Development of a Data Driven Model for EstimatingPollutant Levels

Pollutants accumulated on roads during dry periods are washed off the surface during rainfall events, presenting a potentially hazardous non-point pollution source. Estimation of pollutant loads in these runoff waters is required for developing mitigation strategies, yet the numerous factors involved and their complex interconnected influences make straightforward assessment impossible. Data driven models are especially suitable for this task. The book starts with a review of pollutants in highway runoff and their environmental effects, followed by an account of the sources of these pollutants and the factors affecting their concentrations. The second part describes the development and implementation of a coupled MT-GA (Model Tree - Genetic Algorithm) model for the prediction of event mean concentrations of selected highway runoff pollutants. At first MT and GA approaches are described and the development of the coupled MT-GA model is portrayed. Finally the model's performance is evaluated and compared with literature results and sensitivity analysis is performed to examine the relative significance of the influencing factors and the model’s response to changes in input values.