Assessing the effects of climate change on monthly precipitation: proposing of a downscaling strategy through a case study in Turkey

Abstract

The forecasting of future precipitation can be considered by using the outputs of the General Circulation Models (GCMs). In the study, a downscaling strategy was improved to forecast monthly precipitation over Tahtali watershed in Turkey for climate change scenarios using neural networks. First, predictor variables, which represent the monthly areal precipitation of watershed, were selected from the NCEP/NCAR reanalysis data set with the help of correlation and mutual information analyses. The study of the predictor selection showed that large scale precipitation at surface, air temperature at 850 hPa, and geopotential height at 200 hPa are the explanatory variables for downscaling. When the statistical performance measures were investigated, developed downscaling model trained with selected predictors was found satisfactory and was applied to simulate the future projections of selected two GCMs; namely, ECHAM5 and HADCM3. Finally, the simulation results were examined to assess the possibility of climate change effect on precipitation in the study area.