FIRAT HAVZASI KARASU NEHRİNDEKİ AYLIK AKIMLARIN YAPAY SİNİR AĞLARI YAKLAŞIMINI İLE TAHMİNİ

This study aims to estimate streamflow values with artificial neural networks (ANN) using various meteorological parameters. In developing the ANN model, combinations of precipitation, air temperatures, and potential evapotranspiration were used as inputs, obtained. Meteorological data is divided into 70% train, 15% test, validation. model's design, training algorithms, network architecture, input combinations, number iterations tried, most suitable model was tested. Correlation coefficient (R), determination (R2), absolute error (AE), relative (ARE) coefficients compared, selected. According analysis results, optimal obtained 2000 iterations, architecture 4-4-1 Quasi-Newton algorithm. It determined that ANNs successfully modeled rainfall-runoff relationship produced reliable estimates. addition, it revealed inclusion by Thornthwaite method increases success.