Pielke Senior on tree and thermometer divergence

by Dr Roger Pielke, Sr.

With the McShane and Wyler paper examining and questioning the method, this look at the proxy data and its problems seems like a relevant issue to review.

Comment On Tree Ring Proxy Data and Thermometer Type Surface Temperature Anomalies And Trends

There was an interesting conclusion in a New York Times article on the relationship between tree ring proxy temperature trend analyses and thermometer type measures of temperature anomalies and trends.  The article is

British Panel Clears Scientists by Justin Gillis published on July 7, 2010

The relevant text is on page 2 it is written

“But they were dogged by a problem: Since around 1960, for mysterious reasons, trees have stopped responding to temperature increases in the same way they apparently did in previous centuries. If plotted on a chart, tree rings from 1960 forward appear to show declining temperatures, something that scientists know from thermometer readings is not accurate.”

There are, however, problems with this conclusion. Since the thermometers are not coincident in location with the tree ring data (in the same local area), it would not be surprising that they are different. Indeed, this is yet another example that implies unresolved biases and uncertainties in the surface temperature thermometer type data as we discussed in several of our papers (see and see), as the thermometers are measuring elsewhere then where the proxy tree data is obtained.  This obvious issue has been ignored in the assessment of this so-called divergence between the two methods to evaluate temperature anomalies and trends.

It is possible, of course, that the trees are responding differently due to the biogeochemical effect of added carbon dioxide and/or nitrogen deposition. Nonetheless, to accept the thermometer record as the more robust measurement of spatial representative temperatures is premature.

I have discussed this issue further in the posts

Comments On The Tree Ring Proxy and Thermometer Surface Temperature Trend Data

December 2007 Session ‘The “Divergence Problem’ In Northern Forests

A New Paper On The Differences Between Recent Proxy Temperature And In-Situ Near-Surface Air Temperatures


newest oldest most voted
Notify of
Retired Engineer

Restating the obvious: since tree rings respond to lots of things, like sunlight and moisture, it seems odd to me that they matched temperature trends at all. Particularly in ‘previous centuries’. Just how in the {snip} did they measure temperature in ‘previous centuries’? (what made that proxy so good?)
Seems like a lot of assumptions, and facts not in evidence. Or just Bad Science.

Rob Dawg

Excellent snark sir. May I suggest that the supremacy of trees be vetted by an unbiased panel of Druids?


In retrospect, trees are terrible proxies. Many botanists have commented on the problem. Insect, fungus, bacteria, drought, flood, rock slides, avalanches, wind, are all unrelated to temperature. Frankly, the proxies are likely falsified through the entire course. It was when temperature and correlation became possible the panic started. It was all BS to begin with.

Mike McMillan

I fail to understand the great faith in tree rings as temperature proxies. Trees do better in warm years than in cool years. Trees do better in wet years than dry years. Do they do better in warm dry years than in cool wet ones?
A close look at the rings in Dr Mann’s glamour shot show that a core taken from different azimuths can give you very different ‘temperature’ readings. Or maybe different rainfall readings. They can even give you different ages.
Whatever they show, they show it only for daytime, and only for part of the year. And they sure didn’t mysteriously stop showing it in 1960.


“Nonetheless, to accept the thermometer record as the more robust measurement of spatial representative temperatures is premature.”
Interesting. Perhaps the tree ring results which show cooling in the area should be incorporated into the temperature record “from 1960 forward”.

As for changing the way they (trees) respond to temperature increases, I call upon the concept of uniformitarianism.
Yes, this isn’t pure geology, but I believe the concept is sound.

Bruce of Newcastle

Trees like CO2 rather a lot. Next to a coal fired power station growth rate was found to triple. If the growth rate is x3 then the trees’ rings are thicker. Amazing!

J. Knight

Yes…yes…here we were, going merrily, merrily for thousands of years with the tree rings being the best of the best temperature proxies until in 1960 they weren’t. I blame the thermometers, and the mystery. It’s a mystery, my friends, so methinks a novel is in order. A climate change novel, with a movie based on it. Quite an idea!

Matt E

“Since the thermometers are not coincident in location with the tree ring data (in the same local area)”
Boy, stunning that such close proximity calibration hasn’t been done. How hard would it be to puta thermometer in the woods and actually callibrate tree ring data to exact temp? Or do tree ring analysis in woods near a well-sited thermometer with a long term record?
Not glorious work, but how impotent to do.

