« North Korea saves planet (again) | Main | The President’s gullible base re-acknowledged (#2 update 3apr12) »

01 April 2012

Comments

Russ Steele

George,

Thanks for mentioning NC2012. I have some more thoughts on climate change models HERE, however this is the money quote in my post:

. . . for the models to have any credibility to predict the future climate on decadal time scales . . .:

1. They must accurately simulate (hindcast) the statistics of major atmospheric and ocean circulation features over the last few decades (since real world data is available)
and

2. They must accurately simulate (hindcast) the statistics of the changes in the statistics of these major atmospheric and ocean circulation features over the last few decades.

If they cannot do both #1 and #2, they must be rejected as robust predictive (projection) tools for the coming decades.

The problem is our political leaders in Sacramento, and their rogue agencies like CARB, are using these unvalidated models to make policies that are costing us billions.

George Rebane

For readers unfamiliar with the terminology here: the term for hindcasting in estimation theory is retrodiction as used in the post.

Bonnie McGuire

Lucille's a wonderful lady who usually expresses the way it is pretty well.

Gregory

I'm not sure a hindcast buys anything over what we have now.

The forward-only general circulation models are "parameterized" (ie jiggered) to match exiting data, which is all we have. Reality is 20 20 hindsight.

A bidirectional model would also be parameterized to match the past.

So, where is the advantage? The problem is that each time there's new data, rather than say, hmmm, the GCM are diverging from reality they again jigger the data to fit.

There's nothing per se wrong with doing that. However, until you stop having to jigger the parameters, the model remains speculative and unverified. Or more to the point, determined to be incorrect.


btw, Dr. Happer made a slight error in favor of the GCM. He wrote,

"The direct warming due to doubling CO2 levels in the atmosphere can be calculated to cause a warming of about one degree Celsius. The IPCC computer models predict a much larger warming, three degrees Celsius or even more, because they assume changes in water vapor or clouds that supposedly amplify the direct warming from CO2."

In fact, none of the GCM cited by the IPCC predicted a 3C warming for a doubling of CO2. They were all over the place, and rather than say "we don't know because no two of the GCM agree", they took an arithmetic average and declared the answer to be Three.

George Rebane

Discovered jiggering data is more dangerous to your career as a scientist than jiggering model parameters, which is called research. Basing public policy on jiggered and still jiggering results from such ongoings is lunacy.

Tony Loro

http://thetyee.ca/Blogs/TheHook/Environment/2012/06/28/exxon-fossil-fuel-adapt-climate/

The sound of one pig wing flapping...

The comments to this entry are closed.