Abstract:

According to the Intergovernmental Panel on Climate Change (IPCC), the globe has been warming due to the increase in the amount of CO 2 and other greenhouse gases accumulated in the atmosphere. The effect of climate change result from global warming has been observed in different spatio-temporal forms such as rise in the sea level, melting glaciers, the change in the climate regions' boundaries, meteorological events and increase in disasters (Türkeş et al., 1999a; Türkeş, 2008). It is possible to examine the change in climate by analyzing atmospheric data. In this study, it is aimed to investigate the effect of climate change observed globally, on the boundaries of climate regions of Turkey. To accomplish this aim, after preprocessing, hierarchical cluster analysis is employed on the temperature (e.g. mean, minimum, maximum) and precipitation variables recorded in different stations of Turkey State Meteorology Service General Directorate in the period of 1950-2006. Since precipitation and temperature variables have different characteristics, multivariate cluster analysis is applied after all variables are standardized. In the study, four different hierarchical clustering methods with different distance measures are used. These are single linkage, average linkage, complete linkage and Ward method. To be able to test the stabilities of the clusters obtained using these methods, they are applied to the same variables obtained in two different time span: 1950-2006 and 1970-2006. It has been observed that Ward method creates more stable and meaningful clusters. Depending on the results of Ward method, it can be concluded that Turkey has seven climate regions with respect to temperature and precipitation characteristics. However, these regions are different than the ones known previously.

Keywords: