Addiction of paleoclimate reconstructions to particular proxies has been a longstanding concern at Climate Audit.
One of the battleground issues has been the addiction to Briffa’s Yamal tree ring series, while the nearby update of Polar Urals (with a pronounced MWP) was disappeared. (See CA category.)
Just before Climategate, we raised questions about the Yamal reconstruction – noticing first that, contrary to prior belief, it was not “highly replicated”, having far fewer modern trees than is standard for an RCS reconstruction, and that the reconstruction was highly sensitive to inclusion of a nearby Schweingruber site (Khadyta River, Yamal). Briffa did not deny the validity of this criticism, instead attempting to salvage the reconstruction by adding in cores from some nearby sites (but notably not Polar Urals.)
The elephant in the room remained the disposition of the Polar Urals site. A 1995 Briffa reconstruction from this site purported to show that 1032 was the “coldest” year in the millennium. Updated data had shown elevated ring widths in the MWP. However, Briffa hadn’t reported this. (Schweingruber had archived the updated measurement data at the ITRDB, but no journal article had reported the results.
The Climategate Letters have a teaser here. On April 28, 2006 (almost exactly the same date as I was being stonewalled about Yamal, Taymir and Tornetrask measurement data), Osborn emailed Philip Brohan of the UK Met Office:
To: philip.brohan@xxxxxxxxx.xxx
From: Tim Osborn
Subject: Re: Standardisation uncertainty for tree-ring series
Cc: Keith Briffa ,simon.tett@xxxxxxxxx.xxxHi Philip,
we have three “groups” of trees:
“SCAND” (which includes the Tornetrask and Finland multi-millennial chronologies, but also some shorter chronologies from the same region). …“URALS” (which includes the Yamal and Polar Urals long chronologies, plus other shorter ones). These fall mainly within these 3 boxes:
52.5E, 67.5N
62.5E, 62.5N (note this is the only one not at 67.5N)
67.5E, 67.5N“TAIMYR” (which includes the Taimyr long chronology, plus other shorter ones). These fall mainly within these 4 boxes:
87.5E, 67.5N
102.5E, 67.5N
112.5E, 67.5N
122.5E, 67.5NWe do some analysis at the group scale, and for this we take the JJA temperatures from each box and average to the group scale to obtain a single series from each of SCAND,
URALS and TAIMY.We do some analysis at the overall scale, and for this we take these three group temperature series and average them to get an overall NW Eurasia temperature for boxes
with tree chronologies in them…
So on this occasion, the long Polar Urals chronology, together with other “shorter” chronologies (presumably the Schweingruber chronologies that Gavin Schmidt condemned) were included in a larger regional RCS reconstruction – something that Rob Wilson would have been interested in seeing.
But the subsequent Briffa et al (2008 Phil Trans B) only includes the very small Yamal data – without the long Polar Urals chronology or the shorter chronologies. Wonder why?
Climate Audit 
48 Comments
Vanishingly small point, but a missing “)” in the 4th para “(Schweingruber…”
Also it may be helpful to put a few links to you previous analyses of these series at the end for a bit of relevant background.
“Wonder why?”
No. As other Texas Sharpshooters have pointed out, they are paid to find a story. And simply adding more data adds noise that obscures the real signal.
I still don’t understand why the big names in dendro are not standing behind you, or pretending they saw it first and were outraged. Your efforts to fix the deceptions of Yamal seem like an effort to rescue the science, not maim it.
General heavyhanded backlash against anything associated with warming and the IPCC seems to be gaining momentum. It will be socially difficult for an arrogant socialite to be on the wrong side of it soon.
Yes, it does seem as if everyone has gone to ground…
Why would scientists admit that they didn’t catch the shenanigans that were going on at EAU CRU?
One of the current memes is that even IF CRU were behaving badly, the other researchers in other places have stayed on the straight and narrow.
Isn’t that logically impossible? If you’ve been hewing strictly to the scientific method, shouldn’t you be able to detect shoddy work in your colleages right away? Shouldn’t you be outraged at the first incidence of data being withheld — or worse, being deleted? Shouldn’t YOU have noticed that the computer code was an irretrievable mess from the off and that the data sets were incomplete?
