E.V. Goulart, S. Branford, T.G. Thomas, O. Coceal, S.E ...

14th Conference on Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes ? 2-6 October 2011, Kos, Greece

H14-199 DISPERSION OF A POINT-SOURCE RELEASE OF A PASSIVE SCALAR THROUGH AN ARRAY OF CUBES

WITH DIFFERENT CONFIGURATIONS.

E.V. Goulart, S. Branford, T.G. Thomas, O. Coceal, S.E. Belcher

Abstract: Understanding dispersion from localized sources in urban areas is crucial for modelling releases from accidents or terrorist attacks. Whilst there is a vast literature on urban dispersion in relation to air quality, there is relatively little work on dispersion from localised sources, particularly in the short range, where releases can have the most dangerous impact and the local buildings more strongly affect the dispersion. In this work we investigate the dispersion of a point-source release of a passive scalar by means of direct numerical simulations (DNS) over cubical urban-like obstacles. In order to investigate the influence of geometry on the near-field dispersion, two different arrays of cubes are used: a regular array and a staggered array. For the regular array the ground source is located in an intersection and for the staggered array three simulations with different ground source locations are analyzed. The wind direction is either perpendicular to the buildings or at an angle of 45 degrees. Several similar characteristics in the near-field dispersion can be identified across the different configurations. Firstly, the wake regions behind the buildings are dominated by a recirculation, which cause material to be trapped and rereleased more slowly. This process gives rise to so-called secondary sources in the near-field region. Secondly, updrafts behind the buildings transport material towards the building top and out of the array. Thirdly, an aisle of fast flow across any open channels in the domain removes material. The location of the source in relation to neighbouring buildings determines whether a secondary source occurs. Secondary sources influence the mass transfer at the top of the canopy. Moreover, the presence of secondary sources affects the dispersion above the building canopy and the re-entrainment of material further from the source. Despite the complexity of the flow and dispersion patterns across the domain, temporal and spatial fluctuations in the concentration field within the array are smaller than in open terrain. This is because of efficient mixing of the scalar due to small-scale turbulent fluctuations in the flow within the array.

H14-231 APPLICATION OF THE LAGRANGIAN MODEL GRAL AND THE ONLINE COUPLED METEOROLOGY-

CHEMISTRY MODEL WRF/CHEM TO THE SANTIAGO DE CHILE REGION

Peter Suppan

Abstract: Within the project "Risk-Habitat-Megacity" a chain of coupled models has been implemented to investigate the assessment of traffic emissions in the Metropolitan Area of Santiago de Chile. Based on traffic counts and traffic modeling work (ESTRAUS) for the road network of Santiago (about 8000 street sections), traffic emissions have been calculated traffic emission model MODEM and allocated to the respective road sections. These detailed results were used for micro scale modeling with the Graz Lagrangian model GRAL at the street level. The fully coupled "`online"' meteorology-chemistry model WRF/Chem will supply meteorological input parameters for GRAL as well as regional background information for a urban scale assessment study in this region. The base year 2010 is the initial point for the development of future scenarios which was one of the topics of the "Risk-Habitat-Megacity" initiative. The scenarios consist of three alternatives for the year 2030 defined as business as usual (BAU), collective responsibility (CR) and market individualism (MI) in order to identify future strengths and problem "hot spots". Highest concentrations are found along main access roads and in the central areas of Santiago. In the scenarios a significant decrease of primary pollutants (NOx, CO and PM10) can be observed. Despite the increase in motorization and vehicle miles travelled, the decrease in pollution is caused by substantially improved vehicle emission standards. The general decrease of emission levels of about 50% is accompanied by local increases, especially on the northern and southern access roads. These patterns can also be observed in the CR scenario, but with generally higher reduction levels than in the BAU and MI scenarios. The related health assessment resulted in a daily risk decrease due to pollutants for the scenarios CR and MI. Such reductions (based on the target values of the Directives 1999/30/EC and 96/62/EC of 40 (20) mg/m3 as yearly exposure averages) may decrease the associated mortality risks by around 20-35% in areas of the city with high reduction potentials.

H14-285 VALIDATION OF THE URBAN SCALE POLLUTION MODEL SIRANE-2.0 AGAINST FIELD DATA IN THE

CITY OF LYON

Lionel Soulhac, Pietro Salizzoni, Marie-Laure N'Guyen, Florence Troude, Isabelle Rios

Abstract: SIRANE is an operational urban scale pollution model, developed to simulate atmospheric concentration field over a whole city, at an hourly temporal scale and at the street spatial scale. SIRANE is based on a representation of a city as a network of interconnected streets. Specific parameterizations are used to model flow and dispersion within each street, exchange at each intersection, and a Gaussian plume/puff model is used to deal with dispersion over the roof level. A simple chemical scheme is included in the model to describe the Chapman cycle for NO/NO2/O3. During this work, SIRANE has been applied to the agglomeration of Lyon and hourly simulations have been performed for the year 2008. A network of 21833 street segments is used to describe the geometry of the canopy over a 30km x 40km domain. For each hour and each street, traffic emissions are provided. Industrial point sources and other surface emissions are also considered. The meteorological wind field over the city is calculated by the model from input data recorded in a meteorological ground monitoring station. Results of the model are then compared with ground concentrations hourly measured at 20 monitoring stations during 2008. 75 passive diffusion samplers were also used during 6 weeks to get a detailed spatial distribution over the west part of the city. The results are analysed in order to explain the agreement and the differences between the model and the measurements. A sensitivity analysis to the model parameters and to the input data assumptions has also been performed. Finally, an assimilation algorithm has been tested in order to use concentration measurements to reduce the error between the model and these measurements. This approach allows a significant improvement of the results and will be applied in the future by the local authority to predict concentrations maps and population exposure.

Topic 5: Urban scale and Street Canyon Modelling

505

 ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download