Jimmy Haigh

I was born in 1960 and I deliberately caused the divergence problem because I thought it might be fun.

Dave Wendt

I will consider the possibility that dendroclimatology is worth even looking at when someone can provide me with a description of the exact physical mechanism by which a tree can encode a record of the ambient atmospheric temperature of the vicinity where it grows, which is even remotely plausible. Until then, if you come bearing tree ring proxies, just keep walkin’, cuz I ain’t interested.

son of mulder

Everyone knows the trees were there in the 60’s so no wonder thay can’t remember the temperature…..man.
Which is as supportable as any theory that tree rings are a useful proxy for temperature.

Matt E says:
August 21, 2010 at 11:35 pm
[…]Not glorious work, but how impotent to do.

Uh…is that really what you meant to say?

The logical conclusion is that the tree ring calculations are flawed.
The religious faith based conclusion is that CO2 has changed the trees.

James Sexton

Wow, relevant comments with some humor mixed in. What a nice contrast to the hockey stick thread!
I’m amazed people still give any credence to tree rings as proxies for temp. As stated, there are so many variables that combine for the make up of a tree ring, it is impossible for me to believe anyone can reasonable discern what is temp in the tree rings and what is not. Of course, if I cared to figure out the mystery of divergence, I’d look for an event that occurred near 1960. I don’t buy the CO2 and/or nitrogen theory, because if they can indeed discern temps from the tree rings, surely they’d be able to discern known substances that directly effect tree growth. If they can’t do that, then it is just more evidence that tree rings as proxies are as reliable as reading tea leaves. Further, given the hyperbole surrounding CO2, I think if there were even a remote chance CO2 were the culprit, we’d all already heard about it.

Christopher Hanley

“…..but they were dogged by a problem: Since around 1960, for mysterious reasons, trees have stopped responding to temperature increases in the same way they apparently did in previous centuries….”
Since around 1960, trees have stopped responding to ADJUSTED temperature increases.


Mike McMillan says:
A close look at the rings in Dr Mann’s glamour shot show that a core taken from different azimuths can give you very different ‘temperature’ readings. Or maybe different rainfall readings.
Not only is this tree trunk more “egg shaped” than round it also shows very asymetric growth patterns in some years. Especially when young. Can’t really see how either temperature or rainfall could account for this. An alternative possibility would be shading from other trees restricting where this tree can put its branches and leaves, this needing more mechanical support in the trunk. However as it became more mature factors such as pervailing wind and direction of sunlight became more important.


Bruce of Newcastle says:
Trees like CO2 rather a lot. Next to a coal fired power station growth rate was found to triple.
Not really much of a suprise. Given that market gardeners have discovered that putting carbon dioxide into greenhouses increases yields.

Cold Englishman

Yeah, but it was the sixties, flower power and The Beatles, and they were “cool”, no wonder the trees stopped growing.

Neil Jones

Hide the decline?


“Since around 1960, for mysterious reasons, trees have stopped responding to temperature increases in the same way they apparently did in previous centuries.”
This sort of moebius lie is acceptable? ?
I call ‘loco como mierda de murcielago’

John R. Walker

The idea that tree rings can be used as a temperature proxy is so preposterous it shouldn’t even be on the table.
And anybody who claims to be able to measure rings to a hundredth of a mm should be locked up…
The fact that anybody is still talking about this is a triumph of intellect over intelligence!

John Marshall

The most prominent input to tree ring growth is atmospheric CO2 content. It is also problematic that some scientists insist that global temperatures have increased as shown by thermometers (after much fiddling with the data) and now consider that tree rings are wrong for the last 50 years when temperatures have actually shown a leveling then slight fall. But we must not be diverted from our forward march by data we do not like.

jim hogg

Until there are thermometers co-sited with the trees to be examined we can make no reliable deductions at all Mr Goddard – concerning the so called DP. Once the divergence problem was identified I would have thought that placing thermometers beside subject trees for long term evaluation would have been the first and most obvious response. If it hasn’t been done, now would be a good time dear “scientists”, and a wee bit further down the line you can make your deductions on the divergence problem based on relevant and useful data analysis, and also of course on the long term record. But, surely, this process has already been started. If not then we should pull out all interested climatologists by their rotten roots and have them replaced by proper scientists, or even rational human beings . . . .


