Atmospheric Environment (v.44, #30)
Editorial board (i).
SVOC partitioning between the gas phase and settled dust indoors by Charles J. Weschler; William W Nazaroff (3609-3620).
Semivolatile organic compounds (SVOCs) are a major class of indoor pollutants. Understanding SVOC partitioning between the gas phase and settled dust is important for characterizing the fate of these species indoors and the pathways by which humans are exposed to them. Such knowledge also helps in crafting measurement programs for epidemiological studies designed to probe potential associations between exposure to these compounds and adverse health effects. In this paper, we analyze published data from nineteen studies that cumulatively report measurements of dustborne and airborne SVOCs in more than a thousand buildings, mostly residences, in seven countries. In aggregate, measured median data are reported in these studies for 66 different SVOCs whose octanol-air partition coefficients (K oa) span more than five orders of magnitude. We use these data to test a simple equilibrium model for estimating the partitioning of an SVOC between the gas phase and settled dust indoors. The results demonstrate, in central tendency, that a compound’s octanol-air partition coefficient is a strong predictor of its abundance in settled dust relative to its gas phase concentration. Using median measured results for each SVOC in each study, dustborne mass fractions predicted using K oa and gas-phase concentrations correlate reasonably well with measured dustborne mass fractions (R 2 = 0.76). Combined with theoretical understanding of SVOC partitioning kinetics, the empirical evidence also suggests that for SVOCs with high K oa values, the mass fraction in settled dust may not have sufficient time to equilibrate with the gas phase concentration.
Keywords: Exposure pathways; Flame retardants; Indoor environment; Octanol-air partitioning; Pesticides; Plasticizers;
Global sensitivity analysis of a process-based model for ammonia emissions from manure storage and treatment structures by J. Arogo Ogejo; R.S. Senger; R.H. Zhang (3621-3629).
The development of process-based models to estimate ammonia emissions from animal feeding operations (AFOSs) is sought to replace costly and time-consuming direct measurements. Critical to process-based model development is conducting sensitivity analysis to determine the input parameters and their interactions that contribute most to the variance of the model output. Global and relative sensitivity analyses were applied to a process-based model for predicting ammonia emissions from the surface of anaerobic lagoons for treating and storing manure. The objectives were to compare global sensitivity analysis (GSA) to relative (local) sensitivity analysis (RSA) on a process-based model for ammonia emissions. Based on the first-order coefficient, both GSA and RSA showed the model input parameters in order of importance in process model for ammonia emissions from lagoon surfaces were: (i) pH, (ii) lagoon liquid temperature, (iii) wind speed above the lagoon surface, and (iv) the concentration of ammoniacal nitrogen in the lagoon. The GSA revealed that interactions between model parameters accounted for over two-thirds of the model variance, a result that cannot be achieved using traditional RSA. Also, the GSA showed that parameter interactions involving liquid pH had more impact on the model output variance than the single parameters: (i) temperature, (ii) wind speed, or (iii) total ammoniacal nitrogen. This study demonstrates that GSA provides a more complete analysis of model input parameters and their interactions on the model output compared to RSA. A comprehensive tutorial regarding the application of GSA to a process model is presented.
Keywords: Sensitivity analysis; Global sensitivity analysis; Relative sensitivity analysis; Ammonia emissions; Lagoon ammonia; Emissions;
Seasonal variability and source apportionment of metals in the atmospheric deposition in Belgrade by Z. Mijić; A. Stojić; M. Perišić; S. Rajšić; M. Tasić; M. Radenković; J. Joksić (3630-3637).
The primary objective of this study is to assess anthropogenic impacts on the environment by determination of element atmospheric depositions. Bulk depositions were collected monthly, from June 2002 to December 2006, at three urban locations in Belgrade. Concentrations of Al, V, Cr, Mn, Fe, Ni, Cu, Zn, Cd and Pb were analyzed by atomic absorption spectrometry and the current deposition fluxes of atmospheric metals were established. Fourier analysis was applied in order to investigate seasonal variation of the monthly data set. Nickel, V, Fe and Al showed pronounced seasonal dependence, while seasonal variation of the other elements was not evident. The enrichment factors of Pb, Zn, Cd and Cu were obviously above those who could have been caused by natural processes, indicating a mainly anthropogenic origin. Nickel was intermediately enriched suggesting participation of both natural and anthropogenic sources. The multivariate receptor model, Unmix, was used to analyze a 5-yr element atmospheric depositions data set. Three main source profiles (mixed road dust, oil combustion and metal processing) were identified and the overall average percentage source contributions determined.
