Climate Science Glossary

Term Lookup

Enter a term in the search box to find its definition.

Settings

Use the controls in the far right panel to increase or decrease the number of terms automatically displayed (or to completely turn that feature off).

Term Lookup

Settings


All IPCC definitions taken from Climate Change 2007: The Physical Science Basis. Working Group I Contribution to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Annex I, Glossary, pp. 941-954. Cambridge University Press.

Home Arguments Software Resources Comments The Consensus Project Translations About Donate

Twitter Facebook YouTube Pinterest

RSS Posts RSS Comments Email Subscribe


Climate's changed before
It's the sun
It's not bad
There is no consensus
It's cooling
Models are unreliable
Temp record is unreliable
Animals and plants can adapt
It hasn't warmed since 1998
Antarctica is gaining ice
View All Arguments...



Username
Password
Keep me logged in
New? Register here
Forgot your password?

Latest Posts

Archives

Plant stomata show higher and more variable CO2 levels

What the science says...

Stomatal data is not as direct as ice core measurements and hence not as precise.

Climate Myth...

Plant stomata show higher and more variable CO2 levels
"When stomata-derived CO2 (red) is compared to ice core-derived CO2 (blue), the stomata generally show much more variability in the atmospheric CO2 level and often show levels much higher than the ice cores." (David Middleton)

Shortly after F. Wagner published his stomatal results (here), a response appeared in Science. The key difference in the result can be seen in the figure:


This figure shows that Wagner's data shows a sharp increase to 330ppm at 11,260 years BP (years before 1950), staying there for 500 years, in disagreement with the Taylor Dome and Vostok ice-core records.

In deciding between these results, several items should be noted:

Firstly, ice-core CO2 measurements are direct measurements on air that has been enclosed in bubbles. On the other hand, stomatal density is an indirect measure. Experiments on stomata density showed that "the stomatal response to increasing atmospheric CO2 was identical to that induced by removing water from the plant roots" (Idso et al 1984). In other words, stomatal index data may not be the able to measure the atmospheric concentration as precisely as its proponents would like.

Secondly, several different ice-core data sets are essentially consistent. Artifacts do appear in earlier ice core records - mainly the Greenland drill sites where CO2 was depleted through a chemical reaction - but there are no such indications of this in the Taylor Dome ice core. In any event, this is a known phenomena, and one that can be accounted for. These records all indicate the CO2 concentration from 260 to 280 ppmv during the preindustrial Holocene.

Stomata data, on the other hand, do not show such agreement. For example Beerling et al (D. J. Beerling, H. H. Birks, F. I. Woodward, J. Quat. Sci. 10, 379 (1995)) report largely scattering proxy CO2 values from 225 to 310 ppmv between 9940 and 9600 14C-yr, in disagreement with the data presented by Wagner et al.

In summary, the skeptics claim that stomatal data falsify the concept of a relatively stable Holocene CO2 concentration of 270-280 ppmv until the Industrial Revolution. This claim is not justified.

Last updated on 3 January 2011 by Glenton Jelbert.

Printable Version  |  Offline PDF Version  |  Link to this page

Comments

Comments 1 to 10:

  1. Very nice summary, Glenton Jelbert. Thanks!
  2. Van Hoof et al., 2005 demonstrated that the ice core CO2 data essentially represent a low-frequency, century to multi-century moving average of past atmospheric CO2 levels.



    Van Hoof et al., 2005. Atmospheric CO2 during the 13th century AD: reconciliation of data from ice core measurements and stomatal frequency analysis. Tellus (2005), 57B, 351–355.

    Three separate lines of scientific evidence indicate that Antarctic ice cores underestimate the average global atmospheric CO2 concentration by 20-40 ppmv:

    1) GeoCarb
    2) Plant stomata
    3) NASA AIRS satellite measurements

    See my Debunk House post for an in depth discussion.

    By the way... Wagner, Van Hoof, Kouwenberg and the other botanists publishing papers on plant stomata & atmospheric CO2 are generally not "skeptics." They are looking for a pre-industrial coupling of CO2 & temperature. They aren't trying to debunk AGW.
  3. David Middleton at 07:00 AM on 1 August, 2010

    GeoCarb is a (very nice) model with 10 million year resolution. I don't see how you can use this to say that much about what true [CO2] levels were in the past at high resolution. It certainly doesn't "trump" direct measurements.

