Neural network and multiple regression models for PM10 prediction in Athens: a comparative assessment

Particulate atmospheric pollution in urban areas is considered to have significant impact on human health. Therefore, the ability to make accurate predictions of particulate ambient concentrations is important to improve public awareness and air quality management. This study examines the possibility of using neural network methods as tools for daily average particulate matter with aerodynamic diameter <10 microm (PM10) concentration forecasting, providing an alternative to statistical models widely used up to this day. Based on a data inventory, in a fixed central site in Athens, Greece, ranging over a two-year period, and using mainly meteorological variables as inputs, neural network models and multiple linear regression models were developed and evaluated. Comparison statistics used indicate that the neural network approach has an edge over regression models, expressed both in terms of prediction error (root mean square error values lower by 8.2-9.4%) and of episodic prediction ability (false alarm rate values lower by 7-13%). The results demonstrate that artificial neural networks (ANNs), if properly trained and formed, can provide adequate solutions to particulate pollution prognostic demands.     

A new method to estimate air-quality levels using a synoptic-regression approach. Part I: Present-day O3 and PM10 analysis

In order to make projections for future air-quality levels, a robust methodology is needed that succeeds in reconstructing present-day air-quality levels. At present, climate projections for meteorological variables are available from Atmospheric-Ocean Coupled Global Climate Models (AOGCMs) but the temporal and spatial resolution is insufficient for air-quality assessment. Therefore, a variety of methods are tested in this paper in their ability to hindcast maximum 8 hourly levels of O₃ and daily mean PM₁₀ from observed meteorological data. The methods are based on a multiple linear regression technique combined with the automated Lamb weather classification. Moreover, we studied whether the above-mentioned multiple regression analysis still holds when driven by operational ECMWF (European Center for Medium-Range Weather Forecast) meteorological data. The main results show that a weather type classification prior to the regression analysis is superior to a simple linear regression approach. In contrast to PM₁₀ downscaling, seasonal characteristics should be taken into account during the downscaling of O₃ time series. Apart from a lower explained variance due to intrinsic limitations of the regression approach itself, a lower variability of the meteorological predictors (resolution effect) and model deficiencies, this synoptic-regression-based tool is generally able to reproduce the relevant statistical properties of the observed O₃ distributions important in terms of European air quality Directives and air quality mitigation strategies. For PM₁₀, the situation is different as the approach using only meteorology data was found to be insufficient to explain the observed PM₁₀ variability using the meteorological variables considered in this study.     

A review of statistical methods for the meteorological adjustment of tropospheric ozone

A variety of statistical methods for meteorological adjustment of ozone have been proposed in the literature over the last decade for purposes of forecasting, estimating ozone time trends, or investigating underlying mechanisms from an empirical perspective. The methods can be broadly classified into regression, extreme value, and space–time methods. We present a critical review of these methods, beginning with a summary of what meteorological and ozone monitoring data have been considered and how they have been used for statistical analysis. We give particular attention to the question of trend estimation, and compare selected methods in an application to ozone time series from the Chicago area. We conclude that a number of approaches make useful contributions to the field, but that no one method is most appropriate for all purposes and all meteorological scenarios. Methodological issues such as the need for regional-scale analysis, the nonlinear dependence of ozone on meteorology, and extreme value analysis for trends are addressed. A comprehensive and reliable methodology for space–time extreme value analysis is attractive but lacking.     

