This project addresses the challenging problem of the impact that density stratification has on contaminant dispersion in urban areas. Potential atmospheric hazards include toxic industrial chemical spills, intentional or accidental releases of nuclear, biological and chemical agents, and power plant accidents. Despite its importance, there is a lack of efficient approaches to the problem of determining the parameters necessary to predict the concentration statistics under stably stratified atmospheric conditions. This shortcoming arises because of several factors:
i) scarcity of data due to the difficulty in conducting experimental campaigns for urban areas; ii) the variety of possible building configurations; iii) the multiscale nature of the flow over complex geometries.
This gap will be bridged by combining laboratory experiments and numerical modeling as complementary techniques, with a multi-institution collaboration.
The final goal is to progress towards a more complete understanding of the air pollution dispersion and micrometeorology in urban canopies in both neutral and density stratified conditions. Efforts will concentrate on four interrelated tasks:
i) Set up of a novel water tunnel experiment involving urban geometry and density stratification. ii) Creation of an experimental database of density-stratified urban flow to characterize mean dispersion and fluctuations through the collection of long time series of turbulent velocity and scalar fields. iii) Development of a new high performance Lagrangian computational fluid dynamics code for three dimensional stratified flow. Parallel simulations will be performed on state of the art high performance computer clusters. iv) Development of theory and parameterizations to estimate the impact of stratification on concentration statistics.
Intellectual Merit: The water tunnel data will be used to validate the results of a Lagrangian probability density function (PDF) numerical model specifically developed for urban environments. The analysis of the time series and the development of new parameterizations offer a unique potential for a significant improvement of our understanding of urban dispersion. These data will also be used to extend ongoing research on turbulence to stably stratified flows.
Broader Impacts: This project will have a substantial impact on the development of parameterizations, models and theories necessary for improving capability to predict dispersion of hazardous materials in urban areas. Findings of this project will be presented beyond the traditional university audience through outreaching activities. George Mason University (GMU) organizes the Annual Conference on Atmospheric Transport and Dispersion, regularly attended by various stakeholders such as government agencies, emergency managers and commercial companies as well as international researchers, where the results of this research will be presented and discussed. The multidisciplinary character of this research will have a significant educational impact on the graduate students involved in the project as they work interactively on both experiments and theory. The participation of the students in the GMU conference will give them practical knowledge of the spectrum of research and operational environments related to their projects.
."- , $323776, 9-1-2009 to 8-31-2014. The growth of population and traffic in cities everywhere exacerbates already serious levels of air pollution. The concomitant exposure can cause severe health problems: television news routinely shows images of people wearing face masks as a protective measure against air pollution. Central to the analyses of air quality is knowledge of the rate of dispersion of pollutants in the urban environment. Currently our knowledge of dispersion of pollution in cities or urban environments is less than satisfactory. The work undertaken for NSF Award 0849190 has successfully clarified the physics of dispersion in urban environments. There is now a framework for analysis of dispersion in cities. The problem of air flow in cities is difficult as it involves many length scales. Figure 1 below shows flow past model buildings in a lab experiment in a water tunnel. The left panel shows that flow is deflected around buildings, and that flow "whistles" in the gap between the buildings. (The gap is commonly called the canyon.) The fastest flow occurs in the canyon. However, as shown in the right panel, when the spacing between the buildings is smaller, the flow in the canyon is suppressed and all the flow is deflected around the buildings. So there are completely different flow fields for building configurations that are not too dissimilar. Difficulties in calculating dispersion occur as there are a myriad of building arrangements in any city. Meteorological data in and above cities is expensive and scarce. Any meteorological data is also specific to that city or even that location in the city where data was taken. This makes it difficult to make accurate calculations of dispersion. It also frustrates the development of physics and theories that can be applied to any other city. The control and repeatability of laboratory experiments offers an attractive alternative. Thus the work for NSF Award 0849190 involved laboratory experiments of dispersion in cities in a water tunnel. We have also undertaken extensive analysis of available field studies of dispersion in cities in different countries. Dispersion scenarios can be different for day time and night time. They can also vary between summer and winter. We establish that such differences can often be attributed to the effects of density stratification. The combination of lab experiments, field data and theory has resulted in a general methodology for making predictions of dispersion in any city. The two papers describing lab, field data and models are a framework for analysis of dispersion in cities that has been calibrated against field data. Thus it is now possible to accurately predict dispersion in cities at both daytime and nighttime (i.e. neutral and stratified conditions). We have concluded from our analysis that nighttime stratification in the urban atmosphere is weak but its effect on dispersion is not negligible. We also predicted the existence of a near and a far field dispersion regime. Our data analysis and scaling leads to a good collapse of the data, suggesting that urban dispersion is governed by the characteristic length scales of atmospheric boundary layer turbulence, rather than urban canopy length scales. The latter are more likely to affect dispersion only in the vicinity of the source. An unexpected outcome that evolved from theoretical analysis of turbulence dynamics involved in dispersion is the identification of new avenues of energy transport in turbulent flows. The discovery of the fluid dynamical Lorentz force & Poynting theorem provides new physics for the theory of turbulence. We have also performed numerical simulations of water tunnel experiments (Macdonald 2002) that helped us to obtain a better understanding of how to produce the right level of turbulent kinetic energy in a numerical simulation. We included the spikes and roughness elements used in the experiments in our numerical simulations as can be seen in Figure 2. These elements added in the downwind region of our simulation proved to represent the right amount of turbulence. We were able to match our numerical turbulent kinetic energy results with the experimental results. Our predicted plume footprints match with the experimental plumes as shown in Figure 3. Our research has also contributed to the development of techniques that characterize the source of release using evolutionary algorithms paired with machine learning operators. These techniques have helped to identify the location, size and emission rate of an unknown pollutant source. Source characterization is crucial for cases of accidental or intentional releases where location, size or emission rate are not known.