Until there is a full confirmation by experiment of a Theory for the divergence the only explanation for the divergence is that trees do not react to higher temps post 1960 and never have before 1960. So until this tested theory is available using trees as a temp proxy will not work as high temps are unrecorded hence the Hockey Stick.


For all the ills of being poorly correlated to temperature, tree rings studies in places like the Pacific Northwest, where some species go back 2,000 years, events such as the Roman Warm period, MWP and Little Ice Ages are well depicted. They also show off Super El Ninos, putting 1998 in it’s place…down the list.


Oops, sorry, the above tree ring data series was done by Cook. Couple of nice videos of each year plus a comparison with temp data.
Today’s rough climate is tame by comparison to events of the last 2000 years.


My goodness – trees arn’t thermometers – except for those that show a hockey stick shape.


Why look at the trees being the problem, couldn’t it be the thermometers’ fault?
Maybe since the 1960s thermometers have been giving inaccurate readings of temperatures, after all the tree ring based temperature record is 1500 years long and only had a problem in the last 50 years (about 3% of the record) while the thermometer based temperature record is only about 150 years long so a whole 1/3rd of that record doesn’t match the tree ring record. Overall it seems more reasonable to assume that thermometers are just a poor means of measuring temperature thus the unreliable thermometer data should be thrown out in favor of the tree ring data.

anna v

There is a saying, “putting the cart before the horse”.
Climatology is full of examples.
It seems that they have not heard of the word “calibration”. All thermometers are proxies of the temperature T which enters into equations. ALL
Some are more accurate proxies than others.
Take a mercury thermometer, and calibrate it the standard way : 0 at ice melting point and 100 at water boiling point at 1 atmosphere. That is the Celcius scale, and can be very accurate.
One now has a thermometer to calibrate against for less accurate proxies.
The scientific method for calibrating tree ring width to be used as a thermometer would require a statistically significant sample of tree samples from a region where there are thermometer readings for a number of years.
Where is this calibration curve?
The defect of the tree ring studies is not only that they compare the tree proxy temperatures to average globe temperatures, but that they have not demonstrated a calibration curve .
Putting the horse in front of the cart the fact that there is a decline after 1960, according to the logic of the anomalies, should have shown a decline in the more accurate thermometer readings. The decline in the tree ring proxy data demonstrates that the tree rings are not thermometers.


“…trees have stopped responding to temperature increases in the same way they apparently did in previous centuries…”
No they did not respond that way for “centuries” . Temperature records started only around 1850, and divergence problems before that period may have been just as common as they have been after 1960.
The best answer is still that tree rings are very poor temperature proxies.
The second best answer is that temperature records may be corrupt as well.
But Gillis’ assumption that tree rings responded to temperature variations for a thousand years and then in 1960 suddenly atopped to do so is embarrassing junk.


It really is very simple.
If you “lose temperature sentativity” after a certain date, you cannot under any circumstances be certain that it EVER had ANY temperature sensativity to begin with. I do not understand why that is so difficult to understand.
Trees make horrid thermometers.

Sam the Skeptic

I’m well kitted-up in a kevlar suit and tin helmet before I say this but ….
Is there any mileage in the hypothesis that tree rings may well be a lousy proxy for temperatures but that they could be a fair proxy for climatic conditions in general?
Most people, certainly on the sceptical side of the argument, appear to believe that there is more to climate change than temperatures alone. If there is a “divergence” around 1960, why would that be? And would it tell us anything important about the current state of the planet?
I only asked … 🙁

