Click images for more details



Recent posts
Recent comments

A few sites I've stumbled across recently....

Powered by Squarespace
« Sliding science | Main | Valuing "subsidies" »

Tree ring proxies RIP

Well, well. Look what Steve McIntyre has found. After all those years of sceptics calling for tree-ring series to be updated so as to provide out-of-sample validation of their effectiveness as proxies, and all those years of mainstream climatologists telling us how this couldn't happen because of the cost and difficulty, one of the key series in the Hockey Stick and many other temperature reconstructions has finally been brought up to date.

The series in question is Sheep Mountain, prominently featured in The Hockey Stick Illusion as having a hockey stick shape, the blade of the stick allegedly tracking the rise in northern hemisphere temperatures up to 1980, the end of the Hockey Stick reconstruction. Since 1980 we had another 18 years of temperature rises followed by a decade and a half of the pause.

So what has been happening to tree rings on Sheep Mountain been doing in that time? As McIntyre explains, we already knew about the first few years, because Donald Graybill, who collected the original data, did so in 1987.

The original Graybill Sheep Mountain chronology ended in 1987 (rather than 1980) and, though little discussed previously, actually declined quite sharply in the 1980s.

And now that hint of problems being swept under the carpet has been brought out into the open. The update, by Salzer et al, comes in the form of two different updates to the Sheep Mountain record - the northern and southern series (NFa and SFa):

The updated Salzer SFa chronology...shows a dramatic decline from the closing values of the series used in Mann et al 1998.  While the Salzer NFa chronology...  is slightly elevated relative to the SFa chronology and to the millenium mean, its values are also much lower than closing MBH98 values of the Graybill chronology. Both diverge dramatically from the NH temperature. To have kept pace, SFa and NFa chronology values ought to have reached nearly 3, while the SFa chronology has almost reverted to the long-term mean, with several recent values actually below the long-term mean. Perhaps this accounted for the interest in looking at north-facing exposure separately.

Yet mainstream science and government scientists will not acknowledge a problem with the hockey stick or with reconstructing temperatures from tree rings.

PrintView Printer Friendly Version

Reader Comments (78)

Why would they ?

Isn't Steve Mc brilliant.? Calm, collected, concise and precise.

Dec 5, 2014 at 9:23 AM | Unregistered CommenterStephen Richards

> Isn't Steve Mc brilliant.? Calm, collected, concise and precise.

When it's all over I hope he gets the acclaim he rightly deserves. Maybe they could hand over the IPCC's Nobel Prize to him?

Dec 5, 2014 at 9:33 AM | Unregistered CommenterSean OConnor

Someone ought rediscover where Gavin said the whole CAGW is based on paleo

Dec 5, 2014 at 9:50 AM | Registered Commenteromnologos

Any guesses on SkS repost ;-)

Dec 5, 2014 at 9:52 AM | Unregistered CommenterCurious from Cleathropes

Sean OConnor
Talking of which (Nobel Peace Prizes) one Laureate's prediction about Ice Free Arctic by 2014 has *got long to go before he needs to add the obligatory 5 years.


Dec 5, 2014 at 9:57 AM | Unregistered CommenterSandyS

Can I suggest adding Steve's Figure 2 to the post. As one commenter at CA note - "Devastating".

Dec 5, 2014 at 10:01 AM | Unregistered CommenterDerek

The interesting thing here is NOT the chronology.

The interesting thing is that this paper saw the light of day at all. It says, for instance:

...We have shown that approximately 60–80 m of vertical elevation can be sufficient to create a change in the climate response of bristlecone pine. Trees below this elevation are not as effective temperature recorders as trees at treeline. Such fine-scale sensitivity, if present at other treeline sites around the world, would have important implications for chronology development and inferences of past climate variability. ....


