CA readers know that virtually all of the “independent” IPCC reconstructions purporting to compare modern and MWP temperatures use Graybill strip bark chronologies and/or Yamal. In various posts, problems with strip bark chronologies have been discussed, including discussion of Pete Holzmann’s observation based on our sampling at Almagre that strip bark trees seemed to show a growth pulse after the strip bark event. Indeed, this topic was under discussion in the very thread that the Climategate dossier was first mentioned (though no one noticed this until Gavin Schmidt brought it to our attention.)
In one of the last Climategate emails in October 2003 prior to MM2003, Malcolm Hughes (368. 1065785323.txt Oct 10, 2003) observed that he was “sitting on the bones of a manuscript” reporting the phenomenon of dramatic growth pulses after strip bark formation:
I am sitting on the bones of a manuscript in which I had someone spend
several months checking many hundreds of bristlecone and similar cross-sections and cores in our store. They found only a few dozen – less than 10%, where either pith was present, or the innermost ring could reasonably be described as ‘near pith’. If you have seen these stripbark montane 5-needle pines, and ever tried to core them, you will understand why. A further problem arises from the observation that radial increment may increase rather dramatically in the period after most of the bark dies back, but of course we don’t know when that was.
Now compare this to the CA post:
Note, as reader Erasmus de F observed, the tremendous growth pulse in the surviving part of the trunk immediately following the glacier scar.
If you drilled a core in the center of the surviving “strip” bark in the scarred spruce, you would get a huge growth pulse in the late 19th century; if you drilled a core at the edge of the surviving strip bark, you would get correlated but narrow widths. This is exactly the situation that we hypothesized at Almagre strip bark (our Tree 31 discussed here.) Here’s a ring width plot from the prior post. The glacier-scarred tree would yield a graphic like this:
Hughes observed that the date of the strip bark event was not knowable, but in some cases, dating seemed plausible. In the Miracle post, the event had been dated quite precisely to glacier expansion in the 1840s. It seemed highly plausible that strip barking in a number of Almagre bristlecones had also occurred in the very cold and snowy 1840s. Because the growth pulse was highly nonlinear – six sigma deviations in some cases – it wouldn’t take more than a few such trees in a typical sample to affect the chronology.
Seven years later, Hughes is still sitting on the “bones” of his manuscript describing the post-event growth pulse from strip bark trees.
Climate Audit 
48 Comments
“Seven years later, Hughes is still sitting on the “bones” of his manuscript describing the post-event growth pulse from strip bark trees.”
I have a drawer full of half written manuscripts slowly festering into irrelevance. Most will never get completed. I suspect I am not the only academic with such a collection.
Steve – i have many unfinished projects as well.
Re: richard telford (Nov 11 13:06),
This is one subject which really does need to be published. Perhaps you should offer Hughes the opportunity to publish a paper together on the subject.
Sorry to disappoint, but I have enough of my own manuscripts to write, including some that will be of relevance to those here interested in the robustness of palaeoclimatic methods.
Steve’s remark elsewhere to Jeff Alberts, concerning Ed Cook’comments, is perhaps more in line with the thrust of his article here (imho) :
“[..] it’s as though major people in the field understand the problems, but not if the problems are pointed out by third parties.”
Steve, do you plan to formally publish results of your Almagre data? Serious question; not a TCO nag. The stripbark growth anomaly ought to be known better than it is.
Steve: at some point.
I would like to encourage Malcolm Hughes to complete work on the sudden growth spurts in bristlecones and other trees that they have in their collection. I think this would be a very important work to complete.
Hughes did manage to participate in studies claiming “unprecedented” 20th century growth in bristlecones. As I recall, he didnt mention the strip bark pulse problem. Oversight, I guess.
More likely that he had only a skeletal understanding of the topic.
Look at the AGU abstracts that Hughes and others are presenting this fall. Their PNAS paper is now used to justify using bristlecones as infallible temperature proxies, but relies on comparisons with PRISM data, which are notoriously unreliable for time-series trend analysis. Correlation is still confused with causation in the dendro field.
Spartacus, what makes you say that PRISM is ‘notoriously unreliable for time-series trend analysis’? I use DAYMET rather than PRISM, would you consider it suffers from the same flaws?
PRISM takes any station record available for the particular/month year it’s interpolating. While this works well for spatial analysis, the stations that contribute to the spatial fields for different times can be different. There is also no adjustment for urban effects, station relocations, changes in instruments, etc.
I’m not familiar with DAYMET.
