Michael Mann won’t be happy about this.
A new paper now in open review in the journal Climate of the Past suggests that “modern sample bias “has “seriously compromised” tree-ring temperature reconstructions, producing an “artificial positive signal [e.g. ‘hockey stick’] in the final chronology.”
Basically, older trees grow slower, and that mimics the temperature signal paleo researchers like Mann look for. Unless you correct for this issue, you end up with a false temperature signal, like a hockey stick in modern times. Separating a valid temperature signal from the natural growth pattern of the tree becomes a larger challenge with this correction.
Here is a relevant excerpt:
Much of the work in dendrochronology, and dendroclimatology in particular, relies on accurate, unbiased reconstructions of tree growth long into the past. As a result, a great deal of effort has been put into trying to isolate important trends and identify potential 5 biases. However, one major bias called “modern sample bias”, first identified by Melvin (2004), is still largely neglected in applied studies, despite its potential impact on all regional curve standardization chronologies (Brienen et al., 2012a).
Dendrochronologists observed that the older a tree was, the slower it tended to grow, even after controlling for age- and time-driven effects. The result is an artificial downward signal in the regional curve (as the older ages are only represented by the slower growing trees) and a similar artificial positive signal in the final chronology (as earlier years are only represented by the slow growing trees), an effect termed modern sample bias. When this biased chronology is used in climate reconstruction it then implies a relatively unsuitable historic climate. Obviously, the detection of long term 15 trends in tree growth, as might be caused by a changing climate or carbon fertilization, is also seriously compromised (Brienen et al., 2012b). More generally, modern sample bias can be viewed as a form of “differing-contemporaneous-growth-rate bias”, where changes in the magnitude of growth of the tree ring series included in the chronology over time (or age, in the case of the regional curve) skew the final curve, especially 20 near the ends of the chronology where series are rapidly added and removed (Briffa and Melvin, 2011).
A likelihood perspective on tree-ring standardization: eliminating modern sample bias
J. Cecile, C. Pagnutti, and M. Anand
University of Guelph, School of Environmental Sciences, Guelph, Canada
Abstract
It has recently been suggested that non-random sampling and differences in mortality between trees of different growth rates is responsible for a widespread, systematic bias in dendrochronological reconstructions of tree growth known as modern sample bias. This poses a serious challenge for climate reconstruction and the detection of long-term changes in growth. Explicit use of growth models based on regional curve standardization allow us to investigate the effects on growth due to age (the regional curve), year (the standardized chronology or forcing) and a new effect, the productivity of each tree. Including a term for the productivity of each tree accounts for the underlying cause of modern sample bias, allowing for more reliable reconstruction of low-frequency variability in tree growth.
This class of models describes a new standardization technique, fixed effects standardization, that contains both classical regional curve standardization and flat detrending. Signal-free standardization accounts for unbalanced experimental design and fits the same growth model as classical least-squares or maximum likelihood regression techniques. As a result, we can use powerful and transparent tools such as R2 and Akaike’s Information Criteria to assess the quality of tree ring standardization, allowing for objective decisions between competing techniques.
Analyzing 1200 randomly selected published chronologies, we find that regional curve standardization is improved by adding an effect for individual tree productivity in 99% of cases, reflecting widespread differing-contemporaneous-growth rate bias. Furthermore, modern sample bias produced a significant negative bias in estimated tree growth by time in 70.5% of chronologies and a significant positive bias in 29.5% of chronologies. This effect is largely concentrated in the last 300 yr of growth data, posing serious questions about the homogeneity of modern and ancient chronologies using traditional standardization techniques.
The full paper is here: http://www.clim-past-discuss.net/9/4499/2013/cpd-9-4499-2013.pdf
h/t to The Hockey Schtick
One thing we know about tree rings is that they reflect how much the tree has grown. What we don’t know is if they make an accurate thermometer.
Pete Brown says:
August 16, 2013 at 6:57 am
These people should do the environment a favour and stop cutting all the flippin’ trees down to look at their rings…
>>>>>>>>>>>>>>
Foresters use a coring tool to count tree rings. <a
Darn the google URL was too large for a working link. Try searching “forester coring tool tree rings image’ instead.
