|
Activity Number:
|
271
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Tuesday, July 31, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
| Abstract - #308559 |
|
Title:
|
How Far Can a Physicist Travel Along the Bayesian Way: Probability, Uncertainty, and Adequacy in Climate-Like Modeling
|
|
Author(s):
|
Leonard Smith*+
|
|
Companies:
|
London School of Economics
|
|
Address:
|
Houghton Street, London, WC2A 2AE, United Kingdom
|
|
Keywords:
|
model inadequacy ; physical modelling ; decision support ; nonlinear
|
|
Abstract:
|
The roles of probability and uncertainty in physics based simulation modeling are considered within two hierarchies of physical models. It is argued that Bayesian notions are relevant only when state-of-the-art models are arguably empirically adequate. Given a model class no member of which accurately reflects the dynamics of the system under current conditions, how might one develop rational probabilistic beliefs regarding how the system would behave under altered conditions? What role does the probability calculus (including Bayes' Theorem) play in this situation? A slightly lower target than empirical adequacy, Probabilistic Similarity, is introduced. The current state of climate modeling, the rationality of many climate modelers, and the relevance of probabilistic analysis of current climate model output for decision support are each considered in this context.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2007 program |
