Wind Direction Bias in Generating Wind Roses and Conducting Sector-Based Air Dispersion Modeling

Abstract

Certain widely used wind rose programs and air dispersion models use an overly simple data-transfer algorithm that induces a directional bias in their output products. The purpose of this paper is to provide a revised algorithm that corrects the directional bias that occurs from the aliasing that occurs when the sector widths used to report wind direction data are on the same order of magnitude, but not equal, to the sector widths used in the wind direction summaries. The directional bias issue arises when output products in 16 direction sectors (22.5° each) are produced from wind direction data reported in terms of 36 sectors (10° each). The result directional bias affects the results of simulations of air and surface concentrations using widely applied air dispersion models. Datasets or models with the directional bias discussed here give consistent positive biases (～30%) for cardinal direction sectors (north, south, east, and west) and consistent negative biases for all of the other sectors (around [?10%). Data summary and air dispersion programs providing outputs in direction sectors that do not match the observational sectors need to be checked for this bias. A revised data-transfer algorithm is provided that corrects the directional bias that can occur in transferring wind direction data between different sector widths.     

A Method of Predicting Point and Path- Averaged Ambient Air VOC Concentrations,Using Meteorological Data

A method of predicting point and path-averaged ambient air VOC concentrations is described. This method was developed for the case of a plume generated from a single point source, and is based on the relationship between wind directional frequency and concentration. One-minute means of wind direction and wind speed were used as inputs to a Gaussian dispersion model to develop this relationship.

Both FTIR spectrometry and a whole-air sampling method were used to monitor VOC plumes during simulated field tests. One test set was also conducted using only whole-air samplers deployed in a closely-spaced network, thus providing an evaluation of the prediction technique free of any bias that might exist between the two analytical methods.

Correlations between observed point concentrations and wind directional frequencies were significant at the 0.05 level in most cases. Predicted path-integrated concentrations, based on observed point concentrations and meteorological data, were strongly correlated with observed values. Predicted point concentrations, based on observed path-integrated concentrations and meteorological data, accurately reflected the location and magnitude of the highest concentrations from each test, as well as the shape of the concentration-versus-crosswind distance curve.     

Improving pollutant source characterization by better estimating wind direction with a genetic algorithm

In homeland security applications, it is often necessary to characterize the source location and strength of a potentially harmful contaminant. Correct source characterization requires accurate meteorological data such as wind direction. Unfortunately, available meteorological data is often inaccurate or unrepresentative, having insufficient spatial and temporal resolution for precise modeling of pollutant dispersion. To address this issue, a method is presented that simultaneously determines the surface wind direction and the pollutant source characteristics. This method compares monitored receptor data to pollutant dispersion model output and uses a genetic algorithm (GA) to find the combination of source location, source strength, and surface wind direction that best matches the dispersion model output to the receptor data. A GA optimizes variables using principles from genetics and evolution.The approach is validated with an identical twin experiment using synthetic receptor data and a Gaussian plume equation as the dispersion model. Given sufficient receptor data, the GA is able to reproduce the wind direction, source location, and source strength. Additional runs incorporating white noise into the receptor data to simulate real-world variability demonstrate that the GA is still capable of computing the correct solution, as long as the magnitude of the noise does not exceed that of the receptor data.     

Evaluation of a Gaussian and a Lagrangian model against a roadside data set,with emphasis on low wind speed conditions

The evaluation of the high percentiles of concentration distributions is required by most national air quality guidelines, as well as the EU directives. However, it is problematic to compute such high percentiles in stable, low wind speed or calm conditions. This study utilizes the results of a previous measurement campaign near a major road at Elimäki in southern Finland in 1995, a campaign specifically designed for model evaluation purposes. In this study, numerical simulations were performed with a Gaussian finite line source dispersion model CAR-FMI and a Lagrangian dispersion model GRAL, and model predictions were compared with the field measurements. In comparison with corresponding results presented previously in the literature, the agreement of measured and predicted data sets was good for both models considered, as measured using various statistical parameters. For instance, considering all NO_x data (N=587), the so-called index of agreement values varied from 0.76 to 0.87 and from 0.81 to 1.00 for the CAR-FMI and GRAL models, respectively. The CAR-FMI model tends to slightly overestimate the NO_x concentrations (fractional bias FB=+14%), while the GRAL model has a tendency to underestimate NO_x concentrations (FB=−16%). The GRAL model provides special treatment to account for enhanced horizontal dispersion in low wind speed conditions; while such adjustments have not been included in the CAR-FMI model. This type of Lagrangian model therefore predicts lower concentrations, in conditions of low wind speeds and stable stratification, in comparison with a standard Lagrangian model. In low wind speed conditions the meandering of the flow can be quite significant, leading to enhanced horizontal dispersion. We also analyzed the difference between the model predictions and measured data in terms of the wind speed and direction. The performance of the CAR-FMI model deteriorated as the wind direction approached a direction parallel to the road, and for the lowest wind speeds. However, the performance of the GRAL model varied less with wind speed and direction; the model simulated better the cases of low wind speed and those with the wind nearly parallel to the road.     

