Changes in surface air temperature for Mediterranean climate in Turkey

Abstract

Local climate influences of inland water bodies, complex topography, and surrounding seas cause temperate, arid, and continental climate properties to prevail with local variations in different parts of Turkey. The intra-regional variability of environmental factors creates uncertainties and challenges in climate modeling. Multi-model ensemble analysis is suggested to be used to characterize the uncertainties and minimize the generalization error in projections. This study is part of a research on climate change impacts in Turkey, focusing on the impacts on surface air temperature through a multi-model ensemble analysis of high-resolution climate models. The ensemble set comprises 12 EURO-CORDEX RCMs and two models from the Japan Meteorological Research Institute (MRI). Firstly, historical model data are validated with temperature records from 59 meteorological stations. Furthermore, changes in temperature climatology in the future in short- (2020-2030), medium- (2031-2050), and long-term (2051-2100) horizons are analyzed and compared with the precipitation changes. In the ensemble, two MRI models (MRI-AGCM, NHRCM) and two CORDEX RCMs nested in the HadGEM2-ES (RCA4 and CCLM4-8-17) perform best to replicate the spatial variability of climatology. The 14-member ensemble projects a gradual increase in the temperature up to 4.5 and 6.6 ºC under RCP4.5 and RCP8.5 scenarios, respectively. The projections agree on an inverse relationship between temperature and precipitation changes. More substantial impacts are projected in inland compared to coastal regions.