Modeling of re-aeration coefficient For Nkisa River, Egbema, Rivers State, Nigeria

Abstract

This study aimed at developing a model for re-aeration (K2) that may be utilized in conjunction with de-oxygenation (K1) to calculate the self-purification capacity, f, of Nkisa River. An iterative least squares fitting routine was employed by XLSTAT to produce the optimal goodness of fit between Nkisa data and function. Thus, Nkisa re-aeration models; K22DRY (dry season model), K22RAINY (rainy season model) and K22all-seasons (combined model) were developed and validated statistically resulting in the output with the optimal statistical performance (SSE = 0.0005; R2 = 0.948; Adjusted R2 = 0.937 and RMSE = 0.009), (SSE =0.0483, R2 = 0.81, Adjusted R2 = 0.77, and RMSE = 0.08) and (SSE = 0.1004; R2 = 0.93; Adjusted R2 = 0.92 and RMSE = 0.08) for dry, rainy and all-seasons respectively. Its performance was verified by comparing the model with nine previously existing models and Nkisa re-aeration models gave the least SSE values of 0.0005, 0.0483 and 0.1004; which was the best interpretation of the conditions of Nkisa River in dry, rainy and all-seasons. The self-purification capacity of the studied river was found to be greater than unity during all studied period except in March which implies that, re-aeration is greater than de-oxygenation. HIGHLIGHTS Raw water samples were collected and field measurement of hydraulic and physical parameters of the Nkisa River was done.; Laboratory analyses of the collected raw water samples were done.; Modeling and validating the re-aeration coefficient (K2) for the Nkisa River were carried out.; The models' suitability was verified by comparing them with nine other existing models.; The performance of different K2 models was compared with the following statistics: SSE, R2, adjusted R2, and RMSE.;

Keywords

• dissolved oxygen
• model
• nkisa river
• non-linear regression
• re-aeration coefficient
• standard error