An alternative way to determine the size distribution of airborne particulate matter

We developed and tested a methodology to extract both the size-segregated source apportionment of atmospheric aerosol and the size distribution of each detected element. The experiment is based on the parallel use of a standard low-volume sampler to collect Particulate Matter (PM) and an Optical Particle Counter (OPC). The approach is complementary to size-segregated PM sampling, and it was tested versus a 12-stage cascade impactor. Samples were collected inside the urban area of Genoa (Italy) and their elemental composition was measured by Energy Dispersive-X Ray Fluorescence (ED-XRF). Positive Matrix Factorization (PMF) was applied to time series of elemental concentrations to identify major PM sources, and both PM mass concentration and size-segregated particle number concentration were apportioned. Source profiles and temporal trends extracted by PMF were analyzed together with the OPC data to obtain the size distribution for several elements. The new methodology proved to be reliable for the PM apportionment as well as in providing the elemental concentrations in PM10, PM2.5, and PM1 (PM with aerodynamic diameter, D_ae < 10, 2.5, and 1 μm, respectively). The elemental size distributions are in good agreement with those obtained by the cascade impactor for several elements but some discrepancies, in particular for traffic emissions, are stressed and discussed in the text. The new methodology has two main advantages: it only requires standard semi-automatic sampling equipment and compositional analysis and it provides size-segregated information averaged over quite long periods (typically several months). This is particularly important since campaigns with cascade impactors are generally laborious and thus limited to short periods.     

Respirable particulate matter at an urban and nearby industrial location: concentrations and variability and synoptic weather conditions during high pollution episodes

Continuous data of the concentration measurements of respirable suspended particulates (PM10, particles with aerodynamic diameter smaller than or equal to 10 pm) were analyzed. These measurements were carried out at an urban and nearby industrial location in northern Greece for the 5-year period 1996-2000. The time series concentration trend was examined, the seasonal and diurnal variations were identified, and the lognormality of the daily mean concentration data sets was tested. Over the 5-year data-gathering period, the days on which the U.S. Environmental Protection Agency (EPA) 24-hr PM10 standard was exceeded (episode days) were identified and their relation to prevailing synoptic-scale meteorological conditions was studied. The analysis led to useful information concerning the air quality levels, the contribution of the main pollution sources in this area, as well as some of the mechanisms that influence the PM10 concentrations. It also was proved that the measured PM10 concentrations are a result of a combination of processes including local anthropogenic sources, mesoscale transport, and resuspension. A complex system of sources and meteorological conditions modulate the heavy particulate pollution in the area of interest.     

Evaluation of artificial neural networks for fine particulate pollution (PM10 and PM2.5) forecasting

Multi-layer perceptron (MLP) artificial neural network (ANN) models are compared with traditional multiple regression (MLR) models for daily maximum and average O3 and particulate matter (PM10 and PM2.5) forecasting. MLP particulate forecasting models show little if any improvement over MLR models and exhibit less skill than do O3 forecasting models. Meteorological variables (precipitation, wind, and temperature), persistence, and co-pollutant data are shown to be useful PM predictors. If MLP approaches are adopted for PM forecasting, training methods that improve extreme value prediction are recommended.     

Neural Network and Multiple Regression Models for PM10 Prediction in Athens: A Comparative Assessment

Abstract

Particulate atmospheric pollution in urban areas is considered to have significant impact on human health. Therefore, the ability to make accurate predictions of particulate ambient concentrations is important to improve public awareness and air quality management. This study examines the possibility of using neural network methods as tools for daily average particulate matter with aerodynamic diameter <10 µm (PM₁₀) concentration forecasting, providing an alternative to statistical models widely used up to this day. Based on a data inventory, in a fixed central site in Athens, Greece, ranging over a two-year period, and using mainly meteorological variables as inputs, neural network models and multiple linear regression models were developed and evaluated. Comparison statistics used indicate that the neural network approach has an edge over regression models, expressed both in terms of prediction error (root mean square error values lower by 8.2–9.4%) and of episodic prediction ability (false alarm rate values lower by 7–13%). The results demonstrate that artificial neural networks (ANNs), if properly trained and formed, can provide adequate solutions to particulate pollution prognostic demands.     

