Atmospheric Environment (v.40, #15)

The Pacific 2001 Air Quality Study was undertaken to characterize the physical and chemical properties of particulate matter in the Lower Fraser Valley (LFV) airshed of British Columbia. Aerosol transport in the LFV was observed to occur through several regional scale mechanisms, including seabreeze–landbreeze flows, upslope–downslope flows and north–south valley flows. Enhanced aerosol concentrations were observed in the eastern portion of the LFV, the north–south tributary valleys and the Gulf/San Juan islands convergence zone. Particle composition was found to vary according to proximity to emission sources relative to surface flows. Approximately half of the particle mass was comprised of organic carbon, with the rest being inorganic species. Sodium nitrate was an important component of the coarse particle fraction, while ammonium sulphate was concentrated in the fine fraction. A broad suite of organic substances were detected, and both anthropogenic and biogenic sources were found to contribute to secondary organic aerosol formation. Particle formation and growth was observed to occur via nucleation, condensation and coagulation. In general, the western part of the LFV airshed was dominated by combustion-related compounds, sea salt chemistry and organics (both anthropogenic and biogenic), while the eastern part of the airshed was dominated by NH3 chemistry and biogenic organics. The extent of processing of air masses was found to increase in a west to east direction, although the occasional accumulation of photochemically aged pollutants in the Gulf/San Juan islands convergence zone was observed to reverse this gradient. NH3 was found not to be a limiting species in the formation of fine mode inorganic aerosols, suggesting that NH3 emission reductions would have to be substantial in order to confer a significant improvement in PM and visibility. This coupled with the non-linear relationships between NO x , VOC and secondary particulates, suggests that PM and visibility improvement in the LFV would most likely require concurrent reductions in NH3, SO x , NO x and VOC emissions.
Keywords: Particulate matter; Inorganic aerosols; Organic aerosols; Secondary organic carbon; Visibility;

Hygroscopicity of particles at two rural, urban influenced sites during Pacific 2001: Comparison with estimates of water uptake from particle composition by Y. Aklilu; M. Mozurkewich; A.J. Prenni; S.M. Kreidenweis; M.R. Alfarra; J.D. Allan; K. Anlauf; J. Brook; W.R. Leaitch; S. Sharma; H. Boudries; D.R. Worsnop (2650-2661).
Hygroscopicity of particles was measured at Langley (rural) and Eagle Ridge (semi-rural) as part of the Pacific 2001 field campaign. The measured growth factors at the two sites were comparable. However, differences in particle composition as measured by an Aerosol Mass Spectrometer were evident at these two sites. Sulphate mass concentration was found to be similar at the two sites, while higher nitrate and organic mass were observed at Eagle Ridge. Higher growth factors were observed when the air mass was impacted by SO2 sources, while lower growth factors were observed when the air mass was affected by urban emissions. To examine the hygroscopic role of the different particle components, expected growth factors were calculated from the composition data and compared to measured growth factors. Calculations were done using the Zdanovskii, Stokes and Robinson (ZSR) mixing rule. Sulphate fraction played a dominant role in particle hygroscopicity at both sites. Calculated growth factors were within the uncertainty of the measurements, except when the nitrate fraction was high. The results imply that particulate nitrate takes up much less water than ammonium nitrate, indicating that the ZSR mixing rule fails for nitrate. Small variations of organic growth factors with source regions suggest that secondary organic matter is more hygroscopic than primary organic matter.
Keywords: Hygroscopic properties; Growth factor; Particle organic matter; Particle composition; Field studies;

