Atmospheric Environment (v.40, #1)

Wind fields and turbulence statistics in an urban street canyon by I. Eliasson; B. Offerle; C.S.B. Grimmond; S. Lindqvist (1-16).
This is the first paper of a long-term measurement campaign to explore wind, temperature, radiation and energy fields within an urban canyon. A canyon and a rooftop mast were installed in a canyon with an aspect ratio (Height/Width) of ∼2.1 in Göteborg, Sweden. A number of instruments including sonic anemometers, radiometers and thermocouples were mounted in vertical profiles and across the width of the canyon. The experimental set-up, the characteristics of the canyon flow pattern and mean and turbulence statistics with respect to above canyon flow are examined using data collected under clear-sky conditions in summer and autumn 2003. Results show that under cross-canyon (within 60° of orthogonal) flow, a single helical vortex exists. High temporal resolution analysis suggests that eddies frequently penetrate the shear stress layer at the canyon top disrupting established flow patterns. A combination of complex building roof shapes and local topography may contribute to this effect by maintaining a high degree of turbulence. The profile of mean wind speed within the canyon and the relation with that above canyon depends on the ambient flow direction in relation to the canyon long axis. Turbulence statistics show results similar to other field studies, with turbulence kinetic energy and vertical mixing greatest toward the windward wall.
Keywords: Urban canyon flow; Turbulence; Vortex;

Development of a photo-fragmentation/laser-induced fluorescence measurement of atmospheric nitrous acid by Wei Liao; Arsineh Hecobian; Joe Mastromarino; David Tan (17-26).
Atmospheric nitrous acid (HONO) is an integral part of HO x (OH and HO2) and NOx (NO and NO2) cycles. In the arctic atmosphere measurements of HONO are an important part of understanding the dynamics of snow-air chemistry and atmospheric photochemistry. The low levels of HONO present in such regions necessitate the development of instrumentation with low detection limits. An improved method of detecting HONO is introduced in this paper, using photo-fragmentation and laser-induced fluorescence. HONO is photo-fragmented into OH and NO at 355 nm. OH is then probed using a wavelength of 282 nm and the fluorescence of OH is observed at 309 nm. The instrument is calibrated using HONO produced from the reaction of sulfuric acid and sodium nitrite. The detection limit of this method is 15 pptv for 1-min integration time with 35% uncertainty.
Keywords: HONO; NO y ; Photochemistry; LIF; Trace gas measurement;

Organic marker compounds for surface soil and fugitive dust from open lot dairies and cattle feedlots by Wolfgang F. Rogge; Patricia M. Medeiros; Bernd R.T. Simoneit (27-49).
Fugitive dust emissions from cattle feedlots and open lot dairies are substantial. In order to determine the contribution of intensive cattle operations on ambient PM levels, more knowledge besides the elemental composition is necessary in order to distinguish between airborne PM from nearby agricultural fields, barren lands, or dirt roads. Here, as part of the San Joaquin Valley Fugitive Dust Characterization Study, surface soil samples collected from feedlots and open lot dairy farms are investigated for potential source specific molecular marker compounds. More than 100 organic compounds were quantified including: n-alkanes, n-alkanoic acids, n-alkenoic acids, n-alkanols, n-alkanals, n-alkan-2-ones, steroids, triterpenoids, isoprenoids, and tocopherols (vitamin E) and metabolites. Biohydrogenation of plant lipids and sterols in the rumen results in distinctive alteration products. Animal and plant derived steroids are most abundant. Here, it is shown that 5β-stigmastanol and epi-5β-stigmastanol, two biohydrogenation products of sitosterol and stigmasterol, are the most distinctive molecular marker compounds. While stearic (C18) and palmitic (C16) acids are as individual compounds not source specific, biohydrogenation of the more abundant C18 unsaturated fatty acids, causes the ratio of C18/C16 fatty acids to shift from below 0.5 for vegetation to an average of 3.0±0.7. Consequently, the C18/C16 fatty acid ratio is unique and can be used as well in source apportionment studies.
Keywords: Concentrated animal feeding operations; Tocopheryl acetate metabolites; 5β-stigmastanol; epi-5β-Stigmastanol; Fatty acids; epi-Coprosteroids; Coprostanol; epi-Coprostanol; Phytosterols; Phytostanols;

