Source Apportionment and Risk Assessment of Atmospheric Polycyclic Aromatic Hydrocarbons in Lhasa , Tibet , China

Much attention has been given to the distributions, sources, and health risks of atmospheric polycyclic aromatic hydrocarbons (PAHs) in cities. In this study, a total of 62 suspended particle samples were collected from April 2013 till March 2014 in the city of Lhasa. Positive matrix factorization (PMF) was applied to investigate the source apportionment of 15 priority PAHs, the lifetime carcinogenic risk (LCR) levels of which were assessed. The average annual particle phase PAH concentration was 43.9 ± 60.4 ng m. Evident seasonal variations of PAHs were observed, with the highest concentration observed in winter, followed by autumn, spring, and summer. Fourand five-ring PAHs accounted for the predominant proportion (63.3%–84.4%) throughout the year. Correspondingly, gas phase PAHs showed the opposite variations, with the highest and lowest concentrations observed in summer and winter, respectively; also, three-ring PAHs, especially Ace, Acel, and Flu, were the largest contributors. Compositions of particle phase PAHs varied seasonally, with four-ring PAHs contributing more in winter than in summer and five-ring PAHs exhibiting the opposite trend, thereby reflecting the variety of emission sources. PMF analysis showed that biomass combustion (48.4%) and vehicle emissions (27.9%) were the two main sources, followed by coal combustion and the air–surface exchange. These results were consistent with the diagnostic molecular ratios. The benzo(a)pyrene equivalent (BaPeq) concentration of particle phase PAHs ranged from 1.48 to 24.5 ng m, which exceeds were higher than the new limit in China (1 ng m). The average BaPeq of gas phase PAHs was 6.43 ± 4.15 ng m, which was similar to that of particle phase PAHs. The LCR of the total PAHs (9.08 × 10) was one time higher than that of the particle phase; however, it was slightly lower than the acceptable level, thereby indicating that atmospheric PAHs in Lhasa pose little or no carcinogenic risk to the local population.


INTRODUCTION
Polycyclic aromatic hydrocarbons (PAHs) compose a large group of organic compounds that consist of fused aromatic rings.They are emitted mainly from the incomplete combustion or pyrolysis of organic materials such as coal, oil, gas, waste, and biomass (Ravindra et al., 2008;Zhang and Tao, 2009).PAHs are the subject of considerable concern economic pillars of the TP, whereas industry accounts for only 8% of the gross domestic product of this region (http://www.tibet.stats.gov.cn/).Therefore, the TP is considered a background site on a global scale.Numerous recent studies have reported atmospheric pollutants in this remote region (Cong et al., 2015a;Lüthi et al., 2015;Li et al., 2016); however, such studies have primarily been concerned with the role of the long-range transportation of pollutants emitted from South Asia.Notably, local anthropogenic activity in the TP contributes significantly to the atmospheric environment (Gong et al., 2011;Li et al., 2012;Chen et al., 2015a;Li et al., 2016a).
Lhasa, the provincial capital of Tibet, is currently undergoing rapid urbanization and industrialization (Li and Wang, 2014;Ran et al., 2014;Li et al., 2016b).Previous studies have reported that the atmosphere in this high altitude city has been heavily influenced by local emissions (Huang et al., 2010;Cong et al., 2011;Li et al., 2016b).In addition, as the largest city in the TP, emissions from Lhasa are potential pollution sources to the surrounding areas (Li et al., 2008;Tao et al., 2010); for example, higher elemental and organochlorine pesticide concentrations were reported in aerosol samples from Lhasa than in those from more remote sites in the TP (Li et al., 2008;Cong et al., 2011).Current use and local emissions of organochlorine pesticides may contribute to environmental contamination in the populated agricultural Lhasa River Basin (Li et al., 2008).Radiocarbon isotope measurements of total carbon emissions revealed biomass burning and the incineration of agricultural waste, which contributed more to carbonaceous aerosols in winter than summer (Huang et al., 2010).Fine particulate matter (PM 2.5 ) in Lhasa is characterized by its low organic carbon to elemental carbon ratio, which reflects the heavy influence of vehicle emissions (Li et al., 2016b).Regarding PAHs, two studies have reported atmospheric PAH concentrations and compositions in Lhasa (Gong et al., 2011;Liu et al., 2013).Diagnostic ratios and principal component analysis revealed that atmospheric PAHs in Lhasa were primarily derived from local human activities such as vehicle emissions and incense burning.However, current knowledge of PAH source apportionment and its effects on the health of locals in Lhasa remains limited.
This paper presents the results of particulate-phase PAH observations in Lhasa during 2013 to 2014.The main objectives were as follows: (1) to investigate the atmospheric PAH concentrations and compositions in Lhasa; (2) to identify the seasonal variations of PAH concentrations and apportion the PAH sources by using a positive matrix factorization (PMF) model; and (3) to assess the health risks experienced due to PAHs in Lhasa.