If CRU is shown to have behaved un-scientifically, then that indicts the entire field, which failed to correct bad behavior by the alleged “small but influential” number.
Its not simple to challenge another’s work.
When you read a paper which seems to disagree with your own beliefs, unless you have the same data and fully understand the methods used its very hard to definitively say its wrong.
The data was generally not available. The computer code was generally not available.
The first thing you do is question yourself and your methods. If you have the inclination, time and money you may try to replicate the results.
So I am not at all surprised that no-one called them on their results when the papers were published.
Since climategate there has been some evidence of … shall we say “poor scientific method”, but now is not the time for people to stick their heads up. Its far from clear who is going to win. Academia is a very political place, and now its political^2.
Depending on who wins the political battles I do expect to see some conflicting comment once the dust settles.
“The data was generally not available. The computer code was generally not available.”
I’m wondering why they weren’t protesting the lack of available data. Maybe I’m naive, but I thought if you didn’t understand how one of your peers arrived at conclusion X, the scientific method requires you to obtain the data and methods, then try to replicate your peer’s conclusions.
And unless I’ve been reading the protests from the Team wrongly, the data was supposed to be freely available to other researchers.
PJP
I understand your point as I find I am questioning myself wondering what I am missing. I cannot fathom how a group develops a climate estimating model that they know is not capable of predicting knowns (during the “decline”) but still advocate its use for predicting unknowns from 1000 yrs past and no one in the field notices. Seriously, someone educate me.
Is the Polar Urals site in the 62.5E, 62.5N box? If so, would that be an ostensible reason to exclude it from a reconstruction otherwise located at 67.5′ N?
Steve: It’s higher than Yamal and is a treeline site.
Re: Pat Frank (Jan 4 13:27),
The rationale for inclusion would be a “treeline”, with some species under stress. I imagine outer individuals, at their limit, not any of the other ones that are more protected.
I am not dwelling on how that makes a thermo-meter.
It seems that the selection of additional series IS the real point — whether it’s justified or not, it’s hard to avoid the suspicion that selection bias is at work, however justified.
“Ostensible” is the key word in the original comment by Pat Frank. Cherry picking data points, series, and adjustments is the method used to derive and defend the politicized conclusions.
“Yes, it does seem as if everyone has gone to ground…”
If it weren’t for you meddling kids…
A long Polar Urals chronology…..as in maybe it goes back before 1400, maybe it includes the Medieval Warm Period? Or should I say the flat, relatively cool period once known to mankind as the MWP.
I believe Mann now grudgingly refers to the MWP as the MWA(nomaly). Some anomaly.
But the subsequent Briffa et al (2008 Phil Trans B) only includes the very small Yamal data – without the long Polar Urals chronology or the shorter chronologies.
The point is well made.
There might be some valid innocent explanation why the Polar Urals chronology was excluded, but none has been offered. The most charitable explanation I can think of is that it was overlooked out of carelessness, but this hardly seems reasonable. It is troubling.
Climategate has raised many issues, but the most serious assault on the reliability of the data appears to be the extent of the cherry picking of data involved in the datasets used by the modelers e.g. garbage in garbage out.
Given that we frequently hear about how the CRU data is just one of four temperature reconstructions and they all supposedly support one another, what does this say about the independence of the other reconstructions? How independent are they really? If they really are independent and they broadly agree, doesn’t that indicate there may be some serious integrity / cherry picking issues in the other data sets as well?
Steve: Last point is OT.
The “four temperature reconstructions” must be refering to the Hadley, GISS, and NOAA 150 year global instrumental temperature reconstructions. Polar Urals is a tree ring data set, not instrumental. Steve keeps asking about Polar Urals and Tornetrask because these tree-ring series don’t give the hockeys stick shape, but are ignored by the Team without justification (which is cherry picking, as you mention).
snip – OT sorry.
