Bristlecone Pines: Treemometers or rain gauges ?


Over on Climate Audit, Steve McIntyre has been making a series of posts that have been putting the final nails in the coffin for Michael Mann’s MBH98 paper. This paper was responsible for the famous hockey stick graph which is based on tree ring data from Bristlecone Pine trees. Mann’s work implies them to be excellent proxy indicators of temperature, and due to their age, a profound record of temperature. Problem is,  it looks like most of the results is Mann’s paper have been thoroughly discredited by the work of McKittrick and McIntyre in 2005, plus McIntyre’s more recent work.

At 4600-4800 years old for some of the oldest trees, Bristlecone Pines (BCP) certainly have seen most if not more than all of human recorded history, so it seems logical to look to them for answers about our temperature history.

One of the graphs Steve McIntyre recently produced was this one:

About this graph he notes:

Here’s the MBH98 PC1 (bristlecones) again marking 1934. Given that bristlecone ring width are allegedly responding positively to temperature, it is notable that the notoriously hot 1934 is a down spike.

Since 1934 is generally accepted now to be the hottest year on record in 20th century it is indeed curious that 1934 in Mann’s data shows up as a down spike.

But seeing what happened with 1934, one has to wonder what do these trees really record in their tree ring growths? Is it temperature as Mann speculates? Or is it any number of other things related to plant growth in various combinations?

I was curious about what others had to say about these ancient pines. One of the first articles I came across was by NOVA, the PBS science program. They had an in-depth article on the “Methuselah grove” where the most ancient trees reside in the White Mountains of California’s Inyo National Forest.

What caught my eye in the NOVA article were these passages:

It turns out that the bristlecone pine has evolved survival strategies that might make other, less hardy plants, well, green with envy. These strategies help it cope with one of the most flora-unfriendly environments on the planet.

But the really interesting one is this:

Bristlecone rings, which vary in width year to year, reveal that the trees have an innate ability to endure times of stress, such as a string of drought years. In such periods, the species can go almost dormant. “There is something a little fantastic,” wrote Edmund Schulman in the March 1958 National Geographic, “in the persistent ability of a 4,000-year-old tree to shut up shop almost everywhere throughout its stem in a very dry year, and faithfully to reawaken to add many new cells in a favorable year.”

No where in the NOVA article does it link temperature and tree ring growth for Bristlecone pines, but it seems clear that water is a major factor in BCP growth.

Another article I found on NASA’s Earth Observatory website initially talks briefly about temperature proxies, but then focuses on precipitation for the remainder of the article with a review of early work with BCP tree rings by Andrew Ellicott Douglass.

He wrote:

“Through long-past ages and with unbroken regularity, trees have jotted down a record at the close of each fading year—a memorandum as to how they passed the time; whether enriched by added rainfall or injured by lightning and fire…. So, in the rings of the talkative pines we find lean years and fat years recorded. The same succession of drought and plenty appears throughout the forest.”

The NASA article goes on to say:

In the 1950s one of Douglass’ former students and a respected tree researcher in his own right, Edmund Schulman, headed into the White Mountains to look at the trees rumored to be very old. He discovered Methuselah and the old bristlecone pines surrounding it. Around the trees, even older dead trees remained on the ground. Together, they gave a climate record of the Southwest United States that extends back 9,000 years, the longest record for a single tree species.

Douglas’ rings tell about rainfall in the southwestern United States, but trees also respond to changes in sunlight, temperature, and wind, as well as non-climate factors like the amount of nutrients in the soil and disease. By observing how these factors combine to affect tree rings in a region today, scientists can guess how they worked in the past. For example, rainfall in the southwestern United States is the factor that affects tree growth most, but in places where water is plentiful, like the Pacific Northwest, the key factor affecting tree ring growth may be temperature.

So in the case of the BCP in the USA desert southwest, it appears that since water is it’s scarcest resource, it has the most effect on it’s growth rather than temperature.

This situation fits well with Liebig’s Law, from Wikipedia:

Liebig’s Law of the Minimum, often simply called Liebig’s Law or the Law of the Minimum, is a principle developed in agricultural science by Carl Sprengel (1828) and later popularized by Justus von Liebig. It states that growth is controlled not by the total of resources available, but by the scarcest resource. This concept was originally applied to plant or crop growth, where it was found that increasing the amount of plentiful nutrients did not increase plant growth. Only by increasing the amount of the limiting nutrient (the one most scarce in relation to “need”) was the growth of a plant or crop improved.