Trace-gas mixing in isolated urban boundary layers: Results from the 2001 Phoenix sunrise experiment by Carl M. Berkowitz; J. Christopher Doran; William J. Shaw; Stephen R. Springston; Chester W. Spicer (50-57).
Measurements made from surface sites, from the 50 and 140 m levels (the 16th and 39th floors) of a skyscraper, and from an instrumented aircraft are used to characterize early morning profiles of CO, NO y , and O3 within the mid-morning summertime convective atmospheric boundary layer (CABL) over Phoenix, Arizona. Although mixing was anticipated to produce uniform values of these species throughout the CABL, this was found not to be the case. Background air advected into the upper levels of the boundary layer and entrained air from above appear to be the most likely cause for the lack of well-mixed trace gases. The results show that surface measurements may provide only limited information on concentrations of trace-gas species higher in the boundary layer.
Keywords: Air pollution; CO; Ozone; No y ; Mixing; Convective atmospheric boundary layer;

In October 1957, an accident involving a fire in a nuclear reactor at Windscale Works, Sellafield, on the Cumbrian coast of England released a quantity of radioactive material into the atmosphere. Past studies have examined the resultant ground-level concentrations of radioactivity in the UK and the rest of Europe using trajectory models or the analysis of milk samples taken at the time. Presented here is the latest study into the levels of radioactivity attributed to these releases which has been undertaken using the Met Office's Nuclear Accident Model, NAME. The study examines the direction and timing of the transport, indicating the most likely spread within Europe of radioactive material from the accident. Best estimates are presented of likely time-integrated air concentrations and ground-level air concentrations of radioactivity throughout the initial few days after the accident. Although these estimates were not expected to be very accurate due to the source term being ill-defined, the results have been presented for completeness. The report concludes that NAME manages to successfully recreate the general movement of the radioactive plume and is in broad agreement with previous analyses. Discrepancies between modelled and measured plume arrival and activity for certain locations are primarily associated with errors in the calculation of the wind fields used by the model. Model predictions for this event could be improved by reducing the uncertainties in the release scenario and Met data used by NAME, and this possibility will be explored in future work.
Keywords: Dispersion model; Windscale; NAME; Radionuclides; Time-integrated air concentration (TIAC);

Methyl tert-butyl ether (MTBE) in snow samples in Germany by Axel Kolb; Wilhelm Püttmann (76-86).
In the present study, experimental results of methyl tert-butyl ether (MTBE) analysis in 43 snow samples from 13 various locations during two winter seasons are presented. MTBE could be detected in 65% of the samples. The concentrations ranged from 11 to 631 ng L−1. The comparison to formerly analysed rainwater samples indicates atmospheric transfer of MTBE from urban to rural areas preferentially in winter due to lower atmospheric degradation rates and suggests that MTBE is more effectively scavenged from the atmosphere by snow than by rain. This result is in contrast to calculated distribution coefficients K Rain/Air (−6.8°) and K Snow/Air (−6.8°) using different values for the specific snow surface area. MTBE concentrations measured in the collected snow samples from rural areas did not differ from MTBE concentrations in snow samples collected in the urban area of Frankfurt/M. The explanatory factor for this finding is also the lower atmospheric degradation rate of MTBE in winter. Particular meteorological conditions at the monitoring site Schmücke most probably explain maximum concentrations of 631 and 438 ng L−1 measured at this site. Decreasing MTBE amounts in snow samples collected at the monitoring site Schauinsland in January 2003 might have been due to increasing ambient temperatures. Varying MTBE amounts in snow samples collected at the Taunus Observatorium (Kleiner Feldberg) in February 2002 and in January/February 2003 cannot be explained by ambient temperatures. In future studies concurrent air sampling and morphological characterisation of the sampled snow would be required to clarify and confirm the results of the present study.
Keywords: MTBE; Snow; Organic contaminants; SPME;