And of course Jim Bouldin has been beating this drum for a while now!
http://ecologicallyoriented.wordpress.com/
Lets face it, after 70 odd years of pumping CO2 into the atmosphere any changes it has caused are so small as to be imperceptible. Team AGW have been banging the same drum for 30 years now as if climate catastrophe was round the corner. They have all the credibility of a man with a sandwich board calling out “the end is nigh”. Dealing with their “science” piece by piece is lending them some kind of credibility they don’t deserve, suggesting they at least have an arguable position when in fact they have no such thing.
The climate is not changing.
Greg says:
August 16, 2013 at 3:04 am
Very well said. I am very pleased to see that just about everyone writing a comment understands the empirical problems faced by dendrothermometry. Just a couple of years ago many critics of dendrothermometry could not see the underlying empirical problems.
I am pleased to see that some who practice dendrothermometry, the authors under discussion, are attempting to refine their statistical methods in light of a hard won awareness of systemic bias in tree ring samples. Let the empirical light shine on the statistical magic.
WUWT has been, and will continue to be, the website where the demands of empirical science are taken most seriously and where sophistication about these matters grows steadily.
The link is to an abstract, rather than the complete document. Since the abstract seems to contradict itself, the complete document should be studied to see if the study is useful in the AGW debate. (I have not read it)
{ Alan Mackintosh says:
August 16, 2013 at 1:15 am }
Correct.
Any logger who practices Best Management Practices (i.e. not clear cutting) considers tree age. For example, red oak are most efficient in net dry weight growth until about 30″ Diameter Breast High (DBH). Subsequently, the energy expended to get water and minerals to the top growth foliage requires increasing amounts of respiratory energy for transport, reducing net dry matter accumulation. Therefore when using selective cutting policy, I never felled a tree less than about 30″.
(Well, if it was a uniform, straight tree with few lower branches probably containing a high percentage of First and Second quality valuable wood I sometimes cut it down to 25″, but don’t tell anyone.)
Also, as forests age and trees grow, crowns begin to touch each other and the forest canopy closes. Most of the tree crowns will be unable to grow as rapidly as if they had free space to occupy. The photosynthetic capacity will be spread among a greater number of trees. That means less photosynthesis per tree, which translates into slower growth.
http://plants.usda.gov/java/charProfile?symbol=QURU
Height at 20 Years, Maximum (feet) 36
Height, Mature ~80 yrs (feet) 81.0 .
Do the math.
An Excerpt:
D’Arrigo et al.(2008) suggest that modern sample bias may be responsible for the “divergence problem” in dendroclimatology, the widespread reduction in temperature sensitivity of tree-ring chronologies in recent decades. The generally negative trend (ref 10) induced by modern sample bias in recent years certainly suggests that this may be at least part of the problem.
More generally, the theoretical results of this paper clarify, simplify and extend re-gional curve standardization. Regional curve standardization is a biased implementa-tion of signal-free standardization, while signal-free standardization is itself equivalent
(ref 15) to the new effect regression standardization. Working within a regression framework improves the transparency of the standardization process, allows investigators to use classical regression tools such as AIC and, as demonstrated, facilitates investigation
of alternate underlying models of tree growth.
I certainly support “may be” and “suggests”. The paper is important, and should stimulate other papers on the problem, including “regional curve standardization.” I can imagine a cottage industry growing from this.
@- jim2
“One thing we know about tree rings is that they reflect how much the tree has grown. What we don’t know is if they make an accurate thermometer.”
However lumber companies use the temperature history of a stand of trees to estimate the amount of wood they can extract. There are other factors of course, but at least to a first approximation just the temperature record is a good correlate with tree growth.
REPLY:They also consider rainfall. Bulk production correlation of temperature to thousands of trees is easy, reverse correlation getting temperature from a few trees (like YAD06) not so much, which is the central issue to bollixed paleoclimatology- Anthony
hmmm. I probably ought to have used the new preview facility.
The link is to an abstract, rather than the complete document
follow the hint: the full paper is here etc.
This is amazing. In a previous epoch (how’s that for a start?) I wrote a comment on this site concerning the similarities between Michael Mann’s research and my sister’s, my older sister’s fungal toe.
