Chemical characterization of aerosols at the summit of Mountain Tai in Central East China

PM2.5 and TSP samples were collected at the summit of Mountain Tai (MT) (1534 m a.s.l.) in spring 2006/2007 and summer 2006 to investigate the characteristics of aerosols over central eastern China. For comparison, aerosol samples were also collected at Tazhong, Urumqi, and Tianchi in Xinjiang in northwestern China, Duolun and Yulin in northern China, and two urban sites in the megacities, Beijing and Shanghai, in 2007. Daily mass concentrations of TSP and PM2.5 ranged from 39.6–287.6 μg m −3 and 17.2–235.7 μg m −3 respectively at the summit of MT. Averaged concentrations of PM 2.5 showed a pronounced seasonal variation with higher concentration in summer than spring. 17 water-soluble ions (SO 2− 4 , NO − 3 , Cl , F, PO 4 , NO − 2 , CH3COO , CH2C2O 2− 4 , C2H4C2O 2− 4 , HCOO, MSA, C2O 2− 4 , NH + 4 , Ca 2+, K, Mg2+, Na), and 19 elements of all samples were measured. SO 2− 4 , NO − 3 , and NH+4 were the major water-soluble species in PM 2.5, accounting for 61.50 % and 72.65 % of the total measured ions in spring and summer, respectively. The average ratio of PM2.5/TSP was 0.37(2006) and 0.49(2007) in spring, while up to 0.91 in summer, suggesting that aerosol particles were primarily comprised of fine particles in summer and of considerable coarse particles in spring. Crustal elements (e.g., Ca, Mg, Al, Fe, etc.) showed higher concentration in spring than summer, while most of the pollution species (SO 4 , NO − 3 , K , NO−2 , NH + 4 , Cl , organic acids, Pb, Zn, Cd, and Cr) from local/regional anthropogenic emissions or secondary formation presented higher concentration in summer. The ratio of Ca/Al suggested the impact of Asian dust from the western deserts on the air quality in this region. Correspondence to: G. Zhuang (gzhuang@fudan.edu.cn) The high concentration of K + in PM2.5 (4.41 μg m−3) and its good correlation with black carbon ( r = 0.90) and oxalic acid (r = 0.87) suggested the severe pollution from biomass burning, which was proved to be a main source of fine particles over central eastern China in summer. The contribution of biomass burning to the fine particle at MT accounted for 7.56 % in spring and 36.71 % in summer, and even reached to 81.58 % on a day. As and Pb were two of the most enriched elements. The long-range transport of aerosols spread the heavy pollution from coal-mining/coal-ash to everywhere over China. Anthropogenic air-pollution was evidently rather severe at MT, though it has been declared by UNESCO to be a World Heritage site.

The high concentration of K + in PM 2.5 (4.41 µg m −3 ) and its good correlation with black carbon (r = 0.90) and oxalic acid (r = 0.87) suggested the severe pollution from biomass burning, which was proved to be a main source of fine particles over central eastern China in summer.The contribution of biomass burning to the fine particle at MT accounted for 7.56 % in spring and 36.71% in summer, and even reached to 81.58 % on a day.As and Pb were two of the most enriched elements.The long-range transport of aerosols spread the heavy pollution from coal-mining/coal-ash to everywhere over China.Anthropogenic air-pollution was evidently rather severe at MT, though it has been declared by UNESCO to be a World Heritage site.