Measurements of radon and its progeny concentrations, and conductivity due to both polarities were made simultaneously at a height of 1 m above the ground at Mysore (12°N, 76°E, 767 m above mean sea level), India. Observations show an increase in the concentration of radon and its progeny during nighttime compared to the daytime values with the peak values in the early morning (4–7 h), the observed values are higher in winter than in other seasons. The radon exhalation was observed to be higher in summer than in winter and rainy seasons. The electrical conductivity of the atmosphere that depends on the ionization rate also exhibits similar trends. These results are discussed in terms of ionization rate due to radioactivity and the influence of meteorological parameters on radon and its progeny concentrations.
Keywords: Radon; Exhalation; Ion production; Conductivity; Gerdien condenser;

Particle-bound polychlorinated dibenzo-p-dioxins and dibenzofurans in the atmosphere of Guangzhou, China by Liping Yu; Bixian Mai; Xiangzhou Meng; Xinhui Bi; Guoying Sheng; Jiamo Fu; Ping’an Peng (96-108).
A short-term sampling scheme was devised to determine the levels of particle-bound polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/Fs) in Guangzhou atmosphere and to evaluate the impact of PCDD/Fs on the lives of the inhabitants in Guangzhou. Samples were collected from four different districts of Guangzhou City, and the results indicated that industrial activities had tremendous influence on the levels of atmospheric PCDD/Fs in these districts. The mean PCDD/Fs concentrations (mean I-TEQ values) for Huadu, Liwan, Tianhe and Huangpu districts were 3815 fg m−3 (104.6 fg I-TEQ m−3), 12 777 fg m−3 (430.5 fg I-TEQ m−3), 6963 fg m−3 (163.7 fg I-TEQ m−3) and 10 953 fg m−3 (769.3 fg I-TEQ m−3), respectively, which were higher than the concentration ranges for other cities in the world. The mean concentrations of particulate organic carbon (OC) and elemental carbon (EC) were 17.61±0.94 and 3.17±0.21 μg m−3 for Huadu district, 44.18±2.29 and 13.32±0.75 μg m−3 for Liwan district, 33.53±1.72 and 9.89±0.56 μg m−3 for Tianhe district, and 29.52±1.53 and 8.87±0.53 μg m−3 for Huangpu district, respectively. The relationship between PCDD/Fs concentrations and OC and EC concentrations demonstrated that EC could be a better predictor for the concentration of PCDD/Fs. Detailed source analysis indicated that the PCDD/Fs in Guangzhou were derived from small diffuse combustion sources, e.g., traffic sources, domestic burning of fossil fuels, non-industrial combustion sources and industrial combustion sources. Results of daily intake of PCDD/Fs by inhaling air suggested that the inhalation exposure of PCDD/Fs by the inhabitants in Liwan district is relatively high.
Keywords: PCDD/Fs; OC/EC; Sources; Inhalation; Guangzhou district;

The vertical pollutant dispersion is quite sensitive to the eddy diffusivity, K V . Therefore, good estimations of K V are essential for improving the predictive performance of Eulerian dispersion models; especially in urban areas where literature based K V correlations are not always accurate. Here, we present a methodology to obtain a more accurate, but site-specific, K V correlation. It is based on using artificial neural networks (ANN) to find the best K V function for a particular urban area by minimizing, in a least-squares sense, the difference between ambient measurements of carbon monoxide and dispersion simulations of this tracer species. The resulting ANN-K V correlation is a function of three parameters namely, the stability parameter ( z / L ), the height within the mixing layer ( z / h ), and the scaled height ( z f C / u * )–hence the Monin–Obukhov ( L ), mixing ( h ) and Ekman ( u * / f C ) lengths are used to predict K V across the atmospheric boundary layer.We then assess how such an ANN-K V model improves the capability of a dispersion model (CAMx) to predict peak concentrations of ambient carbon monoxide in a large city. The evaluation has been performed with a set of eight air-quality meteorological stations evenly spread across the city of Santiago, Chile, during springtime. Results show that with the ANN-K V model, CAMx achieved better predictions of peak CO concentration levels than has been hitherto possible. Typically root-mean-square errors are reduced to half their original values. The resulting ANN-K V model—without any additional training—was then used to predict CO ambient concentrations at another period (summertime) and also to predict ambient concentrations of total carbon (PM2.5) at both periods. A much-improved agreement was observed. Furthermore, the ANN formulation allowed for the quality of the urban emission inventory to be critically assessed indicating that the weekend emissions in Santiago are most likely underestimated.
Keywords: Artificial neural network; Vertical eddy diffusivity; Turbulent transport; Urban air pollution; Model performance;