...Similarly, treeline samples from differing aspects should not be mixed to avoid problems and uncertainties related to potential 'divergences' and to 'dilution'. Interpretations of existing bristlecone chronologies need to take this into account, particularly when these ring width chronologies are used in climate reconstructions....

and it does NOT say, in the conclusions, something like ...Of course, this should not be taken to negate earlier papers, which obviously prove the existence of damaging human-caused climate change..

In other words, this is a paper casting doubt on the Mann et al 1998 paper, and it has been allowed to be published!

The world is changing....

Dec 5, 2014 at 10:07 AM | Unregistered Commenterdodgy geezer

The paper is “Changing climate response in near-treeline bristlecone pine with elevation and aspect”. One of the authors of the paper is Malcolm Hughes, who is the H in the MBH papers.

Dec 5, 2014 at 10:09 AM | Unregistered CommenterDouglas J. Keenan


Might need to check your ads - I just got an anonymous one telling me that I am missing an important driver - click here".

No details, so no click for details

Dec 5, 2014 at 10:11 AM | Unregistered CommenterAnother Ian


Might need to check your ads - I just got an anonymous one telling me that I am missing an important driver - click here".

No details, so no click for details

Dec 5, 2014 at 10:11 AM | Another Ian
Clickbait. Or worse. ALWAYS ignore anything telling you out of the blue that your machine is missing something/has something out of date. Don't talk to strange adverts...

Dec 5, 2014 at 10:25 AM | Unregistered CommenterJeremy Poynton

How unexpected. Empirical data contradicts a charlatan prone to hissy fits when contradicted. Can somebody remind me where I can find real world data that actually supports the CAGW scam.

Dec 5, 2014 at 10:25 AM | Unregistered CommenterSteve Jones

Perhaps it is time for The Hockey Stick Illusion 2nd edition.

Must be plenty of new material including a chapter on court wars! And Faux Nobel Laureates!

Dec 5, 2014 at 10:27 AM | Unregistered CommenterCharmingQuark

There are quite a few tree-ring based temperature reconstructions (local and large scale) that do not use BP data.
Let's not cherry pick please

Dec 5, 2014 at 10:45 AM | Unregistered CommenterRob Wilson

"There are quite a few tree-ring based temperature reconstructions (local and large scale) that do not use BP data."

Yes, the ones of no interest for advocacy purposes.

Dec 5, 2014 at 10:51 AM | Unregistered CommenterWill Nitschke

I agree with Rob Wilson.

The story here is good and very important, but the headline is far stronger than what is supported by the story.

Dec 5, 2014 at 10:52 AM | Unregistered CommenterDouglas J. Keenan

I love the implication in the title of the paper that they were right all along but that it's the trees that are suddenly doing something different.
I wonder if there's a scientific explanation for that.

Dec 5, 2014 at 10:52 AM | Registered CommenterMike Jackson

There are quite a few tree-ring based temperature reconstructions (local and large scale) that do not use BP data.
Let's not cherry pick please

Dec 5, 2014 at 10:45 AM | Rob Wilson

Good point. Fair point. However, will "climate world" (for lack of a better phrase) take not of Steve's work, or will it pretend it never happened?

Dec 5, 2014 at 11:05 AM | Unregistered CommenterJeremy Poynton

When we did the analysis for D'Arrigo et al (06) - we did not use the BP data as we were concerned with the ambiguities in their climate interpretation. There is no doubt that some of Steve's observations helped us come to that conclusion although we did our own analyses as well. The BP data are far from useless but there are caveats which need to be kept in mind. Bunn/Salzer are going some great work.
anyway - back to marking!

