Forecasting of Dam Lake Water Level Using M5 Decision Tree and Anfis Models

Abstract

Dam reservoir level prediction is important for dam construction, operation, design and safety. In this study, dam reservoir level change predictions were investigated using the M5 Decision Tree (M5 Tree) and Adaptive Neural Fuzzy Inference System (ANFIS) models. For modeling the daily dam reservoir water level (t), the lagged time of reservoir water level (t-1), stream flow (t) and precipitation heights in the dam basin (t) were used. The model results were compared with the results of conventional multiple linear regression (MLR) models. The models were analyzed with graphical and statistical results. The coefficient of determination (R2), root mean square error (RMSE) and mean absolute error (MAE) performance criteria were taken into account when comparing the prediction models. The results showed that M5 Tree and Anfis model results gave a better performance in predicting the dam reservoir level change.