Wind speed is reduced above urban areas due to their high aerodynamic roughness. This not only holds for above the urban canopy. The local vertical wind profile is modified. Aerodynamic roughness (both roughness length and displacement height) therefore is relevant for many fields within human biometeorology, e.g. for the identification of ventilation paths, the concentration and dispersion of air pollutants at street level or to simulate wind speed and direction in urban environments and everything depending on them. Roughness, thus, also shows strong influence on human thermal comfort. Currently, roughness parameters are mostly estimated using classifications. However, such classifications only provide limited assessment of roughness in urban areas. In order to calculate spatially resolved roughness on the micro-scale, three different approaches were implemented in the SkyHelios model. For all of them, the urban area is divided into reference areas for each of the obstacles using a voronoi diagram. The three approaches are based on building and [+one of them also on] vegetation (trees and forests) data. They were compared for the city of Stuttgart, Germany. Results show that the approach after Bottema and Mestayer (J Wind Eng Ind Aerodyn 74-76:163-173 1998) on the spatial basis of a voronoi diagram provides the most plausible results. [ABSTRACT FROM AUTHOR]