2010-10-19 09:23:59New article in response to John Kehr's header graphic - Need feedback
Rob Honeycutt


I was hoping to post this and get some feedback before posting.  I'm also a little annoyed at Kehr right now so I need to give this article a day or so before I post it.  It reads as follows....


Which Part is the Global Warming Part?

I've always held that science is inherently skeptical.  John Cook's website is a very good expression of that basic tenet of good science.  Question everything.  Twice.  Three times.  Then question it again.  Only then, as more and more evidence is amassed, can we even begin to have any confidence about any given conclusion.  

I had a teacher in High School who once told me, "Believe nothing that you hear and only half of what you see."  This is a statement, for me, that adequately describes the human condition; the whole of who we are.  We are easily fooled by our own eyes and ears.  This is why science works so incredibly well.  Science is the method by which we rise above ourselves and own human shortcomings to acquire a clearer understanding of the world around us.

There is good skeptical science related to climate change.  I have often defended the likes of Drs Spencer and Lindzen because they are putting out opposing ideas, even though I rarely find myself agreeing with their work.  But their work pushes the broader scientific community to respond.  Providing a scientist is putting out strong work challenging the established position the responses only serve to strengthen the science rather than weaken it.  

But, bad science is not this.  It is not skeptical.  Science done poorly only serves to weaken and give into our human folly.  One instance of this I wish to take to task.  In recent weeks we've been joined here at Skeptical Science by The Inconvenient Skeptic, John Kehr.  I've sent several comments to John about the header graphic he uses for his website.

The graphic comes from Richard Alley's GISP-2 ice core data.  John Kehr presents the last 4000 years of the GISP-2 data that ends in 1905 and points to this as the "global warming part."



Fig. 1 - From the header graphic at The Inconvenient Skeptic

The chart in his graphic is quite accurate and many of you may recognize it.  It absolutely does comes from GISP-2.  It shows the data from about 4 thousand years before the present to 0.095... thousand years before the present, and the text also states, "Data are smoothed from original measurements published by Cuffey and Clow (1997), as presented in Figure 1 of Alley (2000).  That means the data ends near the turn of the 20th century.  About 1903-05.

As you can see the header graphic is a fair representation of GISP-2, as seen in Fig 2.



Fig 2 - GISP-2 smoothed data for 4000 ybp to 1905. 


Yes, I know all you regular Skeptical Science readers (even self professed skeptics) are going to roll your eyes and heave a huge sigh looking at John Kehr's header graphic.  YOU, my friends, are applying healthy skepticism to the data being presented to you.  Of course your first thought is, "But what have we seen in central Greenland since 1905?"  Bad science would just present the data, create a misleading marker and rely only the laziness of the audience to not question where that marker is placed or what the data for current warming actually is.

But good skeptical science wants to question it.  Twice.  Three time.  Even more.

Resident ice specialist here at SkS, Robert Way, was kind enough to dig up some information on Greenland local temps since 1880.  The data even surprised my own suppositions on what I had expected.  My own notion was, with polar amplification the 0.8C global average rise in temps would result in something along the line of 3-4C in Greenland temps.  My own notion was that Greenland temps would be far higher than even the high peak of the GISP-2 data at 3500 ybp.  But I was wrong.  Here is what Robert turned up.





Fig 3 - Greenland temperatures from 1880 to the present

The source of this data is from weather stations around Greenland.  Robert attributes the drop in temperature during the 1970's to the great salinity anomaly (GSA), "an atmospheric process that cause[s] irregularly low temperatures in the north atlantic (sometimes 2 degrees celsius colder)."  If we combine the two charts we get something that looks a little more like this:





Fig 4 - Combined graphs

I would be the first to admit that this is not a scientific paper I'm producing and the results are likely to be even more complex that I'm relaying here.  There are any number of ways to rationally compare GISP-2 data with modern temps since 1905.  My point here is merely to suggest that you can not point to 1905 and call it the "global warming part" and expect that to be construed as anything other than a completely disingenuous statement.  This is not skepticism.  It's not science.  This would not make it past a peer review.  This would not make it past an undergraduate science course TA.  

