Changes in Temperature Extremes in High Resolution Simulations of RegCM driven by MPI-ESM2 under RCP8.5 and RCP4.5 Scenarios over Turkey

Wednesday, 16 December 2015
Poster Hall (Moscone South)
Merve Acar and Yurdanur Sezginer Ünal, ITU, Aeronautics and Astronautics, Istanbul, Turkey
In this study, the effect of climate change on temperature extremes over Turkey is investigated in high resolution climate model simulations driven by a global model outputs. MPI-ESM2 earth system model simulations are downscaled to first 50 km coarse resolution over Med-CORDEX domain and then 10 km high resolution over Turkey by regional climate model, RegCM4.3. The simulations cover the periods of 1970-2000 for the reference and 2015-2100 for the future with proposed changes under Representative Concentration Pathways 4.5 (RCP4.5) and 8.5 (RCP8.5).

High resolution gridded surface climate dataset is of great value for the validation phase of the high resolution climate models and the daily temperature observations are interpolated to regular grids which coincide with the simulation’s grids by using PRISM (Parameter-elevation Relationships on Independent Slopes Model) approach. High resolution regional climate model performance is evaluated for the reference period by using gridded observations for both averages and extremes of temperature. The RegCM coupled with MPI-ESM2 shows negative biases over most of Turkey. Hence monthly mean value bias correction is applied to temperature simulations of reference and 2015-2100 periods. Extreme temperature climate indices such as FD0, SU25, TX10p, TN10p, TX90p, TN90p, TX35, WSDI, CSDI and DTR are calculated and compared to reference period. Bias corrected high resolution simulations show good agreement with the observations.

The climate indices are calculated for the intervals of 2015-2040,2041-2070,2071-2100, for RCP4.5 and RCP8.5 scenarios. FD0, TN10p, TX10p and CSDI indices decrease through the end of century for both scenarios and the most dramatic changes occur on the eastern and central part of Turkey. The rate of decrease is more pronounced in RCP8.5 scenario. On the other hand, the indices of SU25, TX90p, TN90p, TX35 and WSDI increase till 2100. 30 year averages of SU25 and SU35 over all Turkey increase from 64 days and 7 days to 131 days and 29 for RCP4.5 and to 157 days and 51 for RCP8.5, respectively. However, the rate of change for warm climate indices does not show uniform distribution. Especially, the changes on the regions of southern Anatolia along the Aegean and Mediterranean Sea are large while there is almost no change on the regions along the Black Sea coast