2003 Progress Report: Bayesian Methods for Regional-Scale Stressor-Response Models
EPA Grant Number: R830887Title: Bayesian Methods for Regional-Scale Stressor-Response Models
Investigators: Lamon, E. Conrad , Stow, Craig A.
Institution: Louisiana State University - Baton Rouge , University of South Carolina at Columbia
EPA Project Officer: Sergeant, Anne
Project Period: May 14, 2003 through April 13, 2006 (Extended to April 13, 2008)
Project Period Covered by this Report: May 14, 2003 through April 13, 2004
Project Amount: $389,168
RFA: Developing Regional-Scale Stressor-Response Models for Use in Environmental Decision-making (2002)
Research Category: Ecological Indicators/Assessment/Restoration
Description:
Objective:The objective of this research project is to use modern classification and regression trees and other Bayesian hierarchical techniques to link multiple environmental stressors to biological responses and quantify uncertainty in model predictions.
Progress Summary:In Lamon and Stow (2004), we demonstrate a Bayesian Classification and Regression Tree (CART) approach to link multiple environmental stressors to biological responses and quantify uncertainty in model predictions. This approach can: (1) report prediction uncertainty; (2) be consistent with the amount of data available; and (3) be flexible enough to permit updates and improvements. Tree-based methods are a flexible approach to variable subset selection and when the analyst suspects global nonlinearity and cannot (or does not want to) specify the functional form of possible interactions a priori. We use the U.S. Environmental Protection Agency National Eutrophication Survey data to fit three models demonstrating the methods and to highlight important differences arising from slightly different model specifications. The Bayesian approach offers many advantages from a decision-theoretic point of view, including the estimation of the value of new information and proper probability distributions on the variable of interest as an output, which can be directly used in risk assessment or decision making.
Composite data at national, regional, and local scales will be used to identify and estimate regional stressor response models and link a large, rich suite of environmental, chemical, physical, hydrologic, and watershed characteristics that can be used as predictor variables for chlorophyll a. We have obtained approximately 656,000 water quality observations from the Nutrient Criteria Database for lakes and reservoirs, of which approximately 98,000 observations had non missing values for chlorophyll, nitrogen, and phosphorus (Freeman, 2004). Model estimation will be accomplished using a training set of data (n = 48,916) randomly chosen from the full dataset assembled for this study. A validation, or test, dataset (n = 49,251) will be withheld from use in model estimation and used to evaluate out-of-sample predictive performance. Mean squared error, median absolute deviation, and cumulative log likelihood will be evaluated and reported for all models using both the training and test data.
A systematic method for the identification and estimation of regional scale stressor-response models in aquatic ecosystems will be useful in the monitoring and assessment of aquatic resources, determination of Total Maximum Daily Loads (TMDLs), and an increased understanding of the differences between regions. Use of the decision-theoretic framework to guide the modeling process will assure proper use of attributes that measure ecosystem value and will avoid the information loss associated with aggregation of attributes into indices.
Future Activities:In the next year, we will expand our list of potential predictor variables.
Journal Articles on this Report: 1 Displayed | Download in RIS Format
Other project views: | All 5 publications | 1 publications in selected types | All 1 journal articles |
Type | Citation | ||
---|---|---|---|
|
Lamon EC III, Stow CA. Bayesian methods for regional-scale eutrophication models. Water Research 2004;38(11):2764-2774. |
R830887 (2003) |
![]() ![]() |
water, lakes, reservoirs, regionalization, scaling, decision-making, Bayesian, limnology, modeling, adaptive implementation modeling, aquatic ecosystem, assessment endpoint mechanistic research, decision analysis, decision support tool, ecological indicators, ecological models, ecological variation, ecology, ecology assessment models, ecosystem assessment, ecosystem modeling, ecosystem stress, environmental decision-making, environmental risk assessment, regional scale impacts, risk assessment, stress response, water monitoring, watershed. , Ecosystem Protection/Environmental Exposure & Risk, Economic, Social, & Behavioral Science Research Program, Water, Scientific Discipline, RFA, Water & Watershed, Social Science, decision-making, Economics & Decision Making, Watersheds, Regional/Scaling, Monitoring/Modeling, Ecology and Ecosystems, risk assessment, water quality, aquatic ecosystem, environmental risk assessment, decision support tool, watershed, ecosystem modeling, decision analysis, ecology assessment models, decision making, regional scale impacts, Bayesian classifiers, TMDL, stress response, ecology, ecosystem assessment, ecological variation, ecosystem stress, watershed assessment, ecological models, water monitoring, assessment endpoint mechanistic research, Bayesian approach, environmental decision making
Relevant Websites:
Progress and Final Reports:
Original Abstract