Modeling Emissions from Elevated Cylindrical Bins

Abstract

This paper demonstrates how wind tunnel modeling data that accurately describe plume characteristics near an unconventional emission source can be used to improve the near-field downwind plume profiles predicted by conventional air pollution dispersion models. The study considers a vertical, cylindrical-shaped, elevated bin similar to large product storage bins that can be found at many industrial plant sites. Two dispersion models are considered: the U.S. Environmental Protection Agency's ISC2(ST) model and the Ontario Ministry of the Environment and Energy's GAS model. The wind tunnel study showed that plume behavior was contrary to what was predicted using conventional dispersion models such as ISC2(ST) and GAS and default values of input parameters. The wind tunnel data were used to develop a protocol for correcting the dispersion models inputs, resulting in a substantial improvement in the accuracy of the dispersion estimates.     

Assessing the representativeness of monitoring data from an urban intersection site in central London,UK

The wind flow field around urban street-building configurations has an important influence on the microscale pollutant dispersion from road traffic, affecting overall dilution and creating localised spatial variations of pollutant concentration. As a result, the “representativeness” of air quality measurements made at different urban monitoring sites can be strongly dependent on the interaction of the local wind flow field with the street-building geometry surrounding the monitor. The present study is an initial attempt to develop a method for appraising the significance of air quality measurements from urban monitoring sites, using a general application computational fluid dynamics (CFD) code to simulate small-scale flow and dispersion patterns around real urban building configurations. The main focus of the work was to evaluate routine CO monitoring data collected by Westminster City Council at an intersection of street canyons at Marylebone Road, Central London. Many monitors in the UK are purposely situated at urban canyon intersections, which are thought to be local “hot spots” of pollutant emissions, however very limited information exists in the literature on the flow and dispersion patterns associated with them. With the use of simple CFD simulations and the analysis of available monitoring data, it was possible to gain insights into the effect of wind direction on the small-scale dispersion patterns at the chosen intersection, and how that can influence the data captured by a monitor. It was found that a change in wind direction could result in an increase or decrease of monitored CO concentration of up to 80%, for a given level of traffic emissions and meteorological conditions. Understanding and de-coupling the local effect of wind direction from monitoring data using the methods presented in this work could prove a useful new tool for urban monitoring data interpretation.     

Using pollution roses to assess sulfur dioxide impacts in a township downwind of a petrochemical complex

This study used pollution roses to assess sulfur dioxide (SO₂) pollution in a township downwind of a large petrochemical complex based on data collected from a single air quality monitoring station. The pollution roses summarized hourly SO₂ concentrations at the Taishi air quality monitoring station, located approximately 7.8–13.0 km south of the No. 6 Naphtha Cracking Complex in Taiwan, according to 36 sectors of wind direction during the preoperational period (1995–1999) and two postoperational periods (2000–2004 and 2005–2009). The 99th percentile of hourly SO₂ concentrations 350? downwind from the complex increased from 28.9 ppb in the preoperational period to 86.2–324.2 ppb in the two postoperational periods. Downwind SO₂ concentrations were particularly high during 2005–2009 at wind speeds of 6–8 m/sec. Hourly SO₂ levels exceeded the U.S. Environmental Protection Agency (EPA) health-based standard of 75 ppb only in the postoperational periods, with 65 exceedances from 0–10? and 330–350? downwind directions during 2001–2009. This study concluded that pollution roses based on a single monitoring station can be used to investigate source contributions to air pollution surrounding industrial complexes, and that it is useful to combine such directional methods with analyses of how pollution varies between different wind speeds, times of day, and periods of industrial development.

