Atmospheric Environment (v.36, #9)

An empirical approach for the prediction of daily mean PM10 concentrations by Gary W Fuller; David C Carslaw; Hamish W Lodge (1431-1441).
An empirical model has been devised to predict concentrations of PM10 at background and roadside locations in London. Factors to calculate primary PM10 and PM2.5 concentrations are derived from annual mean NO X , PM2.5 and PM10 measurements across London and south east England. These factors are used to calculate daily means for the primary and non-primary PM10 fractions for the London area. The model accurately predicts daily mean PM10 and EU Directive Limit values across a range of sites from kerbside to rural. Predictions of future PM10 can be made using the expected reductions in secondary PM10 and site specific annual mean NO X predicted from emission inventories and dispersion modelling. The model suggests that the EU Directive Limit values will be exceeded close to many of London's busiest roads, and perhaps at central background sites should there be a repeat of 1996 meteorological conditions during 2005. A repeat of 1997 meteorology conditions during 2005 would lead to the EU Limit Value being exceeded alongside the busiest central London roads only. The model is applicable for London and south east England but the methodology could be applied elsewhere at a city or regional level. The model relies on the currently observed ratio between NO X and PM10. This ratio has remained constant over the last 4 years but might change in the future. The NO X :PM10 ratio derived from measurements and used in this model, implies that emission inventories might over estimate primary PM10 by more than 50%.
Keywords: NO X ; PM2.5; Receptor model; European Union Daughter Directive; London;

Development of a NO2 scrubber for accurate sampling of ambient levels of terpenes by Linda Pommer; Jerker Fick; Barbro Andersson; Calle Nilsson (1443-1452).
The result of pumping air containing 56 ppb NO2 through a terpene-spiked adsorbent (90–130 ng, 90–100 ml min−1), Tenax TA, for 20 min (1.8–2.0 l) was that 8% of α-pinene, 7% of β-pinene, 21% of Δ 3-carene and 5% of limonene were oxidised. In similar experiments with air containing 56 ppb O3, 3% of α-pinene, 4% of β-pinene, 10% of Δ 3-carene and 38% of limonene were oxidised. Sampling a mixture of a terpene and NO2 using Tenax TA can give unwanted overestimation of the amount of reaction products from the terpene–NO2 reaction or underestimation of the original terpene levels. A scrubber was needed to reduce the problems caused by interfering reactions on the adsorbent of NO2 and to reduce discrimination of reactive compounds due to their relatively fast decay on the adsorbent. Several chemicals have been tested for their ability of removing NO2 and our objective was to develop a well functioning, reusable, easy to handle, easy manufactured NO2 scrubber. The result of the experiments was a scrubber consisting of two glass fibre filters coated with Na2SO3 assembled in a dust collector. The recovery of the terpenes through the scrubber varied between 75% and 97% at 15–75% relative humidity, and the scrubber is a one-use scrubber due to memory effects. The Na2SO3 scrubber could be stored in room air for at least one month without loosing the capacity of removing NO2.
Keywords: Nitrogen dioxide; Scrubber; Tenax TA; Terpene; Sodium sulphite;

The effect of porous wind fences on the wind erosion of small sand particles from a two-dimensional triangular prism pile of sand was investigated experimentally. The mean velocity and turbulence intensity profiles measured at the sand pile location were simulated to fit to the atmospheric boundary layer over open terrain. Flow visualization was carried out to qualitatively determine the movement of the wind-blown sand particles. In addition, the threshold velocity was measured using a digital image processing system at a range of values of the fence porosity ε, particle size D p and sand pile height h. As a result, various types of particle motion were observed according to the experimental conditions. The fence of porosity ε=30% was found to have the highest threshold velocity, indicating a good shelter effect for abating wind-blown sand particles. The threshold velocity was found to increase with increasing sand particle diameter. The threshold velocity was also enhanced when the height of the sand pile was lower than the fence height.
Keywords: Porosity; Fence; Sand pile; Threshold velocity; Atmospheric boundary layer;