Keywords: Metals; Bulk deposition; Fourier analysis; Unmix;
Seasonal variability of formaldehyde production from photolysis of rainwater dissolved organic carbon by M.W. Southwell; J.D. Smith; G.B. Avery; R.J. Kieber; J.D. Willey (3638-3643).
Photochemical production of formaldehyde (HCHO) was measured in rainwater from 13 precipitation events in Wilmington, North Carolina, USA under conditions of simulated sunlight. HCHO concentrations increased in all samples irradiated with no changes observed in dark controls. HCHO photoproduction rates were strongly correlated with dissolved organic carbon (DOC) suggesting HCHO was derived from direct or indirect photolysis of rainwater DOC. The higher photoproduction rates (0.03–2.9 μM h−1) relative to those reported for surface waters suggests that rainwater DOC is more photolabile in terms of HCHO production than surface waters. HCHO photoproduction rates were higher in growing season (1.0 ± 1.0 μM h−1) compared to non-growing season (0.08 ± 0.05 μM h−1) even when rates were normalized for DOC (6.8 ± 3.6 μM h−1 mM C−1 versus 1.8 ± 1.0 μM h−1 mM C−1). The higher growing season rate may be related to seasonal differences in the composition of DOC as evidenced by differences in fluorescence per unit carbon of rainwater samples. Irradiation of C18 extracts of rainwater also produced HCHO, but at lower rates compared to corresponding whole rain samples, suggesting that hydrophyllic components of rainwater play a role in HCHO photoproduction. Our results indicate that photolysis of rainwater DOC produces significant amounts of HCHO, and possibly other low molecular weight organic compounds, likely increasing its reactivity and bioavailability.
Keywords: Formaldehyde; Carbonyl; Dissolved organic matter; Dissolved organic carbon; Photolysis; Photochemical;
Ions in motor vehicle exhaust and their dispersion near busy roads by E.R. Jayaratne; X. Ling; L. Morawska (3644-3650).
Measurements in the exhaust plume of a petrol-driven motor car showed that molecular cluster ions of both signs were present in approximately equal amounts. The emission rate increased sharply with engine speed while the charge symmetry remained unchanged. Measurements at the kerbside of nine motorways and five city roads showed that the mean total cluster ion concentration near city roads (603 cm−3) was about one-half of that near motorways (1211 cm−3) and about twice as high as that in the urban background (269 cm−3). Both positive and negative ion concentrations near a motorway showed a significant linear increase with traffic density (R 2 = 0.3 at p < 0.05) and correlated well with each other in real time (R 2 = 0.87 at p < 0.01). Heavy duty diesel vehicles comprised the main source of ions near busy roads. Measurements were conducted as a function of downwind distance from two motorways carrying around 120–150 vehicles per minute. Total traffic-related cluster ion concentrations decreased rapidly with distance, falling by one-half from the closest approach of 2 m to 5 m of the kerb. Measured concentrations decreased to background at about 15 m from the kerb when the wind speed was 1.3 m s−1, this distance being greater at higher wind speed. The number and net charge concentrations of aerosol particles were also measured. Unlike particles that were carried downwind to distances of a few hundred metres, cluster ions emitted by motor vehicles were not present at more than a few tens of metres from the road.
Keywords: Vehicle emissions; Cluster ions; Charged particles;
Variable emissions of microbial volatile organic compounds (MVOCs) from root-associated fungi isolated from Scots pine by Jaana Bäck; Hermanni Aaltonen; Heidi Hellén; Maija K. Kajos; Johanna Patokoski; Risto Taipale; Jukka Pumpanen; Jussi Heinonsalo (3651-3659).
