Forecasting water surface evaporation with artificial neural networks (ANN) model: Example of Ataturk dam

Considering the increasing population, especially with the effect of global warming, the insufficiency of our water resources in the coming years is among the major problems that are likely to be experienced. One of the important issues discussed today is the shrinking of the existing water potential day by day and taking precautions against water scarcity that may occur in the future. In this context, besides how much of the existing water we can use, it is essential for planning to predict how much the water potential will shrink in the future. In this study, daily minimum temperature, maximum temperature, average temperature, solar radiation, relative humidity, wind speed and evaporation data of a station in Atatürk Dam between 2016-2018 were used. The parameters affecting the evaporation are given to the network as input and it is observed how close the network produces results to the actual evaporation values shown as output. For this, different models were tried and the findings were analyzed. As a result, successful results were obtained in evaporation estimation with ANN with 90% accuracy.