Dec 5, 2014 at 11:16 AM | Unregistered CommenterRob Wilson

Mann snookered as it turns out to be a 'Hickory Stick'

Dec 5, 2014 at 11:16 AM | Unregistered CommenterAnoneumouse

climateaudit is the best there is but gets difficult to access : graphs overflow feeling for the uninitiated
i was wondering for a while what pages2k means??

the site lacks recapitulating articles / whitepapers intros
mci is not "embedded" if he were working in the multiculti corp he would
have had long time ago cardboard woodwork around him.. beautiful wimmin
( but inan empowered good way, tits slightly hidden) african spokesman who come
over authentically andsoforth. more colours where it does not matter.
decorum, quoi. so we can all feel proud and pick a grain on the side (or am i now
describing a quango? )

Dec 5, 2014 at 11:43 AM | Unregistered Commenterptw

Salzers early work clearly showed medieval warm periods in the US while the realclimate crew were still telling us it was just a European phenomenon - a line they irrationally stuck to even when Mann contradicted himself by claiming a relationship between the Carribean MWP and hurricanes. I'd been wondering how fine a line Salzer had to tread with his boss Hughes, since subsequently he actively pushed the CO2-as-a climate-driver dogma in his papers despite his own data telling him the opposite. Of course an academic must protect his/her career by not overtly criticising their mentor but at the same time the less dishonest still want to show the raw data.

So perhaps this mean climateers can now pretty much say anything they like because the catastrophe meme has gained a life of it's own, entirely divorced from the science that ritually contradicts it. Hence the IPCC synthesis report completely contradicts the reports it purports to summarise and no academic or journalist even raises an eyebrow.

It was always ironic that Idso, who co-authored the Graybill paper with the over-weighted hockey-stick proxies, was an ardent skeptic of AGW who, with Graybill, had suggested the anomalous growth that didn't even reflect local temperatures was due to CO2 fertilization. Paleos apparently don't much bother to read the papers that describe the proxies they use, hence the endemic use of proxies upside-down. This has always been more like abstract art than science: Pseuds will always pretend that obvious crap has some worth in order to be in with the in-crowd.

Dec 5, 2014 at 11:48 AM | Unregistered CommenterJamesG

@ Rob Wilson

No cherry picking, the implications of this paper are clear. Nobody should take for granted any claim that any tree rings provide reliable proxy for temperatures.

ALL tree ring-based reconstructions really need extensive decadal, post publication validation., which places all such studies in the dock until the excercise conudcted on the Graybills here is conducted AND PASSED by those used in other studies.

I am sure you will agree.

Dec 5, 2014 at 11:57 AM | Unregistered CommenterGeckko

James G,
Yes, the climateers (I see them as climate obsessed) can pretty much make any claim they wish and as long as it supports the climate catastrophe dogma the claim is OK.

Dec 5, 2014 at 12:00 PM | Unregistered Commenterhunter

In fact, the lesson extends to all proxy measures employed in temperature reconstructions.

Here is a little project I would think would be valuable.

Somebody construct an comprensive matrix of al the major proxy reconstructions against the proxy series employed by them. (imagine, studies listed down the left column and proxies, like the Sheep Mountain Graybills, listed across the top row).

Then identify box by box in the matrix, which proxies have failed a reliable validation like this one, those that have passed such a test and those that haven't had any post publication validation at all. That would give a good stock-take on what studies we can now discard and which studies are still in the "not validated" bucket and which ones have had all proxy inputs validated.

I would love to see the results. My guess would be none fall in the latter catergory.

Dec 5, 2014 at 12:04 PM | Unregistered CommenterGeckko

Tree rings are about a good measurement of past temperatures, as pine cones are a good measurement of future temperatures, in other words rubbish. And once again let’s remember that such proxies are used not because they are known to be good, but because their better than nothing.

Dec 5, 2014 at 12:05 PM | Unregistered Commenterknr

The best proxy for temperature is the clothes people wear :-)

Dec 5, 2014 at 12:16 PM | Unregistered CommenterStacey

Will provide plenty of fodder for Mark Steyn when he questions Mann under oath in the upcoming libel court case, when it finally gets there of course - Mann is doing everything possible to delay/prevent it.

