Atmospheric Environment (v.37, #32)

Computer simulations are described for the homogeneous portion of the chemistry of the ozone and mercury depletion observed in the troposphere during the polar spring. Conditions chosen are similar to those of a clear day (17 March) at the location of Barrow, Alaska. Small amounts of reactive trace gases, representative of the Arctic spring, are present initially with typical background levels of Hg (0.24 ppt) and 50 ppb of O3. The simulations show that ozone and gaseous atomic mercury depletion are both strong functions of two important variables: (1) the concentrations of Br2 and BrCl present during the day; and (2) the concentrations of reactive trace gases such as CH2O present in the air mass. The simulated rate of Hg depletion and the expected product distribution are also functions of the size of the rate coefficient for the Hg+BrO reaction (k 17). The relatively small extent of O3 and Hg depletion rates that result from Cl–ClO chemistry and the major effects that result from the Br–BrO chemistry are rationalized in terms of the significant differences in the reaction mechanisms for the two halogens.
Keywords: Arctic O3 depletion in troposphere; Arctic Hg depletion in troposphere; Computer simulations BrO–Br–ClO–Cl–O3 chemistry;

On the correlation between the depth of the boundary layer and the columnar aerosol size distribution by Monica Campanelli; Luca Delle Monache; Vincenzo Malvestuto; Bruno Olivieri (4483-4492).
Most inversion algorithms used to retrieve the aerosol characteristics from ground-based solar radiation measurements, assume a vertically homogeneous atmosphere and constant aerosol optical characteristics over the whole air column. This assumption leads to correct results only if the extinction of the solar radiation incident on the earth surface depends on the mean optical characteristics of the entire atmospheric column, rather than on the details of the vertical distribution of air particles. The validity of this simplifying assumption is an issue of a certain importance, which requires a more definite answer based on observational evidence. The aerosol vertical profile is significantly affected by the extent of the atmospheric boundary layer, within which the physical properties of the suspended particulate matter show sharp variations. Therefore, if the above assumption were not correct, one should observe a significant correlation between the retrieved optical characteristics and the thickness of the boundary layer measured on the same site over time. In this paper we limit ourselves to study the possible correlations between the aerosol volume size distributions and the mixing height of the atmospheric boundary layer, in order to ascertain whether there exists any dependence between the volume size distributions and the vertical distribution of the aerosol, in a site where there are strong reasons to believe that most of the existing aerosol is confined below the mixing height level. The investigation is based on a dataset of measurements taken at the Southern Great Plains base station of the ARM Program, located in Oklahoma, USA. Measurements of the columnar aerosol characteristics are taken by means of a Cimel sun-sky radiometer of the Aeronet network, while vertical profiles of pressure and temperature, used for the determination of the height of the boundary layer, are taken by vertical sounding systems. In order to find a possible relationship between the columnar aerosol size distribution and the depth of the boundary layer, a statistical analysis has been carried out on this data. The analysis has not shown any significant correlation between those two parameters. This negative result supports the validity, for the Southern Great Plains site, of the assumption that the radiation incident on the earth surface depends only on the mean optical characteristics of the entire atmospheric column, rather than on the details of the aerosol vertical distribution.
Keywords: Aerosol; Atmospheric boundary layer; Remote sensing; Statistical analysis;

Industrial emissions occur frequently in the form of multiple stacks positioned close to each other. The dispersion of the individual plumes may in that case interfere with each other. Here with the help of direct numerical simulation we study the effect of such interference on the concentration statistics and in particular on the variance of the concentration fluctuations. The simulation results show that the combined variance of two nearby sources cannot be obtained by adding the variance of the individual sources. The difference depends on the source separation and the downstream distance from the two nearby sources. Nearby in this context means that the distance between the two sources is smaller than the length scale of the largest turbulent eddies involved in the dispersion process. The numerical results are found to agree with data obtained in laboratory experiments. Finally we propose a model based on the well-known fluctuating plume model, by which we can extend our simulation results to atmospheric situations.
Keywords: Statistics; Correlation; Numerical simulation; Multiple sources; Line source; Fluctuating plume model;

