Feasibility Assessment of Data-Driven Models in Predicting Pollution Trends of Omerli Lake, Turkey

Abstract

Data-driven models are commonly used in a wide range of disciplines, including environmental engineering. To analyze Omerli Lake’s historic water pollution status, this study monitors data for dissolved oxygen, 5-day biochemical oxygen demand, ammonium nitrogen, nitrite nitrogen, nitrate nitrogen, and ortho phosphate. The quality of the lake water is assessed based on measurements of dissolved oxygen. The collected data are analyzed using regression analysis and artificial neural network models. The main goal of this paper is to reveal the best applicable data-driven model in order to gain forward-looking information regarding the dissolved oxygen level of the lake using other pollution parameters. In order to ascertain eutrophic status, total phosphorus loads for each year are represented on a Vollenweider diagram. Results designate an increasing risk of eutrophication for Omerli Lake in recent years. Results of the data-driven models show that the artificial neural networks model constitutes the best relationship between the dissolved oxygen and other parameters.