Sample Collection
A total suspended particle (TSP) filter sampler (flow rate: 100 L min -1 ; KC-120H, Qingdao Laoshan Applied Technology Institute, Qingdao, China) was placed on the rooftop (20 m above ground) of the Institute of Tibetan Plateau Research, Lhasa (E29.65°,N91.03°, 3642 m).A total of 62 samples were collected on prebaked quartz fiber filters (diameter: 90 mm; Whatman plc, Maidstone, United Kingdom) from April 2013 to March 2014 with duration of each sample as 24 h.TSP filter samples were collected continuously every week.For example, we collected 3 samples each week in spring but 1 sample in other seasons.However, several samples could not be collected due to lack of power or equipment breakdown in some periods.Finally,19,17,11,and 15 samples were collected in spring, summer, autumn, and winter, respectively.All filters were prebaked at 550°C for 6 h before sampling.They were equilibrated at constant temperature and humidity (25 ± 3°C, 30 ± 5%) for 72 h and weighed using a microbalance with a sensitivity of ± 0.01 mg before and after sampling.The volume of air passing through each filter was converted into standard atmospheric conditions (25°C, 101.3 kPa).Ten field blank filters were collected once each month by placing in the sampler, which had no air drawn through it.And 3, 3, 2, and 2 blank filters were collected in spring, summer, autumn, and winter, respectively.Lhasa is a famous historic tourist city where considerable seasonal variations in traffic and religious activities occur.It exhibits the typical characteristics of four seasons, namely spring (March-May), summer (June-September), autumn (October-November), and winter (December-February).Meteorological parameters of studying period were recorded with automatic observation instruments at Lhasa Station.According to the observation, the air temperature ranged from -5.7 to 21.2°C with an average of 9.0°C, and the relative humidity ranged from 7 to 75% with an average of 38.7% (Wan et al., 2016).The annual mean precipitation amount is around 400 mm with the majority of precipitation occurred frequently between June and September because of the summer monsoon activities, but it was relatively dry during spring and winter.The largest coal-fired power plant of Tibet (Dongga power plant) and the Dongga cement factory were located about 10 km west of our sampling site.In addition, our sampling site is close to Jinzhu Road which is one of the busiest roads in Lhasa city with large numbers of trucks and other vehicles running.These are potential sources of particles in the atmosphere (Huang et al., 2013).

Extraction and Analysis
Sonication extraction was used as detailed in Chen et al. (2015b).Here, the method is described briefly.A quarter of each filter was cut into pieces, placed into a glass tube, and immersed in 20 mL of dichloromethane (DCM) and nhexane (1:1).The extraction was performed by sonication twice at 27°C for 30 min.Every single sample was spiked with deuterated PAHs (naphthalene-d8, acenaphthene-d10, phenanthrene-d10, chrysene-d12, and perylene-d12) as recovery surrogates.The extracts were evaporated to about 0.5 mL with a rotary evaporator, and transferred to a multilayer column filled with 2 g of activated silica gel, 4 g of neutral alumina, and 1 cm of anhydrous Na 2 SO 4 (presoaked in n-hexane).Then, the column was eluted by a mixture of 10 mL of n-hexane and 20 mL of DCM/nhexane (1:1).The eluent solvent was blown down to a final volume of 1 mL under a gentle stream of nitrogen.Finally, the solution was transferred to a 1.5-mL vial and stored at -20°C for rejection.
Sixteen PAHs (naphthalene (Nap), aecnaphthene (Ace), acenaphthylene (Aecl), anthracene (Ant), fluorene (Flu), phenanthrene (Phe), benzo(a)anthracene (BaA), chrysene (Chr), fluoranthene (Fla), pyrene (Pyr), benzo(a)pyrene (BaP), benzo(b)fluoranthene (BbF), benzo(k)fluoranthene (BkF), dibenzo(a,h)anthracene (DahA), benzo(g,h,i)perylene (BghiP) and indeno(1,2,3-cd)pyrene (IndP)) prioritized by the United States Environmental Protection Agency (US EPA) were analyzed at the State Key Laboratory of Cryospheric Sciences, Chinese Academy of Sciences Northwest Institute of Eco-Environment and Resources, China, by using gas chromatography-mass spectrometry with a 30 × 250-µm ID HP-5MS.High-purity helium was used as a carrier gas at a constant flow rate of 1.0 mL min -1 .The mass spectrometer was operated in 70-Ev electron impact mode.The oven temperature was 100°C, which was held stable for 2 min.Then it was ramped to the final temperature of 260°C with different rate of increase; to 170°C at 25 °C min -1 , to 225°C at 8 °C min -1 , to 235°C at 0.7 °C min -1 , to 260°C at 25 °C min -1 , and finally held at 260°C for 2 min.The temperature of the injector was 250°C and that of the transfer line was 280°C (Chen et al., 2017).Nap was not analyzed as it was detected with high concentration in the laboratory and field blanks.

