|
Activity Number:
|
37
|
|
Type:
|
Invited
|
|
Date/Time:
|
Sunday, August 3, 2008 : 4:00 PM to 5:50 PM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
| Abstract - #300378 |
|
Title:
|
Combining Computer Model Output and Observational Data: Forecasting Climatic and Weather-Related Processes
|
|
Author(s):
|
Mark Berliner*+
|
|
Companies:
|
The Ohio State University
|
|
Address:
|
Department of Statistics, Columbus, OH, 43210,
|
|
Keywords:
|
Climate change ; Hierarchical Bayesian analysis ; Experimental design
|
|
Abstract:
|
Modern science relies on both computer modeling and continuing improvements in observational assets. However, both of these information sources are subject to error. Hence, combining these sources in a fashion that reflects their uncertainties is critical. The approach suggested is a hierarchical Bayesian formulation involving the statistical modeling of computer model output. The approach also enables direct formulation of experimental design for combining computer models and observations. The main notions are illustrated in an extended example involving climate and climate change analyses.
|
