Urban Bias in Temperature Time Series – a Case Study for the City of Vienna, Austria
- Cite this article as:
- Böhm, R. Climatic Change (1998) 38: 113. doi:10.1023/A:1005338514333
- 375 Downloads
Compared with other large cities Vienna shows different urban development characteristics. The city has had a zero population growth during 1951–1995, a period of rapid growth elsewhere. In spite of its stagnating population of about 1,6 million Vienna has had development in other areas: a doubling of living floor space, a two and a half-fold increase in total energy consumption, a 60% rise of traffic area. In contrast, forests have been reduced by 20% and grasslands within the city borders by 30%. Of the 34 temperature recording stations in the study area of 1450 km2, nine series passed the quality tests after careful homogenization. Three of these were in the rural environment and were used as reference series for the urban temperature excess at the other six stations in the urbanized area. The urban excess temperatures vary from site to site: from 0.2 K in suburban areas up to 1.6 K in densely built-up areas. The Vienna case study illustrates two features of more than local interest which should be considered in urban climatology as well as in time series studies where the urban temperature excess is regarded as a bias. Firstly, in a city with constant population the urban heat excess shows significant to strongly significant trends of up to 0.6 K in 45 years due to changes in urban morphology and energy consumption. Secondly, the urban heat island and its trend cannot be regarded simply for the city as a whole. There are different absolute levels, different annual variations and different increases of the urban temperature excess in different parts of a city. The urban effect is more strongly influenced by the local surroundings of the site than by the city as a whole. So, if possible, urban heat islands should not be described by a two station approach only (the typical airport-downtown comparison), nor should it rely on regression between population number and heat island.