Dependence of ozone sensitivity analysis on grid resolution by Daniel S. Cohan; Yongtao Hu; Armistead G. Russell (126-135).
We examine the extent to which the sensitivity of ground-level ozone concentrations to precursor emissions varies with grid resolution in a regional air quality model. Nested grids of 36-, 12- and 4-km resolution are applied to model an air pollution episode in Georgia. A direct sensitivity analysis method is used to compute the response of ozone to emissions of its precursors, nitrogen oxides (NO x ) and volatile organic compounds (VOCs). All three grids predict that ozone production is limited primarily by the availability of NO x , and yield similar predictions of average ozone sensitivity to both regional and local emissions of NO x . However, in urban areas, the coarsest (36-km) resolution fails to capture the extent and magnitude of VOC sensitivity and underpredicts the nonlinearity of ozone response indicated by the finest resolution. Medium (12-km) resolution is sufficient to reproduce most features of ozone response, except for the texture of local variability. Despite the consistency of average results, discrepancies in wind fields at times lead to inconsistencies in the predicted location of ozone enhancements, especially for a power plant plume.
Keywords: Decoupled direct method (DDM); Photochemical modeling; Regional air pollution; Nitrogen oxides (NO x ); Nonlinearity;

A study on trace elemental composition of atmospheric aerosols at a semi-arid urban site using ICP-MS technique by P. Chandra Mouli; S. Venkata Mohan; V. Balaram; M. Praveen Kumar; S. Jayarama Reddy (136-146).
A study has been carried out on trace elemental composition of atmospheric aerosols (PM10) over a 1-year period i.e. from October 2001 to September 2002 at a semi-arid urban site, Tirupati, southern peninsular India. The samples were collected on paliflex filters with a frequency once a week using Mini-Partisol air sampler equipped with PM10 inlet. Aerosol loaded filters were extracted by the hot extraction method and the elemental concentrations were determined using inductively coupled plasma mass spectrometry (ICP-MS). Detection limits were found to be in a sub-ppt range and the precision about <5% RSD with comparable levels of accuracy. ICP-MS results are in good agreement with the certified data (<20% RSD). Elemental composition of PM10 showed a higher contribution of Fe followed by Mn and the lowest contribution of Be. Multivariate receptor model results (PC matrix) demonstrated that the trace elemental components of atmospheric aerosol were largely contributed from crustal material and road dust (69.41%), followed by metallurgical and other industrial processes (11.76%) and fuel oil combustion (6.52%). Regression models suggested the significant impact of meteorological factors on elemental concentration, which ranged between 14% and 53.6%. ANOVA results showed a clear trend of atmospheric levels of elements with variation in climatological conditions.
Keywords: Aerosol; Trace elements; ICP-MS; Sources; Meteorological factors; India;

Emission rates of regulated pollutants from on-road heavy-duty diesel vehicles by Sandip D. Shah; Kent C. Johnson; J. Wayne Miller; David R. Cocker (147-153).
Emissions from heavy-duty diesel (HDD) vehicles are affected by many factors. Changes in engine technology, operating mode, fuel properties, vehicle speed and ambient conditions can have significant effects on emission rates of regulated species. This paper presents the results of on-road emissions testing of 11 HDD vehicles (model years 1996–2000) over the ARB Four Phase driving schedule and the urban dynamometer driving schedule (UDDS). Emission rates were found to be highly dependent on vehicle operating mode. Per mile NO x emission rates for vehicle operation at low speeds, in simulated congested traffic, were three times higher per mile emissions then while cruising on the freeway. Comparisons of NO x emission factors to EMFAC baseline emission factors were within 5–40% for vehicles of various model years tested over the UDDS. A comparison of NO x emission factors for a weighted average of the ARB four phase driving schedule yielded values within 17–57% of EMFAC values. Generally, particulate matter (PM) emission rates were lower than EMFAC values.
Keywords: Diesel emissions; NO x ; Oxides of nitrogen; On-road emissions; Regulated pollutants; Heavy-duty diesel;

