Search

The number of cities, or parts of cities, where air quality has been computed using the PMSS 3D model now appears to be sufficient to allow assessment and understanding of performance. Two fields of application explain the growing number of sites: the first is the long-term air quality assessment required in urban areas for any building or road project. The geometric complexity found in such areas can justify the use of a 3D approach, as opposed to Gaussian ones. However, these studies have constraining rules that can make the modelling challenging: several scenarios are needed (current, future with project, future without project), the long-term impact implies a long physical time period to be computed, and the spatial extension of the domain can be large in order to cover the traffic impact zone of the project. The second type of application is dedicated to services and, essentially, to forecasting. As for impact assessments, the modelling can be challenging here because of the extension of the domain if the target area is a whole city. Forecast also adds the constraint of time, as results are requested early, and the constraint of robustness. The CPU amount needed to meet all these requirements is important. It is therefore crucial to optimize all possible parts of the modelling chain in order to limit cost and delay. The sites presented in the article have been modelled with PMSS for long periods. This allows feedback to be provided on different topics: (a) daily forecasts offer an opportunity to increase the robustness of the modelling chain, (b) quantitative validation at air quality measurement stations, (c) comparison of annual impact based on a whole year, and based on a sampling list of dates selected thanks to a classification process, (d) large calculation domains with widespread pollutant emissions offer a great opportunity to qualitatively check and improve model results on numerous geometrical configurations, (e) CPU time variations between different sites provide valuable information to select the best parametrizations, to predict the cost of the services, and to design the needed hardware for a new site.

Urban air quality in cities is strongly influenced by road traffic emissions. Micro-scale models have often been used to evaluate the pollutant concentrations at the scale of the order of meters for estimating citizen exposure. Nonetheless, retrieving emissions information with the required spatial and temporal details is still not an easy task. In this work, we use our modelling system PMSS (Parallel Micro Swift Spray) with an emission dataset based on Floating Car Data (FCD), containing hourly data for a large number of road links within a 1 × 1 km2 domain in the city of Rome for the month of May 2013. The procedures to obtain both the emission database and the PMSS simulations are hosted on CRESCO (Computational Centre for Research on Complex Systems)/ENEAGRID HPC facilities managed by ENEA. The possibility of using such detailed emissions, coupled with HPC performance, represents a desirable goal for microscale modeling that can allow such modeling systems to be employed in quasi-real time and nowcasting applications. We compute NOx concentrations obtained by: (i) emissions coming from prescribed hourly modulations of three types of roads, based on vehicle flux data in the FCD dataset, and (ii) emissions from the FCD dataset integrated into our modelling chain. The results of the simulations are then compared to concentrations measured at an urban traffic station.

In a complex environment such as an urban area, accurate prediction of the atmospheric dispersion of airborne harmful materials such as radioactive substances is necessary for establishing response actions and assessing risk or damage. Given the variety of urban atmospheric dispersion models available, evaluation and inter-comparison exercises are vital for assessing quantitatively and qualitatively their capabilities and differences. To that end, the European Commission/Directorate General Joint Research Centre with support from the European Commission/Directorate General-Migration and Home Affairs, and with the contribution of the U.S. Defense Threat Reduction Agency, launched the Urban Dispersion INternational Evaluation Exercise (UDINEE) project. Within UDINEE, nine atmospheric dispersion models are evaluated and intercompared. Sulphur hexafluoride concentrations from puffs released near the ground during the Joint Urban 2003 (JU2003) field experiment are used in UDINEE to evaluate atmospheric dispersion models. The JU2003 experiment is chosen because UDINEE aims at the better understanding of modelling capabilities for radiological dispersal devices in urban areas, and the neutrally-buoyant puff releases performed in the JU2003 experiment are the closest scenario to this purpose. The present study evaluates the capability of models at simulating the presence and concentration levels of the tracer at sampling locations. The fraction of predicted concentrations and time-integrated concentrations within a factor-of-two of observations are less than 0.36 and 0.4 respectively. The analysis reveals an improvement in the performance of models by using time-varying inflow conditions. Since the simulation of the dispersion of puff release is particularly challenging, the results of UDINEE could constitute a benchmark for future model developments.

