Friday, November 06, 2009

Roy Spencer debunks Lindzen and Choi

Roy Spencer has weighed in here (thanks tb) with some analysis of the Lindzen and Choi study that I briefly covered before. It seems that RS has investigated the difference between CMIP (coupled atmosphere-ocean) and AMIP (atmosphere with prescribed sea surface temperature) runs and found that they give completely different answers. In other words, the analysis of AMIP output that LC performed is not relevant to diagnosing the properties of the fully coupled climate system. Which is what I suspected but had not checked. RS also has various other criticisms about how the data were processed, and his alternative analysis shows a much closer agreement between models and data. Although his wording tries to be gentle (because he wants to believe LC's overall conclusion that the models are too sensitive) it is quite clear that he thinks the LC paper is wrong.

The sad thing about this is to see Lindzen getting his claws into some young post-doc and teaching them how to do (and get published) shoddy analyses without doing obvious checks. I hope this person learns how to not fool himself so easily in future.

Update

Comments continue on the earlier post, and Rob in particular seems to have pointed to some sleight-of-hand in how LC dealt with the analysis of SW radiation (unrelated to the CMIP/AMIP issue). There are so many different definitions of feedback that I'm not 100% sure on this though.

6 comments:

Hank Roberts said...

Do you think the paper in press he that mentions covers these issues? I can't tell from the blog if he's saying he will or won't be publishing this criticism in a journal.

I guess it'd be hard for someone else to publish the same criticism in a science journal article, now that Spencer has blogged it.

James Annan said...

I think he only mentions a paper in review, which is Spencer + Braswell and has been kicking about for some time on the blogs (IIRC it's regarded as pretty dodgy). Since he says "as yet unpublished" for this latest work, it sounds like he is at least thinking about submitting it.

Steve Bloom said...

James, I seem to recall that you have in your files a text-basded irony meter from usenet days. Time to dust it off, I think, as it doesn't get more ironic than this.

David B. Benson said...

Steve Bloom --- I fear I'll have to buy a heavy duty model. Mine 'sploded upon reading this thread.

Tom C said...

"...because he wants to believe LC's overall conclusion that the models are too sensitive..."

If he "wanted to believe" this he probably would have kept quiet or communicated privately with LC. Seems to me he addressed the issues in a straightforward, constructive way with malice toward none. Kind of a good role model for others in this field, eh?

James Annan said...

Tom,

Spencer said explicitly: "I tend to agree with the Lindzen and Choi position". But yes, I do approve of the way he was prepared to state his disagreement openly.