Fair enough, thanks. DAYMET is similar (based on MTCLIM, with a daily time step). I’ve been using it to identify which regions in Austria where some signals of climate change may be unequivically present in forest growth histories. Not an easy search! 😉
I see your points about using interpolations with inconsistent bases for time-series work, but interestingly enough the regions in Austria that seem to show greater relative warming are reasonably homogenous (with many station inputs), and not particularly urban. At the National scale, the trend for the ‘forest’ climate from DAYMET is near enough to identical to the trend from a set of ‘high quality’ stations, and to the downscaled CRU stuff. The ‘forest’ mean is considerably cooler than the ‘high quality station’ mean though (altitude differences).
Logic would suggest that where cool, alpine areas are warming, forests should grow faster and in regions where the warming doesn’t happen, there’s no increase in growth rates over the past 50 years (all things considered). So far, the evidence is still elusive. With so many complicating factors, it’s hard to tie apparent effects to a single cause. Maybe I should be a climate scientist, instead of a forest scientist. 🙂
Are the stations used for the interpolations in Austria located at the same elevations as the forests? Usually that’s not the case (especially if the forests have not been “managed”), and although these exercises can be useful to generate hypotheses, the proper way to test them is to place instruments in the same areas as the trees/forests. It’s a lot of work, and it takes years to get answers, but there are people doing it, mainly in Canada.
Elevation is factored into the interpolations. MTCLIM was developed in Montana as a mountain microclimatology tool, and DAYMET uses similar logic, adapted for the Austrian Alps. I can dig up the reference in AgForMetr journal if you like.
I’ve used it as you suggest, as a hypothesis-generator. The next step is to test it against forest growth responses (treerings and/or inventories). The tricky part will be seperating out the non-climatic influences.
There is also a big European group doing in-forest monitoring including climate (‘ICP Forests’), but they only have a couple of hundred sites across the continent and the time frame is rather short so far.
I don’t think logic tells us much about how plants respond to very slight temp changes, especially when you don’t know all the other limiting factors (precip, nitrogen, phosphorous, potassium, etc…)
The tree physiology guys seem to think that they have a reasonable understanding of how it all works, but for sure there are a lot of other factors that (depending on the site) will have far more impact than temperature. Forest management is the most important, followed by (in Central Europe) atmospheric nitrogen deposition. I’ve modelled Austrian forest growth changes with the BIOME-BGC carbon flux model, and at a national level the simulations suggest that the climate (temperature and precipitation) response over the past 50 years is near enough to zero. There are some regional differences, so now the challenge is to see if there are geographic patterns in forest growth response (measured, not modelled) that match the patterns in climate trends. Ask me again in another two or so years!
Not according to the CRU emails. They explicitly tell us that we only know anything about climate variability via tree response for the last 100 or so years (meaning we have instrumental records of both temp and precip to say what works and what doesn’t), but beyond that, as Ed Cook says, we know f*** all.
I’m sure that SOME of the time trees respond to temp moreso than other factors, but only when temp is the most limiting factor at any given time. Wiithout accurate climatic conditions known for the life of the tree, you simply can’t say whether a tree was responding to temp, water, industrialization, whatever. And if you have that information, then you don’t need the tree rings. The divergence problem is a prime example of why tree rings can’t be relied upon as historic, or pre-historic, thermometers.
That’s a new approach; ‘Arguing from Authority’, based on the CRU emails. 😉 Sorry, couldn’t resist…
I agree completely, that tree rings only measure tree growth. When temperature is the limiting factor to growth, then perhaps they could conceivably be a useful palaeo-temperature proxy. As you say though, other factors must be known or else it’s all rather dubious. But…
It’s reasonably well established that forest growth in central Europe accelerated from about 1960 to the mid 1990s. This is data-driven, and well-reflected by modelling. Our model also reproduces the big hit to forest growth in the 2003 heatwave/drought year (most models do, or else they don’t work very well).
Let’s say we accept that climate (temp and precip) interpolations such as DAYMET are reasonably accurate (open to challenge for sure, but bear with me…). Rising atmospheric CO2 will increase forest growth, but we can say that that is geographically consistent. We know the age of the trees on our study sites, so we can correct for that effect. We have some interpolations of nitrogen deposition in the mid 1990’s (again, open to challenge), and we assume that N deposition rose from almost nothing in pre-industrial times and followed the same curve as CO2 increases (both being products of industrialisation). If we have enough sufficiently precise forest growth data, then it should (I hope) be enough to quantify the effect of each of these influences over the past 50 years. You could consider it perhaps as a multivariate analysis, because each of our 2000-odd study sites has a different combination of impactors. It may be that the temperature response is tiny, perhaps too tiny to seperate from the noise in the data, but we’ll see.