    Plant stomata frequency estimates of past [CO2] have large uncertainties (e.g. +/- 30-60 ppm; see a recent analysis of reconstruction uncertainties in Betula nana leaves). I think these studies are fine, and useful for estimating broad atmospheric [CO2] levels (or changes in [CO2] levels) in the deeper past, but one should accept that these are not precise measures.

    The NASA AIRS comparison is misleading since these are generally shown as snapshots. If one averages a full years worth of AIRS data, then the yearly averaged difference between polar and equatorial (say) [CO2] is only a few ppm. It's not reasonable to compare AIRS snapshots with ice core data (or stomatal data for that matter!) which is significantly temporally averaged.

    Obviously the mechanism for sealing off atmospheric samples in ice cap or glacial firn results in a considerable multiyear averaging of the atmospheric [CO2]. In the high resolution Law Dome core the averaging is smallish, whereas in the deep Antarctic cores the averaging may encompass a large number of years (can't remember off hand, but this may be a hundred years or more???).

    However considering the high resolution Law Dome data and the last couple of thousand years, I don't see any basis for concluding that the ice core data is biased low as you suggest. Yes, it's smoothed (it's something like a 10 year running mean); but (just like contemporary [CO2] variation), we expect rather low amplitude variability in [CO2] at high resolution. Yes, the natural variability (likely largely ENSO-related, with perhaps some significant wildfire variability) encompassing a few ppm will have been smoothed out. But it's not reasonable to think that we are missing large jumps and falls in [CO2], apart from anything else, because largish non-ENSO-related increased [CO2] levels take a long time to drop, and so they should stil be observed in cores. In any case if we're not seeing them (i.e. high resolution, large amplitude jumps and falls in [CO2]) during the last 50 years of very high resolution measurement, what is the basis for expecting that these occurred in the past?

    Incidentally, I don't understand the reference to "skeptics" in your last paragraph. These guys/gals are just scientists working to improve their methodologies and obtain insight into the past. I don't think one should adopt the false notion that science is composed of groups of people that have one view of the science and others that are "skeptics". If the group of people you refer to were "trying to debunk AGW" they would be wasting their careers. Science simply doesn't work like that...


  4. David Middleton at 07:00 AM on 1 August, 2010

    GeoCarb is a (very nice) model with 10 million year resolution. I don't see how you can use this to say that much about what true [CO2] levels were in the past at high resolution. It certainly doesn't "trump" direct measurements.


  5. The Antarctic ice cores are not "direct measurements" of global atmospheric CO2. They are direct measurements of gas that filtered into snow and were eventually trapped in ice. They are an indication of what the atmospheric CO2 was in the air, near the ground over Antarctica.

    GeoCarb is a very low frequency model-derived function. The resolution is low; but it is far better than 10 million years in the Neogene.

    It shouldn't "trump" the ice cores. But it should be incorporated with the ice cores and the stomata data to a more complete "spectrum" of the CO2 "signal."
    Plant stomata frequency estimates of past [CO2] have large uncertainties (e.g. +/- 30-60 ppm; see a recent analysis of reconstruction uncertainties in Betula nana leaves). I think these studies are fine, and useful for estimating broad atmospheric [CO2] levels (or changes in [CO2] levels) in the deeper past, but one should accept that these are not precise measures.

    Plant stomata are a lot noisier than ice cores. One doesn't just throw out the high frequency data in signal processing just because it's noisy.
    The NASA AIRS comparison is misleading since these are generally shown as snapshots. If one averages a full years worth of AIRS data, then the yearly averaged difference between polar and equatorial (say) [CO2] is only a few ppm. It's not reasonable to compare AIRS snapshots with ice core data (or stomatal data for that matter!) which is significantly temporally averaged.

    Then maybe you can show me an AIRS image that shows the polar regions to have higher CO2 levels than the mid and low latitudes. If it all averages out over the year, the polar regions would have to have higher CO2 levels at some point during the year. Every daily AIRS image I've seen, shows the polar regions to have 15-20 ppmv less CO2 than the mid and low latitudes.

    Getting the monthly average differential down to 5-10 ppmv and the annual average differential down to a few ppmv is a neat trick, considering how sparsely sampled the polar regions are.


    Obviously the mechanism for sealing off atmospheric samples in ice cap or glacial firn results in a considerable multiyear averaging of the atmospheric [CO2]. In the high resolution Law Dome core the averaging is smallish, whereas in the deep Antarctic cores the averaging may encompass a large number of years (can't remember off hand, but this may be a hundred years or more???).