Introduction
Aerosols have potential impact on the global atmospheric chemistry, cloud properties and precipitation development (Tegen et al., 1996;Arimoto, 2001;Rastogi andSarin, 2005, 2006).Anthropogenic aerosols, including the primarily emissions and the secondary aerosols, are mainly in fine mode, which has much more adverse impact on climate and hydrologic cycling (Kaufman et al., 2002), visibility (Chan et al., 1999), and human health (Dockery et al., 1993).
Eastern China, including provinces of Hebei, Shandong, Jiangsu, Zhejiang, and the mega-city, Shanghai, is the area with rapidest growth in economy, e.g.Shandong is one of the two provinces in China, whose GDP exceeded 3 × 10 11 RMB in 2008 (http://finance.people.com.cn/), which resulted in the increasing emissions of SO 2 , NO 2 , and particulate matter, and, in turn, the severe acidic precipitation (Wang et al., 2008).Mountain Tai (200 × 50 square kilometers), with the highest altitude (1534 m high) in central-eastern China, C. Deng et al.: Chemical characterization of aerosols at the summit of Mountain Tai is located in Shandong and surrounded by Jiangsu, Anhui, Henan, and Hebei provinces.Aerosols, from the summit of MT could be the representative of the regional pollution.Previous study found that CO and O 3 at the top of MT exhibited high in summer and low in winter, which was attributed to the seasonal changes of meteorological conditions and seasonal variations of sources (Gao et al., 2005;Wang et al., 2001).In addition, VOCs, O 3 , and CO were all higher than those observed at other rural mountainous sites (Suthawaree, et al., 2010), which might be due to the strong sources surrounded.Average concentration of peroxides at MT was much lower than that measured at some rural mountain sites, suggesting that significant removal processes took place in this region (Ren et al., 2009).O 3 and CO play key roles in determining the oxidizing capacity of the atmosphere in the presence of sunlight, and they are ideal tracers for anthropogenic pollutions (Novelli et al., 1994(Novelli et al., , 1998)).VOC, O 3 , CO, and peroxide are all related to the formations of secondary aerosols in ambient air, which suggested that the characteristics of the aerosols at MT might be different from those at other sites.
Mineral aerosols through long-range transport may directly or/and indirectly affect the properties of air mass by providing surfaces for many chemical and physical processes and serving as carriers of anthropogenic substances, which would affect the global climate/environmental change (Guo et al., 2004;Dentener et al., 1996;Sun et al., 2004;Liu et al., 2002).Northwestern China is one of the main source areas of Asian dust that can be transported to hundred and thousand miles away, passing through central and eastern China and even to the Pacific.The composition of mineral aerosols would subject to transform due to adsorbing gaseous species, surface reactions, and coagulation with anthropogenic aerosol on the pathway during transport.
Little information on aerosols at the summit of MT is available.The studies on aerosols reported in literatures mostly represented those samples collected from ground level, and limited knowledge has been acquired on the aerosols at high elevation over the world.MT is just in the downstream of Asian Dust from northwestern China.Therefore, the summit of MT is an ideal site to examine the longrange transport of Asian dust and to observe the mixing of dust with anthropogenic aerosol.This paper presents the characteristics, sources, formation processes, and the relation with the long-range transport of aerosols collected at MT, which would reveal the air quality in PBL (Planetary Boundary Layer) over central east China.

Sampling
TSP and PM 2.5 aerosol samples were simultaneously collected at the meteorological observation station located at the summit of MT (36.25 • N, 117.10 • E) in spring (14 March-18  Ti  and Shanghai (Fig. 1).Information of all samples is list in Table 1.The detailed analytical procedures were given elsewhere (Zhuang et al., 2001).

Ion analysis
One-fourth of each sample and blank filter was extracted ultrasonically by 10 mL deionized water (18 M cm −1 ).After passing through microporous membranes (pore size, 0.45 µm; diameter, 25 mm; made by the affiliated plant of Beijing chemical school), the filtrates were determined for pH with a pH meter (model, Orion 818).Each filtrate was stored at 4 • C in a clean tube for IC analysis.12 anions  Ion Chromatography (Model: Dionex 3000), which consists of a separation column (Dionex Ionpac AS11 for anion and CS12A for cation), a guard column (Dionex Ionpac AG 11 for anion and AG12A for cation), a self-regenerating suppressed conductivity detector (Dionex Ionpac ED50), and a gradient pump (Dionex Ionpac GP50).The gradient mobile phase generated by EG-3000 was used for anion detection, while the weak acid eluent (20 mM MSA) for cation detection.The recovery of each ion was in the range of 80-120 %.
The relative standard deviation of each ion was less than 5 % for reproducibility test.The limits of detection (S/N = 3) were less than 0.04 mg L −1 for anions and 0.006 mg L −1 for cations.The quality assurance was routinely carried out by using Standard Reference Materials (GBW 08606) produced by National Research Center for Certified Reference Materials, China.Blank values were subtracted from sample determinations.The details were given elsewhere (Yuan et al., 2003).

