|
Activity Number:
|
266
|
|
Type:
|
Invited
|
|
Date/Time:
|
Tuesday, July 31, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
WNAR
|
| Abstract - #308060 |
|
Title:
|
Bayesian Spatial Modeling of Vegetation Coverage Data in the Missouri Ozark Forest Ecosystem Project
|
|
Author(s):
|
Zhuoqiong (Chong) He*+ and Xiaoqian Sun
|
|
Companies:
|
University of Missouri-Columbia and Clemson University
|
|
Address:
|
146 Middlebush Hall, Columbia, MO, 65210,
|
|
Keywords:
|
Forest Ecosystem ; Spatial Models ; Markov Chain Monte carlo ; Bayesian
|
|
Abstract:
|
This talk presents a new spatial model that takes account of the special data structure and treats a cluster of measurements as repeated measurements in one location. The model is applied to the analysis of the total vegetation coverage data in the Missouri Ozark Forest Ecosystem Project. An MCMC algorithm based on the shrinkage slice sampler is developed. The results show that the soil depth is an important factor while the aspect class is less important when modeling the total vegetation coverage. In addition, the strong spatial effect does exist in the data and four measurements in quadrats of a subplot are not strongly correlated but are not independent. Prediction of the total vegetation coverage at unmeasured locations is developed. Finally, possible generalizations are discussed.
|
