Boring old climate sensitivity

JA is bored with climate sensitivity - because he knows the answer, 3 oC, and he may well be right. But other people don't seem to have realised. And (via James again, I think) I ran across Tung and Camp on climate sensitivity, and Knutti et al.. They too think its 3 oC (well 2.8 +/- 0.9; and about 3). K et al. are doing this via CP.Net in J Climate; T&C via ERBE data in, they hope, Nature.

K et al. use the plausible idea that there could well be a relationship between the size of the seasonal cycle and the climate sensitivity of the model. And so they use 2500 perturbed-physics ensembles from the CP.Net experiment to derive a relation between seasonal cycle and sensitivity, and then say that predicts such-and-such given that we know the real cycle. Its interesting; not being my speciality I don't really judge it, but I notice that: models with the wrong seasonal cycle are... wrong. Its not clear that you can usefully judge stuff from them. Another point on the same lines is that one criticism of the original CP.Net stuff was that they used a very thin set of model-reliability criteria, and it was suspected that including seasonal cycle might well filter out some of the very high (implauibly high) sensitivities. That now looks like being true. Secondly, this is all in HadSM3 perturbed. Secondly, if you look at their fig 2 then it seems that the relation is rather different amongst the different AR4 runs.

T&C use ERBE data, and some kind of fit to discriminate which bits of the sfc T is due to solar. I haven't read it all yet; presumably this also gives you how much of the sfc T is solar, if you believe it, which would be interesting in itself. JA appears to promise his opinion in due course, so I may wait for that. Or I may perhaps find time and brain to read it myself.

Exercise (for the reader? for JA?): redo JA's latest estimates, including the independant C&T values.

[A reader complains about excessive abbreviations. Sorry. I have excuses... anyway: CP.Net is climate prediction .net. ERBE is earths radiation budget experiment - a satellite (series). JA is James Annan. HadSM3 is the Hadley Centre climate model, version 3, slab version. sfc T is surface temperature. Wikipedia should expand those if you need more -W]

More like this

Tuesday morning has at least 4 sessions I could have been interested in. Leave EPICA for later and start off (cos I happen to pass the room) with Latif on MOC; which to me provides more evidence not to worry about it. Thence into the climate sensitivity session, which is packed. Matt Collins talks…
OK, the Knight et al. paper is here, thanks folks. Clearly they have had some jolly fun dividing the runs up into trees, but the paper is a disappointment to me, as it doesn't really deal with the main issue, which is the physical plausibility of some of the runs. It *does* talk about "Our findings…
I was at the NCAS conference today (since it was in Cambridge it would have been impolite not to go). Tim "Da Man" Palmer spoke about, ermm, sort of a merge of NWP and climate scales. But thats not the point... the point is that he showed a stratification of the Staniforth CP.net PDF in terms of…
OK, so I missed it, not that I was looking for it. Why did everyone else miss it? Because cp.net is obsolete? Since I don't have a subs to PNAS I can't read it except the abstract. Here is the abstract so you don't have to follow the link and I show up on google: In complex spatial models, as used…

Your use of abbreviations, acronyms, and shortcuts reallys gets in the way of understanding what you're trying to say.

Link and snippet below are from a newspaper article; it refers to a study I haven't seen; wonder if you have?

"... Harte, in collaboration with Margaret Torn, a staff scientist at the Lawrence Berkeley National Laboratory, added insight to global climate models in May in a paper based on the ice-core data and published by the American Geophysical Union.

"They concluded that temperatures by the end of the century will be even hotter than the models currently predict because of the heretofore uncalculated feedback effects ...

"If emissions from the burning of fossil fuels continue as expected, and carbon dioxide levels reach 560 parts per million by 2060 as projected, temperatures could increase by 12 degrees Fahrenheit above pre-industrial global temperatures, the study found. Current models project a rise of 8 degrees."

http://sfgate.com/cgi-bin/article.cgi?file=/c/a/2006/07/30/WARMING.TMP

By Hank Roberts (not verified) on 29 Jul 2006 #permalink

I agree with Rufus ... I struggle sometimes with the acronyms. Could you give the terms in full the first time they're mentioned in a post?

Am I right that there's no clear and simple "climate sensitivity" as an absolute, it will be a number that depends both on where the continents happen to be at the time, and the time span over which it's considered, and much else?

So when we have studies like the one above saying the temperature may be higher -- that doesn't change the "climate sensitivity" numbers, it refers to outcomes that are outside what the climate sensitivity numbers consider?

I'm trying to imagine myself as blinkered as some of the people I've worked for in the business world, and still trying to make sense of the superficially contradictory things the press is reporting the scientists as saying.

Sensitivity always seems to come out about the same; all I can figure is that we're coming up with more and more possible feedbacks and talking about widely different time spans (decades vs. centuries vs. millenia) -- so the temperature numbes vary. But climate sensitivity calculations are defined to not include those, yet?

.... hold my hand ....

[It will definitely depend on basics like where the continents are. Since its *equilibrium change* at 2*CO2 is doesn't depend on the time span, since there isn't one. It could depend on base CO2 level. Read James Annan - he thinks its 3 -W]

By Hank Roberts (not verified) on 01 Aug 2006 #permalink

Yep, I am reading James Annan's work and most appreciative of it.

The reminders that climate sensitivity is
-- a measure at equilibrium (time span doesn't matter); and
-- at 2*CO2 (a fixed number)

both are helpful. Thank you.

That helps explain that "climate sensitivity" does not express "how sensitive our climate is" in the lay sense.

Is there any clear word that expresses how the rate of change may change?

For example, melt more deep permafrost; clathrates gasify; ocean acidity reaches the threshold where aragonite dissolves in surface waters.

I don't know a climatology term that expresses something like "vector" or "slope" -- something expressing both direction and rate of change.

Which is what lay people ask about all the time -- "where's it going, how fast?"

[Melting permafrost would count as a feedback. As for the other - how about dT/dt? In oC/decade? - W]