Dec 5, 2014 at 12:45 PM | Unregistered CommenterIlma

It has been obvious for a long time that the vast majority of raw proxies showed nothing at all and that only a very few had a hockey-stick - each with a reason; used upside-down, contaminated, strip-bark whatever. If your reconstruction shows a hockey-stick when virtually none of the inputs do then it must be due to biased data-mining. This is of course bleeding obvious to engineers but not apparently to paleos - who seem to be largely innumerate.

Dec 5, 2014 at 12:49 PM | Unregistered CommenterJamesG

While he is to be commended for engaging in this blog, I'm afraid I find the comment above by Rob Wilson horribly disingeneous. He comments as if this discovery by Salzer et al and highlighted by Steve McIntyre is no big deal, and that in reconstructions he co-authored, brstlecone pine trees were not used.

That is to miss the point by a mile. Surely Dr Wilson understands that the fraudulent hockey stick became a poster child for global warming. We did not see D'Arrigo et al on the front of IPCC documents, we saw the Mann et al version.

Does he condone the montage of a reconstruction achieved through contrived methodology together with modern temperature data, various other undocumented tricks, and no acknowledgement of the divergence issue or does he not?

Will he stand up and defend proper science or not?

Does he condone the use of such dubious work for the creation of energy policy that is costing individuals in such a way?

As has been said many times, "All that is necessary for the triumph of evil is that good men do nothing".

Would it not be honourable to advocate clean work in one's own field?

Dec 5, 2014 at 1:34 PM | Unregistered CommenterKeith

The thrust of Steve Mc's post is that "out-of-sample" validation of the tree-ring proxies was needed to validate the models which have been used to create a temperature reconstruction from them. What must be remembered is that the proxy record is a model and the proxies have not just been added up, but been tuned by differential weighting to a certain time period (the overlap with the instrumental record), and then extended into the past to build a reconstruction of temperatures prior to the record. This is a model - not data - and it needs to be validated with data collected out of the tuning range. This is what has now been done for the most important data set (by most important, I refer to the proxy which was given the highest weighting in the model). Steve has previously shown that the backwards extension of the model was done using poor/inappropriate statistical methodology, but now the actual model has been falsified by using new data.

Now, you could split hairs and say that the new data does not invalidate tree rings as a proxy, but this new data showing that one data set does not fit the model "out-of-sample" invalidates the model on which the hockey stick was based. In that sense, I am quite happy with the headline - one can allow a little hyperbole here and there.

Dec 5, 2014 at 1:44 PM | Unregistered CommenterRob

Interesting. We should also note that hockey sticks occur in lots of non-tree ring proxies too (permafrost temperatures, deep sea cores using forams, glacier length records etc.).

Based on these from a huge range of different groups using a range of different techniques and from a range of different locations, it looks like Mann was generally right after all!

Dec 5, 2014 at 2:01 PM | Unregistered CommenterMonty


"it looks like Mann was generally right after all!"

A wonderfully surreal conclusion. Is your surname "Python" perchance?

Dec 5, 2014 at 2:17 PM | Unregistered CommenterJames Evans

James: You do accept the main point of my post? That " We should also note that hockey sticks occur in lots of non-tree ring proxies too (permafrost temperatures, deep sea cores using forams, glacier length records etc.).

Based on these from a huge range of different groups using a range of different techniques and from a range of different locations".

Do you accept this?

Dec 5, 2014 at 2:21 PM | Unregistered CommenterMonty

Dec 5, 2014 at 2:01 PM | Unregistered CommenterMonty

"Right answer", wrong method = bad science. N'est ce pas?

Dec 5, 2014 at 2:27 PM | Unregistered CommenterRobert Austin

As per usual their are multiple thrusts to Steve McIntyre's piece. The most damning is the parallels drawn between those who attempt to model the stock market and paleoclimatologists, i.e. data snooping and dust-bowl empiricism.