Soils emit a large variety of volatile organic compounds. In natural ecosystems, measurements of microbial volatile organic compound (MVOC) exchange rates between soil and atmosphere are difficult due to e.g. the spatial heterogeneity of the belowground organisms, and due to the many potential sources for the same compounds. We measured in laboratory conditions the MVOC emission rates and spectra of eight typical fungi occurring in boreal forest soils. The studied species are decomposers (Gymnopilus penetrans, Ophiostoma abietinum), ectomycorrhizal (Cenococcum geophilum, Piloderma olivaceum, Suillus variegatus, Tomentellopsis submollis) and endophytic fungi (Meliniomyces variabilis, Phialocephala fortinii). The MVOC emissions contained altogether 21 known and 6 unidentified compounds whose emission rates were >0.1 μg g(DW)−1 h−1. The most abundant compounds were the short-chain carbonyl compounds (acetone and acetaldehyde). The greatest carbonyl emissions were measured from P. olivaceum (1.9 mg acetone g(DW)−1 h−1) and P. fortinii (0.114 mg acetaldehyde g(DW)−1 h−1). Terpenoid emissions (isoprene, mono- and sesquiterpenes) were detected from some fungal cultures, but in relatively small amounts. We conclude that soil micro-organisms can potentially be responsible for significant emissions of volatiles, especially short-chain oxygenated compounds, to the below-canopy atmosphere.
Keywords: Boreal soil fungi; Microbial volatile organic compounds (MVOCs); Carbonyls; Isoprenoids;
Evaluation of land-use regression models used to predict air quality concentrations in an urban area by Markey Johnson; V. Isakov; J.S. Touma; S. Mukerjee; H. Özkaynak (3660-3668).
Cohort studies designed to estimate human health effects of exposures to urban pollutants require accurate determination of ambient concentrations in order to minimize exposure misclassification errors. However, it is often difficult to collect concentration information at each study subject location. In the absence of complete subject-specific measurements, land-use regression (LUR) models have frequently been used for estimating individual levels of exposures to ambient air pollution. The LUR models, however, have several limitations mainly dealing with extensive monitoring data needs and challenges involved in their broader applicability to other locations. In contrast, air quality models can provide high-resolution source–concentration linkages for multiple pollutants, but require detailed emissions and meteorological information. In this study, first we predicted air quality concentrations of PM2.5, NOx, and benzene in New Haven, CT using hybrid modeling techniques based on CMAQ and AERMOD model results. Next, we used these values as pseudo-observations to develop and evaluate the different LUR models built using alternative numbers of (training) sites (ranging from 25 to 285 locations out of the total 318 receptors). We then evaluated the fitted LUR models using various approaches, including: 1) internal “Leave-One-Out-Cross-Validation” (LOOCV) procedure within the “training” sites selected; and 2) “Hold-Out” evaluation procedure, where we set aside 33–293 tests sites as independent datasets for external model evaluation. LUR models appeared to perform well in the training datasets. However, when these LUR models were tested against independent hold out (test) datasets, their performance diminished considerably. Our results confirm the challenges facing the LUR community in attempting to fit empirical response surfaces to spatially- and temporally-varying pollution levels using LUR techniques that are site dependent. These results also illustrate the potential benefits of enhancing basic LUR models by utilizing air quality modeling tools or concepts in order to improve their reliability or transferability.
Keywords: Air pollution; Exposure assessment; Intra urban scale; Dispersion models; Health effects assessment; Land-use regression models;
Seasonal and spatial trends in the sources of fine particle organic carbon in Israel, Jordan, and Palestine by Erika von Schneidemesser; Jiabin Zhou; Elizabeth A. Stone; James J. Schauer; Radwan Qasrawi; Ziad Abdeen; Jacob Shpund; Arye Vanger; Geula Sharf; Tamar Moise; Shmuel Brenner; Khaled Nassar; Rami Saleh; Qusai M. Al-Mahasneh; Jeremy A. Sarnat (3669-3678).
A study of carbonaceous particulate matter (PM) was conducted in the Middle East at sites in Israel, Jordan, and Palestine. The sources and seasonal variation of organic carbon, as well as the contribution to fine aerosol (PM2.5) mass, were determined. Of the 11 sites studied, Nablus had the highest contribution of organic carbon (OC), 29%, and elemental carbon (EC), 19%, to total PM2.5 mass. The lowest concentrations of PM2.5 mass, OC, and EC were measured at southern desert sites, located in Aqaba, Eilat, and Rachma. The OC contribution to PM2.5 mass at these sites ranged between 9.4% and 16%, with mean annual PM2.5 mass concentrations ranging from 21 to 25 ug m−3. These sites were also observed to have the highest OC to EC ratios (4.1–5.0), indicative of smaller contributions from primary combustion sources and/or a higher contribution of secondary organic aerosol. Biomass burning and vehicular emissions were found to be important sources of carbonaceous PM in this region at the non-southern desert sites, which together accounted for 30%–55% of the fine particle organic carbon at these sites. The fraction of measured OC unapportioned to primary sources (1.4 μgC m−3 to 4.9 μgC m−3; 30%–74%), which has been shown to be largely from secondary organic aerosol, is relatively constant at the sites examined in this study. This suggests that secondary organic aerosol is important in the Middle East during all seasons of the year.
