2012-01-11 17:21:45Gillett et al. Estimate Human and Natural Global Warming
Dana Nuccitelli

Post on the Gillett paper.

Gillett et al. Estimate Human and Natural Global Warming

2012-01-11 18:44:07
Ari Jokimäki


I ran your post through Word spell checker:

"In another new global warming attritubion study"

attritubion -> attribution

"...which refers to the global mean temperature change that is realised at the time of CO2 doubling in ascenario in which CO2 concentrations increase by 1% per year."

ascenario -> a scenario

"...namely that humans have been the domnant cause of the global warming over the past 25-150 years (Figure 1)."

Domnant -> dominant

"...Gillett et al. use use five-member ensembles..."

use use -> use

"A linear trend fitted to the obs in Fig 1a would show less warming than a difference between the first and last decades."

obs -> observations?, the Fig 1a should be 2a, I think.

"The result stil remains counter-intuitive,..."

stil -> still

2012-01-12 03:01:34
Dana Nuccitelli

Hmm seems my typing skills were a bit off last night.  Thanks Ari.

2012-01-12 09:19:21


Fig. 1 caption: in "the sum of the trends" and "observed trends" it's not trends but temperature changes

Transient Climate Response: in fig. 3 the two time frames shown are 1851-2010 and 1901-2000; the discussion (right before the graph) on the 1961-2000 intervall doesn't fit or I misunderstood what you're saying.

I think you should ask permission before including private messages. You may also let him read the post before publishing it and eventually include his comments.

Strange how the RCP 2.6 projections are flat after about mid century.

2012-01-12 09:33:09
Julian Brimelow


As annoying as they are, I would suggest reading some of the dimwit comments at WUWT.  For example,

"John N says:

John, I read it the same way. The “Key Points” further express preference for the 1851-2010 time period as “Using 1851-2010 observations gives lower AND LESS UNCERTAIN projected warming”

So they prefer the 1850-2010 window b/c it has a lower uncertainty in the scaled projections...at least that is my take. And that is reasonable, but using a longer data set will by default probably lead to that (Dikran where are you!?), it does not necessarily imply greate skill or future skill. Anyhow, I have a feelign that this higher confidence is an artifact of the nuimber of data points or natur eof the data for the to windows rather than something physical.


1) I read the post and the second part , to me, sounds a little defensive right now.  For example,

"The relatively low warming estimate in the Gillett study appears to influence their somewhat surprising TCR results."

I don't think their findins surprising. I'd leave out such references.

Remember, as I keep harping on about,  the median TCR of the models listed in Table 8.6 in AR4 isd 1.6 and that is in the range determined using the 1850-2010 data, and that those models estimate the EQS to be near +3 C.  [The range of TCR using the 1901-2000 data is higher 1.7-2.5 and that is the upper bound for the AR4 models.]


2) "Given that there is greater uncertainty associated with the HadCRUT data prior to 1900"

Can you qiuantify this or provide a link?  I think this is key here.   What was the global coverage by HadCRUT before 1900?

3) "Obviously the dotted bars (based on the 1901-2000 regression) match the simulated projected temperatures (dashed lines) much more closely than the solid bars (based on the 1851-2010 regression)."

Be careful here--tas this stands it could (rightly) land yuin hot water. What is to say the unscaled version of the model is truth? It is dangerous to validate a model  against a model, especially for a projection.  Also, that projection was run using the unscaled model data and the authors admit that the model is overly sensitive to all forcings (not just CO2).  What I do not is that for the 4.5 RCP there is quite a bit of overlap between the raw model run and the model run scaled using the 1901-2000 scaling.  Perhaps the more likely answer is where there is overlap between the dotted and solid bars-- that would indiscate about 2.5 C warming by 2100 for RCP 4.5.  Alternatively their data suggest between 1.8 and 3.5 C warming by 2100, depending on whicj scaling they use. 

Anyone suggesting that this paper means that we are off the hook is delufding themselves and their readers.


4) "that the TCR 1901-2000 better matches simulated projected temperatures"

It could be that those evil IPCC assessments are wrong though Dana, and here lies th truth! ;)  I would not say this is avalid reason. I still like my reasoning that the TCR range in this study encompasses the range of the models used in AR4, it is honest and fair comparison-- and shows that this paper is probably not a game changer.

Maybe note that it will be interesting to see how the reults changed when their method is applied it multiple models for multiple time periods using more than one surface temperature record.

2012-01-12 09:44:30
Julian Brimelow


"I think you should ask permission before including private messages. You may also let him read the post before publishing it and eventually include his comments."


2012-01-12 10:07:20
Dana Nuccitelli

I did ask permission - just haven't heard back yet.  If Gillett doesn't respond I'll take his quote out and paraphrase it.

Thanks for the comments, keep 'em coming.

2012-01-12 10:54:48
Julian Brimelow

People are going to go over tis with a fine toothcomb after your appearanc eat WFUWT today Dana.  So I encourage people to be super critical (sorry Dana, but it ifs for the greater good), check for hints of "cheerleading" etc.  I didn't find any evidenc eof cheerleading and appearing to dismiss the lower TCR values, it would be nice if ThingBreak would have a good read.

I am tired and can't spend any more time on this, but know this is in good hands.  Good luck Dana.

2012-01-12 11:03:11
Dana Nuccitelli

Dang, Gillett asked me not to directly quote him.  I'll have to paraphrase his explanation.  Ah well, not a big deal.

Thanks for your input Alby.

2012-01-12 14:26:35
Julian Brimelow

No worries Dana, you are welome as always, I enjoy working with you.

2012-01-12 22:14:57Yeah, gotta be careful quoting scientists...
John Cook

I corresponded with a scientist once then quoted him in a blog post without permission. Then I got an angry email from him, saying never to email him again. I nearly dislocated a joint in my hurry to remove the quote and sent a gushing apology. That mollified him although I've been too scared to email him since and I will never quote a scientist without permission ever again.