Meteorological data, fire spot map and trace gases
The meteorological data, including temperature, relative humidity (RH), dew point, wind speed, wind direction, atmospheric pressure, visibility etc., were collected from http://www.wunderground;Data of SO 2 , NO 2 in Shanghai and MT were collected from http://www.envir.gov.cn and http://www.tahb.gov.cn;Fire spot data were got from MODIS Global Fire Mapping Service (http://firefly.geog.umd.edu/firemap/); O 3 and CO were detected with a commercial UV photometric analyzer (Thermo Environment Instruments Inc., Model 49) that had a detection limit of 2 ppbv and a 2-sigma (2-s) precision of 2 ppbv for a 2-min average.CO was measured with a gas filter correlation, a nondispersive infrared analyzer (Advanced Pollution Instrumentation Inc., Model 300) with a heated catalytic scrubber for baseline determination, which was conducted every 2h.The detection limit was 30 ppbv for a 2-min average, with a 2-s precision of about 1 % for a level of 500 ppbv (2-min average).The overall uncertainty was estimated to be 10 %.Measurement method of peroxide: A 26-turn coil was used to strip the peroxides out of the air and into a liquid phase by water (0.42 mL min −1 ), which was sucked through a 10-turn coil (0.42 mL min −1 ) at which buffer (potassium hydrogen phthalate (KHP, Fisher Scientific) and tetrasodium ethylenediaminetetraacetate (Na4EDTA, Fluka Scientific)) and fluorescence (POPHA (Fluka Scientific), horseradish peroxidase (Sigma Chemical Co.), KHP, and Na4EDTA) solutions were added, then the liquid entered a pH booster cell that contained 30 % ammonium hydroxide (Sigma Chemical Co.) to adjust solution pH from 6 to approximately 10.After going through a debubbler (0.42 mL min −1 ), where the bubbles were removed, the fluorescence of the dimmer was detected using a fluorimeter (Spectrovision FD-100, Groton Technology Inc., Concord, MA) with excitation and emission wavelength of 326 nm and 400 nm, respectively.

Results and discussion
3.1 Overview of particle matters at summit of MT

Mass concentrations of PM and size distribution
Temporal variations of mass concentrations and the corresponding deviations of PM 2.5 and TSP at MT and other sampling sites are summarized in Fig. 2 and Tables 1-2 14March-6 May), which was likely attributed to the different sampling periods and it will be interpreted in Sect.3.2.1.TSP from MT was comparable with those from other sampling sites in ground level, and there was no obvious decrease at a height of 1534 m compared to those at ground level, as expected.Both TSP and PM 2.5 at MT were much higher than those at Tianchi that is also located at a high land (1900 m).Aerosol pollutions at MT were as severe as or even worse than those at ground level, e.g.PM 2.5 at MT were much higher than those at other sites, including the sites in megacities, Beijing and Shanghai, at ground level in summer.The average ratio of PM 2.5 /TSP was 0.37 in spring and 0.91 in summer in 2006, indicating that fine particles dominated in summer while coarse particles in spring.The seasonal variation of fine particles should be firstly attributed to the seasonally different meteorology at MT (Table 3), which showed typical seasonal variations with higher temperature, lower windy speed, lower atmospheric pressure and more solar radiation in summer than those in spring.The meteorology at MT could strengthen the vertical convection of the air, and the regional anthropogenic pollutants on the ground surface could easily transport upward, resulting in the increase of the height of PBL, that  was even higher than the summit of MT.On the contrary, the height of the PBL in spring was so compressed that it used to be below the summit of MT, and the Mountain-valley breezes could not fully develop due to the weak solar radiation and the strong regional winds in ground level.Hence, the transport upward of the regional pollutants to the summit was less frequent in spring than in summer, which resulted in lower concentrations of fine particles in spring.Besides, dust storms mostly occur in spring and MT is located on the very pathway of its long-range transport from central Asia to the northern America (Arimoto et al., 1996;Zhang et al., 1997;Sun et al., 2010).Thus, in those days with high wind speed in spring, higher ratio of those coarse particles to the total aerosol mass was often observed even at elevation of more than 1500 m.

