Atmospheric Environment (v.37, #24)

A single framework is developed for describing the equilibrium gas/particle partitioning of organic and inorganic constituents wherein all partitioning molecules are handled in the same manner, and multiple phases may be present in the particulate matter (PM). The model assumes an equilibrium distribution of all constituents between the gas phase and all of the condensed PM phases. The partitioning species may include neutral non-ionizing compounds (e.g., alkanes) as well as acids, bases, and metal ions. Acid-related species are assumed to be volatile only when present in the neutral “free-acid” form. Base-related species are assumed to be volatile only when present in the neutral “free-base” form. Metal ions are assumed to be non-volatile. The behavior of systems involving multiple liquid and solid phases are discussed using analogs of equations that have been developed, tested, and verified in prior efforts. Schemes are discussed for seeking the overall equilibrium composition of a gas and multi-phasic PM system. The framework developed thus provides a basis for solving a wide variety of complex partitioning problems. Because effective algorithms exist for predicting liquid-phase activity coefficients for neutral molecules as a function of composition and temperature, the framework can currently be used to solve a wide variety of gas/particle partitioning problems that involve non-ionizing constituents in a gas and multi-phasic PM system. Application of the framework to more general problems involving both ionic constituents and neutral molecules in liquid PM will become possible when reliable algorithms are developed for predicting activity coefficients of ions and neutral molecules in a general liquid mixture of such species.
Keywords: Aerosol modelling; Particulate matter; PM; Gas/particle partitioning; Multi-phase PM;

Intercomparison and evaluation of four semi-continuous PM2.5 sulfate instruments by F. Drewnick; J.J. Schwab; O. Hogrefe; S. Peters; L. Husain; D. Diamond; R. Weber; K.L. Demerjian (3335-3350).
The development and evaluation of time-resolved (minutes) measurement technologies to characterize the physical and chemical make up of ambient aerosols/particulate matter in the atmosphere are essential to our improved understanding of aerosol process science, source attribution, and population exposure. During the PMTACS-NY summer 2001 campaign in Queens/New York, a wide variety of on-line aerosol analysis instruments were deployed together with gas-phase and filter sampling techniques. Here, we report on the intercomparison of four semi-continuous PM2.5 sulfate instruments and evaluation of these instruments with one set of 6 h and three sets of 24-h filter measurements, collected at the same site. The semi-continuous instruments were an aerosol mass spectrometer, a particle-into-liquid sampler coupled with ion chromatograph, a Rupprecht & Patashnick Sulfate Monitor (R&P 8400S), and a continuous sulfate monitor developed by George Allen at Harvard School of Public Health and built in the field by one of us (J.J.S.). We found an excellent almost one-to-one correlation between the four semi-continuous instruments with typical multiple R-squared values >0.9. In addition, the correlations of the semi-continuous data with the filter measurements are also highly linear (R 2 0.86–0.98) but the semi-continuous instruments recover only about 85% of the sulfate mass collected by the filter techniques. The most likely explanation for this deviation is a combination of positive sampling artifacts on the filters (collection of particles with diameter greater then 2.5 μm, oxidation/condensation processes on the filters) with negative biases of the semi-continuous measurements (inlet line losses, limited collection efficiency for small particles below ca. 0.1 μm).
Keywords: Particulate matter; Sulfate; Air quality; Air pollution; Instrument methods;

Photochemistry in the arctic free troposphere: NO x budget and the role of odd nitrogen reservoir recycling by Craig Stroud; Sasha Madronich; Elliot Atlas; Brian Ridley; Frank Flocke; Andy Weinheimer; Bob Talbot; Alan Fried; Brian Wert; Richard Shetter; Barry Lefer; Mike Coffey; Brian Heikes; Don Blake (3351-3364).
The budget of nitrogen oxides (NO x ) in the arctic free troposphere is calculated with a constrained photochemical box model using aircraft observations from the Tropospheric O3 Production about the Spring Equinox (TOPSE) campaign between February and May. Peroxyacetic nitric anhydride (PAN) was observed to be the dominant odd nitrogen species (NO y ) in the arctic free troposphere and showed a pronounced seasonal increase in mixing ratio. When constrained to observed acetaldehyde (CH3CHO) mixing ratios, the box model calculates unrealistically large net NO x losses due to PAN formation (62 pptv/day for May, 1–3 km). Thus, given our current understanding of atmospheric chemistry, these results cast doubt on the robustness of the CH3CHO observations during TOPSE. When CH3CHO was calculated to steady state in the box model, the net NO x loss to PAN was of comparable magnitude to the net NO x loss to HNO3 (NO2 reaction with OH) for spring conditions. During the winter, net NO x loss due to N2O5 hydrolysis dominates other NO x loss processes and is near saturation with respect to further increases in aerosol surface area concentration. NO x loss due to N2O5 hydrolysis is sensitive to latitude and month due to changes in diurnal photolysis (sharp day–night transitions in winter to continuous sun in spring for the arctic). Near NO x sources, HNO4 is a net sink for NO x ; however, for more aged air masses HNO4 is a net source for NO x , largely countering the NO x loss to PAN, N2O5 and HNO3. Overall, HNO4 chemistry impacts the timing of NO x decay and O3 production; however, the cumulative impact on O3 and NO x mixing ratios after a 20-day trajectory is minimal.
Keywords: Atmospheric chemistry; Arctic troposphere; TOPSE; Peroxynitric acid; Acetaldehyde;

