Research Information System
Global Nest Journal, vol.23, no.2, pp.232-240, 2021 (SCI-Expanded)
Researches to foresee the possible effects of climate change on the environment and living beings for taking necessary precautions on time have increased in recent years. In the improvement of these studies, especially the reduction of estimation errors by downscaling the outputs of global climate models played an important role. In this study, the effect of the statistical downscaling method on improving the prediction accuracy of global climate models (GCM) was investigated. For this purpose, a statistical downscaling method based on multiple linear regression was applied to improve monthly precipitation estimates of 3 different GCM (CanESM2, GISS-E2H, and CSIRO Mk 3-6-0) used in future climate predictions. The effect of this method on improving GCM prediction accuracy was determined by comparing the results obtained as a result of scale reduction with the results obtained from the observation station. The predictive parameters for global climate models were determined using downscaling methods by applying correlation analysis for the study area. As a result of this analysis, it was seen that the air temperature and specific humidity values at the pressure level of 925 hPa and the geopotential height value at the 300 hPa pressure level had the best correlation for the years 1970-2005. The usability of three different global climate models for the forecast of future precipitation in the Antakya district of Hatay province was investigated using multiple linear regression analysis, one of the downscaling methods. As a result of the statistical analysis, it was seen that the use of the downscaling method increased the accuracy of all prediction models.