The capabilities of nine atmospheric dispersion models in predicting near-field dispersion from puff releases in an urban environment are addressed under the Urban Dispersion INternational Evaluation Exercise (UDINEE) project. The model results are evaluated using tracer observations from the Joint Urban 2003 (JU2003) experiment where neutrally-buoyant puffs were released in the downtown area of Oklahoma City, USA. Sulphur hexafluoride concentration time series measured at ten sampling locations during four daytime and four night-time puff releases are used to evaluate how the models simulate the puff passage at the measurement locations. The neutrally-buoyant puff releases in the JU2003 experiment are the closest scenario to radiological dispersal device (RDD) releases in urban areas, and therefore, UDINEE is a first step towards improving the emergency response to an RDD explosion in the urban environment. We investigate for each puff and sampler the model capability of simulating the peak concentration; the peak and puff arrival times; and time duration, defined as the period over which concentrations exceed 10% of the peak concentration. This analysis points out differences on the performance of models: the fraction within a factor-of-two values ranges from 0.10 to 0.6 for peak concentration, from 0 to 1 for the peak and arrival times, and from 0 to 0.8 for the time duration. The results reveal that the characteristics of the release site largely influence the model simulation as it affects initial puff size and the initial downwind spread of the puff.

The role of atmospheric dispersion models is becoming increasingly relevant to assess air pollution urban population exposure for epidemiological studies. Estimating urban air quality is challenging, because of the intrinsic characteristics of cities atmospheric structure, such as high density of primary emissions and presence of local dispersion processes, that produce strong concentration gradients. Therefore, very high spatial resolution simulations may often be required to improve the accuracy of estimations. The objective of this study is developing a microscale hybrid modelling system (HMS) to carry out, in a reasonable computational time, long-term simulations providing hourly concentration fields at building-resolving scale in extended urban areas in order to calculate annual indicators to evaluate exposure. The proposed system couples two atmospheric dispersion models suited for different scales: a Eulerian chemical transport model, FARM (Flexible Air quality Regional Model), accounting for dispersion phenomena due to regional and local emission sources, and a Lagrangian particle micro-scale dispersion model, PMSS (Parallel Micro Swift Spray), used to compute concentrations induced by vehicular traffic inside the city. The HMS has been applied on 12 × 12 km2 domain in Rome with a horizontal resolution of 4 m for calculating NO2 and PM10 concentrations for all year 2015. This study has been carried out in the frame of project BEEP (Big data in Environmental and occupational Epidemiology), that is an Italian research project in epidemiological field. Results show that the combined use of the two models reproduces the spatial and temporal variability of the observed atmospheric pollutants with a good agreement. The statistical analysis performed on daily average concentrations proves that the HMS suits the standard acceptance criteria for urban dispersion model evaluation, with a FAC2 of 0.92 and 0.80 and a Fractional Bias of −0.03 and −0.2 for NO2 and PM10 respectively. Furthermore, the implementation of an innovative kernel method to calculate concentrations in PMSS has made possible to reduce the computational time by 80%, leading to an average computational time of 3 h per simulated day on an HPC (High Performance Computing) system with 180 cores.

A multi-scale modelling system was developed to provide hourly NOx concentration fields at a building-resolving scale in the urban area of Modena, a city in the middle of the Po Valley (Italy), one of the most polluted areas in Europe. The WRF-Chem model was applied over three nested domains and employed with the aim of reproducing local background concentrations, taking into account meteorological and chemical transformation at the regional scale with nested resolutions of 15 km, 3 km and 1 km. Conversely, the PMSS modelling system was applied to simulate 3D air pollutant dispersion, due to traffic emissions, with a very high-resolution (4 m) on a 6 km × 6 km domain covering the city of Modena. The methodology employed to account for anthropogenic emissions relies on two different strategies. Traffic emissions were based on a bottom-up approach using emission factors suggested by the European Environmental Agency with traffic fluxes estimated by the PTV VISUM model in the urban area of Modena, combined with direct traffic flow measurements performed between October 28 and November 8, 2016 which was used for the hourly vehicle modulation. Other anthropogenic emissions were taken from the TNO-MACC III inventory at the scales resolved by the WRF-Chem model. Simulations were performed for the same period whereby the traffic measurement campaign was carried out. 2 m temperature and 10 m wind speed were captured quite well by the WRF-Chem model with statistical metrics in line with similar case studies related to the Northern Italy. The NOx concentrations reproduced in the Po Valley area by WRF-Chem were on average simulated reasonably well with a general negative bias in almost all the examined rural background monitoring stations. Additionally, the deployment of an emission inventory at the original resolution (7 km) highlighted that increasing resolution from 3 km to 1 km does not generally improve the model performance. Nevertheless, simulated and observed NOx hourly concentrations in the urban area of Modena exhibit a large agreement in particular for urban traffic site where detailed traffic emission estimations proved to be very successful in reproducing the observed NOx trend. At urban background stations, despite a general underestimation of the observed concentrations, the combination of WRF-Chem with PMSS provided daily pattern in line with observations. The analysis of the modelled NOx daily cycle pointed out also that at both traffic and background urban stations the morning NOx peak concentration was on average underestimated. This could be explained with an overestimation of mixing phenomena between 07:30 a.m. and 10:00 a.m. by WRF-Chem which leads to a greater dispersion of NOx along the vertical and thus a morning underestimation. The statistical analysis showed finally that PMSS combined with WRF-Chem at both the resolutions (3 km and 1 km) and at both traffic and background sites fulfilled standard acceptance criteria for urban dispersion model evaluation, confirming that the proposed multi-modelling system can be employed as a tool to support environmental policies, epidemiological studies and urban mobility planning.