    It can vary from a few decades to more than two thousand years.
    However considering the high resolution Law Dome data and the last couple of thousand years, I don't see any basis for concluding that the ice core data is biased low as you suggest. Yes, it's smoothed (it's something like a 10 year running mean); but (just like contemporary [CO2] variation), we expect rather low amplitude variability in [CO2] at high resolution. Yes, the natural variability (likely largely ENSO-related, with perhaps some significant wildfire variability) encompassing a few ppm will have been smoothed out. But it's not reasonable to think that we are missing large jumps and falls in [CO2], apart from anything else, because largish non-ENSO-related increased [CO2] levels take a long time to drop, and so they should stil be observed in cores. In any case if we're not seeing them (i.e. high resolution, large amplitude jumps and falls in [CO2]) during the last 50 years of very high resolution measurement, what is the basis for expecting that these occurred in the past?

    We don't see those jumps in the plant stomata data either over the last 50 years. The stomatal response is consistent with a steady increase in atmospheric CO2 over the last 60 years...

    ]

    Wagner F, Dilcher DL, Visscher H (2005) Stomatal frequency responses in hardwood swamp vegetation from Florida during a 60-year continuous CO2 increase. Am J Bot 92:690–695.

    While species were less responsive to CO2 changes, the stomatal response of M. cerifera very closely tracked the actual changes in atmospheric CO2.

    Hopefully the West Antarctic Ice Sheet Divide Ice Core Project will yield higher resolution data (similar to GISP2) over the last 50,000 years or so. This may yield a CO2 "signal" similar to the stomata or something in between the lower frequency cores and the stomata.
    Incidentally, I don't understand the reference to "skeptics" in your last paragraph. These guys/gals are just scientists working to improve their methodologies and obtain insight into the past. I don't think one should adopt the false notion that science is composed of groups of people that have one view of the science and others that are "skeptics".

    The first post in this thread says:
    The skeptic argument...

    Plant stomata show higher and more variable CO2 levels


    That's actually what the science says. Hence, my reference to "skeptics."

    If the group of people you refer to were "trying to debunk AGW" they would be wasting their careers. Science simply doesn't work like that...

    Actually, that is exactly how science works. To quote Einstein, "No amount of experimentation can ever prove me right; a single experiment can prove me wrong." Science is all about skepticism and debunking (AKA testing).
  6. David Middleton at 02:06 AM on 4 August 2010

    Geocarb:

    The most up to date versions of Berner’s excellent model for long term carbon cycle processes [***] have a temporal resolution of 1 million years. Geocarb is a model. It isn’t a measure of [CO2] in the deep past, and it certainly can’t be used to assess [CO2] during the Neogene ice age cycles represented in Antarctic cores (see following paragraph). In any scientific analysis we use the appropriate tools. Geocarb, however nice a model, doesn’t give us insight into atmospheric [CO2] during the last 700,000-odd years where we have highish resolution direct measures of atmospheric [CO2] from ice cores.

    I wonder whether you have read the Geocarb papers? These show zero variation of [CO2] during the Neogene ice age cycles. If you look at the Geocarb papers you will see that the [CO2] levels in the deep past (Phanerozoic) are defined as “RCO2 is the ratio of the mass of atmospheric CO2 at a past time to that at present (weighted mean for the past million years”) (see e.g. legend to Figure 18 of the linked paper).

    David, if Geocarb sets the baseline for atmospheric [CO2] by averaging the last million years, how can we possibly use Geocarb to assess specific atmospheric [CO2] in ice cores?Your argument simply doesn’t make sense. Geocarb is a model for assessing our ability to constrain [CO2] levels in the deep past through our understanding of long term (multi-million year) carbon cycling. Please read the paper(s).


    Plant stomata:

    ”Plant stomata are a lot noisier than ice cores. One doesn't just throw out the high frequency data in signal processing just because it's noisy.”

    The problem is that the noise (+/- 30-60 ppm as indicated in the paper I linked to in my post you’re responding to) often overlaps with the difference between ice core [CO2] data and apparent stomatal [CO2] data. It’s not a question of “throw(ing) out high frequency data”. The question is whether the apparent differences between some (but not all) stomatal data and ice core data is statistically significant. I think we both agree that the ice core [CO2] data is temporally averaged due to the variably slow rates of firn sealing. However there isn’t any strong evidence that the ice core data is biased low as you insinuate. If the stomatal frequency boffins come at some point to a conclusion as to a reliable means of determining historic [CO2] with highish precision, then that will be great. In any case the high resolution Law Dome data is only temporally averaged on the decadal time scale.