Ionic and elemental composition of the particles at MT
Mass concentrations of ions in PM 2.5 and TSP at MT are listed in Table 4. Water soluble ions contributed 10.82 % of TSP and 23.99 % of PM 2.5 in mass in spring, while 40.89 % and 41.08 % in summer.This result revealed evidently that the secondary aerosol possessed much larger part in summer than in spring.The ratios of total ions to the total mass in PM 2.5 were higher than those in TSP both in summer and spring, suggesting that the pollution components, such as SO 2− 4 , NH + 4 , NO − 3 , existed more in fine mode.These three major water soluble ions accounted for 61.50 % in spring and 72.65 % in summer of the total ions measured in PM 2.5 , while 69.20 % and 71.47 % in TSP, respectively (Table 5).Also, higher concentration of K + , the tracer of the biomass source, in summer was observed, which could account for 8.26 % of total ions in TSP.Whereas, higher concentration of Ca 2+ , likely more from the Asian dust source, in spring, accounted for 21.93 % of the total ions in TSP.
Water-soluble ions are proved to play key roles in many atmospheric processes, such as cloud formation, visibility degradation, solar radiation, acidification of cloud, rain, and fog, and haze formation because of their affinity with water (Tsai et al., 1999;Novakov et al., 1993;Matsumoto et al., 1997;Facchini et al., 2000).Haze mostly occurred in stable, warm, and moist air, and was characterized by the high level of fine particles with high percentage of certain water soluble ions, such as SO 2− 4 , NH + 4 , and NO − 3 , in the aerosols.Certain characteristics of the aerosols collected at MT in summer and spring and at Beijing in dust, haze, and clear days are summarized in Table 5. C IC /C P , C (S+N+A) /C IC , SOR and NOR at MT in summer were very similar to those in haze days in Beijing, suggesting that haze could occur frequently in summer at the summit of MT, which could be proved by the lower visibility (9.0 km) on average at MT.
The concentrations of nineteen elements and black carbon (BC) in aerosols at MT are listed in Table 6, which showed clearly that crustal element (Ca, Ma, Al, Mn, Ti, Sr, and Na) were higher in spring, while pollution elements (Pb, Cr, Cd, Zn, Ni, S, BC), except As and Cu, were higher in summer.The elements could be classified into four groups according to their enrichment factors (EFs = (X/Al) aerosol /(X/Al) crust ): high enriched pollution elements (Pb and As), medium enriched (S and Zn), slightly enriched (Ni, Cu and Cr), and non-enriched crustal metals (Ca, Ma, Al, Mn, Ti, Sr and Na).EFs of all crustal elements (see Fig. 3) were higher in TSP

Acidity of the aerosols at MT
Figure 4 shows the variations of pH of the filtrates of aerosols at MT, and it indicated that pH decreased obviously from spring to summer.The mean pH values of the aqueous filtrates of aerosols collected at MT and other sampling sites are compared in Table 7.The aerosols at MT showed higher acidity for both PM 2.5 (pH = 4.62) and TSP (pH = 4.92) in summer in comparison with the weak acidity of PM 2.5 (pH = 5.92) and slight alkalinity (pH = 7.22) of TSP in spring.
The pH of aerosols at MT in summer was the lowest, while TSP in spring showed slight alkalinity, which was similar to those from Tazhong (pH = 7.39, original source of dust storm) and from Beijing in supper dust day (pH = 7.25, Wang et al., 2005).The higher acidity of aerosols at MT in summer was further supported by the fact that the pH of rain samples in summer at MT was obviously lower than those in other seasons (Wang et al., 2006c  aerosols is determined by the presence and the proportions of the cations and anions in the filtrates of aerosols.In addition, the ratio of equivalent concentrations of total cations to total anions (C/A) should be equal to 1, if all the cations and anions in the aerosol were measured.However, CO 2− 3 and HCO − 3 were not measured in this study due to the limitation of ion chromatography, hence, the difference of the total anions and cations could be used to roughly estimate the amount of CO 2− 3 and HCO − 3 unmeasured (Wang et al., 2005).C/A of aerosols in spring at MT reached 1.6, while only 1.1 in summer, suggesting that much more carbonate or bicarbonate presented in the aerosols at MT in spring.The remarkable increase of SO 2− 4 , NO − 3 and organic acids in summer shown in Table 4 indicated that in summer much more aerosols at MT was from secondary pollution, which would lead to the lower pH of the filtrates.Compared to the pH of those aerosols collected from Shanghai in summer (5.29 in PM 2.5 and 6.37 in TSP), the aerosols at MT showed even higher acidity, suggesting that the secondary air pollution at MT in summer was even severer than that in Shanghai.

