Three dimensional flow structures and turbulence distribution in an urban environment

Understanding and controlling the dispersion of pollutants and contaminants in urban areas has become a major focus recently. Field measurements, numerical studies, and wind tunnel experiments have increased in number. Specifically, there is a growing need for a spatio-temporal description of such complex flow fields under well-controlled conditions, typically obtained in wind tunnel experiments. The reduced scale model of interest is a 4 by 3 array of cuboid blocks in an experimentally modeled, neutrally stratified, atmospheric boundary-layer. The use of Stereoscopic Particle Image Velocimetry (SPIV) allows for a three-dimensional description of this urban flow. A large amount of SPIV data is collected upstream and in each middle street of the urban environment allowing for a study of the flow evolution from street to street. Valuable information about the flow structures are presented along with the mechanisms responsible for contaminant transport and dispersion. The effects of small incidence angles of the incoming flow with respect to the urban array and the effects of streamwise spacing between streets on the flow characteristics are investigated. A major observation from this work is that a strong channeling effect is observed for incidence angles as small as 4.5° and is found to be comparable in strength to that observed in other investigations for much larger angles. A coupling between this channeling effect and the structures responsible for contaminant transport is revealed.;An innovative method using sparse measurements to estimate the continuous temporal evolution of the dominant structures in the flow is investigated. Proper Orthogonal Decomposition is used to obtain a reduced-order representation (ROR) of the flow field. Sparse velocity measurements within the domain serve as input to measurement models that provide an estimation of the ROR of the velocity field. This ROR of the flow field could be regarded as the first that provides a temporal evolution of a spatially well-resolved flow field in a complex geometry. Finally, a linear state-space model is used to describe the continuous temporal evolution of the ROR of the velocity field which is of primary importance with respect to contaminant tracking at the urban scale.