Validation of the lagrangian particle dispersion model FLEXPART against large-scale tracer experiment data

A comprehensive validation of FLEXPART, a recently developed Lagrangian particle dispersion model based on meteorological data from the European Centre for Medium-Range Weather Forecasts, is described in this paper. Measurement data from three large-scale tracer experiments, the Cross-Appalachian Tracer Experiment (CAPTEX), the Across North America Tracer Experiment (ANATEX) and the European Tracer Experiment (ETEX) are used for this purpose. The evaluation is based entirely on comparisons of model results and measurements paired in space and time. It is found that some of the statistical parameters often used for model validation are extremely sensitive to small measurement errors and should not be used in future studies. 40 cases of tracer dispersion are studied, allowing a validation of the model performance under a variety of different meteorological conditions. The model usually performs very well under undisturbed meteorological conditions, but it is less skilful in the presence of fronts. The two ETEX cases reveal the full range of the model’s skill, with the first one being among the best cases studied, and the second one being, by far, the worst. The model performance in terms of the statistical parameters used stays rather constant with time over the periods (up to 117 h) studied here. It is shown that the method used to estimate the concentrations at the receptor locations has a significant effect on the evaluation results. The vertical wind component sometimes has a large influence on the model results, but on the average only a slight improvement over simulations which neglect the vertical wind can be demonstrated. Subgrid variability of mixing heights is important and must be accounted for.     

Modelling plume rise and dispersion from pool fires

This paper describes an investigation into the behaviour of smoke plumes from pool fires, and the subsequent generation of empirical models to predict plume rise and dispersion from such a combustion source. Synchronous video records of plumes were taken from a series of small-scale (0.06–0.25m²) outdoor methanol/toluene pool fire experiments, and used to produce sets of images from which plume dimensions could be derived. Three models were used as a basis for the multiple regression analysis of the data set, in order to produce new equations for improved prediction. Actual plume observations from a large (20.7 m×14.2 m) aviation fuel pool fire were also used to test the predictions. The two theoretically based models were found to give a better representation of plume rise and dispersion than the empirical model based on measurements of small-scale fires. It is concluded that theoretical models tested on small-scale fires (heat output ≈70 kW) can be used to predict plume behaviour from much larger combustion sources (heat output ≈70 MW) under near neutral atmospheric conditions.     

Validation of odor concentration from mechanical-biological treatment piles using static chamber and wind tunnel with different wind speed values

The basic principle of odor sampling from surface sources is based primarily on the amount of air obtained from a specific area of the ground, which acts as a source of malodorous compounds. Wind tunnels and flux chambers are often the only available, direct method of evaluating the odor fluxes from small area sources. There are currently no widely accepted chamber-based methods; thus, there is still a need for standardization of these methods to ensure accuracy and comparability. Previous research has established that there is a significant difference between the odor concentration values obtained using the Lindvall chamber and those obtained by a dynamic flow chamber. Thus, the present study compares sampling methods using a streaming chamber modeled on the Lindvall cover (using different wind speeds), a static chamber, and a direct sampling method without any screens. The volumes of chambers in the current work were similar, ~0.08 m³. This study was conducted at the mechanical-biological treatment plant in Poland. Samples were taken from a pile covered by the membrane. Measured odor concentration values were between 2 and 150 ou_E/m³. Results of the study demonstrated that both chambers can be used interchangeably in the following conditions: odor concentration is below 60 ou_E/m³, wind speed inside the Lindvall chamber is below 0.2 m/sec, and a flow value is below 0.011 m³/sec. Increasing the wind speed above the aforementioned value results in significant differences in the results obtained between those methods. In all experiments, the results of the concentration of odor in the samples using the static chamber were consistently higher than those from the samples measured in the Lindvall chamber. Lastly, the results of experiments were employed to determine a model function of the relationship between wind speed and odor concentration values.

Implications: Several researchers wrote that there are no widely accepted chamber-based methods. Also, there is still a need for standardization to ensure full comparability of these methods. The present study compared the existing methods to improve the standardization of area source sampling. The practical usefulness of the results was proving that both examined chambers can be used interchangeably. Statistically similar results were achieved while odor concentration was below 60 ou_E/m³ and wind speed inside the Lindvall chamber was below 0.2 m/sec. Increasing wind speed over these values results in differences between these methods. A model function of relationship between wind speed and odor concentration value was determined.     