Dec 5, 2014 at 2:28 PM | Unregistered Commenterbernie1815


I'm sure there are all sorts of data series with all sorts of shapes. There's certainly an eclectic variety of shapes that get bundled into the palaeo studies that we've all become so fascinated by.

Dec 5, 2014 at 2:42 PM | Unregistered CommenterJames Evans


Good point. For years financial/economics types have been pointing out the folly of those proxy studies based on their knowledge of time series analysis and modelling natural processes (in their case prices typically). And Alarmists have always poo-pooed them as not knowing about "climate" - when of course th eprocess is irellevant on the knowledge of data processing and time series was.

The next are where Alarmists make the same claims? Climate model. Economists have for years attempted to build realistic macroeconomic models and know it is impossible to do so to the extent that they can tell us anything reliable about the future. But claimte "scientists" insist that their models "are different", when the only difference is that it is a different natural process being modelled, but the impossiblity of doing so with any degree of qunatitiative accuracy or reliability is the same.

Dec 5, 2014 at 2:48 PM | Unregistered CommenterGeckko

Great work, as usual, by Steve McIntyre..

In any other branch of science, apart from paleoclimatology, these observations would be a show-stopper.

I guess I shouldn’t get my hopes up, though.

The final sentence in the abstract sums it up nicely.
“This suggests the possibility that the climate-response of the highest South-facing trees may have changed and that temperature may no longer be the main limiting factor for growth on the South aspect. These results indicate that increasing warmth may lead to a divergence between tree growth and temperature at previously temperature-limited sites.”

Lovely bit of genuflection to the global warming paradigm.

In other words “business as usual”.

Dec 5, 2014 at 3:32 PM | Unregistered CommenterBitter&Twisted

Rob Wilson observes: "There are quite a few tree-ring based temperature reconstructions (local and large scale) that do not use BP data. Let's not cherry pick please." I firmly reject ROb's suggestion that it is "cherrypicking" to comment on out-of-sample validation of Graybill chronologies as a temperature proxy.

I have had a longstanding interest and challenge in out-of-sample performance of proxies used in important multiproxy reconstructions and it is highly reasonable for me to comment on an article providing fresh information on one of the most important such proxies. In the scope of a single blog article commenting on Salzer et al 2014, I did not attempt to survey all multiproxy reconstructions.

Yes, there are some multiproxy reconstructions that do not rely on stripbark BCPs, but many of them have other equally troubling problems. For example, Mann's nodendro recosntruction relied on upside-down contaminated varve sediments. Many other recent multiproxy reconstructions rely on varve series from Kaufman et al 2009 and similar studies. I have criticized varve series in many CA articles. There are many interpretation problems with these series - such that even specialists can make mistakes on which way is "up" - as occurred in the recent PAGES2K multiproxy study and previously in Mann et al 2008, Kaufman et al 2009 and Tingley and Huybers 2013. I will be astonished if any of these reconstructions validated out of sample in a resampling 20 years from now.)

Rob's comment also did not make clear that problems with bristlecones impact numerous canonical IPCC studies, not just Mann et al 1998-99. Other studies impacted by stripbark bristlecones and foxtails include: Hegerl et al 2007, Esper et al 2002, Mann and Jones 2003, Rutherford et al 2005, Juckes et al 2007, Mann et al 2008 CPS-L, Mann et al 2008 EIV-F, Mann et al 2008 EIV-L, Mann et al 2009 EIV, Shi et al 2013 -PCAR. Stripbark BCPs even impact Southern Hemisphere reconstructions in the AR5 spaghetti graph.

Rob's comment in no way refutes or undermines the conclusion of my post: the graybill bristlecone chronologies do not validate out of sample as a temperature proxy. I further observe that, in 2006, the NAS panel recommended that stripbark chronologies not be used in temperature reconstructions, but IPCC and many specialists thumbed their noses at the NAS panel recommendation and use of stripbark chronologies subsequently increased. Perhaps it's time for Rob and other specialists to speak out against the practice.