Liebig used the image of a barrel—now called Liebig’s barrel—to explain his law. Just as the capacity of a barrel with staves of unequal length is limited by the shortest stave, so a plant’s growth is limited by the nutrient in shortest supply.

So when looking at trees that grow in a desert environment, you could naturally conclude that water, and not temperature is the shortest barrel stave.

No wonder 1934 is a negative on Man’s graph above, it was hot and dry that year. It’s been said that even grasshoppers were starving during that drought.


Above: USA Palmer drought index for 1934 see original source here

The White Mountains where Mann and other researchers took core samples from BCP’s is marked with the arrow above. Notice that the location is in the “extreme drought” area. Below is the map of all tree samples for the MBH98 paper:


It looks like, at least for 1934, BCP’s in the USA desert southwest are better at being rain gauges than “treemometers”. Given “Liebig’s barrel”, it makes one wonder whether BCP’s are a good proxy for temperature at all. Perhaps “Mann’s rain barrel” would be a better name for the MBH98 paper.

0 0 votes
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
March 19, 2008 10:55 pm

Trees have never been a good temperature proxy. The response of a tree to it’s environment is modified by a tremendous number of variables including biotic and abiotic damage, rainfall, fire, etc. You can create a “Drakes Equation” to list the unknowns and come up with any answer you like. At best, the range and amplitude of historic tree species occurence “may ” give a very general indication of historic temperature patterns. For instance, we know the world is warmer than the past, by the fact that trees now exist at the mouth of the Mackenzie River above the arctic circle. Ten or twenty thousand years ago, this area was under a kilometer of ice. We also know that the world is now warmer because that historic ice sheet no longer exists.

Alan Chappell
March 20, 2008 2:22 am

bobonthebellboy wrote;
” this area was under a kilometer of ice.”
If you go back a bit further than your 20,000 years you will see the same area was semi tropical.
If someone could explain the “how’s” of the ‘Piri Reis Map’ we would surely have a better understanding of life.

Alan Chappell
March 20, 2008 2:31 am

If anybody is interested in working out the possible explanation of the ‘Piri Reis Map’
is a good start.

March 20, 2008 2:34 am

How is tree ring growth affected by CO2 levels?
– more CO2 = more growth, right?

March 20, 2008 3:14 am

I would really like to see Steve get this published. Or at least try…

March 20, 2008 4:04 am

Isn’t it true that CO2 enrichment has been shown to cause most plant to grow better.
So isn’t it possible, or even likely, that recent “growth spurts” could be due to CO2 enrichment, and not temp fluctuations?

March 20, 2008 5:19 am

And yet today, there’s still huge momentum to make public policy based on MBH98. Has the IPCC ever apologised for the hasty use and resulting alarm it set off when it published its TAR?

March 20, 2008 5:21 am

I wonder if the people at the “public” broadcasting station’s NOVA realize the contradiction in their programming, one discussing the relationship of BCP rings to water and the other touting the upcoming disaster of AGW. Fill in the blanks with Mann’s study using the BCP rings as a cornerstone, it all falls apart. My guess is that they realize that a majority of people will simply believe the latest thing they hear and that they will program whatever their largest contributor prefers to hear.

Gary Gulrud
March 20, 2008 5:29 am

As McIntyre has progressively uncovered issues with the PCA calculations of the synthetic temperature proxy he is now turning to analytic probems.
Even the corrected synthetic proxy diverges from the smoothed global temp with post 1980 data. The MBH graph conflated post 1980 data with the proxy generation to hide this fact.
MBH discarded 17th century data and replaced it with constituent proxies.
Beside the manifest moisture-limited growth of the Bristlecones, they are CO2 limited as a high altitude species as well. MBH did no PCA work on this constituent proxy. Even so, PCA only establishes the relative contribution of the factors chosen for study to the dependent (tree ring growth) variable.
Moreover as McIntyre has also pointed out the Bristlecone series is heavily overweighted in the synthetic proxy; ironic, as a prime motivation for the proxy was to provide a global rather than NH-specific temperature record.
But more fundamental issues remain: Were the PCA competently undertaken, would it be praticable to, by trial and error, construct a polynomial of higher order, to generate an accurate trendline for global temperature, a chaotic non-linear function?
If this is the case, what is its utility since a reasonable facsimile of the temperature data cannot be generated from a trendline?