    AIRS and polar [CO2]:

    We're quibbling (or you are I should say!) over a few ppm of [CO2]. You're happy to use a model for [CO2] in which [CO2] is averaged over the last 1 million years to attempt to counter the Antarctica ice core data, and yet you are fussing about a possible few ppm difference between Antarctic [CO2] and global [CO2].

    I think your quibbling is misplaced. Atmospheric [CO2] has been measured in Antarctica (South Pole) since the early 1990's. We can compare the directly measured South Pole [CO2] with the [CO2] measured at Mauna Loa or from the globally averaged sea surface sites (or with globally and yearly averaged AIRS data). The difference is small (a few ppm).


    How science works:

    The idea that scientists set out to "debunk" something is silly. Scientists set out to find stuff out. With careful experiment and analysis the real world leads them towards reliable interpretations of natural phenomena. If this happens to lead to a robust conclusion that is at odds with other interpretations then that's just great.

    I can't imagine a real world example where your assertion "Science is all about skepticism and debunking (AKA testing)." might have any meaning! Science is surely about the formulation of hypotheses, testing these with experiments/analyses and seeing where the latter lead.

    [***] Berner RA (2006) GEOCARBSULF: A combined model for Phanerozoic atmospheric O-2 and CO2 Geochim. Cosmochim. Acta 705653-5664.
  7. If you "can't imagine a real world example where" skepticism and debunking have meaning, you've probably never been an exploration geologist or geophysicist in the oil industry. Our "experiments" (exploration wells) are rather expensive. Every "hypothesis" (prospect) is subjected to a lot of skepticism and debunking before we run the main experiment.

    GeoCarb is based on a model derived from weathering rates and other geological factors. It says that the atmospheric CO2 concentration, averaged globally over the last 10 million years, has been about 267 ppmv. Contemporaneous plant stomata studies show CO2 oscillating between 270 and 360 ppmv over that same time period. Which is very similar to their oscillation range in the Sangamonian and Holocene.

    The average CO2 level from the ice cores is about 231 ppmv over the last 800,000 years (generally oscillating between 230 and 310 ppmv). This is 36 ppmv below the GeoCarb global Neogene average.

    Both GeoCarb and plant stomata show higher average CO2 levels than the ice cores do over the Neogene. The stomata data show much more variability in CO2 levels during the two most recent interglacials. We can quibble all day long about the AIRS data... But the daily images show that the polar regions have 10-20 ppmv lower CO2 levels than the mid to low latitudes and the monthly averages show the polar regions to have 5-10 ppmv lower CO2 levels than the mid to low latitudes.

    "The Science Says"

    Plant stomata show higher and more variable CO2 levels than ice cores.

    "The Science Says"

    GeoCarb and plant stomata show higher CO2 levels than ice cores.

    "The Science Says"

    AIRS shows higher mid and low latitude CO2 levels than ice cores show for Antarctica.

    "The Science Says"

    That the ice cores are not resolving decadal and century scale CO2 variations very well and that Co2 levels recorded in Antarctic ice cores should yield lower values than just about any other method used to estimate past global CO2 levels.
  8. David Middleton at 07:00 AM on 1 August, 2010

    “AIRS shows higher mid and low latitude CO2 levels than ice cores show for Antarctica”

    Slightly: Check out these recent visualizations of global CO2 measurements, but look at the Arctic values.

    You then state:

    “That the ice cores are not resolving decadal and century scale CO2 variations very well and that CO2 levels recorded in Antarctic ice cores should yield lower values than just about any other method used to estimate past global CO2 levels”

    I disagree. The following data from six different cores post dates that used in Van Hoof somewhat.


    (Source NOAA Paleo Ice core data).
  9. David, I'd say the science is pretty straightforward, especially with respect to Geocarb and AIRS:

    Geocarb:

    Geobarb has essentially nothing to say about absolute atmospheric [CO2] levels during the ice age cycles represented in deep Antarctic cores. Here’s what Berner says about his Geocarb model with respect to absolute values of his modelled [CO2], and the temporal resolution [***]:
    ” Results are expressed as RCO2 which is defined as the ratio of mass of CO2 in the atmosphere at time t divided by the mass at present, and the results are compared to a standard run, where best estimates of the various input parameters are used. To convert RCO2 to CO2 concentration, because of appreciable errors inherent in this kind of modeling, a rough value of 300 ppm can be used to represent “the present.””
    And:
    ”It should be emphasized that GEOCARB modeling has only a long time resolution. Data are input into the model at 10 my intervals with linear interpolation between. In the case of rock abundance data, averages for up to 30 my time slices are sometimes used. Thus, shorter term phenomena occurring over a few million years or less are generally missed in this type of modeling.”

