Air Pollution Modeling and its Application XXI

Urban sprawl is altering the landscape and leading to changes in emissions and air quality. This study analyses the impacts of alternative urban growth patterns on air quality, through a 1-year application of a modeling system to Porto urban region (Portugal). MM5-CAMx was evaluated to define an adequate setup for the purpose of the study, namely the land use dataset and the spatial distribution of emissions. Two land use scenarios for Porto urban area were defined (SPRAWL and COMPACT). Modeling results revealed that SPRAWL leads to an aggravation of PM10 concentrations. For O

3

, differences between scenarios were smaller. Finally, population exposure to pollutants is higher for COMPACT because more inhabitants are found in areas of higher concentrations.

The goal of this work is to investigate, compare and evaluate three different modelling approaches to describe pollutant dispersion at the microscale and in presence of obstacles. The observed data used as reference for the comparison are those of the Mock Urban Setting Test (MUST) field campaign, providing flow and dispersion data measured within an idealized urban roughness. A case in neutral conditions was chosen.

We present the outcome of the international conference ‘Local Air Quality and its Interactions with Vegetation’, which took place in Antwerp, Belgium on January 21-22, 2010. Results of international CFD studies, measurement campaigns and experimental studies show that vegetation can have an important effect on dispersion patterns determining local air quality. However, there are many parameters involved (vegetation structure, local meteorology, urban canopy characteristics, mechanical turbulence properties) and the results show that the complexity of the mechanisms of vegetation affecting local air quality are often underestimated.

Aerosols have a significant impact on the Earth’s radiation budget by scattering and absorbing solar radiation thus decreasing the radiation absorbed at the surface and increasing low-level static stability. Recent developments in urban air pollution modelling have focused on the introduction of two-way coupling between aerosol modules and the driving meteorological models, aiming to a more accurate description of aerosol-induced radiative forcings in the atmosphere. In this study, the feedback mechanism due to the aerosol direct effect is examined with the aid of a mesoscale Eulerian meteorological model and a chemical transport model, as part of an on-line coupled scheme. For the assessment of the performance of the coupled model system, pollutant dispersion over the Paris area is simulated over a summer period in 2005 identified in the framework of the MEGAPOLI project. A comparison between the standalone and coupled calculations reveals that the radiative forcing due to the direct effect has a substantial impact on the calculated meteorological fields and leads to an improved representation of the pollutant concentration patterns.

This paper presents the results of two recent case studies which were evaluated with the CFD model ENVI-met. In a first case, the impact of a vegetation barrier along a highway on the local air quality is studied. In a second case, the dispersion of total UFP concentrations within a street canyon is examined. In both cases, ENVI-met model output is compared with measurements resulting in an overall acceptable agreement of the model performance.

The projected increase in distributed power generation has given rise to the need for a model to examine the air quality impact of small power plants located in urban areas. This model needs to treat plume rise and dispersion of a buoyant release in an inhomogeneous urban boundary layer whose structure is governed by complex surface characteristics. Models such as AERMOD [Cimorelli AJ, Perry SG, Venkatram A, Weil JC, Paine RJ, Wilson RB, Lee RF, Peters WD, Brode RW (2005) AERMOD: A dispersion model for industrial source applications. Part I: General model formulation and boundary layer characterization. J Appl Meteorol 44(5):682–693] are not designed for such sources, although they do include methods to account for the effects of isolated buildings on dispersion. This paper describes progress in developing a dispersion model that treats the governing processes relevant to an urban boundary layer. The model is being developed using data from a field study conducted in the vicinity of a distributed generator (DG). The data from the field study was supplemented with data from simulations conducted in a water channel.

This numerical work presents a computational fluid dynamics (CFD) model to simulate wind flow over complex site SIRTA which is located in Palaiseau, 20 km south of Paris (France) in a semi urban environment. The purpose is to study the ability of an atmospheric CFD model (Code_Saturne) to reproduce the micro scale heterogeneities of wind and turbulent kinetic energy (TKE), using a 16 month data set collected at SIRTA site. The forested areas were modeled either according to the classical roughness law or with a drag porosity model, in which trees effects are modeled with additional terms in momentum, TKE and dissipation rate equations. The results obtained in this work show that the drag porosity model represents well the special heterogeneities of the site and gives results much closer to experimental results than the classical roughness law in the sector where the influence of forested areas is predominant.

An alternative method for the buoyant plume rise computation is proposed. It is based on the same idea described in a recent paper related with chemical reactions in Lagrangian particle models (Alessandrini S and Ferrero E, Phys A 388:1375–1387, 2009) for simulating the ozone background concentration. A fictitious scalar transported by the particles, the temperature difference between the plume portions and the environment air temperature, is introduced. As a consequence no more particles than those inside the plume have to be released to simulate the entrainment of the background air temperature. In this way the entrainment is properly simulated and the plume rise is calculated from the local property of the flow.

A large-eddy model is developed for a street canyon flow and dispersion of scalars when building/street surfaces are heated by solar radiation. A wall function for temperature and scalar is proposed at the building/street surfaces. The model is applied to a canyon, with the aspect ratio of unity and background wind blowing across the canyon axis, for two idealised heating scenarios: (1) the roof and the upstream wall (UW) are heated, and (2) the roof and the downstream wall (DW) are heated. In Case (1), the mean primary vortex (MPV) driven by the boundary layer wind is

assisted

by the thermal updrafts near the heated UW and the turbulent kinetic energy (TKE) inside the canyon is also enhanced with the increasing wall temperature. In Case (2), however, the MPV is

opposed

by the thermal updrafts near the heated DW and its intensity is inhibited remarkably. On the other hand, the TKE inside the canyon for Case (2) increases with the wall temperature, more significantly in comparison with that for Case (1). Consequently the advective flux driven by the MPV is comparable to the turbulent flux driven by TKE at the roof level for Case (1), but the advective flux is negligible for Case (2). These results are useful for proposing a new parameterisation scheme for a street canyon air quality model.

The stable boundary layer (SBL) is one of the key modeling problems for air quality predictions at local and regional scales. This paper presents an SBL model for the mean wind, temperature, and turbulent flux profiles over a range of stability based on an eddy-diffusivity approach. Model results agree well with large-eddy simulations and observations in both weakly- and very-stable boundary layers.

The composition of air within a street canyon is determined by the composition of background air mixed in from above, advection of air into and out of the canyon, emissions from within the street, and the mixing and chemical processing of pollutants within the canyon. This paper describes a modelling study of street canyon atmospheric composition, integrating the combined effects of emissions, dynamics and chemistry. The research builds upon an existing dynamical model of canyon atmospheric motion (Large Eddy Simulation model) by adding a detailed chemical reaction scheme.

Paper presents three new methods to improve the computational performances of air-pollution modelling methodology based on Lagrangian particle dispersion model. New methods are developed in such a manner that properties of the original air-pollution model are preserved in their original form. All methods are complementary one to another and each one can be integrated into existing methodology separately as a standalone method. The first method is application of the clustering method for a reduction of the computational cost. The second method is the cell concentration kernel density estimation method adaptation that is used to substitute the box counting concentration estimation method. The Lagrangian particle-dispersion control method based on artificial neural networks is presented as third method to control the parameters of the Lagrangian particle model. All methods are applied to Spray Lagrangian particle model (product of Arianet s.r.l.) and validated on two field data sets on highly complex terrain in Slovenia. The validations proved the hypothesis that controlling the number of particles in the simulation using proposed new methods is actually preserving the quality of results at a constant level.

In order to take into account atmospheric radiation and the thermal effects of the buildings in simulations of atmospheric flow and pollution dispersion in urban areas, we have developed a three-dimensional atmospheric radiative scheme in the atmospheric module of the open-source CFD model

Code_Saturne

. This paper describes our ongoing work on the development of this model. The radiative scheme has been previously validated with idealized cases and the results of a real case. Here we present results of the full coupling of the radiative and thermal schemes with the 3D dynamical flow model and compare the results with simpler approaches found in the literature.

Thompson (1991, 1993) measured ground-level concentrations profiles (GLCPs) for 330 combinations of stack height, building type and distance between stack and building. Cosemans and Lefebvre (Dispersion parameters in a wind tunnel and in the field: analysing Thompson’s 1991 wind tunnel data for isolated stacks with IFDM, and its application to building downwash. In: Preprints of the 13th international conference on harmonisation within atmospheric dispersion for regulatory purposes, Paris, pp 304–308, 2010) derived dispersion parameters needed by the bi-Gaussian transport and dispersion equation to reproduce the GLCPs measured for nine isolated stacks. Now, we present a formula to reproduce the 321 other GLCPs, measured. Basically, the plume affected by building downwash is replaced by a set of plumes that have a modified lognormal distribution of height and pollutant mass. The different phenomena that influence plume growth subject to building downwash are modelled by virtual origin functions. Incorporating this in a regulatory plume model could strongly increase the capacity of such model to predict the pollutant concentrations in the vicinity of buildings neighbouring emission sources

In Ukraine, to predict the area of chemical pollution after accidents with toxic substances standard model which is adopted by Government is used. This is the empirical model which possibilities are very restricted. This paper presents numerical models to simulate the flow field and pollutant concentrations in the regions with complex topography. The hydrodynamic model of inviscid flow is used to predict the velocity field over complex terrain and that is the reason of main advantage of these models, because of they do not consume much of computing time. The models can be used for express (quickly ) air quality modeling for cases of air pollution after accidents when there is lack of initial information about meteorological parameters, mass of ejection, etc.

The CUDM system provides transport and dispersion modeling capability at the urban scale. It simulates the mean flow, turbulence and concentration fields in urban areas. The system involves a cascade of meteorological models from the operational regional model to an urbanized meso-scale model that in turn drives a CFD flow model running at the urban scale. This CFD model simulates mean wind and turbulence fields that can feed dispersion models. The presentation will briefly describe the current state of the CUDM system, present results obtained over Vancouver during the Olympics and over Toronto during the G8/G20 summit meetings, and compare the different configurations.

The modeling of chemical reactions between two species at small scales where the chemical equilibrium is not yet attained and the segregation effect may not be negligible depends on the knowledge of the covariance of the concentration field. Unfortunately, while the mean concentration field determination is almost a settled matter, the evaluation of the second moments of the concentration field is still an open problem. In this framework, the fluctuating plume model has proved to be a versatile and powerful tool to evaluate the whole concentration field, nevertheless the Lagrangian simulation of the plume barycenter meandering in real cases is still a challenging task. Following Cassiani and Giostra (Atmos Environ 36:4717–4724, 2002) we use the hypotheses of the fluctuating plume model to develop a simple and fast method to derive the concentration field in real situations without using a Stochastic Lagrangian model, but using a mean concentration field, either measured or numerically simulated. The model is tested against a wind tunnel experiment data-set and the results are presented and discussed.

Ground level ozone levels were simulated for Istanbul during a summer episode in June 2008, using MM5/CMAQ model system. Two sets of base runs were carried out in order to investigate the impact of biogenic emissions on ozone concentrations in Istanbul. The inclusion of biogenic emissions significantly improved the performance of the model, particularly the temporal variation of ozone concentrations. It was found out that biogenic NMVOCs emissions enhanced maximum ozone concentrations in the GIA up to 25 ppb. The sensitivity analysis pointed to NO

x

-sensitive chemistry in the region

An integrated forecasting system, using artificial neural networks (ANNs) and cokriging technique, has been developed to forecast daily mean PM10 concentrations 2 days in advance. The test case has been performed over Milan metropolitan area in northern Italy where PM10 concentrations are continuously monitored by 14 monitoring stations. In the first step, ANNs are identified for each monitoring stations to forecast and in the second step, the forecasted values are interpolated using cokriging algorithms over the whole domain. The use of the MODIS derived PM concentration maps as a secondary variable for cokriging, account for the local spatial patterns of PM10 where no measurements are available. The results are validated in terms of statistical and forecast exceedance indexes. The validation has been performed in two steps: first, the Neural Network model performances have been investigated comparing the point forecast with observations, and then the cokriging algorithm has been validated using leave one out cross validation method. The validation results show good agreement in terms of statistical indexes. The proposed forecasting methodology represents a fast and reliable way to provide decision makers and general public with PM10 forecast over an urban area.

