Rainfall-runoff process, most important component of the hydrologic cycle, has a significant place in hydrologic analysis and engineering of water resources. Recently, artificial intelligence methods which are in demand due to bringing successful outcomes to complex problems are preferred in modelling of hydrological events as in many fields. In this study, monthly average streamflow values that belongs to two stations which is inside of the boundary of the Tigris basin, is tried to determine by using precipitation data which has been obtained from meteorological observation stations. Rainfall-runoff relationship was evaluated by setting up for used stations with Adaptive Neuro-Fuzzy Inference System (ANFIS), Artificial Neural Network (ANN), Gene Expression Programming (GEP) and Multiple Linear Regression (MLR). It was seen that GEP performed better performance compared to other methods between observed streamflow and estimated streamflow at one of the used stations. At another station, it was observed that ANFIS is rather successful at predict the streamflow with high sensitive. The outcomes not only corroborate the feasibility of the intelligence methods, but also show the usability of GEP that is a mathematical method to determine the rainfall-runoff relationship.
: February 23, 2021
|APA||Gerger, R , Gümüş, V , Dere, S . (2021). Dicle Havzasının Yağış Akış İlişkisinin Belirlenmesinde Farklı Yapay Zeka Yöntemlerinin Değerlendirilmesi . Bilecik Şeyh Edebali Üniversitesi Fen Bilimleri Dergisi , 8 (1) , 300-311 . DOI: 10.35193/bseufbd.885644|