A methodology for the quantification of the net African dust load in air quality monitoring networks

This paper proposes and validates a methodology for the quantification of the daily African PM load during dust outbreaks in southern Europe. The daily net dust load in PM10 attributable to an African episode in a given region can be obtained by subtracting the daily regional background (RB) level from the PM10 concentration value at an RB station. As demonstrated in this paper, the daily RB level can be obtained by applying a monthly moving 30th percentile to the PM10 time series at an RB station after a prior extraction of the data of the days with African dust transport. The daily PM10 RB levels obtained can be subtracted from the daily PM10 levels recorded at the same RB site only on days when the occurrence of African dust outbreaks was demonstrated, the difference being the daily net African dust load. It is thus possible to quantify the African dust contribution during an African PM event in southern Europe without the need for PM speciation.The validation of this methodology was performed by comparing the estimated net dust load during African dust outbreaks at three RB stations and the crustal load determined by chemical speciation of PM10 filters. The correlation (r²>0.6) and the equivalence (correlation lines’ slopes ∼1) were significant in the three cases.     

Spatiotemporal Features of Severe Air Pollution in Northern Taiwan (8 pp)

Background, Aims and Scope This research attempted to identify the dominant factors simultaneously affecting the airborne concentrations of five air pollutants with principal component analysis and to determine the meteorologically related parameters that cause severe air-pollution events. According to the definition of subPSI and PSI values through the U.S. EPA, the historical raw data of five criteria air pollutants, SO2, CO, O3, PM10 and NO2, were calculated as daily subPSI values. In addition to the airborne concentrations, this study simultaneous collected the surface meteorological parameters of the Taipei meteorological station, established by the Central Weather Bureau. Methods Principal component analysis was conducted to screen severe air pollution scenarios for five air pollutants: SO2, CO, O3, PM10 and NO2. The concentrations of various air pollutants measured at 17 air-quality stations in northern Taiwan from 1995 to 2001 were transformed into daily subPSI values. The correlation analysis of the five air pollutants and four meteorological parameters (wind speed, temperature, mixing height and ventilation rate) were included in this research. After screening severe air pollution scenarios, this study recognized the synoptic patterns easily causing the severe air-pollution events. Results and Discussion Analytical results showed that the eigenvalues of the first two principal components for SO2, CO, O3, PM10 and NO2 were greater than 1. The first component of five air pollutants explained 64, 64, 67, 76 and 63% of subPSI variance for SO2, CO, O3, PM10 and NO2, respectively. Only the correlation coefficient of NO2 and CO had statistically significant positive values (0.82); other pollutant pairs presented medium (0.4 to 0.7) or low (0 to 0.4) positive values. The correlation coefficients for air pollutants and three meteorological parameters (wind speed, mixing height and ventilation index) were medium or low negative values. In northern Taiwan, spring was most likely induced high concentrations and the component scores of the first component for SO2, CO, PM10 and NO2; summer was the worst season that caused high O3 episodes. Consequently, the analytical results of factor loadings for the first principal component and emission inventory of various sources revealed that mobile sources were dominant factors affecting ambient air quality in northern Taiwan. Conclusion According to the results of principal component analysis for the five air pollutants, the first two of 17 components were cited as major factors and explained 71% of subPSI variance. Based on the inventory of NOx emissions and the isopleth diagram of factor loading for the first component, mobile sources in the southwest Taipei City accounted for the highest factor loading values and emission inventory values. Synoptic analysis and principal component analysis demonstrated that three types of weather patterns (high-pressure recirculation, prefrontal warm sector and the southwesterly wind system) easily caused the severe air-pollution scenarios. In summary, if severe air-pollution days occurred, the average meteorological parameters experienced adverse conditions for diffusing air pollutants; that is, the average values of wind speed, mixing height and ventilation index were lower than 2.1 ms-1, 360 m and 800 m2s-1, respectively. If one of the three synoptic patterns were to occur in combination with adverse meteorological conditions, severe air-pollution events would be developed. Recommendation and Outlook By utilizing synoptic patterns, this work found three weather systems easily caused severe air-pollution events over northern Taiwan. Analytical results showed, respectively, the wind speed and mixing height were less than 2.1 m/s and 360 m during severe air-pollution events.     