The influence of the approach flow direction on contaminant spreading and ventilation within an intersection in an idealised symmetrical urban area was investigated in this study. Advective horizontal and vertical scalar fluxes are computed from measured data for five flow directions. The highest advective contaminant fluxes are measured in the bottom parts of street-canyons. The important role of the vertical turbulent scalar flux in ventilation of intersection is expected. Quadrant analysis of vertical flux of longitudinal momentum is used to determine a domination of sweep or ejection events above the intersection.

Meteorological forecasts have been analyzed in detail during an episode involving substantially high concentrations of PM10 that occurred in Istanbul on 18–20 November 2009. The episode addressed here is associated with areas of high pressure over the Northwest of Turkey. These conditions caused poor air quality, especially in the Istanbul Metropolitan Area. Simulations were executed with version 7.2.1 of the HIRLAM model. The simulations were able to predict the temperatures, wind speeds and relative humidity at near surface.

WRF Modelling System (ver. 2.2) was running for several days during summer and winter seasons within four nested domains with horizontal resolution of 36, 12, 4 and 1.33 km, respectively. Short-term 24 h simulations, as the input meteorological data for operative air quality forecast, were realized using the NCEP global model forecast data. Simulations were fulfilled for different PBL schemes. The calculated standard meteorological parameters, such as wind speed and direction, temperature and relative humidity, were compared with the ground-level observation data from automatic meteorological stations within the urban area and outside the city. The main object of the study was the urban heat island (UHI) effect and, as its consequence, modification of the wind field and surface layer turbulent structure within and over the urban canopy. The intensity of the UHI effect as function of urban surface characteristics, anthropogenic heat release and season meteorological conditions was evaluated.

Degradation processes of organic substances in wastewater treatment plants give rise to aerosolised microorganisms (bioaerosols) that are generated mainly during aeration phases. The dispersion of such bioaerosols in the air may cause adverse health effects not only to the workers in wastewater plants but also to the population living or working in the adjacent areas. In order to prevent health risks to the population, Italian national laws prescribe a buffer zone (at least 100 m from the plant) regardless of the local situation, i.e. not considering the dimension of the plant, local meteorological conditions etc. In this paper an evaluation of the actual impact of bioaerosols produced by a small wastewater plant located in Northern Italy is reported. The main aim of the study is an estimation of the real extent of the areas affected by the bioaerosols dispersion.

Several experiments of dispersion of radioactive particulate matter after an explosion have been performed by the National Radiation Protection Institute of the Czech Republic. These experiments were set up to simulate an attack by a so-called “dirty bomb”. Many pieces of experimental equipment were used to measure aerosol concentration, deposited activity, dose rates and basic meteorological parameters. In this contribution we perform a computer simulation of this experiment using our in-house CFD (LES) code. We mainly concentrate on deposition of activity in the area of interest and the time evolution of concentration. These experiments are also used as a model short scale scenario for Working group 9 of project EMRAS II coordinated by the IAEA.

The quality of the simulation of the motion and spread of a passive substance in a puff model depends on the ability to estimate the speed of the wind at the source height and standard deviation of concentration both in time and space. The wind profile is calculated using the standard Monin-Obukhov similarity theory and standard deviation of concentration, (“sigmas” in the further text) is calculated using a scheme based on the second-order closure theory. To asses the success of the model comparison was made with two other approaches for the calculation of the “sigmas”, the Briggs empirical method and a semi-empirical scheme based on the statistical theory. All of these methods are based on the various parameters of turbulence which were calculated from standard measurements using the so called meteorological pre-processing. Finally the results were compared with the results from the measurements from Tracer Dispersion Experiments in the Copenhagen Area during 1978/1979. First we have analyzed the wind extrapolation results and when we were satisfied with those we did comparisons with the observed concentration.

There is a discrepancy in data quality between the highly detailed concentration measurements in the surroundings of industrial plants emitting heavy metals and the registered emission data at these sites. When simulating the concentration fields in the direct vicinity of the emitting plants by using the bi-gaussian model IFDM and the reported emissions, the simulated concentrations were much lower than the measured concentrations. Originally, this was thought to be due to diffuse, wind-fugitive emissions not reported in the official inventories. Therefore, inverse modeling was performed to get the emission data and wind dependency of these emissions. It was expected that the emissions coming out of the inverse modeling would follow a power law of the wind speed except for very low and very high wind speeds. In the latter case, a constant emission was expected, while in the former case, no emissions were expected to be found. However, this lower threshold did not seem to exist in the modeled emissions. Furthermore, these emissions seemed to have their source in spots not used for storage of heavy metals such as parking lots. Detailed analysis of these results showed that another effect, known as building downwash, is responsible for this behavior. Thereafter, it was shown that it is possible for a bi-gaussian model that lacks a building downwash module, to simulate correct concentration levels by putting in virtual sources just behind the buildings causing the building downwash phenomenon. By using half of the available concentration data for the inverse modeling and half for the validation, it was shown that this technique can be used to produce detailed and validated concentration maps of the surroundings of the industrial site. Finally, it was shown that in this case study building downwash has an important effect on local concentrations and that a better representation of building downwash is needed in bi-gaussian models to describe the complex dispersion patterns in the wake of industrial sites.

We modelled the air quality in Switzerland for summer and winter periods in 2006 using the CAMx model with the focus on aerosols. The contributions of various source regions to aerosol concentrations at receptor sites were investigated using the Particulate Matter Source Apportionment Technology (PSAT) tool of CAMx. The source regions were defined as Switzerland, surrounding countries and other countries in Europe. Organic aerosols (OA) and particulate nitrate (PNO

3

) are the main components of winter aerosols in the modeled domain. In summer, organic aerosols dominate the aerosol composition and they are mainly secondary. In general more than 50% of PM2.5 was predicted to come from Swiss sources except for a few sites at the border regions. Primary particles originate mostly from local sources in both seasons, except at Basel where significant contribution from France and Germany was predicted. In the south, Italy is the main contributor to aerosols. On the other hand, about half of the secondary aerosols originate from Germany and France for receptors in the north and from Italy for receptors in the south.

During July 2005, a particularly high ozone episode (values above 350 μg m

−3

) occurred at Lamas d’Olo, a rural station in northern Portugal. The main objective of the work is to identify the origin of this ozone-rich episode. Data analysis together with air quality numerical simulations were applied. The results indicate that long-range transport of ozone and its precursors instead of local chemical production is responsible for the high ozone concentrations. The sea-breeze circulation turns out to be the driving force for the transport of pollutants from the coastal urban and industrialized areas.

Ammonia is an important contributor to particulate matter in the atmosphere and can significantly impact terrestrial and aquatic ecosystems. Surface exchange between the atmosphere and biosphere is a key part of the ammonia cycle. Agriculture, in particular, is a large source of ammonia emitted to the atmosphere, mostly from animal operations and fertilized crops, while dry and wet deposition are the primary sinks of atmospheric ammonia. Although, current air quality models consider all of these source and sink processes, algorithms for emissions from fertilized crops and dry deposition are too simplistic to provide accurate accounting of the net surface fluxes. New modeling techniques are being developed that replace current ammonia emission from fertilized crops and ammonia dry deposition with a bi-directional surface flux model. Comparisons of the ammonia bi-direction flux algorithm to field experiments involving both lightly fertilized soybeans and heavily fertilized corn are presented and discussed. Initial tests and evaluation of CMAQ modeling results for a full year (2002) at 12 km grid resolution including implementation of a soil nitrification model and the ammonia bi-directional flux algorithm result in improved NH

x

wet deposition.

The adaptation of the Community Multiscale Air Quality (CMAQ) modeling system to simulate O

3

, particulate matter, and related precursor distributions over the northern hemisphere is presented. Hemispheric simulations with CMAQ and the Weather Research and Forecasting (WRF) model are performed for the year 2006 using identical projections and grid configurations. The ability of the model to represent long-range transport of air pollutants is analyzed for selected cases through comparison with available surface, aloft and remotely sensed observations. These demonstrate the feasibility of extending the applicability of the CMAQ modeling system to hemispheric scales to provide a conceptual framework to examine interactions between atmospheric processes occurring at various spatial and temporal scales in a consistent manner.

The building effect parameterization (BEP) module [3] was implemented in a high resolution version of the COSMO model (DWD) in order to take into account the urban impact on the airflow and the radiation budget. Based on urban structure data of Dresden, relevant input parameters of the BEP module were developed. By means of this model setup it is possible to investigate the interactions between the city structure and the meteorological variables with different types of artificial cities, ranging from densely built-up areas to suburban areas in order to illuminate the impact of the city type on the dynamical and thermal properties of the atmosphere.

The National Air Quality Forecast Capability (NAQFC) is based on the EPA Community Multiscale Air Quality (CMAQ) model driven by meteorological data from the NOAA North American Mesoscale (NAM) Non-hydrostatic Meso-scale Model (NMM). Currently, NMM meteorological data on Arakawa E-grid are interpolated on a CMAQ’s Arakawa C-grid using the processors PRODGEN and PREMAQ to handle map-projection transform, vertical layer collapsing, and other emission and meteorological data feed issues. The FY11 pre-implementation version of NAM has undergone significant changes in the vertical layering, horizontal grid projection and improved science components for its FY11 upcoming major upgrade release. This provides an opportunity to improve the coupling methodology between NMM and CMAQ that reduces uncertainties both in the meteorological and emission inputs for the off-line air quality modeling and helps development of on-line NMM-CMAQ version. Three major tasks are needed to achieve a tighter coupling between them: (1) Adapt to NAM’s vertical hybrid pressure and grid structure; (2) Change CMAQ to use the same rotated latitude longitude B staggered horizontal grid structure as NAM, (3) Modify emission model to provide generic inputs for the B staggered grid and hybrid vertical structure of NAM. The first task achieves consistent matching of dynamics between the two systems, despite the possible necessity of layer-collapsing to fit within operational time-lines. The second task removes unnecessary interpolation of meteorology data for air quality simulations. The third task involves modification of the U.S. EPA Sparse Matrix Object Kernel Emission (SMOKE) model to handle the staggered B grid. At this time only the first of these three steps has been accomplished, and the test result from this test focusing on the selected test period has been compared to that produced by the operational NAQFC. Further work with all these three modifications concurrently in place is underway.

The Dutch National Precipitation Chemistry Monitoring Network measures the wet deposition fluxes of acidifying and eutrophying compounds and heavy metals over the Netherlands since 1978. Recent measurements of sulfate, ammonium and nitrate of this precipitation network are used to validate the outcome of the new wet deposition module implemented into the regional air quality model LOTOS-EUROS. The old wet deposition module only included below-cloud scavenging, whereas the new wet deposition module takes into account below-cloud

and

in-cloud scavenging. Both processes are implemented such that the effect of the saturation of rain droplets on the scavenging efficiency is included. A simple parameterization, based on the liquid water content data from ECMWF, is used to distinguish between regions in which respectively below-cloud or in-cloud scavenging occurs. In general, it is found that the combination of below-cloud and in-cloud scavenging increases the net scavenging process in the LOTOS-EUROS model simulations; hence the wet deposition fluxes are increased. These increased wet deposition fluxes in the model simulations are evaluated at the sites of the precipitation network in the Netherlands and show a better agreement as compared to results of simulations performed with the old wet deposition module. Finally the wet deposition fluxes are also evaluated at the sites of the European EMEP monitoring stations: a similar improvement is found again between the measurements and model results.