Ionic composition and size characteristics of particles in the Lower Fraser Valley: Pacific 2001 field study by Kurt Anlauf; Shao-Meng Li; Richard Leaitch; Jeff Brook; Katherine Hayden; Desiree Toom-Sauntry; Alan Wiebe (2662-2675).
The size distribution of aerosol chemical compositions was studied over a 2-week period in August 2001 at three locations (one urban and two semi-rural) in the Lower Fraser Valley (LFV) of British Columbia, Canada. The size distributions varied diurnally and were bimodal with a pronounced peak at about 0.3–0.55 μm (accumulation mode) and 2–7 μm (coarse mode). Generally, on an equivalent basis, the aerosol was dominated by SO4 2− (balanced by NH4 +) in the accumulation mode range; SO4 2− often exhibited a diurnal variation that was primarily the result of fog deposition overnight and photochemical formation during the day. SO4 2− in the 3.1–6.2 μm particles was mostly of marine origin. Most of the smaller particle SO4 2− was from anthropogenic sources and it is estimated that about 75% of the SO4 2− in the smaller particles were due to secondary processes. Oxalate, C2O4 2−, although at low concentrations (generally <0.1 SO4 2−), displayed a bi-modal size distribution, the accumulation mode being similar to SO4 2− and the coarse mode peaking at 1–1.8 μm. Frequently at night, there was also a significant increase in NO3 in the accumulation mode that is attributed to the co-condensation of gas phase NH3+HNO3 to form aerosol NO3 and the reactive uptake of N2O5 onto aerosols. In the coarse mode, the sea salt Cl was frequently deficient relative to Na+ yet accompanied by a corresponding increase in NO3 ; this was attributed to reaction of NO3 precursors (e.g., HNO3 and N2O5) with sea salt aerosol Cl.
Keywords: Ambient aerosols; Nitrate; Sulphate; Size distributions;

Aerosol sulphate and its oxidation on the Pacific NW coast: S and O isotopes in PM2.5 by Ann-Lise Norman; Kurt Anlauf; Katherine Hayden; Bruce Thompson; Jeffrey R. Brook; Shao-Meng Li; Jan Bottenheim (2676-2689).
Sources of sulphate and oxidation pathways important to the formation of PM 2.5 during the Pacific 2001 field campaign in Canada, were studied in diurnal samples from an urban coastal and a rural inland site in the Fraser Valley, B.C. Ion and elemental characteristics of aerosols as well as the sulphur and oxygen isotopes in sulphate were compared and related using multiple linear regressions. Sources of sulphate in PM2.5 were distinguished at each site and included: 1. sulphate from DMS oxidation during the day-time in Vancouver that coincided with high sodium and oxalate concentrations inland, 2. well-mixed pollutant SO2 from the Strait of Georgia that was oxidized to sulphate as it moved into the Vancouver area, 3. vehicle exhaust or a mixture of biomass burning and vehicle exhaust which has a larger fraction of primary sulphate than other sulphate sources and is possibly associated with Cu, and 4. isotopically light SO2 from a refinery in the US that was strongly associated with higher Zn concentrations in aerosols. Primary and secondary sulphate were quantified using an oxygen isotope apportionment model. On average 19–42% and 58–81% of the non-sea salt sulphate in PM2.5 was primary and secondary, respectively. More primary sulphate was present in Vancouver (0.21–0.47 μg m−3) PM2.5 aerosols than at the site further inland (0.13–0.33 μg m−3) but there was no difference between the amount of primary sulphate present during the day and night at each site.
Keywords: Sulphate; PM2.5; Oxygen isotopes; Aerosols;

