Dam Lake Water Level Estimation Using Different Artificial Intelligence Methods.

Dam reservoir level estimation is important for the construction, operation, design and safety of dams. In this study, dam reservoir level change estimations were investigated using Multi Linear Regression (MLR), M5 Decision Tree (M5 Tree) and Artificial Neural Network (ANN) models. Tuscaloosa Lake in Alabama, USA was selected as the study area. In determining the daily dam reservoir water level , daily stream flow and daily proecipitation height parameters were used. Models were analyzed using graphical and statistical results. In the comparison of prediction models, performance criteria such as corrletion coefficient (R), Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) were used. When the model results were examined; ANN methods showed best performance in predicting the dam reservoir level change.