A state-of-the-art regional European emission data base is combined and cross-checked with bottom-up emission inventories of Paris, London, Rhine-Ruhr area (Germany) and the Po-valley (Italy). It is shown that the allocation of the emission in the regional top-down inventory can deviate substantially from the megacity bottom-up inventories. For example, the PM10 and NOx in local inventories are respectively 26% and 62% (London), 33% and 95% (Paris), 55% and 108% (Rhine-Ruhr) and 110% and 107% (Po valley) of the emission allocated to the same area in the regional inventory. The match for the Po Valley is reasonable but if we zoom in on a city level (Milan) similar problems as seen in Paris and London surface. We conclude that the European scale inventory is consistent with official reported national emissions but local-national-regional scale inventories are not consistent. Since the discrepancies are large, predicted concentrations and population exposure estimates may be significantly different. Our work shows the importance of regionalization of emissions for model input and argues that consistency between emission inventories at different scales deserves more attention.

Using a combination of models, high resolution air quality maps for Flanders (Belgium) have been made. First of all, the Eulerian air quality model AURORA has simulated for a complete year the air pollutant concentrations over the region on a 3 × 3 km² resolution. These results are calibrated using the RIO-interpolation model on air quality station data. Thereafter, an extra simulation using the bi-Gaussian IFDM model is made on a resolution of 1 × 1 km², with a finer resolution (up to 25 m) close to the major roads. The nesting methodology of IFDM in AURORA is designed to avoid double counting of the roads. The results are highly detailed PM

10

, PM

2.5

, NO

2

and EC maps for Flanders. Using station data and data from several measurement campaigns, the maps have been validated and it has been shown that the maps show indeed a highly detailed picture of the air quality in Flanders. These data will be used in assessing the air quality and human exposure to it in Flanders, and in assessing policy scenarios designed to improve the air quality.

Atmospheric ammonia (NH

3

) plays an important role in fine-mode aerosol formation. Accurate estimates of ammonia from both human and natural emissions can reduce uncertainties in air quality modeling. The majority of ammonia anthropogenic emissions come from the agricultural practices, such as animal operations and fertilizer applications. The current emission estimates at the U.S. Environmental Protection Agency (U.S. EPA) are based on the annual National Emission Inventory (NEI). However, accurate estimation of ammonia emissions in space and time has been a challenge. For instance, fertilizer applications vary in the date of application and amount by crop types and geographical area. With the support of the U.S. EPA, we have responded by an agricultural fertilizer modeling system for use with a newly developed ammonia bi-directional flux algorithm in the Community Multiscale Air Quality (CMAQ) model. This modeling system will simulate NH

3

emissions from fertilizer applications on agricultural lands rather than from emission estimates based on pre-defined emission factors. The goal for this paper is to demonstrate how this agricultural fertilizer modeling system is developed for a continental U.S. CMAQ 12-km modeling domain and the tools we developed in this system.

The European project MACC (Monitoring Atmospheric Composition and Climate) is developing a pre-operational system for forecasting the chemical composition of the atmosphere, both on a global scale and on a regional scale for Europe. The regional scale forecasts in MACC are based on an ensemble of seven regional air quality (RAQ) models, including the Dutch LOTOS-EUROS model. These RAQ models have been set up independently to deliver forecasts of trace gases on a daily basis, up to three days ahead. In the presentation I will introduce the MACC approach to air quality forecasting for Europe. The performance of the LOTOS-EUROS model will be compared with the other contributing models (EMEP, EURAD, MATCH, SILAM, MOCAGE, CHIMERE), and the evaluation procedure in MACC will be discussed. Within MACC a data assimilation capability is built for all seven models. We will show first results of the assimilation of both European surface observations as well as OMI NO

2

satellite observations.

Model simulations were performed using the CB05 chemical mechanism containing the base and updated toluene mechanisms for the eastern US. The updated toluene mechanism increased monthly mean 8-h ozone by 1.0–2.0ppbv in urban areas of Chicago, the northeast US, Detroit, Cleveland, and Cincinnati compared to those with the base toluene chemistry. The updated chemistry reduced mean bias and root mean square error in ozone predictions when compared with observations greater than 85ppbv and increased mean secondary organic aerosol from toluene by a maximum of 4%.

Because of rapidly growing emission sources of photochemical smog precursors in continental countries in East Asia, people are concerned about possible strong contribution of long range transport to episodic high ozone in Japan. This study evaluates relative importance of the long range transport and local production to the episodic high surface ozone in Central Japan by using high resolution chemical transport calculation with 1 km horizontal grid.

The European Chemical Weather Forecasting Portal (ECWFP) has been developed within the COST (European Cooperation in Science and Technology) ES0602 action, “Towards a European Network on Chemical Weather Forecasting and Information Systems”. The portal provides access to the predictions of a substantial number of chemical weather forecasting systems and may be used to find out which services are available for specific (1) areas, (2) time periods and (3) pollutants. The portal serves as a “one stop shop” of chemical weather modeling services and associated information, and is currently expanding its functionalities to allow for a harmonized presentation and inter-comparison of the various available forecasts, as well as for the computation of model ensemble predictions.

Accurate assessment of air quality on a regional scale and understanding the contribution of various sources is critical for developing mitigation strategies to improve air quality and protect human health. This study is a first attempt to use air quality modeling in characterizing the impact of anthropogenic emissions from urban sources in Russia on air pollution in Northern Europe. We used CMAQ to simulate concentrations of key air pollutants (CO, NO2, O3 and PM) in winter-spring of 2006. We used WRF to simulate meteorology, and SMOKE to prepare emission inputs using the official emission inventory for Russian Federation. CMAQ results were compared with the NO2 ground-based measurements of NO2, CO and O3 at Saint-Petersburg and its suburbs. We found that Saint-Petersburg is the most significant factor impacting air quality in the North-Western region of Russia. Specifically, concentrations of NO2 and CO within the lower layer in the troposphere may exceed background concentrations by 20 and 6 times (correspondingly) for St. Petersburg and its suburbs. Model results suggest that anthropogenic emissions from Saint-Petersburg can impact levels of pollutant concentrations more than 300 km away from megacity (e.g. in border countries Finland and Estonia).

The aim of this contribution is to study the aerosol load in Buenos Aires in order to help characterize the contribution of distant sources and the role of the flow patterns in Southeastern South America (SESA) as dispersion mechanisms. The regional pollution in SESA is mainly due to biomass burning and the related smoke plume could be frequently observed by remote sensors from space. Observations and modeling results are combined to assess the relative importance of the biomass burning contribution to the aerosol pollution in the megacity. Buenos Aires aerosol load and derived quantities have episodic contributions from biomass burning. An increase in aerosol optical depth and Ångström coefficient and an impact in the particle size distribution are observed. A dispersion model coupled on-line with a regional atmospheric model is able to reproduce the plume pattern and evolution and the mean aerosol load.

The Canadian Model for Environmental Transport of Organochlorine Pesticides (CanMETOP) is used to simulate concentration and deposition of Endosulphan for a period of 2years over North American region. The preliminary model results are compared with the surface based monitored air measurements around the Great Lakes. The interannual variation and seasonal variation of endosulphan concentration are reasonably simulated and possible explanation is offered.

The present paper demonstrates some preliminary results of the set up and testing of a system for modelling of toxic air pollution due to possible accidents in industrial sites. Risk assessment for the region of “Vereya Him” factory, Yambol, Bulgaria is performed. The modelling tools used for this study are: WRF, CMAQ and SMOKE. The CB05 toxic chemical mechanism, including chlorine reactions, is used. The shown numerical results demonstrate the practical value of the preparedness mode of the planned emergency response system.

An evaluation of mesoscale meteorological simulations using MM5 and WRF in respect to air pollution conditions during a winter episode in January/February 2004 in a steep Alpine valley, the Inn Valley, is carried out. Static input fields, e.g. landuse, delivered with the model have been evaluated and compared to latest available data. Comparison of e.g. landuse data indicate that an update to the existing CORINE landuse data would be useful. Results of the MM5 simulations show that grid resolutions of at least 1km or higher are needed to realistically represent the meteorological conditions in steep and narrow valleys.

The strategies for industrial emissions control depend on, amongst other variables, the best available techniques, their economical feasibility, and their positive impact over air quality; i.e., advances in power generation technology are making older coal and fuel oil power plants to be replaced by natural gas combined cycles. In ITM 2007, the estimation of the impact of the application of best available techniques (BATs) in industrial emissions control in Galicia (NW Spain) was presented [Rodríguez R et al (2008) Modelling the impact of best available techniques for industrial emissions control in air quality: setting up inventories and establishing projections. In: Air pollution modeling and its application XIX. Springer, Dordrecht, pp 677–678]. In this work, the effect of these changes over the tropospheric ozone levels in this region is simulated using the CAMx model, for three different ozone episodes previously characterised.

The FRAME model was used to assess the changes in contribution of domestic emissions and transboundary transport to national deposition budget of oxidised sulphur and nitrogen in Poland during the period 2005–2020. It was found that the role of the transboundary transport will increase in 2020, especially over E and NE areas of Poland, due to relatively small decrease in emissions from eastern European countries, which are not obliged to meet EU emission standards.

Air quality in urban areas is often dominated by vehicle emissions but pollutants such as volcanic ash, dust from bush fires and particles from dust storms also contribute. This paper presents a semi-empirical methodology for quantifying the relative contribution of these sources. The technique is based on the underlying premise that vehicles follow a diurnal pattern of emissions (whereas the other sources do not) but that both are modulated by the surface wind flows. The methodology is evaluated using rainfall washout as a removal mechanism (rather than an additional source). Whereas gaseous pollutants are relatively insoluble in water, particulate matter is readily removed. The extent of the washout is observed as the difference between the model predicted and observed concentrations. The same methodology can be applied to quantify the relative contribution of pollution sources such as ash.

The assessment of the anthropogenic and natural contributions to PM10 concentrations is a key issue for the development of air quality policies in Europe. In areas such as the Mediterranean basin, a consistent fraction of the natural contribution to PM10 concentration is given by dust particles transported from Sahara. This study presents an estimate of the dust contribution to PM10 concentrations in years 2003–2005 at six Italian locations. The reduction (%) in the number of daily exceedances of the PM10 limit value (50μgm

−3

) after subtraction of the African dust contribution is also presented.The reduction varies with station between 20% and 50% in 2005 and between 5% and 25% in 2003 and 2004.

Preliminary results of two annul simulations of the air quality in a region of the Northern Italy are presented. The simulations, which refer to two different year, were performed using a modeling system based on a meteorological model, an interface code and a transport-chemistry model. Comparisons with measured data are presented and discussed.

Sea-salt aerosol (SSA) could influence the Earth’s weather and climate acting as cloud condensation nuclei. In spite of the importance of SSA effects on the Earth’s climate and weather, there were no measurements of sea-salt aerosols in the open sea. At Tel-Aviv University, the DREAM-Salt prediction system has been producing daily forecasts of 3-D distribution of sea-salt aerosol concentrations over the Mediterranean model domain 20 W–45E, 15 N–50 N (

http://wind.tau.ac.il/salt-ina/salt.html

). In order to evaluate the model performance in the open sea, daily modeled sea-salt aerosol concentrations were compared directly with sea-salt ground-based measurements taken at the tiny island of Lampedusa, in the Central Mediterranean. In order to further test the robustness of the model, the model performance over the open sea was indirectly verified by comparing modeled SSA concentrations with wave height measurements collected by the ODAS Italia 1 buoy. Model-vs.-measurement comparisons show that the model is capable of producing realistic SSA concentrations and their day-to-day variations over the open sea, in accordance with observed wave height and wind speed.