Implications: The pollution roses summarize SO₂ concentrations by wind direction and to investigate source contribution to air quality. Percentile statistics can catch pollution episodes occurring in a very short time at specific wind directions and speeds. The downwind areas have already exceeded regulated 1-hr SO₂ standard since the operation of the complex.     

Using prognostic model-generated meteorological output in the AERMOD dispersion model: an illustrative application in Philadelphia, PA

In this study, we introduce the prospect of using prognostic model-generated meteorological output as input to steady-state dispersion models by identifying possible advantages and disadvantages and by presenting a comparative analysis. Because output from prognostic meteorological models is now routinely available and is used for Eulerian and Lagrangian air quality modeling applications, we explore the possibility of using such data in lieu of traditional National Weather Service (NWS) data for dispersion models. We apply these data in an urban application where comparisons can be made between the two meteorological input data types. Using the U.S. Environment Protection Agency's American Meteorological Society/U.S. Environmental Protection Agency Regulatory Model (AERMOD) air quality dispersion model, hourly and annual average concentrations of benzene are estimated for the Philadelphia, PA, area using both hourly MM5 model-generated meteorological output and meteorological data taken from the NWS site at the Philadelphia International Airport. Our intent is to stimulate a discussion of the relevant issues and inspire future work that examines many of the questions raised in this paper.     

Assessment of the nested grid model estimates for driving regional visibility models in the southwestern United States

The Nested Grid Model (NGM) is a primitive-equation meteorological model that is routinely exercised over North America for forecasting purposes by the National Meteorological Center. While prognostic meteorological models are being increasingly used to drive air quality models, their use in conducting annual simulations requires significant resources. NGM estimates of wind fields and other meteorological variables provide an attractive alternative since they are typically archived and readily available for an entire year. Preliminary evaluation of NGM winds during the summer of 1992 for application to the region surrounding the Grand Canyon National Park showed serious shortcomings. The NGM winds along the borders between California, Arizona and Mexico tend to be northwesterly with a speed of about 6 m/sec, while the observed flow is predominantly southerly at about 2-5 m/sec. The mesoscale effect of a thermal low pressure area over the highly heated Southern California and western Arizona deserts does not appear to be represented by the NGM because of its coarse resolution and the use of sparse observations in that region. Tracer simulations and statistical evaluation against special high resolution observations of winds in the southwest United States clearly demonstrate the northwest bias in NGM winds and its adverse effect on predictions of an air quality model. The "enhanced" NGM winds, in which selected wind observations are incorporated in the NGM winds using a diagnostic meteorological model provide additional confirmation on the primary cause of the northwest bias. This study has demonstrated that in situations where limited resources prevent the use of prognostic meteorological models, previously archived coarse resolution wind fields in which additional observations are incorporated to correct known biases provide an attractive option.     

Mapping air quality zones for coastal urban centers

This study presents a new method that incorporates modern air dispersion models allowing local terrain and land–sea breeze effects to be considered along with political and natural boundaries for more accurate mapping of air quality zones (AQZs) for coastal urban centers. This method uses local coastal wind patterns and key urban air pollution sources in each zone to more accurately calculate air pollutant concentration statistics. The new approach distributes virtual air pollution sources within each small grid cell of an area of interest and analyzes a puff dispersion model for a full year’s worth of 1-hr prognostic weather data. The difference of wind patterns in coastal and inland areas creates significantly different skewness (S) and kurtosis (K) statistics for the annually averaged pollutant concentrations at ground level receptor points for each grid cell. Plotting the S-K data highlights grouping of sources predominantly impacted by coastal winds versus inland winds. The application of the new method is demonstrated through a case study for the nation of Kuwait by developing new AQZs to support local air management programs. The zone boundaries established by the S-K method were validated by comparing MM5 and WRF prognostic meteorological weather data used in the air dispersion modeling, a support vector machine classifier was trained to compare results with the graphical classification method, and final zones were compared with data collected from Earth observation satellites to confirm locations of high-exposure-risk areas. The resulting AQZs are more accurate and support efficient management strategies for air quality compliance targets effected by local coastal microclimates.

Implications: A novel method to determine air quality zones in coastal urban areas is introduced using skewness (S) and kurtosis (K) statistics calculated from grid concentrations results of air dispersion models. The method identifies land–sea breeze effects that can be used to manage local air quality in areas of similar microclimates.     

