LES Estimation of Environmental Degradation at the Urban Heat Island due to Densely-Arrayed Tall Buildings

For the mitigation of heat island effect on coastal cities, it is sometimes expected that the sea breeze comes into the inland area of a city, where its cold air mingles with the hotter air over and inside the urban canopies. In the case of Tokyo, there exists a coast line at the southeast boundary, so the sea breeze is observed to reach into about 20 km inland. But recently several very tall buildings have been constructed at Shiodome near this coast line between the center of Tokyo and the Tokyo bay. We are concerned about that these tall buildings block a passage of sea breeze into the downtown of Tokyo. Of course, it is generally thought that the size of heat island phenomenon itself ranges to several tens square kilo-meters with meso-scale behavior, which is ten times larger than the Shiodome area of some square kilo-meters. Such a large discrepancy tends to introduce this local effect to be not serious but trivial. Therefore the treatment to aspect of the ground surface has been thus far very poor. But for the analysis of the heat island, the surface treatment and the area considered should be selected by a significance of phenomenon that each issue encompasses. In such a convection-dominant phenomenon at Shiodome, it is important to focus on details of the very local flow characteristics inside both of the roughness layer and the urban canopies, and to clarify the effect of cold air contained within the sea breeze on heat environment in the downtown of Tokyo. To solve this issue, we have to formulate the physical model or the numerical model which represents correctly aspect of the ground surface condition consisting of buildings, structures and vegetation et al. In order to get the data for building shapes in a local area, the electronic mapping information is utilized. The present study uses height data of surface roughness and expresses directly surface shapes of urban area for the numerical simulation. The previous numerical simulation of the boundary layers usually has employed the log law model, or sometimes a little sophisticated version such as the wall layer model for the treatment of the ground surface condition. It means that the integral quantities are utilized for representing total boundary effects, but not local effects. Recent data for building height determining the surface roughness are much improved and have a horizontal resolution with about 2 or 2.5 m, so it might be enough to simulate roughness shapes, even an individual residential house, placed on the ground surface in urban areas. Further, in the case of the numerical simulations for winds in urban canopies, which need to deal with details of existing flows throughout spaces between buildings, so the RANS (Reynolds averaged Navier-Stokes) technique might have advantage to compute the flow field highly resolved by the grid from standpoints for computer capability. However, in the case that we focus on the roughness layer or inside the urban canopies, the flow in this near-wall region is very complicated and unsteady due to separation, vortex-shedding, rapid flow contraction and extension among roughness elements. Nevertheless, the complete RANS modeling of turbulence has not been developed yet for such a complex flow. Also, for the estimation of mitigation of heat island effect by utilizing the local area circulation, the numerical model which can predict time sequential turbulence is appropriate, because the convection by fluctuating behavior of turbulent flows represents directly and strictly a scale or a range of heat transport. Hence this study employs the LES (Large eddy simulation) technique, which is applied to the wind flow over actual roughened ground surface in Tokyo area. For the horizontal inflow boundary condition of the computational domain at the Shiodome area of 1 km by 1.5 km, turbulent flow data such as wind velocity or temperature are imposed by taking into consideration the predictive results of meso-scale meteorological model with 1 km resolution. For the generation of inflow turbulence, we additionally set up the driver domain of 1 km by 2.5 km and solve a spatially developing turbulent boundary layer over uniformly rough surface by applying the quasi-periodic condition based on the rescaling technique for the rough-wall boundary layer. According to the previous study, the field measurement around the Shiodome area showed that the extreme reduction of wind velocity behind a group of tall buildings. We compare the LES results with these field measurement data and validate the LES model. This study employs the dynamic procedure for determining a coefficient of the sub-grid scale modeling for turbulence, so the applicability of the present procedure for boundary layer over quite rough wall consisting of actual buildings and houses with various shapes, trees and vegetation is also discussed. Next, we show the wind flow characteristics around tall buildings at the Shiodome area. Based on the flow visualization, it can be seen that very complicated but specified wind flow patterns are formed in the wake of a group of tall buildings, and these patterns vary in the vertical direction. These flow patterns are thought to strongly affect the formation of thermal environment. Also, we bring into focus the turbulence structures in the roughness layer and the urban canopies, especially examine the spectral density functions and spatial coherence in the near region of the roughened surface and discuss the universality of the wind characteristics in the actual urban area. In view of transport process, we should note the occurrence of high wind inside the street canyon. It can be imagined that the heat easily moves along the street by convection effect of wind flow. Finally, based on these results at the Shiodome area, we provide the numerical data which is utilized to estimate the environmental degradation at the urban heat island due to densely-arrayed tall buildings.