The purpose of this work was to assess the influence of emission data on CHIMERE, by means of the study of episodic ozone levels in 2007 at Galicia (NW of Spain). As input data to the simulations both EMEP emissions inventory data and a detailed bottom-up inventory for the EMEP (51,14) grid cell were applied, in order to compare their impact in the air quality results.

The steps taken to expand the Updatable MOS (UMOS) methodology to air-quality forecasting (UMOS-AQ) will be presented. This methodology has shown great ability to improve direct model output with weather elements. Air-quality data (observational and numerical) are used to produce hourly spot concentration forecasts of ozone (O

3

), particulate matter 2.5 μm(PM

25

) and nitrogen dioxide (NO

2

), up to 48 h. Numerical data originating from one source, an air-quality model (GEM-MACH15), were used to produce a set of direct and calculated predictors. The UMOS-AQ system produces one equation per station, per predictand, per model run and per forecast hour. Verifications on an independent sample show encouraging results with significant improvement over the model’s direct forecast. The UMOS-AQ system effectively reduces the model’s bias and error while increasing its variance. Important conclusions can be derived from the order of the chosen predictors that can lead to a better understanding of site-specific conditions.

At the Royal Meteorological Institute of Belgium we use the chemical transport model (CTM) CHIMERE. In a first phase it has been coupled to meteo fields from ECMWF. The emission database used is EMEP. The simulation domain covers Western Europe with a spatial resolution of 50 km. RMI also runs the Limited Area NWP model ALADIN operationally four times a day at a spatial resolution of 7 km. The meteorological fields from ALADIN have been used to apply a one way nesting to the CTM model running at a spatial resolution of 7 km on a domain of about 680 × 680 km

2

covering Belgium and The Netherlands. The emission database used for this high resolution is the TNO/GEMS emission database. We present the first results of this coupling for ozone and PM10 for two different episodes. Also the influence of running ALADIN and its updated version ALARO with improved physical parameterizations will be investigated.

This study aims to improve the performance of the Flexible Air quality Regional Model FARM by including dust concentrations from a dust transport model (SKIRON) into Lateral Boundary Conditions (LBCs). A sensitivity test has been performed in order to assess the impact of SKIRON Saharan dust on PM

10

modelled concentrations. A dust episode (25th–30th July 2005) has been simulated running FARM on a 20 × 20 km

2

resolution domain over Italy. Preliminary results from three simulations: NDC (“no dust” case), DC (“dust” case) and DC1.3 (“dust multiplied by a factor 1.3” case) have been compared to PM

10

ground measurements from the Italian Air Quality Network.

In this presentation, the contribution of the European aviation on the air pollution of the Mediterranean region is discussed using the new modeling tool, the Integrated Community Limited Area Modeling System (ICLAMS). The modeling system uses the approach of “online coupling” of meteorological and chemical mechanisms which studies the processes that take place in the atmosphere in an integrated way and in the same spatial and temporal resolution. The model was tested for July 2005 for Europe and the Mediterranean Region. Two simulations have been performed, one with emissions from all anthropogenic activities and the second excluding the emissions from aviation. The area that is influenced by the emissions from aviation is very large, and the most affected region is the Eastern Mediterranean and several areas in North Africa. The typical summer circulation with the prevailing west – northwest wind field over West and Central Europe favors the transport of pollutants towards East, South East Europe and North Africa leading to perturbations in the atmospheric composition especially up to 8 km above surface. The alterations in the atmospheric concentrations of O

3

, are discussed.

We develop a physically motivated statistical model for regional ozone air pollution by separating the ground-level pollutant concentration field into three components, namely: transport, local production, and large-scale mean trend mostly dominated by emission rates. The model is novel in the field of environmental spatial statistics in that it is a combined deterministic-statistical model, which gives novel perspectives on the modeling of air pollution. The model is presented in a Bayesian hierarchical formalism, and explicitly accounts for advection of pollutants, using the advection equation. We apply the model to a specific case of regional ozone pollution – the Lower Fraser Valley of British Columbia, Canada. As a predictive tool, we demonstrate that the model outperforms existing, simpler statistical modelling approaches. Our study highlights the importance of simultaneously considering different aspects of air pollution as well as taking into account the physical bases that govern the processes of interest.

Performance of an air quality prediction may suffer with problems due to uncertainties in meteorological forecasts, lack of event-based real-life emissions variations such as forest fires, and inaccurate dynamic boundary conditions representing effects of long-range transport events. The main objective of this article is to demonstrate the combined synergistic effect of incremental reduction of uncertainties in the model inputs for improving air quality predictions. Effects of improved meteorological inputs, emissions, and initial and boundary conditions, through the use of in-situ measurements and satellite-retrieved information are studied for the continental US domains utilizing the Community Multiscale Air Quality (CMAQ) model. In this extended abstract, only the impacts of improved initial and boundary conditions are discussed.

Both at the national and international level there is a large interest for improved air quality forecasts of ozone, NO

2

and aerosols. Data assimilation of observations into chemistry transport models is a means to achieve these improved forecasts. For improvement of the ozone and NO

2

forecasts over Europe tropospheric NO

2

columns from the OMI instrument on board the NASA’s AURA satellite will be assimilated into the LOTOS-EUROS model. As a first step the tropospheric NO

2

columns from the model are compared with the values from the OMI satellite instrument. The results show a seasonal bias which is largest in summer and over Eastern and Southern Europe. A similar bias has been observed in an intercomparison study of 9 atmospheric chemistry models with OMI. In this paper we will discuss the results from the intercomparison between the LOTOS-EUROS model and the OMI tropospheric NO

2

column data. We will provide possible reasons for the detected bias and discuss suggestions for model improvements. Finally we will present some first results from the assimilation of OMI tropospheric NO

2

columns into the LOTOS-EUROS model.

Detailed, time-varying spatial fields of air contaminant concentrations are valuable to public health professionals seeking to identify relationships between human health and ambient air quality, and policy makers interested in assessing compliance with air quality regulations. In this paper PM

25

fields are created from a linear model that predicts PM

25

at unmonitored grid points from observed PM

25

concentrations, CMAQ model outputs, and satellite estimates of aerosol optical density. The dimensionality of the input data set is first reduced using projection onto latent structures. Parameters of the linear model are mapped to the CMAQ model domain, permitting estimation of PM

25

at unmonitored sites.

Methods that include a combination of weather forecasting and atmospheric chemistry simulations are here referred to as chemical weather forecasting (CWF). We have selected 18 operational CWF models on regional and continental scales in Europe for a more detailed analysis. We have collected the information in a structured form, and inter-compared and evaluated the mathematical structure of these models. This information makes it possible to evaluate the relative advantages and limitations of the various modeling systems, modeling approaches and sub-models. We have also surveyed the most prominent gaps of knowledge in this field, and suggested potential priorities for future research directions. There are substantial gaps of knowledge, especially in the following fields: emission inventories, the availability and exchange of observations, the evaluation of the boundary conditions for the CWF models, the integration of numerical weather prediction and atmospheric chemical transport models, the data assimilation of the various chemical species, the understanding of several chemical and physical processes, the construction of model ensembles, and the scientific evaluation of the CWF models, including their evaluation against data.

Within the European CITEAIR II project, deterministic and statistical forecasting approaches are combined to forecast an air quality index in European cities. The idea is to benefit both from chemistry-transport modelling, which provides large scale information about the evolution of photo-oxidizing pollution, and from local measurements of pollution levels. In addition to usual predictors, numerical outputs from the PREV’AIR system are used in multiple linear regression to forecast O

3

, NO

2

and PM

10

concentrations at urban stations and derive the city-level forecast index. The methodology was applied to Rotterdam, Sevilla and Prague with 2008 data set and validated against 2009 one. It can be extended to any European city involved in CITEAIR II, in particular to those having not implemented their own forecasting system yet.

Atmospheric composition modelling requires the handling of a very wide range of scales from local, where anthropogenic emission are concentrated, to global where we are interested to study the impact of these sources. The need to allow different scales to interact stimulates a variety of approaches to bridge them. Here we describe a nudging-based method to realize such bridging. This technique permits models of different design and resolution to interact allowing, for instance, to force global models using regional or local models without a direct dynamic two-way interaction. In this study we describe the general approach, the experimental setup and the results of a numerical experiment performed looking at the Po Valley hot spot. For sake of simplicity this first experiment is performed running the same model BOLCHEM, a regional air quality model, at different spatial resolutions.

The US National Air Quality Forecast Capability (NAQFC), developed by the National Oceanic and Atmospheric Administration (NOAA) in partnership with the Environmental Protection Agency (EPA), currently provides next-day operational predictions for ground level ozone and smoke for 50 US states. Ozone predictions are produced with the Community Multiscale Air Quality (CMAQ) model driven by NOAA’s operational North American Mesoscale weather forecast Model (NAM); routine verification is conducted with monitoring data compiled by the EPA. Smoke prediction relies on satellite detections of smoke sources, US Forest Service emission estimates, with transport and dispersion simulated by the HYbrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model driven by NAM; routine verification is conducted with satellite observations of smoke. Quantitative predictions of fine particulate matter (PM2.5) are in development. Inventory based simulations using CMAQ with aerosol modules show seasonal biases: overestimating in wintertime and underestimating in summertime. Current testing focuses on including intermittent aerosol sources directly emitted by wildfires and dust storms within the forecast domain; longer-range transport of dust is incorporated through lateral boundary conditions. For example, simulations of trans-Atlantic transport of Saharan dust, injected into the prediction domain, contribute enhanced surface PM2.5 concentrations in the southern US, as observed in surface monitoring. Simulated PM2.5 concentrations are being evaluated with speciated observations in order to improve seasonal biases in predictions. Research on assimilating PM2.5 surface observations shows potential to improve predictions. Furthermore, analysis of discrepancies between observations and model predictions that are produced during assimilation can provide insight on impacts of proposed improvements to PM2.5 predictions.

The study presents the Fire Assimilation System (FAS) based on Level MODIS Collection 5 Active Fire Products. The FAS estimates the emission fluxes originated from the wild-land fires and provides this information to the atmospheric composition modelling system SILAM. Presently, the FAS incorporates the emission factors for three main land-use types (grass and agriculture, forest and mixed), which values were verified for several European episodes. Current work is a part of the FAS calibration for totally different geographical regions, Portugal and Australia, which nonetheless have several common environmental features.

The height of the Planetary Boundary Layer (PBL) is an important quantity for certain applications such as dispersion modeling. A dedicated two-dimensional PBL height analysis has been developed as an additional component of NCEP’s Real-Time Mesoscale Analysis. As for other meteorological analysis applications, the quality of the output is dependent on the quality of the input, including the observation. Here we assess the quality and potential for use in the PBL height analysis of a series of candidate observations, including Radiosonde Observations (RAOBS), Aircraft Communications Addressing and Reporting System (ACARS), Cooperative Agency Profilers (CAP), COSMIC satellite Radio Occultation and NWS Next-Rad radar reflectivities. The quality is assessed both by physical plausibility of the measurements and by comparison of the observations and the resulting analysis with independent observations not used in the analysis.