Rainwater samples were collected in a rural region in Northern Jordan using 24-h sampling periods from December 1998 to April 2000. All samples were analyzed for major ions (Na+, K+, Ca2+, H+, Mg2+, NH4 +, Cl, NO3 and SO4 2−) and trace metals (Pb, Cd, Zn, Cu, Al, Fe, Mn, Mo, Ni, Sb and V). The majority of the rain samples collected had pH values higher than 5.6. The average pH was 6.4±0.9. High values of pH were attributed to the neutralization by natural alkaline local dusts which contain large fractions of calcite. The annual average SO4 2−-to-NO3 ratio is 1.8, which is close to that observed in more polluted regions. Concentrations of measured species were lower than those reported for other rural sites worldwide. Elements of anthropogenic origins (Zn, Pb, As, Sb, Ag and Cd) were highly enriched with respect to crustal composition. Factor analysis permitted the identification of four source groups, namely crustal dust, sea-salt spray, road traffic and combustion and secondary aerosol formation processes.
Keywords: Precipitation; Acidity; Trace metals; ICP-MS; Enrichment factors; Factor analysis; Jordan;

Rate constants for the gas-phase reaction of NO and HO2 radicals with 33 peroxy radicals are presented. The peroxy radicals are derived from the addition of either OH, Cl, or NO3 radicals, followed by addition of O2, to a series of alkenes: tetrachloroethene, ethene, 2,3-dimethyl but-2-ene, butadiene, 2,3,4,5-tetramethyl hexa-2,4-diene, 1,1,2,3,4,4-hexachlorobutadiene, but-1-ene-3-one (methyl vinyl ketone) and 2,3-dimethylpen-2-ene-4-one. The rate constants were predicted using a correlation between the singly occupied molecular orbital (SOMO) energy of the peroxy radical and the logarithm of the rate constant for reaction with NO or HO2. A discussion of the accuracy of the method and the trends in the reactivity of the titled peroxy radicals is given. Peroxy radicals derived from halogenated alkenes have larger values of rate constants for reaction with NO relative to reaction with HO2, indicating that they are more likely to react with NO, rather than HO2, in the atmosphere. The reverse is true for peroxy radicals derived from alkylated alkenes.
Keywords: HOMO energy; SOMO energy; Halogenated alkenes; Alkylated alkenes; Troposphere oxidation; Peroxy radical;

Raman lidar and sunphotometric measurements of aerosol optical properties over Thessaloniki, Greece during a biomass burning episode by D.S. Balis; V. Amiridis; C. Zerefos; E. Gerasopoulos; M. Andreae; P. Zanis; A. Kazantzidis; S. Kazadzis; A. Papayannis (4529-4538).
The influence of biomass burning smoke on the aerosol loading in the free troposphere over Thessaloniki, Greece (40.5°N, 22.9°E) is discussed in this paper. A selected case during summer 2001 is presented, when very high aerosol optical depth values were observed, benefiting from the synergy of various remote sensing instruments. The data that were collected allow the characterization of the optical properties of the aerosols in this region, where only little information has so far been available. Four-day back trajectories indicated that air masses were advected from Bulgaria and the northern coast of the Black Sea, where strong forest fires occurred in early August 2001. In order to investigate the optical properties of biomass burning aerosols, we used a two-wavelength lidar system that combines Raman and elastic-backscatter observations, in addition to a Brewer spectrophotometer, a nephelometer and a multi-filter rotating shadowband radiometer. The lidar measurements on 9 August 2001 recorded an integrated aerosol optical depth at 355 nm of the order of 1.35 during cloud-free conditions. The estimated mean extinction–to-backscatter ratios from the Raman lidar were 60 sr for 355 nm and 50 sr for 532 nm. Estimated values of the single scattering albedo, using spectral UV measurements and modeling were of the order of 0.90, consistent with previous findings, indicating a weak contribution of absorption to the total extinction. The Angstrom exponent, calculated from the multi-filter rotating shadowband radiometer exhibited also high values around 1.78, indicating the presence of rather small particles.
Keywords: Lidar; Aerosol; Biomass burning;