A comparison between a puff atmospheric dispersion model (hereafter: PuM) and a Lagrangian particle model (hereafter: LPM) was conducted for a real case of emissions from an industrial plant, in the context of a complex and coastal site. The PuM’s approach is well-known and widely adopted worldwide, thanks to the authoritative suggestions by the US-EPA for regulatory use as, according to the definitions included in its guidelines, an “alternative” to “preferred” models; LPMs are more advanced models and have gained reliability over the last two decades. Therefore, it is of interest to provide insights into the decision to adopt or recommend, in the field of atmospheric impact assessment, a more advanced, but more knowledge- and resource-intensive, modeling tool, rather than an established albeit less accurate one. An inter-comparison of the two approaches is proposed based on the use of various statistical and comparative parameters with the goal of studying their differences in reproducing maps of ground-level ambient concentration statistics for assessment purposes (annual means, hourly peaks). The models were tested under a year-long simulation. The dispersion from both a point and a volume source, belonging to an existing industrial plant, was analyzed separately. The inter-comparison was performed through the analysis of 2D ground concentration maps, scatterplots, and three classical indices from the 2D maps of annual concentration statistics. To correlate the differences among models with site characteristics, the statistics were analyzed not only globally, but also according to distance from the source, the elevation, and the land-use classification. The analysis shows that around-its-axis plume dispersion in LPM is lower than in PuM over all the land-use types except water surfaces, in agreement with the theoretical basis provided by the models. Because of its more advanced theoretical formulation, e.g., in the interaction of the plume with the complex terrain and the three-dimensional wind field, an LPM used as a comparison term allowed us to highlight the weaknesses of a more traditional approach, such as PuM, in reproducing effects such as plume up-sloping, deflection, channeling, confinement, and wind shear diffusion.

When modelling the dispersion of pollutants in the atmosphere, uncertainty in the simulation results that depend on the available data used as input conditions is a critical issue, particularly in the context of emergency response to the accidental release of harmful substances. In the framework of the UDINEE Project, a Lagrangian particle dispersion model is used to simulate puff emissions in four different test cases, using input wind velocity data from two different datasets, both representative of the flow at the time of the release. The effect of the choice of the input data on the final concentration distribution at ground level is discussed and compared with observations. A statistical analysis is applied to estimate the deviation between the results of the two runs. The bias between the two concentration fields, connected to the variability and the uncertainty in the input flow, is found to be of similar magnitude to the typical bias between model predictions and observations.

After the summer 2015 chlorine releases during the Jack Rabbit II field experiment at Dugway Proving Ground, a special sonic anemometer study was carried out in October 2015 and March 2016. The goal was to provide documented wind fields in similar conditions to the actual releases and to study building wakes without risk for the instruments. This study included 30 sonic anemometers that were placed around these "buildings" to best capture the recirculating regions. Building wakes are important for the dispersion of pollutants in cities and around industrial plants and therefore we have selected two time periods during March 2016, in order to present model to model and model to data comparisons. There are four models involved in the comparison ranging from a diagnostic wind model with wake parameterization and Lagrangian particle dispersion and a CFD model with a k-epsilon turbulence closure and solving the Eulerian dispersion equation. Detailed comparisons are carried out for the wind speed and for the turbulence level. Six virtual releases are added in the simulations to investigate how the wake resolving method can affect the dispersion. For the wind speed comparison, the models give fair to good comparison with the sonic measurements. However, for the turbulence comparison the results clearly show a need for improvement, even for the k-epsilon closure, which still provides the best results. For the concentrations, for which there is no corresponding measurements, we find that the simulations are much closer than for the turbulence comparison.