The MM5-CHIMERE regional modelling system is implemented over Italy and validated against available ground-based observations of atmospheric composition in summer. MM5 meteorological model is run on two nested domain covering Europe and Italy at respectively 36 and 12 km horizontal resolution. CHIMERE chemistry transport model is used to simulate gas and aerosol composition on the two domains at 0.5° and 0.15° horizontal resolution. Anthropogenic hourly emissions of primary pollutants over Italy are derived from the CTN-ACE inventory developed by national environmental agencies. Biogenic volatile organic compound emissions are calculated with MEGAN model driven by MM5 radiation and temperature fields. Ozone (O

3

) daily maxima are simulated with good correlation (0.76) and overestimated by 10%. Model overestimation decreases with increasing observed O

3

, while correlation increases. PM

10

is underestimated by 30–40% and reproduced with a correlation >0.5. Over Italy, the model capture enhanced level of O

3

, but not that of PM. Model skills in the 3-day ahead forecast do not degrade rapidly. Graphical output of forecast is operationally available on the ForeChem web site:

http://pumpkin.aquila.infn.it/forechem/

.

The paper discusses the operational experience of European-scale AQ hind- and fore-casting with the SILAM dispersion model and compares the performance of the two setups. Two parallel lines of daily AQ assessment in Europe have been established: the 72 h long stand-alone forecast and a previous-day 24 h hindcast followed by a 72 h forecast. In the hindcasting mode, 3D-VAR based data assimilation is used to refine the initial fields for O

3

, NO

2

and SO

2

. The standalone forecast, on the other hand, is initialized from the previous forecast without data assimilation. According to preliminary statistics, the best score improvement in the forecast is obtained for O

3

, while only minor or no improvement is seen for SO

2

and NO

2

. In addition, the data assimilation of gas-phase species has a visible effect on predictions of secondary inorganic aerosols.

Volatile organic compounds (VOCs) are key species for the production of tropospheric ozone, and biogenic VOCs (BVOCs) emissions are significantly important in rural areas because of their amount and the photochemical ozone creation potential (POCP) of biogenic species.Galicia is an Atlantic coastal region located in northwestern Spain, with forestry as the most important land use, including both deciduous (oak, chestnut, birch) and evergreen (pine, eucalyptus) species. In this region, because of its changeable weather conditions and its local Atlantic forest species, BVOCs emissions are highly variable, depending on meteorological (temperature, PAR and moisture) and seasonal parameters (leaf area index and daylight hours).This work is focused on the sensitivity analysis of the rural tropospheric ozone levels due to influence of BVOCs emissions from the European Atlantic forest.

Model-based air quality forecasts are affected by uncertainties in both meteorology and emissions. In this study, we analyze the impacts of these uncertainties on ozone (O

3

) forecasts over the Northeastern U.S. during a 45 day period in the summer of 2008. The meteorological uncertainty is prescribed through the use of 12 different weather forecasts from a short-range ensemble forecasting system differing in their initial conditions and physics options. To quantify the effects of emission uncertainties, we computed O

3

sensitivities towards variations in anthropogenic NO

x

and VOC emissions using the Direct Decoupled Method (DDM). Results show that the perturbations in meteorology tend to have a larger impact on O

3

forecasts than the emission perturbations considered in this study.

We investigate the possibility of using different metrics for the evaluation of multi-model ensembles, in the attempt to find the optimal representation of the ensemble spread and bias. We present basic properties of different metrics and we discuss the consequences of applying them in atmospheric dispersion multi-model ensemble systems. We show also how we can obtain relevant information equivalent to different statistical treatments of an ensemble by combining the application of various metrics for calculating the ensemble spread and bias. A digression is presented on the use of the optimal combination of model results within an ensemble Kalman filter application for data assimilation.

A model evaluation has been undertaken by considering a diagnostic evaluation of the distance dependent structure of a single source footprint.

Ensemble construction and treatment are considered in application to the atmospheric composition problem. One emergency and one air quality test case are discussed. Performances of several approaches to the ensemble treatment are compared: (a) static methods based on mean, median, upper and lower percentiles, (b) adaptive weighting based on dynamic solutions of minimization problems for selected statistical scores, (c) heuristic methods using the model scores as the weighting parameters, (d) spatially homogeneous and inhomogeneous weighting coefficients. For the adaptive weighting of ensemble members, the predictability of the scaling coefficients for the future times and the possibility of their extrapolation outside the area where they are established via model-measurement comparison are considered.

A 5year (2002–2006) simulation of CMAQ covering the eastern United States is evaluated using principle component analysis in order to identify and characterize statistically significant patterns of model bias. Such analysis is useful in that in can identify areas of poor model performance across space and time; facilitate understanding of the probable mechanisms (emissions, meteorological, and/or chemical) responsible for said poor performance; and designate specific locations that can be used in providing in depth diagnostic analysis.

The Community Multiscale Air Quality (CMAQ) model was evaluated for its ability to reproduce observed changes in ambient concentrations of ozone (O

3

) for two seasons: the summer of 2002 and the summer of 2005 covering the eastern United States. These two summer periods were distinguished by large emissions reductions stemming from controls mandated by the NO

x

State Implementation Plan (SIP) after 2002. CMAQ was evaluated for the robustness of its response in ambient O

3

levels to changes in NO

x

emissions. Furthermore, uncertainties in the NO

x

emissions inventory were propagated through the model using a direct sensitivity approach. Considering a 50% uncertainty in the area and mobile source NO

x

emissions, the model was able to replicate changes in O

3

concentrations between 2002 and 2005 at most observation sites.

The aim of this study is to evaluate the results from using parallel observation and numerical weather prediction data as inputs to the urban dispersion model SIRANE applied to traffic scenarios of NO

X

in the city of Florence. A detailed sensitivity analysis of the air quality modelling system against measurements from monitoring stations of the city of Florence has been carried out in order to identify the advantage and disadvantages of the use of meteorological observation and numerical weather prediction data.

The Chemistry Transport Model REM-Calgrid (RCG) has been improved by implementing a more detailed description of aqueous-phase chemistry and wet deposition processes including droplet pH. A sensitivity study on cloud and rain droplet pH has been performed to investigate its impact on model sulphate production and gas wet scavenging. Air concentrations and wet deposition fluxes of model runs applying differing droplet pH have been analysed and compared to observations. It was found that droplet pH variation within atmospheric ranges affects modelled air concentrations and wet deposition fluxes significantly.

The chemistry transport models REM_Calgrid and LOTOS-EUROS have been used to simulate annual total deposition fluxes on German ecosystems for the whole year 2005. To evaluate these results only indirect measurements are available, i.e. wind turbulence, air pollution concentrations and deduced deposition fluxes over specific ecosystems. The comparison of friction velocities processed from COSMO_EU and from ECMWF has shown good agreement with measurements with correlation coefficients around 0.8. Nitrogen deposition fluxes calculated with the two models for the year 2005 were comparable giving the same spatial distribution and summed over the whole area similar amounts. Comparisons with measured values at a forest site in Augustendorf for the year 2003 have shown substantial differences between the two models, both underestimating the measured nitrogen deposition flux.

The transport and transformation of PM is mainly forced by meteorological processes. Therefore, an appropriate description of these processes is of essential interest in chemistry transport modelling. In the paper, the influence of two different meteorological drivers on the simulated particle concentrations is analyzed. For this purpose, the chemistry transport code MUSCAT was online-coupled with WRF as well as with the COSMO model of the German Weather Service. Furthermore, WRF-Chem simulations are also considered in the model comparison. The combination of two meteorological and two chemistry-transport models has a great potential for a detailed analysis of the meteorological dependencies and its impact on the aerosol processes. The simulation results were compared with a comprehensive set of ground-based and profile measurements. The influence of several meteorological parameters (e.g., PBL height, humidity, precipitation) on the simulated concentration fields was analyzed. Main differences are caused by deviations in PBL and cloudiness.

Is an air quality model acceptable for use (fit for purpose)? We now have several state-of-the-art statistical methodologies in place for calculating model performance measures and for determining the statistical significance of the results. But a more difficult question is whether, based on the calculated performance measures, the model is acceptable or not. This question is being addressed in several of the authors’ recent studies, and an example carried out for the Joint Effects Model (JEM) is the focus of the current paper. The rationale for selecting acceptance criteria for air quality models is described, and the results of applications of the method to four urban field experiments are presented. The values of the acceptance criteria are based on the authors’ experiences with a wide variety of air quality models and observations. For example, an acceptable FAC2 is>0.3 for urban scenarios. Furthermore, since a model is unlikely to fulfill all acceptance criteria at every field experiment site, we require that the model meet the individual criteria for at least 50% of the performance measures and field experiments and input options used in the study. It is shown here that the JEM model meets the 50% criterion.

The Po valley in Northern Italy has been identified as a hot spot area in Europe where pollutant levels are expected to remain problematic in the years to come despite the application of the legislation devoted to air pollution control. High anthropogenic emissions in combination with frequently occurring stagnant atmospheric conditions cause very high PM (particulate matter) concentrations in winter. The POMI project (PO valley Model Inter-comparison exercise) has been set-up to evaluate the capabilities of current air quality models to reproduce the concentration levels reached in this area. One of the outcomes of the project has been the confirmation that air quality models do under-estimate quite significantly the observed PM concentrations in this region. A series of sensitivity test have been run in order to understand the reasons for this underestimation and in particular to highlight the impact of meteorology on modeled concentrations. Different nudging techniques of observations from regulatory networks into MM5 (meteorlogical model) are explored in an attempt to improve the simulation of frequent low wind regimes in the Po valley. The sensitivity of CHIMERE (chemical transport model) is evaluated for January 2005, a period with high PM concentrations. Strengths and weaknesses of the approaches on simulating PM in one of the most polluted and complex areas in Europe are discussed.

This study shows part of the results obtained during the operational evaluation of MINNI atmospheric modeling system over Italy. MINNI is the Italian Integrated Assessment Modelling System for supporting the International Negotiation Process on Air Pollution and assessing Air Quality Policies at national/local level sponsored by the Italian Ministry of the Environment. The evaluation was carried out for both meteorology and air quality for the years 1999 and 2005. Changes of meteorological variables and of ozone concentrations in relation to the change of horizontal grid resolution were also investigated. The results show the capability of the modelling system to reconstruct the meteorological and ozone fields over Italy.

An important new emphasis in meteorological models evaluation is the thorough discussion of vertical profiles of meteorological parameters. This discussion contributes to the understanding of why different mesometeorological models calculate quite different wind and temperature profiles and atmospheric boundary layer heights (even when using the same method) while all show good agreement between simulated and measured surface data. The wind, temperature and turbulence profiles influence significantly the transport and diffusion of pollutants in the air. The calculations of a number of crucial parameters in air quality models, such as deposition and biogenic emissions, also depend on meteorological parameters. The availability of 3D measured wind fields provided by wind profile radars and lidars give a new challenge for such studies. The 3D model to measurements comparisons should consider new performance statistics. This study presents few examples from the COST 728 intercomparison exercise for the winter of 2003 in Europe.

The CMAQ modeling system has been used to simulate the CONUS using 12-km by 12-km horizontal grid spacing for the entire year of 2006 as part of the Air Quality Model Evaluation International Initiative (AQMEII). The operational model performance for O

3

and PM

2.5

for the simulation was assessed. The model underestimates O

3

mixing ratios in the winter, which is likely due to low O

3

mixing ratios in the middle and lower troposphere from the lateral boundary conditions. PM

2.5

performance varies seasonally and geographically, with PM

2.5

overestimated in the winter and fall, while performance in the spring and summer is generally good, especially in the summer. PM

2.5

concentrations are systematically higher in the AQMEII CMAQ simulation than in previous CMAQ simulations, primarily due to higher concentrations of TC and unspeciated PM

2.5

mass, which may also be due to differences in the lateral boundary conditions.