Source contributions to fine particulate matter in an urban atmosphere

This paper proposes a practical method for estimating source attribution by using a three-step methodology. The main objective of this study is to explore the use of the three-step methodology for quantifying the source impacts of 24-h PM2.5 particles at an urban site in Seoul, Korea. 12-h PM2.5 samples were collected and analyzed for their elemental composition by ICP-AES/ICP-MS/AAS to generate the source composition profiles. In order to assess the daily average PM2.5 source impacts, 24-h PM2.5 and polycyclic aromatic hydrocarbons (PAH) ambient samples were simultaneously collected at the same site. The PM2.5 particle samples were then analyzed for trace elements. Ionic and carbonaceous species concentrations were measured by ICP-AES/ICP-MS/AAS, IC, and a selective thermal MnO2 oxidation method. The 12-h PM2.5 chemical data was used to estimate possible source signatures using the principal component analysis (PCA) and the absolute principal component scores method followed by the multiple linear regression analysis. The 24-h PM2.5 source categories were extracted with a combination of PM2.5 and some PAH chemical data using the PCA, and their quantitative source contributions were estimated by chemical mass balance (CMB) receptor model using the estimated source profiles and those in the literature. The results of PM2.5 source apportionment using the 12-h derived source composition profiles show that the CMB performance indices; chi2, R2, and percent of mass accounted for are 2.3%, 0.97%, and 100.7%, which are within the target range specified. According to the average PM2.5 source contribution estimate results, motor vehicle exhaust was the major contributor at the sampling site, contributing 26% on average of measured PM2.5 mass (41.8 microg m-3), followed by secondary sulfate (23%) and nitrate (16%), refuse incineration (15%), soil dust (13%), field burning (4%), oil combustion (2.7%), and marine aerosol (1.3%). It can be concluded that quantitative source attribution to PM2.5 in an urban area where source profiles have not been developed can be estimated using the proposed three-step methodology approach.     

Model for forecasting expressway fine particulate matter and carbon monoxide concentration: application of regression and neural network models

The Borman Expressway is a heavily traveled 16-mi segment of the Interstate 80/94 freeway through Northwestern Indiana. The Lake and Porter counties through which this expressway passes are designated as particulate matter < 2.5 microm (PM2.5) and ozone 8-hr standard nonattainment areas. The Purdue University air quality group has been collecting PM2.5, carbon monoxide (CO), wind speed, wind direction, pressure, and temperature data since September 1999. In this work, regression and neural network models were developed for forecasting hourly PM2.5 and CO concentrations. Time series of PM2.5 and CO concentrations, traffic data, and meteorological parameters were used for developing the neural network and regression models. The models were compared using a number of statistical quality indicators. Both models had reasonable accuracy in predicting hourly PM2.5 concentration with coefficient of determination -0.80, root mean square error (RMSE) <4 microg/m3, and index of agreement (IA) > 0.90. For CO prediction, both models showed moderate forecasting performance with a coefficient of determination -0.55, RMSE < 0.50 ppm, and IA -0.85. These models are computationally less cumbersome and require less number of predictors as compared with the deterministic models. The availability of real time PM2.5 and CO forecasts will help highway managers to identify air pollution episodic events beforehand and to determine mitigation strategies.     

