2. I suggest doing a lit review, rather than relying too much on blog authors for probabiliites.

3. I think it will be very tricky to do much useful wrt Prob AGW, but perhaps the attempt will have some use.

4. I do think a review paper of ill (anjd good) effects which gives probabilites as well as source references has some usefulness.

5. Seems awfully long-widned in making the asic point (which I have also noticed) of the prevalent confusion of ill effects depenadnt on AGW with AGW iteslf.

6. It actually lacks concrete examples of the sin in comment 5.

It is clearly written and easy to understand. The realities of uncertainties have been clearly underestimated.

Maybe the mention of http://www.surfacestations.org/ can help reinforce some point about problem with location of measurement.

Roger Pielke sr also has a nice post today about the inadequacy of surface temp:

http://climatesci.org/2008/03/28/a-short-tutorial-on-global-warming/

Maybe this can help.

Sylvain

The conclusion made by B&P reminds me of an example I often use in my class: “kids who can play piano do better in school”. This statement should not imply that there are no other factors (or measurements) that could possibly affect a kid’s performance in school (stage 1). Nor does it imply that the model employed to make such conclusion (stage 2) is perfect. Moreover, one cannot conclude a causal relation (stage 3) between “being able to play piano” and “doing better in school”.

And students ask ?Do your kids play piano?? Yes, they do? just in case there is indeed a causal relation.

Thanks for your comments. I have done a poor job defining significant, no question about it. I tried to state it in terms of the IPCC forecast, which would certainly be significant. The problem is that I want to avoid circularity: AGW can’t just be significant if and only if it causes some undesirable effect.

I am also trying not to be too specific on any one item because, obviously, you can write papers and papers on any of them. I was aiming more for a review-paper type thing.

Also no question that it needs a lot of work. I agree with your Orrell comment, and included that book only because of the many references therein.

Lastly, your guess that
Pr{chaos not important} = 0.2
is welcome.

Thanks again.

Well to answer your question, one number that is completely wrong is Pr{chaos not important} = 0.9. A figure of 0.2 might be a better estimate.

The irregular oscillations of the climate over many different time scales, combined with the failure of prediction (eg the recent leveling off) are strong signatures of chaos.

Your reference Orrel (2005) is not the best - that paper is about weather not climate, and uses some very simple toy model equations which have no relation to weather or climate. A better choice would be Palmer (2000), cited by Orrel. That paper even has ‘uncertainty’ in the title so you should like it!

Unfortunately a google search for “climate chaos” turns up mostly junk (as is the case for either of these words alone!) a good example being the blog
http://www.pbs.org/kcet/wiredscience/blogs/2007/11/chaos-part-2-chaos-doesnt-matt.html
where I have just posted a comment.

Here are some that appear to be sensible-
http://www.sciencemag.org/cgi/content/full/sci;296/5567/483
http://blog.sciencenews.org/mathtrek/2007/08/cloudy_crystal_balls.html
http://www.nonlin-processes-geophys.net/8/201/2001/npg-8-201-2001.html

Noting the recent work by Anthony Watts to confirm a possible 66 year cycle, the concept of even GW may not be as secure as we are lead to believe. That is beyond the uncertainty with proving the subset of AGW.

John M Reynolds