CO2: Ice Cores vs. Plant Stomata

David,
thank you for the very interesting post. Could you please modify the images to have links to individual pages? I find that the graphs extend out into the right sidebar and are partially covered by the items in the sidebar. For example, I cannot read the caption on the Wagner et. al. image because it is partially covered by the right sidebar.
Links that open the images in their own pages would solve this minor problem and would be greatly appreciated.
Best regards,
Tom Moriarty
ClimateSanity.wordpress.com

Might we be able to reconstruct the greening of the planet during these episodes of stomata changes? I am guessing there is a correlate lag similar to CO2 reconstruction. Fossil remnants would be useful to measure this, especially around the supposed edges of such “greening”.

Interesting and informative and obviously took some time and effort to put together.
Thankyou.

Nice breakdown. I’ve been questioning the ice core data for a while. You do have to marvel at how the major core reference used by the Alarmist side is usually Lonnie Thompson’s, but no one can find his actual data to confirm his work. Makes getting Phil Jones’ homework look simple.

Good post David.
Just one comment: The stomata-based CO2 estimates seem to be generally accurate, but they do exhibit a lot of varibility which means there is a large error margin in the methodology for individual estimates. They should probably be averaged over some longer time period.

Wow! So thorough, so methodical, very impressive. You should publish it in a peer reviewed publication.

Pamela Gray says:
December 26, 2010 at 8:58 am
Written at a level and without jargon, that most can readily understand.
Yes, very well done.
Thanks David.

From one geoscientist to another: nice piece of work.
Thanks for putting this together.

@ tommoriarty: If you will right click on the image, you may select view image or open image. That works in Firefox, Opera and Chrome/Safari. I have no clue how to do it in IE.
cheers,
gary

@Middleton
The amount of anthropogenic CO2 released into the atmosphere is one of the better known quantities in this debate. The directly measured increase in atmospheric CO2 if only about half of the anthropogenic emissions. Clearly something is sequestering about half the anthropogenic CO2 emissions. For you to say that the increase is the result of ocean outgassing instead of anthropogenic emission frankly makes very little sense. Environmental sinks do not discriminate between CO2 molecules by source. It almost seems writ in granite that the oceans are not currently a source of CO2 but are rather a sink of CO2 and they aren’t sinking it as fast as anthropoids are sourcing it.
There are a few things in the global warming controversy which are almost beyond reasoned debate. The measured rise in CO2 being due to human activities is one of those few things.

Thanks for the very informative post. The general point– that CO2 may have had more variability than ice cores data can show is a valid one, but one should use caution when talking about the accuracy of stomatal densities as their response to CO2 can be quite variable and nolinear and may also be related to other factors such as humidity and temperatures. This is a new area of research with even the lead researchers acknowledging the uncertainties of the science. I am fairly certain however that CO2 was more varible in the past than assumed by the IPCC but the fidelity of the ice core measurements would not yeild such data, however, the longer term TRENDS of atmospheric CO2 over the past 800,000 years as displayed is the ice cores is very likely correct. As stated, stomatal density as a proxy for past CO2 levels is still a relatively young science with known issues and problems that are well documented and caution is warranted before making large assumptions based this new field.
A few conclusions that are probably safe to make, drawn from studies such as this:
http://www.pnas.org/content/early/2008/10/03/0807624105.full.pdf
“A coherent scenario explaining preindustrial atmospheric CO2
changes of the last millennium and their possible temporal link
with changes in terrestrial and marine carbon uptake or release
still needs to be established. 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. The results of our study
therefore underscore the need to understand anthropogenic
global warming within the context of rates and amplitudes of
natural CO2 variability of the last millennium. A stomata-based
CO2 record may provide an important observational constraint
on the sensitivity of climate models.”

@David: I’ll consider this a late X-mas present. Are you planning on taking this report forward to a scientific journal?

Dave, the natural closed systems (such as human population growth as a direct contributor to the recent rise in CO2) have well known closed system lags and unbalances. This is well established in proxies such as ice cores and stomata data mined for CO2 data. The current rise might very well be due to a closed system on the rising side of the closed equation that includes a lag time in response. It may also be due to a faster out of control rise on one side of the closed equation with a slower catching up on the downside of a closed equation. No where is it scientifically stated that a closed system is always instantaneously balanced. Nature is necessarily unbalanced and is always under pressure to change, which is likely what makes it so robust.
Logically, eventually as the human animal population increases as it is currently doing in dramatic fashion, the ability to sustain such a growth will crumble. Left unchecked, we are breeding a future population of increasing starvation and eventual population deflation. But that is the nature of closed systems. We are no different than a bloom of locusts if viewed from such a lens.
In my opinion, your view of CO2 cycles and data is myopically biased by emotional belief.

@Pamela
“In my opinion, your view of CO2 cycles and data is myopically biased by emotional belief.”
I’m about as emotionally biased as an eggplant. You’re projecting.

Warren,
When diagrams are used from previous reports (note the plurals) they are usually used as originally constructed; maybe with something new added. To reverse them side-to-side (mirror) or top-to-bottom (flip) requires (a) the original data, or (b) tedious measurements from the original, and considerable graphics capability –any of which can lead to mistakes, and problems going back to the original source text of clarification.
Try doing such a thing. Take the Pleistocene CO2 vs Temperature chart with four lines (one straight orange one) and five balloon style text boxes and (mirror) it so the x-axis has 0 on the right and 1 on the left. This is easy in a graphics program but the letters and numbers in the chart and in the caption will have to be recreated. And, then as mentioned, the original referenced text (perhaps) will be scrambled.
———————————————————
David, Nice post. Thanks.
Good comments, also. Thanks to you folks.

@Pamela
“Logically, eventually as the human animal population increases as it is currently doing in dramatic fashion, the ability to sustain such a growth will crumble. Left unchecked, we are breeding a future population of increasing starvation and eventual population deflation.”
Paul Erlichism appears to rise and fall in correlation with dress hem length. In the meantime technological advances continue to outpace population growth. Average life expectancy continues to increase even as the number of individuals living those lives increases. Make a logical case for why technological advance cannot continue to outpace population growth and I’ll consider the merits of it. Good luck.

One should be reminded that there is a fourth method of measuring CO2, as was discussed here the compilation of the late Ernst-Georg Beck of the 200 year record of chemical measurements. These also show large variations of CO2.
The way I have been looking at these discrepancies is not in terms of noise, but in terms of geographical and height variations, as happens also with temperature. In contrast to the method of getting a global average from the surface temperatures, the CO2 is measured at high and remote places, and with an orthodox method of analysis outlined by all the Keeling et al publications.
Ice cores are far away from biological sources. Stomata by construction are not, and will reflect values similar to the values seen in the Beck compilations.
Mauna Loa measure next to a volcano, and volcanoes are known sources of CO2. They have a funny way of throwing away outliers in the analysis. All these were discussed previously.
As one does not go to the top of a mountain and measure the temperature and call it the global temperature one should not declare the Mauna Loa values as global either.

