Another reason trees don't make good treemometers – new tree ring data bias discovered

[Note: My first post in which I had written commentary mysteriously lost all of its content, posting nothing but white space. This is some sort of internal wordpress error, but has never happened before. I have some elements restored below, but my original commentary is lost. -Anthony]

I’ve written before about the difficulties associated with extracting a valid temperature signal due to all of the confounding variable in Liebigs law of the minimum, which I describe in detail here:  A look at treemometers and tree ring growth

Now a new confounding variable has been introduced that does not bode well for tree ring studies such as Mann et al.

Bishop Hill writes:

A new paper by Brienen et al in the journal Global Biogeochemical Cycles suggests that there may be a whole new set of biases in tree ring studies.

Tree ring analysis allows reconstructing historical growth rates over long periods. Several studies have reported an increasing trend in ring widths, often attributed to growth stimulation by increasing atmospheric CO2 concentration. However, these trends may also have been caused by sampling biases. Here we describe two biases and evaluate their magnitude. (1) The slowgrower survivorship bias is caused by differences in tree longevity of fast- and slow-growing trees within a population. If fast-growing trees live shorter, they are underrepresented in the ancient portion of the tree ring data set. As a result, reconstructed growth rates in the distant past are biased toward slower growth. (2) The bigtree selection bias is caused by sampling only the biggest trees in a population. As a result, slow-growing small trees are underrepresented in recent times as they did not reach the minimum sample diameter. We constructed stochastic models to simulate growth trajectories based on a hypothetical species with lifetime constant growth rates and on observed tree ring data from the tropical tree Cedrela odorata. Tree growth rates used as input in our models were kept constant over time. By mimicking a standard tree ring sampling approach and selecting only big living trees, we show that both biases lead to apparent increases in historical growth rates. Increases for the slow-grower survivorship bias were relatively small and depended strongly on assumptions about tree mortality. The big-tree selection bias resulted in strong historical increases, with a doubling in growth rates over recent decades. A literature review suggests that historical growth increases reported in many tree ring studies may have been partially due to the big-tree sampling bias. We call for great caution in the interpretation of historical growth trends from tree ring analyses and recommend that such studies include individuals of all sizes.

Presumably, this new source of bias applies just as much to tree ring studies where the increase in growth is ascribed to temperature.

==================================================================

Here is the abstract from GLOBAL BIOGEOCHEMICAL CYCLES, VOL. 26, GB1025, 13 PP., 2012 doi:10.1029/2011GB004143

Detecting evidence for CO2 fertilization from tree ring studies: The potential role of sampling biases

Key Points

  • Observed increases in tree ring widths may be caused by sampling biases
  • Standard sampling methods lead to spurious trends in historical growth rates
  • Reported increases in ring width may often not be due to CO2 fertilization

Roel J. W. Brienen

School of Geography, University of Leeds, Leeds, UK

Programa de Manejo de Bosques de la Amazonía Boliviana, Riberalta, Bolivia

Emanuel Gloor

School of Geography, University of Leeds, Leeds, UK

Pieter A. Zuidema

Programa de Manejo de Bosques de la Amazonía Boliviana, Riberalta, Bolivia

Ecology and Biodiversity, Institute of Environmental Biology, Faculty of Science, Utrecht University, Utrecht, Netherlands

Forest Ecology and Forest Management, Centre for Ecosystem Studies, Wageningen, Netherlands

Tree ring analysis allows reconstructing historical growth rates over long periods. Several studies have reported an increasing trend in ring widths, often attributed to growth stimulation by increasing atmospheric CO2 concentration. However, these trends may also have been caused by sampling biases. Here we describe two biases and evaluate their magnitude. (1) The slowgrower survivorship bias is caused by differences in tree longevity of fast- and slow-growing trees within a population. If fast-growing trees live shorter, they are underrepresented in the ancient portion of the tree ring data set. As a result, reconstructed growth rates in the distant past are biased toward slower growth. (2) The bigtree selection bias is caused by sampling only the biggest trees in a population. As a result, slow-growing small trees are underrepresented in recent times as they did not reach the minimum sample diameter. We constructed stochastic models to simulate growth trajectories based on a hypothetical species with lifetime constant growth rates and on observed tree ring data from the tropical tree Cedrela odorata. Tree growth rates used as input in our models were kept constant over time. By mimicking a standard tree ring sampling approach and selecting only big living trees, we show that both biases lead to apparent increases in historical growth rates. Increases for the slow-grower survivorship bias were relatively small and depended strongly on assumptions about tree mortality. The big-tree selection bias resulted in strong historical increases, with a doubling in growth rates over recent decades. A literature review suggests that historical growth increases reported in many tree ring studies may have been partially due to the big-tree sampling bias. We call for great caution in the interpretation of historical growth trends from tree ring analyses and recommend that such studies include individuals of all sizes.