A wind tunnel study of the flow structure and dispersion from sources upwind of three-dimensional hills

The flow fields around moderately steep hills of triangular cross section and varying crosswind aspect ratio and around a bell-shaped hill were examined by using models immersed in a simulated neutral atmospheric boundary layer in a meteorological wind tunnel. The triangular hills ranged from an axisymmetric cone to a two-dimensional ridge. Concentration patterns resulting from sources of three heights placed upwind of each of these hills were examined to determine plume deformations and terrain amplification factors. The separated flow fields, increasing in size with increasing aspect ratio, appeared to have dominating influences on the entire flow structure; changes in several flow parameters were plausibly explained in terms of the notion that the effective hill shape was the hill-plus-recirculation region rather than the actual hill shape. The concentration measurements showed strong distortions of plume shapes effected by the hills, with convergence in vertical planes and divergence in horizontal planes. Plumes from elevated sources approached the hill surfaces much more closely the smaller the aspect ratio; streamline displacements were generally within the limits suggested by potential flow theory. The terrain amplification factor A, defined as the ratio of the maximum surface concentration in the presence of the hill to the maximum in flat terrain, was found to decrease with increasing aspect ratio. For the half-hill-height sources, values of A ranged from 4 for the bell-shaped hill down to 1.5 for the two-dimensional ridge; for the hill-height source, from 1.8 for the cone down to 0.6 for the two-dimensional ridge. The latter value is suspect, however, because larger concentrations are expected downwind of the reattachment point, a region not probed in the current study.     

An ensemble Kalman filter for atmospheric data assimilation: Application to wind tunnel data

In the previous work (Zheng et al., 2007, Zheng et al., 2009), a data assimilation method, based on ensemble Kalman filter, has been applied to a Monte Carlo Dispersion Model (MCDM). The results were encouraging when the method was tested by the twin experiment and a short-range field experiment. In this technical note, the measured data collected in a wind tunnel experiment have been assimilated into the Monte Carlo dispersion model. The uncertain parameters in the dispersion model, including source term, release height, turbulence intensity and wind direction have been considered. The 3D parameters, i.e. the turbulence intensity and wind direction, have been perturbed by 3D random fields. In order to find the factors which may influence the assimilation results, eight tests with different specifications were carried out. Two strategies of constructing the 3D perturbation field of wind direction were proposed, and the result shows that the two level strategy performs better than the one level strategy. It is also found that proper standard deviation and the correlation radius of the perturbation field play an important role for the data assimilation results.     

A wind tunnel study of the effect of roadway configurations on the dispersion of traffic-related pollution

In this paper we examine the effect of different roadway configurations, including noise barriers and roadway elevation or depression relative to the surrounding terrain, on the dispersion of traffic-related pollutants for winds perpendicular to the roadway. A wind tunnel experiment modeling 12 different configurations was performed to study the flow fields and the concentration distributions resulting from emissions from a simulated six-lane highway. All of the configurations examined here reduced the downwind ground-level concentrations relative to that for a flat, unobstructed roadway; however, the degree to which the concentrations were reduced varied widely depending on the particular situation.Ground-level concentration data from the cases considered in this research indicate that a constant entrainment velocity can be used over the region beginning downwind of any initial disturbance to the flow resulting from the roadway configuration (e.g., a recirculation region behind a noise barrier) and extending at least to the end of our measurements. For example, for the case of a single noise barrier on the downwind side of the road, this region extends from approximately four barrier heights downwind of the roadway to 40 barrier heights. It was also found that the virtual origin concept is useful in describing the initial mixing created by the particular roadway configuration. To effectively model the influence of the roadway configuration on the dispersion, a combination of a virtual origin and an entrainment velocity may be effective. The magnitude of the virtual origin shift appears to depend on the particular roadway configuration, while the entrainment velocity appears to be a function of the friction velocity and the roadway geometry. These results suggest that road configuration must be taken into account in modeling near-road air quality.     

A comparison of tailpipe, dilution tunnel, and wind tunnel data in measuring motor vehicle PM.

