in Climate Prediction
Charles Jackson (1)
Mrinal Sen (1)
Gabriel Huerta (2)
Yi Deng (1)
Ken Bowman (3)
(1) Institute for Geophysics, The University of Texas at Austin
(2) Department of Mathematics and Statistics, University of New
(3) Department of Atmospheric Science, Texas A&M University
Surface air temperature
(AchutaRao et al., 2004)
Where can clouds go wrong?
Are current approaches to climate
model development convergent?
Address question using:
• Bayesian inference
• Stochastic sampling
– Simulated annealing to focus sampling
– Multiple search attempts for uncertainties
Posterior probability density for
Target: Match observed climate
One 11-year climate model
integration takes 11 hours over 64
processors of an Intel-based
• Analysis of top six performing model
• T42 CAM3.1, forced by observed SST
March 1990 to February 2001.
• ~400 experiments completed (so far).
Histogram of configurations
with Improved skill
Convergence in predictions of global
means does not imply predictions
Much improved simulation of
rain intensities over tropics.
27,000 experiments completed in past year on 10,000 personal computers
(Stainforth et al., Nature 2005)
• Stochastic optimization of CAM3.1
suggests the model may provide
convergent results of global mean
– Assumes parameters tested are key sources
– Hadley Center model supports inference.
– Unanticipated gains in model skill.
• Important differences at regional scales
Each parameter affects many aspects of the model
There are multiple ways to combine
parameter values to yield better model skill.
Definition of model-observational data mismatch
E (m) (dobs g (m))T C1(dobs g (m)) i
(dobs g (m)) a j EOF j
N K a2
2 N i 1 j 1 2
Villagran-Hernandez et al. (in prep)