Excellent presentation!
As we know CO2 is a good thing; the more their is, the more life on the planet and the inverse. As I understand, at about 150ppm terrestrial plant growth stops, and being the basis of the food chain, would be a big problem.
We know that atmospheric CO2 has been in the long term decline. The Phanerozoic CO2 vs Temp chart clearly shows the slow long term decline over the last ~530MY. Extending the trend line of the peaks shows, that unless the system will naturally bounce off the 150ppm point, it may break through in the next few MYs (1-10?). Not that it is a personal worry and our species may be long gone anyway.
However, as has been shown here and in multiple presentations, as temp goes down, CO2 follows 230-800 years later.
Of more pressing concern should be the unavoidable imminent climate shift back to the normal state of full-on glacial advance and the following large CO2 and food production decline. I just fail to understand why that is not the big debate; how will civilization survive in any recognizable form when the planet quickly shifts from a carrying capacity of 8 Bil people to 2 Bil people and large regions must be abandoned; such as the UK, Northern Europe, Canada, New York, Chicago….? What is the plan? This should be the focus of even marginally responsible governments. It probably secretly is, but since time is running short, and once the shift is clearly underway, too late; it needs to be made public; what the hard decisions will be. Ultimately people are responsible for themselves, but how the future population will best fit and work together should not be left as a State secret. I probably won’t live to see it, but our decedents, at some point, will.

Excellent and extremely useful synthesis and summary of this topic David, many thanks. The figure on Phanerozoic temperature (Veizer) against CO2 by stomata and GEOCARB is hugely significant – in a way a mutual validation of stomata and GEOCARB.
One small comment – figure numbers would be helpful.
The conclusions are solidly grounded and hard-hitting.

DEL says:

Of more pressing concern should be the unavoidable imminent climate shift back to the normal state of full-on glacial advance and the following large CO2 and food production decline. I just fail to understand why that is not the big debate; how will civilization survive in any recognizable form when the planet quickly shifts from a carrying capacity of 8 Bil people to 2 Bil people and large regions must be abandoned; such as the UK, Northern Europe, Canada, New York, Chicago….?

I have one word for you: TIMESCALES. The rate of warming out of the glacial periods was on the order of 0.1 C per century and the descent into the glacial periods was even slower. So, even if the human emissions of CO2 weren’t already enough to at least delay the next glacial period, we would have plenty of time to get ready for it.
The temperature record does show some more rapid changes occurring but these were probably more regional in nature (i.e., some regions warming, some cooling, some not changing much) and seemed to occur during the glacial periods or during the periods when the climate was already changing rapidly. At best they are irrelevant for current considerations and at worst they warn us that one applies forcings to a very complex and nonlinear climate system at one’s peril!

David, there are far more severe problems with the stomata data and the historical measurements (pre-Mauna Loa) than with the ice core data.
The main problem is that stomata by definition are from leaves of growing vegetation on land. That means that the local CO2 levels are highly variable from day to day (and night) and year by year. Even if the stomata (index) data are calibrated (+/- 10 ppmv) against ice cores (!) over the past century, there is not the slightest reason to expect that the same calibration is valid over previous centuries: Take e.g. one of the main datasets of The Netherlands: St. Odiliënberg, South Netherlands: in todays main wind direction a lot of industry and trafic of one of the densiest populations of the world. In the previous centuries: a lot of changes from woods and marshes to agriculture and industry/traffic. Even the main wind direction (and CO2 levels) might have changed from the MWP over the LIA to today.
The same problem for a lot of the late Beck’s historical data: taken at places where there were a lot of local CO2 sources and sinks. E.g. the 1942 peak in his reconstruction is mainly based on two places: Poona (India), most measurements within crops, and Giessen (Germany), where modern measurements show a huge bias and extreme much variability. BTW, the 1942 “peak” doesn’t show up in stomata (index) data, neither in high resolution ice cores or (as d13C) in coralline sponges (2-4 years resolution).
Here the modern data from Giessen, compared to the same days measured (raw data!) at Mauna Loa, Barrow and the South Pole:
http://www.ferdinand-engelbeen.be/klimaat/klim_img/giessen_background.jpg
Do you think that historical data from Giessen (3 samples/day taken at 7 AM, 2 and 9 PM, that alone gives a bias of +40 ppmv) have any value for historical background CO2 levels?
Further, the ice cores CO2 measurements of Greenland are unreliable, because of the frequent acid dust from Icelandic volcanoes deposited on the ice. Together with seadust carbonate deposits that gives extra CO2.
Ice core problems like the ice age – gas age difference are difficult to resolve, but have no influence on CO2 levels at all. But there is an averaging over several (10-600) years of the gas composition during narrowing of the pores with depth. Thus at bubble closing time the gas composition is not of today, but a mix of some to many years.
The oceans are NOT the cause of rising CO2: the direct solubility of CO2 in seawater changes with about 16 ppmv/°C, by far not enough to explain the 100 ppmv rise since 1850 or even the 60 ppmv rise since 1958. Because of the negative feedback from vegetation, the real response of CO2 on temperature is about 8 ppmv/°C (ice ages – interglacials, MWP-LIA) on long term, some 4 ppmv/°C for short term variations around the trend.
And the geocarb is based on proxies, while the inclusion of air in ice cores allows a direct measurement, thus is not a proxy, quite accurate, but smoothed, if handled with care.
See further my take on the CO2 increase at:
http://wattsupwiththat.com/2010/08/05/why-the-co2-increase-is-man-made-part-1/ about the mass balance of man-made CO2 emissions and the increase in the atmosphere.
http://wattsupwiththat.com/2010/08/20/engelbeen-on-why-he-thinks-the-co2-increase-is-man-made-part-2/ about the reliability of ice cores.
http://wattsupwiththat.com/2010/09/16/engelbeen-on-why-he-thinks-the-co2-increase-is-man-made-part-3/ about additional evidence from isotope compositions.
http://wattsupwiththat.com/2010/09/24/engelbeen-on-why-he-thinks-the-co2-increase-is-man-made-part-4/ about where one finds “background” CO2 and further discussion of Beck’s historical data and stomata data.