We are living in a day where good skeptical science is being undermined by bad science lacking anything in the way of skepticism.  You don't have to love the Roy Spencer's and the Richard Lindzen's and the roles they play on this critical issue that humanity faces.  But they do serve a clear purpose, and though they may sometimes play on the very edge of solid science, they generally play within the boundaries of the established playing field with papers that take a lot of work to prepare and pass peer review.  It's an essential driving element for good science.

But I have to take the most severe issue with deliberate attempts to derail good science with misleading information.  I'm sorry John Kehr, 1905 is not the "global warming part."  There is no way to spin it.  There is no way to rationalize it.  You can only fix it.


2010-10-19 09:50:37Comment
Robert Way

Hey Rob,
Im gonna send you a revised version of the temperature. I have a better model for it. Also note that although the 1970s was probably cold with the GSA as you mentioned, I would not call it the driver of the cold. My opinion is that the driver is the AMO. Atlantic Multidecadal Oscillation. It is a predominate driver of north atlantic sea surface temperatures. Gimme a minute and i'll dig something up.

Robert attributes the drop in temperature during the 1970's to the great salinity anomaly (GSA), "an atmospheric process that cause[s] irregularly low temperatures in the north atlantic (sometimes 2 degrees celsius colder)."
2010-10-19 10:09:04Cleaned up the HTML
John Cook


I hope you don't mind, Rob, but I went in and edited the HTML of your post, just removed some of the double line spacing to clean it up. I also fixed the broken image of Figure 3 but I expect you'll want to update that with Robert's new model anyway.

I don't suppose you could replace "John Cook's website" with "Skeptical Science" in the first paragraph. I try to be as invisible as possible - I want the site to be about the science and the community effort.

2010-10-19 10:23:59Comment
Robert Way

Yeah Rob, I have to be honest, I don't have very much faith in the temperature data I gave you there. Not that it is completely wrong but I forgot to iterate what I did (AND YOU NEEED TO) so essentially I would need to iterate my station combination method a few times to get the actual result. Unfortunately I lost all the data a while back (USB issue) so if you have the table I could complete the process, otherwise I would consider asking for someone else to do it again. The best choice perhaps would be to tell me where geographically GISP2 was cored and I would find the stations nearest and use their data because GISP2 is not a representation of Greenland as a whole anyways. I don't mean to take the wind out of your sails here bud I admit its my mistake, I completely forgot to iterate and its only the other day i thought about it and remembered.

Once again if I sent you the data I can fix it but otherwise we have to consider other sources.

Also note what I said earlier about the AMO. The AMO is the dominant driver of temperatures in the North Atlantic Basin so it is likely a driver there in Greenland. The AMO had a negative phase when greenland temperatures dropped, likely explaining the cooling (or some of it). And was in a positive warming phase during the previous warm period (mid century) and the current warm period. The AMO is a periodic oscillation of sea surface temperatures in the north atlantic, it occurs over an 80 year period and alternates between warm and cool sea surface temperatures in that region. A positive AMO implies that the thermohaline circulation has intensified bringing more warm water from the south, a negative AMO implies that the themohaline circulation has slowed down bringing less warm water from the south.

Hope this helps, but yeah the take home point is that temperature data cannot be used unless we can find the original data.
2010-10-19 10:37:31another issue
Robert Way

Just clued into something else regarding the graph, umm temperatures in the 1970s were not colder than the whole little ice age?
2010-10-19 10:58:46
Rob Honeycutt



Well, I think I messed up the fixes you just made.  I'm not the best html guy by a long shot.  Not sure what I've done to bold the type.