Vehicular traffic is an ubiquitous source of air pollution in developed nations, yet relatively few epidemiology studies have considered its long-term health effects. This paper uses an areal measure of traffic density as a surrogate index of exposure to vehicular traffic. We present associations between county-level traffic density (annual vehicle-km traveled km−2), ambient air quality, and mortality in a cohort of about 70,000 male US veterans (the Washington University-EPRI Veterans Cohort) who were enrolled in 1976 and followed through 2001. Traffic density is seen to be a significant and robust predictor of survival in this cohort, more so than ambient air quality, with the possible exception of ozone. Stronger effects of traffic density are seen in the counties that have ambient air quality monitoring data, which also tend to have higher levels of traffic density. These proportional-hazard modeling results indicate only modest changes in traffic-related mortality risks over time, from 1976–2001, despite the decline in regulated tailpipe emissions per vehicle since the mid-1970s. This suggests that other environmental effects may be involved, such as particles from brake, tire, and road wear, traffic noise, psychological stress, and spatial gradients in socioeconomic status.
Keywords: Traffic; Air quality; Exposure; Mortality; Cohort study;

Since 1987, the Environmental Protection Service of Environment Canada has operated a field program for measuring 1,3-butadiene and other volatile organic compounds (VOCs) in ambient air. With the co-operation of provincial and municipal environmental agencies, samples have been collected at urban and rural monitoring sites across the country. Samples were collected in evacuated canisters and analyzed by gas chromatography with a mass-selective detector. Using data from all sites, the composite average 1,3-butadiene concentration at urban sites in Canada for the years 1995–2003 was found to be 0.22 μg m−3 and the composite median was 0.17 μg m−3. In general, 1,3-butadiene concentrations were higher at commercial and industrial sites than at residential sites, particularly at sites close to street level in the urban core. The highest 24-h 1,3-butadiene concentration of 2.58 μg m−3 was measured at the Sarnia, Ontario, site in 2001; this site was significantly influenced by industrial point sources. Mean 24-h 1,3-butadiene concentrations at rural sites were typically an order of magnitude lower than at urban sites. 1,3-Butadiene concentrations varied by season, particularly in western Canada, and by day of week. Concentrations of 1,3-butadiene decreased from 1995 to 2003 in response to reduced rates of emissions from mobile and point sources. At most urban monitoring sites, the concentrations of 1,3-butadiene and acetylene were highly correlated; this supports the hypothesis that 1,3-butadiene is emitted primarily by mobile sources.
Keywords: 1,3-Butadiene; Acetylene; Ambient concentrations; Seasonal; Trends; Sources;

This study presents partitioning data from eight locations in the Laurentian Great Lakes region collected by the Integrated Atmospheric Deposition Network (IADN) over periods ranging from 1 to 6 years. Particle/gas partitioning varies sufficiently between sites in the Great Lakes region to preclude the use of a uniform temperature dependence for its description. Site-specific parameters for describing partitioning as a function of inverse temperature are presented.Relationships between partitioning of appreciably semivolatile PAHs and saturated vapour pressure at Chicago (IIT) and Sturgeon Point (STP) demonstrate that anthracene, benz[a]anthracene and retene behave differently than phenanthrene, fluoranthene, pyrene and chrysene+triphenylene. Possible reasons for these differences include interspecies variations in the fraction of atmospherically non-exchangeable, though analytically extractable, PAHs on particles and differences in soot–air partition coefficients as they relate to saturated vapour pressure. The observed interspecies differences are not consistent with sampling artefacts such as filter adsorption or sorbent breakthrough. At IIT, but not at STP, values of the slope of the relationship between the log partition coefficient and log vapour pressure vary in a manner opposing the annual temperature cycle. A comparison of partitioning calculated by a combined absorption/adsorption model shows good predictability at Chicago but underestimates values at a rural site (Eagle Harbor, EGH) by an order of magnitude.
Keywords: Polycyclic aromatic hydrocarbons; Partitioning; Absorption; Adsorption; Soot;