March 20, 2008 5:31 am

City folks don’t have much contact with trees — So spinning stories about what trees do is easy, depending on the target audience. Trees respond much more aggressively to water and fertilizer than they do to temperature.

March 20, 2008 6:01 am

Why did Mann pick Bristlecone Pines to be included in the MBH98 graph?
Well the simple reason was because of convenience. The tree is easy to access and because its is one of the oldest living tree in the world, it gives long time series going back at least 2000 years.
But, as a proxy for temperature it is one of the worst possible trees to choose from on the planet.
Of these three factors temperature, precipitation and CO2 levels, the temperature is probably the factor with least influence on tree ring growth for type of tree.
What factors are most important?
Humidity is important as the tree live in a dry mountain area.
The fertilization effect from rising CO2 level is important for nutrition.
Increasing levels of CO2 is also important because it reduces the water the tree uses when it sucks in CO2 and releases Oxygen and water vapor. This is because the openings the “stomata” in the needles or leaves of plants and trees become reduced with increasing levels of CO2.
The result is that the trees consume less water with increasing levels of CO2. This effect has led to a greening of desert like areas world wide.
The calibration of the tree ring growth in relation to the temperature has then been made by Mann by comparing tree ring growth and real temperature recordings from the last 100 years.
His temperatures are at least partly contaminated by the urban heat island effect and the tree ring growth variations are strongly affected by increasing levels of CO2.
No wonder then, that the Medieval Warm Period and the Little Ice Age have been wiped out in his graph. Thus,this has led to the support of global warming hysteria and claims that the last century saw unprecedented warming.
How this junk science has been regarded as real climate science is beyond me.
BTW Does anybody of you know the percentage in the MBH98 graph which is taken The Bristlecone Pines Trees?

March 20, 2008 6:25 am

It occurs to me that trying to find temperatures in tree rings is akin to trying to find coded messages in music tracks played backwards. Not much signal in the noise and any discovery is biased toward the human propensity to find patterns in data. “I buried Paul….”

Stan Needham
March 20, 2008 6:29 am

One really has to wonder what the state of the climate change/GW debate might be had the fraudulent Hockey Stick graph not been such a prominent feature of the IPCC’s TAR 2000.

Person of Choler
March 20, 2008 6:32 am

Astounding that nobody has pointed this out before. I often wondered about it myself, but assumed that the settled scientists would have figured out corrections for the obvious fact that rainfall (not knowable from the far past) has much to do with plant growth.

David S
March 20, 2008 7:10 am

Please pardon my ignorance, but what exactly is being plotted in MBH98 PC1?
If its tree ring width how can it have negative values? And what are the units?

March 20, 2008 7:13 am

I expect an entry anytime now at Realclimate thoroughly, in their mind at least, addressing and trumping any and all criticisms Steve brings up. I also predict that the entry will contain:
1) A disparaging reference to the Heartland Institute and the conference a few weeks ago.
2) Ad Hominem attacks on Steve.
3) At least one huge Straw Man and/or red herring that has nothing to do with the topic.
4) A minimum of 300 posts of backslapping and fist pumping from the gallery stating “You tell ’em Gavin…”

March 20, 2008 7:29 am

I thought that the justification for using Bristlecones as a temperature proxy _was_ exactly that their growth was tied to rainfall. It’s an indirect route to hemispheric temperature: Bristlecone growth =>rainfall, which in the Western US is tied to ENSO, which in turn is related to hemispheric temperature shifts.

March 20, 2008 7:41 am

I thought that all the science the IPCC relied on was peer reviewed. This doesn’t seem to have caught the mistakes in this case.
Perhaps the wrong people are reviewing articles? If a physicist reviewed Mann’s article, I would not expect that a mistake regarding tree growth to be caught. Climate science involves many disciplines. Do we know who is reviewing these things?
When I started reading about global warming the first thing that caught my eye was that the IPCC estimates of warming were being revised downward. This lead me to these conclusions:
1. The science is not settled
2. The original IPCC estimates of CO2 driven warming are high

March 20, 2008 8:05 am

Strangely enough I posted a poem about Bristlecone Pines a couple of days ago.

Dan Evens
March 20, 2008 8:10 am

David S, as to what is being plotted: It’s the anomaly. That is, it’s the difference between the calibration period and the value at any given time. So if the temps go up, it’s positive, and if they go down, it’s negative.

