The study aims to determine the location with optimal run-of-river hydroelectric power plants (RoRHPP) production potential by evaluating meteorological data from different locations with the Analytic Hierarchy Process (AHP), T echnique for O rder P reference by S imilarity to I deal S olution (TOPSIS), and VlseKriterijuska Optimizacija I Komoromisno Resenje (VIKOR) methods in annual and seasonal periods. Rize, Giresun, and Trabzon provinces were determined as the application area. Hourly, real-time meteorological measurement data for 2022 and RoRHPP production data for these provinces were obtained. The study consists of four stages. First, missing hours in meteorological data sets are identified and filled using Seasonal Autoregressive Integrated Moving Average (SARIMA). Secondly, the target plants to be used in the provinces are determined with XGBoost regression analysis to compare the results of the multi-criteria decision-making method with real data. Third, alternative scores are created by interpolating meteorological data using AHP, TOPSIS, and VIKOR methods for annual and seasonal levels. Lastly, the scores out of AHP, TOPSIS, and VIKOR are measured against productions obtained by Extreme Gradient Boosting (XGBoost) regression analysis, RoRHPP. The solutions of AHP-TOPSIS-VIKOR are consistent with the province's average production of RoRHPP, and verify that it is also credible to determine the high-potential RoRHPP areas by applying meteorological rules into the MCDM base model with a combination of machine learning based validation. Based on these findings, the best and worst RoRHPP production potentials are in Rize and Trabzon, respectively. The proposed approach offers an interpretable and data-driven tool for hydropower planning. • Integration of real-time meteorological and RoRHPP production data into MCDM. • Determination of the best RoRHPP energy production location using MCDM methods • Target plant selection with XGBoost regression for MCDM ranking comparisons.
İsrafil Karadöl (Sun,) studied this question.