Predicting urban heat island circulation using CFD

Abstract

Most existing microscale computational fluid dynamics (CFD) models and mesoscale meteorological models cannot consider multi-scale urban wind flows, as neither can completely take into account the mesoscale and microscale physics, and their interactions. Here we suggest a CFD model which has good potential for development of the meso-micro scale models for predicting and designing multi-scale urban airflows. The principal idea is to use CFD numerical methods to solve a set of governing equations with appropriate boundary conditions that can govern both major mesoscale and microscale flow characteristics. Based on the coordinate transformation method proposed by Kristóf, Rácz and Balogh (2009), we derived a similar set of governing equations with some improvements and used an existing porous turbulence model for modeling the urban canopy layer. The approach was then successfully implemented using a commercial CFD package (Fluent) for studying urban heat island circulation (UHIC), which is considered to be one of the most difficult problems in CFD. Our predicted mean quantities agree well with existing data in the literature obtained from large eddy simulations, mesoscale models, and laboratory experiments. Our predicted results also reveal the effects of the different heat fluxes and urban height of a city on UHIC characteristics.