The use of QB (Minimum Flow) and ANN (Artificial Neural Network) methods on semi-arid area for determining the environmental flows

Environmental flow as defined in the literature implies the Minimum Level of flow either to achieve the aspired or required level of low in a river bed. The subject of this thesis is to assess the use of Environmental Flow to maintain a healthy river ecosystem along with rehabilitation of habitat for endangered species. Environmental Flows have been calculated with the data available from flow monitoring stations located in Konya Closed Basin using Tennant Method to calculate 7Q10, Q95, minimum flow and artificial neural networks. In total eleven flow monitoring stations have been studied and long-term data sets of these current monitoring stations have been used. For each flow monitoring station, the results obtained from Tennant, 7Q10, Q95, Qb methods were evaluated and artificial neural network method was used to tolerate different environmental flow values resulting from the methods and to obtain healthier results. 10% of the average flow accepted by the State Hydraulic Works as the safety limit and 20% of the median flow used for this study were compared with the ANN results. Since ANN is obtained by blending all methods, the low environmental flows in other methods have led to the calculation of low environmental flows in ANN methods. In stations that do not have current data or have very low and irregular currents, it is seen that the ANN method does not provide 10% of the average current recommended by DSI. In flow monitoring stations with regular currents, it provides the most appropriate results which are thought to be necessary on the basis of the water mass and to ensure the minimum flow. In the Qb method, drought or low flow in two to three months during the year weakens the applicability of this method and it is observed that healthy results cannot be obtained. The most important outcome of this study is that the environmental flow assessment for each station is evaluated and presented as basin or even water body and the ANN method yields high performance results. Keywords: Environmental Flow, Konya Closed Basin, Tennant Method, Artificial neural networks, Qb (minimum flow)