EPA Science Inventory

Advanced Decision-Support for Coastal Beach Health: Virtual Beach 3.0

Citation:

Description:

Virtual Beach is a free decision-support system designed to help beach managers and researchers construct, evaluate, and operate site-specific statistical models that can predict levels of fecal indicator bacteria (FIB) based on environmental conditions that are more readily measured, estimated, or forecast. Since its first operational use in 2009, Virtual Beach has been updated several times based on feedback from users in the field. These updates have improved the efficiency of the model-building and evaluation process and made the process of operating a predictive model faster and more user-friendly. Virtual Beach 3.0 includes significant enhancements, including new statistical methods and the integration of hydro-meteorological data accessed through the Web-based Environmental Data Discovery and Transformation (EnDDaT) system. New statistical methods include partial least squares (PLS) regression and gradient boosted modeling (GBM), which provide alternatives to the standard multiple-linear 67 2013regression (MLR) approach, optimized using a genetic algorithm. This poster will illustrate the functionalities of the enhanced software and highlight applications in the Great Lakes.

Purpose/Objective:

Presented at the Annual Conference of the International Association of Great Lakes Research. June 2-6, 2013. West Lafayette, IN

URLs/Downloads:

http://www.iaglr.org/   Exit

Record Details:

Record Type: DOCUMENT (PRESENTATION/POSTER)
Start Date: 08/02/2013
Completion Date: 08/02/2013
Record Last Revised: 12/13/2013
Record Created: 08/02/2013
Record Released: 08/02/2013
OMB Category: Other
Record ID: 258391

Organization:

U.S. ENVIRONMENTAL PROTECTION AGENCY

OFFICE OF RESEARCH AND DEVELOPMENT

NATIONAL EXPOSURE RESEARCH LAB

ECOSYSTEMS RESEARCH DIVISION