Forecasting water quality parameters in Wadi El Rayan Upper Lake, Fayoum, Egypt using adaptive neuro-fuzzy inference system

Abstract

Wadi El-Rayan depression, located southwest of Cairo, has two artificial lakes; the upper (first) and lower (second) lakes that are used as reservoirs for the drainage of agricultural wastewater through El-Wadi Drain. Wadi El-Rayan Upper Lake, which is connected directly to El-Wadi Drain, is considered among the most important water-resource lakes in Egypt. In this study, the water quality parameters, such as chemical oxygen demand, biochemical oxygen demand, ammonia, and nitrate of the lake in 2030 is predicted. Hybrid artificial intelligence techniques are considered effective procedures for defining optimal solutions for many problems. To forecast the water-quality parameters of Wadi El-Rayan Upper Lake, various adaptive neuro-fuzzy inference system (ANFIS) models were implemented herein using MATLAB SIMULINK and MATLAB ANFIS GUI. Most of the measurement data were used in the training processes and the rest were used for model testing and validation. The developed models were compared to determine the most accurate one, which was in turn used for predicting the physicochemical parameters of the Upper Lake in 2030.