Identifying potential pollution sources in river basin via water quality reasoning based expert system

Indentifying potential pollution sources from the monitoring water quality data is desirable for water quality management. A simple expert system module was developed in this paper to infer the latent pollution sources and explain variation in water quality parameters. It contains three parts. The first part is water quality mining and analyzing component implemented by multivariate statistical techniques such as cluster analysis and factor analysis. The second part is knowledge base constructed by 53 emission standards for water pollutant of particular industries. The last part is the inference mechanism designed through calculating the similarity between a specific pollution factor and these particular industry emission standards. The Songhua River Harbin Region case study demonstrated the goodness of the expert system module. Three main pollution source factors in low pollution area were mainly related to organic pollution and nutrients (some point industry or animal husbandry and agriculture activities), heavy metal pollution (point sources: Industries) and toxic pollution (point sources: Pharmaceutical industries). The module will help managers make decisions with more strong confidence to improve the water quality.