Türkiye'nin nehirlerinde eksik akım verilerinin tamamlanması için çeşitli veri odaklı tekniklerin performans değerlendirmesi

Missing data with gaps is always an obstacle to effective planning and management of water resources. Complete reliable hydrological time series are necessary for the optimal design A study was conducted fill in missing streamflow 54 observation stations across Turkey. This process done aid various statistical estimation methods. Estimations were performed by using Linear regression (LR), Artificial neural network (ANN), adaptive neuro-fuzzy inference system (ANFIS), Support vector machine (SVM), Multivariate Adaptive splines (MARS), K-nearest neighbor (KNN) Performances infilling methods evaluated based on four performance criteria; namely, root mean squared error (RMSE), coefficient determination (R2), absolute (MAE), Kling–Gupta efficiency (KGE) during training test periods. Reliable long from surrounding selected as input output station. The results revealed that a single method cannot be specified best-fit area. During phase, R2 ranged 0.54 0.99, KGE range between 0.62 0.98. showed especially SVM MARS suitable estimating Turkey’s rivers. These findings will provide can used modeling resources management.