Tobacco smoking simulation experiment was conducted in a test room under different conditions such as cigarette brands, smoking number, and post-smoke decay in forced ventilation or in closed indoor environments. Thirty-seven chemical species were targeted and monitored, including volatile organic compounds (VOCs) and environmental tobacco smoke (ETS) markers. The results indicate that benzene, d-limonene, styrene, m-ethyltoluene and 1,2,4/1,3,5-trimethylbenzene are correlated well with ETS markers, but toluene, xylene, and ethylbenzene are not evidently correlated with ETS markers because there are some potential indoor sources of these compounds. 2,5-dimethylfuran is considered to be a better ETS marker due to the relative stability in different cigarette brands and a good relationship with other ETS markers. The VOCs concentrations emitted by tobacco smoking were linearly associated with the number of cigarettes consumed, and different behaviors were observed in closed indoor environment, of which ETS markers, d-limonene, styrene, trimethylbenzene, etc. decayed fast, whereas benzene, toluene, xylene, ethylbenzene, etc. decayed slowly and even increased in primary periods of the decay; hence ETS exposure in closed environments is believed to be more dangerous. VOCs concentrations and the relative percentage constituent of ETS markers of different brand cigarettes emissions vary largely, but the relative percentage constituent of ETS markers for the same brand cigarette emissions is similar.
Keywords: Environmental tobacco smoke; Volatile organic compounds; ETS marker; 2,5-dimethylfuran;

The future changes of acid deposition characteristics in North East Asia are investigated by comparing the simulation results of the comprehensive acid deposition model for the year 1996 with those for the year 2020. The SO2 emissions are estimated to increase less than 20% until 2020 in most of the North East Asian regions, whereas much greater emission increases are expected for NO x . The calculated future surface concentrations of the primary pollutants such as SO2 and NO x show a linear response to their future emission increases while being smoothed by the transport effects. In contrast, the calculated future surface concentrations of the secondary pollutants such as sulfate and HNO3 respond non-linearly to the future emission changes of the corresponding primary pollutants.The ratio of future wet deposition to dry deposition is calculated to decrease for sulfate but to increase for nitrate, implying that the importance of dry deposition will increase for sulfate but decrease for nitrate in 2020 compared to 1996. In addition, the S/N ratio of the wet and dry deposition is predicted to decrease substantially in 2020 due to the aggressive control policies on SO2 emissions expected in the region. Finally, the long-range transport from China to Korea is estimated to be less important for sulfur-containing acids but more important for nitrogen-containing acids in 2020 compared to 1996.
Keywords: Future acid deposition; North East Asia; Wet and dry deposition; Sulfur and nitrogen ratio; Long-range transport;

Semi-volatile compounds distribute mass between gas and particle phases in air. Gas–particle partition coefficients [K p=(F/TSP)/A] are used to predict the fate and transport of these compounds in the atmosphere. Measurements of gas–particle partition coefficients are biased if: (1) a fraction of mass is not quantified due to mass penetration through the sampler or reactions of collected species during sampling, (2) a fraction of particle-phase mass is measured erroneously as gas-phase, or (3) a fraction of gas-phase mass is measured erroneously as particle-phase. The latter two biases can lead to serious errors in measured partition coefficients. Generalized equations were developed to predict the impact of these errors for three sampling methods: filter–adsorbent, filter–filter–adsorbent, and denuder–filter–adsorbent. Two cases were considered: (1) the level of bias is equal for compounds of varying K p and (2) the sampling bias is compound dependent and, hence, varies with K p. These artifacts cause deviations to the slope and intercept in log–log plots of compound partition coefficients vs. subcooled, liquid vapor pressures that are similar to the deviations seen from non-equilibrium effects, temperature and concentration changes, and the presence of non-exchangeable material in the particle.
Keywords: Filter; Semi-volatile; Denuder; Adsorption; Evaporation; Artifact; K p;