The Environment Canada regional air quality modelling system, AURAMS, is used to simulate two summer periods in 2004 and 2007, coinciding with two air quality measurement campaigns over eastern North America. The two summers are quite distinct in weather and air quality conditions. The model results are compared with various surface based monitoring air and precipitation chemistry measurements to examine model’s capability in capturing the impact of meteorology on air quality and explore the roles of different processes affecting ozone and PM in the region.

Dynamic model evaluation assesses a modeling system’s ability to reproduce changes in air quality induced by changes in meteorology and/or emissions. In this paper, we illustrate various approaches to dynamic model evaluation utilizing 18years of air quality simulations performed with the regional-scale MM5/SMOKE/CMAQ modeling system over the Northeastern U.S. for the time period 1988–2005. A comparison of observed and simulated weekly cycles in elemental carbon (EC) and organic carbon (OC) concentrations shows significant differences, indicating potential problems with the magnitude and temporal allocation of traffic-related emissions and the split between primary and secondary organic aerosols. A comparison of the observed and simulated interrelationships between temperature and ozone over the 18-year simulation period reveals that the high end of the modeled ozone concentration distribution is less influenced by interannual variability in the high end of the temperature distribution as compared to the observations.

A near real-time assimilation and forecast system of aerosols has been developed by integration in the ECMWF IFS code within the GEMS project. The GEMS aerosol modeling system is novel as it is the first aerosol model fully coupled to a NWP model with data assimilation. Aerosol optical depth (AOD) data of the MODIS instrument on Terra and Aqua satellites was assimilated. The performance of the aerosol model was evaluated by the means of case studies. The assimilation of MODIS AOD improved the subsequent aerosol predictions when compared with observations, in particular concerning correlations and AOD peak values. The assimilation is less effective in correcting a positive or a negative bias.

Data assimilating and pooling the model predictions in the multi-model ensemble, described in this paper, are based on the techniques of approximation and regularization of multidimensional vectors in the linear Euclidean space with the use of the non-orthogonal vector basis. This approach has been successfully applied to (i) the time series of the annual temperatures averaged over the globe and northern hemisphere corresponding to the last 100years, and (ii) the fields of concentrations of atmospheric pollutants over Europe. Quantitative estimates of the efficiency of the proposed technique are presented in the paper.

Temporal and spatial distributed emissions are an essential input parameter for Chemical Transport Models (CTM). In order to obtain consistent emissions for long-term CTM runs the US EPA emission model SMOKE has been adapted and modified. The modified version of the SMOKE emission model (SMOKE-EU) uses official and publicly available data sets and statistics to create emissions of CO, NOx, SO2, NH3, PM2.5, PM10, NMVOC. Currently it supports several photochemical mechanisms and a PM2.5 split. Additionally emissions of benzo[a]pyrene have been modelled. The temporal resolution of the emissions is 1 h. The resolution of the surrogates used for spatial disaggregation is 1 × 1 km². Vertical distribution is done via plume rise calculations. The area covered by the emission model is Europe including eastern Russia, North Africa and Turkey the currently implemented datasets allow for the calculation of emissions between 1970 and 2020. Emissions for the year 2000 on a 54 × 54 km² domain were evaluated by comparison to datasets from three commonly used emission models. Additionally SMOKE/EU emissions were used as input for the CMAQ4.6 CTM and the calculated air concentrations of Ozone, NH4, NO3 and SO4 were compared to EMEP measurements. O3: (NMB 0.71) (SD 0.68) (F2 0.83) (CORR 0.55) 48 Stations (hourly). NH4: (NMB 0.25) (SD 1.01) (F2 0.55) (CORR 0.53) 8 Stations (daily).NO3: (NMB 0.42) (SD 0.60) (F2 0.40) (CORR 0.45) 7 Stations (daily). SO4: (NMB 0.34) (SD 0.84) (F2 0.65) (CORR 0.55) 21 Stations (daily). Abbreviations: Normalized Mean Bias (NMB), Standard Deviation (SD), Factor of 2 (F2), Correlation (CORR). Using three different emission datasets as input for CMAQ showed that the SMOKE-EU model produces results comparable to those of commonly used European emissions data sets.

The SILAM model (version 4.5.4) is applied for retrospective modelling of wintertime deposition fluxes of fly ash and sulphate near Estonian oil-shale-based industrial complex. The fluxes are validated against snow-based deposition measurements that have been carried out during nine winters since 1985. Also, the data from the Lahemaa (Estonia) and Virolahti (Finland) EMEP stations are applied for regional-scale validation. In this test case SILAM tends to systematically underestimate the deposition of fly ash and overestimate the deposition of sulphate at distances up to 30 km from the pollution sources. Overestimation of sulphate is severe during recent decade, whereas at more remote (100 km) EMEP stations the fluxes are nearly correct. As SILAM is basically developed and validated for meso-scale applications, some scale-dependent effects near the sources may appear. The recent changes in oil shale combustion technology may also play a role in increasing overestimation trend for sulphate.

Performance and efficiency of the general purpose air-pollution dispersion modeling system based on Lagrangian particle model Spray is evaluated on highly complex terrain over Šaleška region. The evaluation is made to determine if modeling system complies with the requirements of the Slovenian legislation for industrial air pollution control which requires efficient modeling systems for small domains over complex terrain. Evaluation has proved that the modeled concentrations show significant agreement in time and place with the observations. After this proof was available the model was used for regulatory planning of the new block stack.

The results of parcel model studies seem to indicate that increasing particulate pollution and decreasing solubility suppresses rain formation. In individual and short time cloud simulations this behaviour was confirmed in our 3D model studies. However, taking into account entire cloud fields over longer periods of time yields the strong spatial and temporal variability of the results with isolated regions of inverse correlation of the effects. Even though in general the expected behaviour was found, after several hours of simulation, the integrated precipitation of the more polluted cases caught up. This suggests that a changing pollution will affect the spatial and temporal pattern of precipitation, but will probably not reduce the overall long term precipitation amount which might be entirely governed by the moisture state of the atmosphere.

The new generation fully coupled “online” WRF/Chem model is implemented over Europe on a coarse grid and validated against ground-based observations of meteorological variables and atmospheric composition. Anthropogenic emissions are derived from the EMEP database. We studied its sensitivity to “Dudhia” and “Goddard” short waves radiation schemes. The model satisfactory reproduces the observed meteorological fields, particularly with the Goddard scheme. For chemical variable, they are not much difference between the two scheme. Ozone (O

3

) daily maximum is reproduced with a correlation of 0.81 and an root mean square error (RMSE) of 23 μg/m

3

. Nitrogen dioxide (NO

2

) is overestimated by a factor of 2, and reproduced with a correlation of 0.47. Particulate matter shows poor correlation with observations, however the model captures the magnitude of PM

2.5

concentrations (bias −10%, RMSE 7 μg/m

3

). Simulated PM

10

is affected by a too strong local dust source over dry areas.

This paper seeks to calculate marginal external costs from human health impacts of transport air pollution. We focus on PM2.5 and NO

x

emissions in Belgium. Our analysis builds upon the widely used impact pathway approach. The great novelty compared to previous work is that both future emissions and background concentrations are modelled now, taking into account different destinations of a tonne of emitted pollutant now versus in the future. Moreover, the model allows for demographic evolution. We find similar results from two separate exercises: one for Belgian and one for Flemish PM2.5 and NO

x

emissions.

The amount, size distribution and chemical properties of cloud condensation nuclei (CCN), depend on the type of the prevailing air mass and also on local production and transportation of natural and anthropogenic particles. The aerosol properties vary in both space and time and impose significant amount of uncertainty on atmospheric research and also on future climatic projections. Cloud microphysics are very complex in nature and most of the times it is difficult to separate and identify the links and feedbacks between air quality and atmospheric processes under real atmospheric conditions. This presentation focuses on the interactions between air quality, clouds and precipitation for the area of greater Mediterranean. Mixtures of Saharan desert dust with sea-salt or with particles of anthropogenic origin, mainly sulphates and nitrates, may lead to the formation of new aerosols with different physiochemical properties. The effectiveness of airborne particles to act as CCN is examined towards their chemical composition and size distribution. Several modeling experiments, in-situ and airborne observations are analyzed and first results on the role of natural and anthropogenic particles on cloud formation and precipitation are discussed.

In this study, the shortwave cloud forcing (SWCF) and longwave cloud forcing (LWCF) are estimated with the newly developed two-way coupled WRF-CMAQ over the eastern United States. Preliminary indirect aerosol forcing has been successfully implemented in WRF-CMAQ. The comparisons with the observed PM

2.5

at the AIRNow sites indicates that the models captured a majority of observed daily PM

2.5

within a factor of 2, but generally underestimated the observations in the high PM

2.5

concentration range. The domain means of CERES satellite observations, WRF-CMAQ/CAM and WRF-CMAQ/RRTMg for SWCF (LWCF) are −48.1 (31.9), −31.9 (22.6), −19.8 (15.5) watts m

−2

, respectively. This means that the WRF-CMAQ model generally underestimated the cloud field for the 12-km resolution simulations.

Crustal materials are mainly emitted by anthropogenic and windblown fugitive dust, but also may potentially include some fly ash and industrial process emissions which are chemically similar to crustal emissions. Source apportionment studies have shown that anthropogenic fugitive dust emissions contribute on the order of 5–20% of PM

2.5

(particulate matter with an aerodynamic diameter less than 2.5 μm) and 40–60% of PM

10

(particulate matter with an aerodynamic diameter less than 10 μm) in urban areas that either have been or potentially may be unable to attain the National Ambient Air Quality Standards (NAAQS) for PM

2.5

and/or PM

10

. On the other hand, air quality models suggest vastly higher contributions from current fugitive dust emission inventories, with contributions ranging from 50% to 80% for PM

2.5

and 70–90% for PM

10

. This paper uses an improved speciation of PM to include, in addition to the current species, eight trace metals as well as separate non-carbon organic matter to assess potential improvements to the emission estimates of anthropogenic fugitive dust (unpaved and paved road dust, dust from highway, commercial and residential construction and agricultural tilling).

Elevated levels of PM

2.5

have been observed in North America in all seasons, underlining the need for year-round air-quality (AQ) forecasts. Wintertime AQ forecasting, however, poses unique challenges given well-known seasonal variations in emissions, meteorology, chemistry, and removal processes. In the case of PM

2.5

, both systematic and episodic overpredictions have been noted in the wintertime for current AQ forecast models, including Environment Canada’s GEM-MACH15 AQ forecast model. GEM-MACH15 is the regional forecasting configuration of the multi-scale, in-line AQ model GEM-MACH. Performance evaluations of GEM-MACH15 predictions for several recent winter periods have pointed to several sources of forecast error, including the spatial and temporal allocation of primary PM

2.5

emissions and the treatment of vertical diffusion. This paper describes recent improvements to input emissions for the GEM-MACH15 modelling system and their resulting impacts on forecast performance.

The Po Valley in Northern Italy is frequently affected by high PM10 concentrations, where both natural and anthropogenic sources play a significant role. This work was aimed at giving a proper account of the contribution of Saharan dust to air quality in the Po Valley. This was carried out by improving the boundary conditions of the mesoscale 3D deterministic Transport Chemical Aerosol Model (TCAM) on large-scale transport of Saharan dust, daily predicted over the Mediterranean region. A case study of the integration of the TCAM and DREAM models was implemented over the Po Valley. The results show that the use of improved boundary conditions leads TCAM to better performances, both in terms of correlation and mean errors: TCAM was able to more accurately reproduce high PM10 concentrations observed in the Po Valley during particular dust events.