Dec 5, 2014 at 3:46 PM | Unregistered CommenterSteve McIntyre

Steve - my comment was not specific to your post - more to individuals using one study to vindicate their blind belief that tree-rings cannot be used to reconstruct past temperatures.

Dec 5, 2014 at 4:09 PM | Unregistered CommenterRob Wilson

Monty, the only way other proxy series show the so-called hockeystick is if they are processed through methodologies similar to the Mann. So no. There are no meaningful temperature records that, processed legitimately, indicate a troubling climate experience ocurring on in the current era.

Dec 5, 2014 at 4:37 PM | Unregistered Commenterhunter

Dec 5, 2014 at 4:09 PM | Unregistered CommenterRob Wilson

I didn't see a comment cherry picking a study before yours (Steve Jones maybe) so you must have been replying to the post itself.

Do you agrre with Steve's main point??

"Rob's comment in no way refutes or undermines the conclusion of my post: the graybill bristlecone chronologies do not validate out of sample as a temperature proxy."

Dec 5, 2014 at 5:28 PM | Unregistered CommenterRob Burton

James Evans: Monty 'Python'? Nah, I think it should now be Monty - Owned.

Funny thing though, I read SMc's thread over at CA: I don't recall seeing Monty over there?

Dec 5, 2014 at 5:28 PM | Unregistered CommenterHarry Passfield

typical hypothesis in climatology : hypothesis is true... well hypothesis is consistent with we can conclude hypothesis is true..
The less evidence the more hypothesis is valid...

Dec 5, 2014 at 5:37 PM | Unregistered Commenterlemiere

Andrew, a fascinating and illuminating conclusion.

Mind you, it wasn't as if ...we didn't expect it and after all, Steve told us it would be thus................

Dec 5, 2014 at 6:16 PM | Unregistered CommenterAthelstan.

I greatly appreciate your dropping by.
You said that your comment was directed at those "individuals using one study to vindicate their blind belief that tree-rings cannot be used to reconstruct past temperatures". I am not sure it is a blind belief, but for the sake of argument can you point me to a recent published study of tree-rings that have met the requirements of an out of sample validation? This approach is certainly a best practice in my area of psychometrics, i.e., test development - though it is also frequently ignored.

Dec 5, 2014 at 6:35 PM | Unregistered Commenterbernie1815

The fact that the world didn't end on Dec 21 2012 can't be used to vindicate the blind belief that the Mayan calendar cannot be used to predict when the world will end

Dec 5, 2014 at 6:50 PM | Registered Commenteromnologos

The problem, surely, lies with those who have a 'blind belief' that tree rings CAN be reliable proxies for temperature. It is for them to demonstrate the reasonableness of their belief. A good start would be for Rob Wilson to explain why, in the case under discussion, anyone should accept the tree rings as a reliable proxy pre 1980 when they clearlly have not been since.

Dec 5, 2014 at 7:23 PM | Unregistered CommenterDaveS

Rob Wilson, thanks for participating. I'll keep an open mind regarding correlations between temperature and tree rings. And between tree rings and CO2. insolation, elevation, snow accumulation, microclimate, local hydrology, biota decomposition, mineral content, volcanism, forest fires, bacteria, fungi, insects, ozone, precipitation, humidity, deadfall, and moose piss. Keep coming back!

Dec 5, 2014 at 8:33 PM | Unregistered Commenterjorgekafkazar

We should also note that hockey sticks occur in lots of non-tree ring proxies too (permafrost temperatures, deep sea cores using forams, glacier length records etc.).downwards pointing ones are ignored in paleo studies. (Or flipped, in the most egregious cases.)

Dec 5, 2014 at 9:53 PM | Unregistered CommenterMooloo

PostPost a New Comment

Enter your information below to add a new comment.

My response is on my own website »
Author Email (optional):
Author URL (optional):
Some HTML allowed: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <code> <em> <i> <strike> <strong>