Comparison between particle size distributions recorded directly at the tailpipes of both diesel and gasoline vehicles and measurements made using a conventional dilution tunnel reveals two problems incurred when using the latter method for studying particle number emissions. One is the potential for particulate matter (PM) artifacts originating from hydrocarbon material stored in the transfer hose connecting the tailpipe to the dilution tunnel, and the other is the particle coagulation (as well as condensation and chemical changes) that occurs during the transport. Both are potentially generic to current PM emissions measurement practices. The artifacts typically occur as a nanoparticle mode (10-30 nm) that is 2-4 orders of magnitude larger than what is present in the vehicle exhaust and can easily be mistaken for a similar mode that can arise from the nucleation of hydrocarbon or SO4(2-) components in the exhaust under appropriate dilution rates. Wind tunnel measurements are in good agreement with those made directly from the tailpipe and substantiate the potential for artifacts. They reveal PM levels for the recent model port fuel injection (PFI) gasoline vehicles tested that are small compared with the ambient background particle level during steady-state driving. The PM emissions recorded for drive cycles such as the Federal Test Procedure (FTP) and US06 occur primarily during acceleration, as has been previously noted. Light-duty diesel vehicle emissions normally exhibit a single lognormal mode centered between 55 and 80 nm, although a nonartifact nanoparticle mode in some cases appears at a 70-mph cruise up a grade.     

A wind tunnel study of dense gas dispersion in a stable boundary layer over a rough surface

Measurements of the vertical entrainment velocity into two-dimensional dense gas plumes over fully rough surfaces were carried out as part of a co-operative research programme with wind tunnel facilities in the USA. This paper presents results obtained for stable boundary layer conditions in the EnFlo wind tunnel at the University of Surrey; a companion paper treats the neutral boundary layer case. Mean velocity and temperature, turbulent normal and shear tresses, temperature fluctuations and heat fluxes were measured and used to demonstrate that a moderately stable atmospheric boundary layer had been successfully simulated in the tunnel. Entrainment velocities, W_E, were then deduced from the streamwise development of the concentration field, non-dimensionalised with respect to the friction velocity in the undisturbed flow, u^*, and correlated with the plume Richardson number, Ri^*. Higher non-dimensional entrainment speeds, W_E/u^*, were observed for Ri^*>5 in the stable boundary layer than in the neutral boundary layer, the difference growing with increasing Richardson number. Emission velocity ratios, W₀/u^*, were however larger in the stable experiments, and exceeded one at about Ri^*=18. Entrainment in the stable boundary layer appeared therefore to be more sensitive to emission velocity ratio than in the neutral case. Entrainment behaviour for Ri^*⩽5 followed that found in the neutral boundary layer. In this regime, use of the neutral boundary layer entrainment speed correlation is unlikely to lead to the over-prediction of plume dilution rates in moderately stable boundary layers.     

A wind tunnel study of dense gas dispersion in a neutral boundary layer over a rough surface

Measurements of the vertical entrainment velocity into two-dimensional dense gas plumes over fully rough surfaces were carried out as part of a co-operative research programme with wind tunnel facilities in the USA. This paper presents results obtained for neutral boundary layer conditions in the EnFlo wind tunnel at the University of Surrey; a companion paper treats the stable boundary layer case. Entrainment velocities, W_E, were deduced from the streamwise development of the concentration field, non-dimensionalised with respect to the friction velocity in the undisturbed flow, u^*, and correlated with the plume Richardson number, Ri^*. Results for Richardson numbers in the range Ri^*<15 were found to be well fitted by the empirical expression: W_E/u^*=0.65/(1+0.2Ri^*). Flow visualisation studies showed layered plume structures with a sharp upper interface at higher Richardson numbers and in this regime turbulent motion below the interface became progressively more intermittent as Ri^* increased. Measured turbulence levels collapsed within such high Richardson number plumes and flow and dispersion were significantly affected by molecular processes. Up-welling above the source was observed when the emission speed exceeded the approach flow friction velocity, though there was no clear evidence that this affected plume behaviour away from the immediate vicinity of the source.     

Modelling pollutants dispersion and plume rise from large hydrocarbon tank fires in neutrally stratified atmosphere

Petrochemical industries normally use storage tanks containing large amounts of flammable and hazardous substances. Therefore, the occurrence of a tank fire, such as the large industrial accident on 11th December 2005 at Buncefield Oil Storage Depots, is possible and usually leads to fire and explosions. Experience has shown that the continuous production of black smoke from these fires due to the toxic gases from the combustion process, presents a potential environmental and health problem that is difficult to assess. The goals of the present effort are to estimate the height of the smoke plume, the ground-level concentrations of the toxic pollutants (smoke, SO₂, CO, PAHs, VOCs) and to characterize risk zones by comparing the ground-level concentrations with existing safety limits. For the application of the numerical procedure developed, an external floating-roof tank has been selected with dimensions of 85 m diameter and 20 m height. Results are presented and discussed. It is concluded that for all scenarios considered, the ground-level concentrations of smoke, SO₂, CO, PAHs and VOCs do not exceed the safety limit of IDLH and there are no “death zones” due to the pollutant concentrations.     