Paul Coppin

As a biologist, I will say for the umpteenth time, individual trees in the wild cannot be used for temperature proxies. Even a first year biology student, with a little reflection, can explain why. The short answer is that every tree grows in a unique ecology, sufficiently different from its neighbour to render the proxy data unquantifiable for small populations. A sufficiently large population of tree samples can point to gross local events such as large scale fires, droughts or similar events, but to look at climatic variables you have to look at forests, not trees.
This underpins the entire problem with using complex organisms as proxies for anything. Their inherent adaptability makes them unfit for the role. Its not until you can begin to evaluate the response of the population can you begin to approach any kind of statistical significance to certain measured results, and even then, with great risk. Complex biological systems are inherently chaotic. Identifying confounding variables is often a bigger problem then accurately measuring certain parameters. Complex organisms and their behavior is the result of a bewildering number of multi-causal, multi-variable effects.
From my very first year uni programs, its been clear that the physical sciences pursued by my student colleagues have always had a poor scientific appreciation of the nature and complexity of biological systems (aided, I suppose, by the late night beer sessions explaining to my engineering buddies just how it was their girlfriend got pregnant…). Mostly I chalked this up to significant lack of training in the biological sciences as they pursued their interest in the relatively simplistic (if not indeed, simple) narrow world of physical and mathematical problems. The result is the kind of worthless science produced by the current crop of “climate scientists”, which is at best , simply intellectually naive, and at worst, dangerously agendized. It pains me to see colleagues climbing on the agenda bandwagon promoting the thin-on-the-ground attitudinal bio-proxy science that is turning up. Science education is, I believe, in crisis. Not only so in the agendas that drive much of it, but in the fundamental training biology students are receiving – they simply are not getting an adequate foundational basis on what life is, a knowledge that comes from things like the “old” disciplines of comparative anatomy, phylogenetic analysis, physiology etc.
In order to model anything, you first have to know what it is you are modeling. I believe science education has failed miserably in preparing students how to come to know about what they are doing, preferring instead to dazzle them with the bright lights of technology, rather then the bright spark of intellectual realization. The end result is the present morass of intellectual fluff being posited as grand revelations on the future of the planet.


Hi there!
Well I wonder why a layman as me react to the most obvious basic contradictions
in cliamate proxies.For six month I wonder what the climatehistory would look with out tree ring proxies removed. What does the hockeystick and all the climatic conclusions look like without them? For me this is defenitely a “its much vorse than I thought” that this obvious circumstances doesnt hit the table until now!

Joe Lalonde

Proxies are theories for replacement of missing data that cannot be recovered by the past as measurements were not available.
Widely used by scientists and a good source of funding to continue the theoretical area.


The idea about treemometers is that only trees which are growing in certain limited an more or less stable conditions may respond predominantly to temperature as a main factor. The bristlecones are constantly water and nutrient-limited, the Siberian trees have a constant abundance of water (no limit there) but their growth occurs only during one month in summer.
BTW – the Russians, when they write themselves about their tree rings, are very careful to point out, that the trees are a proxy for summer Siberian temperature. Not much talk about global temp here…


My simple point has always been to point at the elephant in the room.
If tree ring data was adequate prior to 1960 then why are they not adequate after 1960? [diverged from thermometer data]. Fine, so what degree of confidence should I have on tree ring data prior to 1960?
In my simple mind one of the measurements is most likely producing erroneous data. UHI, “biogeochemical effect of added carbon dioxide and/or nitrogen deposition” and geographic locations only complicates matters further.

Ian W

Botanists have told climatologists that tree growth is not dependent on temperature so the use of tree rings as temperature proxies is unsound.
Statisticians have told climatologists that their use of statistics is flawed both in methodology and sample sizes.
Yet climatologists continue despite this expert advice to use tree ring temperature proxies and flawed statistics based on insignificant samples to support wild unvalidated claims.
Is there any other branch of science where these behaviors would be accepted?

Just being a Simple Red Neck, I am easily puzzled.
If you have one set of temperature records taken in old growth forests (I.e. tree ring proxies).
And a second set of temperatures taken in areas that have, over the years, become Urban Heat Islands.
And you then compare them.
Why in the world would you expect them to match?
Befuddled in Texas,
Steamboat Jack (Jon Jewett’s evil twin)

Henry chance

Bad news for Mann and Briffa. We just aren’t into the tree ring readings.


So let’s start with the hypothesis that trees are really better thermometers than thermometers. Now someone has to explain the following to me:
My mother has 2 red maple trees in her yard. They were the same size when they were purchased at the same nursery and planted 20 feet apart about 10 years ago. Today one is twice the diameter of the other and the larger one is nearly 3 times as tall.
Which one is the better thermometer? Why?
PS The real reason one is bigger than the other is because the biggest one was planted where the old outhouse used to be, and the smallest one wasn’t.

old construction worker

It is my understanding with wide enough error bars anything can be a thermometer.