2010-10-19 11:03:30
Rob Honeycutt

Kehr is also making that claim that GISP2 is a perfectly fine proxy for global temps over this 4000 year period because the ice core data is from evaporated oxygen isotopes originating from lower latitudes.  It's a claim I can't dispute right now.  If anyone has any clues on that one I'd like to hear about them.
2010-10-19 11:13:54Comment
Robert Way

Not to be a nitpicker, but I have a lot of concerns that I mentioned previously, plus the fact that the most violent swing in climate over the last 4000 years occurred in the data you added on makes me question the combination technique. Could I get some more information perhaps, and a response to my previous concerns? I can't afterall lend my name to something if I have these concerns.
2010-10-19 11:17:37



I believe he's off base: The ratio of O-16 to O-18 in the ice cores is a function of LOCAL temperature, because it has to do with the molecular dynamics of condensation. There IS of course a dependence on the ratio in the gas phase; but to first order, you derive the temperature at the time of freezing by assuming that the gas-phase isotope ratio is the same, regardless of latitude.

Check out:


and note that both O-16 and O-18 are stable (non-radioactive) isotopes.

2010-10-19 12:49:39


Am I becoming a hopeless curmudgeon? I hope not. 

A decent amount of effort and attention put into this, it's true and good, but it's a gift of fame to Kehr, who is a pipsqueak and presently invisible.

How often has SkS devoted an article to a takedown of WUWT?

For the most part, SkS seems at its best when it's tackling untruth without pointing fingers. Sure, the occasional exception is practically demanded (Jo Nova comes to mind) but there's a risk here of getting involved in a slagging match via blog posts with a distinctly counterproductive outcome, that of boosting John Kehr in the constellation of contrarians. 

Just a thought to consider.  

2010-10-19 17:24:01
Ari Jokimäki


Can anybody point me to the proof that the GISP2 data ends at 1905? "Before present" usually refers to before 1950 which would make GISP2 data end at 1855. However, Alley (2000), which is cited as original reference to GISP2 temperature reconstruction in NOAA paleoclimate website, is kind of vague on that. He has "before 1950" in one of his figures and one other figure says "before present". The additional reference, Cuffay & Clow (1997) doesn't seem to be available for free.




2010-10-19 18:35:34
Rob Painting

I agree with Doug, you're shining the spotlight on some LIM (lying incompetent moron - Doug Mackie trademarked). I'm sure Kehr would be delighted if this were posted, he'll bask in the attention. His blog doesn't appear to attract much in the way of reader comments (so far), so maybe not much traffic?.

As it stands the post doesn't deliver some killer blow anyway. What would the poorly informed, but rational, individual take away from this, reading this post and the exchanges at Kehr's blog?. Some ongoing personal squabble?.

2010-10-19 19:49:15

This sounds like something that should be fixed "on spot", but not responded to by an SkS posting.
2010-10-20 02:00:30
Rob Honeycutt


Robert...  I think you're totally right.  I definitely wouldn't post this with your name attached or the data you provided if you didn't feel comfortable with it.  That's exactly why I wanted to post this on the forum before posting it.

I'm sure I'm going to have to completely rewrite most of this piece.  It was written in the heat of the moment after reading Kehr's post yesterday defending his uses of his graphic.  Not an ideal time to write things.  But it's serving as motivation to approach this issue.

I've seen GISP2 used in a number of places to make the claim that current warming is nothing compared to the past 4000 years.  They always use that little tick up at the end of the GISP2 graph and claim THAT is current warming, or somehow related to current warming. 

2010-10-20 02:18:39
Rob Honeycutt


Doug, Dappledwater, nealjking...  I definitely see what you're saying.  It seems the only traffic Kehr is getting at his site is people from SkS since he started posting here.  

Maybe the more appropriate course of action is to continue to pepper him with calls to change his graphic at his site, in comments there.

At the same time, though, this is not the first time I've seen the GISP2 data used in this manner.  There are about 6 different youtube videos that's I've seen that use the graphic in an almost identical manner.  So, I think there is something here to address.

It would be probably better to address this whole issue sans Kehr.  Don't give him the benefit of fame at SkS while still addressing the fundamental faulty use of the GISP2 data. 

I had started to dig into this a while back and, not having a science degree, felt a bit overwhelmed at the task.  I'm a very good writer (I think), have a lot of enthusiasm for the general topic of climate change and would like to apply my skills to the issue.  My art school days just don't seem to lend too much aid in tackling some of this data.  But once I understand something I'm highly capable of producing unique and compelling copy.

