Kato, 2006).They were used also for their large flow rate, which is important for limited sampling durations due to problems with security and an available supply of electricity.Sized particles could provide more detailed information on air pollution (e.g., PM 10 and PM 2.5 samples), as well as nano-aerosol particles smaller than 0.1 µm (now being tested with results to be reported soon).At each sampling site, the sampler was installed on a chair at about a 50 cm sitting height, along with a temperature-hygro sensor and a data logger (SATO KEIRYOKI, SK-L200TH II), which recorded during the entire sampling period.
Simultaneous samplings of TSP were conducted in Sites 1, 2 and 3 (monument area) in May In order to minimize the degradation and evaporation of chemical compounds in particles due to the UV irradiation and temperature fluctuation, the samples were covered with aluminum foil immediately after sampling and kept in a freezer in the hotel during sampling activities.All filters were kept in a freezer (-20 o C) until they were analyzed after transport to Japan (about one day of travel time).Travel blank filters were handled in the same manner.

Analysis of chemicals
Filters were weighed after 72 hours of conditioning (room temperature ~25 o C, RH approximately 50%); then they were analyzed to obtain concentrations of PAHs and heavy metals.
Heavy metals were analyzed by ICP-AES (HITACHI, P-4010) after decomposing the samples with nitric, hydrofluoric and perchloric acids in a microwave digester (CEM, Mars X) following a previous procedure (Vandecasteele and Block, 1993).The recovery efficiency was confirmed as 0.97 ± 0.18 (n = 7), which is the average for all analyzed heavy metals, using standard materials (NIST SRM1648).Following the same procedure used in a previous paper (Toriba et al., 2003), acetonitril/ultra-pure water mobile phase after ultrasonically dissolving the samples on the filter in an ethanol/benzene (1:3) solution and evaporated on a rotary vacuum evaporator.The recovery efficiency was confirmed to be 0.82 ± 0.12 (n = 3) by adding a standard reagent [Accustandard 0.2 mg/mL in CH 2 Cl 2 :MeOH (1:1)] to dissolved samples (Tang et al., 2005).Travel blank values of heavy metals and PAHs were subtracted from analyzed values: Al (2.06 ± 0.2 µg/cm 2 ), Ca (1.36 ± 1.6 µg/cm 2 ), Fe (0.34 ± 0.2 µg/cm 2 ) (n = 3), 2-3 ring PAHs (8.5 ± 5 pg/cm 2 ), and 4-6 ring PAHs ( 5.5 ± 5 pg/cm 2 ) (n = 3).These blank values are less than the concentrations of each element and compound in all samples used in the following discussion.For example, in the Angkor Wat samples, the average travel blank concentration of PAHs (2-15 rings) was estimated as 0.10 ± 0.1% of the average particle-bound PAH concentration, which indicates that PAH blank values were reasonably small.Blank values of Al, Ca and Fe correspond to fractions of 11.1 ± 6, 0.18 ± 0.05 and 3.2 ± 1.4%, respectively, of the average value of Angkor Wat and Phnom Krom samples.Blank values may slightly reduce the data reliability, although they were constant.

Meteorological data
Data for meteorological conditions, such as wind direction and velocity, solar radiation, temperature, and humidity were measured at two monitoring sites in the Siem Reap and monument areas (shown in Fig. 1).Site W1, the monitoring site managed by the Agricultural Village Engineering Institute, Japan, was located near the northern cost of Lake Tonle Sap ("Chong Khneas").The sampling equipment was installed on a tower in the lake (13 o 13'N, 103 o 50'E) at two elevations (10 and 15 m) from the lake bottom (Tsujimoto, 2006)  Fig. 2 shows the seasonal changes in wind direction and monthly precipitation for the lake area (Site W1, 2005).Precipitation reached its peak during September and October in 2005.This pattern was similar to 2004(Tsujimoto, 2006)).The seasons in this area are classified customarily as either rainy (May-Oct.)or dry (Nov.-April)(Hori, 1996;Garami and Kertai, 1993).The authors use this terminology in the following discussion despite a recent (last 3-4 years) pattern of precipitation decreasing in May and increasing in November (Tsujimoto, 2006;Mekong River Commission (MRC), 2002;MRC, 2003).Generally speaking, the northeast winter monsoon from central Asia dominates in the dry season while the southwest summer monsoon dominates in the rainy season (Hori, 1996).However, Fig. 2 shows more-detailed characteristics: south-southwest-southeast and south-southwest winds are prominent January-April and May-August, respectively, while the frequency of a north wind increases from September, then becomes prominent October-December, as previously reported (Tsujimoto, 2006).Fig. 3 shows wind velocity in the monument area (Site W2) compared with that in the lake area (Site W1).Data are available only for June through September 2006 at Site W2 and from June 2003 through November 2006 at Site W1.The average wind velocity in the monument area was higher (1/4-1/5) than that in the lake area, while the direction was similar.The wind velocity difference at elevation may be explained roughly by the logarithmic velocity profile law (Pal Arya, 1999), taking into account the average water level during these months (~5 m) (Tsujimoto, 2006).