In-stack emissions of heavy metals estimated by moss biomonitoring method and snow-pack analysis by D Čeburnis; J Šakalys; K Armolaitis; D Valiulis; K Kvietkus (1465-1474).
Data from the chemical analysis of moss growing close to a thermal power station and snowpack have been used for the estimation of heavy metal deposition close to the point pollution sources. A semi-empirical model was proposed to describe atmospheric trace metal deposition close to the point pollution source. Model parameters were derived from experimental data, and nickel and vanadium quantities, washed out with snow and rain, were calculated. Using long-term meteorological observation data of rain and snow duration and metal uptake efficiencies in moss, the average emission rates of vanadium and nickel from the stack were calculated. The coincidence between data from emission inventory and model results was within 25%. It was estimated that in the vicinity of pollution source (within 30 km) about 15% out of total emitted metals were washed out by rain and snow events. Metal concentrations in the environment become indistinguishable from the background at a distance of about 20 km from the stack.
Keywords: Heavy metals; Emission; Moss; Deposition;

Estimates of the emission rates of ammonia from light-duty vehicles using standard chassis dynamometer test cycles by Thomas D Durbin; Ryan D Wilson; Joseph M Norbeck; J.Wayne Miller; Tao Huai; Sam H Rhee (1475-1482).
Emissions rates of ammonia (NH3) are reported for a fleet of 39 in-use light-duty gasoline-fueled vehicles. The fleet consisted of both light-duty passenger vehicles and light-duty trucks with various levels of emission control technologies, ranging from non-catalyst vehicles to those that were certified at the ULEV standard for California. NH3 measurements were performed using Fourier transform infrared spectroscopy and the federal test procedure (FTP) driving cycle. The FTP NH3 emission rate for this fleet of vehicles averaged 54 mg mi−1 with a range from <4 to 177 mg mi−1. For this fleet of vehicles, NH3 emissions did not decline as significantly as the regulated pollutants with improvements in emission control technology. A subset of 5 vehicles was tested over the US06, the New York City Cycle (NYCC), and a high-speed freeway cycle for comparison with the FTP cycle. NH3 emissions showed a strong cycle dependence, with increased emissions under more aggressive driving conditions. These results show that NH3 emissions formed during more aggressive driving conditions should be considered in the development of NH3 emission factors. The onset of NH3 emissions typically occurred after catalyst light-off, near when the catalyst reached its equilibrium temperature. Initial studies showed that NH3 emissions increased as the sulfur content in the fuel was decreased.
Keywords: Automotive exhaust; Automotive emissions; Vehicle emissions; Chassis dynamometer; Emission rates; Ammonia;

Estimating the vapor pressures of multi-functional oxygen-containing organic compounds using group contribution methods by William E Asher; James F Pankow; Garnet B Erdakos; John H Seinfeld (1483-1498).
A UNIFAC-based method for estimating the vapor pressure (p L o) values of oxygen-containing compounds of intermediate-to-low volatility has been developed as an aid in modeling the formation and behavior of organic aerosols. This UNIFAC-p L o method was constructed using a set of 76 compounds with experimentally determined p L o values. The compounds chosen are of intermediate-to-low volatility and contain multiple oxygen-containing functionalities. For test and development purposes, the 76 compounds were divided into a basis set of 43 compounds used to generate the coefficients required in the UNIFAC-p L o method and a second set of 33 compounds that was used to test the coefficients generated using the basis set. Both the basis and test sets contained compounds that possessed similar structures and functionalities. For the 33 compounds in the test set, on average UNIFAC-p L o predicted the p L o values to within a factor of 2 over the temperature range 290–320 K. Furthermore, the UNIFAC-p L o method did not show any correlation in prediction error with p L o so that it was equally likely to underpredict as overpredict p L o regardless of volatility. For comparison, three other vapor pressure estimation methods were applied to the test set of compounds. On average, these other methods all predicted the test set p L o values to within a factor of 3 over the temperature range 290–320 K. In contrast to the UNIFAC-p L o method, the prediction errors from the methods were found to be correlated with p L o so that the other methods overpredicted p L o as volatility decreased.

