Abstract
This study aims at establishing a numerical simulation method for obtaining high wind data in actual urban area. Height data of surface roughness measured by remote sensing technique are used and can directly express geometrical boundary of the urban areas on the numerical model. We employ LES (large eddy simulation) model for complex turblulent flows among various shapes of buildings, so time-dependent fluctuations of wind velocities and pressures can be predicted. Vertical profiles of mean wind velocity and turbulent intensity and power spectrum density functions of wind velocity fluctuation are examined. Based on the computed data, the potentiality of prediction for high wind in actual urban areas is also discussed.
