About This Item

DEVELOPING MATHEMATICAL MODELS FOR PREDICTING Escherichia coli CONCENTRATIONS AT WISCONSIN BEACHES USING REGRESSIONS

Show full item record

File(s):

Author(s)
Yerram, Sashidhar R.
Advisor(s)
McDermott, Colleen
Degree
MS, Biology
Date
Aug 2009
Subject(s)
Water pollution; Regression analysis; Lakes monitoring; Escherichia coli
Abstract
The traditional methods to quantitatively determine Escherichia coli concentrations in beach water are the membrane filtration and defined substrate methods. These methods take 18-24 hours for enumeration of E. coli concentrations. This may result in improper beach closures and openings, as authorities base their decisions on previous day E. coli concentrations. To overcome these problems mathematical models were developed, using the data collected from the 2007 swimming season, to predict the E. coli concentrations using various explanatory variables. Beaches from Douglas, Ashland, Bayfield and Door Counties, Wisconsin were chosen for developing the mathematical models. Mathematical models were developed using the United States Environmental Protection Agency (USEPA) ?Virtual Beach? software, an application that uses multiple linear regressions. The developed mathematical models were tested with model fitting (i.e. model is fitted with E. coli concentrations used for developing the predictive model). Mathematical models for six beaches were able to predict the loge E. coli concentrations with 100% accuracy. Explanatory variables that were included in the predictive models were unique for each beach. Overall it was concluded that predictive models should be beach specific. In future more robust mathematical models should be developed using larger data sets and mathematical models should be tested with real time E. coli concentrations.
Description
A Thesis Submitted In Partial Fulfillment of the Requirements For the Degree of Master of Science-Biology
Permanent link
http://digital.library.wisc.edu/1793/46819 
Export
Export to RefWorks 

Part of

Show full item record



Advanced Search

Browse

Deposit materials

About MINDS@UW