Stable isotope measurements of carbon fractions (OC/EC) in airborne particulate: A new dimension for source characterization and apportionment by L. Huang; J.R. Brook; W. Zhang; S.M. Li; L. Graham; D. Ernst; A. Chivulescu; G. Lu (2690-2705).
A method to measure 13C/12C ratios of individual carbon fractions of airborne particular matter (PM) from filter samples using a stepwise thermal desorption/combustion OC/EC analyzer (via thermal optical transmission, (TOT) coupled with gas chromatography separation, followed by isotopic ratio mass spectrometer (GC-IRMS) analysis has been developed. In the TOT instrument, carbon fractions are released at different temperature ranges and different redox conditions. Organic carbon fraction (OC) was released at a relatively low temperature ( T = 550  °C), whereas, elemental carbon or black carbon fraction (EC or BC) was released at a high temperature (T>800 °C) via combustion. A temperature step of 870 °C without oxygen was chosen to remove the impact of carbonate carbon (CC) and possible cross-impact from OC and EC. All the fractions were collected cryogenically and subject to carbon isotope measurements via GC-IRMS. To evaluate the precision, accuracy and linearity range of the measurements, the different types of blanks and standards were investigated, including OC (i.e. glucose, sucrose, n-Alkanes and polycyclic aromatic hydrocarbons (PAHs), CC (i.e. carbonates) and EC (i.e. carbon black and graphite). The overall precision and the accuracy of the method is ∼0.3‰. The method was applied to Pacific2001 aerosol samples from the Greater Vancouver area in Canada. The results show that good baseline separations in thermographs can be achieved for individual carbon fractions (i.e. OC and EC) using optimized temperature plateau and retention times; relative small difference in carbon isotopic composition between OC and EC ( Δ 13 C OC – EC = δ 13 C OC – δ 13 C EC ) were found in tunnel samples, whereas, the largest Δ13COC–EC were obtained in forest air samples; the Δ13COC–EC in ambient PM is likely dependant upon the dominant sources present in the vicinity of the sampling sites; the distribution of 13C/12C ratios of OC/EC can provide useful information for source characterization and apportionment of ambient particulate matter.
Keywords: 13C/12C; δ13C; OC; EC; Black carbon;

Ninety daytime/nighttime PM2.5 aerosol samples were collected at 5 sites in forest, tunnel, urban, rural, and mixed forest/urban areas in the Lower Fraser Valley during the Pacific 2001 Air Quality Study. Solvent-extractable organic matter, such as n-alkanes (C14–C33), n-alkan-2-ones (C10–C31), and 6, 10, 14- trimethylpentadecan-2-one on the fine aerosols, were quantified. The concentrations of total n-alkanes from primary sources were 45.5–112 ng m−3 at the tunnel site, 3.3–34.6 ng m−3 at the urban site, 0.6–18.1 ng m−3 at the rural site, and 1.7–16.9 ng m−3 at the forest and the mixed areas. The homologue distributions of the n-alkanes displayed different patterns at the 5 sites, showing day–night differences and reflecting their primary source types and impacts of episodes. The carbon preference index (CPI) values of the n-alkanes showed highest value (average of 2.39±0.47) at the forest, lowest (1.15±0.11) at the tunnel. The CPI showed higher values in night samples at all sites except the urban site which was impacted by specific episodes such as biomass burning and/or fuel burning occurring during the nighttimes, the higher nighttime values of CPI, with consistent lower n-alkane concentrations, suggested that weaker anthropogenic emissions during night were a more likely cause . The total n-alkan-2-ones on the aerosols were 1.8–12.6 ng m−3 at the tunnel site and 0.2–7.2 ng m−3 at the other 4 sites. Low-molecular weight n-alkan-2-ones (<C22) were observed at all sites with the highest level at the tunnel. High molecular weight n-alkan-2-one (>C23) were consistently higher at the tunnel but varied with date at the forest site. The n-alkan-2-ones, both low and high molecular weight, could have multiple sources including vehicular emissions and oxidation processes. The branched ketone, 6, 10, 14- trimethylpentadecan-2-one, an oxidative product of phytol on the fine aerosols, was enriched in the forest with the average of 7.6±7.1 ng m−3.
Keywords: n-alkanes; n-alkan-2-ones; Branched ketone; Distributions; Source fingerprints; Fine aerosols;