Well, what a coincidence. It was just this very week, this very week, that my sister, my older sister (ok, I’m getting repetitious), went again for treatment for her fungal toe, her first follow-up treatment since my previous comment. And, lo and behold, what occurs almost simultaneously to that treatment is another story on this site about Michael Mann. I mean it’s weird but it’s almost as if the fungus in her toe has a relationship with Michael Mann. And, what the heck, why not? Why shouldn’t two distinct living organisms which exist (or are funded) solely through the blood and sweat of host organisms, which never go away but instead keep coming back and back again, why shouldn’t they have a relationship. (Izen, are you listening?)
While there’s not much difference between a fungal toe and Michael Mann’s research there is a big difference between Michael Mann’s research and my older sister. You see, when my older sister wants to intimidate me she waves her contagious fungal toe at me, but I don’t think the IPCC waves Michael Mann’s tree ring research at anybody anymore.
BTW: Welcome back Richard Courtney.
I made a comment a few years back at Climate Audit that if one were going to use tree rings for temperature proxies, it needed to so in a manner in which any year’s measurement had to come from trees that were of the same age. So, if one were going to measure a temperature in 1560, for example, then one would measure it a tree that was X years old in 1560, and measure temperature for 1980 in a tree that was X years old in 1980.
We’ve had many thousands of thermometers around the planet measuring temperature and still can’t agree entirely on what its been doing the last 15 years or even 100+ years based on how you interpret the data or the location of the instrument shelter/weather station(with respect to urbanization).
And trees, being greatly effected by more elements than temperature(precip is often more important) from hundreds of years ago can do this?
The uncertainty over what caused the size of each ring long ago is too great to be making assumptions. Guesses yes.
Regarding recent data. How much error is being introduced by not properly dialing in the huge growth of woody stemmed plants/trees in response to increases in CO2?
If you’re a scientist that believes CO2 is bad, it’s highly unlikely that you will acknowledge and fully/accurately account for massive new growth based on just the CO2 fertilization factor.
Doing so would be evidence of the most powerful benefit of CO2 that overwhelms any effect on temperature.
I proposed this theory many years ago (http://www.nrcresearchpress.com/doi/abs/10.1139/x88-032) in1988 and showed that data support it. The basic idea is that slower growing species (or trees within species) will live longer because the reason they grew slower was better investment in defenses such as thick bark and defensive chemicals. The oldest trees thus have a bigger percentage of the slower growing trees and the young trees are a more random sample, with more fast growing individuals. This creates the bias and the hockey stick.
Most of these studies seem to be of old trees in temperate and colder climates. Alan above notes ” … Birch … maturity … 70-80 years ….Oak, … several hundred … 40 – 50 years …. Spruce …” Conspicuous by their absence are any studies of trees in tropical climates.
Here at 19° 11′ 38″ S 146° 40′ 31″ E I have two tree stumps from a local Acacia species. One planted 23 years ago as a “whip” eg 1m+ high. In 2011 it was about 15m high, 12m diameter crown, diameter of trunk at base maybe 700mm. Benefited from profligate irrigation next door. Another, planted from tube-stock about 15 years ago, might be a better candidate for analysis as this had no artificial irrigation or nutrients. In 2011 height was at least 20m, trunk diameter maybe 500mm. Along with other mature trees, these were severely damaged by Cyclone Yasi, I ended up with about 15 tonnes of felled timber from a mere 1347m2. More tree-ring specimens than would fit in Mann’s office. Significant increase in tree ring width after about 2004, and nothing to do with increase in temperature, precipitation or insolation, because according to local BoM monitoring (20km ESE on the same coastal plain) there wasn’t any.
Looks like slowing growth in old trees scotches the left wing fetish for old growth forests as a carbon sink. And why no interest in increasing the organic material in soil? Far better to leave the plant material in the soil.
-T
“Basically, older trees grow slower, and that mimics the temperature signal paleo researchers like Mann look for.”
‘ralfellis says:
August 16, 2013 at 1:14 am
>>Unless you correct for this issue, you end up with a
>>false temperature signal.
But will you end up with a temperature signal at all? Here are a few scenarios for you…
a. Tree experiences a very cold summer, with no growth. (a temperature signal?)
b. Tree experiences a very hot but too dry summer, with no growth. (a temperature signal?)
c. Tree experiences a cool but nicely moist summer, with good growth. (a temperature signal?)
d. Tree experiences a very hot and wet summer with good growing conditions but a rampant pest infestation, with no growth. (a temperature signal?)