The evaluation of the spatial representativeness of air quality monitoring stations is of fundamental relevance when observed concentration levels are used in air quality assessment. Since no reference method is provided, there is a need to develop tools for its quantitative assessment. In this paper we test a recently developed methodology for spatial representativeness area assessment, based on the analysis of time series of model concentrations by means of a Concentration Similarity Frequency (CSF) function, on the Taranto-via Machiavelli industrial monitoring station, located in a mid-size city in Southern Italy. The complex territorial context, the peculiar anthropogenic emissions features, dominated by the contribution of the largest steelworks in Europe, and the critical situation of human health in the city make this application an interesting case study to assess the portability of the CSF approach, so far applied only to background stations, to industrial sites that experience high concentration variability. A comprehensive characterization of the main anthropogenic emissions of the area was carried out, with detailed treatment of dust emission by wind erosion from industrial mineral piles; a Lagrangian modelling approach was chosen to simulate PM10 dispersion patterns, to achieve a reliable and high resolution description of concentration variability around the site. The resulting representativeness area of the station is 0.067 km2, fulfilling EU prescriptions for industrial stations. The comprehensive evaluation results, through the comparison with the observed data, showed good performances pointing out the reliability of the estimated concentration fields around the site and consequently of the assessment of its representativeness area.

Unexpected releases of hazardous airborne materials may happen in various circumstances, more frequently in accidental situations, less usually due to malicious activities. In all cases, the risk for the population and the rescue team is real and must be evaluated in order to take appropriate protection measures. Numerical models are more and more used in emergency response tools for predicting the atmospheric dispersion of pollutants and estimating impacted areas. Still, flow and dispersion modelling in a built-up environment may be very challenging given the complex geometry and meteorological conditions. Here, we present and evaluate a modelling approach that is a compromise between accuracy and timeliness of the computations. It is based on a 3D Lagrangian Particle Dispersion model driven by 3D flow and turbulence input provided by either a diagnostic mass-consistent model or a model solving the Reynolds-Averaged Navier-Stokes equations. This modelling system has been validated on a panel of experimental test cases from the Cooperation in Science and Technology (COST) Action ES1006 in both idealized and realistic urban mock-ups, wind tunnel and field trials, continuous and puffs releases. The sensitivity of the model to the meteorological input data is assessed. The results are discussed comparing predicted concentrations to measurements. The performances of the modelling system are evaluated through a statistical analysis. They were found to depend on the type of test case, type of release and location of the source in the built-up domain. Results show that the model is compliant with the validation criteria, established in literature for the reference metrics, in the large majority of the test cases. The modelling system proved to be robust enough to be used in the context of emergency response, when fast but still reliable results are needed.

This paper evaluates the relative impact on air quality of harbour emissions, with respect to other emission sources located in the same area. The impact assessment study was conducted in the city of Taranto, Italy. This area was considered as representative of a typical Mediterranean harbour region, where shipping, industries and urban activities co-exist at a short distance, producing an ideal case to study the interaction among these different sources. Chemical and meteorological field campaigns were carried out to provide data to this study. An emission inventory has been developed taking into account industrial sources, traffic, domestic heating, fugitive and harbour emissions. A 3D Lagrangian particle dispersion model (SPRAY) has then been applied to the study area using reconstructed meteorological fields calculated by the diagnostic meteorological model MINERVE. 3D short term hourly concentrations have been computed for both all and specific sources. Industrial activities are found to be the main contributor to SO2. Industry and traffic emissions are mainly responsible for NOx simulated concentrations. CO concentrations are found to be mainly related to traffic emissions, while primary PM10 simulated concentrations tend to be linked to industrial and fugitive emissions. Contributions of harbour activities to the seasonal average concentrations of SO2 and NOx are predicted to be up to 5 and 30 μg m−3, respectively to be compared to a overall peak values of 60 μg m−3 for SO2 and 70 μg m−3 for NOx. At selected urban monitoring stations, SO2 and NOx average source contributions are predicted to be both of about 9% from harbour activities, while 87% and 41% respectively of total concentrations are predicted to be of industrial origin.

