Stomata atmospheric CO2 reconstructions and hockeysticks

[Andre]

The number of stomata on leafs (breathing pores) of certain plant species appears to be sensitive to the CO2 level in the atmosphere. If it is higher, it appears that these plants does not bother to make many stomata, whereas CO2 is scarce, it simply makes more stomata to extract the CO2 from the atmosphere.

So if we have fossil leafs of these species, with some accurate dating, it might be possible to reconstruct CO2 levels from the past and perhaps much better than ice cores, since there are several problems associated with the accuracy of ice cores.

A first thread about that concept is here. The University of Utrecht in the Netherlands appears to be the mark leader in this specialty and http://www.pnas.org/content/early/2008/10/03/0807624105.full.pdf:

Thomas B. van Hoof, Friederike Wagner-Cremer, Wolfram M. Ku¨
rschner†, and Henk Visscher 2008, A role for atmospheric CO2 in preindustrial climate forcing
PNAS September 30, 2008, 105 (39)

Abstract

Complementary to measurements in Antarctic ice cores, stomatal frequency analysis of leaves of land plants preserved in peat and lake deposits can provide a proxy record of preindustrial atmospheric CO₂ concentration. CO₂ trends based on leaf remains of /Quercus robur/ (English oak) from the Netherlands support the presence of significant CO₂ variability during the first half of the last millennium. The amplitude of the reconstructed multidecadal fluctuations, up to 34 parts per million by volume, considerably exceeds maximum shifts measured in Antarctic ice. Inferred changes in CO₂ radiative forcing are of a magnitude similar to variations ascribed to other mechanisms, particularly solar irradiance and volcanic activity, and may therefore call into question the concept of the Intergovernmental Panel on Climate Change, which assumes an insignificant role of CO₂ as a preindustrial climate-forcing factor. The stomata-based CO₂ trends correlate with coeval sea-surface temperature trends in the North Atlantic Ocean, suggesting the possibility of an oceanic source/sink mechanism for the recorded CO₂ changes.

The study refers to earlier work in Washington on Tsuga needles, which is reported in this PhD thesis. Note that the graphs of fig 1c and fig 5.4 on page 57 (Pdf:61) of the thesis appear to correlate reasonably both with maximums around 1000AD and 1350AD and a minimum around 1200AD giving it a certain robustness.

Furthermore, they make a temperature reconstruction based on the alleged dependence of temperature on CO2 (Fig 2) with a 0,1-0,2C order of magnitude temperature variation. See a concluding remark:

Reconstructed multidecadal changes are not as prominent as man-made CO2 increases since the onset of industrialization. Yet it seems obvious that a dynamic CO2 regime with fluctuations of up to 34 ppmv implies that CO2 can no longer be discarded as a forcing factor of preindustrial air-temperature changes.

I find it curious however that they did not test their work to other temperature reconstructions in that period, If we do that, for instance with the new & improved hockey stick discussed here, with an minimum at around 1350AD, instead of a maximum, one would tend to think that it would support the CO2 cooling hypothesis rather than the CO2 warming hypothesis.

 

[wolram]

Ice core deposits, has anyone taken into account of which way the wind blows and how fast?

 

[Andre]

No ice cores here Wolram, but that would be a factor

Anyway, Comparing the stomata CO2 to climate reconstruction of Van Hoof et al 2008 (fig 2) with Mann et al 2008

http://www.pnas.org/content/early/2008/10/03/0807624105.full.pdf

Digitized, the average of the two reconstructions of fig 2 of Van Hoof et al in red below Mann's graphs with scales adjusted:

Not much to say here. No support for a strong correlation between temps and CO2

 

[Bored Wombat]

Ice core deposits, has anyone taken into account of which way the wind blows and how fast?

Yeah, they analyse those ice cores with everything you can think of and then some. They reconstruct the pattern of wind direction and speed from dust and pollen in the cores.

 

[Bored Wombat]

So if we have fossil leafs of these species, with some accurate dating, it might be possible to reconstruct CO2 levels from the past and perhaps much better than ice cores, since there are several problems associated with the accuracy of ice cores.

This is a really interesting paper.

I'm impressed how well the ice core from D47 (light blue) lines up with the smoothed CO2 reconstruction from the stomatal index. (looking at panel D)

I'm somewhat surprised how much smoothing there is in the ice core data. It looks like a 1 or 2 century low pass filter. I assume this is from the voids connecting in the snow on the top tens of metres.

The Law Dome (purple line) looks dead flat over the whole 500 years, like is has a lowpass filter longer than 500 years, or perhaps about 500 years. As you might expect, since it is bollocksing cold up there ... you couldn't get much precipitation so the top tens of metres would represent a long time period than most any other part of the planet.

