Investigation of possible effects of climate change on seawater temperatures: a case study for Black Sea

The aim of this study is to investigate the possible effects of climate change on the seawater temperatures (SWT) in the case of the Black Sea. The monitoring data from five meteorological stations operated along the coast of Turkey was for the period 1963-2014. The SWT values for the future period (2023-2052 and 2053-2082) were obtained for each station by downscaling the optimistic (SSP2-4.5) and pessimistic (SSP5-8.5) scenario outputs, belonging to four general circulation models (GCMs) chosen from the CMIP6 archive, to regional scale. The coarse-resolution GCM outputs were downscaled to a regional scale using two statistical downscaling techniques: conventional regression analysis (CRA) and artificial neural networks (ANNs). Four performance statistics, namely root mean square error, mean absolute error, scatter index, and Nash-Sutcliffe efficiency coefficient, were used to compare the downscaling efforts of the CRA and ANNs techniques. The downscaling capability of the ANN-based models was higher than that of the CRA-based models based on the performance statistics. For this reason, the ANN-based models were preferred to forecast the future SWT values. Considering the average annual SWT values, the variation of the values estimated under the SSP2-4.5 and SSP5-8.5 scenarios with the monitoring data was studied. It is predicted that the average annual values may rise by 0.67 to 1.11 °C under the SSP2-4.5 scenario and 0.56 to 1.03 °C under the SSP5-8.5 scenario for the period 2023-2052. It is also expected that these values may rise by 1.23 to 1.71 °C under the SSP2-4.5 scenario 2.06 to 2.80 °C under the SSP5-8.5 scenario for the period 2053-2082.