|
Activity Number:
|
492
|
|
Type:
|
Invited
|
|
Date/Time:
|
Thursday, August 10, 2006 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Statistics and the Environment
|
| Abstract - #305280 |
|
Title:
|
Ensemble Smoothing for Understanding Geophysical Processes
|
|
Author(s):
|
Douglas W. Nychka*+
|
|
Companies:
|
National Center for Atmospheric Research
|
|
Address:
|
P.O. Box 3000, Boulder , CO, 80307,
|
|
Keywords:
|
data assimilation ; Kalman filter ; spatial statistics
|
|
Abstract:
|
Much of the understanding and prediction of geophysical processes comes from the construction and use of large numerical models that simulate natural phenomenon. A key area for statistical science is when these models are confronted with data. Part of the challenge is computational, and straight forward maximum-likelihood or Bayesian approaches are not feasible. The Ensemble Kalman Filter (EKF) is a Monte Carlo--based approximation to a Gaussian-Bayesian posterior that can handle high-dimensional state vectors and processes that evolve nonlinearly over time. Examples will be given using the EKF in the Data Assimilation Research Testbed and NCAR community atmospheric model to tune model parameters and estimate the state based on sparse observations.
|
