Abstract
Urban, suburban, and rural air temperature data for clear and calm nights were analyzed to detect typical thermal variations found within a single city. Measurements taken over a three-year period formed the basis of the analysis. The measurement sites were four permanent climatic stations that were similar in geomorphological and topographical respects but exhibited different building structures. In the city center, the street geometry was shown to influence the air temperature at a small but constant rate, larger in the summer than in winter. The suburban site resembled the rural area with regard to vertical temperature distribution and cooling rates. However, the existence of a suburban heat island of low magnitude and short duration was proven and attributed to geometry. The mean air urban heat island was preserved for a longer time and demonstrated an increase of approximately 1° C in all seasons when data from a canyon site of SVF = 0.5 were used rather than data from an open urban area. The results suggest that street geometry does influence intraurban air temperature variations to some extent, but this effect should not be exaggerated. The effect of differences in thermal properties also is discussed.