|
Activity Number:
|
235
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Tuesday, August 8, 2006 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Statistics and the Environment
|
| Abstract - #306725 |
|
Title:
|
Models for Multivariate Spatial Lattice Data and Assessing Climate Change
|
|
Author(s):
|
Stephan Sain*+
|
|
Companies:
|
University of Colorado at Denver and Health Sciences Center
|
|
Address:
|
P.O. Box 173364, Denver, CO, 80217-3364,
|
|
Keywords:
|
Markov random field ; conditional autoregressive model ; hierarchical model
|
|
Abstract:
|
The U.S. Climate Change Science Program Strategic Plan has recognized the need for regional climate modeling to assess climate impacts. This is the focus of the newly formed North American Regional Climate Change Assessment Program that seeks to study a number of regional climate models. In this talk, we will discuss the details of the specification of a spatial hierarchical model based on a multivariate Markov random field (also referred to as a conditional autoregressive, or CAR, model) and take a close look at the output of a particular regional climate model that uses a "business as usual" scenario to model climate over the western United States. In particular, we seek to explore how the modeled climate---seasonal measures of temperature and precipitation in particular---changes over the next 50 years.
|
