Exploratory Analysis of Applicability of Artificial Neural Networks for Prediction of Fecal-Indicator Bacteria Concentrations

An artificial neural network (ANN) will be evaluated to determine if it is capable of predicting fecal-indicator bacteria concentrations in receiving waters during dry and wet weather flows using possible functional relations between hydraulic, hydrologic, and water-quality data collected at five stations during the summers of2001, 2002, and 2003. The proposed study is of exploratory nature. The task is designed as a proof-of-concept demonstration. No written report will be prepared summarizing the results of the project. Progress and results will be discussed during the monthly project meetings conducted at ALCOSAN's office. It is assumed that work will begin on July 19 and conclude September 30, 2004