CR 70:77-93 (2016)  -  DOI: https://doi.org/10.3354/cr01410

ABSTRACT: In this study, the climate zones of Turkey were re-examined using different objective statistical tests based on the differences in the behaviour of meteorological variables, and a comparative analysis of 2 consecutive periods was performed statistically. The data consisted of total precipitation, and minimum, maximum and mean air temperature series recorded from 1950-2010 at 244 climatological/meteorological stations operated by the Turkish Meteorological Service. K-means and hierarchical clustering methods were applied separately to each variable to obtain surface air temperature and precipitation patterns in Turkey for the periods of 1950-1980 and 1981-2010. Paired-samples Student’s t-test (paired t-test) and Pitman-Morgan (P-M) t-test were used to detect possible changes in the mean and variance of the series in the transition from one period to the other. The results of the analysis reveal that the climate characteristics of Turkey are generally similar for the temperature series under study. However, there are some changes in the existing geographical patterns of the climate regions. Statistical tests show that all 3 air temperature series increased after 1980. The major changes appeared in the precipitation regions of Turkey: there were significant changes in the continental central, central-west and central-east Anatolia regions, and in the continental north and eastern Anatolia region. It was also apparent that precipitation amounts increased in the northern and eastern regions of Turkey after 1980, but amounts decreased in the west, central and southern regions, most of which are generally characterized as having a dry summer subtropical Mediterranean climate.

KEY WORDS: Turkey · Mediterranean climate · Abrupt climate change · k-means · Hierarchical clustering · Statistical comparison tests