Quality Control
All analytic procedures were carried out using same method with strict quality assurance and control measures with Chen et al. (2017).Laboratory and field blanks were extracted and analyzed in the same way as the samples.The recoveries in field samples were 74-93%, 80-97%, 83-105%, and 89-109% for acenaphthene-d10, phenanthrene-d10, chrysene-d12, and perylene-d12 as inferred standards, respectively.The PAH concentrations were not corrected for the recoveries.

Gas/Particle Partitioning Estimation
Octanol-air partition coefficient (K OA )-based model has been proved applicable to estimate the gas-particle partitioning of PAHs (Cheruiyot et al., 2015;Cheruiyot et al., 2016;Wang et al., 2018).The temperature-dependent K OA values can be measured directly using the GC retention time method with an equation as Eq. ( 1) (Harner and Bidleman, 1996): The regression parameters (A and B) were given by Harner and Bidleman (1998) and Odabasi et al. (2006).The difference of K OA values of PAHs among four seasons were calculated by adjusting the equations to the average  Odabasi et al. (2006).ambient temperature (pre-monsoon: 5°C, monsoon: 16°C, post-monsoon: 14°C, and winter: 1.5°C) (Table 1).Then the gas-particle partition coefficient (K P ) can be predicted by Eq. ( 2) if the organic matter fraction fom of the aerosols are known and assuming that all of the particle organic matter is available to absorb gas phase compounds.log K P = log K OA + log fom -11.91 (2) In this study, the average organic carbon concentrations were 13.2, 10.2, 30.6, and 35.9 µg m -3 for these four seasons in aerosols of sampling site.And organic compounds contributed 18.3%, 16.9%, 25.7%, and 23.9% to total suspended particles in four seasons.Finally, we can estimate the gas phase PAHs using Eq.(3): where C p and C g are the PAH concentrations in the particulate and gas phases, respectively, and C TSP is the concentration of TSP in the air.

PMF Receptor Model
PMF is a powerful factorization method used to calculate the source profiles and contributions of pollutants in the environment.It has been used extensively for atmospheric source identification (Paatero and Tapper, 1994;Dvorská et al., 2012;Callén et al., 2014).PMF analysis is described in detail in Chen et al. (2016).A 62 × 15 (62 samples each containing 15 PAHs) dataset was input into the PMF 5.0 model to conduct source apportionment of atmospheric PAHs in Lhasa.A random seed mode and random starting point were selected.After testing three to six factors, a four-factor solution was adopted.Correlation indices between the estimated and measured concentrations ranged from 0.75 (Ace) to 0.92 (IcdP and BghiP), which suggested that the measured concentrations were fully explained by the four selected factors.

Cancer Risk Assessment
The carcinogenic potency of PAH exposure can be estimated as the sum of each individual BaP equivalent (BaP eq ) based on the toxic equivalency factor (TEF) of each individual PAH including both particle and gas phase PAHs (Nisbet and LaGoy, 1992;Petry et al., 1996;US EPA 2010).The TEF data in Nisbet and LaGoy (1999) were used in this study; BaP eq was calculated as follows: where C i is the concentration of an individual PAH and TEF i is its TEF.Inhalation was determined to be the main exposure pathway for PAHs in the air.We applied the following inhalation exposure assessment model recommended by the US EPA to calculate the daily exposure dose (DED) of BaP eq : where DED is the DED of BaP eq through the inhalation pathway (mg (kg day) -1 ), C is the airborne BaP eq concentration (ng m -3 ), IR is the air inhalation rate (m 3 day -1 ), EF is the exposure frequency (day year -1 ), ED is the exposure duration (year), BW is the body weight (kg), AT is the averaging time (day), and cf is the conversion factor (in this study, cf = 10 -6 ).
The carcinogenic risk (CR) associated with inhalation exposure was calculated using Eq. ( 6), adapted from the US EPA (1989) and expressed as follows: where CR is the probability of developing cancer over a lifetime because of exposure to atmospheric PAHs and CSF is the cancer slope factor, which quantitatively defines the relationship between carcinogen exposure dose and degree of CR.The CSFs used were derived assuming a body weight of 70 kg, which deviates from the actual conditions of the exposure populations of various age groups.Thus, the CSF values were extrapolated to actual body weights in various age groups by multiplying the conversion factor (BW/70) 1/3 , recommended by the US EPA (2004).

PAH Concentrations and Compositions
During the study period, the particle phase PAH concentrations varied between 9.18 and 211 ng m -3 with an average of 45.0 ± 60.4 ng m -3 (Table 2).Although the annual average PAH concentration was considerably lower than those reported in numerous areas of northern China such as Shanxi, Shandong, and Beijing (Zhang et al., 2016) and South Asia such as Kathmandu (Chen et al., 2015) and Delhi (Sarkar and Khillare, 2013), it was higher than those observed at background sites in the TP.For example, PAHs in Lulang on the southeastern TP ranged from 0.06 to 2.53, with a mean value of 0.59 ng m -3 (Chen et al., 2014).In Zhongba and Nyalam, the mean PAH concentrations were 8.78 and 5.60 ng m -3 , respectively (Wang et al., 2014;Chen et al., 2017;).This indicated that although the atmosphere in Lhasa is relatively clean compared to other Asian cities, it is affected by local anthropogenic activities.In addition, the PAH concentration had evidently increased since the two studies conducted in Lhasa in 2006 and 2007 (Table 2), implying that emissions might have increased as a result of urbanization (Gong et al., 2011;Liu et al., 2013).
Evident seasonal variations in the total PAH concentrations were observed, with the maximum and minimum concentrations occurring in winter and summer, respectively (Fig. 2).Based on the Octanol-air partition coefficient (K OA ) model, the gas phase PAH concentrations were calculated, which also demonstrated clear seasonal variation with the highest concentrations occurred in summer, decreasing to minimum concentrations in winter (Table 3).Previous studies have reported that seasonal variations of some atmospheric pollutants in Lhasa were mainly influenced by vehicle emissions in summer and biomass burning, power plants, and religious activities during the other seasons (Huang et al., 2010;Gong et al., 2011).During the cold season, biomass or coal combustion for heating increased as the temperature drops (Huang et al., 2010).And the incense burning in the temples for religious activities might be another important emission source of PAHs (Liu et al., 2013).Furthermore, low temperature (-7 to 9°C) strengthens the condensation of gas phase PAHs onto atmospheric particles, leading to higher concentrations of particle phase PAHs in winter (Gong et al., 2011).A greater abundance of PAHs from late autumn to winter implied that PAH contributions from anthropogenic activities increase in the cold season, whereas in summer, although vehicle use and the frequency of tourist activities increase, high rainfall (approximately 70% of the total annual precipitation) washes most suspended particles and particle-   bound pollutants out of the atmosphere (Cong et al., 2011;Wan et al., 2016).In addition, the high summer temperature (9-22°C) enhances the evaporation of particle phase to gas phase PAHs, which leads to lower PAH concentrations in summer (Tham et al., 2008;Liu et al., 2017b); for example, the contribution of PAHs with low molecular weights (e.g., Ace, Acel, Ant, Flu, and Phe) was twice as high in winter as in summer, indicating that temperature is a crucial factor possibly resulting in seasonal variations between PAH concentrations.
According to the number of aromatic rings, the aforementioned 15 PAHs were classified into four groups: three-ring (Acel, Ace, Flu, Phe, and Ant), four-ring (Fla, Pry, BaA, and Chr), five-ring (BbF, BkF, BaP, and DahA), and six-ring (IndP and BghiP) PAHs (Kaur et al., 2013).The seasonal composition patterns of the particle phase 15 PAHs over four seasons are shown in Fig. 3; four-and five-ring PAHs contributed the most, followed by six-and three-ring PAHs.Correspondingly, gas phase PAHs showed different composition patterns with three-ring compounds especially Aec, Acel, and Flu contributed most to the total particles (Table 3).However, the profiles of particle phase PAHs in Lhasa varied across all four seasons; for example, the contributions of four-ring PAHs were highest in winter at 56.5%, which decreased to 31.5% in summer.By contrast, the contributions of five-and six-ring PAHs exhibited the opposing trend.Previous studies have reported that coal combustion and biomass burning release an abundance of four-ring PAHs, whereas five-and six-ring PAHs mainly originate from high-temperature combustion processes such as vehicle exhaust (Dachs et al., 2000;Moon et al., 2008).The variations in PAH compositions evidently reflect the aforementioned changes in source contributions during the sampling period (Huang et al., 2010;Gong et al., 2011).Meanwhile, the partitioning of PAHs between the particleand gas phase is temperature sensitive and consequently fourring compounds making an apparently greater contribution in winter as opposed to summer.