An Efficient Gaussian-Plume Multiple-Source Air Quality Algorithm

The information presented in this paper is directed to air pollution scientists with an interest in applying air quality simulation models. RAM is the three letter designation for this efficient Gaussian-plume multiple-source air quality algorithm. RAM is a method of estimating short-term dispersion using the Gaussian steady-state model. This algorithm can be used for estimating air quality concentrations of relatively stable pollutants for averaging times from an hour to a day in urban areas from point and area sources. The algorithm is applicable for locations with level or gently rolling terrain where a single wind vector for each hour is a good approximation to the flow over the source area considered. Calculations are performed for each hour. Hourly meteorological data required are wind direction, wind speed, stability class, and mixing height. Emission information required of point sources consists of source coordinates, emission rate, physical height, stack gas volume flow and stack gas temperature. Emission information required of area sources consists of south-west corner coordinates, source area, total area emission rate and effective area source height. Computation time is kept to a minimum by the manner in which concentrations from area sources are estimated using a narrow plume hypothesis and using the area source squares as given rather than breaking down all sources to an area of uniform elements. Options are available to the user to allow use of three different types of receptor locations: 1 ) those whose coordinates are input by the user, 2) those whose coordinates are determined by thé model and are downwind óf significant point and area sources where maxima are likely to occur, and 3) those whose coordinates are determined by the model to give good area coverage of a specific portion of the region. Computation time is also decreased by keeping the number of receptors to a minimum.     

A simple model for the dispersion of pollutants from a road tunnel portal

The dispersion of pollutants from a roadway tunnel portal is mainly determined by the interaction between the ambient wind and the jet stream from the tunnel portal. In principal, Eulerian microscale models by solving the conservation equations for mass, momentum, and energy, are thus able to simulate effects such as buoyancy etc. properly. However, for engineering applications such models need too much CPU time, and are not easy to handle by non-scientific personnel. Only a few dispersion models, applicable for regulatory purposes, have so far appeared in the literature. These models are either empirical models not always applicable for different sites, or they do not capture important physical effects like buoyancy phenomena. Here, a rather simple model is presented, which takes into account most of the important processes considered to govern the dispersion of a jet stream from portals. These are the exit velocity, the buoyancy, the influence of ambient wind direction fluctuations on the position of the jet stream, and traffic induced turbulence. Although the model contains some heuristic elements, it was successfully tested against tracer experiments taken near a motorway tunnel portal in Austria. The model requires relatively little CPU time. Current limitations of the model include the neglect of terrain, building, and vehicle effects on the dispersion, and the neglect of the horizontal dispersion arising from entrainment of ambient air in the jet stream. The latter could lead to an underestimation of plume spreads for higher wind speeds. The validation of the model will be the focus of future research. The experimental data set is also available for the scientific community.     

Relating error bounds for maximum concentration estimates to diffusion meteorology uncertainty

The intent of this paper is to relate the magnitude of the error bounds of data, used as inputs to a Gaussian dispersion model, to the magnitude of the error bounds of the model output, which include the estimates of the maximum concentration and the distance to that maximum. The research specifically addresses the uncertainty in estimating the maximum concentrations from elevated buoyant sources during unstable atmospheric conditions, as these are most often of practical concern in regulatory decision making. A direct and quantitative link between the nature and magnitude of the input uncertainty and modeling results has not been previously investigated extensively. The ability to develop specific error bounds, tailored to the modeling situation, allows more informed application of the model estimates to the air quality issues.In this study, a numerical uncertainty analysis is performed using the Monte-Carlo technique to propagate the uncertainties associated with the model input. Uncertainties were assumed to exist in four model input parameters: (1) wind speed, (2) standard deviation of lateral wind direction fluctuations, (3) standard deviation of vertical wind direction fluctuations, and (4) plume rise. For each simulation, results were summarized characterizing the uncertainty in four features of the ground-level concentration pattern predicted by the model: (1) the magnitude of the maximum concentration, (2) the distance to the maximum concentration, and (3) and (4) the areas enclosed within the isopleths of 50% and 25% of the error-free estimate of maximum concentration.The authors conclude that the error bounds for the estimated maximum concentration and the distance to the maximum can be double that of the error bounds for individual model input parameters. The model output error bounds for the areas enclosed within isopleth values can be triple the error bounds of the input. It was not our intent to cover all possible combinations for the error in the input parameters. Ours was a much more limited goal of providing a lower bound estimate of model uncertainty in which we assume the input is reasonably well characterized and there is no bias in the input. These results allow estimation of minimum bounds on errors in model output when considering reasonable input error bounds.     

