Comparison of conceptual monthly rainfall-runoff models

Conceptual rainfall-runoff models are one of the most widely used methods in the construction of hydrological enterprises today. Considering factors such as climatic changes and permanent water consumption, rainfall-runoff models have a great place in the preliminary studies and reconstructions of the reservoirs. Due to the uncertainties of single models, it is thought that the use of multiple models will give better results. The prepared study was carried out by using 8 different conceptual monthly rainfall-runoff models of the Çağlayan and Gördes Dams which belong to the Gediz Basin, which is not ready for use yet. These models include Abcde, Awbm-6p, Dynwb, Gr2m, Guo-3p, Guo-5p, Temez and Twwb. Multi models were performed with the Bayesian Model Averaging (BMA) method. The ensemble models in the versions created for both basins with many different combinations were compared with the single models in terms of performance criteria. As a result of this study, it was seen that the Gr2m gave the best performance among the individual models, while the ensemble version of the triple combination models in ensemble models gave the best performance. It was observed that multi-model combinations have performed more efficiently than single models have done. However, it was concluded that the determination of the single model with the best performance of BMO is open to improvement. It was aimed to adapt and use the obtained data to different scenarios in the operation of dams and under climatic changes.