Atmospheric particulate matter, notably desert dust, can have important impacts on air quality and climate. In this work we present a new parameterization scheme of dust production in the atmospheric chemistry general circulation model EMAC. The new scheme extends previous dust emission parameterizations, making use of online vegetation, meteorological and soil particle size distribution fields basic to the simulation of dust emissions. It aims at a more detailed and accurate representation of the dust production worldwide and in specific targeted regions. Data from satellites and in situ measurements are used to evaluate the model performance showing the first preliminary results of the new implementation.

The Marine Aerosol Concentration Model is presented. It is a two-dimensional unsteady model describing the evolution of aerosol concentration in the marine coastal area. It is validated with the help of experimental data measured in the Mediterranean coastal zone. Then, it is used to quantify the effect of the thermal stability on aerosol concentration measured at 10 m height in the MABL.

Black carbon (BC) is a key component of fine particulate matter that is emitted from aircraft and other combustion sources at airports. Understanding the contribution of aviation activity to ambient BC is important given the projected growth in aviation transport and decrease in emissions from other anthropogenic sources, but there are limitations in the various approaches conventionally used for source attribution. In this study, we evaluate three contrasting approaches to assess the contribution of aviation activity to BC concentrations near a small regional airport in Rhode Island, USA. First, we apply a previously developed regression model utilizing continuous BC concentrations measured at five monitoring sites and predictors such as real-time flight activity (departures and arrivals) and meteorological data, including mixing height, wind speed and direction. Second, we used AERMOD to model all airport sources of BC. Third, we used the Community Multiscale Air Quality (CMAQ) model, a comprehensive chemistry-transport air quality model. We included emissions estimates of all airport activity using an enhanced representation within the lowest 10,000 ft. inclusive of the landing and takeoff (LTO) cycle (the lowest 3,000 ft.) in the modeling system, along with other emissions from all other background sources. Median contribution estimates from the regression model indicate that flight activity contribute to approximately 24–28% of the observed BC concentrations at these five locations. The CMAQ and AERMOD based results show a much smaller contribution, in the range of 2–5% of the observed BC concentrations. We present possible explanations for these discrepancies, using a detailed analysis at different temporal scales at all five locations, and consider possible refinements to each approach. Our findings help to highlight the strengths and weaknesses of various source attribution approaches in the context of aviation emissions and other settings where it is of interest to infer local contributions to ambient concentrations.

We used an off-line, regional, model of atmospheric transport and chemistry, CTM, to investigate current and future levels of near-surface ozone and deposition of oxidized nitrogen in Europe. To assess changes in concentrations and deposition due to climate change, the CTM was run with recent (2000) emissions using meteorology from a regional climate model simulating control (1990–2009) and future (2030–2049) climates. The sensitivity to the choice of global climate model was explored by using output from a range of global climate models as forcing to the regional climate model. The analysis shows that results are more robust for surface ozone than for deposition of oxidized nitrogen with increase in summer surface ozone simulated over southern and western Europe. The analysis presented clearly demonstrates the need of exploring a range of different climate scenarios before drawing conclusions on the possible implications of climate change for future regional air pollution.

For the purpose of qualifying and quantifying the climate forcing due to atmospheric chemistry/aerosols on regional scale, the regional climate model RegCM3 has been coupled with the chemistry/aerosol model CAMx. Experiments with the couple have been run for EC FP7 project MEGAPOLI assessing the impact of the megacities and industrialized areas on climate. New domain have been settled in 10 km resolution including all the European “megacities” regions, i.e. London metropolitan area, Paris region, industrialized Ruhr area, Po valley etc. TNO emissions are adopted to resolve urban areas. A sensitivity test of the resolution effect is presented to reveal whether the concept of effective emission indices could help to parameterize the urban plume effects in lower resolution models. The sensitivity test to switch on/off Paris area emissions is analysed as well.

Assuming steady emissions and background concentrations, we investigate how changes of synoptic meteorology alone in future climates affect ozone episodes in the Lower Fraser Valley, Canada. We perform synoptic typing of combined sea-level pressure and 500 hPa geopotential heights for June to September for 1961–2000 from NCEP reanalysis. Five clusters provide a qualitatively good representation of typical synoptic conditions and stratify exceedance days into one cluster with more than half of all exceedances. Independent cluster analyses for climate model output from CGCM3.1 T63 1961–2000 control runs and 2046–2065 SRES A1B scenario runs give clusters qualitatively similar to NCEP. When CGCM output is mapped to the NCEP clusters, the CGCM control run cluster frequencies are almost identical to NCEP frequencies, while CGCM 2046–2065 output shows only small frequency changes. However, CGCM predicts substantial increases in daily maximum temperatures in the Lower Fraser Valley across all five clusters. An analysis of exceedance probabilities suggests that the predicted temperature increase will more than double the number of exceedance days per year.

We tried to simulate local climatic elements (temperature, wind speed and wind direction) in a lake watershed by using the turbulence closure model. The study site is Lake Biwa watershed, which is the largest lake in Japan and the most important water resource for 12 million people in Kinki region including Osaka and Kyoto Prefectures. The lake watershed is surrounded by mountains and covered mainly by lake, forest, and paddy field. Then, we conducted scenario analysis of global warming by rising the surface temperature of the lake according to IPCC reports and estimated the impact of global warming on the regional climate adjacent to the great Lake Biwa. In addition, we selected some local areas, which are typical land use, in the watershed and conducted sensitivity analysis for global warming in the local areas. The effect of global warming on the lake watershed appears clearer in the daytime, when wind speed in the upper boundary is higher. In this case, the energy from the great lake may be rapidly distributed to the surrounding region.

The mission of most environmental organizations around the world is to safeguard human health and the environment. In the United States, air quality management practice relies on the ambient concentrations of pollutants of concern and the success of a regulatory strategy is assessed by determining whether pollutant emissions have been reduced and ambient concentration levels have decreased. Some countries (e.g. United Kingdom) assess the success of their regulatory strategy by determining whether exposure reduction have occured. Exposure is defined as the contact of a stressor with a receptor for a given duration of time. It is now well-recognized that ambient concentrations and human exposure levels are not the same. Hence, it is important to better understand exposure levels since exposure is the link between environmental pollution and human and ecosystem health. Exposure science deals with comprehensive understanding of the relationship among pollutant emissions, pollutant transport and fate, pollutant levels that people breathe in and stressor levels to sensitive ecosystems, and associates health effects. Given the increasing complexity of current and emerging environmental problems, exposure science must play a pivotal role in developing and implementing most meaningfull and cost-effective emission control policies. Theis paper provides exposure framework to identify and minimize exposures to harmful pollutants, thereby better protectinf human ecosystem health.

Recent estimates of the growth in demand for aviation indicate that passenger counts may double or even triple by the year 2025, with a corresponding projected increase in emissions from the aviation sector. This would contribute to approximately proportional increases in concentrations and health effects if other factors were unchanging, but background emissions from non-aviation anthropogenic sources are generally expected to decrease due to several emissions control measures that are likely to be in place, and population size and age distribution will change over time. In this study, we evaluated changes in air quality and health risk due to growth in aviation activities from the year 2005 to 2025, focusing on 99 major U.S. airports with aircraft activity data for landing and takeoff (LTO) based on Terminal Area Forecasts (TAFs) developed for a sample growth scenario under the Next Generation Air Transportation System (NextGen) for 2025. We performed four annual simulations using the MM5-SMOKE-CMAQ modeling system for the year 2005 and 2025, with and without aircraft emissions. We obtained non-aviation emissions for 2005 from EPA’s 2005 National Emissions Inventory (NEI), and projected these to the year 2025. In performing the health risk analyses, we applied the EPA’s Speciated Modeled Attainment Test (SMAT) to the CMAQ results, and compared the air quality and health risk results on a pre-SMAT and post-SMAT basis. Our findings illustrated that each of the time-varying components – background concentrations, emissions, and population patterns – contributes significantly to growth in projected health risks over time, with significant differences in trends by particle constituent and region of the country. The relative importance of various particle constituents also depended significantly on the SMAT process, although secondary sulfate and nitrate particles dominated health risk in all scenarios. These conclusions provide an indication of the factors influencing health risk over time and the resulting areas in which interventions may be most effective.

Atmospheric transport models may be applied to run historic emissions scenarios, which are important in order to assess the correlation between monitored change in ecosystem health and biodiversity and changes in nitrogen inputs. The ability of models to calculate the response of nitrogen deposition to future emissions scenarios is of importance for policy makers to assess the benefits of implementing controls on atmospheric nitrogen emissions. Here we apply a relatively simple Lagrangian model, the Fine Resolution Atmospheric Multi-pollutant Exchange (FRAME) model to estimate the spatial distribution of nitrogen deposition in the UK, past and estimated future changes and the exceedance of critical loads. A second important issue concerns the grid resolution of the model simulation. We consider the difference in environmental impact criteria obtained with model simulations at 5 km and 1 km.

The effectivity of abatement strategies concerning elevated PM 10 and PM 2.5 levels in Germany have been determined by establishing detailed emission data bases reflecting the strategies, and using CTM’s to calculate the resulting concentrations up to the year 2020. The concentrations have been weighted by population density, and the uncertainty of the calculated yearly averaged concentrations have been estimated to be +/− 30%.

Population-based epidemiologic studies of air pollution have traditionally relied upon imperfect surrogates of personal exposures, such as area-wide ambient air pollution levels based on readily available outdoor concentrations from central monitoring sites. This practice may introduce exposure misclassification in epidemiologic analyses for pollutants that are spatially heterogeneous, including those associated with traffic emissions (i.e., carbon monoxide, elemental carbon, nitrogen oxides, and particulate matter). To investigate the potential impact of misclassification within observed health risk estimates, U.S. EPA in collaboration with Emory University is developing and evaluating several tiers of exposure metrics for ambient traffic-related and regional pollutants that vary in their approaches for modeling pollutant spatial heterogeneity. The following tiers of exposure metrics are examined: (1) central site monitoring data, (2) local scale modeling (AERMOD), and (3) combined regional and local scale modeling. Each metric is applied in two extensive, ongoing epidemiologic studies conducted by Emory University to examine ambient air pollution and acute morbidity in Atlanta, GA. We hypothesize that using the more refined exposure estimates will provide greater power to detect epidemiologic associations of interest, particularly for heterogeneous, traffic-related pollutants. This research will be useful for improving exposure assessment in future air pollution epidemiology studies, by providing alternative methods as well as by providing a further understanding of to the situations that might require refined exposure metrics.

Long term exposure estimates over large areas can be made using a combination of air quality models and population density data. However, the grid resolution of such models is often limited to 25–50 km and there may be a significant level of unresolved variability within the grids that will impact on the exposure estimates. In this paper the sub-grid variability is assessed using air quality monitoring (AirBase) and population data, concentrating on the covariance of concentration and population, which is the defining term in estimating sub-grid population exposure. The error that occurs when calculating the urban background exposure is assessed. The assessment shows that the error made in the exposure calculation for all of Europe is small for typical CTM resolutions of 50 km. The error is largest for NO

2

, where the average European urban background exposure is underestimated by 16%. Particulate matter is also underestimated, but only by 6%. Conversely, estimates of ozone exposure (SOMO35) are overestimated by a factor of 15%.