Vehicle wake factor for heterogeneous traffic in urban environments

Vehicles' 'wakes' are generated as a result of vehicular movements. They are one of the dominant factors in dispersing the pollutants in 'calm' meteorological conditions when wind velocity is <1 m/sec (Chock, 1978). They are used as a wind-speed-correction factor in several air quality models considering the effects of traffic movements on the pollutant dispersion. In this study, the vehicle wake factor (VWF) has been estimated using the inverse general finite line source model, GFLSM, (Luhar and Patil, 1989) for heterogeneous traffic conditions at one of the busiest traffic intersections of the Delhi city, near the Income Tax Office (ITO). The results show that in 'unstable' conditions, the VWF varies between 1.63 and 0.3 (for wind direction, θ = 90°) and 2.5 and 0.8 (for wind direction, θ = 180°). During 'neutral' and 'stable' conditions, it is in the range of 0.84–0.4, 1.91–0.85, (for wind direction θ = 270°) and 1.7–0.7, 3.1–0.3 for wind direction θ = 360°, respectively.     

Frank A. Chambers

Abstract

This work assessed the usefulness of a current air quality model (American Meteorological Society/Environmental Protection Agency Regulatory Model [AERMOD]) for predicting air concentrations and deposition of perfluorooctanoate (PFO) near a manufacturing facility. Air quality models play an important role in providing information for verifying permitting conditions and for exposure assessment purposes. It is important to ensure traditional modeling approaches are applicable to perfluorinated compounds, which are known to have unusual properties. Measured field data were compared with modeling predictions to show that AERMOD adequately located the maximum air concentration in the study area, provided representative or conservative air concentration estimates, and demonstrated bias and scatter not significantly different than that reported for other compounds. Surface soil/grass concentrations resulting from modeled deposition flux also showed acceptable bias and scatter compared with measured concentrations of PFO in soil/grass samples. Errors in predictions of air concentrations or deposition may be best explained by meteorological input uncertainty and conservatism in the PRIME algorithm used to account for building downwash. In general, AERMOD was found to be a useful screening tool for modeling the dispersion and deposition of PFO in air near a manufacturing facility.     

The Transport of Suspended Particulates as a Function of Wind Direction and Atmospheric Conditions

Linear regression of high volume air sampler data and various meteorological parameters was used to determine a suspended particulate air pollution climatology for Albany, NY. A new method for exhibiting associations between wind direction and pollutant levels using correlation coefficients is presented. Correlations between wind direction distribution frequency and other meteorological parameters is employed to help explain differences in correlations for direction with suspended particulate levels. Results show that high particulate concentrations correlate well with southerly wind flow throughout the study area, regardless of relative location of receptor to local sources. This suggests that ambient background concentrations inherent in different air masses more consistently affected suspended particulate levels than did the diffusion from local sources during the study period. Maximum particulate advection occurs under conditions of good mixing of the boundary layer and moderate wind speeds and is enhanced further in the absence of removal processes such as rainout and washout. Trajectory analysis of selected days indicates a definite relationship between path and origin of the wind flow and regional average particulate concentration.     

Assessment of the Nested Grid Model Estimates for Driving Regional Visibility Models in the Southwestern United States

ABSTRACT

The Nested Grid Model (NGM) is a primitive-equation meteorological model that is routinely exercised over North America for forecasting purposes by the National Meteorological Center. While prognostic meteorological models are being increasingly used to drive air quality models, their use in conducting annual simulations requires significant resources. NGM estimates of wind fields and other meteorological variables provide an attractive alternative since they are typically archived and readily available for an entire year. Preliminary evaluation of NGM winds during the summer of 1992 for application to the region surrounding the Grand Canyon National Park showed serious shortcomings. The NGM winds along the borders between California, Arizona and Mexico tend to be northwesterly with a speed of about 6 m/sec, while the observed flow is predominantly southerly at about 2-5 m/sec. The mesoscale effect of a thermal low pressure area over the highly heated Southern California and western Arizona deserts does not appear to be represented by the NGM because of its coarse resolution and the use of sparse observations in that region. Tracer simulations and statistical evaluation against special high resolution observations of winds in the southwest United States clearly demonstrate the northwest bias in NGM winds and its adverse effect on predictions of an air quality model. The “enhanced” NGM winds, in which selected wind observations are incorporated in the NGM winds using a diagnostic meteorological model provide additional confirmation on the primary cause of the northwest bias. This study has demonstrated that in situations where limited resources prevent the use of prognostic meteorological models, previously archived coarse resolution wind fields in which additional observations are incorporated to correct known biases provide an attractive option.     