Chemical characterisation of PM episodes in NE Spain

The chemical composition of ambient particulate matter (PM) varies widely as a function of its main emission sources and of the chemical reactions which take place in the atmosphere. The aim of this study is to obtain the chemical profile of PM10 and PM2.5 during peak PM episodes, thus identifying the main emission sources and/or atmospheric processes which originate the PM episodes. To this end, cluster analysis was applied to a set of PM10 and PM2.5 data collected throughout 2001 in two urban and industrialised areas in NE Spain. As a result of this analysis, five clusters were identified for each site, and the interpretation of their chemical profiles lead to the identification of five types of peak PM episodes for each site: industrial, traffic and regional re-circulation episodes at both sites, plus crustal episodes in Barcelona, and peak traffic and industrial episodes (T+I) in Tarragona. Traffic episodes are characterised by daily means of 23 and 10 microg/m3 of OM+EC in Barcelona and Tarragona in PM10. Levels of secondary inorganic aerosols reach average daily means of 19 and 11 microg/m3 in Barcelona and Tarragona in PM10 during industrial episodes. High levels of sulphate (>5 microg/m3) and ozone (up to 77 microg/m3 daily mean) are good tracers of regional re-circulation episodes. During crustal episodes daily means of crustal elements reach up to 34 microg/m3 in Barcelona. Special attention has been drawn to the composition of the mineral matter during the different PM episodes.     

Urban atmospheric pollution: personal exposure versus fixed monitoring station measurements

We initiated the PETER (pedestrian environmental traffic pollutant exposure research) project to investigate pedestrians' exposure to traffic related atmospheric pollutants, based on data obtained with the collaboration of selected categories of pedestrian urban workers. We investigated relations between roadside personal exposure levels of volatile aromatic hydrocarbons (including benzene) and particulate matter <10 microm (PM10) among traffic police (n = 126) and parking wardens (n = 50) working in downtown Bologna, Italy. Data were collected from workshifts throughout four 1-week periods in different seasons of 2000-2001. For benzene and PM10, comparisons were made with measurements by fixed monitoring stations, and influence of localized traffic intensity and meteorological parameters was examined. Roadside personal exposure to benzene correlated more strongly with other volatile aromatic hydrocarbons (toluene, xylenes and ethylbenzene) than with PM10. Benzene and PM10 personal exposure levels were higher than fixed monitoring station values (both p<0.0001). At multivariate analysis, benzene and PM10 data from fixed monitoring stations both correlated with meteorological variables, and were also influenced by localized traffic intensity. Plausibly because of the downtown canyon-like streets, weather conditions (during a period of drought) only marginally affected benzene personal exposure, and moderately affected PM10 personal exposure. These findings reinforce the concept that urban atmospheric pollution data from fixed air monitoring stations cannot automatically be taken as indications of roadside exposures.     

Prediction of maximum of 24-h average of PM10 concentrations 30 h in advance in Santiago,Chile

We have developed a neural network based model that uses values of PM10 concentrations measured until 6 p.m. on the present day plus measured and forecasted values of meteorological variables as input in order to predict the level reached by the maximum of the 24-h moving average (24MA) of PM10 concentration on the next day. We have adjusted the parameters of the model using 1998 data to predict 1999 conditions and 1999 data to forecast 2000 maximum concentrations. We have found that among the relevant meteorological input variables, the forecasted difference between maximum and minimum temperature is the most important. Due to the fact that local authorities impose restrictions to emissions on days when the maximum of 24MA of PM10 concentration is expected to exceed 240 μg/m³, we have corrected the measured concentrations on these days before evaluating the efficacy of the forecasting model. Percent errors in forecasting the numerical value are of the order of 20%. The performance of the neural network is better than that of a linear model with the same inputs, but the difference being not great is an indication that the right choice of input variables may be more important than the decision to use a linear or a nonlinear model.     

Spatial heterogeneity of PM10 and O3 in São Paulo, Brazil, and implications for human health studies

