Click images for more details



Recent comments
Recent posts
Currently discussing

A few sites I've stumbled across recently....

Powered by Squarespace
« Environmentalists trashing the environment | Main | Mathematical models for newbies »

A whole new bias

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 slow-grower 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 big-tree 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.

(H/T Hockey Schtick)

PrintView Printer Friendly Version

Reader Comments (20)

Similar ideas are discussed in section 5.4.3 of Briffa and Melvin 2011

Also try googling Lee's phenomenon

Mar 20, 2012 at 2:25 PM | Registered CommenterJonathan Jones

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

And if CO2 produces the same results, how does one tell them apart?

Mar 20, 2012 at 2:32 PM | Registered Commenterjamesp

As soon as we found out about the divergence problem we should have dumped the idea of tree rings as a proxy for anything except, just maybe, the passage of time. We should not be worrying now about how to tweak some signal out of them, we should forget them entirely.

Mar 20, 2012 at 2:33 PM | Unregistered CommenterRhoda

"There are things we know that we know. There are things that we now know we don't know. But there are also things we do not know we don't know. And each year, we discover a few more of those unknown unknowns."

Donald Rumsfeld

Sounds a lot like climate science.

Mar 20, 2012 at 2:54 PM | Unregistered CommenterJack Maloney

It's a frigging useless model again. Stop It.

Mar 20, 2012 at 3:13 PM | Unregistered Commenterstephen richards

Get out there, in the field, and measure. You dolts.

Mar 20, 2012 at 3:13 PM | Unregistered Commenterstephen richards

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.

Mar 20, 2012 at 3:44 PM | Unregistered CommenterKnR

'Slow-growing survivorship'
"As a result, reconstructed growth rates in the distant past are biased toward slower growth. "

Which means, networks with slow-growers will not capture warmer temperatures of the past.

Mar 20, 2012 at 3:45 PM | Unregistered CommenterShub

You don’t need to know anything about dendrochronology to see that the possibilities of sample bias are numerous. The fact that trees are sampled at their “Limits to Growth” i.e. on tops of mountains or at the northernmost limit of their occurence means that you’re losing any trees which were too young to survive a particularly harsh winter. Since you need a certain minimum sample size, you’re almost certainly going to sample where trees are sufficiently numerous, and miss trees which were to young to survive that year, and therefore underestimate short term temperature variability.
It’s like doing a survey of obesity on the top deck of a bus. A lot of your potential sample just won’t make it up the stairs.
The scandal is that this paper wasn’t written before dendrochronoloy was used to overthrow history.

Mar 20, 2012 at 4:10 PM | Unregistered Commentergeoffchambers

Frankly, about the only use I see for tree rings in relationship to warming is in my fireplace.

Mar 20, 2012 at 4:30 PM | Unregistered CommenterDon Pablo de la Sierra

Another problem with treeline sampling is the fact that it’s based on a circular argument. You collect samples at the treeline because that’s where the treerings are temperature-sensitive. But if temperatures were different in previous centuries, the treeline was either higher (in warmer times), and the trees you want no longer exist, or lower (in colder times) and you’ve only got young trees to sample. So you can only do the experiment somewhere where temperatures haven’t changed.

If I’m talking rubbish, would someone please tell me. Otherwise I shall continue to cite this as an example of the fact that, very often, no particular expertise is necessary to see through the “science”.

Mar 20, 2012 at 7:57 PM | Unregistered Commentergeoffchambers

Funny things these tree rings.

"the world had been warming for centuries" - now thats not supposed to have happened - it was only supposed to start about 50 years ago

Mar 20, 2012 at 8:03 PM | Unregistered CommenterBarry


Well said - I've looked at quite a few backgrounds for the Team and can't find any evidence of plant physiology training.

I'd hypothesise that people who know anything about the variables that go into plant development or agronomy couldn't be simple-minded enough to believe that a single variable could be tracked over the course of a millennium or so through any parameters of an annual growth ring.

Those who think that the fourth vector of an fraudulently-calculated principal component analysis might be able to magically pick out this variable are snake oil merchants of the lowest order. It's exceedingly challenging to pick out a defining variable that impacts on crop yield from year to year with any degree of confidence, even when you've got a weather station within a few metres of your crop making several measurements every minute of temperature, relative humidity, incident and intercepted photosynthetically active radiation (PAR), windspeed, rainfall, soil moisture deficit and a host of other factors, and in conjunction with this you might be measuring crop development through growth stage analysis, yield unit development (e.g. sequential harvesting of seeds, leaves, etc.) and so forth.

Ignorance with a blend of wishful thinking, delivered with a blast of patronising sophistry.

Mar 20, 2012 at 8:16 PM | Registered Commenterflaxdoctor

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

And if CO2 produces the same results, how does one tell them apart?

And if increased precipitation produces the same results, how does one tell them apart?
And what about Mycorrhiza? No biologists seems to be involved in this tree-ring nonsense, only statisticians.

Mar 20, 2012 at 9:17 PM | Unregistered CommenterJohn Silver

Root and branch reform of palaeoclimatology is necessary.

Mar 20, 2012 at 10:51 PM | Registered CommenterPharos

Brienen is clearly a shrill paid for by the Fossil Fuel Denialist Industry.

Michael E, Mann.

Mar 20, 2012 at 11:24 PM | Unregistered CommenterDon Keiller

One might go through official AGU channels to 'get' Brienen

The divergence problem has been well discussed in the literature. Why is there any need for further work on it?

Brienen is 'strange'

This is part of a 'desperate ongoing disinformation campaign'.

This is part of a 'well-funded organized fossil fuel campaign'.

This paper could be legally actionable.

The peer-review process at this journal is completely broken

We should let this journal wither in disrepute

This is a 'biased attack paper'

Mar 21, 2012 at 11:27 AM | Unregistered CommenterShub

just for fun, revisit this post & comments

Mar 22, 2012 at 1:13 AM | Unregistered Commenterdougieh

ps. never noticed before, but on the above link 2nd catoon has caption "then a miracle occurs" spooky.

Mar 22, 2012 at 1:23 AM | Unregistered Commenterdougieh

Not being a scientist I am at the mercy of that species of terrestrial beings. I have seen no reference to the validity or otherwise of the Chinese tree ring study of Tibetan junipers. See the study summary on JoNova's blog. Cheers from still very soggy Downunder.

Mar 22, 2012 at 4:38 AM | Unregistered CommenterTommo

PostPost a New Comment

Enter your information below to add a new comment.

My response is on my own website »
Author Email (optional):
Author URL (optional):
Some HTML allowed: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <code> <em> <i> <strike> <strong>