Activity Number:
|
208
|
Type:
|
Contributed
|
Date/Time:
|
Tuesday, August 13, 2002 : 10:30 AM to 12:30 PM
|
Sponsor:
|
Section on Statistics & the Environment*
|
Abstract - #300703 |
Title:
|
Multivariate Lattice Models for Spatial Environmental Data
|
Author(s):
|
Stephan Sain*+ and Noel Cressie
|
Affiliation(s):
|
University of Colorado, Denver and Ohio State University
|
Address:
|
P.O. Box 173364, Denver, Colorado, 80217-3364, USA
|
Keywords:
|
Markov random field ; Bayesian hierarchical model ; MCMC ; environmental equity
|
Abstract:
|
In many environmental problems, more than one variable is measured at each spatial location or lattice point. We explore the use of a multivariate Markov random field to build multivariate spatial models and analyze such data. Specific attention is given to building models that incorporate the correlations between variables at different sites. The methodology is demonstrated using a Bayesian hierarchical model to assess environmental equity: The relationship between the racial distribution of residents of Louisiana and the location of sites listed with the Environmental Protection Agency's Toxic Release Inventory is studied.
|