Effects of wind shear on pollution dispersion

Using an accurate numerical method for simulating the advection and diffusion of pollution puffs, it is demonstrated that point releases of pollution grow into a shape reflecting the vertical wind shear profile experienced by the puff within a time scale less than 4 h. For distances beyond several 10 s of kilometers from a release point, shear-related dispersion effects are probably the dominant mechanism affecting the area and magnitude of surface impacts. For assessing long-range pollutant dispersion, the common assumption that pollutants disperse as horizontally spherical “puffs” in the atmosphere is inherently inaccurate since shear-induced horizontal spreading of pollution is not a homogeneous “turbulent-like” diffusion process. A Lagrangian puff model can simulate an area impacted by a pollution puff only if larger shear-dependent horizontal puff dispersions are assumed. However, even if impacted areas are reasonably simulated, peak concentrations will be severely underestimated since atmospheric puffs influenced by even small amounts of wind shear are nonspherical. If horizontal dispersion coefficients in a Lagrangian puff model are adjusted so that peak concentrations are correctly simulated, then the calculated pollution impact area will be severely skewed. In shear environments, no choice of horizontal dispersion coefficients in a single-puff Lagrangian model will yield reasonable correlations with puffs that are skewed into nonspherical shapes by atmospheric wind shear.     

A vegetation modeling concept for Building and Environmental Aerodynamics wind tunnel tests and its application in pollutant dispersion studies

A new vegetation modeling concept for Building and Environmental Aerodynamics wind tunnel investigations was developed. The modeling concept is based on fluid dynamical similarity aspects and allows the small-scale modeling of various kinds of vegetation, e.g. field crops, shrubs, hedges, single trees and forest stands. The applicability of the modeling concept was validated in wind tunnel pollutant dispersion studies. Avenue trees in urban street canyons were modeled and their implications on traffic pollutant dispersion were investigated. The dispersion experiments proved the modeling concept to be practicable for wind tunnel studies and suggested to provide reliable concentration results. Unfavorable effects of trees on pollutant dispersion and natural ventilation in street canyons were revealed. Increased traffic pollutant concentrations were found in comparison to the tree-free reference case.     

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

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

Validation and application of obstacle-resolving urban dispersion models

The development of micro-scale meteorological models has progressed in recent years. Some of them are already commercially available. With little hesitation, consulting engineers apply them to complex real-world problems. How accurate are the results? Using the example of urban dispersion models, the paper tries to give a critical assessment of the present ‘state of application’.     

A comparison of mesh-adaptive LES with wind tunnel data for flow past buildings: Mean flows and velocity fluctuations

In this paper we address two important aspects of micro-scale urban airflow model evaluation: (a) the identification of key flow features as dictated by the physics of the problem and as captured by the simulations, and (b) the comparison of important model output parameters (mean flows and fluctuations) with experimental data. A series of mesh-adaptive large eddy simulations (LES) was carried out for the study of air flows within two intersecting street canyons with varying building configurations. The novelty of the approach lies in the combination of LES with mesh adaptivity, which allows a variable-filter length and the implementation of an anisotropic eddy-viscosity model. Both coarse and fine-mesh simulations were carried out, using single and parallel-processor systems respectively. The simulations showed clearly that the expected flow patterns such as the street canyon recirculation and the street-mouth vortices, as well as the exchange of air flow at the street intersections, can readily be captured by the mesh-adaptive LES.In addition, the detailed comparisons of mean flows and fluctuations of the resolved velocity field with the measured data showed that the simulation results agreed well with the patterns and trends of the wind tunnel measurements. In most cases the finer-mesh simulations improved considerably the accuracy of the mean flows, especially for the symmetrical configuration. The improvement in the predicted fluctuations was less obvious, with several detector locations underpredicting the measured values, although the overall comparison was also satisfactory. The typical errors for the mean flows for all three building configurations were less than 30%, whilst for the velocity fluctuations less that 40%. Both the simulated means flows and turbulence levels were generally more accurate in the streets parallel to the wind (streamwise direction) than in the streets normal to the wind.     