“The advantage of Antarctic ice cores is that they can provide a continuous record of relative CO2 changes going back in time 800,000 years, with a resolution ranging from annual in the shallow section to multi-decadal in the deeper section.”
That’s not true.
Here is the Historical CO2 Record from the Vostok Ice Core. As you can see the difference between Age of ice and Mean age of air in it is anywhere between 1879 and 6653 years (at depth 506.4 m and 3119.51 m respectively).
Therefore it takes several millennia for carbon dioxide to get enclosed in Antarctic ice. It means the resolution is much worse than several decades, even century scale spikes are smoothed out completely.
Temperature resolution is better, since in this case stable oxygen isotope ratio of ice itself is used as a proxy and water in ice crystals don’t move much once frozen.
If you want better resolution for carbon dioxide, you should choose a site where snow accumulation is fast (South-Eastern Greenland?). This way you’ll have shorter records with gas bubbles getting enclosed faster. Depth of firn-ice transition zone (about 90 m in Antarctica) mostly depends on pressure, so it does not vary much between sites.
But even in this case multi-decadal resolution (for CO2) is like pie in the sky.

jorgekafkazar says:
December 26, 2010 at 1:54 pm
The emissions are based on a global inventory of fossil fuel sales (taxes!) per country:
http://www.eia.doe.gov/iea/carbon.html
Maybe somewhat underestimated, but rather accurate for the past decennia, less reliable farther back in time.
The increase in the atmosphere is measured at some 70 “background” stations all over the world, where a few near sea-level stations are averaged to give a “global” trend.
Mauna Loa many times is referenced, because it has the longest record (the South Pole started before MLO, but misses a few years). But for the trend it hardly matters which station you use.
The increase in the atmosphere is about 53% of the emissions. This is near the same percentage over a 100+ years period (data before 1958 are less reliable):
http://www.ferdinand-engelbeen.be/klimaat/klim_img/acc_co2_1900_2004.jpg
For the most recent (post-Mauna Loa) period:
http://www.ferdinand-engelbeen.be/klimaat/klim_img/dco2_em.jpg
where the year by year variability is caused by temperature variations, which influence the uptake by nature, but hardly influences the trend.

Berényi Péter says:
December 26, 2010 at 1:55 pm
“The advantage of Antarctic ice cores is that they can provide a continuous record of relative CO2 changes going back in time 800,000 years, with a resolution ranging from annual in the shallow section to multi-decadal in the deeper section.”
That’s not true.
As you can see the difference between Age of ice and Mean age of air in it is anywhere between 1879 and 6653 years (at depth 506.4 m and 3119.51 m respectively).
Therefore it takes several millennia for carbon dioxide to get enclosed in Antarctic ice. It means the resolution is much worse than several decades, even century scale spikes are smoothed out completely.
There is a correlation between ice age – gas age difference and gas age smoothing (both depend of the accumulation rate), but the resolution of the gas age mainly depends on how long it takes to migrate through the pores vs. the speed of narrowing the pores. For the coastal Law Dome ice cores, the resolution is about a decade, for Vostok and Dome C about 600 and 560 years. A spike of 120 ppmv during 10 years or an increase of 2 ppmv during 600 years would be noticed in the Vostok ice core, but even a spike of 20 ppmv over one year would be measured in the Law Dome ice core.
The Greenland ice cores are not reliable for CO2, as the Icelandic volcanoes frequently added acid deposits over the sea salt (carbonate) inclusions.

Just happen to be reading Solomon’s ‘The Deniers’ at the moment. He explains that it was only possible to get the nice splice between the Siple ice core data and the Mauna Loa data by shifting the former forward by 83 years. Did the Law Dome data have to be adjusted also?

The article mentions uncertainties with ice cores, but it doesn’t mention any of the uncertainties and problems with stomata CO2 data.
The conclusion “The rise in CO2 that began in 1860 is most likely the result of warming oceans degassing.” is based on the second to last graph. This graph doesn’t make a lot of sense. The arrows draw with lag times of 250 years seem rather arbitrarily sketched and with only two cases coincidence is quite likely. The value 250 years seem to be based on anything other than the graph itself.
The graph does not make a case for “The rise in CO2 that began in 1860 is most likely the result of warming oceans degassing.”. The temperature fall from the MWP looks to be about 0.6C and the CO2 decrease looks to be only about 30ppm. Yet CO2 has risen far more than 30ppm since then. Basically the recent CO2 rise is far greater than you’d expect even from the temperature/CO2 relationship seen in that graph.
Additionally the ocean is currently absorbing more CO2 than it emits so the cause of the ongoing rise cannot be ocean degassing.
As such the conclusion “The anthropogenic contribution to the carbon cycle since 1860 is minimal and inconsequential.” is not supported.
The anthropogenic contribution to the carbon cycle since 1860 is minimal and inconsequential.

David, you also reran the good ole “temperature change preceded the CO2” story. I believe this has been debunked but you did not address the problems with this particular debating point.
Personally I am suspicious of the “compare the peaks of two curves” technique you use here because the relationships are not necessarily linear and noise makes the correct matching of leaks unreliable.
In addition it is widely believed that CO2 and temp are tied together in a feedback loop. This means that
1. changing the temp will change the CO2
2. Changing the CO2 will change the temp
So you have made an argument that temp affected CO2 in the past. Fine, sounds plausible. But that was then, this is now. We are dumping a whole lot of CO2 into the atmosphere NOW.

Ferdinand Engelbeen says:
December 26, 2010 at 3:10 pm
For the coastal Law Dome ice cores, the resolution is about a decade, for Vostok and Dome C about 600 and 560 years.
Would you explain if difference between mean age of air and age of ice in the same layer is more than two thousand years in the Vostok ice core, how the resolution is supposed to be as good as 600 years?

Engelbeen must be corrected here. The oceans contain 38000 Gt carbon and the atmosphere 700GtC. Henry’s law insists that about 1100GtC outgasses per ºC at 15ºC, which equates to over 600ppmC/ºC. Not 16ppmv/ºC as Engelbeen states. http://www.seafriends.org.nz/issues/global/acid.htm
Of course the deep ocean reacts slowly, as the 250 year delay between temperature and CO2 shows.
Lance Endersbee found a short term relationship between CO2 and temperature, at 150ppmv/ºC, still ten times higher than Engelbeen’s estimate. http://www.seafriends.org.nz/issues/global/acid2.htm
Most of the ocean’s outgassing is sequestered by terrestrial plant life, so what remains in air is just enough to speed terrestrial sequestration. http://www.seafriends.org.nz/issues/global/climate5.htm
The carbon budgets have some severe conflicts as explained in http://www.seafriends.org.nz/issues/global/climate4.htm which leads to only one conclusion: the oceans are outgassing and some of it remains in air. Human emissions play only a very small part, and that part has no whatsoever influence on temperature.

