Graphical Models for Spatial Data: Interpretations and Estimation

• Activity Number: 428
• Type: Contributed
• Date/Time: Tuesday, August 3, 2010 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics and the Environment
• Author(s): Alix I. Gitelman*+ and Kathryn Mary Irvine and Xuan Che
• Companies: Oregon State University and Montana State University and Oregon State University
• Keywords: Graphical models ; multivariate spatial models ; ecological processes

Abstract

Graphical models provide an important tool for facilitating communication between scientists, decision-makers, and statisticians. Indeed, many complicated ecological processes can be described in terms of ``box-and-arrow'' conceptual diagrams (e.g., Shipley, 2000; Clark and Gelfand, 2006). In this work we examine graphical models for spatial data in an effort to catalog how (or whether) statistical independences codified by the graphical models translate into meaningful ecological processes. Having drawn a box-and-arrow diagram, an important question becomes---how does one estimate it? We make connections between some specific graphical models and existing parametric models for multivariate spatial data, and we raise questions about how those connections might be made for more complicated graphical models.