Having said that, a brief google failed to reveal to me where, by comparison the D47 core is. Is there a map of these things somewhere? Or a layer than can be imported into google earth?

 

[Andre]

The D47 core is Antarctica for sure but I need to look for the exact location.

The low bypas filter effect of gasses in ice cores is well documented. Roughly the upper 90 meters of a ice sheet is open and poreous. This snow/ice transition is known as "firn" So the gasses are free to diffuse in this areas and spikes of trace gasses tend to smoothen out. The rate at which this happens is highly dependent on the accumulation rate of the snow, ranging from meters to centimeters per year. The same effect is also causing the difference in ice age and gas age in it's bubbles, ranging from a few dozen years to a few millenia.

 

[Bored Wombat]

The low bypas filter effect of gasses in ice cores is well documented. Roughly the upper 90 meters of a ice sheet is open and poreous. This snow/ice transition is known as "firn" So the gasses are free to diffuse in this areas and spikes of trace gasses tend to smoothen out. The rate at which this happens is highly dependent on the accumulation rate of the snow, ranging from meters to centimeters per year. The same effect is also causing the difference in ice age and gas age in it's bubbles, ranging from a few dozen years to a few millenia.

Right.

But other data sets that one sees look a lot better than this one in terms of temporal resolution in ice cores. The top chart from this graphic shows CO2 from the Law dome (DE08, DE08-2, and DSS ice cores), Siple Station, and the Taylor Dome, Dome C and DML ice cores.

http://www.globalwarmingart.com/wiki/Image:Carbon_History_and_Flux_Rev_png"

The temporal resolution looks to be within a few years, with the data that coincides with the flask measurements from Mauna Loa (1959 to 1978) exhibiting none of the lag that one would expect from a 100 year low pass filter.

What am I missing? Further mathematical treatment of the CO2 concentration data to achieve higher temporal resolution isn't mentioned in http://cdiac.ornl.gov/trends/co2/lawdome.html" discussion of the Law Dome cores.

 

[Andre]

What am I missing?

Good point, but the question is, how was the age of the ice core gas obtained? It is suggested that this was done by tuning the CO2 records to the Mauna Loa records. In that case it is circular reasoning. Obviously the current higher CO2 levels are slowly diffusing down in the firn of the ice core and tuning that way makes the air appear younger than it actually is.

So this is an interesting statement from that link:

Etheridge et al. (1996) determined the sealing depth at DE08 to be 72 m where the age of the ice is 40±1 years; at DE08-2 to be 72 m depth and 40 years; and at DSS to be 66 m depth and 68 years.

Meanwhile, realize that the ice at 70 meters with accumulation rates of centimeters per year at Dome C is still in the millenium scale.

 

[Bored Wombat]

Good point, but the question is, how was the age of the ice core gas obtained? It is suggested that this was done by tuning the CO2 records to the Mauna Loa records. In that case it is circular reasoning. Obviously the current higher CO2 levels are slowly diffusing down in the firn of the ice core and tuning that way makes the air appear younger than it actually is.

Right. If the estimate of the mean age of the trapped air is about right, then there would be no lag. (Except because of curvature of the CO2 concentration vs time, but that would be much less significant).

And they got out all their boffins to make that estimate pretty good, and I should certainly not expect a lag.

"The effects of diffusion in the firn on the CO2 mixing ratio and age of the ice core air were determined by analyzing air sampled from the surface down to the bubble close-off depth." (From the http://www.agu.org/pubs/crossref/1996/95JD03410.shtml").

(I wonder if they mean http://en.wikipedia.org/wiki/Thermophoresis" as well?)


An abrupt air-temperature change causes a temperature difference between the snow surface and the bubble-trapping depth, and this temperature difference then relaxes over a century or so as the deeper layers adjust to the new surface temperature. Temperature gradients cause gas-isotope fractionation by the process of thermal diffusion, with heavier isotopes migrating toward colder regions. Diffusion of gases through pore spaces in firn is faster than diffusion of heat, so the isotope signal reaches the bubble-trapping depth before the heat does, and the isotope anomaly is recorded as the air is trapped in the bubbles (http://www.pnas.org/content/97/4/1331.full")

So this is an interesting statement from that link:

Meanwhile, realize that the ice at 70 meters with accumulation rates of centimeters per year at Dome C is still in the millenium scale.

Ah, yes, you remind me that it is Dome C and not the Law dome, as I was thinking in the GGGP post, that is the cold one with not much precipitation. The Law dome in near the coast towards Western Australia, and has a high accumulation rate.

Yes, you would expect Dome C cores to have a very low temporal resolution, by comparison. (But still; what a phenomenal resource of climate history these Ice cores have been! Who would have dreamed that we could have actual air samples from 300 000 years ago to analyse?)