The capability of two 3D Lagrangian particle models (ARCO and SPRAY) in simulating the airborne dispersion in highly complex terrain is tested. The simulations are compared to the tracer experiment TRANSALP-I performed on the 19th of October 1989 in the Alpine region (Canton of Ticino, Switzerland). TRANSALP exercises were part of the EUROTRAC-TRACT Subproject. Tracer was released during one hour (11.00–12.00) near the ground at Iragna, a few kilometres upwind from the bifurcation of two valleys (Leventina and Blenio Valleys); about 40 samplers, located in the two valleys in the range of 40 km, measured 1/2 h average mean concentrations near the ground for five hours. The results of the simulations are presented and discussed. A statistical model evaluation aimed at quantifying the performance of the two models was carried out. The simulations reproduced with a reasonable degree of accuracy the general behaviour of the observed phenomena: the split of the plume between the two valleys, the space and time distribution of the concentration, the concentration maxima locations.

Recently, Thomson and Montgomery (1994, Atmospheric Environment 28 , 1981–1987) stated the correct method of treating the reflection of particle velocity at the boundaries in Lagrangian particle diffusion models for non-Gaussian turbulence. Unfortunately, this method does not have an analytical solution. Two different approximated analytical solutions are proposed and compared. It is concluded that both of them satisfy the well-mixed condition and do not appreciably depart from the correct solution. The one consuming less time is proposed.

Spray, a 3D Lagrangian particle model for the simulation of complex flow dispersion, is presented. Its performance is tested against the Environmental Protection Agency wind tunnel concentration distribution of passive tracer released from elevated point sources, located in the lee region of a two-dimensional schematic hill, in a neutrally stratified boundary layer. Based on the measured values of the first two moments of the turbulent flow velocity, the mean fields are computed over a regular grid using a mass-consistent model, whereas the turbulence structure is simply interpolated. From these fields, trajectories of tracer particles are computed using a linear formulation of the Langevin equation, with a correlated, skewed forcing. The self-consistence test (well-mixed condition), aimed at maintaining an initially well-mixed particle distribution uniform in time, has shown satisfactory results in the region under study. The computed concentration field turns out to be in good agreement with th...

A Lagrangian stochastic model (MicroSpray), able to simulate the airborne dispersion in complex terrain and in presence of obstacles, was modified to simulate the dispersion of dense gas clouds. This is accomplished by taking into account the following processes: negative buoyancy, gravity spreading and the particle's reflection at the bottom computational boundary. Elevated and ground level sources, continuous and instantaneous emissions, time varying sources, plumes with initial momentum (horizontal, vertical or oblique in any direction), plumes without initial momentum are considered. MicroSpray is part of the model system MSS, which also includes the diagnostic MicroSwift model for the reconstruction of the 3-D wind field in presence of obstacles and orography. To evaluate the MSS ability to simulate the dispersion of heavy gases, its simulation performances are compared in detail to two field experiments (Thorney Island and Kit Fox) and to a chlorine railway accident (Macdona). Then, a comprehensive analysis considering several experiments of the Modelers Data Archive is presented. The statistical analysis on the overall available data reveals that the performance of the new MicroSpray version for dense gas releases is generally reliable. For instance, the agreement between concentration predictions and observations is within a factor of two in the 72% up to 99% of the occurrences for the case studies considered. The values of other performance measures, such as correlation coefficient, geometric mean bias and geometric variance, mostly set in the ranges indicated as good-model performances in the specialized literature.

Two Lagrangian particle models, APOLLO and MILORD, were used to simulate the first ETEX experiment. The role played by wind field, mixing height h and horizontal diffusivity KH appeared to be the most important aspects to be studied. The sensitivity to the accuracy of the input advection field was studied through the application of APOLLO using different ECMWF data sets differing in space and time resolution and in being forecasted or analysed, corresponding to the real-time, emergency-like condition, and to the a posteriori benchmark simulation. The role of h and KH was investigated by running both APOLLO and MILORD with different parameterisations, and comparing the model results between them and with the available observations. The model evaluation was carried out through a set of statistical indexes computed on three hourly average concentrations paired in space and time and time-integrated concentrations. It was found that the quality of the input wind field plays a major role in predicting with sufficient accuracy the plume position and extension after the first 24 h from the beginning of the release. The best-model results are obtained with large values of KH (in the range of 2.5×104–4.5×104 m2 s-1), which confirms the need to enhance the horizontal diffusion, in order to include the advection fluctuations unresolved by large-scale meteorological fields. A fixed value of h in the range 1000–1500 m seems to be more efficient than space and time variable h computed with standard algorithms. A reasonable explanation for this result is given, based on the consideration that in the long range, particles diffuse also in the residual layer above the stable nocturnal boundary layer.