Advertisements

  Subscribe  
newest oldest most voted
Notify of
trbixler

Where is the treemometers talk about hide the decline. WordPress?

Justthinkin

Oooookay. Am I going nuts here,Anthony? I see a title,tags,posted in,etc,but NO story. Oy.Do I need more caffiene or watt?

All that is showing right now:
Another reason trees don’t make good treemometers – new tree ring data bias discovered
Posted on March 20, 2012 by Anthony Watts
Ads by Google
Digi-Key
Distributor of Quality Electronic Components. Order from Digi-Key® .
http://www.digikey.com
Data Center Service
100% SLA, SSAE 16 Type II ITIL Facilites – Carrier Diverse
go.SunGardAS.com
Solar Panel Blowout Sale
Solar Panel Up to 53% Off 2012 Winter Sale On Now!
Comparestores.net/Solar-Panel
now that’s positivenergy™
Tell Us How You’re Making the World a Positive Place & Inspiring Others
nowthatspositivenergy.com/Energizer

Did you forget your content?
cheers,
gary

…and it’s worse than we thought.
Hey, *somebody* had to say it.

I think you got ahead of yourself and posted before givng us something to read.
REPLY: I think you got ahead of yourself and jumped to conclusions – Anthony

dorsai123

by my count tree growth is effected by at least 4 variables: sunlight, temperature, water and soil/nutrients … since we have no historical data for sunlight, water or nutrients it is impposible to calculate historical temperature using tree growth rates … impossible … Why is this still being debated by anyone ?

in a previous thread, a poster worried that all science might be tarnished by the impacts of the bogosity that has overrun the field in the last 20 years.
Well, all science DESERVES to be tarnished for having let this nonsense go on this far. This has happened because the scientific “establishment”, such as it is, has allowed it to happen and left it up to a handful of outsiders to point out what is wrong. Until “science” starts taking serious steps to point out how compromised this entire warmista project has been, they will be rightly seen as nothing but another special interest group packed with rent-seeking apparatchik’s willing to say and do anything for money and status.

KNR

It’s worth noting that for all their self declared ‘vast knowledge ‘ there is not one amongst ‘the team’ who actual understand plant physiology well.
We been here before of course with statistics, were people with far more expertise in an area can show us how the ‘the Team’ can be totally out of its depth and its only the arrogance ,that seems to be part of being a ‘climate scientists ‘, that stops them admitting it and accepting advice form those outside their little club.

Michael D Smith

Anthony,
Try the “Lazarus” add-in for firefox. If you are doing your wordpress editing from firefox, you will be able to recover text from various text boxes even if you have navigated away. The length of time to save is configurable. You want the status bar visible to do this. To recover text, just right click the same text box and see your choices of what has been saved. Mine is set to save for 4 weeks.
Strangely enough, I just had to use it for this post since I wasn’t logged in to wordpress…

spence

If it’s too cold, trees have sluggish growth.
If it’s too hot, trees have sluggish growth.
That sort of sums it up, trees are not thermometer material

Doug Proctor

Will Mann et al review this peer-reviewed paper? He should be e-mailed a copy and asked for his comments.

KenB

Wait for the push-back, nothing these days must be allowed to challenge or interfere with our carefully constructed and cultivated tree rings. The ready response team’s “bogus is as bogus does team”, will fix this even if they have to turn the tree ring data upside down!!

Chuck

So many variables influence tree growth that one has to wonder if tree rings can be interpreted as proxies for anything. Nature is seldom as simple as some would like to think it is.

Bill Marsh

I think we’ll get the standard “The science is settled” statements and the authors can look forward to people going through their trash looking for their connection to Big Oil. I suppose though that they’ll just use the new methodology and make up a fake document to ‘prove’ the link.

Given the wide variety of influences on any growth of anything over time; it is simply a fools journey to do anything but the most generalizing about it. Add such confounding elements as variations in genetics and the variations of genetic responses to variations in stimulation… this is all simply a waste of time and time is money.

Dave

Not sure what impact it has… but in reference to the figure, an R2 value of 0.39 indicates little/no correlation between the trendline and the data.