The anthropogenic contribution to the carbon cycle since 1860 is minimal and inconsequential.
Sorry David but this is incorrect. CO2 rising from 1860 until now is from human consumption of fossil fuels. 1 gallon of gasoline produces about 20 pounds of CO2. Humans are the reason why CO2 keeps creeping up (27 billion tons a year of CO2 released into the atmosphere).

David says:
December 26, 2010 at 7:27 pm
“Humans are the reason why CO2 keeps creeping up (27 billion tons a year of CO2 released into the atmosphere).”
No. They aren’t. The increase in CO2 concentration bears only a superficial resemblance to the human production of CO2. The two series correlate poorly in the low frequency regime, and not to any level of significance at all in the higher frequency realm. The natural sequestering of carbon provides a negative feedback which, absent a much larger forcing than anthropogenic CO2, regulates atmospheric CO2 in a narrow band. Indeed, atmospheric concentrations have been decelerating for the last decade, even as human production has increased. Humankind has been convicted in this regard on a post hoc ergo propter hoc basis. In time, it will be exonerated.

G.L. Alston:

Meanwhile nobody but nobody seems to be able to explain the MWP warmth nor the LIA coldness. CO2 was constant, according to the ice cores, yet the temp varied. Well, then, how can that be? I thought temp went up due to CO2.

“If A then B” does not logically-imply “if B then A”. Temperature changes due to changes in the radiative balance of the earth, which we call “radiative forcings”. Right now, the primary radiative forcings are the rapidly rising CO2 levels (along with other greenhouse gases and aerosol levels). However, in the past, there have been other important radiative forcings. (For the LIA and MWP, it is generally believed to be some combination of solar variations and variations in volcanic forcings…due to variations in the frequency of significant eruptions, although the fact that the global temperature changes were fairly small and the forcings are uncertain to sufficient accuracy during these times, it is hard to pin it down very precisely.) Over timescales of many thousand of years (i.e., glacial -interglacial cycles), variations in the earth’s orbit become significant; these variations do not produce a very significant forcing on a global level but they do produce significant variations in the seasonal and latitudinal distribution of solar insolation which then produce significant radiative feedbacks due the growth and decay of ice sheets and the variation in greenhouse gases in the atmosphere.

Plant stomata data appears to me to be nearly as suspect as ice core data for a yard stick for past atmospheric data CO2. Far too many assumptions as to the validity of the science. Kind of like the science of CO2 caused run away global warming.
PNS is the science of faking the assumptions in the questions so you can get the result you want. Just like the ” science ” question in school that poses the Tree falling in the primal woods that has no listener. Is there sound? A philosophy question that is presented as science. But is mainly to prove that normal people are too stupid to really understand science.
No one has proved to me that 0.03% CO2 can cause any measurable warming and that such warming might be bad for man or animal. I do know for sure that increased CO2 results in greater plant growth and better efficiencies in the use of water and fertilizers. I further know that cooling is deadly to all life. pg

Joel Shore says:
December 26, 2010 at 12:13 pm
they tell you what you want to believe even though there is no good explanation of what could possibly cause such rapid variations in CO2 levels in the past.

I hope skeptics do not fall into the same trap the AGW proponents have, that of following beliefs and ignoring data.
One of the data ignored is that CO2 is a gas. Gas is transported by air. Air means winds. So, even though I do not know what caused the variations in CO2 levels in Becks compilations, I can treat CO2 levels as long term proxies of wind directions. This means that I would have to go where the measurements are done and verify the wind records.
Winds are persistent over decades and change over decades too. I live in Greece and though we have periodic winds called “meltemia” coming from the north in the summer, the past ten years they are mostly missing. We are getting westerly winds and southern ones. North is a great source of green, and CO2. South is the Sahara and water, west is the industrial west.
There were forest fires in Russia this year. We would have gotten lots of CO2 if the meltemia had worked, which they didn;t.I expect if we measured CO2 this decade chemically we would see large variations from the previous decade.
Then there are volcanoes , land and undersea. Their time constants are in decades, and certainly not constant over millenia.
These are hypothesis to be checked and viable until refuted but one cannot make the blanket statement “CO2 cannot change fast”.
Tim Folkerts says:
December 26, 2010 at 12:24 pm
Do your truly think that the concentrations changed that dramatically over such a short time, and if so, what might have caused this change?
see my arguments above.

I’d contend that they are hiding inconvenient data…re high freq resolution CO2 vs temp. Otherwise where is the GISP2 CO2 data, and why the low res Vostok data. They did pretty good with the sampling rate during the previous interglacial, but somehow the sampling rate drops for this interglacial? WUWT? I would think it would be rather important to increase the samplig rate during this interglacial…just saying…something smells.

Oh, snap! These pesky facts spoil the whole, wonderful theory of AGW!
Can we hide the decline, please? Thus, I would feel a lot warmer while removing the freshly fallen, foot-high blanket of wonderful, white Global Warming from my doorstep.
I want my Anthropogenic Global Warming back!

People should have a look at the GOSAT data too, discussehere before .
and also their more recent publication here.
There are over 30ppm variations in column averaged CO2.
What surprises me in these plots is that though we see the variation with seasons in the northern and southern hemisphere, there is continuous large input from Africa. Even the Sahara. This must mean that hot soil also exudes CO2, not just the oceans, something I have seen only in passing and not seriously considered. Thus it is not only the oceans that theoretically contribute excess CO2 as they are heated ( in these plots that effect is not measurable, oceans are mostly white). Also the colors show that the land location is more important in the level of CO2 exuded than the industrial concentrations.
It upsets the CO2 apple cart, in my opinion.

If we for a second step back from details and look at it logically Temp vs. CO2, just what this high correlation between the two can mean.
CO2 and Temperature seem to be highly correlated through millenia. Even more than that, they nearly match movement to eachother. There were great variations in Earth’s temperature through millenia and CO2 ALWAYS tracked it closely. That much we know for a fact afaik.
If this is not coincidence, then one of two statements is true.
1. Amount of CO2 is the only determining factor of Earth’s temperature. Sun’s activity, volcanos, cloud cover, sea currents, … have or may not have an effect, they influenced temperature at some point or other, but based on nearly perfect Temp to CO2 correlation through millenia, I conclude CO2 is the ONLY factor of all the variables that actually matter for earth’s temperature.
2. Amount of CO2 is the result of Earth’s temperature. There are countless variables that define Earth’s thermodynamics and thus temperature such as sun’s activity, volcanos, cloud cover, sea currents,… but the amount of CO2 in the air reacts to the variable of temperature and just follows it closely.
Hope i was clear in what i wanted to say. Personally i find it a bit funny if CO2 was the ONLY determining factor in what variations in the temperature of Earth will be. And i find it highly probable that CO2 simply mirrors Earth’s temperature. But that’s just me.
Regards,
Sasha

Excellent post!
But what is taking place with the images? I would like to unleash a Czech discussion on it. The graphs disappear irregularly in inscrutable pattern. They are like the very climate.

p.s. Does this mean corals will disappear as they will not be able to sequester enough carbonate to build their skeletons?
Also, this at least suggests what we can do with all that CO2 which is to be removed from coal-fired emissions – just pump it into the deep oceans to replace what is outgassing at the surface!