Influence of biomass burning
Biomass burning includes those from grassland, forest, and crop residue, among which more than 60 % are from crop residue burning in the world (Streets et al., 2003).With the rapid development of economy in China, crop residue is increasingly being burned openly in the field.In harvest every year ∼5.182 × 10 7 ton crop residues (accounted for 40.0 % of the total crop residues in China) were directly burnt openly, especially in central eastern China, including Shandong, Jiangsu, Henan, Hebei provinces.Shandong, where MT is located, was in number 1 among those provinces mentioned above with the largest amount of crop residues (1.798 × 10 7 t) to be burned openly (Cao et al., 2007).Biomass burning emissions are known to contribute a considerable amount of PM and gaseous pollutants to the air, and crop residues contribute more fine particles and gaseous than other biomaterial residues (Cao et al., 2005;Zhu et al., 2005).Water soluble potassium is a good tracer for aerosols formed from biomass burning (Andreae, 1983).Sources of total potassium ion (K + Total ) in aerosols might be attributed to soot from biomass burning, sea salt, and crustal dust.Elemental Al and Na + could be the good markers of crustal dust and sea salt respectively, as the composition of Na, K, and Al in sea water and in crustal dust were found to be: Na: 31 %, K:1.1 %, Al:0 %, and Na: 2.6 %, K:2.9 %, Al:7.7 %, respectively (Wedepohl, 1995), which showed that the percentage of K + were much lower than Na + in sea salt and also lower than Al in crustal dust.This means that the concentrations of Na + and Al in the aerosols would increase more than K + if the aerosols were from sea salt and crustal dust.Figure 5 showed that K + on certain days in summer was significant higher than Na + and Al with remarkable increase from the normal days, which suggested that the high concentrations of PM 2.5 in summer could likely from biomass burning rather than sea water or crustal dust.It www.atmos-chem-phys.net/11/7319/2011/C. Deng et al.: Chemical characterization of aerosols at the summit of Mountain Tai was reasonably assumed that the total potassium (K + Total ) is the sum of dust-derived potassium (K + Crust ), sea-salt-derived potassium (K + SS ), and biomass burning-derived potassium (K + BB ) i.e.K + Total = K + Crust + K + SS + K + BB (Virkkula et al., 2006), we can calculate the biomass burning potassium (K + BB ) as K + BB= K + Total -K + Crust -K + SS .Assuming Na + in aerosol is from crust and sea salt only, i.e.Na + Total = Na + Crust + Na + SS , where Na + Total is the concentration of Na + in the aerosol sample, and Al in aerosol is originated from crust only, thus, K + Crust and K + SS can be estimated through the ratios of K + /Al and Na + /Al in the aerosol sample and the ratio of K + /Na + in sea salt, i.e.K + Crust = (K + /Al) crust × Al crust and K + SS = (Na + Total -(Na + /Al) crustl × Al crustl ) × (K + /Na + ) sea−salt .The ratio of K + /Na + in those particles from sea salt is constant (0.037) (Chester, 1990), whereas the ratios of K + /Al and Na + /Al in crustal dust were different between the samples collected from different sites because of the different background and the different influence of the human activities on each site.We now have two methods to estimate the biomass burning derived potassium (K + BB ): (1) Using the minimum ratios of K + /Al of 0.152 and Na + /Al of 0.240 in the PM 2.5 aerosol sample collected at MT on 26 April 006 among all samples collected to calculate the background value of K + and Na + in crustal dust, then the biomass burning derived potassium (K + BB ) is given from K + Crust = 0.152 × Al Aerosol , and K + SS = (Na + Total -(Na + /Al) Aerosol × Al Aerosol ) × (K + /Na + ) sea−salt , i.e.K + SS = (Na + Total -0.240 × Al Aerosol ) × 0.037, (2) Using the ratios of K + /Al of 0.107 and Na + /Al of 0.031 measured in the soil samples collected at MT to calculate the background value of K + and Na + in crustal dust, then the biomass burning derived potassium (K + BB ) is given from .107 × Al Aerosol , and The concentrations of K + BB (biomass burning-derived K + ) calculated with two methods mentioned above in PM 2.5 collected at MT in 2006 are shown in Fig. 6.It can be seen clearly that very similar concentrations of K + BB were acquired by the two methods.The averaged K + BB were 0.40 and 4.30 µg m −3 in spring and summer, respectively, with the method of using the minimum ratios of K + /Al and Na + /Al in the aerosol sample to calculate the background value of K + and Na + in crustal dust, while the K + BB were 0.32 and 4.30 µg m −3 with the method of using the ratios of K + /Al and Na + /Al in the soil samples.The high level of K  (4.30 µg m −3 ) in summer indicated that there were the significant part of the aerosols from biomass burning emissions at MT during the summer time.This result could also be confirmed by the fire spot data in this region, as shown in Fig. 7a-f.The amount of those fire spots distributed in the region surrounding MT increases obviously in May (Fig. 7c), and highly active fire disturbance appeared in June (Fig. 7d).
The fire spots spread mainly to the south to MT during 1-9 June (Fig. 7e), and then extended to the northern places to MT during 10-19 June (Fig. 7f), which was in accordance to the harvest time of wheat/rice from south to north in this region.
The contributions of biomass burning, crustal dust, and other sources to the fine particles, PM 2.5 , in spring and summer 2006 at MT are shown in Fig. 8.To assess quantitatively the contributions of biomass burning openly from the agriculture residues to the formation of PM 2.5 at MT, we used the ratio of K + /PM 2.5 of 9.56 % (wt %) from the agriculture residues, i.e. biomass burning-derived PM 2.5 = K + BB /0.0956, to calculate the mass of PM 2.5 derived from biomass burning, for both Li et al. (2007) and Cao et al. (2008a) reported a very similar content of K + in PM 2.5 emitted from agriculture residues: 9.94 ± 11.8 or 9.56 ± 9.01(wt %) from wheat straw, and 11.38 ± 8.49(wt %) from rice straw.The crustal dust-derived PM 2.5 was calculated with the formula of Al/0.08.The results revealed that the contribution of biomass burning to the fine particle at MT accounted for 7.56 % in spring and 36.71% in summer, and even reached to 81.58 % on the day of 12 June.In addition, the concentrations of K + at different sampling sites are summarized in 8. K + at was much higher than those at other sites, and it showed strong seasonal variation with 4.41 µg m −3 in summer and 0.48 µg m −3 in spring in PM 2.5 .Also, K + in PM 2.5 correlated well to other species that are related to biomass burning, such as BC, C 2 O 2− 4 , etc. (see Table 9).All of these results demonstrated evidently that biomass burning was one of major contributor to the aerosol pollution in summer over central eastern China, where MT is located.).  9. The results showed that these secondary ions in aerosols in summer at MT were ∼5 times higher of those in spring, while there was no big change in Beijing, less change in Shanghai between the two seasons, and they were evidently greater in spring than those in summer at Urumqi.The sum of the three ions at MT in summer was higher than that at any other site, indicating much heavy secondary pollution at this region.Besides, at MT in summer the concentration of SO 2− 4 was much greater than NO − 3 , while in Beijing and Shanghai the concentrations of NO − 3 was close to that of SO 2− 4 .Average concentrations of  Duolun, MT, and Beijing.The ratio of Ca/Al has been proved to be a good tracer for different dust original areas (Wang et al., 2005;Sun et al., 2005), for Ca/Al showed remarkable dependence of the source regions for both dust aerosol and soil samples.Ratios of Ca/Al in aerosol or soil from different sampling sites are shown in Table 10, and it could be seen clearly that Taklimakan desert, the western source of Asian dust, is characterized by high ratio of Ca/Al of greater than 1.5, while the northern source of Asian dust, such as Duolun and Hunshandake Sandland, is characterized by low ratio of Ca/Al of around 0.5.The ratio of Ca/Al (1.37 ± 0.22) in the aerosol at MT indicated that the dust aerosol at MT in spring could be from the long-range transport of the dust from western or northwestern high-dust sources.The facts mentioned above that the coarse particles dominated in spring (PM 2.5 /TSP of 0.37) and the alkalinity (pH = 7.22) of the TSP in spring further confirmed that the long-range transport was the major source of the aerosols in spring at MT.