Scroll down to page 18 (of the pdf file) and take a look at the symmetry of the bristle cone pine and strip bark juniper which Mann used to create his proxies. Does anyone know what kind of tree Mann is holding in his glamor shot?
Yes, this is a tree you can trust!

oops, here is McIntyre’s file referred to above. How it all started.


Trees like CO2 rather a lot.
I thought trees responded rather well to increased levels of CO2, that should mean that trees should show greater growth since 1960 not less, OR DO THEY.
The Parker and McMahon’s paper focuses on the drivers of the accelerated tree growth. The chief culprit appears to be climate change, more specifically, the rising levels of atmospheric CO2, higher temperatures and longer growing seasons.
During the past 22 years CO2 levels at SERC have risen 12%, the mean temperature has increased by nearly three-tenths of a degree and the growing season has lengthened by 7.8 days. The trees now have more CO2 and an extra week to put on weight. Parker and McMahon suggest that a combination of these three factors has caused the forest’s accelerated biomass gain.


From an earlier article:
“Ultimately, any alternative explanation must fit all the observations. If the alternative hypothesis fails even only one of the observations, then the alternative is rejected.” Ferdinand Engelbeen
Ok, tree rings as a temperature proxy, rejected.

amicus curiae

tree rings for Age of the tree, fine.
for wet or dry seasons, fine.
for fires, fine,
for bugs, fungi,pollen, soil type and nutrition of soils, fine
for the immediate surrounding area only, fine
for temp?
no way!


What do you think would be more consistent over the centuries?
1. Trees in forests
2. Thermometers in villages and at airports
The NY Times blindly accepts that #2 is more accurate gauge of temperature. Most objective observers would consider the possibility that the thermometers were measuring increased heat produced by human development.

Harry Lu

trees do not grow at -40C
trees do not grow at +100C
trees grow very well at 15C
This cap shape must have a dependence on temperature. It may not be linear but it must be there.
Somewhere between 15C and 100C the growth must start declining Did trees pass the optimum in the 60s?
Uncontrolled emissions in the 60s, 70s and 80s was known to cause acid rain (to an extent that some countries were forced to add lime to lakes to prevent damage) there was plenty of evidence that trees were being damage also.
Is it not true to say Damaged trees=slow growth
There are many factors that can slow tree growth but apart from over temperature these effects will be diminished by limited industrialisation (before 1900?).
Trees are rubbish thermometers, but in all the noise there MUST be a temperature signal. A large local sample will lower the noise from sickness, or damage. A large global sample will lower the noise from changes in soil fertility, etc.
Nothing will remove the noise from CO2 fertilisation, or other global events.
Some trees growing at the limit of their water needs may be negatively affected by rises in temperatures from their minimum growing value – growing in heat requires more water. these will always show a negative growth increase with temp. But if averaged with enough positive responders then these will be insignificant.
But the signal that remains must, when averaged contain a temperature signal (not necessarily linear)
“Overall, the Program’s cap and trade program has been successful in achieving its goals. Since the 1990s, SO2 emissions have dropped 40%, and according to the Pacific Research Institute, acid rain levels have dropped 65% since 1976.[16][17] However, this was significantly less successful than conventional regulation in the European Union, which saw a decrease of over 70% in SO2 emissions during the same time period.[18]
In 2007, total SO2 emissions were 8.9 million tons, achieving the program’s long term goal ahead of the 2010 statutory deadline.[19]
The EPA estimates that by 2010, the overall costs of complying with the program for businesses and consumers will be $1 billion to $2 billion a year, only one fourth of what was originally predicted.[16]”
“However, the issue of acid rain first came to the attention of the international community in the late 1960s, having been identified in certain areas of southern Scandinavia, where it was damaging forests. The matter quickly became an international issue when it was discovered that the acid deposits in these areas were a result of heavy pollution in the UK and other parts of northern Europe.
Acid rain and air pollution emerged from the industrial boom of the early 1900s onwards and the increasing levels of chemical production associated with these processes. The building of taller industrial chimneys from the 1960s onwards was largely held to be responsible for pollutants generated in the UK blowing as far as Scandinavia. “