Extensive evaluation of neural network models for the prediction of NO2 and PM10 concentrations, compared with a deterministic modelling system and measurements in central Helsinki by Jaakko Kukkonen; Leena Partanen; Ari Karppinen; Juhani Ruuskanen; Heikki Junninen; Mikko Kolehmainen; Harri Niska; Stephen Dorling; Tim Chatterton; Rob Foxall; Gavin Cawley (4539-4550).
Five neural network (NN) models, a linear statistical model and a deterministic modelling system (DET) were evaluated for the prediction of urban NO2 and PM10 concentrations. The model evaluation work considered the sequential hourly concentration time series of NO2 and PM10, which were measured at two stations in central Helsinki, from 1996 to 1999. The models utilised selected traffic flow and pre-processed meteorological variables as input data. An imputed concentration dataset was also created, in which the missing values were replaced, in order to obtain a harmonised database that is well suited for the inter-comparison of models. Three statistical criteria were adopted: the index of agreement (IA), the squared correlation coefficient (R 2) and the fractional bias. The results obtained with various non-linear NN models show a good agreement with the measured concentration data for NO2; for instance, the annual mean of the IA values and their standard deviations range from 0.86±0.02 to 0.91±0.01. In the case of NO2, the non-linear NN models produce a range of model performance values that are slightly better than those by the DET. NN models generally perform better than the statistical linear model, for predicting both NO2 and PM10 concentrations. In the case of PM10, the model performance statistics of the NN models were not as good as those for NO2 over the entire range of models considered. However, the currently available NN models are neither applicable for predicting spatial concentration distributions in urban areas, nor for evaluating air pollution abatement scenarios for future years.
Keywords: Neural network; Urban air; NO2; PM10; Model evaluation;

Methane emission from wetlands in Taiwan by Tsan-Chang Chang; Shang-Shyng Yang (4551-4558).
To investigate methane emission from wetlands in Taiwan, soil properties, environmental conditions and methane emission were determined at three wetlands in northern Taiwan from September 1995 to April 1999. Soil pH values ranged from 6.15 to 7.80. Total organic carbon and total nitrogen contents were high in Kuan-du wetland and Kang-nan lake area, where the soils were sandy loam. In Kang-nan wetland, total organic carbon and total nitrogen contents were low and the soil was clay loam. Soil redox potential of Kuan-du wetland and Kang-nan lake area was lower than that of Kang-nan wetland. Atmospheric methane concentration was 0.8–2.7, 0.7–1.6 and 0.8–1.7 ppm (mg kg−1) in Kang-du wetland, Kang-nan wetland and Kang-nan lake area, respectively. Average methane emission rate was 1.82, 0.14 and 0.23 mg m−2  h−1 and annual methane emission was about 1.59×10−1, 1.23×10−2 and 2.02×10−2  ton ha−1, respectively. Annual methane emission from 11,896 ha of nine wetlands is estimated around 340 ton in Taiwan.
Keywords: Methane flux; Wetland; Soil properties; Environmental conditions;

Vehicle-based road dust emission measurement: I—methods and calibration by V. Etyemezian; H. Kuhns; J. Gillies; M. Green; M. Pitchford; J. Watson (4559-4571).
Testing re-entrained aerosol kinetic emissions from roads (TRAKER) is a vehicle-based method for measuring road dust emissions. Particulate matter is sampled in front and behind a vehicle's tire and the difference in PM concentration is related to emissions. This paper describes the most recent developments of TRAKER. The loss of particles within the inlet lines, the response of the TRAKER signal (differential PM concentration) to vehicle speed, and the relationship between the TRAKER signal and unpaved road dust PM10 emissions were examined. Losses of particles to the walls in the inlet lines were similar for the left and right inlets and were less than the inter-instrument precision for particles between 0.56 and 7.3 μm in diameter. When summed over the PM10 size range, losses were less than 20%. Line losses were also characterized when a dilution system was used to sample emissions from unpaved roads. Two independent tests indicated that the TRAKER signal increases as the cube of the speed for a given road dust loading. Simultaneous measurement of PM10 dust emitted behind the tires by TRAKER with PM10 flux measured using upwind/downwind towers suggested that the emission factor for road dust was proportional to the cube root of the TRAKER signal. The results also showed a linear relationship between unpaved road dust PM10 emissions and vehicle speed.
Keywords: TRAKER; Spatially resolved; Real-time; Fugitive; Silt loading; PM10;

