Authors: Francesco Granata, Stefano Papirio, Giovanni Esposito, Rudy Gargano and Giovanni de Marinis
 
Journal Title: Water
 
Publisher: MDPI AG
 
Abstract
 
Stormwater runoff is often contaminated by human activities. Stormwater discharge into  water bodies significantly contributes to environmental pollution. The choice of suitable treatment  technologies is dependent on the pollutant concentrations. Wastewater quality indicators such as  biochemical oxygen demand (BOD5), chemical oxygen demand (COD), total suspended solids (TSS),  and total dissolved solids (TDS) give a measure of the main pollutants.
 
The aim of this study is to  provide an indirect methodology for the estimation of the main wastewater quality indicators, based  on some characteristics of the drainage basin. The catchment is seen as a black box: the physical  processes of accumulation, washing, and transport of pollutants are not mathematically described.  Two models deriving from studies on artificial intelligence have been used in this research: Support  Vector Regression (SVR) and Regression Trees (RT). Both the models showed robustness, reliability,  and high generalization capability.
 
However, with reference to coefficient of determination R2 and  root‐mean square error, Support Vector Regression showed a better performance than Regression  Tree in predicting TSS, TDS, and COD. As regards BOD5, the two models showed a comparable  performance. Therefore, the considered machine learning algorithms may be useful for providing  an estimation of the values to be considered for the sizing of the treatment units in absence of direct  measures.
 

Illustration Photo: Stormwater Outfall Into the Hudson River in Albany, New York, United States (credits: Andy Arthur / Flickr CC BY 2.0)

Comments

No comments to display.