Mann is not involved in creating the instrumental climate reconstructions, only the proxy ones. The issue of cherry picking for the instrumental data involves things like the drop out of rural stations around the world after about 1990 (the data are there, but no longer included) and the failure to adjust for the Urban Heat Island effect, as well as adjustments of rural stations to create a warming trend with no documentation of the reasons. Instrumental and proxy are two different things, but both suffer from hidden computer codes, undisclosed data sources, and unknown adjustments.
Treerings
http://star.arm.ac.uk/press/2009/treerings_2009dec.html
Dear Mr McIntyre
Not sure where to post this but please delete when you have considered.
The following is posted at the BBC web blog of Richard Black.
http://www.bbc.co.uk/blogs/thereporters/richardblack/2009/12/2010_intruders_on_a_biodiverse.html#comments
I have complained that the post is defamatory and libellous. My view is that if the poster had brains he would be dangerous and if he had two they would both be lonely.
Please delete when you have read.
389. At 7:07pm on 04 Jan 2010, U14260427 wrote:
“does anyone think the beebs bias might stem from the fact that their pension trust is a signatory invester of the CDP?”
And Steve McIntyre has money from fossil fuel companies.
And Inhofe has a large number of funders from the fossil fuel companies.
Please prove that that causes bias, as opposed there not being any reasonable theory disproving AGW in the same way as there’s no reasonable theory showing that the 11/9 bombings were caused by the Bush administration.
Or are the BBC biased against the truthers?
End of Rant.
Further to my earlier post I have received the following from Richard Blacks blog
Dear BBC Reader,
Further to your complaint about some of the content on a BBC blog (reference number P27106603), we have decided that it does not contravene the House Rules and are going to leave it on site.
Please note that this response refers to a complaint about one specific piece of content. If you have alerted us about a member’s nickname or behaviour or other ongoing issues relating to the blog then we may still be taking action.
The complaint button should only be used to alert the moderators to content that may contravene the House Rules.
If you disagree with another poster’s viewpoint, please add your own views to the blog.
General questions, comments and other feedback about the BBC cannot be responded to or passed on via the moderation process.
Regards,
The BBC Blog Team
It is getting late now and I am so ashamed that the BBC would let this stand, but I will pick up on it tommorrow.
Nos Da
I was curious as to why all the attention to a fragmented tree ring chronology when the aono/amoto 1994 cherry blossom reconstruction for Kyoto extends all the way back to 1000 AD and is pretty much continuous. So I looked at it and it seems to show a variable but steady climate with hockey stick style warming starting in the mid 1900’s. I read the methodology in detail and concluded that an error had been introduced to the data prior to 1583 due to improper calendar conversion. I approximated the graph correction pre 1583 and also adjusted the tail end of the graph to reflect urban warming as calculated in Aono’s own follow up report in 1998. An overview and the complete document are here http://knowledgedrift.wordpress.com
The complete document contains a comparison of the corrected reconstruction with IPCC 1990 and a European long term reconstruction that I obtained from Burt Rutan’s Oshkosh presentation. If Siberian larch trees claim there was no Little Ice Age or Medieval Warming Period then the Kyoto cherry trees are calling them liars.
Sigh … nothing’s ever straightforward, is it ?
I accessed the original paper, Aono & Amoto, because I wanted to know how blossom time could match temperature so precisely, to a degree or so – and the paper is in Japanese. Sorry, no can read, not at all
Perhaps there’s a fluent translation for it ?
sorry, forgot about that. there’s an english abstract, a number of english explanations, and the calculus is in well… calculus. I’ve read enough about their work that I may have inferred some things from other articles. I’ll have to go back and check and see what else I can post links to (nothing is ever straightforward, is it? 🙂 )
In brief full flowering of cherry blossoms is very dependent on over all heat over the previous few weeks. you can watch the blossom festivals roll in waves across Japan from south to north and from low altitude to high over a short period of time. Aono/Amoto took the March mean weather station temps in Kyoto and they oscillated in concert with cherry blossom festival dates calculated in days from Jan1. Kyoto weather station records ran only back to 1930, but the correlation was so close that it was statisticaly accurate to a fraction of a degree. Pure genius if you ask me.