I must disagree though, that there is only a temperature response when temp is the limiting factor. Trees are just a big collection of chemical reactions, most of which work faster when it’s warmer. Also, physiological responses such as spring budburst are very highly temperature sensitive, and increasing the length of the ‘growing’ season could (maybe) have a noticable impact on annual growth. I guess in that sense, temperature is ALWAYS a limiting factor, because we always have a winter when there’s no growth. Maybe, shorter winter = more growth…
We need to define “warmer”. We’re supposedly only talking about a half degree C average over time. Is that enough of a difference to affect tree growth? Considering the daily min and max will vary by 20 or more times that, I just find it difficult to believe that tree rings are going to be that sensitive to that small an average change.
Don’t forget that the tree growth response is non-linear and that there is a temperature optimum: above and below the optimum, tree growth is less. Maybe that is part of the “divergence problem”. The optimum is species dependent and may be dependent of other limiting items (drought, fertiliser, minerals,…). Trees that show the divergence problem now may have had the same problem during the MWP, which makes the 1,000 years proxy quite unreliable…
Oh, and most trees respond really well to CO2!!!! That would mean that other issues would have to get really bad to balance the CO2 fertilization. It certainly hasn’t gotten several degrees warmer to get too hot for optimum, so, maybe it really IS getting cooler?
Interesting that this is an obvious conclusion that is immediately rejected by so-called scientists!!
Now, why might I come to this conclusion? Well, aren’t many rings gathered at the EDGE of the growth area to alledgedly find those most sensitive to the temps? And aren’t those edges generally the COOL edges rather than the downslope or lower latitude WARM edges??? Yup, sounds to me like the rings are telling them something they don’t want to hear!!
Although many variables can affect tree growth, Dr. Parker said he had ruled out all causes for the sustained nature of the recent growth except for warmer temperatures, a longer growing season and the rising level of carbon dioxide in the atmosphere.
Carbon dioxide levels around the research center have increased 12 percent in the last 22 years.
Thanks Ford, yes, I remember this newspaper article (and the journal paper it describes). But also see Foster et al 2010 ‘Evidence for a recent increase in forest growth is questionable’, PNAS vol 107(21) pp E86-E87, in reply.
I know something of the complications of measuring growth in today’s forests – it makes me glad I don’t have to make guesses about the growth of trees long-dead! There’s still not really any ‘consensus’ on why trees seem to be growing faster now, so perhaps it’s rather a bold call at this stage to say that a few centuries-old tree rings are conclusive evidence for recent climate shifts .
And what is the cause of the difference in growth between trees on adjacent mountains?
One is closer to Starbucks?
Furthermore, if the concern is global average temperature on some kind of annualized scale (as commonly articulated), and if, as you say, trees are not growing at a particular time of year, or are more sensitive to temperatures at particular times of year, then in any particular year we cannot know that the entire year was warmer or cooler, simply based on growth for the year.
One could of course argue that this should balance out over many years and over a wide enough sample set, but it should at least be acknowledged that that is an assumption, not something that follows from the data itself.
That’s a very good point. Trees don’t care about global average temperatures (whatever they mean exactly), and probably don’t much care about local annual averages either. At certain times of the year a temperature anomaly or trend could (maybe) be detectable in growth, but to conclude that this anomaly or trend is indicative of an change in the annual mean temperature seems to me to be unjustified. At first glance it seems reasonable that a shorter winter could imply a higher annual mean temperature, but I agree that this should be tested rather than assumed. Maybe it’s also plausible that shorter winters imply a lower intra-annual variation, and hence cooler or shorter summers. I dunno, you’ll have to ask a climate scientist. 😉
I’ve stratified my climate interpolations by season, which my help a little if the tree-growth data is precise enough. Parts of Austria show ‘seasonal warming’ trends in the linear approximations over the past 50 years, others don’t. Apparent warming in one season may be accompanied by warming in another, or it may not. I have a paper on this in the latest issue of the ‘Austrian Journal of Forest Science’, but I guess that may be hard to access for most people.