March 20, 2008 8:39 am

In addition to a relatively wet 20th century in Eastern California, we must also consider additional factors, such as sheep grazing, and, the rise in imported nutrient rich dust generated by various human activities in the Owens Valley, Westernmost Nevada, and beyond.

March 20, 2008 8:44 am

Ian, go learn about Tonopah Lows. As well as Siberia Express, and Inside Sliders. Also, learn about the SW Monsoon. All of these are actually stronger during a La Nina. Also, go learn about the impact of the Sierra Nevada on mid latitude cyclones on zonal tracks. Bottom line – the amount of available moisture, during the critical growth months, depends largely on late snow pack volume and persistence – the later the snow presents its moisture, the better the trees do. SW monsoon strength and location are secondary controls. You would find very poor correlation between E Calif BCP growth, and say, rainfall in Monterey or Fresno.

March 20, 2008 8:55 am

coaldust says:
“I thought that all the science the IPCC relied on was peer reviewed. This doesn’t seem to have caught the mistakes in this case.”
The peer review process in climate science is hopelessly broken. Papers that support the IPCC view of the world are passed with little or no scrutiny. Papers that go against the IPCC view are carefully examined and trashed. CA has numerous examples of double standards applied by RC and others when it comes to reviewing papers. For example, all of the criticisms of Loehl’s paper applied equally well to the IPCC approved Moberg paper yet Moberg is considered to be acceptable by the AGW folks.

Stan Needham
March 20, 2008 9:17 am

Do we know who is reviewing these things?
Coaldust, this article in the Financial Post a couple years ago by Steve McIntyre and Ross McKitrick will give you a good overview of the process of debunking Mann’s hockey stick. I found this to be the most telling portions of the entire article:

In February, 2006, the NAS appointed a panel of 12 eminent academics involved in climate science but not directly involved in the temperature reconstructions of the past 1,000 years. They were not an entirely “independent” panel, as some were occasional co-authors with the Hockey Stick authors. But even this limited independence was a major departure from procedures of the IPCC, which permits authors actively involved in scientific controversy to summarize the research — even if they end up acting as reviewers of their own work.
The IPCC lead author who selected Mann’s reconstruction for prominent display in the review of millennial temperature history was none other than Mann himself. At the time, he was a fresh and ambitious PhD, an odd choice to write the “consensus” review of climate history.
The system that allows such conflicts of interest has been severely criticized by some senior climate scientists, including Hans von Storch of Germany. However, the flawed process remains unchanged for the next IPCC assessment report, due in January, 2007. As reviewers of that report, we have expressed concerns to the IPCC about prominent use of graphics and empirical results from the lead authors’ own freshly published papers, which have not been in print long enough to have undergone adequate, independent review and assessment and, in some cases, not even long enough to meet IPCC publication deadlines. (emphasis added)

March 20, 2008 9:27 am

I’m not sure I follow you – I wasn’t claiming that precip for BCPs was automatically higher during El Nino, or that ENSO was the only factor determining their growth. Are you saying that BCP growth is completely unconnected with ENSO?

March 20, 2008 9:54 am

No doubt the New York Times will run the story about the end of Mann made global warming. In my dreams anyway.

March 20, 2008 10:02 am

It’s not completely unconnected to ENSO, but, only weakly connected. Years with the best late season snow pack can be either ones with a very strong El Nino (rare, see below) or ones where moisture got delivered late, down from the north (typically but not always La Nina years). The secondary moisture control, summer monsoon, is so late in the season (typically after late July) that it’s almost too late. Plus. monsoon T-storms are hit or miss. Also, if the monsoon points west or, north, instead of NW, it may miss this area completely. Put is all together, and the Mannian notion of an indirect correlation, with greater precip driven growth being teleconnected to a general trend of more dominant El Nino ergo “global warm period / trend” sort of falls apart. If there is any teleconnection it would probably be with bad winters in the intermountain West and Northern High Plains. BTW – this year is such a year and snow pack is good in the White / Inyo Mountains.

March 20, 2008 10:04 am

Sorry, forgot to mention, storms coming from the W or SW (typical of El Nino) drop a lot of precip W of the Sierra, and on the W Slope of the Sierra. But east of the Sierra, their impact is limited. Only the strongest El Ninos will result in much moisture making it past the crest. The best “approach route” for snow bearing systems, in the White – Inyos is from the north. Which is typical of La Nina or neutral.

