Abstract River ecosystems are increasingly impacted by human activities such as industrial development and agriculture, leading to shifts in water quality variables and threatening ecological sustainability. Here, we present a modeling approach that integrates reaeration processes with hydraulic parameters to improve prediction of the Water Quality Index in river systems. We developed a genetic programming model to estimate the reaeration coefficient, expressed as a second-order rate constant, and use it as a key predictor of water quality. The model shows that this coefficient can be reliably derived from the Froude number, a dimensionless indicator of flow regime. Through iterative regression, less influential variables are removed, resulting in a simplified equation that maintains high predictive accuracy. Turbidity, temperature, and the reaeration coefficient are identified as primary drivers of water quality. The proposed framework is computationally efficient, cost-effective, and suitable for real-time monitoring across diverse river systems.
Arzhangi et al. (Wed,) studied this question.