PAH Sources Assessed By Diagnostic Ratios
Diagnostic ratios are used frequently to identify PAH origins (Yunker et al., 2002;Rajput et al., 2014).In this study, IndP/(IndP + BghiP) and Fla/(Fla + Pyr) ratios were used simultaneously to cross-check the results and reduce uncertainty.The IndP/(IndP + BghiP) ratios were below 0.5 in all seasons, implying a strong contribution from petroleum combustion (Table 4).The mean Fla/(Fla + Pyr) ratios were 0.54 ± 0.03, 0.48 ± 0.02, 0.49 ± 0.02, and 0.52 ± 0.03 in spring, summer, autumn, and winter, respectively, with lower values in summer and autumn, thereby reflecting the impact of vehicles as PAH sources (Table 4).The tourist activities mainly happen from May to October especially in summer.And vehicle emission is higher during this period than the cold season.Higher values were observed in winter and spring, implying that the biomass contribution increased in these seasons because of low temperatures.Biomass energy consumption was considered as the largest parts (about 40%) of total energy consumption, compared to coal and liquid fossil fuel, 5% and 16%, respectively (Hua, 2009).These results are consistent with those of the PMF model in Subsection 3.3.

Source Apportionment of All PAHs Determined Using PMF
Fig. 4 shows the four source contributions of all samples and PMF factors of all profiles and contributions.Each column corresponds to the concentration profile of one PAH.The first factor accounted for 9.30% of all PAHs.The profile contains more volatile PAHs, which are similar to those reported as air-ground and air-soil emissions (Nelson et al., 1998;Cheng et al., 2012).Accordingly, the first factor was classified as air-surface exchange.The second factor accounted for 14.3% of all PAHs and showed high loadings of Fla and Pyr, which are typical markers of coal combustion (Huang et al., 2010;Qin et al., 2014).Thus, this factor was classified as coal emission.The third factor accounted for 48.4% of all PAHs and had high loadings of four-ring PAHs such as Fla, Pyr, and Chr and moderate contributions from BaA, Bbf, Bkf, and Bap.This type of profile is mainly the result of biomass burning (Lin et al., 2011;Wang et al., 2015).In Lhasa, biomass varieties such as wood and yak dung are burned year-round for heating and cooking, especially in winter (Huang et al., 2010).Thus, this factor was assigned as biomass burning.The fourth factor  (Grimmer et al., 1983;Yunker et al., 2002) Fla/(Fla + Pyr) 0.54 (0.03) 0.48 (0.02) 0.49 (0.02) 0.52 (0.03) < 0.4 Petrogenic 0.4-0.5 Petroleum combustion > 0.5 Coal and biomass (Sicre et al., 1987;Tsapakis and Stephanou, 2003;Tang et al., 2005;Bari et al., 2011) Fig. 4. PMF 5.0-generated factor profiles and their contributions to the 15 examined PAHs.contributed 27.9% of all PAHs.High loadings of five-and six-ring PAHs were observed.IndP and BghiP are typical markers of traffic emissions (Simcik et al., 1999) and IndP is associated with diesel emissions (Li and Kamens, 1993).A similar profile was observed in aerosols at a site downwind of East Asia (Wang et al., 2014).As a famous tourist city, Lhasa has undergone a rapid increase in vehicle use in recent years, especially in summer and autumn.Therefore, this factor was deemed to be caused by vehicle emissions.
In summary, the main source of atmospheric PAHs in Lhasa is biomass combustion, followed by vehicle emissions, coal combustion, and air-surface exchange.However, notably, determining precise source contributions based on the PMF model and diagnostic ratios only is difficult because of high uncertainty.In addition, the PAH source profiles for coal combustion and residential biomass combustion are usually similar, thereby engendering difficulties in separating these two factors.Thus, further research integrating other methods or evidence is required.

