CFD simulation of urban microclimate: Validation using high-resolution field measurements.

Abstract

Heat stress in urban areas can have detrimental effects on human health, comfort and productivity. In order to mitigate heat stress, Computational Fluid Dynamics (CFD) simulations of urban microclimate are increasingly used. The validation of these simulations however requires high-quality experimental data to be compared with the simulation results. Due to lack of available high-resolution high-quality experimental data, CFD validation of urban microclimate for real urban areas is normally performed based on either a limited number of parameters measured at a limited number of points in space, or on experiments for idealized generic configurations. In this study, CFD simulations of urban microclimate are performed for a dense highly heterogeneous district in Nicosia, Cyprus and validated using a high-resolution dataset of on-site measurements of air temperature, wind speed and surface temperature conducted for the same district area. The CFD simulations are performed based on the 3D Unsteady Reynolds-Averaged Navier-Stokes (URANS) equations and the simulated period covers four consecutive days in July 2010. It is shown that the CFD simulations can predict air temperatures with an average absolute difference of 1.35\u202f°C, wind speed with an average absolute difference of 0.57\u202fm/s and surface temperatures with an average absolute difference of 2.31\u202f°C. Based on the comparative results, conclusions are made regarding the performance of URANS for the selected application and possible reasons for deviations between measured and simulated results are discussed.