March 20, 2008 2:08 pm

My exposure to tree rings was back in 76-77 when a visiting tree growth ring expert from the US gave a seminar on the water flows in the Colorado river. He thought that in the Colorado catchment tree growth rings going back 400 years were an excellent paleo rain gauge. Didn’t say anything about temperatures.

Bill Illis
March 20, 2008 5:59 pm

Just look at the picture of those trees. They are typical of the bristlecone pine trees we are talking about.
How can you possibly use a core sample to detect any kind of signal from that. Half the tree is dead at any one time and there is no bark on it on the other half of it.
The tree might be long-lived but any particular ring taken from a core drilled through any particular angle on the tree cannot tell you what the temperature was in 1716. Half the tree was dead in 1716, an maybe one little section was growing rapidly while nearby sections were trying to recover from insects in 1715.
Great tree, but the overlapping jumble of live and dead sections are not the “sweetspot’ for temperatures over the past 1,000 years. Mann picked them to use as his heaviest weighted data source because the jumbled mess of data produced can easily be manipulated.

March 20, 2008 8:08 pm

Hey, I just came across this thread. I’ve been commenting over at Climate Audit on this very subject. Between my comments over there and finding this thread, I came across some information which everyone might find interesting. It’s a discussion on BCP ring widths by Harold C. Fritts. The paper can be found here:

Chuck Bradley
March 20, 2008 9:13 pm

The excluded sites seem to be mostly in the very wet Pacific Northwest.
It would be interesting to see what would happen if they were included.
Even better, to see what would happen if they were the only data.
It is good to know the size of the cherry picking effect.

March 21, 2008 1:19 am

Sometimes very simple things have a vast impact on science. Therefore SteveMcIntyre deserves a Nobel Prize.

steven mosher
March 21, 2008 4:27 am

i pointed this out before to stmc. look at 1850. now google drought 1850

Jeff Alberts
March 21, 2008 7:18 am

The excluded sites seem to be mostly in the very wet Pacific Northwest.
It would be interesting to see what would happen if they were included.
Even better, to see what would happen if they were the only data.
It is good to know the size of the cherry picking effect.

The PNW isn’t that wet. Seattle gets less annual rainfall than Dallas and New York City. It’s just that it tends to be an all-day, slow drizzle affair when it does happen. I’m sure there are some parts of the PNW that are very wet, but it’s not everywhere up here.

March 21, 2008 9:12 am

The excluded PNW sites are in various places. Some of them are west of the Cascades, and therefore, do not experience a very pronounced lack of summer moisture. Of course, believe it or not, east of the Cascades, in many years, the SW monsoon, while not reaching the area per se, injects enough moisture into N Nevada / the southern Columbia Plateau, that there is enough available for subsequent local convective events. So, even east of the Cascades, there is at least an even chance in any given year that summer moisture may be enough to prevent slowing of growth owing to drought stress.
One site I like is on Vancouver Island. Absolutely no moisture limitation. If there are treemometers anywhere, they are in the Marine West Coast climate zones.

Jim Arndt
March 21, 2008 10:56 am

Sorry but this is a little OT. Seeing Tamino like the BCP for his PCA. I is best described like this.

Jeff Alberts
March 21, 2008 2:26 pm

Sorry, to differ, Steve. But I live very west of the Cascades, and my grass is always brown in the summer, green in the winter. The year after I moved out here (2002) we went 88 days without rain on Whidbey Island, and that wasn’t very out of the ordinary. It may be different closer to the Cascades where clouds coming in from the west and north pile up against the mountains and produce more precip, but along the I-5 corridor, we don’t get a lot of rain in the summer. Like I mentioned before, Dallas, NYC, just about any major city gets more ran than Seattle annually.

March 22, 2008 5:28 am

One gets the feeling that examining tree rings is about as scientifically valid as examining the entrails of goats.

Jeff Alberts
March 22, 2008 10:05 am

One gets the feeling that examining tree rings is about as scientifically valid as examining the entrails of goats.
Only if those goats are grazing near Bristlecone Pines. 😉

Evan Jones
March 22, 2008 4:58 pm

According to Liebig’s Law, might it be prudent instead to use trees located in the cold climates where heat, not water would be the limiting factor?
BTW Does anybody of you know the percentage in the MBH98 graph which is taken The Bristlecone Pines Trees?
Per: It’s worse than you think. The 2005 paper, and on Mac’s site (ClimateAudit), it is claimed that the BCP samples are weighted 390 times as much as the others.

Verified by MonsterInsights