Evaluation of Ergene River water quality and determination of chemical oxygen demand with artificial neural networks

Keywords: Ergene River, Pollution, Water quality, Climate, Precipitation, Artificial Neural Networks. Ergene Basin is located in the middle of Trakya region. It is one of the 13 sedimentary basins in our country which has a very important place as a location. The Ergene River, which was used as drinking water for almost half a century ago, poses great threat not only for itself but also for its environment. The pollution created by the industrial organizations developing in the surrounding area has negatively affected the agricultural land in the fertile soil of Ergene Basin and therefore human health. In this study, in order to determine water quality classifications monthly data, between years 1999-2014 belonging to İnanlı Station and Lüleburgaz measurement stations of The General Directorate of State Hydraulic Works on Ergene River, which are Biological Oxygen Demand (BOD₅), Calcium (Ca₂⁺), Magnesium (Mg₂⁺), Sodium (Na⁺), Chlorine (CI⁻), Chemical Oxygen Demand (COD), Nitrite Nitrogen (NO₂-N), Nitrate Nitrogen (NO₃-N), Phosphate (PO₄), Sulfate (SO₄), Hydrogen power (pH), Temperature (T), Alkalinity (m-Al), Total Dissolved Solids (TDS), Ammonium Nitrogen (NH₄-N), Electrical Conductivity (EC), Dissolved Oxygen (DO), Potassium Permanganate (pV), Total Hardness (TH) , Suspended Solid (SS) parameters were examined and water quality class was determined. In addition, for COD estimation BOD₅, Ca₂⁺, Mg₂⁺, Na⁺, CI⁻, NO₂-N, NO₃-N, PO₄, SO₄, pH, T, M-Al, TDS, NH₄-N, EC, DO, pV, SS, TH parameters were analyzed by ANN and as a result of the analysis, as variable group of when R² value is 0.989, M-Al (d1), TDS (d2), pV (d3), NO₄-N (d4), DO (d6) are identified. In the literature, no studies have been found to estimate COD by artificial neural networks using the water quality parameters of Ergene River and our study is also important in this respect.