Health Risk Values
The BaP eq values for particle phase PAHs during the sampling period varied from 1.48 to 24.5 ng m -3 with an average of 6.27 ± 8.55 ng m -3 .For gas phase PAHs, the BaP eq values were 4. 28, 11.6, 6.23, and 3.64 ng m -3 for spring, summer, autumn, and winter, respectively with an similar average value (6.43 ± 4.15 ng m -3 ) with particle phase PAHs.The most recent ambient air quality standard for China (GB3095-2012) showed an improvement in the limited value of atmospheric BaP from 10 to 1 ng m -3 .All BaP eq concentrations calculated in Lhasa were higher than the limit value, which indicates the potential for adverse health effects caused by atmospheric PAHs in this region.
People of three age groups (children, teenagers, and adults) were examined to effectively estimate the exposure levels of BaP eq in various age ranges.Table 5 shows the parameters of these three age groups used in the exposure assessment.Based on the data, the DEDs and CR of BaP eq in Lhasa were characterized for all three age groups.The CRs of particle phase PAHs were estimated as 9.20 × 10 -7 , 7.20 × 10 -7 , and 2.84 × 10 -6 for children, teens, and adults, respectively.While for gas phase PAHs, the CRs were similar with particle phase PAHs for three groups, with values as 9.43 × 10 -7 , 7.38 × 10 -7 , and 2.91 × 10 -6 .The CR values of total (gas and particle phase) PAHs were 1.86 × 10 -6 , 1.46 × 10 -6 , and 5.76 × 10 -6 , respectively, which were one time higher than those of particle phase PAHs.PAHs are present in the atmosphere in both particle and gas phases.In this study, the calculated gas phase PAH concentrations were much higher than those of particle phase.And CR values doubled after adding gas phase PAHs.Therefore, additional detailed surveys of gas phase PAHs should be considered 3.14 3.14 3.14 a cited from the US EPA; b citied from statistical data from China; c cited from Chen and Liao (2006). in further study.Most regulatory programs employ a limiting CR value of 1.00E-05 (De Miguel et al., 2007).The lifetime CR (LCR) in Lhasa was determined by calculating the CR values of two types of PAHs for all three age groups.The results showed that the LCR (9.08 × 10 -6 ) was lower than the acceptable target risk value, indicating that atmospheric PAHs in Lhasa pose no or little potential CR to locals.

CONCLUSION
The PAH concentrations in Lhasa were higher than those at other remote sites in the TP, indicating that the atmospheric environment in Lhasa has been influenced by local anthropogenic activities.Clear seasonal variations of PAH concentrations were observed, with maximum and minimum values occurring in winter and summer, respectively.The concentrations of gas phase PAHs were much higher than those of particle phase PAHs, and the two phases exhibited opposite seasonal trends.Particle phase PAH profiles, especially those of four-ring compounds, changed significantly between winter and summer, thereby reflecting the variation in emission sources across different seasons.However, three-ring species, especially Ace, Acel, and Flu, contributed the most to gas phase PAHs.The particle phase PAH sources were quantified using the PMF model, and the results revealed that atmospheric PAHs in Lhasa originated mainly in the combustion of biomass fuels (48.4%) and vehicle emissions (27.9%), which was supported by the diagnostic molecular ratios.The average BaP eq of the PAHs was higher than the new limit in China, thereby indicating potential adverse health effects on local inhabitants.However, a probabilistic health risk assessment showed that atmospheric PAHs in Lhasa posed little or no CR.

Table 2 .
Summary of particle-bound PAH concentrations (ng m -3 ) in Lhasa and other regions.16 PAHs of TSP samples; b 16 PAHs of PM 10 samples; c 15 PAHs of TSP samples; d 17 PAHs of TSP samples; e 15 PAHs of TSP samples; f 22 PAHs of TSP samples. a

Table 3 .
Calculated gas phase PAH concentrations in four seasons of Lhasa.

Table 4 .
Diagnostic ratios of PAHs in Lhasa aerosols and their source profiles.Numbers in brackets represent standard deviations.

Table 5 .
Values and probability distributions of parameters used in the exposure assessment.