Determination of the effect of climate change on the occupancy rates of dams used for drinking water supply using artificial neural network modeling

For the key concepts in the thesis to be in the form of a pattern from general to specific, leadership must first be examined. The study includes climate change, dams used for drinking water purposes and artificial neural network packages. In this doctoral thesis, in order to offer a solution to the problem of climate change, the method of revealing the damage caused by the problem is first applied. This method reveals that the rainfall flow to the reservoir of dams used for drinking water is different. For this, a new system, artificial neural network, is used. Dams are reservoirs built to store water using surface or domestic water resources to meet water needs. In addition, dams are one of the important structures of water supply systems. Dams ensure that the maximum amount of water is filled during periods of abundant rainfall and the required flow rate during dry periods. In this doctoral thesis, the climate climate has an impact on the occupancy rates of dams used for drinking water supply in Ankara, Istanbul and Izmir. Research on the human brain has become an example for completing the information processing process. Programs that can create new data and results as a result of modeling the nervous system created by the human brain by using biological neural records or unlearned data are called artificial neural networks. As a structure, it is aimed to establish a connection between the model times of biological neural networks and artificial neural networks. In artificial neural networks, there are process elements called nervous system, which consist of 5 basic elements: weights, activation functions, outputs, inputs and summation functions. Implementation of artificial neural networks in the computer environment can be done in different ways. What needs to be done initially is to select the artificial neural network model appropriate to the problem. Technology selection is as important a criterion as determining the appropriate model. Software can be created to solve the problem, or ready-made artificial intelligence libraries can be used. The widely used library is the Matrix Laboratory application. It expands the rainfall area of the dam, which is used primarily for drinking water purposes, for this amount. Then, meteorological stations in the rainfall area are determined and the dams used for drinking water are entered and the distribution of their output data is determined. These data were investigated by modeling artificial neural networks. At the end of the findings section, in the regression analysis performed separately for the dams used for drinking water purposes, a difference emerged between the expected rainfall to ensure R2 rights and the actual rainfall.