Review of Japanese Urban Models and a Scale Model Experiment - Meteorological and Air Quality Models for Urban Areas - page 32

that incorporate simple urban canopy models for investigating the urban heat island
(UHI) (Kondo and Kikegawa, 2003; Kusaka and Kimura, 2004a, b; Tokairin et al.,
2006; Kusaka and Hayami, 2006; Ohashi et al., 2007). A distinct improvement
commonly observed in simulations that include a simple urban scheme is a bet-
ter reproduction of the nocturnal urban temperature field because of the predicted
large fraction of heat storage. Increased urban heat storage is a crucial process that
must be considered. However, it is currently not clear what level of complexity to the
representation of the urban surface is necessary for mesoscale simulations. Model
comparison will be useful to identify the dominant physical processes and the most
suitable representation for urban mesoscale simulations (e.g. Best, 2006).
5.1.2 Computational Fluid Dynamics Models in Japan
An alternative approach is to explicitly resolve the urban infrastructure using com-
putational fluid dynamics (CFD) technology. Murakami et al. (1999) suggested the
use of CFD analysis winds from human to urban scales. CFD methods have been
used to examine flows around a single building (e.g., Murakami et al., 1990; Kogaki
et al., 1997; Sada and Sato, 2002). Murakami et al. (1996) reviewed the strengths
and weaknesses of various modelling approaches, including Reynolds-Averaged
Navier-Stokes (RANS) models and Large Eddy Simulations (LES). Several numer-
ical investigations have considered single street canyon flows (see review Li et al.,
2006).
Limitations of computational resources, however, have so far restricted CFD
applications to simple cases such as turbulent flows within and above a group
of buildings. Turbulent flows within and above regular obstacle arrays have been
investigated by direct numerical simulations (Miyake et al., 2001; Nagano et al.,
2004) and LES (Kanda et al., 2004; Kanda, 2006a). Results from LES applica-
tions (Kanda et al., 2004; Kanda, 2006a) suggest there are some important physical
aspects that are currently not emphasized in simpler CFD simulations i.e., large dis-
persive momentum flux within the urban canopy layer due to a mean stream such
as recirculation; intermittent urban canyon flow; non-persistent stream patterns; and
longitudinally elongated streaks of low speed over building arrays with a scale an
order of magnitude larger than individual buildings. Dispersive flux contributions
should be included in sink/source term approaches (e.g., Lien and Yee, 2005). Cou-
pling of very large turbulent eddy motions and street canyon flows should be con-
sidered in the physical interpretation of turbulent flows within single street canyons.
Application of CFD technologies to real cities is a promising breakthrough in
studies of urban meteorology (Fig. 5.1). CFD technologies do not require the con-
cept of roughness, the Monin-Obukhov Similarity Theory (MOST), or simplified
representative buildings, but instead explicitly include turbulent flows at multiple
scales from individual buildings to the boundary layer, thereby representing more
realistic effects of cities onto the atmosphere. Such applications are currently still
in the trial stage because of difficulties in the boundary conditions and limitations
in computational resources (e.g., Tamura et al., 2002; Ashie et al., 2005; Ashie and