    Clearly the model (which is constructed to assess weathering-dependent carbon cycling during the Phanerozoic) doesn’t attempt to determine precise [CO2] levels (”because of appreciable errors inherent in this kind of modeling, a rough value of 300 ppm can be used to represent “the present.”). And since it has 10 million year resolution with linear interpolation (up to 1 mya resolution in the Geocarbsulf version I linked to above), it can’t really say very much about the [CO2] levels during the ice age cycles. That’s pretty clear isn’t it David? We should take account of what the inventor of the model says about his model!

    AIRS and graces:

    You point out that daily AIRS retrievals can show lower mid-tropospheric [CO2] at the poles than at mid-latitudes. In monthly averages the differences are smaller (and it depends on the month of course, sometimes the polar regions have higher [CO2] than mid-latitudes).

    So why not take your point to its logical conclusion? Averaged yearly the differences are small (a few ppm). Since, in assessing fossil leaf stomata, or ice cores, the data is at least yearly averaged, we really should consider yearly-averaged AIRS data in assessing any difference between globally averaged and polar [CO2], yes? In which case your argument based on daily or monthly AIRS retrievals has little merit.

    stomata:

    Yes, there are apparent differences in some of the stomatal [CO2] reconstructions compared to the [CO2] measured in Antarctic cores. We all agree that [CO2] in ice cores are temporally averaged. For the high resolution Law Dome core, some stomatal data are in broad agreement with the ice core data, and some differ. In all cases the stomatal scientists highlight rather large uncertainties in their data. At some point the stomatal scientists will no doubt come to some consensus about the best methods for addressing the fossil stomatal data.

    Otherwise, one needs to be very careful not to over-interpret stomatal data. They have very large associated uncertainties (this is obvious from inspection of the training sets used to generate calibration curves). Since there is no absolute relationship between [CO2] and stomatal indices, the data can only be used for species that still exist today (in order to calibrate the particular analysis), and one has to make the assumption that the physiological response to varying [CO2] is the same 100’s of 1000’s to millions of years ago, as now. Fossil plant stomatal analysis is a nice method for estimating low resolution [CO2] levels and changes in the deep past. But there isn't really a basis for discounting ice core data in favour of stomatal data. Apart from anything else it’s difficult to imagine a process that would give rise to a depth-independent 20% reduction in [CO2] levels that is exactly constant between numerous different high resolution and low resolution cores through 1000’s of metres of ice.

    [***] R. A. Berner and Z. Kothavala (2001) Geocarb III: A Revised Model of Atmospheric CO2 over Phanerozoic Time American Journal of Science 301, February 182-204.
  10. (Reader Matthew posted this in a different thread at 12:54 PM on 10 January 2011;
    as a courtesy I'm reposting it here.)



    Using plant stomata to determine carbon dioxide concentrations over the past 15,000 years

    A skeptic posted this and I went to know what you think of it. Thanks.
  11. Last time this stomata question came up, the geocraft.com site featured prominently (but that was on a different thread). Oddly enough, one point never gets mentioned by the deniers: The error bars around stomata reconstructed CO2 values tend to be +/-50ppm. See Figure 6 in Jessen et al 2005 (one of the sources for the graph in #9).

    Put those error bars on the orange line around 300ppm in the graph in #9 and it is no different from the ice core data.

    And the deniers claim tree ring data is suspect!

Post a Comment

Political, off-topic or ad hominem comments will be deleted. Comments Policy...

You need to be logged in to post a comment. Login via the left margin or if you're new, register here.

Link to this page



The Consensus Project Website

TEXTBOOK

THE ESCALATOR

(free to republish)

THE DEBUNKING HANDBOOK

BOOK NOW AVAILABLE

The Scientific Guide to
Global Warming Skepticism

Smartphone Apps

iPhone
Android
Nokia

© Copyright 2014 John Cook
Home | Links | Translations | About Us | Contact Us