None of the images are working for me-
“photobucket bandwidth exceeded”
any ideas?

Joel Shore — “If A then B” does not logically-imply “if B then A”.
No kidding. 🙂
Let’s look at a few things:
1. We’re constantly told that the rise from 1800 onward (that’s 200+ years) is anthropogenic. CO2 is up. Temps must therefore rise. It’s so well known that this results in the skeptics reminding all that correlation isn’t causation.
2. Advocates tell us there’s a relationship between temp and CO2. To explain ice core lags (or handwave or whatever) the idea is that if temp goes up CO2 will go up, PERIOD. There is no uptick in temp without an uptick in CO2, and it doesn’t matter which one drive which — there’s a correlation. When one goes up, the other one does.
3. OK, so it’s been pounded in my head by everyone from Gavin S to Andy Lacis to other hangers on that there is a correlation between temp and CO2. No matter what they will follow each other, that they MUST follow, else radiative physics simply doesn’t work.
4. OK, let’s accept this as being true: radiative physics works as explained.
5. The MWP was a warming period. It lasted a couple of centuries. The ice cores show no corresponding CO2 increase (despite the radiative physics people claiming that this MUST happen.)
6. Joel Shore posits a 200 year long volcano or planetary wobble event, claiming “other” forcings.
(Dude. Please.)
7. Conclusion:
a) if temp and Co2 are always intertwined as per radiative physics experts like Lacis, then the ice cores MUST be lying.
b) if the ice cores tell the truth then Lacis is completely wrong.
c) We simply don’t know enough to tell.
Choose.

******
Leif Svalgaard says:
December 26, 2010 at 6:37 pm
With the CO2 in ice cores so uncertain what is then with the oft repeated statement that CO2 lags temperatures by 800 years? It would seem that that now is not on firm ground.
******
Yeah, that’s a pretty important question. Maybe Ferdinand has some idea — he seems to know some of the estimated inclusion time-lags at the various ice-core sites.

G.L. Alston says:
December 27, 2010 at 5:37 am

Let’s look at a few things:
1. We’re constantly told that the rise from 1800 onward (that’s 200+ years) is anthropogenic. CO2 is up. Temps must therefore rise. It’s so well known that this results in the skeptics reminding all that correlation isn’t causation.

Temps must therefore rise only if everything else remains constant. It isn’t correlation. It’s fundamental physical law known for 200 years and experimentally demonstrated by John Tyndal 150 years ago. The problem is everything else doesn’t remain constant. Tyndal had to be very careful to dry the gases in the thousands of experiments he ran because water vapor was such a good absorber of thermal radiation he couldn’t otherwise detect the thermal absorption properties of any other gases when water vapor at normal atmospheric concentration was present. On a planet where 70% of the surface presents liquid water and surface pressure is 14psi the water cycle caps how warm the surface can become when incoming radiation at the top of the atmosphere remains nearly constant. Non-condensing greenhouse gases put a bottom on how cold the surface can become when ice becomes the dominant surface presentation. Today we are at a tipping point but that tipping point is towards a surface presentation dominated by ice. CO2 partial pressure is far too low to keep the ice away. It’s insane to want it even lower. The periods of great biological fecundity and many tens of millions of years at a stretch without an ice age are marked by CO2 partial pressures of 2000 ppm or more.
The CAGW boffins are the ones in denial. They are in denial of 500 million years of history contained in the geologic column. They are in denial of the water cycle’s action in capping how warm the surface can become. They are in denial of the biological fecundity that happens when CO2 partial pressure is vastly higher than today.

“Wow. Least squares curve fits are robust. What a revolutionary discovery!”
But, for all that, the fit is still poor.

“Do a cross correlation.”
By that, I mean a cross spectral analysis. I apologize for slipping into colloquial jargon which may tend to confuse.

…”there is only one tiny one with a roughly 20 year period which appears that it could correlate.”
If memory serves. I only found one harmonic which was close enough that it could be common to the two, and I think it was in the low 20’s. The fact that none of the other harmonics shows up dictates that the input CO2 response is severely attenuated below the noise floor. Which means that the system exhibits either an incredibly efficient low pass response, or more likely, that it is simply insensitive to anthropogenic CO2 across the board.

Larry D says:

IIRC Dr. Henson’s “greenhouse runaway” tripping point is at CO2 = 450ppm.
The GEOCARB data shows this value exceeded by most of the time in the last 600 mya, by a large margin. Two Ice ages occurred with CO2 > 2000 ppm. This is why I’ve been a CO2 doom skeptic for a long time.

You’ve invented a strawman. To my knowledge, Hansen has never claimed that there is a runaway above 450ppm. What he has claimed is that it will bring us to a world that is quite different than the one we are inhabiting now in terms of temperature, sea levels, and so forth. And, in fact, over the geologic time, there have been quite large changes in climate and sea levels.
As for the claim of two ice ages with CO2 levels above 2000ppm: The temporal resolution and precision of the data (or simulation results…I am amused at how amazingly reliable models seem to suddenly become when one likes the results that they give) for both temperature and CO2 become worse as you go back further in time. So, one cannot make such blanket statements. Furthermore, as one goes back over geological time, other factors have to be considered, including the locations of continents and mountain ranges, etc., etc. Changes in CO2 are not the only forcing that determines the earth’s climate. It is simply the forcing that is changing most rapidly at the moment.

“The carbon cycle lags behind the climate cycle and thus does not drive the climate cycle.”
“The anthropogenic contribution to the carbon cycle since 1860 is minimal and inconsequential.”
Two more nails in that coffin!
Thanks for the brilliant article, and a Happy New Year!