Keywords: Middle East; Source apportionment; Aerosol; Carbonaceous; Organic carbon;
Isotopes as tracers of sources of lead and strontium in aerosols (TSP & PM2.5) in Beijing by David Widory; Xiande Liu; Shuping Dong (3679-3687).
Even after its being phased out in gasoline in the late 90s, lead (Pb) is still present at relatively high levels in the atmosphere of Beijing, China (0.10–0.18 μg m−3). Its origin is subject to debate as several distinct sources may contribute to the observed pollution levels. This study proposes to constrain the origin(s) of Pb and strontium (Sr) in aerosols, by coupling both Pb and Sr isotope systematics. The characterisation of the main pollution sources (road traffic, smelters, metal refining plants, coal combustion, cement factories, and soil erosion) shows that they can unambiguously be discriminated by the multi-isotope approach (206Pb/204Pb and 87Sr/86Sr). The study of total suspended particulates (TSP) and fine particles (PM2.5) from Beijing and its vicinity indicates that both size fractions are controlled by the same sources. Lead isotopes indicate that metal refining plants are the major source of atmospheric lead, followed by thermal power stations and other coal combustion processes. The role of this latter source is confirmed by the study of strontium isotopes. Occasionally, emissions from cement plants and/or input from soil alteration are isotopically detectable.
Keywords: Air pollution; Lead isotopes; Strontium isotopes; Aerosol sources;
Evaluation of an aerosol optical scheme in the chemistry-transport model CHIMERE by J.C. Péré; M. Mallet; V. Pont; B. Bessagnet (3688-3699).
This paper presents an aerosol optical scheme developed in the chemistry-transport model CHIMERE dedicated to calculate optical properties of particles. Such developments are very helpful as they complement the usual validation with PM (Particulate Matter) ground-based measurements by using surface (AERONET/PHOTONS network) and satellite (MODIS) remote sensing observations. To reach this goal, Aerosol Optical Thickness (AOT), column-averaged Single Scattering Albedo (SSA) and asymmetry parameter (g) are calculated at 440 nm, 675 nm, 870 nm and 1020 nm (AERONET wavelengths) under three hypotheses on the particle mixing state (external, internally homogeneous and core-shell). Furthermore and in addition to optical calculations, an original development has been made to estimate column volume size distributions in CHIMERE, directly comparable with AERONET retrievals. Comparisons between simulations and observations are made over Western Europe for the year 2003 but also for one specific case focused on ammonium nitrate aerosols. Observed AOT display a seasonal cycle (with highest values during summer) rather well reproduced by the model but biases with observational data have been found depending on seasons. In fall, winter and early spring, modeled AOT values agree well with AERONET retrievals with small negative biases. Focus on a pollution episode of ammonium nitrate origin during March 2003 reveals that CHIMERE is able to well reproduce the fine mode volume size distribution retrieved by AERONET, leading to good agreements between modeled and observed AOT. In late spring and summer, AERONET AOT values are underpredicted by the model, which could be due to uncertainties in modeling secondary species.
Keywords: Aerosol modeling; Mixing state; Optical properties; Volume size distribution;
Atmospheric trace element concentrations in total suspended particles near Paris, France by Sophie Ayrault; Abderrahmane Senhou; Mélanie Moskura; André Gaudry (3700-3707).