Now if only someone had told them that for a 82 years Jan 10 was playing the part of Jan 1, and then for 100 years Jan 9 was playing the part of Jan 1 and then for 100 years Jan 8 was playing the part of Jan 1…
there are two papers linked to in my paper. the second one is almost all Jaspanese, the first one has the english abstract.
Thanks for that, very interesting
As for my curiosity about the temperature/blossom time correlation”
“Pure genius if you ask me.”
Pure genetic serendipity + genuine inspiration from the authors. I tend to believe them on this because when they published they actually thought they had confirmed the HS, so there was no need to “hide the decline”. Your re-assignment of calendar time is remarkable.
Now we have seemingly hard evidence of a MWP & LIA in Japan ! 🙂 A little more widespread than Finland, I think. Probably you should proffer this to the “litachur”
For the record, I don’t think they were trying to sell a hockey stick. Their reports are very factual and make no outrageous claims about future predictions. They even allude to urban warming as a possible contributing factor and four years later went so far as to calculate it and publish it. All of which makes their data very credible. I’ve tried to e-mail Aono but getting the private e-mail address of a major researcher isn’t easy. I bet that he would correct the data himself if he saw my paper.
Click to access kyoto-cherry-blossom-re-analysis-ix.pdf
I think it’s quite worthwhile to keep trying – it’s a very interesting and credible data set, plus your calendar corrections
If I find one I will post it. I’ve been working on another article though. Were you aware that the clouds and the tree rings have been in a conspiracy to mess up the climate data? Never mind which data was used or not or changed or not… there is a hole in the methodology that you could slide an asteroid through.
http://knowledgedrift.wordpress.com/2010/01/05/tree-rings-how-the-clouds-made-them-lie/
David – I suppose you tried here
http://www.envi.osakafu-u.ac.jp/atmenv/aono/AonoHo-e.html
and
http://www.envi.osakafu-u.ac.jp/atmenv/aono/AonoWe.html
containing references to very recent papers
http://www3.interscience.wiley.com/journal/114298147/abstract?CRETRY=1&SRETRY=0
2008
http://www.springerlink.com/content/e13r57081qk84930/
Oct 2009
(email address listed as: aono(at)envi.osakafu-u.ac.jp)
[Sorry Steve – plz delete his email – it’s an academic address given on the publishers sites – if it’s not appropriate]
Actually, I there may be some support for Briffa’s identification of 1032 as the coldest year in the millenium in the Aono/Amoto cherry blossom reconstruction. If you use their original reconstruction or my corrected reconstruction, there is a single outlier on the graph that roughly corresponds to 1032. The scale on the graph isn’t that precise and the data point represents several years data, but it is clear that there was an out of the ordinary (way out) cold year or two about that time. That said, the rest of the data points before and after are screaming MWP as loud as they can.
I wonder why. Call me naive but let me ask a dumb question. This isn’t about a small group that built some boxes and then shoe-horned in the data to fit them, is it? I don’t mean that to be a merely rhetorical question but let me at least assume a null hypothesis. I wonder why.
Does this email relate to (shed more light on) this issue?
http://www.eastangliaemails.com/emails.php?eid=780&filename=1172776463.txt
-Chip
Steve: If we had access to the Powerpoint, maybe.
I mentioned in an earlier reply that the cherry blossom reconstruction for Kyoto is not fragmented like the tree rings are, and provided that you put an off set into it to account for Gregorian calendar implementation, it pretty much tracks the MWA and LIA. I just discovered a reconstruction for Europe based on grape harvest dates. Apparently the Kyoto cherries and the European grapes were in synch:
http://knowledgedrift.wordpress.com/2010/01/06/grapes-of-wrath-the-cherry-blossoms-have-friends/
But the subsequent Briffa et al (2008 Phil Trans B) only includes the very small Yamal data – without the long Polar Urals chronology or the shorter chronologies. Wonder why?
Calling Dr Wilson Re: this idea that the Polar Urals were screened out based on high variance during MWP. Recall K. Fritsch’s demonstration that mean and variance are correlated. What’s your conclusion?
Re: bender (Jan 9 13:15),
My non-quantitative assessment of why I wouldn’t do it that way: too much exclusion, not enough inclusion.
