Observed and simulated effects of urban canopy on air temperatures in summer Tokyo

This study investigates the influences of urban canopy structure on surface air temperatures in Tokyo based on meteorological observations and numerical models. The surface air temperatures observed by high density meteorological monitoring system named METROS were used for the investigation. METROS stands for Metropolitan Environmental Temperature and Rainfall Observation System, which is composed of 20 stations (METROS20) settled on buildings' rooftops and 100 ground stations (METROS100) for surface observation with high horizontal resolution up to 2.5km over Tokyo 23 Wards area. In this study, hourly surface air temperatures from METROS100 in July 2002 were used, and their relations to local urban canopy structure around stations were analyzed using the sky view factor and dominant building type in each canopy as indices of canopy structure. As a result, diurnal temperatures indicated no obvious dependency on the local canopy structure possibly due to larger scale meteorological effects caused by prevailing winds or unstable atmospheric conditions. Contrastively, nocturnal temperatures showed systematic dependency indicating their increases at urban canopies with smaller sky view factors. Additionally, that nocturnal dependency was found to be more remarkable at canopies composed of concrete constructions with larger heat capacities like office building area. Those nocturnal results were supposed to imply thermal effects of urban canopy.

In order to clarify the physical processes that caused the observed effects on air temperatures, numerical models were applied to Tokyo metropolitan area. As a numerical weather prediction model, Weather Research and Forecasting (WRF) model was used without urban canopy parameterization, and reasonably found to underestimate the nocturnal surface air temperatures in Tokyo. Then, to take account of dynamic and thermodynamic effects of urban canopy, our original Canopy and Building energy Model (AIST-CBM) was incorporated, and one-way coupling computation from WRF to AIST-CBM was done. As a result, WRF & AIST-CBM finally reproduced METROS observed temperatures with their actual dependency on urban canopy structure.