A Bayesian Probability Network Approach To Predictive Modeling in Support of Environmental Decision Making

EPA Grant Number: U915590
Title: A Bayesian Probability Network Approach To Predictive Modeling in Support of Environmental Decision Making
Investigators: Borsuk, Mark E.
Institution: Duke University
EPA Project Officer: Lee, Sonja
Project Period: August 1, 1999 through August 1, 2002
Project Amount: $102,000
RFA: STAR Graduate Fellowships (1999) RFA Text |  Recipients Lists
Research Category: Academic Fellowships , Environmental Justice , Fellowship - Environmental


The objective of this research project is to demonstrate that probability network models represent an improved approach to predictive modeling used for environmental management.


A probability network model is being developed and applied to the problem of eutrophication in the Neuse River, North Carolina. Also called a Bayesian probability network, or a ?Bayes net,? this model consists of the set of variables of interest in the system being modeled as well as a set of assertions concerning the probabilistic relationships among the variables. These relationships are quantified using historical data, models, and expert judgment. Probabilistic predictions of model endpoints are then made that are based on the entire set of conditional probabilities that have been assessed for each system variable. Not only does this network structure provide a more integrated approach to uncertainty analysis, but it also allows easy updating of prediction and inference when observations of model variables are made. This capability is particularly important when applied to a natural system in which additional monitoring is likely to occur concurrent with the modeling effort.

Expected Results:

This study will determine if probability network models are more effective than predictive modeling for environmental management. The probability network method could remedy many of the traditional shortcomings of science used for decisionmaking.

Supplemental Keywords:

integrated modeling, probabilistic prediction, stakeholder involvement, publicly meaningful endpoints, water quality modeling, Bayesian inference., RFA, Economic, Social, & Behavioral Science Research Program, Scientific Discipline, decision-making, Environmental Statistics, Social Science, Economics & Decision Making, Bayesian approach, decision analysis, environmental decision making, decision making, Bayesian method, integrated modeling, probability network models, stakeholder

Progress and Final Reports:

  • 2000
  • 2001
  • Final