Dam Water Levels Prediction Using Different Time Series Models for Yusufeli and Deriner Dams

In this study, water level prediction was conducted for the strategically important Yusufeli and Deriner Dams, located on the Çoruh River in Turkey. With a height of 275 meters, Yusufeli Dam is the tallest dam in Turkey and plays a crucial role in energy production and flood control. Similarly, Deriner Dam, standing at 249 meters, is the second tallest dam in the country and significantly contributes to the region’s hydroelectric energy capacity. Accurate water level prediction is critical for ensuring the sustainability of hydroelectric power generation, effective flood management, and efficient water resource utilization. The study employed time series analysis techniques, utilizing ARIMA, SARIMA, ARIMAX, and SARIMAX models to forecast water levels, and their predictive performances were compared. The analysis for Deriner Dam demonstrated that the SARIMAX model achieved the highest accuracy with an R² score of 99.05%. Conversely, for Yusufeli Dam, the ARIMA model exhibited the best performance with an R² score of 99.32%. Furthermore, the study compared the predictive performance of the employed models with similar studies in the literature, revealing that the proposed models offer high accuracy. The findings indicate that time series models provide reliable and effective results in hydrological forecasting. For future research, it is recommended to test model performance in different hydrological systems and assess the long-term impacts of climate change on water level predictions.