Testing Re-Entrained Kinetic Emissions from Roads (TRAKER) is a new technique to infer paved and unpaved road dust PM10 emission potentials based on particulate matter (PM) measurements made onboard a moving vehicle. Light scattering instruments mounted in front and behind the vehicle's tires measure the differential particle concentration of dust suspended by the vehicle's tire in contact with the road surface. Through empirical regressions relating the differential concentration (i.e. TRAKER signal) with the vehicle speed and the downwind flux of PM10 particles from the road, an equation is derived to infer the speed independent road dust emission potential from the measured TRAKER signal. Measurements from TRAKER offer a new perspective on the processes that affect road dust emissions. The system was used to investigate temporal changes in emission potentials from paved roads in both the winter and summer in the Treasure Valley in Southwest Idaho. During the 3-week wintertime sampling period, the residential road dust PM10 emission potential decreased by ∼50%. Summertime PM10 emission potentials were similar to those observed at the end of the winter sampling and showed no upward or downward trends. Wintertime unpaved road emissions increased consistently with the number of days since the last rainfall. Measurement of road dust emission potentials after road sanding on dry roads indicated a 75% increase in PM10 emissions after 2.5 h. This effect was short lived and emission potentials returned to their pre-sanding levels within 8 h of the sand application. Street sweeping with mechanical and vacuum sweepers was found to offer no measurable reduction in PM10 emission potentials. On several roads, the PM10 emission potentials actually increased immediately after vacuum sweeping. Long term effects of street sweeping on road dust emissions were not evaluated as part of this study and may offer some overall reduction in PM emissions from paved roads.
Keywords: TRAKER; Road dust; Emissions; Road sanding; Street sweeping; Paved road; Unpaved road; PM10;

Vehicle-based road dust emission measurement (III): by V. Etyemezian; H. Kuhns; J. Gillies; J. Chow; K. Hendrickson; M. McGown; M. Pitchford (4583-4593).
The testing re-entrained aerosol kinetic emissions from roads (TRAKER) road dust measurement system was used to survey more than 400 km of paved roads in southwestern Idaho during 3-week sampling campaigns in winter and summer, 2001. Each data point, consisting of a 1-s measurement of particle light scattering sampled behind the front tire, was associated with a link (section of road) in the traffic demand model network for the Treasure Valley, ID. Each link was in turn associated with a number of characteristics including posted speed limit, vehicle kilometers traveled (vkt), road class (local/residential, collector, arterial, and interstate), county, and land use (urban vs. rural). Overall, the TRAKER-based emission factors based on location, setting, season, and speed spanned a narrow range from 3.6 to 8.0 g/vkt. Emission factors were higher in winter compared to summer, higher in urban areas compared to rural, and lower for roads with fast travel speeds compared to slower roads. The inherent covariance between traffic volume and traffic speed obscured the assessment of the effect of traffic volume on emission potentials.Distance-based emission factors expressed in grams per kilometer traveled (g/vkt) for roads with low travel speeds (∼11 m/s residential roads) compared to those with high travel speeds (∼25 m/s interstates) were higher (5.2 vs. 3.0 g/vkt in summer and 5.9 vs. 4.9 g/vkt in winter). However, emission potentials which characterize the amount of suspendable material on a road were substantially higher on roads with low travel speeds (0.71 vs. 0.13 g/vkt/(m/s) in summer and 0.78 vs. 0.21 g/vkt/(m/s) in winter). This suggested that while high speed roads are much cleaner (factor of 5.4 in summer), on a vehicle kilometer traveled basis, emissions from high and low speed roads are of the same order.Emission inventories based on the TRAKER method, silt loadings obtained during the field study, and US EPA's AP-42 default values of silt loading were compared. PM10 paved road dust emission inventories calculated with the TRAKER method were 61% higher in winter and 180% higher in summer than inventories calculated from on-site silt loading measurements. Emissions calculated from silt loading measurements conducted on-site indicated that the AP-42 default values are too low for the Treasure Valley by a factor of 1.5 for summer conditions and by a factor of 3.8 for winter. Both silt loading and TRAKER are techniques that were calibrated against the horizontal flux of dust, which was estimated by the difference in PM10 concentration between instruments located upwind and downwind of an unpaved road. The upwind/downwind method, and therefore both silt loading and TRAKER, gives a measure of the dust emitted near the source, and not the dust that can be transported on a regional or air shed scale. Correcting the measured dust emissions for deposition and removal near the source is outside the scope of this work, but is a continuing area of research among dispersion modelers.
Keywords: TRAKER; Road dust; Emissions; Fugitive; Silt loading;

Corrigendum to “Receptor modeling of ambient and personal exposure samples: 1998 Baltimore particulate matter epidemiology-exposure study” by Philip K. Hopke; Ziad Ramadan; Pentti Paatero; Gary A. Norris; Matthew S. Landis; Ron W. Williams; Charles W. Lewis (4595).