Data-driven Modeling for Water Quality Parameters Prediction of the Drainage System Associated with Lake Manzala, Egypt

Lake Manzala is the largest of the Egyptian lakes along the Mediterranean coast and the most productive for fisheries. It is located on the Eastern margin of the Nile Delta between the Damietta Branch of the River Nile and Port Said on the Suez Canal. Two main drains namely, Bahr El-Baker Drain system and Bahr Hadous Drain system flow into Manzala Lake and drain a considerable volume of drainage water into the lake. Water body of Lake Manzal is affected qualitatively and quantitatively by drainage water that flow into the lake. In most cases of water quality modeling, mathematical models is a helpful tool that might be used to predict the parameters of water quality. Deterministic models are commonly used to simulate the system behavior. Most of these models can only be applied to simplified cases or to situations where the models are strictly calibrated and validated, with no adequate accuracy when applied to unrestricted conditions. However most ecological systems are so complex and unstable. Data-driven models are computing methods that are capable of extracting different system states without using complex relationships. The main objective of the present study is to exploit the capabilities of the adaptive neuro-fuzzy inference system (ANFIS) for drains system quality predictions with emphasis on total phosphorus (TP) and total nitrogen (TN). Discharge, pH, total suspended solids (TSS), electrical conductivity (EC), total dissolved solids (TDS), Temperature, dissolved oxygen (DO) and turbidity (FTU) were the input parameters of the ANFIS models. The models were calibrated and validated against the measured drainage water data for the period from year 2001 to 2010. Predicted and observed values of water quality parameters were evaluated using several common evaluation criteria. Results of model performance showed that the proposed ANFIS models were capable of simulating the water quality parameters and provided reliable prediction of TP and TN. Results showed also that the proposed models could be a good tool for Onsite Water Quality evaluation.