Evaluation of İznik Lake water quality parameters by artificial neural networks method summary

It is important to water quality parameters in the planning and management of lakes. In this study, the measurement results of water quality parameters were obtained from Karsak Stream, including 4 inlet Sölöz, Olukdere, Karasu, Kırandere and 1 outlet in Lake Iznik for the years 2017-2021. pH, conductivity (EC), temperature (T), chemical oxygen demand (COD), biological oxygen demand (BOD), dissolved oxygen (DO), total phosphorus (TP), phosphate (PO4-P), total nitrogen of the entry points (TN), nitrate nitrogen (NO3N) and nitrite nitrogen (NO2N) parameters were taken and their effects on TN, TP and DO at the exit point were investigated. Various attempts were made to predict TN, TP and DO, and an artificial neural network (ANN) method was adopted for model studies. IBM SPSS statistical 23 software was used as the model. The sum of squares of error (SSE) and coefficient of determination (R2) were used to evaluate the amount of error in the performance evaluation of TN, TP and DO values. The results showed that the coefficient of determination was mostly close to 1, and the ANN analysis showed that the pollution estimation of the total nitrogen, total phosphorus and dissolved oxygen parameters could be realized. Thus, it has been seen that ANN models are an estimation tool that can be used effectively to describe the suitability of lake water quality. The developed models can be used for prediction in the future to support the decision process for pollution reduction through sustainable environmental management.