Forget about tree rings, too many factors.
Use the changes in the geomagnetic field
http://www.vukcevic.talktalk.net/HmL.htm

My early career was a series of “discussions” with the PTB, at the time, that most of their sampling methods introduced so much error in their studies that they were worthless…(I was not all that popular, as you can imagine.) Once I started to produce reports that showed ways to save money by accurate and precise tests that were NOT biased by sampling error, it fixed a lot of problems. 🙂
That was in natural products so I can appreciate the difficulty inherent in dealing with anything that is hard to randomize when selecting samples. That does not, however, condone the practices of the GISS gang and their cohorts.

Well, darn, I like counting tree rings. I remember being astonished once when a wind storm toppled two douglas firs that were the same size on my property. They were next to each other. One had thirty rings, one had sixty. I still don’t know how that happens.

Note the article isn’t talking about temperature at all (or precipitation, etc.) — just CO2. Presumably this is because it can be pitched to counter arguments that CO2 is beneficial, and hence it qualifies as OK Sceance with the granters-that-be. Also note that it is modelling; no trees were actually harmed (cut or cored) to crank out this study.

Gary

Let’s be scupulous here, this study is based on models: “We constructed stochastic models to simulate growth trajectories based on a hypothetical species with lifetime constant growth rates and on observed tree ring data from the tropical tree Cedrela odorata.” If we criticize the climate models for their weaknesses, we need to do the same here. This model very well may point out a new bias but we need much more proof that in fact it makes a difference. One might hypothesize that slow-growing old trees of the same species in the same climatic zone respond to environmental factors in relatively the same ways. If so, the bias of fast-growing young trees may make no substantial difference.

Color me crazy but an interesting experiment would be to place a high resolution thermometer by a tree for about 20 years with a data recorder and then compare the derived vs actual temperatures?
Just saying….

Kelvin Vaughan

I’m surprised we haven’t seen newspaper headlines: “Global warming causes trees to put on weight”?

woodNfish

“Several studies have reported an increasing trend in ring widths, often attributed to growth stimulation by increasing atmospheric CO2 concentration.”
That is the first time I’ve ever heard that. Ring width was always attributed to precipitation. What this means to me is that none of these “scientists” know what they are talking about and it is just another reason to be skeptical about anything they say.

Grumpy Old Man

Tree ring data have many problems, not least biased sampling. The bias may be induced by the researcher or the very nature of the material examined as explained so well in this post. But tree ring data can be useful. In Baille’s work, he shows a narrow tree ring occurring over several areas in the first half of the 6thC which suggest a major event leading to global cooling (and destruction) for one year. This may be volcanic but Baille argues for an extra terrestial event backed up by ammonia readings in the Greenland icecores. The use of tree rings depends very much on the nature of sampling and the context (other links) into to which they are examined. Don’t reject tree rings out of hand. They must be considered scientifically (which I have to admit, certain persons have not).

DonS

wws says:
March 20, 2012 at 8:47 am: HEAR, HEAR. The unvarnished truth.

pat

Couldn’t they have just asked a botanist? And I find this intriguing: “Reported increases in ring width may often not be due to CO2 fertilization”. The effect of CO2 on biomass is well established. Are they saying the rings, however, are not affected or the reportage is flawed?
I get the feeling that the treemometer readings are about to get ‘homogenized” much the same way thermometers were. And the underlying data will disappear and the past will have been much colder.

was said, “Until “science” starts taking serious steps to point out how compromised this entire warmista project has been, ”
Unlikely to ever happen as they would also be admitting to knowingly wasting (stealing) many billions. We would and should have to sue them for every penny to be returned, even if they have to hock their future generations that they were planning to cook with global warming.

Hoser

spence says:
March 20, 2012 at 8:55 am

On the contrary, trees might work as a themometer if you find a location where conditions are always on only one side of the tree growth rate optimum (inverted U shape response curve). Danger Will Robinson. If temperature transgresses during some period of time to the other side of the optimum, the data may appear to indicate the opposite condition. Thus, you would need to compare multiple locations further away from the optimum growth conditions to determine whether warming or cooling occurred at some time in the past. Gosh this isn’t so hard. Let’s just model it. Why should we bother with sampling? ;->

Rob Dawg

It goes without saying wordpress must die. That said the evidence is equally as compelling that climate alarmism must die. Publication bias exactly parallels the tree ring data. The fast growing but unrepresentative elements of emerging climate science get undue attention while the longer lived yet less attention grabbing elements such as solar output, cloud cover, orbital divergence, etc. continue to explain ever more of the totality.

