Karlı Dağlık Bir Havzada Günlük ve Aylık Akım Değerlerinin SWAT Modeliyle Değerlendirilmesi: Bitlis Çayı Havzası Örneği

Abstract

In this study, the performance of a SWAT model constructed with daily weather observations of regional meteorological stations and coarse-scale topography, land cover, and soil data in simulating and forecasting daily and monthly streamflow rates in a snowy mountainous watershed is assessed with the example of Bitlis Creek. The SWAT model developed to simulate the flow rates of Bitlis Creek, located in the Euphrates-Tigris Basin, at the location of the Baykan stream-gauging station is calibrated and validated separately for daily and monthly flow simulations by using the records of the Baykan station. Simulation statistics that can be judged as good for daily flow rates and very good for monthly flow rates are obtained in both of the calibration and validation periods. When it is examined how accurately the developed model can predict the daily and monthly flows using only meteorological data without flow measurements, it is observed that the model can provide satisfactory forecast performances in the context of both statistical evaluation indices and flow-duration relations. Although a strong correlation between the daily flow estimates and observed data in terms of annual maximum time series cannot be obtained, it is observed that this situation does not result in significant differences in terms of the calculated recurrence flood peaks. The results indicate that the developed model can enable the investigations of possible changes in the flow regime of Bitlis Creek in the coming decades on daily and monthly time bases.