The pollution elements As and Pb
The pollution elements, As and Pb, were highly enriched in the aerosols with the EF of 1541 and 679, respectively, in spring, while 1470 and 1969 in summer in PM 2.5 at MT.As and Pb were of different seasonal variations with summerhigh/spring-low of Pb, while spring-high/summer-low of As.
Correlation analysis (see Table 11) revealed that both Pb and As were highly correlated to those crustal elements in spring with the correlation coefficients of 0.701 for Pb to Al and 0.873 for Pb to Fe, while of 0.837 for As to Al and 0.778 for As to Fe.Furthermore, the correlation coefficient between Pb and As was as high as 0.949.These results indicated that both Pb and As were highly associated with the mineral components, which was long-range transported from northwestern China to MT in spring.Those primary dust aerosols from northwestern China would mix with As and Pb, which were emitted from coal-mining/coal-ash and widely distributed over northwestern China, and act as a carrier for As and Pb.The pollution elements, As and Pb, would gradually be enriched in the dust aerosol during its longrange transport.The results revealed that the long-range transport of aerosols spread the heavy pollution from coal burning everywhere over China.However, in summer the correlation coefficients of Pb to Al and Pb to Fe decreased between Pb and As decreased to 0.494.These results indicated that in summer the source of Pb and As was not mainly from long-range transport, instead, they could be from those local/regional sources.It could be seen that Pb and As were both correlated well to Cr, Cu, and Zn in summer, suggesting that the source of Pb and As in summer would likely from the same local/reginal anthropogenic discharge as the pollution elements, Cr, Cu, and Zn, did.The strong convections of air mass would result in the elevation to the summit of MT from the ground level of the pollutants.The emission of Pb from the local/regional anthropogenic source in summer could be much more than that from the long-range transport in spring, which led to the higher enrichment of Pb in summer than that in spring.