Numerical modeling on air quality in an urban environment with changes of the aspect ratio and wind direction

Due to heavy traffic emissions within an urban environment, air quality during the last decade becomes worse year by year and hazard to public health. In the present work, numerical modeling of flow and dispersion of gaseous emissions from vehicle exhaust in a street canyon were investigated under changes of the aspect ratio and wind direction. The three-dimensional flow and dispersion of gaseous pollutants were modeled using a computational fluid dynamics (CFD) model which was numerically solved using Reynolds-averaged Navier–Stokes (RANS) equations. The diffusion flow field in the atmospheric boundary layer within the street canyon was studied for different aspect ratios (W/H?=?1/2, 3/4, and 1) and wind directions (θ?=?90°, 112.5°, 135°, and 157.5°). The numerical models were validated against wind tunnel results to optimize the turbulence model. The numerical results agreed well with the wind tunnel results. The simulation demonstrated that the minimum concentration at the human respiration height within the street canyon was on the windward side for aspect ratios W/H?=?1/2 and 1 and wind directions θ?=?112.5°, 135°, and 157.5°. The pollutant concentration level decreases as the wind direction and aspect ratio increase. The wind velocity and turbulence intensity increase as the aspect ratio and wind direction increase.     

Assessing sensitivity of source term estimation

Source term estimation algorithms compute unknown atmospheric transport and dispersion modeling variables from concentration observations made by sensors in the field. Insufficient spatial and temporal resolution in the meteorological data as well as inherent uncertainty in the wind field data make source term estimation and the prediction of subsequent transport and dispersion extremely difficult. This work addresses the question: how many sensors are necessary in order to successfully estimate the source term and meteorological variables required for atmospheric transport and dispersion modeling?The source term estimation system presented here uses a robust optimization technique – a genetic algorithm (GA) – to find the combination of source location, source height, source strength, surface wind direction, surface wind speed, and time of release that produces a concentration field that best matches the sensor observations. The approach is validated using the Gaussian puff as the dispersion model in identical twin numerical experiments. The limits of the system are tested by incorporating additive and multiplicative noise into the synthetic data. The minimum requirements for data quantity and quality are determined by an extensive grid sensitivity analysis. Finally, a metric is developed for quantifying the minimum number of sensors necessary to accurately estimate the source term and to obtain the relevant wind information.     

VOC source–receptor relationships in Houston during TexAQS-II

During the TRAMP field campaign in August–September 2006, C₂–C₁₀ volatile organic compounds (VOCs) were measured continuously and online at the urban Moody Tower (MT) site. This dataset was compared to corresponding VOC data sets obtained at six sites located in the highly industrialized Houston Ship Channel area (HSC). Receptor modeling was performed by positive matrix factorization (PMF) at all sites. Conditional probability functions (CPF) were used to determine the origin of the polluted air masses in the Houston area. A subdivision into daytime and nighttime was carried out to discriminate photochemical influences. Eight main source categories of industrial, mobile, and biogenic emissions were identified at the urban receptor site, seven and six, respectively, at the different HSC sites. At MT natural gas/crude oil contributed most to the VOC mass (27.4%), followed by liquefied petroleum gas (16.7%), vehicular exhaust (15.3%), fuel evaporation (14.3%), and aromatics (13.4%). Also petrochemical sources from ethylene (4.7%) and propylene (3.6%) play an important role. A minor fraction of the VOC mass can be attributed to biogenic sources mainly from isoprene (4.4%). Based on PMF analyses of different wind sectors, the total VOC mass was estimated to be twofold at MT with wind directions from HSC compared to air from a typical urban sector, for petrochemical compounds more than threefold. Despite the strong impact of air masses influenced by industrial sources at HSC, still about a third of the total mass contributions at MT can be apportioned to other sources, mainly motor vehicles and aromatic solvents. The investigation of diurnal variation in combination with wind directional frequencies revealed the greatest HSC impact at the urban site during the morning, and the least during the evening.