Developing exposure estimates is a challenging aspect of investigating the health effects of air pollution. Pollutant levels recorded at centrally located ambient air quality monitors in a community are commonly used as proxies for population exposures. However, if ample intraurban spatial variation in pollutants exists, city-wide averages of concentrations may introduce exposure misclassification. We assessed spatial heterogeneity of particulate matter with an aerodynamic diameter < or = 10 microm (PM10) and ozone (O3) and evaluated implications for epidemiological studies in S?o Paulo, Brazil, using daily (24-hr) and daytime (12-hr) averages and 1-hr daily maximums of pollutant levels recorded at the regulatory monitoring network. Monitor locations were also analyzed with respect to a socioeconomic status index developed by the municipal government. Hourly PM10 and O3 data for the Sāo Paulo Municipality and Metropolitan Region (1999-2006) were used to evaluate heterogeneity by comparing distance between monitors with pollutants' correlations and coefficients of divergence (CODs). Both pollutants showed high correlations across monitoring sites (median = 0.8 for daily averages). CODs across sites averaged 0.20. Distance was a good predictor of CODs for PM10 (p < 0.01) but not O3, whereas distance was a good predictor of correlations for O3 (p < 0.01) but not PM10. High COD values and low temporal correlation indicate a spatially heterogeneous distribution of PM10. Ozone levels were highly correlated (r > or = 0.75), but high CODs suggest that averaging over O3 levels may obscure important spatial variations. Of municipal districts in the highest of five socioeconomic groups, 40% have > or = 1 monitor, whereas districts in the lowest two groups, representing half the population, have no monitors. Results suggest that there is a potential for exposure misclassification based on the available monitoring network and that spatial heterogeneity depends on pollutant metric (e.g., daily average vs. daily 1-hr maximum). A denser monitoring network or alternative exposure methods may be needed for epidemiological research. Findings demonstrate the importance of considering spatial heterogeneity and differential exposure misclassification by subpopulation.     

Extreme value modeling for the analysis and prediction of time series of extreme tropospheric ozone levels: a case study

The occurrence of high concentrations of tropospheric ozone is considered as one of the most important issues of air management programs. The prediction of dangerous ozone levels for the public health and the environment, along with the assessment of air quality control programs aimed at reducing their severity, is of considerable interest to the scientific community and to policy makers. The chemical mechanisms of tropospheric ozone formation are complex, and highly variable meteorological conditions contribute additionally to difficulties in accurate study and prediction of high levels of ozone. Statistical methods offer an effective approach to understand the problem and eventually improve the ability to predict maximum levels of ozone. In this paper an extreme value model is developed to study data sets that consist of periodically collected maxima of tropospheric ozone concentrations and meteorological variables. The methods are applied to daily tropospheric ozone maxima in Guadalajara City, Mexico, for the period January 1997 to December 2006. The model adjusts the daily rate of change in ozone for concurrent impacts of seasonality and present and past meteorological conditions, which include surface temperature, wind speed, wind direction, relative humidity, and ozone. The results indicate that trend, annual effects, and key meteorological variables along with some interactions explain the variation in daily ozone maxima. Prediction performance assessments yield reasonably good results.     

Origin of the exceedances of the European daily PM limit value in regional background areas of Spain

The origin of the daily exceedances of 50 μg PM10 m⁻³ (daily limit value or DLV of the EU air quality directive) and of an arbitrary daily value (DV) 35 μg PM2.5 m⁻³ recorded in 2001–2003 in 13 regional background stations of the Iberian Peninsula were interpreted. This was carried out by means of back-trajectory analysis, available PM model outputs, satellite data and meteorological maps. This allows the detection of high PM episodes on a regional scale and the study of their seasonal and geographical variability.The number of exceedances of the PM10 DLV ranged in 2001–2003 from 6 to 41 depending on the monitoring site. For the selected PM2.5 DV, the range of daily exceedances was 0–10 in the study period.The majority of the PM10 (>70% in most stations) and PM2.5 (17–55% in most stations) exceedances in regional background monitoring stations are caused by African dust outbreaks. These exceedances were less frequent in winter than in summer due to: (a) the frequent long range transport of dust in the warm seasons over Iberia, (b) the re-suspension associated with convective atmospheric dynamics, and (c) the relative low rainfall favouring re-suspension and high residence time of PM. Moreover, a regional contribution of secondary aerosols derived from the efficient photochemical transformation of gaseous precursors may coincide with African transport in summer.Episodes with lack of advective conditions caused 2–29% and 20–50% of the PM10 and PM2.5 exceedances. These occurred mainly in summer due to poor renovation of air masses, increased convective re-suspension, dispersion of pollutants towards rural areas and regional re-circulation and aging of air masses which result in the proliferation of secondary inorganic species.Long-range transport of PM from continental Europe caused exceedances (9–40% and 18–38% of the PM10 and PM2.5 exceedances, respectively), only in northern Iberia because, as the European air masses evolve towards the south, the pollutants suffer dispersion/dilution. Local exceedances are associated with the advection of the clean Atlantic air masses, which cannot increase PM levels to a great extent without the influence of a local source of PM. The proportion of local exceedances of PM10 and PM2.5 ranged 6–33% and 17–40%, respectively.     