Summary
Aerosols over central eastern China showed significantly season variation, with fine particles dominated in summer while coarse particles in spring.The summit of MT was suffering from the invasion of long-range transported dust from northwestern China, the heavy local/regional anthropogenic emissions from the surrounding areas, and the severe secondary pollutions.The contribution of biomass burning to the fine particle at MT accounted for 7.56 % in spring and 36.71% in summer, and even reached to 81.58 % on a day.The high levels of peroxides and ozone, as well as the favourable meteorological conditions in the ambient air were in favour of the transformation of SO 2 and NO x to sulfate and nitrate, which resulted in the even severe secondary pollution in this region.As and Pb were two of the most enriched elements.The long-range transport of aerosols spread the heavy pollution from coal-mining/coal-ash to everywhere over China.Anthropogenic air-pollution was evidently rather severe at MT.

Fig. 1 .
Fig. 1.Map of the sampling sites in this study.

Fig. 1 .
Fig. 1.Map of the sampling sites in this study.

Fig. 2 .
Fig. 2. Daily variations of TSP and PM 2.5 from 14 March to 30 June in 2006 at MT.

Fig. 6 .Fig. 6 .
Fig. 6.Concentrations of biomass burning derived K + calculated with two methods in PM2.5 in 2006 at M Fig. 6.Concentrations of biomass burning derived K + calculated with two methods in PM 2.5 in 2006 at MT.

Table 1 .
Mass concentrations of aerosols (µg m −3 ) at the summit of MT and other sampling sites.

Table 2 .
Mass concentrations (µg m −3 ) and size distributions of aerosols at the summit of MT and other sampling sites.

Table 3 .
Meteorological conditions and concentrations of gases (O 3 , CO, Peroxide) in ambient air at MT.

Table 4 .
Water-soluble ions in PM 2.5 and TSP and the corresponding ratios of summer/spring at MT.

Table 6 .
Concentrations (µg m −3 ) of elements in PM 2.5 and TSP and the corresponding ratios of summer/spring at MT. Unit after BC: ng m −3 .and lower in PM 2.5 , while pollution elements were higher in PM 2.5 and lower in TSP.EFs of As and Pb exhibited different seasonal variations with higher EF of Pb in summer and higher EF of As in spring.Daily variations of TSP and PM2.5 from 14 March to 30 June in 2006 at MT.

Table 7 .
pH of aqueous filtrates of PM 2.5 and TSP aerosols at different sampling sites.
Daily variations of pH of aqueous filtrates of aerosols at MT, 2006.Daily variations of K + , Na + , and Al in PM 2.5 from 14 March to 30 June in 2006 at MT. Concentrations of biomass burning derived K + calculated with two methods in PM 2.5 in 2006 at MT Fig. 5. Daily variations of K + , Na + , and Al in PM 2.5 from 14 March to 30 June in 2006 at MT. Daily variations of K + , Na + , and Al in PM2.5 from 14 March to 30 June in 2006 at MT.

Table 8 .
Concentrations of K + in PM 2.5 at different sampling sites.Wang et al. (2006b); b Wang et al. (2006a). a

Table 10 .
Ratio of Ca/Al in aerosol or surface soil in different sampling sites over China.

Table 11 .
Correlation coefficients among the certain species in spring and summer at MT. −0.184 and −0.194, while of As to Al and As to Fe to 0.469 and 0.456, respectively, and the correlation coefficient to