2010-10-20 02:21:23
Rob Honeycutt


nealjking...  Here is the reference Kehr provided for the O18 ratios.


2010-10-20 03:13:18


The TIS example is admittedly extra comical in a dark way, Rob. 

One way to thread the needle might be to avoid actually mentioning Kehr or "The Inconvenient Skeptic" while still using his graphic as a leit motif, with the caption "An ironically branded example." Or, use his similarly pseudoanonymized as well as another, to illustrate the breadth of the infection.

I've got to say, Kehr's header is an extraordinarily arrogant deception. 

2010-10-20 03:30:19
Rob Honeycutt


Doug...  I agree about the arrogant deception.  That's what has me so irked.  And the fact that he so blithely defends his use of GISP2 data.  

The other thing I want to do with this article is come up with a better, broader metaphor.  Something is going to click and pull it all together.  

It helps having all the comments here.  Thanks! 

Here is another commonly used graph that I just noticed also contains the GISP2 data (light blue line) and points out how misleading it is to use one line of evidence. 

2010-10-20 08:07:57


Rob Honeycutt,

Thanks for the reference: Very interesting.

My first impression is that they're saying that the Oxygen isotope ratios in the ice reflect the local temperature, but the ratios in the ocean reflect a kind of summation of what's happening in all the ice regions: If it's really cold, and more than the usual amount of O-16 is being frozen in, then there will be more O-18 in the oceans that is usual. (That's what I was kind of thinking about when I originally made the remark about "to first order"; to second order, I would think about the effect of depletion.)

But apparently the whole issue of pinning down the temperature is fairly complicated: Depends on latitude, altitude, etc. See http://www.scholarsandrogues.com/2009/09/22/oxygen-proxy-errors-corrected/ . Anyway, getting a global indication out of an ice-core measurement is going to be tough, since even different nearby measurements have to be relatively calibrated for these characteristics.



2010-10-20 09:41:22Reframing this post
John Cook


Perhaps the idea of reframing this post so it's not about TIS is a good one. Instead, make it a teachable moment, talking about the proper use of data and the general misuse of GISP2. Rob, you're excellent at creating a compelling narrative so this would be a good approach for you.

The argument I am repeatedly pushing is that genuine skepticism means you look at the fully body of evidence. Cherry picking select pieces of data and rejecting the rest of the evidence is the antithesis of skepticism. It's part of a gradual campaign to take back the term skepticism.

2010-10-20 10:47:38
Rob Honeycutt


You know what is interesting about this whole thing?  Here, this self professed "skeptic", John Kehr is putting forth what I see to be inaccurate information regarding climate.  He's not a complete dummy.  He does some research, but his conclusions are frightfully wrong.  

What it's causing me to do is do MORE research.  I'm learning more about this issue because I'm ticked off at the guy.

His lazy skepticism is making me smarter.  Kinda the point I was making in my draft post above.

2010-10-20 12:10:40
Rob Painting

Rob, definitely worth debunking this rubbish, it's one of the prime objectives of Sks afterall. Just cool down first. Kehr strikes me as well practiced troll. He undoubtedly knows he winding you up. Don't reference him at all.

As for figure 3. Isn't the period from 1940(ish) to the 70's (cooling) strongly influenced by human sulphate emissions too?. 




2010-10-20 15:54:46GISP-2
Dana Nuccitelli

I agree that it's not worth focusing on TIS.  It will just give him traffic that he doesn't deserve.  But particularly if you've seen this data used in the same way from other sources, you could just discuss how it's frequently misused, and then discuss why.  Just keep it vague without naming any specific sites or skeptics.

I'd also put more emphasis on the fact that it's Greenland-only data, and I really don't buy using it as a global temperature proxy.  But that's worth discussion in the blog post.

2010-10-21 02:20:51
Rob Honeycutt


Thanks Dana....  I had my brilliant stroke of metaphor for this article about 3am last night.  I think everyone will like it.  I'm completely rewriting the post.  It's getting fun now.  I'll post a draft on the forum here maybe today or tomorrow.

You're right, though.  I've seen the same data used a number of places.  I'm planning to only reference TIS in a link as one misuse of the data among many.