Development and preliminary evaluation of a particulate matter emission factor model for European motor vehicles

Although modeling of gaseous emissions from motor vehicles is now quite advanced, prediction of particulate emissions is still at an unsophisticated stage. Emission factors for gasoline vehicles are not reliably available, since gasoline vehicles are not included in the European Union (EU) emission test procedure. Regarding diesel vehicles, emission factors are available for different driving cycles but give little information about change of emissions with speed or engine load. We have developed size-specific speed-dependent emission factors for gasoline and diesel vehicles. Other vehicle-generated emission factors are also considered and the empirical equation for re-entrained road dust is modified to include humidity effects. A methodology is proposed to calculate modal (accelerating, cruising, or idling) emission factors. The emission factors cover particle size ranges up to 10 microns, either from published data or from user-defined size distributions. A particulate matter emission factor model (PMFAC), which incorporates virtually all the available information on particulate emissions for European motor vehicles, has been developed. PMFAC calculates the emission factors for five particle size ranges [i.e., total suspended particulates (TSP), PM10, PM5, PM2.5, and PM1] from both vehicle exhaust and nonexhaust emissions, such as tire wear, brake wear, and re-entrained road dust. The model can be used for an unlimited number of roads and lanes, and to calculate emission factors near an intersection in user-defined elements of the lane. PMFAC can be used for a variety of fleet structures. Hot emission factors at the user-defined speed can be calculated for individual vehicles, along with relative cold-to-hot emission factors. The model accounts for the proportions of distance driven with cold engines as a function of ambient temperature and road type (i.e., urban, rural, or motorway). A preliminary evaluation of PMFAC with an available dispersion model to predict the airborne concentration in the urban environment is presented. The trial was on the A6 trunk road where it passes through Loughborough, a medium-size town in the English East Midlands. This evaluation for TSP and PM10 was carried out for a range of traffic fleet compositions, speeds, and meteorological conditions. Given the limited basis of the evaluation, encouraging agreement was shown between predicted and measured concentrations.     

Innovative approach for estimating fugitive gaseous fluxes using computed tomography and remote optical sensing techniques.

This paper presents a new approach to quantify emissions from fugitive gaseous air pollution sources. The authors combine Computed Tomography (CT) with Path-Integrated Optical Remote Sensing (PI-ORS) concentration data in a new field beam geometry. Path-integrated concentrations are sampled in a vertical plane downwind from the source along several radial beam paths. An innovative CT technique, which applies the Smooth Basis Function Minimization method to the beam data in conjunction with measured wind data, is used to estimate the total flux from the fugitive source. The authors conducted a synthetic data study to evaluate the proposed methodology under different meteorological conditions, beam geometry configurations, and simulated measurement errors. The measurement errors were simulated based on data collected with an Open-Path Fourier Transform Infra-Red system. This approach was found to be robust for the simulated errors and for a wide range of fluctuating wind directions. In the very sparse beam geometry examined (eight beam paths), successful emission rates were retrieved over a 70 degrees range of wind directions under extremely large measurement error conditions.