Urban climate experimental results from the Comprehensive Outdoor Scale Model (COSMO) were used to estimate roughness lengths for momentum and heat. Two different physical scale models were used to investigate the scale dependence of the roughness lengths; the large scale model included an aligned array of 1.5-m concrete cubes, and the small scale model had a geometrically similar array of 0.15-m concrete cubes. Only turbulent data from the unstable boundary layers were considered. The roughness length for momentum relative to the obstacle height was dependent on wind direction, but the scale dependence was not evident. Estimated values agreed well with a conventional morphometric relationship. The logarithm of the roughness length for heat relative to the obstacle height depended on the scale but was insensitive to wind direction. COSMO data were used successfully to regress a theoretical relationship between B Ϫ1 , the logarithmic ratio of roughness length for momentum to heat, and Re*, the roughness Reynolds number. Values of B Ϫ1 associated with Re* for three different urban sites from previous field experiments were intercompared. A surprising finding was that, even though surface geometry differed from site to site, the regressed function agreed with data from the three urban sites as well as with the COSMO data. Field data showed that B Ϫ1 values decreased as the areal fraction of vegetation increased. The observed dependency of the bulk transfer coefficient on atmospheric stability in the COSMO data could be reproduced using the regressed function of Re* and B Ϫ1 , together with a Monin-Obukhov similarity framework.

A simple urban energy balance model for mesoscale simulations (SUMM) was tested using results from an outdoor scale-model experiment. The model geometry is assumed to be an infinitely extended regular array of uniform buildings, each of which is composed of six faces (roof, floor, and four vertical walls). The SUMM explicitly considers the three dimensionality of surface geometry and theoretically predicts the energy balance at each face without time-consuming iterations. The SUMM was compared with outdoor scale-model experiments. The simulated energy balance and surface temperatures agree well with the values measured on a reduced-scale hardware model corresponding to the numerical model geometry.

A Simple 3-dimensional Urban canopy Model for Meso-scale simulation (SUMM) was proposed.The model predicts the energy balance of constituent faces (roof, floor and 4 vertical walls) without time-consuming iterations.In the model, complex 3-dimensional radiation processes (multi-reflection processes) are theoretically solved, resulting in a large reduction of computational time. As for the turbulent transfers, the network of resistance is adopted. We also conducted outdoor scale model experiments having the similar geometric structure with SUMM, to construct the databases (energy balances and surface temperatures) for model validation, and to examine unknown model parameters (e.g. turbulent transfer coefficients).

No abstract