A wind tunnel test of newly developed personal bioaerosol samplers

In this study the performance of two newly developed personal bioaerosol samplers was evaluated. The two test samplers are cyclone-based personal samplers that incorporate a recirculating liquid film. The performance evaluations focused on the physical efficiencies that a personal bioaerosol sampler could provide, including aspiration, collection, and capture efficiencies. The evaluation tests were carried out in a wind tunnel, and the test personal samplers were mounted on the chest of a full-size manikin placed in the test chamber of the wind tunnel. Monodisperse fluorescent aerosols ranging from 0.5 to 20 microm were used to challenge the samplers. Two wind speeds of 0.5 and 2.0 m/sec were employed as the test wind speeds in this study. The test results indicated that the aspiration efficiency of the two test samplers closely agreed with the ACGIH inhalable convention within the size range of the test aerosols. The aspiration efficiency was found to be independent of the sampling orientation. The collection efficiency acquired from these two samplers showed that the 50% cutoff diameters were both around 0.6 microm. However the wall loss of these two test samplers increased as the aerosol size increased, and the wall loss of PAS-4 was considerably higher than that of PAS-5, especially in the aerosol size larger than 5 microm, which resulted in PAS-4 having a relatively lower capture efficiency than PAS-5. Overall, the PAS-5 is considered a better personal bioaerosol sampler than the PAS-4.     

Droplet evaporation from porous surfaces; model validation from field and wind tunnel experiments for sand and concrete

The evaporation model of Roberts and Griffiths (1995 Atmospheric Environment 29, 1307–1317) has been subjected to an extensive validation exercise based on a major campaign of field experiments on evaporation from surfaces composed of sand and of concrete. This complements the previous validation which was limited to wind tunnel experiments on sand surfaces. Additionally, the validation using wind tunnel data has been extended to include concrete surfaces. The model describes the constant-rate and falling-rate periods that characterise evaporation from porous media. During the constant-rate period, the evaporation is solely determined by the vapour transport rate into the air. During the falling-rate period, the process in the porous medium is modelled as a receding evaporation front, the overall evaporation rate being determined by the combined effects of vapour transport through the pore network and subsequently into the air. The field trials programme was conducted at sites in the USA and the UK, and examined the evaporation of diethyl malonate droplets from sand and concrete surfaces. Vapour concentrations at several heights in the plume were measured at the centre of a 1 m radius annular source (of width 10 cm) contaminated by uniformly sized droplets (2.4 or 4.1 mm in diameter), key meteorological data being measured at the same time. The evaporation was quantified by coupling concentration and wind speed data. In all, 22 trials were performed on sand and concrete; a further 8 were performed on non-porous surfaces (aluminium foil and slate) as references. The model performance was evaluated against the experimental data in terms of two quantities, the initial evaporation rate of the embedded droplets, and the mass-fraction remaining in the substrate at intervals over the evaporation episode. Overall, the model performance was best in the case of the field experiments for concrete, and the wind tunnel experiments for sand; the performance for wind tunnel experiments for concrete was reasonably good; in the case of the field experiments for sand there was significant underprediction of evaporation rates, though the trends with the determining variables were well predicted.     

Wind tunnel investigation of the effects of a rectangular-shaped building on dispersion of effluents from short adjacent stacks

In a wind tunnel study, the influence of the highly turbulent region found in the lee of a model building upon plumes emitted from short stacks was examined through smoke visualization and tracer gas concentration mappings. A thick, simulated atmospheric boundary layer was used to provide background dispersion. A rectangular-shaped building with its length equal to twice its height and width was oriented with the long side perpendicular to the approaching wind. The stack was placed midway along the lee side of the building. In all phases of the study, each smoke or tracer release from the stack was repeated with the building removed. This allowed for a simple demonstration of the building wake effects. A simple mathematical model was developed that provided good estimates of concentrations in the building wake. The building influence was found to be reduced with increases in the effective source height. Application of the “2.5 times rule”, i.e. an effective source height at 2.5 times the height of the building, resulted in maximum ground-level concentration in the wake being approximately 20 per cent higher than found in the absence of the building. A stack 1.5 times the height of the building resulted in maximum ground-level concentrations in the wake being 250 per cent higher, a far more significant effect.