Well, yes, that’s just one more reason why people should regard treeometers as a work of fiction. Biases, environmental factors for growth, and the fact that the trees only grow 6-8 weeks out of the year and cannot possibly be representative of the year except by chance, means that all of this madness needs stricken from the books and people should be made mockery of to the point of shame.

Paul Coppin

You know, I have had internet hook-up problems through Firefox, ever since I let it talk me into updating to FF10. Haven’t done the upgrade to 11 yet. Symptoms are – can’t hold or loses logins, won’t go off to urls, etc , having to reload and recycle urls to get the browser to address the web. Additionally, I recently added Facebook hookups too, and when Facebook is online with me, it gets in the way of a lot of link completions. Facebook and wordpress are not happy together – WP admits it. You might be running into issues if you’re autoposting to Facebook. I do on my systems, depending on which email address is logged in to FB (paging purposes).

Paul Coppin

Nothing particularly wrong with wordpress – its all of the corporate data miners it has to interract with that cause most of the problems.

Bill Parsons

Re:

Max Hugoson says:
March 20, 2012 at 8:09 am
All that is showing right now:
go.SunGardAS.com
Solar Panel Blowout Sale
Solar Panel Up to 53% Off 2012 Winter Sale On Now!
Comparestores.net/Solar-Panel
now that’s positivenergy™
Tell Us How You’re Making the World a Positive Place & Inspiring Others
nowthatspositivenergy.com/Energizer

You’re doing better than I. Last weekend, Mark Zuckerberg’s smiling, youthful visage popped up onscreen below the opening post, urging me to get in on some deal “before Facebook goes public”… thought I was in a time warp. [With all the money Zuckerberg made, you’d think he could clean up some of that cr*p!]

Michael Mann's Mum

This is a baseless and fraudulent attack on my son. The conclusions the authors have come to are without foundation. My son’s data, which the authors sought repeatedly and which he refuseed to give them, shows clearly that they are mistaken. I can only assume that this is a witchunt against my little boy planned and paid for by Big Al (sorry, Oil). I can assure you that my son will immediately seek the dismissal of these authors from any posts that they have in which they might earn their meagre living. He will also be asking Phil Jones to email his son to see if he knows these charlatans so that pressure can be applied from above. There is a silver lining, I suppose. My boy will now be able to tour lecturing on why he is a victim of these nasty nasty people.
I myself will personally get straight on to that nice water expert and ask him to get to the heart of this scurrilous slur as quickly as possible. He has proved quite reliable recently in obtaining dirt on evil people likek this.
Did I mention that my little cherub is a climate change hero?

Matt Skaggs

The decline of “hide the decline” fame was almost certainly the result of two things, (1) some sort of sampling bias very much like what is described here, and (2) a growth curve standardization algorithm that apparently overcompensated for the bias on the outermost rings of the youngest trees in the population. Layman Lurker and Dr. Mauri Timonen have done quite a bit of excellent work on this puzzle, but it remains somewhat less than fully resolved AFAIK.

Paul Coppin

Dennis Wingo (@wingod) says:
March 20, 2012 at 9:30 am
Color me crazy but an interesting experiment would be to place a high resolution thermometer by a tree for about 20 years with a data recorder and then compare the derived vs actual temperatures?
Just saying….

If you want to do this properly, find a statistically valid number of similar pine plantations – we use to have them up here, but they’ve fallen out of favour – plenty still around. Equip about 100 trees in about five(?) plantations with instrumentation that measures continuously ambient temp, absolute humidity, rainfall, and photometry, and run it for about 5 years. Core them all at the start and at the end in the same location. Pick trees of the same DBH within each plantation, and with the same ambient light level, and no tree should be further than 50 feet from one another – ideally the array of measured trees would be spatially even.
Then, you might have some idea of what correlates with what.

Justthinkin

“Presumably, this new source of bias applies just as much to tree ring studies where the increase in growth is ascribed to temperature.”
Not really. The bias attributed to temperature is for 2 reasons only….money and power. Or maybe the fact the “scientists” tried to use ONE sample out of how many billions of billions of samples possible/available?
My only gripe…why the heck did they have to use a model(for above study)? sigh

Steve McIntyre
NW

I proposed this kind of sampling bias mechanism on a CA thread more than two years ago. (Pats self on back.) I’m happy to see that someone had the same idea and actually developed it into a paper.

GeologyJim

This article will have no effect on Mann’s perception of the significance of his own work.
After all, he didn’t do the sampling (It’s soooo difficult to update chronologies, ya know, unless you’re Steve McIntyre with a Starbucks in hand)
And he can still rely on the Tiljander sediment proxies, even if he did use them upside-down
“Robust! Robust, I tell you!!”

