Estimation of Gediz basin irrigation water demands under possible climate change scenarios

In the study, it was aimed to prepare net irrigation water demand projections under different climate change scenarios and projections were examined through an exemplary application on irrigation unions in Gediz Basin. In the study, downscaling models based on artificial neural networks (ANN) were established for rainfall and temperature projections of Manisa, Menemen, Salihli and Turgutlu meteorological stations in the basin. In the models, NCEP/NCAR re-analysis variables were used as predictors. All possible lineer regression relations and performance criteria have been determined from these variables to be used in the downscaling models. The downscaling models calibrated with the optimal predictors convert the coarse resolution results of both reference period (20C3M) and future period (A2, A1B and B1) scenarios of ECHAM5 climate model to the station scale rainfall and temperature forecasts. Correction of biases in the forecasts are achieved by using cumulative distribution functions. Finally, corrected rainfall and temperature forecasts were taken as input in the Blaney-Criddle method and then net irrigation water needs were calculated. According to the A2, A1B and B1 scenarios, the mean temperatures of 2021-2100 period could increase by 2.9, 3.2 and 2.5 oC respectively and the precipitation could decrease by 17.8%, 29.2% and 16.7%, respectively, when the obtained results were evaluated over study region. Under these probable impacts, it is expected that the average net irrigation water need will increase by 14.9%, 17.5% and 12.5% according to the scenarios A2, A1B and B1, respectively.