A photocatalytic oxidation (PCO) reactor model was developed to analyze the removal of volatile organic compounds (VOCs) in indoor air. Two parameters, the fractional conversion, ε, and the number of mass transfer units (NTUm), were found to be the main parameters influencing the photooxidation performance of PCO reactors. The factors of affecting the effect of removing VOCs of PCO reactor are discussed. The predicted photocatalytic reaction rate coefficient, fractional conversion, ε, and NTUm for a typical PCO reactor agreed well with experimental data. The study shows that the effectiveness—NTU method used for heat exchangers is also an effective tool for designing PCO reactors and for evaluating their VOC removal effect.
Keywords: Indoor air quality; Titania; Nanometer material; Photooxidation; Volatile organic compounds;

Urban aerosol samples for PM10 and PM2.5 were collected during summer (August) and winter (December) 2000 in southern Taiwan (Tainan City) to demonstrate the temporal variations of Hg and As in particulate matter (PM). The mean mass concentrations with standard deviations were 80.0±26.8 μg m−3 for PM10 and 50.6±16.6 μg m−3 for PM2.5. The average PM2.5/PM10 mass ratio for the two periods combined was 63%, indicating that fine particles were a large portion of PM10. Particulate samples of Hg and As were analyzed within 2 days following sampling and weighing, because of the highly volatile nature of PM Hg and As. The average Hg and As values in PM10, PM2.5 and PM2.5–10 in summer were significantly lower than those in winter. PM2.5 Hg constituted 0.34 to 5.8 ng m−3 and PM2.5–10 Hg 0.05 to 3.1 ng m−3. PM2.5 As constituted 1.09 to 9.51 ng m−3 and PM2.5–10 As 0.18 to 4.14 ng m−3. In summer and winter PM10, the Hg contents showed regular daily variation, with the higher values at daytime and lower values at nighttime, indicating conversion of gaseous Hg to the particulate phase by reaction with atmospheric oxidants under strong solar radiation during the daytime in both summer and winter. PM As behaved similar to Hg in the summer, but in the winter higher concentrations were observed during the nighttime than during the daytime, implying that the stable temperature inversion during winter nighttime caused the accumulation of PM As near the ground. In summer, SE–WSW winds carried As from an As-emitting fossil power plant to the sample area. In a similar vein, NE–WNW winter winds contributed to aerosol Hg, especially in PM2.5, originating from a waste incinerator located NW of Tainan City.
Keywords: Arsenic; Mercury; PM10; PM2.5; Temporal variations; Storage;

Temperature dependence of secondary organic aerosol formation by photo-oxidation of hydrocarbons by Hideto Takekawa; Hiroaki Minoura; Satoshi Yamazaki (3413-3424).
Photo-oxidation experiments on hydrocarbons were performed with a temperature-controlled smog chamber to study the temperature dependence of secondary organic aerosol (SOA) formation. A higher SOA yield was obtained at lower temperature and with a higher concentration of SOA generated. The relationship of SOA yield to temperature and SOA concentration is expressed by a gas/particle partitioning absorption model considered with temperature dependence. Under the condition of the same SOA concentration, the SOA yield at 283 K was approximately twice that at 303 K. It has been clarified experimentally that temperature is one of the most important factors in SOA formation. The experiments were performed not only with three aromatic hydrocarbons (toluene, m-xylene and 1,2,4-trimethylbenzene) and one biogenic alkene (α-pinene), but also with one alkane (n-undecane) on which few experiments for SOA formation have been performed. n-Undecane indicates a lower SOA yield than any other hydrocarbon investigated in this study.
Keywords: Gas/particle partitioning; Aerosol yield; Secondary organic aerosol; Aromatic hydrocarbons; α-Pinene; n-Undecane;

On heavy dustfall observed with explosive sandstorms in Chongwon-Chongju, Korea in 2002 by Yong-seung Chung; Hak-sung Kim; Jugder Dulam; Joyce Harris (3425-3433).
Continuous monitoring of sand and duststorms (SD) and associated heavy dustfall (HD) is made in Korea. In particular, accurate measurements of atmospheric dust loadings have been carried out with the tapered element oscillating microbalance method, and satellite detection of dust clouds is included in the analysis. In 2002, we found three gigantic dust clouds that moved over the Korean Peninsula, and associated HD occurred with PM10 values of 1106–3006 μg m−3. In Beijing, China much higher concentrations were recorded, while in SW Japan measured values were up to 986 μg m−3. Two SD occurred in March and April, while the third one occurred unusually in November. During the year, there were nine cases of reddish-brown sand with 18 dusty days. The intensity of HD was extraordinary for the recent decade. It was observed that with invading SD, the higher the PM10 values the lower the PM2.5 loadings. Also, variations of visibility were more depending on PM2.5 variations than PM10 values. It is shown that satellite detection is a useful technique in monitoring SD and HD.
Keywords: Gigantic sandstorms; East Asian duststorms; Satellite dust detection; PM10 and PM2.5 measurements; Heavy dustfall; Reddish-brown sand;