A newly developed system (Mist-UV sampling system) to simultaneously trap and detect metal compounds in the atmosphere is described. Metals present in atmospheric samples as volatile metal compounds (VMCs), including those associated with condensation mode particles, are simultaneously oxidized by ultraviolet (UV) irradiation (low-pressure mercury lamp) and dissolved in nitric acid mist in an oxidation chamber. Air is first passed through a 0.2 μm filter and then mixed with a nitric acid mist and exposed to UV radiation after which it is passed through a nitric acid trap. After trapping, the inorganic metals resulting from the VMCs and particulates are quantitatively determined using inductively coupled plasma-mass spectrometry (ICP-MS). After development, this sampling system was applied directly to the quantitation of VMCs in ambient air in Auckland, New Zealand. The trapping efficiency of the Mist-UV sampling system was over 69% for each of the seven different metals that were detected in the atmosphere (Mg, Al, Ga, Sr, Ba, Hg and Pb). Key results obtained were: Mg concentrations were found to be higher than concentrations associated with particulate in remote air (range: 0.36–1.6 μg m−3); Al and Sr existed as VMCs unaffected by UV; UV oxidized Hg provided the total concentration of Hg detected.The characteristics of the Mist-UV sampling system suggest that it can be used as a tool to determine volatile metals, metal clusters and metals associated with particulates <0.2 μm in diameter especially in polluted atmospheres. This is a relatively new area, and the information gained is of particular relevance to metal transport and deposition.
Keywords: Metalloids; Nanoparticles;

The importance of NO x production by lightning in the tropics by Donald W Bond; Scott Steiger; Renyi Zhang; Xuexi Tie; Richard E Orville (1509-1519).
The importance of NO x production by lightning in the tropics has been assessed by using satellite lightning measurements from the Lightning Image Sensor (LIS). The lightning data from LIS over the period of 1998–2000 are analyzed and partitioned based on the latitude to obtain the numbers of cloud-to-ground (CG) and cloud-to-cloud (intracloud or IC) flashes. The average annual lightning counts over the 3-yr period are 3.0×108 CG flashes and 1.1×109 IC flashes between 35°N and 35°S. The resulting lightning distributions are employed to calculate the production of NO x . We obtain a lightning NO x production of 6.3 Tg N yr−1 over this latitudinal region, using representative production values of 6.7×1026 and 6.7×1025 NO molecules for each CG and IC flash, respectively. NO x production by lightning varies slightly on a seasonal basis in accordance with the lightning distribution, with the maximum production occurring in the months of September, October, and November. The geographical and seasonal production of NO x by lightning is compared to NO x emissions from other sources (i.e., from anthropogenic activity, biomass burning, and soil emissions). The results indicate that production of NO x by tropical lightning is significant throughout the year. Lightning accounts for almost all of the NO x emitted over the oceans and 50–90% of NO x emitted over some continental areas on a seasonal basis. On the annual basis, the contribution of lightning to total NO x production is 23% in the tropics, and globally the lightning NO x production occurs predominately in this region. The uncertainty in estimating NO x production by lightning over the tropics is discussed.

Aerosol deliquescence plays a significant role in atmospheric phenomena such as acidic precipitation, formation of clouds and fogs, and changes in visibility. Important factors including aerosol composition and atmospheric relative humidity govern the aerosol deliquescence. However, recent work has revealed that certain aerosols form a metastable composition, which cannot be elucidated by measuring the solubility of bulk-phase solution, nor can it be predicted by traditional thermodynamic theory. This composition, however, contributes to the difference between the predicted and measured deliquescence of atmospheric aerosol. This work applied the newly developed semi-automated GC-TCD (gas chromatograph equipped with thermal conductivity detector) aerosol water mass measurement system to measure the hygroscopic characteristics of aerosols of NaCl, Na2SO4, NH4NO3, NH4Cl and mixed aerosols of NaCl–Na2SO4 and NH4NO3–NH4Cl in various proportions. The measurements are compared with those in the literature to validate the GC-TCD method and investigate the metastable solid form of the Na2SO4 aerosol, which cannot be predicted by traditional thermodynamic theory. The comparisons confirm that measurements of the characteristics of single- and mixed-salt aerosols, obtained by GC-TCD, are satisfactorily accurate. Furthermore, the Na2SO4 aerosol is found to form an anhydrate other than the dehydrate predicted by traditional thermodynamic theory. This result may explain the difference between the measured and modeled results at the DRH of NaCl–Na2SO4 mixed aerosol.
Keywords: Aerosol water mass; Deliquescence relative humidity; Efflorescence relative humidity; Single salt; Mixed salts;