Joel Shore says:
December 27, 2010 at 11:07 am
“That is not what is relevant. What is relevant is that if you look at the level of CO2 over time and compare it to the cumulative emissions (scaled by ~50% to account for the CO2 that partitions into the biosphere and ocean mixed layer) …”
No, Joel, a thousand times, no. If part of the input to a system is unobservable, then it either had to have been filtered out by some mechanism, or the entire hypothesized input-output mechanism is in error. You can claim this isn’t true all you like, but experienced systems analysts will only laugh at you.
“…then the agreement between the two graphs is very good…”
Not to my eyes. When the residual is a significant fraction of the signal itself, that shows that there are unmodeled processes which could easily account for the entire signal, but the fit still “fits” to some extent because that is what least squares fits do, i.e., as I said before, they are robust.
“…especially once you get beyond the earlier years when land use changes were likely more important than fossil fuel emissions.”
Ah, I see you found it necessary to do some hand-waving yourself, so perhaps my aged eyes are not so bad after all.
“No…The point is not that there are not relaxation processes but that these relaxation processes…”
It does not matter. You are merely approaching the limit of a relaxation-less system, and you will begin to see the same type of behavior on timescales less than the relaxation time. This is why the whole thing just doesn’t gel. You and the rest have jury-rigged a system to behave as you want, but you have not taken account of secondary characteristics which would be exhibited by such a system.
“He has links to his data sources. Check it yourself and then demonstrate to us if you find something different.”
An utter waste of time. I know the man’s works, and their demonstrable lack of rigor, willful omission, and fudging. I have performed the analysis myself. I know. If you perform the analysis yourself, then you will know, too. Otherwise, I suffer no delusion that I can teach you what you do not wish to learn. The only way is for you to demonstrate it to yourself.

Changes in CO2 are not the only forcing that determines the earth’s climate. It is simply the forcing that is changing most rapidly at the moment.
Exactamundo.
“Indeed, atmospheric concentrations have been decelerating for the last decade, even as human production has increased. Humankind has been convicted in this regard on a post hoc ergo propter hoc basis. In time, it will be exonerated.”
The growth rate doesn’t appear to be decelerating at the moment.
1991 0.96
1992 0.46
1993 1.37
1994 1.92
1995 1.94
1996 1.22
1997 1.93
1998 2.97
1999 0.91
2000 1.75
2001 1.57
2002 2.59
2003 2.30
2004 1.57
2005 2.53
2006 1.73
2007 2.19
2008 1.66
2009 1.86

cohenite says:

in fact this is what Hansen has said:
“…Given the solar constant that we have today, how large a forcing must be maintained to cause runaway global warming? Our model blows up before the oceans boil, but it suggests that perhaps runaway conditions could occur with added forcing as small as 10-20 W/m2.”
http://climatechangepsychology.blogspot.com/2008/12/james-hansens-agu-presentation-venus.html
Any contribution you have in the real climate debate Joel [which is not AGW but the extent to which humans should intefere with natural process], is mitigated by these obvious little white lies you resort to.

tony b says:

Hansen did claim there was a tipping point at 450ppm. He also claimed sea levels would rise by 20metres.
http://webcache.googleusercontent.com/search?q=cache:5yxOVZMzX40J:www.giss.nasa.gov/research/news/20070530/+hansen+450+ppm+runaway&cd=7&hl=en&ct=clnk&gl=uk

My recommendation to both of you is that you learn to read a little more carefully, both what I wrote and what Hansen has said. The statement that I was responding to was the claim that Hansen said there would be a runaway if CO2 levels went above 450ppm. I pointed out is that what Hansen said is that it would bring us into a very different world (and I think he has in fact used the word “tipping point”)…but not a true runaway…and thus it was irrelevant to point out that we had had CO2 levels this high before and the world survived. My point was, yes, the world has survived, but sea levels and temperatures have been quite a bit different from what they were today…and thus this past evidence did not contradict Hansen’s claim of a very different world if CO2 levels go this high.
cohenite: A forcing of 10-20 W/m^2 that Hansen feels may be large enough to trigger a true runaway would correspond to more than a quadrupling of CO2 levels…i.e., something on the order of 1500 (for 10 W/m^2) and much higher for 20 W/m^2. Hansen is worried that we could reach such forcings, especially if we really go to town burning our conventional and non-conventional fossil fuels. I have also stated in the past here on WUWT that I remain skeptical of this claim of Hansen’s, given that (to my knowledge) it has not yet appeared in a peer-reviewed publication and given that other scientists have stated that they don’t think a runaway is possible with current solar irradiance. That is not to say that I know Hansen is wrong…but just that I think he needs to explain in more detail how he thinks this can occur and what those who think it can’t occur are not considering. (I think that he has vaguely said something about carbon cycle feedbacks and the inability of biogeochemical feedbacks that in the past would have prevented such runaways from occurring when greenhouse gas levels were so high to operate on the timescales fast enough to prevent it this time…and also something about the fact that the sun was fainter back in the times hundreds of millions to billions of years ago when the CO2 levels were believed to be really high.)

Middleton (article)
That’s why oil and gas are almost always a lot older than the rock formations in which they are trapped. I do realize that the contemporaneous atmosphere will permeate down into the ice… But it seems to me that at depth, there would be a mixture of air permeating downward, in situ air, and older air that had migrated upward before the ice fully “lithified”.
Caleb’s arguments add the issue of kinetic motion and atmospheric pressure changes.
To this I want to ask if during the process of closing off the pores there is not a significant molecular filtering process in place. In the 200 years it takes to pack snowflakes into an ice cube with “air” bubbles, the molecules are not equally free to migrate through the pores.
At first glance, it might seem that the trapped air may be richer in CO2 than in the atmosphere at deposition because the CO2 molecule is larger and heavier than N2 and O2. But with a variable pressure and diurnal temperature, as that constricting bubble breathes, once a CO2 escapes, it is much less likely to reenter than an O2 or N2.
The other point I did not see mentioned here is the relative solubility of the air gases in Ice and water. When dealing with ppm’s, is the solubility of CO2 on the surface of an ice crystal inconsequential?

As one of the “stomata: people and author ofd the cited Tellus paper, I want to draw attention to one of the most interesting outcomes of our research. That is that for the past thousand years the stomata records seem to match with respect to timing to two Antarctic ice core records which are not often cited…. Matching variabilities between ice cores of such resolution has not been achieved yet… well, ice core people claim that they reproduce their flat liners, but if you zoom into detail the small fluxes never match wit respect to timing… The lone fact that stomata data of the USA and Europe have the same timing of a CO2 wiggle which has also been recorded (but with a much lower amplitude) in two Antarctic ice cores is evidence enough that Co2 variability has been larger in the past millennium then assumed. If the variability would have been as small as the ice cores tell us, plant would hav e never ever picked this signal up on two different continents on another hemisphere…

David says:
December 28, 2010 at 7:07 am
“Bart—yes it looks negative on the quadratic (for the recent 2-4 years), but linear is still up…”
As Doc Brown would admonish, you’re not thinking four dimensionally. Which term will dominate over time?
More importantly, yearly emissions in the last two decades, at a time when India and China, with a combined population of more than 1/3 of the planet, have been rapidly industrializing, show no such slackening.