(I know the question is why it was done the way it was)
The year 1032 is an interesting anomaly. In H.H. Lamb’s book, “Climate History of the Modern World”, he has a reference to the Nile River freezing in 1029 AD, presumably from Egyptian sources. Does anyone out there in the peer-to-peer science community know of a volcano that may have popped off in this era?
Re: Dennis Wingo (Jan 9 18:38),
Supposedly, (I wasn’t there), Baitoushan.
Of course, there’s an email.
Aono’s cherry blossom record for Kyoto is posted on his site. It shows the blossoms being late by (if I am reading it correctly) 11 days in 1029 and 6 in 1032. No data for 1030 but the adjacent years they are early by up to a week. Aono’s model which is very good estimates March mean temps vary by about 0.26 degrees per day early or late from average flowering date. So yeah, something nasty happened to the weather in both 1029 and 1032.
Guys
Interesting. However, this paper here from 2007 puts that Volcano at 935 +8/-5 or 912-972 with less exact methods.
http://www.bioone.org/doi/abs/10.3959/1536-1098-63.1.37?cookieSet=1&journalCode=trre
Interestingly enough, Lamb’s book does detail another Nile freezing in the early 900’s, I forget the exact date.
I did a bit of research on this one and no one has made a connection between those eruptions and global climate. If indeed the Nile did freeze in the winter in these years, one would tend to think that as a significant climatic risk. Here is a paper that made a weak connection to that event.
Click to access 1889.full.pdf
Thanks JIm for the reference.
Here is a paper that references this event in Western USA Bristle Cones.
Very interesting dendro information in here as well.
Click to access Salzer_Hughes_2007.pdf
I think what you are looking for is Crowley et al 2003. I don’t have a copy but I’ve seen it referenced in other works. Reconstruction of both solar constant and volcanic modifiers going back to 1000AD. The graphs are reproduced in this report though and I don’t see anything bizarre http://nzc.iap.ac.cn/ewea/images/stories/reference/2009/7/703_peng_qi.pdf
Re: Dennis Wingo (Jan 10 00:43),
1032 CE ‘seemed cold’ to some. Was it the coldest? Right in the middle of the MWP?
Here is an email describing an attitude about a reference like your bioone ref above.
I don’t know that it is less exact per se. I read the abstract. Too small a sample? No tree rings? Correct assessment of error?
I won’t comment on the Salzer-Hughes paper.
Now, what did happen to those Urals?
To me what is interesting here is that these “cool” events in the middle of the MWP are used as a more generalized falsification of the entire concept of the MWP. It is unfortunate that Lamb’s reference for the Nile freezing is incomplete. I think that there is a huge over reliance on tree rings and other proxies and not enough consideration of historical sources of climate information. These can be used as an independent check on these proxies, whether it is the Polar Urals, Yamal, the Western USA bristlecones, or ice cores. Here is an interesting statement from a historical source about Islam of all things.
http://www.thelatinlibrary.com/imperialism/notes/islamchron.html
and the widespread use and perfection of new farming devices. Technological innovations include the use of the heavy plow, the three-field system of crop rotation, the use of mills for processing cloth, brewing beer, crushing pulp for paper manufacture, and the widespread use of iron and horses. With an increase in agricultural advancements, Western towns and trade grow exponentially and Western Europe returns to a money economy.
snip – I discourage attempts to prove climate scienc in a paragraph or two, but do this sort of thing in Unthreaded.
moved to unthreaded but the “proof” was in the link not in the two paragraphs.
What Happened to Polar Urals?
Null Hypothesis: Total Inclusion, All data is ‘there’.
Criteria for Exclusion: 95% confidence interval.
Now, I need to design the experiment. I couldn’t do anything until I got that far.
Expectation of experimental outcome: Null Result
After that, “What Happened to Polar Urals?”
Next experiment.
2 Trackbacks
[…] What Happened to Polar Urals? […]
[…] What Happened to Polar Urals? Addiction of paleoclimate reconstructions to particular proxies has been a longstanding concern at Climate Audit. One […] […]