Detection of biomass burning smoke in satellite images using texture analysis by Koji Asakuma; Hiroaki Kuze; Nobuo Takeuchi; Takashi Yahagi (1531-1542).
Classification results using texture analysis is presented for forest fire smoke from satellite remote sensing data. Texture analysis is carried out for normalized difference images calculated from visible and thermal infrared images of the Indonesian forest fire in 1997. Smoke regions are identified by assuming threshold values for the resulting texture feature as well as for radiances in the original and difference images. It is found that when the thresholds are chosen appropriately for GMS visible and infrared spin scan radiometer, 94% pixels exhibit agreement between the classification results using the texture analysis and the supervised Euclidean classification. Agreement is found for 96% pixels in mutual verification using the VISSR image and a concurrent NOAA advanced very high resolution radiometer image. A correlation coefficient of 0.91 is obtained between the results from the two sensors in the variation of the number of smoke pixels accumulated for 12 days in September 1997. Additionally, it is confirmed that as the threshold value of the texture feature is increased, the variation range of the aerosol optical thickness is also increased. As a whole, this study indicates that texture analysis provides quite reasonable results in the smoke detection when appropriately combined with the spectral information.
Keywords: Indonesian forest fire; Unsupervised classification; Multi spectrum classification; Aerosol optical thickness; GMS VISSR; NOAA AVHRR;

Respirable suspended particulate matter and nitrogen oxides concentrations were measured inside and outside a student office in an urban location during a 9-month period. Direct reading tapered-element oscillating microbalance (TEOM) instruments and passive sampling techniques were used to provide indoor and outdoor hourly averages of the two pollutants during the sampling period. The variations and correlations of the pollutant concentrations and the indoor–outdoor (IO) ratio against various outdoor meteorological factors, namely temperature, humidity, pressure, wind speed and solar irradiation were studied. Using a data-mining procedure, suitable sets of data were amassed and grouped together to study the effect of these individual weather parameters on the IO ratio. It is found through statistical regression techniques that temperature, humidity and solar irradiation play a vital role in the variation of the IO ratio. In fact, the IO ratio shows convincing tendency of increase with increase of these three weather parameters. On the other hand, both pressure and wind speed seems to have relatively little effect on the IO ratio.
Keywords: Indoor air; Indoor/outdoor relationship; Regression models; Meteorological conditions; Respirable suspended particulate; Nitrogen oxides;

Formic, acetic, oxalic, malonic and succinic acid concentrations and their contribution to organic carbon in cloud water by M Löflund; A Kasper-Giebl; B Schuster; H Giebl; R Hitzenberger; H Puxbaum (1553-1558).
The carbon content of cloud water at a continental background site in Austria was studied during two intensive field campaigns in spring 1999 and 2000. Six carboxylic acids, total (TC) and black (BC) carbon as well as major inorganic ions were determined. Organic carbon (OC) was calculated as the difference between TC and BC. The most abundant carboxylic acids were acetic (average: 0.93 μg ml−1) and formic (0.61) followed by oxalic (0.38), succinic (0.15) and malonic (0.20) acids. Pyruvic acid was below the detection limit (<0.08) in all samples. The BC concentration was 1.15 and OC 4.81 μg ml−1 on average. Relating carboxylic acid concentrations to OC, the monocarboxylic acids alone represent 9.3% of OC. Adding the dicarboxylic acids, this average value increases to 11%. Although they are major components, no general trend could be seen between carboxylic acid and OC concentrations.
Keywords: Cloud water; Organic carbon; Black carbon; C1–C2 monocarboxylic acids; C2–C4 dicarboxylic acids;

This paper describes parallelization of a 3-D Lagrangian stochastic atmospheric dispersion model using both distributed- and shared-memory methods. Shared-memory parallelism is implemented through the use of OpenMP compiler directives. Distributed-memory parallelism relies on the MPI message-passing library. One or both (using MPI for inter-node and OpenMP for intra-node communication) of the parallel modes can be used depending upon the requirements of the problem and the computational platform available. The distributed-memory version achieves a nearly linear decrease in execution time as the number of processors is increased. As the number of particles per processor is lowered, performance is limited by the decrease in work per processor and by the need to produce one set of output files. The shared-memory version achieves a speedup factor of ∼1.4 running on machines with four processors.
Keywords: Parallel processing; Atmospheric dispersion; Lagrangian stochastic particle model; Random walk model; OpenMP; MPI;