Estimate the sediment concentration in the Euphrates basin using machine learning methods

Knowing the amount of sediment carried by streams is important for determining the economic life of dams, the design and operation of accumulation reservoirs, as well as determining the erosion characteristics of the basin. Therefore, it is necessary to measure or determine the amount of sediments transported accurately. Artificial Neural Network methods, which have been used frequently in predicting complex events in recent years, were used to estimate the sediment concentration value measured at seven different river stations in the Euphrates basin, one of the most important basins of Turkey. It can be said that the relationship between sediment concentration value and stream flow is quite high in all stations considered. By creating different input combinations, the success of the prediction methods was determined by the coefficient of determination (R2), Mean Absolute Percentage Error (MAPE) and Root Mean Square Error (RMSE) values. As a result of the study, it was determined that among the methods used, the Artificial Neural Networks (ANN) method provided a closer prediction compared to the other methods used in the thesis study. Forecast methods gave the most successful results at the 2115 Göksu River Malpınar station among the stations used. Data obtained from artificial neural networks were compared with observation values obtained from observation stations, and the results showed that machine learning methods could be used in modeling and predicting sediment amounts.