To be honest, I don’t much care about global average temperatures either. I’m interested in why trees seem to be growing faster. Perhaps my work will add something to the climate debate, but it’s not my focus. As for which side of the debate would wave my papers around as evidence and which side would denounce me as an idiot, I don’t know yet. I guess I’ll find out, when the analyses are finished. 🙂
Here is the index at CO2Science.org.
http://www.co2science.org/subject/g/subject_g.php
Any idea how reliable the papers on CO2 related growth are that they list?
Hi Kuhnkat. I had a quick flick through some of the papers on the website, and there was nothing ‘odd’ that jumped out at me. To properly study a paper and decide whether or not the conclusions are likely to hold outside the boundaries of that particular study is not a simple thing though, so to definitively state whether the papers on your list are ‘reliable’ or not will probably take the global forest science community another few decades. If you’d like to hurry things up, all research grant donations will be gratefully accepted. 🙂
Most people accept that raising CO2 increases growth, but most of the quantification studies have been done on seedlings so there’s still discussion about how this can be scaled up to full-sized forests, and whether there is an ‘upper limit’ for CO2 fertilisation. Another hot topic is if/when the CO2-inspired growth increase will run into speed limits (often water-related). There’s also evidence that increasing CO2 increases trees’ water use efficiency.
This is all a very current research topic. There are a number of rather sexy forest CO2-fertilisation projects happening around the world now (FACE experiments), but it’s still early days. I would say that anyone who unequivicably says that they know all about it, certainly doesn’t.
Re: Eric Anderson (Nov 21 02:20),
Chris,
It’s kind of ironic that the Graybill stripbarked bristlecone pines were collected in response to Itso’s guess that they would be extra responsive to CO2 concentrations. But as it turns out, the stripbark growth pulses are actually primarily mechanical in cause and peter out as the tree gradually heals. But the real irony is that the CAGW people have latched onto them and use them as a supposed proof for global warming when they’re not, even though more study has justified Itso’s contention that CO2 does indeed act as a fertiliser for trees.
Dave, it’s “Idso”, not “Itso”.
Hughes and colleagues are really fighting to justify using bristlecones, based on one presentation this summer
Click to access TolwinskiWard_session3.pdf
Indeed, he is working all the way to show that these are good temperature proxies, probably by mixing normal growth and growth spurt trees to the right extent to match the “localised” temperature record:
Click to access 0903029106.full.pdf
(this all was mentioned at CA: https://climateaudit.org/2009/11/17/salzer-et-al-2009-a-first-look/ )
The same species shows a positive correlation with temperature (and a weak one with precipitation) at the current tree line, but a negative (!) correlation with temperature (and a positive one with precipitatin), 150 meters below the current tree line. No explanation is given for that difference in behaviour.
But have a look at another study by Millar e.a. of different species in SW USA:
http://www.fs.fed.us/psw/publications/millar/psw_2006_millar027.pdf (also mentioned in CA and the NAS panel report), where was found that the MWP was over 3°C warmer that current, with tree stumps well above the current tree line…
The latter study also mentions the change in tree line between the MWP-LIA-CWP. Question is, if the near current tree line species show a positive correlation with temperature, but the same species shows a negative one when 150 meters below tree line, the current treeline trees were certainly a few hundred meters below the MWP tree line, thus in the “negative” zone during the MWP. Thus underestimating the warmer MWP…
As an aside:
Steve M, is it possible to update the proxy use graph at https://climateaudit.files.wordpress.com/2009/01/indepe1.gif with the reconstructions after 2006? And with some indication of which series are suspect HS-shapes and/or updated showing a different behaviour, or used upside down in the reconstruction…
IMO if any corrections are made to problems such as this, they will not come from Hughes and others who are so heavily invested in a particular narrative. The corrections will come from others, possibly young entrants to the field, if the journals will permit publication of papers that do not adhere to the narrative. The journals are a big problem.
There is an earnest desire to find reliable proxies for as far back as the past allows. The key word is ‘reliable’.
In many types of projects, there is a built-in review point – often several – where a go/nogo decision is made. There could be such a point in dendroclimatology, but I am not aware of it.
Would it be seen as a valid audit function of CA to compose a project to draw together the dendroclimatology findings of CA contributors, to an agreed framework, then make a count on a go/nogo? Somebody has to do it, sooner rather than later.
Personally, I have seen enough in the formal literature to cast a nogo vote in the broad sense. But, I also fear that CA might be stepping too far into the policy area with such a project. Any guidance, Steve?
Dem bones, dem bones, dem dry bones …
Dey heard der word of der MM and dey just lay dere.