Meteorological characteristics
Taking into account the meteorological characteristics, there are two points of view concerning the locations of emission sources and wind transport of air pollutants.The city area (Sites 2 and 4), which is the prominent emission source, is located upwind from Zones I and II and from Angkor Wat (Site 1), where emissions come from sources such as tourism traffic, biomass burning for household use, and a number of small restaurants from May through August.
However, during some periods of the year, particularly October through December, the city is downwind of the monument area.In addition, Phnom Krom (Site 3), where very few emission sources exist, is located upwind of the city from May through August, but downwind October to December.

TSP and heavy metal concentrations
The monthly average for total suspended particulate (TSP) concentration at the four monitoring sites is shown in Fig. 4 where the number of samples for each month is different for each site.No value means no sample (see Table 1).There seems to be a weak correlation between TSP concentration and precipitation at Angkor Wat.However, this correlation is not so clear at other sites.For a more reliable discussion, more data needs to be gathered from all sampling sites, although logistical difficulties abound.

Concentration of particle-bound PAHs
Figs. 6 and 7 show the summed mass concentration and mass fraction of particle-bound PAHs, respectively, with 4-6 aromatic rings in different seasons at the four sampling sites.The concentrations of these compounds may be an index of anthropogenic emissions from fuel burning (Environmental Health Criteria 202, 1998;Seinfeld and Pandis, 1998;Finlayson-Pitts and Pitts, 1999;Spurny, 1999).PAH concentration was always greater in the city than in other sites, but the mass fraction of PAHs was lower, or similar, to the other sites.This is because of the larger contribution of soil or cement particles to the particle concentration, as shown in Fig. 5.
The seasonal variation in PAH concentration and mass fraction at Angkor Wat were not so clear, but values were larger than Phnom Krom in the rainy season (June 2006), where these values can be regarded as background, since the southwest monsoon wind is prominent and few emission sources are located nearby.In the dry season north/northeast monsoon (December 2005 and November 2006), they are still larger than the background in spite of the fact that Angkor Wat is located upstream from the city area.This suggests the presence of emission sources from nearby Angkor Wat, such as restaurants and illegal food stands, nearby parking, electric power generators, tourist road traffic, and fuel burning in small villages spreading over Zone I (APSARA, 2006b).Most inhabitants there use wood fuels for cooking and heating.No electricity or gas has been supplied to this area in order to reduce the number of people living there illegally.
The emission from these sources may increase during the tourism high season (Nov.-Feb.).There might be an influence of emissions from the city area on PAH concentration in Phnom Krom in November.compounds produced by biomass burning (Environmental Health Criteria 202, 1998;Furuuchi et al., 2006b;Tekasakul et al., 2006a), are definitely large in Phnom Krom, where bush burning may be a unique source of smoke; although the numbers are very limited and the amount is small.
In contrast, the fractions of BaP, BghiPe and IDP, typical compounds emitted from anthropogenic sources, such as diesel engines used in the traffic and power generation (Finlayson-Pitts and Pitts, 1999;Bernaudat et al., 2006), are of importance in the city and Angkor Wat.However, the compounds thought to be caused by biomass burning (Flu and Pyr) increased in the dry season (Nov.-Dec),both in the city and Angkor Wat.
During the rainy season southwest monsoon, particle-bound PAHs, the concentration of which was an increase in fine particles (Furuuchi et al., 2005b;Furuuchi et al., 2006a;Tekasakul et al., 2006a), may be transported from the city areas to Zones I and II in the north, increasing PAH concentration in Angkor Wat to a similar profile found in the city area.

Comparison with other Asian cities
Figs. 9(a) and 9(b) show the summed concentrations of 4-6 ring PAHs and the mass fractions of PAHs in particles, respectively, compared with those in other Asian cities in Cambodia (Furuuchi et al., 2005c;Murase et al., 2006), Thailand (Furuuchi et al., 2002;Kato et al., 2003;Tekasakul et al., 2006a;Furuuchi et al., 2006bFuruuchi et al., , 2006c) ) and Japan (Furuuchi et al., 2002).The total concentration of PAHs with 4-6 rings was about five times higher in Angkor Wat than in Kanazawa, Japan in May, and 2-3 times than in Tokyo.It was almost the same as in Bangkok, yet 1/8 of that found in Phnom Penh.The mass fraction of total PAHs in Angkor Wat (0.074 ng/µg) was similar to that found in Bangkok (0.070 ng/µg).

