Exploring city-scale thermal and wind environments

Abstract

All buildings and streets are connected in terms of climate and environment in large cities, which establishes a complex urban climate problem. Currently, one neither should design a building in a city in isolation nor include all buildings when designing a new one. A new numerical approach is thus required, which can simulate the flows around several buildings, the whole urban domain, and beyond. This is a great challenge as it needs to consider the influence of the mesoscale atmosphere, microscale physics, and their interactions. Existing mesoscale or microscale models alone cannot achieve this. Integration of the two models is unsatisfactory for two-way coupling. In this thesis, I propose a new approach referred to as city-scale computational fluid dynamics (CSCFD), which modifies the governing equations and turbulence models in traditional CFD to include both length scales, and which adopts a coordinate transformation to consider the compressibility effect. A porous turbulence model is used to simultaneously investigate the wind environment in the entire city and predict detailed flow field around several buildings. An absorbing layer, common in mesoscale models, has been introduced to suppress numerical instability, while an analytical surface temperature is calculated for the new surface temperature boundary condition to save computational time, compared to coupling the heat balance equation in CFD. The CSCFD approach was evaluated by comparing it with existing results from laboratory experiments, field experiments and appropriate numerical models, including prediction of the low-aspect-ratio plumes in an ideal city, the overall air flow of the “porous” model, and the diurnal thermal flows driven by buoyancy forces. The approach successfully simulated complex urban heat island circulation (UHIC) in full scale. A computational domain with substantial resolution variations was created and evaluated, which mainly consisted of the fully-resolved and porous areas. Although all the buildings and their intervals were not fully resolved, the major characteristics of the airflow were reasonably provided. In this study, I analysed the structure and evolution of the UHIC when the city is a homogeneous flat surface and a homogeneous porous medium. In the ‘porous city’, wind decreased, but it was also induced when synoptic wind is calm. The existence of the porous layer changes the UHIC-induced plume due to the following: The porous layer height affects the entrainment of the lower part, and the porosity alters its shape. The sensible heat flux is dominant in influencing the strength of UHIC. A comprehensive case study was carried out to demonstrate the viability of the CSCFD in simulating the diurnal environment around a building in the mesoscale domain. The quality of the performance and its moderate complexity demonstrate the feasibility of CSCFD for urban planning and for analysis of a range of urban environmental problems. Additionally, the street air temperature was measured by the vehicle traverse method in Hong Kong for 15 days in summer and winter. A street air warming phenomenon was observed and analysed. However, I haven’t yet validated CSCFD using this data. Further fundamental development and evaluation of the CSCFD is still needed.