|
Activity Number:
|
137
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 4, 2008 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Statistics and the Environment
|
| Abstract - #302121 |
|
Title:
|
Estimation with Extended CAR Models for Large Lattices
|
|
Author(s):
|
Ernst Linder*+
|
|
Companies:
|
University of New Hampshire
|
|
Address:
|
Dept. of Math. & Statistics, Durham, NH, 03824,
|
|
Keywords:
|
Gaussian random fields ; spatio-temporal analysis ; large data ; approximate inference ; regional climate models
|
|
Abstract:
|
We propose an extension of the usual CAR model for spatial lattice data. The extension incorporates a second spatial parameter that governs smoothness of the underlying spatial field. This model is defined via the spectral decomposition of the precision matrix of the data. For regular rectangular lattices, assuming a circulant structure, the spectral decomposition is the Fast Fourier transform (FFT) and thus provides a computationally feasible method for very large data. We consider strategies to applying this model for a variety of space - time regression and functional ANOVA type models. We provide an implementation within a hierarchical Bayesian estimation framework applied to regional climate model output data over a large grid covering North America.
|