Maximum likelihood cost functions for neural network models of air quality data by Stephen R Dorling; Robert J Foxall; Danilo P Mandic; Gavin C Cawley (3435-3443).
The prediction of episodes of poor air quality using artificial neural networks is investigated, concentrating on selection of the most appropriate cost function used in training. Different cost functions correspond to different distributional assumptions regarding the data, the appropriate choice depends on whether a forecast of absolute pollutant concentration or prediction of exceedence events is of principle importance. The cost functions investigated correspond to logistic regression, homoscedastic Gaussian (i.e. conventional sum-of-squares) regression and heteroscedastic Gaussian regression. Both linear and nonlinear neural network architectures are evaluated. While the results presented relate to a dataset describing the daily time-series of the concentration of surface level ozone (O3) in urban Berlin, the methods applied are quite general and applicable to a wide range of pollutants and locations. The heteroscedastic Gaussian regression model outperforms the other nonlinear methods investigated; however, there is little improvement resulting from the use of nonlinear rather than linear models. Of greater significance is the flexibility afforded by the nonlinear heteroscedastic Gaussian regression model for a range of potential end-users, who may all have different answers to the question: “What is more important, correctly predicting exceedences or avoiding false alarms?”.
Keywords: Neural network; Ozone; Modelling exceedences;

Air/sea gas exchange of PCBs in the southern Baltic Sea by Regina Bruhn; Soenke Lakaschus; Michael S McLachlan (3445-3454).
Air/sea gas exchange is an important process for the environmental fate of many semivolatile organic compounds (SOCs). The ability to quantify this flux can be very useful, e.g. for understanding the role of seas as a sink or a source for SOCs. In this work the air/sea gas exchange of polychlorinated biphenyls (PCBs) was investigated in the Baltic Sea, a particularly contaminated marine ecosystem. Air and water samples were collected during three cruises in the southern Baltic Sea and analysed for eight PCB congeners. Water/air fugacity ratios were calculated using literature values for the air/water partition coefficients. A sensitivity analysis indicated that it was not possible to identify the direction of net air/sea exchange of gaseous PCBs with 95% certainty due to the uncertainty in the air/water partition coefficients. Even when this uncertainty was reduced by using a recent set of internally consistent physical–chemical properties of PCBs, the direction of the net flux was only ascertainable for PCB 28/31. The uncertainty in the fugacity ratio rendered a quantification of the air/sea exchange flux impossible. Nevertheless, the observation that the PCBs in the air/sea system were relatively close to a partitioning equilibrium was useful in identifying discrepancies and weaknesses in existing estimates of the PCB mass balance for the Baltic Sea. Furthermore, the observation of lower water/air fugacity ratios for the more lipophilic PCB congeners suggests that vertical export with particulate organic material can cause a fugacity deficit of lipophilic substances in surface water of the Baltic Sea.
Keywords: Air/sea gas exchange; PCB; Baltic Sea;

Intake fraction of primary pollutants: motor vehicle emissions in the South Coast Air Basin by Julian D Marshall; William J Riley; Thomas E McKone; William W Nazaroff (3455-3468).
The intake fraction is defined for a specific species and emission source as the ratio of attributable population intake to total emissions. Focusing on California's South Coast Air Basin (SoCAB) as a case study, we combine ambient monitoring data with time-activity patterns to estimate the population intake of carbon monoxide and benzene emitted from motor vehicles during 1996–1999. In addition to exposures to ambient concentrations, three microenvironments are considered in which the exposure concentration of motor vehicle emissions is higher than in ambient air: in and near vehicles, inside a building that is near a freeway, and inside a residence with an attached garage. Incorporating data on motor vehicle emissions estimated by the EMFAC2000 model, we estimate that the 15 million people in the SoCAB inhale 0.003–0.009% (34–85 per million, with a best estimate of 47 per million) of primary, nonreactive compounds emitted into the basin by motor vehicles. This population intake of primary motor vehicle emissions is approximately 50% higher than the average ambient concentration times the average breathing rate, owing to higher concentrations in the three microenvironments and also to the temporal and spatial correlation among breathing rates, concentrations, and population densities. The approach demonstrated here can inform policy decisions requiring a metric of population exposure to airborne pollutants.
Keywords: Exposure assessment; Microenvironment; Carbon monoxide; Benzene;