###################
before folks reject tree rings out of hand, or merely by waving their hands its important to get the argument correct. In other words it’s best to describe your opponents argument fairly and then go to work on the weakest parts.
The growth of trees is limited by several factors. When selecting a tree as a treemometer the scientist will try to
A) select a species that is particularly temperature sensitive.
B) select a stand of trees where the growth is temperature limited
Even wikipedia knows this.
“Climate factors that affect trees include temperature, precipitation, sunlight, and wind. To differentiate among these factors, scientists collect information from “limiting stands”. An example of a limiting stand is the upper elevation treeline: here, trees are expected to be more affected by temperature variation (which is “limited”) than precipitation variation (which is in excess). Conversely, lower elevation treelines are expected to be more affected by precipitation changes than temperature variation. This is not a perfect work-around as multiple factors still impact trees even at the “limiting stand”, but it helps. In theory, collection of samples from nearby limiting stands of different types (e.g. upper and lower treelines on the same mountain) should allow mathematical solution for multiple climate factors. However, this method is rarely used.”
http://en.wikipedia.org/wiki/Dendroclimatology
a good blog comment starts with a little research.
Then its important to realize that you can actually test whether a temperature signal exists.
1. Take cores from tree rings going back to 1850
2. Use the instrument record from 1900 to present to calibrate
3. Predict the temperature from 1850 to 1900 using the calibration ( called verification)
4. compare your prediction against the actual.
So, when you see St. McIntyre slamming Mann because of verification statistics you see the Stronger argument against the result. The weaker argument just dismisses all tree rings.
The weaker argument says “we cant know”. the weaker argument is anti science.
The stronger argument accepts that we might be able to tease out a temperature signal and then systematically challenges all the assumptions and methods used to extract that signal.
the stronger argument does not presuppose that we cannot know. The stronger argument is a specific argument against a specific case and not a general dismissal based on a specific cases.
Just what are they measuring, when looking at tree rings? Temperature? Moisture? Infestations? Micro-climates? Forest canopy competition?
Steven Mosher says:
August 16, 2013 at 9:12 am
————————————————–
1. Take cores from tree rings going back to 1850
2. Use the instrument record from 1900 to present to calibrate
Regarding (1), how do you obtain a global sample?
Regarding (2), doesn’t the proxy record diverge from the 1960-present? If so you’re already in trouble without going back much more than 50 years. If we can’t go back 50 years how can we go back 150 years?
It would make more sense to compare different proxies and not let any one proxy (e.g. tree-rings) dominate the so-called science. Where is the hockey-stick in other proxies?
Having cut down many trees of known age from my youth I’ve always been suspect of tree ring data. Water has just as much influence on a tree ring as does sunlight, perhaps even more.
He’s an artist! He knows not what he spakes.
2. Use the instrument record from 1900 to present to calibrate
====
Then it will never be right, will it Mosh?
Steve from Rockwood says:
————————————————–
1. Take cores from tree rings going back to 1850
2. Use the instrument record from 1900 to present to calibrate
Regarding (1), how do you obtain a global sample?-
You don’t- you’re measuring local temperature changes- Alan McIntire
Regarding (2), doesn’t the proxy record diverge from the 1960-present?
That won’t be a problem because by definition the new proxy will match temperatures 1900 to present- Alan McIntire
It would make more sense to compare different proxies and not let any one proxy (e.g. tree-rings) dominate the so-called science. Where is the hockey-stick in other proxies?
Steven Mosher says:
August 16, 2013 at 9:12 am
“The growth of trees is limited by several factors. When selecting a tree as a treemometer the scientist will try to
A) select a species that is particularly temperature sensitive.
B) select a stand of trees where the growth is temperature limited”
I sincerely hope that you do not think that your comment does justice to the empirical questions that underlie the use of tree rings as temperature proxies. When used as a temperature proxy, a tree must be calibrated in just the same way that a thermometer is calibrated. Because trees are not manufactured you must do experiments on the individual trees to determine how they respond to changes in local environment. Does this mean that tree rings are unsuitable as proxies for temperature? Most historical records are useless because neither the tree nor its environment are available for experimentation. Some short historical records might be useful. However, because the “scientists” who use tree rings as proxies cannot bring themselves even to consider that experimentation is necessary, the question is moot.