Levoglucosan and dehydroabietic acid: Evidence of biomass burning impact on aerosols in the Lower Fraser Valley by Amy Leithead; Shao-Meng Li; Ray Hoff; Yu Cheng; Jeff Brook (2721-2734).
As part of the Pacific 2001 Air Quality Study in August 2001, aerosol samples were collected at three sites in the Lower Fraser Valley, BC twice daily over a 2-week period. In this paper, the results for two compounds in the samples, levoglucosan and dehydroabietic acid (DHAA), both tracers for biomass burning plumes, are presented. Concentrations of the compounds were generally low throughout the study. Average values for levoglucosan were 14.4, 14.7, and 26.0 ng m−3 for the urban, mixed urban/forest, and rural sites respectively. Elevations in the concentrations during two periods indicate evidence of biomass burning. Satellite images from 15 to 18 August show the transport of smoke plumes from forest fires in northern Washington and southern British Columbia into the Lower Fraser Valley. The residues of the smoke plumes in the river valleys probably contributed to the elevated ground-level levoglucosan levels. Emission ratios from reported source studies, 4.5–10% for levoglucosan/organic carbon and 4% for DHAA/organic carbon, were used to estimate the contributions of biomass burning to aerosol organic carbon. Average contributions of biomass burning to organic carbon varied significantly throughout the region. Contributions were between 5 and 10%, 12 and 27% and 5 and 12% for the urban, rural and mixed urban/forest sites respectively, but contributions up to 75% were found.
Keywords: Smoke plume; PM2.5; Organic carbon; Forest fire; Transport; Satellite;

Evaluation of CMAQ O3 and PM2.5 performance using Pacific 2001 measurement data by Steve C. Smyth; Weimin Jiang; Dazhong Yin; Helmut Roth; Éric Giroux (2735-2749).
Measurement data from the Pacific 2001 air quality field study was used to evaluate the performance of Models-3/CMAQ for the period of 9–20 August in 2001, for a domain centered about Vancouver and the Lower Fraser Valley in British Columbia, Canada. The model predicted O3 concentrations reasonably well with a normalized mean bias (NMB) and normalized mean error (NME) of 13.3% and 51.2% over all measurement sites and hours. It showed an excellent performance in predicting daily O3 peaks with a NMB and NME of −2.2% and 24.3%. The model also did well in predicting the time at which the peaks occurred. The model captured the diurnal variations and spatial distributions of O3 concentrations very well, with some difficulty at locations deep in the valley. Model predictions of PM2.5 concentrations over all sites and hours had a moderate NMB and NME of 30.9% and 66.2%. The model did well in predicting the general temporal trend of PM2.5 concentration levels. However, the high hourly variability in PM2.5 measurements was not possible for the model to capture. In addition, the temporal variations of modelled concentrations change more dramatically than the measured. The model performed poorly in predicting PM sulphate, ammonium, and nitrate concentrations. The extreme over-prediction in nitrate concentrations was partially related to an overabundance of total ammonia in the system. Sensitivity tests of NH3 emission reductions showed that reducing NH3 supply to the system could help reduce the over-predictions of both ammonium and nitrate concentrations, although the model science also needs to be improved. To evaluate the PM organics performance, carbon multiplication factors (CMFs) of 1.4 and 1.9 were applied to an urban site and an urban-to-rural transition site, respectively. The model showed satisfactory performance for PM2.5 organics with a NMB and NME of −6.0% and 50.5%.
Keywords: Ozone; Particulate matter; Air quality; Modelling; Aerosol;

Heat flux, urban properties, and regional weather by P.A. Makar; S. Gravel; V. Chirkov; K.B. Strawbridge; F. Froude; J. Arnold; J. Brook (2750-2766).
The flux of heat from human activities such as commercial energy use, renewable source combustion, and the human metabolism has been incorporated into a variable resolution global weather forecast model, along with improved urban surface roughness and radiative properties. Sensitivity studies of these changes were used to show that the addition of anthropogenic heat improves the accuracy of surface air temperature forecasts. The addition of urban surface radiative properties has a secondary effect on the forecast temperature, and the addition of urban surface roughness changes has a minimal effect. Comparisons between observed and forecast boundary layer heights suggest that this parameter is poorly predicted by the model employed here, but that the impact of anthropogenic heating is likely to be a substantial increase in PBL heights over urban regions. Decreased atmospheric stability is also evidenced by comparisons of the diffusion constants for heat and moisture between the original and modified models, which show increases from a factor of 2 to a factor of 16 near the surface, depending on the size of the city. An examination of the effects of spatial averaging on heat flux suggests that significant sub-gridscale anthropogenic heating effects may occur, and implies that the results of the current simulations represent lower bounds. The simulations suggest that anthropogenic heat flux has a large local impact, with important implications for simulations of air-quality and radiative balance on regional and global scales.
Keywords: Heat island; Boundary layer; Air pollution; Climate change;