CONCLUSION
The TSP concentration was shown to be proportional to the summed concentrations of Al, Ca and Fe, which were the major metals in particles, regardless of location and season.The mass fraction of total concentration of PAHs with 4-6 rings in Angkor Wat was almost equal to or larger than that found in the Siem Reap city area.During the rainy season's southwest monsoon, the profile of PAHs in Angkor Wat was similar to that found in the city area, but during the dry season north/northeast monsoon, it had rural area characteristics.This indicates that during the southwest monsoon, the transport of air pollutants from the city is an important contributor to air pollution in the monument areas.
As shown by the fact that PAH concentration in Angkor Wat, a remote area, is similar to that in Bangkok city central, emissions from possible anthropogenic sources, such as traffic, generators and biomass burning, are contributing to a very serious situation.Similar threats from other air pollutants are suspected, although they are not discussed in the present study.Since this situation may be related to the explosion of tourism in a region devoid of any environmental protection policies, a much more serious situation will develop soon if nothing is done to control pollution.
The effective application of methodology for the reduction of such environmental loads, and to obtain data for each possible emission source, such as the number of emissions sources, and emission amounts, profiles and distributions in this area, it is important to continue monitoring and to increase statistical reliability.As shown for Phnom Penh, Cambodia (Murase et al., 2006), PAH concentrations during the nighttime are larger than during the day, probably owing to related differences in sources and emitted amounts.Such hourly observations should provide more detailed information on emission sources in Siem Reap, also.Particle size dependency of the pollutants is also important information since anthropogenic pollutants are enriched by smaller fractions (Spurny, 1999), and since particle size affects wind transportation.These studies will be the next step and must soon begin.

Fig. 1 .
Fig. 1. is a map of the Siem Reap and Angkor monument areas showing sampling site locations.

Fig. 1 .
Fig. 1.Map of Siem Reap and the Angkor monument areas showing the locations of sampling and weather monitoring sites.
2005, and in June and November 2006.Continuous monitoring for TSP samples occurred from March through November 2006 at Site 1, and at Site 4 every two weeks.Security considerations and the availability of electricity at Site 1 and 3 dictated that samplings be conducted there only during the daytime.Conditions for samples used for PAHs analysis are summarized in Table 1, which includes both conditions during preliminary surveys in 2005 by the authors, and continuous sampling started in March 2006 by APSARA personnel after their training on sampling methodologies.The sampling at Site 3 (2005) was accidentally stopped after a short period due to problems with electrical supply, thus only TSP values are discussed.Limited numbers of samples are due mostly to logistical difficulties, such as equipment shortages and security concerns.These problems cannot be easily solved in a developing country like Cambodia, but should improve with time.Nevertheless, even a broad discussion of the present status of air pollution is an important first step.

Fig. 2 .
Fig. 2. Meteorological characteristics in the Siem Reap region measured at a site in Lake Tonle Sap (Site W1) (2005).

Fig. 4 .
Fig. 4. Seasonal change in TSP concentration at each sampling location.

Fig. 5 .
Fig. 5. TSP concentration in relation to: (a) the summed concentration of major metal elements in particles (Al, Ca and Fe), and (b) the mass fraction of these elements.

Fig. 6 .Fig. 7 .
Fig. 6.Seasonal change in the summed mass concentration of PAHs with 4-6 aromatic rings at each sampling location.

Fig. 8 .
Fig. 8. Profiles for 15 PAH compounds on a mass fraction basis, respectively, for: (a) rainy season southwest monsoon, and (b) dry season north/northeast monsoon comparing Siem Reap, Angkor Wat and Phnom Krom.

Fig. 9 .
Fig. 9. Concentrations (a) and mass fraction (b) of PAHs with 4-6 rings in the city area (Site 2 and 4) and Angkor Wat (Site 1) compared with other Asian cities.

05 11:40 ～ 19/12/2005 16:45 305 fine 12 City Area 06/08/06 7:45 ～ 08/06/2006 16:20 515 fine 13 City Area 17/11/2006 7:00 ～ 17/11/2006 18:00 660 fine 14 Phnom Krom 27/05/2005 8:28 ～ 27/05/2005 9:05 21 fine 15 Phnom Krom 07/06/2006 7:45 ～ 07/06/2006 16:48 543 fine/partly cloudy-wind 16 Phnom Krom 17/11/2006 8:24 ～ 17/11/2006 17:55 571 Fine
Emission sources nearAngkor Wat contribute, but pollutants transported from the city northward may significantly influence air pollution in Zones I and II during this period.During the dry season north/northeast monsoon, however, the wind direction should not increase air pollutants in Zones I and II north of the city, since very few emission sources exist in the northern forest area.The PAH profiles obtained for Angkor Wat during the dry season are very similar to Phnom Krom during the wet season monsoon, indicating a contribution from wood burning.In the city, compounds both from fossil (BaP, BghiPe and IDP) and biomass (Flu, Pyr) fuels contain large fractions while the amount of biomass pollutants transported from Zones I and II to the city area may not be so important.This may be due to the fact that wood burning is extensive for heating during the dry season (Nov.-Dec.)where wood burning is the main energy source in Cambodia -96.1% of cooking fuel used in Siem Reap is wood (National Institute of Statistics, Ministry of Planning,