Bert, we just had a major global recession, and now we are in La Nina with cooler water temps (less co2), yet this last year 2010 shows a strong rise in CO2.
http://www.esrl.noaa.gov/gmd/webdata/ccgg/trends/co2_trend_gl.png

tonyb says:

(Ps Happy New Year)

You too! (meant to include that in my last post)
Bart says:

“Bart—yes it looks negative on the quadratic (for the recent 2-4 years), but linear is still up…”
As Doc Brown would admonish, you’re not thinking four dimensionally. Which term will dominate over time?

If you are really think that you can fit a quadratic over only 2 or 4 years of the data and draw a serious conclusion then you deserve whatever answer you get!

@ David Middleton… well actually for the somewhat older stomata data ( I focussed on the past 1000 yrs but my colleages on the whole Holocene) there are Greenland iced-core records which match pretty well… However, we can’t use them for publictions as the ice community officially redrew them as soon as the Antarctic records became available.. they claimed the records are contaminated by too much dust in the ice….
Furthermore I want to mention that we fully understand there are uncertainties with the stomata data. what bothers me is that for our records the scientific community focusses on these uncertainties in exact prediction while all the flaws and errors in ice data are ignored… furthermore it is quite amusing for me as a biologist to read the papers where physicists try to attack the proxies by playing plant physiologist…. I am very surprised the scientific community does not have a very warm welcome for new innovative techniques when those techniques put question marks at established ideas.., I always learned that these discussions are the fundamental backbone for science… therefore my hope that climate science will ever become a fullgrown scientific discipline is lost as long as politics (read funding) keeps intermingling..

Sorry for the late reply, I am currently travelling in France, where I have no access to my files…
1DandyTroll says:
December 26, 2010 at 3:14 pm
@ferdinand meeus Engelbeen
‘CO2 levels are measured in “background” conditions at some 70+ places over the world + satellites.’
We only have 30 years of readings from the latest 30 years and since the sat readings don’t correlate to the down to earth readings on a 1:1 basis, there’s discrepancies to account for, and explain, to boot, and until such time who can say with enough certainty what’s what in that department.
We have over 50 years of direct data and a few years of satellite data, which are comparable to the local data near ground. Plus an overlap of some 20 years between ice core data (Law Dome) and South Pole direct data.
Take into account all the “70+” places, those doesn’t begin to cover the whole planet by even the most liberal statistical concoctions
CO2 levels are quite rapid mixed: within days to weeks for the same latitude/altitude. Within weeks to months for different latitudes/altitudes and about a year between the hemispheres. Even one station would be enough to represent the trend in global CO2.
Which is a problem of definition. The higher the definition the better it was supposed to get, but it didn’t did it? We can no more today with even higher definition than yesterday account for even a decades precision in any age, let alone ten years ago from today. So how can we be certain of the readings at all, really.
The resolution of the ice cores really is about a decade for the fastest accumulating ice cores. The ice at closing depth is about 40 years old, the average gas age at the same depth is about 10 years older than at the atmosphere. The (not important) ice age – gas age difference thus is 30 years. The diffusion rate was measured top-down in the firn. At closing depth the composition of air in still open bubbles and already closed bubbles was identical…

cohenite says:
December 26, 2010 at 4:26 pm
Drake reached a similar conclusion about the ice cores underestimated historical concentrations of CO2:
Drake does “adjust” the ice core CO2 levels for ice age – gas age differences. That has not the slightest physical meaning: ice layers and gas bubbles are independent of each other, except that the number of layers and the averaging of the air composition at closing depth both depend of the snow accumulation speed. Thus the error that Drake made is from the type:
A causes B
A causes C
that makes that there is a good correlation between B and C, even if there is no physical relationship between B and C. The correlation (and thus the correction) is completely spurious.

Floor Anthoni says:
December 26, 2010 at 6:03 pm
Engelbeen must be corrected here. The oceans contain 38000 Gt carbon and the atmosphere 700GtC. Henry’s law insists that about 1100GtC outgasses per ºC at 15ºC, which equates to over 600ppmC/ºC. Not 16ppmv/ºC as Engelbeen states.
Sorry, but you are wrong on this point: No matter how much CO2 is in the (deep) oceans, Henry’s Law is a matter of differential pressure between CO2 in the atmosphere and CO2 in the oceans. If the ocean surface (as a whole) increases with 1°C, then an increase of 16 ppmv (32 GtC) in the atmosphere is enough to compensate for the temperature increase, no matter how much CO2 resides in the upper or deep oceans…
Further, I have seen several errors in your pages, good for several corrections… To mention one: the oxygen use that you have calculated is based on what rests in the atmosphere, but should have been calculated on the full use of fossil fuels and not only 1:1 (for coal) but also 1:1.5 (oil) and 1:2 (natural gas)…
Note: I don’t know if I will have more time to respond, but I am back home tomorrow (late) night…

It has a much greater averaging in that the data points (in Vostok core) are generally thousands of years apart.

I was incorrect to call that averaging. It is not , it is missing data.
This record is incredibly long and is useful for that reason but it is a false assumption to believe it is complete. When you have several kilometers of ice core you have take selective samples along it’s length. The gaps are huge.
So there is century scale smoothing due to gas diffusion in the snow and firn plus millenium scale gaps in the data.
This must be taken into account when trying to infer the magnitude of past variation.

P. Solar says:

Here is the beginning of the Vostok record:

Vostok is a slowly-accumulating core and is thus good at providing lower frequency information that goes back far in time. However, there are other cores that have faster accumulation and provided considerably higher-frequency information. For example, here is the Taylor Law Dome data ftp://ftp.ncdc.noaa.gov/pub/data/paleo/icecore/antarctica/taylor/taylor_co2-holocene.txt , where they have about 1 data point every century. There may be others that are higher frequency than this.

The most significant thing to note is that in past epochs CO2 was TEN TIMES what it is today and the world did not blow up or evaporate into space. Life on Earth did not end.

If your criterion concerning whether or not a problem is important enough to take action is whether or not it will cause the world to blow up or evaporate into space, then there would be very few problems that we would need to concern ourselves with. For example, do you think terrorists flying planes into a few buildings is at all a danger to life on earth, let alone comparable to some events from geologic times such as major asteroid impacts? Even such terrorists getting their hands on a nuclear weapon is not going to come anywhere close to threatening life on earth. Despite these facts, I would certainly be surprised if you were to express the opinion that nothing should be done to try to prevent this from happening!
Yes, CO2 has been much higher in past epochs (although there are quite large error bars on the estimates of exactly how much higher CO2 levels were), but it is also true that climate and sea level were very different. Such changes may be something that we don’t want to impose on ourselves and the current flora and fauna adapted to the current climate, sea level, etc.
Finally, as I noted in posts above, the few scientists like Hansen who are really concerned about the possibility of a true Venus-like runaway if we really go to town burning our fossil fuel reserves have arguments why a runaway could happen this time even though it never did before. I am not claiming that these arguments are correct, but I don’t think you can just say, “It hasn’t happened before” and be done with it. Hansen understands that, but he also notes ways in which the current situation may not really have a previous precedent.