Identification and characterization of inland ship plumes over Vancouver, BC by Gang Lu; Jeffrey R. Brook; M. Rami Alfarra; Kurt Anlauf; W. Richard Leaitch; Sangeeta Sharma; Daniel Wang; Douglas R. Worsnop; Lisa Phinney (2767-2782).
There is increasing concern regarding the impact of marine vessel emissions on the air quality of coastal areas and their relative impact is increasing as emissions from other sources decrease and shipping activities increase. Marine vessels contain a variety of large diesel engines and in a relatively large number of areas they are currently not restricted from using fuels with a high sulphur content.In August 2001 during the Pacific 2001 study, which included the port city of Vancouver, British Columbia, a large suite of gas and particle measurements were obtained with high time resolution. Among a total of 29 SO2 episodes observed >5 km inland during a period of 15 days, eight were caused by local emissions sources and four were identified as relatively fresh ship plumes. These ship plumes were indicated by an increase of SO2 above 9 ppbv typically lasting for a few hours. They were accompanied by increases in NO x , NO, CO, VOCs, particle counts (5–200 nm), black carbon and PM2.5. Only one plume occurred when an Aerodyne aerosol mass spectrometer (AMS) was in operation and this event is studied in detail. Ultrafine (<100 nm) sulphate was one of the most unique features of this plume, which also contained significant amounts of ultrafine particulate organic matter. The distribution of AMS organic mass fragments for this case strongly resembled those measured directly in the effluent of an ocean-going ship, suggesting a signature for marine diesel engine emissions. During the event studied in detail, which occurred at night, the meteorological measurements indicated the same plume or puff of high concentrations moved over the measurement site on two instances with peak concentrations separated by about 3 h. From the first to the second occurrence all species decreased in concentration except particle sulphate and VOCs. This is considered to be direct observation of nighttime gas-to-particle conversion of SO2. This process was likely facilitated by SO2 dissolving into the wet surface of particles given the high humidity observed on that night.
Keywords: Ship emissions; Particulate matter; Urban air quality; Formation of sulphate; Sulphur in diesel fuel; Ship plume; Aerosol mass spectrometer;

Data management and archiving for the Pacific 2001 Air Quality Study by William B. Sukloff; Robert J. Vet; Shao-Meng Li (2783-2795).
The data management and archiving activities of the Pacific 2001 Air Quality Study were handled by the Pacific 2001 Data Centre which was run by the National Atmospheric Chemistry (NAtChem) Database and Analysis Facility of Environment Canada. To ensure that the Pacific 2001 Air Quality Study data were archived in a common way, the NARSTO Data Exchange Standard (DES) was used as the mandatory format for the data files, partially because it allowed for the inclusion of metadata within the data files and partially because it provided the necessary flexibility for handling the many measurement types used in the study. Described in detail in the paper, the DES is now readily available to the scientific community. After each DES data file was submitted to the Data Centre, a read-and-verify program was run to check its conformity to the DES and to detect incorrect and problematic data. The errors detected by the read-and-verify program were automatically documented and an error report was sent to the data originators for data correction and resubmission. Statistical summaries and data plots were created for all data files and subsequently sent to the data originators for review and further error detection. Of the 125 data files submitted to the Data Centre, only 5 were error-free upon first submission. A test of 17 randomly selected files determined that all but two required at least four iterations of the submission–error checking–resubmission cycle in order to produce final error-free files. It was therefore concluded that both data originators and data centres alike should assume that errors exist in all submitted data files until proven differently by a set of automated error-checking programs. It was also concluded that data visualization plots and statistical summaries are highly effective tools for detecting errors in data files. Metadata associated with the measurement data were documented in Quality Assurance Project Plans that were archived in the Data Centre with the DES data files.
Keywords: Data quality assurance; Data management; Data archive; Data exchange standard; Metadata;