OT
M 7.6 earthquake hits Mexico near Acapulco.
http://earthquake.usgs.gov/earthquakes/recenteqsww/Maps/10/260_15.php
one more strong earthquake in the wake of the recent solar storms.
http://www.vukcevic.talktalk.net/EQM7.htm

JJ

Gary says:
Let’s be scupulous here, this study is based on models: “We constructed stochastic models to simulate growth trajectories based on a hypothetical species with lifetime constant growth rates and on observed tree ring data from the tropical tree Cedrela odorata.” If we criticize the climate models for their weaknesses, we need to do the same here.

You are talking about two completely different kinds of “models”. Climate models are intended to replicate real world conditions from the past, and are held to be predictive of the future. That is what we normally think of when we say “modeling” in the context of climate discussions.
The “models” discussed here are entirely different. They are not attempting to replicate anything in the real world, and certainly not to predict anything about the future. What they are doing is creating a hypothetical circumstance, and using it to test the denrdro sampling methods. It is very much like feeding Mann’s algorithym with Major League Baseball scores, and checking to see if it produces hockey sticks.
This model very well may point out a new bias but we need much more proof that in fact it makes a difference. One might hypothesize that slow-growing old trees of the same species in the same climatic zone respond to environmental factors in relatively the same ways. If so, the bias of fast-growing young trees may make no substantial difference..
Their point has not anything to do with comparisons between slow growing trees. The question is: what if fast growing trees have shorter life spans than slower growing trees? How would that effect dendro reconstructions? Answer: Fast growing trees would tend to be underrepresented in the early time periods, as those trees would have died before the sampling event. The closer to the time of sampling you get, the higher the proportion of fast growing trees that get included in the sample. This would make it appear that growth rates are increasing, even when they are not.
Ditto the question about big tree bias – if your sampling scheme only uses trees that are larger than some minimum size, how might that effect results? Answer: Disgarding small trees from your sample removes young trees that are slow growing, but keeps young trees that are fast growing. This causes fast growing trees to be overrepresented in the more recent time periods of the sampling. This would make it appear that growth rates are increasing, even when they are not.
The only purpose of the “stochastic models” used here is to quantify the potential magnitude of the bias in the results produced by these sampling methods, by feeding those methods a dataset with known properties. They aren’t “models” in the same sense that we use when referring to climate models.

JimInIndy

More than a decade ago, I read a Nat Geo article on wet and dry cycles in the American Southwest, using bristlecone rings as proxies; wide = wet. By chance, the same month, I saw a study (don’t recall the peer-review pub.) using the same area, same trees as proxies for heat cycles; wide =hot. I haven’t trusted dendro-proxies for anything since.

Dennis Wingo (@wingod) says:
March 20, 2012 at 9:30 am
Color me crazy but an interesting experiment would be to place a high resolution thermometer by a tree for about 20 years with a data recorder and then compare the derived vs actual temperatures?
Just saying….

This has been done. This is what “The Team” refers to as “the divergence problem.”

Central England annual & tree growing season temperatures
http://www.vukcevic.talktalk.net/ATg.htm
However this doesn’t take into account other important factors affecting rate of trees growth such as degree of the soil moisture and air humidity.

Michael D Smith said @ March 20, 2012 at 8:51 am

Anthony,
Try the “Lazarus” add-in for firefox.

Thanks for that tip 🙂

woodNfish said @ March 20, 2012 at 9:34 am

“Several studies have reported an increasing trend in ring widths, often attributed to growth stimulation by increasing atmospheric CO2 concentration.”

That is the first time I’ve ever heard that. Ring width was always attributed to precipitation. What this means to me is that none of these “scientists” know what they are talking about and it is just another reason to be skeptical about anything they say.

Try these:
Graybill, D.A. and Idso, S.B. 1993. Detecting the aerial fertilization effect of atmospheric CO2 enrichment in tree-ring chronologies. Global Biogeochemical Cycles 7: 81-95.
Idso, S.B. 1995. CO2 and the Biosphere: The Incredible Legacy of the Industrial Revolution. Third Annual Kuehnast Lecture. Department of Soil, Water and Climate, University of Minnesota, St. Paul, MN.
LaMarche Jr., V.C., Graybill, D.A., Fritts, H.C. and Rose, M.R. 1984. Increasing atmospheric carbon dioxide: Tree ring evidence for growth enhancement in natural vegetation. Science 225: 1019-1021.