To evaluate today’s trace element atmospheric concentrations in large urban areas, an atmospheric survey was carried out for 18 months, from March 2002 to September 2003, in Saclay, nearby Paris. The total suspended particulate matter (TSP) was collected continuously on quartz fibre filters. The TSP contents were determined for 36 elements (including Ag, Bi, Mo and Sb) using two analytical methods: Instrumental Neutron Activation Analysis (INAA) and Inductively Coupled Plasma Mass Spectrometry (ICP-MS). The measured concentrations were in agreement within the uncertainties with the certified values for the polycarbonate reference material filter SRM-2783 (National Institute for Standard Technology NIST, USA). The measured concentrations were significantly lower than the recommended atmospheric concentrations. In 2003, the Pb atmospheric level at Saclay was 15 ng/m3, compared to the 500 ng/m3 guideline level and to the 200 ng/m3 observed value in 1994. The typical urban background TSP values of 1–2, 0.2–1, 4–6, 10–30 and 3–5 ng/m3 for As, Co, Cr, Cu and Sb, respectively, were inferred from this study and were compared with the literature data. The typical urban background TSP concentrations could not be realised for Cd, Pb and Zn, since these air concentrations are highly influenced by local features. The Zn concentrations and Zn/Pb ratio observed in Saclay represented a characteristic fingerprint of the exceptionally large extent of zinc-made roofs in Paris and its suburbs. The traffic-related origin of Ba, Cr, Cu, Pb and Sb was demonstrated, while the atmospheric source(s) of Ag was not identified.
Keywords: Total suspended particulate; Dust; Metals; INAA; ICP-MS; Air pollution;
Anhydrosugar and sugar alcohol organic markers associated with carboxylic acids in particulate matter from incense burning by Ying I. Tsai; Pei-Ling Wu; Yu-Ting Hsu; Chi-Ru Yang (3708-3718).
Aerosol from the burning two types of sandalwood-based incense, Hsing Shan and Lao Shan, was analyzed to characterize the chemical profile of total particulate matter emitted. The total particulate matter (PM) mass emission factors were 46.3 ± 2.68 mg g−1 of Hsing Shan incense and 43.7 ± 1.08 mg g−1 of Lao Shan incense. Chemical analysis of emissions from the two types of incense revealed that of the 25 components in four groups characterized, anhydrosugars formed the major group, at 46.7–52.2% w/w of the identified particulate and 1078.3–1169.8 μg g−1 of incense, followed by inorganic salts at 30.4–31.8% w/w of identified particulate and 681.6–734.0 μg g−1 of incense, carboxylic acids at 12.0–17.1% w/w of the identified particulate and 268.6–392.8 μg g−1 of incense, and sugar alcohols at 4.44–5.38% w/w of the identified particulate and 102.3–120.6 μg g−1 of incense. More anhydrosugars and sugar alcohols were emitted from Lao Shan incense than from Hsing Shan incense whereas more carboxylic acids and organic salts were emitted from Hsing Shan than from Lao Shan. These differences were due to structural and functional differences in the young sandalwood used to make Hsing Shan and the aged sandalwood used to make Lao Shan. The anhydrosugar levoglucosan, used as a marker of biomass burning, was always the most abundant species in emitted PM for both incenses (Lao Shan 21.7 mg g−1 of PM and Hsing Shan 18.7 mg g−1). K+ and Cl− were the second most abundant components (K+ and Cl− were summed), accounting for 10.6 mg g−1 of Hsing Shan PM and 9.85 mg g−1 of Lao Shan PM. The most abundant carboxylic acids in the emissions were formic, acetic, succinic, glutaric and phthalic acid. The latter is a fragrance ingredient and a potential health hazard and was twice as prevalent in Lao Shan emissions. Xylitol was the most prevalent of the sugar alcohols at 35.7–36.6% w/w of total identified sugar alcohols. These abundant species are potential markers for incense burning. K+, levoglucosan, mannosan and xylitol are already reported in discriminator ratios for wood burning and it is proposed here that these can and should also apply to incense burning. The calculated discriminator ratios for two types of incense burning reported here are 0.229–0.288 for K/Levo, 12.5–13.5 for Levo/Manno, and 21.5–23.7 for the novel discriminator ratio Levo/Xylitol.
Keywords: Incense burning; Organic marker; Levoglucosan; Xylitol; Oxalic acid; Emission factor; Emission strength index;
Understanding of regional air pollution over China using CMAQ, part II. Process analysis and sensitivity of ozone and particulate matter to precursor emissions by Xiao-Huan Liu; Yang Zhang; Jia Xing; Qiang Zhang; Kai Wang; David G. Streets; Carey Jang; Wen-Xing Wang; Ji-Ming Hao (3719-3727).