Interesting to compare with the testimony of David MacKay, chief scientist at our Department of Energy and Climate Change, in his recent reply to Matt Ridley, as tracked by Bishop Hill:
What gives? That’s a genuine question.
If they know how big the uncertainties are, they certainly aren’t telling the public.
But read the comment from Tony B at
“It is an oversimplification to say that dendrochronology is ring counting based on rainfall and the physiology of trees. Many other factors are considered. This is especially true with the old bristlecones, as their growth can be affected by slope gradient, sun, wind, soil properties, temperature and snow accumulation.”
The fellows who invented dendrochronology never imagined it would become a temperature record! Anyone (AnyMann?) who claims this is robust science would have been laughed out of my high school physics class…
Steve: I prefer that people don’t over-editorialize on this. The points been made lots of times.
There’s a danger that the subtlety and tact with with Steve discusses these issues may lead people to assume there is some validity in the science of tree ring thermometry. Such unreflecting trust would be misplaced.
Most amazing to me, tallbloke, is that so many scientists/em> think there is some validity in the (so-called) science of tree ring thermometry.
Congrats to you and McKitrick on the Canadian defeat of the comprehensive climate bill today. I know you had something to do with that!
Steve, I’m posting here because I couldn’t find your e-mail. I request that you reply to the following letter which appeared in The Australian 19 Nov 2010. I might reply, but less effectively.
e-mail address: letters@theaustralian.com.au
Michael Cunningham, posting on CA as Faustino.
“(“Climate scientist ‘regrets’ emails”, 17/11) .
The article asserted that Jones was responsible for “tricking the science to ‘hide the decline’ in global temperatures”. The imputation is that he deliberately manipulated his data to give a false impression about global temperatures and that he is therefore a dishonest person who engages in scientific fraud.
This imputation is made in the article despite the fact that three official inquiries have examined this claim and exonerated Jones, including one by the House of Commons, which concluded: “Critics of CRU have suggested that Professor Jones’s use of the word ‘trick’ is evidence that he was part of a conspiracy to hide evidence that did not fit his view that recent global warming is predominately caused by human activity. The balance of evidence patently fails to support this view. It appears to be a colloquialism for a ‘neat’ method of handling data.
“We are content that the phrases such as ‘trick’ or ‘hiding the decline’ were colloquial terms used in private emails and the balance of evidence is that they were not part of a systematic attempt to mislead.”
It has been apparent for a long time that The Australian is hostile to environmentalism, editorialising about the need to “destroy the Greens”. But regarding the reputation of a distinguished scientist as collateral damage in your campaign is beyond the pale.
Clive Hamilton, Professor of Public Ethics, Charles Sturt University”
Some background: Hamilton was a public service economist, very poorly regarded by his peers; I had public arguments with him in the 1980s on strategic trade theory, where Hamilton misinterpreted Paul Krugman’s work. Hamilton founded the Australia Institute with physicist Ian Lowe, who publicly abused me around 1990 for suggesting that economists should be involved in assessing environmental issues, including AGW. The AI was IMHO a left-wing activist body, for some time mainly promoting the AGW thesis but branching out into wider issues, with Hamilton shifting from economics to being a moral arbiter.
Malcolm Hughes has a new textbook out on Dendroclimatology available on
Amazon. Have not had time to review it yet:
I’d rather read a book on tea leaf reading.
Reading this, I can’t help but reflect on the fact that six sigma growth spurt tree ring data is considered a “robust” scientific method for determining past temperature by some(when everyone knows that it most probably shows rainfall and sunshine amounts). However when actual temperature data measured by high tech thermometers used by the ARGO project shows a profound lowering of the ocean heat content the last several years, the data is “adjusted” because according to the people running the project, some of the readings were just too cold, so they were obviously due to defective instruments and had to be excluded. Were these excluded ARGO readings a six sigma deviation? Were any temp readings that seemed too high excluded?
Makes you want to weep.
TAG (Nov 19 21:22),
Bear poop. Somewhat similar to bullsh*t.
Re:
I’d also like to point out that all life forms live in a series of ‘nows’, which include its local environment. The carbon dioxide, water, nitrogen and other nutrient levels that matter to trees are those where the leaves (and roots) are, not the “background’ level (or regional or global averages) and these levels for carbon dioxide and water are more variable than the background level is seasonally. Also forgotten is that all life forms modify said local environment to enhance their own survival. Do the reconstructions and/or the people who tout them take survival bias into account? Mann, et al, ignore this to their peril, I think.