The use of modeling to support environmental authorities to plan air quality control policies is now quite widespread in different parts of Europe. At the sub-national level the most common way to face the problem is through the scenario analysis approach, using Chemical Transport Models to assess the impact of emission reductions on pollution concentrations. In this paper a multi-objective approach to define air quality policies is proposed. The considered objective function is a vector including an Air Quality Index and an Internal Costs Index. In this work the Air Quality Index (AQI) is the yearly mean PM10 concentration over the study domain. It is estimated processing source-receptor models linking emission and concentrations over the domain cells, taking into account nonlinear phenomena and using Artificial Neural Networks (ANNs). The ExternE methodology is then applied to estimate health impacts and external costs of optimal control policies up to 2020. The data used to identify the source-receptor models has been provided by a set of 10 simulations computed through the TCAM (Transport Chemical Aerosol) model. The Internal Cost Index (CI) describes the costs to implement a particular emission reduction policy. This index is computed by means of ANNs linking emission reductions and their relative implementation costs, for different CORINAIR macro sectors. Simulations with IIASA’s GAINS model have been used to calibrate the emission-to-cost model. The solutions of the decision problem represent cost-effective policies at the sectoral level. The methodology is applied to Northern Italy, one of the most polluted areas in Europe, to select optimal control policies up to 2020.

This study investigates the impact of ship emissions on secondary aerosol formation in coastal areas of the North Sea under conditions of the year 2000. The regional chemistry transport model CMAQ is used with different emission data sets as input to study their impact on aerosol concentrations. It is shown that the largest effect occurs in summer, when sulphate, nitrate and ammonium concentrations may be increased by more than 50% in large areas of Denmark and Northern Germany. The use of sulphur reduced fuels results in a significant reduction of sulphate aerosols. However at the same time, nitrate aerosol concentrations increase.

Optimization algorithms for control of industrial emission parameters with environmental and economic criteria are considered. Combination of the vector relaxation with external penalty function methods is used for multi-objective optimization. The optimization algorithms are developed for air quality management in decision support information systems and to generate some necessary emission scenarios in integrated air pollution models too.

The Clean Air Act (CAA) requires that the United States (U.S.) Environmental Protection Agency (EPA) set National Ambient Air Quality Standards (NAAQS) for pollutants considered harmful to human health and the environment. Previous research has shown that high ambient ozone levels are harmful to human health (e.g., Bell ML, Dominici F, Samet JM, Epidemiology, 16(4):436–445, 2005; Ito K, De Leon SF, Lippmann M, Epidemiology, 16(4):446–457, 2005, [4]). While ozone is not directly emitted, the formation of ozone is driven by chemical interactions in the presence of sunlight involving nitrogen oxides (NO

x

) and volatile organic compounds. Prevailing weather conditions in the Northeastern U.S. transport the relatively long-lived NO

x

(NO and NO

2

) and the secondarily-formed ozone downwind, contributing to pollutant levels at locations much farther from the emission source regions. In this study, we investigate associations between polluted air parcels transported from the Ohio River Valley (ORV) in the Midwestern U.S. and respiratory-related hospital admissions in New York State (NYS). We also examine whether better characterization of exposure in an epidemiology model would improve the discernment of this health signal.

Land-use regression (LUR) models have emerged as a preferred methodology for estimating individual exposure to ambient air pollution in epidemiologic studies in absence of subject-specific measurements. Although there is a growing literature focused on LUR evaluation, further research is needed to identify strengths and limitations of LUR modeling and strategies for improvement. In particular, LUR models have several limitations and among these are the needs for comprehensive monitoring data from a large number of sites, and the inability to link sources of emissions with measured elevated concentrations. In contrast, air quality models are designed to provide this linkage and have a long history of use by regulatory agencies in developing pollution mitigation strategies. Thus, the linkage of LUR techniques with available air quality modeling tools may benefit evaluation and enhancement of LUR techniques. In this study, we evaluated the fitted LUR models in several different ways and examined the implications of alternate LUR development strategies on model performance for benzene, particulate matter (PM2.5), and nitrogen oxides (NOx).

In recent years, an attention to the gene flow problem of dispersal of artificially modified genes to the natural environment by airborne pollen is increasing rapidly. Especially for wind-pollinated crop, there is a possibility that pollen diffuses quite widely depending on meteorological conditions. In order to deal with such problems, it is necessary to develop the model that can estimate the pollen dispersal and the hybridization mating appropriately. In this paper, I present an aerobiological mechanistic model for assessing pollen dispersal and hybridization using hourly data. This model considers hourly change of the meteorological conditions and daily change of biological conditions. And it was constructed for estimating the spatial distribution of hybridization percentage in a recipient field. The effectiveness of the model was certified by the field experiments. The model was constructed in consideration of physical processes and biological processes. The algorithms presented here can be applied to estimate the total pollen deposition and hybridization mating for many kinds of plants.

An integrated methodology evaluating the impact of the urban air pollution on human health is presented. From traffic emission data, background pollution level and meteorological data the pollutant concentration distribution within the street network is calculated by the urban scale dispersion model SIRANE and the potential health impact on population is evaluated by the traditional toxicological approach.

The Air Quality Modeling and Application Section of Environment Canada (EC) is transitioning its policy modeling platform from base year 2002 to base year 2006. The motivation behind this transition is to take into account the latest technological and scientific information upon which sound advice can be given to policy management. The latest data available at the beginning of the transition process includes 2006 emission inventories, 2006 meteorology inputs, and latest tools such as the meteorological and chemical transport models, interpolators, etc. The development of such a modeling platform encompasses the meteorology generation and interpolation, the emissions inventory and processing tools, post-processing of the modeling outputs, preparation of inputs for health and environmental benefits valuation models as well as the performance verification. The new system also addresses some of the technical weaknesses of the previous platform such as portability for different users, domain nesting capabilities, flexibility in emissions scenarios, more robust post-processing tools and better system diagnostic tools (reporting, error traceability). These changes will facilitate easier exchange of scenario configurations, data and results, allowing for improved coordination and collaboration between EC modelers. This paper provides an overview of the new policy modeling platform. It first outlines the general model configuration follow by preliminary results of the 2006 annual base case evaluation.

flexRISK studies the geographical distribution of the risk due to severe accidents in nuclear facilities, especially nuclear power plants (NPP) in Europe. Starting with source terms and accident frequencies, the large-scale dispersion of radionuclides in the atmosphere is simulated for about 1,000 meteorological situations. Together with the subsequent calculation of resulting radiation doses the consequences of severe accidents can be estimated. In this contribution, a description of the flexRISK project is provided.

Numerous health studies have used measurements from a few central-site ambient monitors to characterize air pollution exposures. Relying on solely on central-site ambient monitors does not account for the spatial-heterogeneity of ambient air pollution patterns, the temporal variability in ambient concentrations, nor the influence of infiltration and indoor sources. Central-site monitoring becomes even more problematic for certain air pollutants that exhibit significant spatial-heterogeneity. Improving characterization of air pollution exposures involves novel approaches to estimating ambient concentrations, a better understanding of the personal-ambient relationship, and personal exposure modeling. Estimates of ambient concentrations and human exposure estimates have been enhanced by utilizing both measurements and modeling tools. Statistical interpolation techniques and passive monitoring methods can provide additional spatial resolution in ambient concentration estimates. In addition, spatio-temporal models, which integrate GIS data and other factors, such as meteorology, have also been developed to produce more resolved estimates of ambient concentrations. Hybrid modeling approaches, which integrate regional scale models with local scale dispersion models, provide new alternatives for characterizing ambient concentrations. Estimating actual personal exposures requires an understanding of factors that impact personal-ambient relationships. Publically available data on housing characteristics (e.g., age and size of home) and commuting patterns can be utilized to understand these personal-ambient relationships. In addition, personal exposure modelling approaches are being developed, such as the Stochastic Human Exposure and Dose Simulation (SHEDS) model, which provides estimates of population exposures, and the Exposure Model for Individuals (EMI), which provide individual specific estimates of exposure. Many of these exposure characterization approaches are currently being applied and evaluated in epidemiological investigations. This paper describes the novel exposure assessment approaches discussed above and will briefly present results from the application of these tools. In addition, the paper will discuss ongoing and future applications of these approaches to evaluate their use in health studies.

This study is attempting to identify and quantify trends over the past 20 years in measured air quality in the U.S. Rocky Mountain area, including the states of Colorado, New Mexico, South Dakota, Utah, and Wyoming. Long-term well-maintained air sampling sites in the area include those from the IMPROVE and CASTNET monitoring networks, which were intentionally located in remote areas. Analysis of the observations from the ten available sites in the region focuses on ozone, fine particulates, sulfates and nitrates. Statistical trend analysis methods have been applied to the CASTNET ozone sites, with the conclusion that there is no consistent trend found at a majority of the sites. At the Mesa Verde site, there is a significant upwards trend in ozone, as well as a downwards trend in sulfates, with the latter due to emission controls in that area. Preliminary analyses of monthly mean MDA8 ozone at collocated sites at Rocky Mountain NP have revealed biases in that are being investigated further.

Biofumigation is a sustainable strategy to manage soil-borne pathogens, nematodes, insects, and weeds instead of methyl Bromide in developing countries including Egypt. Initially it was defined as the pest suppressive action of decomposing Brassica tissues, but it was later expanded worldwide to include animal and plant residues. Most data on the efficacy of biofumigation are from in vitro studies using fungal pathogens. Biofumigation also attracted the interest of nematologists, and plant pathologists. Moreover, this is a new trend to find out a substitute for Methyl Bromide that was banned worldwide since 2005, and no one interested in that time to find new and environmentally safe compounds that can work efficiently like the Methyl bromide did. We face this ecological problem in Egypt. However, concerns about the negative impact of synthetic nematicides on the environment and on general public health led to a re-evaluation of these products. For example, high use of the soil fumigant methyl bromide and resulting contamination of ground, surface and drinking water in The Netherlands led to a ban on its use in the 1980s. Later, methyl bromide was listed as an ozone depleting compound at the 4th meeting of the Montreal Protocol in Copenhagen, 1992, and in accordance with the US Clean Air Act its use as a fumigant is now banned in several nations. In last two decades, there is a worldwide awareness and interest to Air Pollution problems. In 1992 methyl bromide (MB) was added as controlled substance to the Montreal Protocol. The problem in developing countries including Egypt is more difficult to solve, as there is still a weak and not rigid environmental laws to respect and there is no multiple alternatives to follow by our Egyptian farmers. On the other hand, the developed countries must phase out most uses by 1 January 2005, with the possibility of critical use exemptions where alternatives have yet to be implemented. All the signed parties could not resolve differences in opinion on the amount and time period for these critical use exemptions which led to the scheduling of the first Extraordinary Meeting of the Parties (ExMOP1) in March 2004. They accord together to the amount of MB for most of the CUEs for 2005, the level of production necessary to satisfy the CUEs (taking stocks into account), and conditions for granting and reporting on CUEs. One of the most relevant to replace this toxic compound is biofumigation. Biofumigation refers to the suppression of soil-borne pests and pathogens by biocidal compounds released by Brassicaceous green manure and rotation crops when glucosinolates (GSLs) in their tissues are hydrolysed. We studied the effect of different brassicaceous green manure on the biological environment in soil and in their ecosystem. We concluded that, biofumigation can easily replace the mis/excessive use of methyl bromide and with some subsidizes programs for the poor farmers in developing countries. We have to enhance and encourage poor farmers especially in developing countries to make greater reductions in MB consumption. We have to mention that, with this climate change problem, we have to focus on this modifications in developing countries which focusing in their economics on agriculture and hence using this methyl bromide with weak control measures and hence we have to force them to use this safe, environmentally-friendly measures like biofumigation. Moreover, we have to add this item as an extension to the Montreal Protocol in 1992, and COP10. Finally, I invite the European Community to modify this Regulation (EC) 2037/2000 on ‘Substances that Deplete the Ozone Layer’, by using this biofumigation as an alternative measure to control this fatal problem. Finally, I encourage the UNEP to find out and to finance an innovative program for the new alternatives for this highly toxic compound especially for those farmers and consumers in the developing countries.