Following model evaluation in part I, this part II paper focuses on the process analysis and chemical regime analysis for the formation of ozone (O3) and particulate matter with aerodynamic diameter less than or equal to 10 μm (PM10) in China. The process analysis results show that horizontal transport is the main contributor to the accumulation of O3 in Jan., Apr., and Oct., and gas-phase chemistry and vertical transport contribute to the production and accumulation of O3 in Jul. Removal pathways of O3 include vertical and horizontal transport, gas-phase chemistry, and cloud processes, depending on locations and seasons. PM10 is mainly produced by primary emissions and aerosol processes and removed by horizontal transport. Cloud processes could either decrease or increase PM10 concentrations, depending on locations and seasons. Among all indicators examined, the ratio of P HNO 3 / P H 2 O 2 provides the most robust indicator for O3 chemistry, indicating a VOC-limited O3 chemistry over most of the eastern China in Jan., NOx-limited in Jul., and either VOC- or NOx-limited in Apr. and Oct. O3 chemistry is NOx-limited in most central and western China and VOC-limited in major cities throughout the year. The adjusted gas ratio, AdjGR, indicates that PM formation in the eastern China is most sensitive to the emissions of SO2 and may be more sensitive to emission reductions in NOx than in NH3. These results are fairly consistent with the responses of O3 and PM2.5 to the reductions of their precursor emissions predicted from sensitivity simulations. A 50% reduction of NOx or AVOC emissions leads to a reduction of O3 over the eastern China. Unlike the reduction of emissions of SO2, NOx, and NH3 that leads to a decrease in PM10, a 50% reduction of AVOC emissions increases PM10 levels. Such results indicate the complexity of O3 and PM chemistry and a need for an integrated, region-specific emission control strategy with seasonal variations to effectively control both O3 and PM2.5 pollution in China.
Keywords: CMAQ; Process analysis; Indicators for O3 and PM2.5 chemistry; China;
Transport of dusts from East Asian and non-East Asian sources to Hong Kong during dust storm related events 1996–2007 by Y.C. Lee; Xun Yang; Mark Wenig (3728-3738).
Over a twelve year period from 1996 to 2007, 76 dust storm related events (as days) in Hong Kong were selected for study, based on Aluminium and Calcium concentrations in PM10. Four of the 76 events reach episodic levels with exceedances of the Hong Kong air quality standards. The purpose of the study is to identify and characterize dust sources impacting Hong Kong.Global distribution of aerosols in NASA’s daily aerosol index images from TOMS and OMI, are compared to plots generated by NRL(US)’s Navy Aerosol Analysis and Prediction System. Possible source areas are assigned by computing air parcel backward trajectories to Hong Kong using the NOAA HYSPLIT model. PM10 and elemental data are analyzed for crustal mass concentrations and element mass ratios.Our analysis reveals that 73 out of the 76 dust events (96%) involve non-East Asian sources-the Thar, Central/West Asian, Arabian and Sahara deserts (Saharan influence is found in 63 events), which are previously not known to affect Hong Kong. The Gobi desert is the most frequent origin of dust, affecting 68 dust events while the Taklamakan desert impacts only 30 of the dust events. The impact of the Gobi desert in March and December is apparently associated with the northeast monsoon in East Asia.Our results also show a seasonal pattern in dust impact from both East Asian and more remote sources, with a maximum in March. Dust event occurrences are conspicuously absent from summer. Dust transport to Hong Kong is commonly associated with the passage of frontal low-pressure systems.The coarse size fraction of PM10 concentrations were, as indicated by Al, Ca and Fe concentrations, about 4–8 times higher during dust events. The mean Ca/Al ratios of sources involving the Taklamakan desert are notably higher than those for non-East Asian sources owing to a higher Ca content of most of the East Asian deserts. The Fe/Al ratios follow a similar trend.Contributions from the desert sources are grossly estimated where possible, by using the average Al abundance of 8% in the upper continental crust to convert the Al mass in the PM10 to dust concentrations. This is done for the six events identified with air mass purely of non-East Asian origin and the two events related only to the Thar/Arabian/Sahara deserts. Results reveal that the average contribution from the non-East Asian sources (including C/W Asia) is approximately 10% and, that from the Thar/Arabian/Sahara deserts is about 8%.
Keywords: Desert dust transport; Hong Kong; Non-East Asian sources; Seasonality; Estimated contribution;