P. Solar says:
December 30, 2010 at 12:49 am
Interesting accumulation of data but this is meaningless without showing the associated uncertainties.
Ice cores tell us nothing about the changes on a timescale comparable to the industrial period.
The available data is not able to prove or disprove that this is in anyway different or “unprecedented”. The whole discussion is totally without foundation.
Please, before writing such things, have a look at the existing data! The data of many ice cores are available at:
http://www.ncdc.noaa.gov/paleo/icecore/current.html
There are data from ice cores with very high accumulation (1.2 meter ice equivalent per year) like Law Dome and very low accumulation (a few mm per year) like Vostok and Dome C. The highest accumulation ice cores have a resolution of about a decade and span some 150 years. Some others have a 40 years resolution and go back 1,000 years and Dome C has a resolution of 560 years but goes back some 800,000 years.
For the past 1,000 years, that gives the following combination of datapoints from different ice cores with available points:
http://www.ferdinand-engelbeen.be/klimaat/klim_img/antarctic_cores_001kyr.jpg
While the accumulation rate and temperature and salt/dust inclusions are quite different for the different cores, the CO2 levels found are quite identical for the same gas age (+/- 5 ppmv). The highest accumulation cores of Law Dome (with an accuracy of +/- 1.2 ppmv – 1 sigma) even have an overlap of some 20 years with direct measurements of the South Pole:
http://www.ferdinand-engelbeen.be/klimaat/klim_img/law_dome_sp_co2.jpg
Thus while Mann’s HS was certainly fabricated, the CO2 HS is quite real, at least for the past 10,000 years and probably 800,000 years: one need quite rapid and large changes of temperature to induce rapid and large changes of CO2… The CO2 : temperature ratio in the long run was about 8 ppmv/°C. To obtain a 100 ppmv increase in 150 years by natural causes, one need an increase of 12°C in the same time span. Not completely unprecedented (the end of the Younger Dryas), but quite uncommon and not global. And not lasting for long: even the 560 years resolution of Dome C would notice a 100 ppmv increase lasting over a 20 years time span.

tallbloke says:
December 31, 2010 at 4:05 am
Hi Ferdinand,
Do you have a plot of seasonal variation at Giessen?
Should be visible in the monthly averages, but is hardly detectable in the huge noise over more than a decade of data:
http://www.ferdinand-engelbeen.be/klimaat/klim_img/giessen_mlo_monthly.jpg
I have no idea what caused the few extreme peaks in the Giessen data, but in general CO2 levels taken over land are very noisy and influenced by crops, industry, traffic, heating,… and what/how much of these sinks and sources are present in the main wind direction of that particular month. Or growing season for stomata data: according to Dr. Tom Van Hoof, stomata density of new leaves is based on CO2 levels in the previous growing season.
Here another plot of land based monthly CO2 averages near Kennedy Space Centre (Florida, USA), compared to several “baseline” stations over a few years:
http://www.ferdinand-engelbeen.be/klimaat/klim_img/month_2002_2006_ken.jpg
Again hardly any seasonal variation detectable in the huge noise…
The Kennedy Space Center Scrub Oak site is located within the Merritt Island National Wildlife Refuge at the Kennedy Space Center (KSC) on the east coast of central Florida, within a 10 ha park. See:
http://public.ornl.gov/ameriflux/Site_Info/siteInfo.cfm?KEYID=us.ksc_scruboak.01
That is part of the Ameriflux network, aimed to obtain more details about the back and forth fluxes of CO2 over vegetation (and human sources) over the seasons. Not an easy task… See more details at:
http://public.ornl.gov/ameriflux/dataproducts.shtml
Pieter Tans of NOAA has a few plots of the Wisconsin tall tower of summer/winter CO2 taken at different heights at:
http://www.esrl.noaa.gov/gmd/ccgg/about/co2_measurements.html

beng says:
December 31, 2010 at 5:29 am
Ferdinand Engelbeen, I understand it’s hard to wade thru all these comments, so I’ll ask again.
How does the time-lag of CO2 diffusion affect the apparent time-lags of CO2 vs temp? It’s been assumed to be ~800 yrs at the interglacial start and ~1200 yrs at the beginning of the glacial period. Or is the data not detailed enough to determine this?
I looked at your site & couldn’t find anything about this directly.
Indeed, I am just half way the comments now…
The main problem of the diffusion for low accumulation cores is to determine the average age of the air in the gas bubbles and the averaging period that the air represents. The main way is by firn densification models, which were validated by direct measurements of the CO2 levels in open pores of the firn top-down.
An additional problem is that during the cold glacial periods, the accumulation (and thus the firm densification) is even less than during an interglacial.
For some periods, there is some help from known volcanic eruptions (dust, sulfate deposits) and comparisons with ocean sediments and overlapping periods in other ice cores (Antarctic and Greenland). And changes in isotopic composition of the different gases are used to determine the timing of land changes from ice covered to vegetation covered and back.
Nevertheless, there are several corrections issued for different ice cores, where better methods of analyses show (small to important) changes in gas age timing. That is less of a problem for the determination of the age if the ice layers, where the d18O and dD ratio’s gives a quite good information of nearby (for coastal ice cores) to hemispheric (for the inland ice cores) ocean temperatures.
Thus the timing of the changes in the gas composition vs. the temperature (proxies) remains problematic, the more that the resolution of the Vostok ice core is around 600 years. One can only be sure that there is a lag, but less sure if that is small or huge.
For the last end of the glacial period, there are more data from a higher sampling density at Dome C: the resolution still is around 560 years, but samples were taken on small parts of the ice, allowing for more datapoints (about 6 per millenium). That gives a somewhat better insight in the variability of the lag, but in absolute value to be taken with a grain of salt:
http://www.ferdinand-engelbeen.be/klimaat/klim_img/epica5.gif
The graph was made by André van den Berg, based on the Dome C data.
Some more literature:
http://www.sciencemag.org/content/283/5408/1712.abstract
http://www.ncdc.noaa.gov/paleo/taylor/indermuehle00grl.pdf
The latter looked at the D.-O. events and found a lag of 1200 +/- 700 years by a Monte Carlo analyses for a best fit of the temperature-CO2 correlation.

David Middleton says:
December 27, 2010 at 1:01 pm
Knorr, 2009 showed that about 60% of the annual anthropogenic emissions are taken up by sinks. The sinks can’t tell the difference between this year’s and last year’s CO2 emissions. It’s a roughly 60% decay rate. This fits the well-established residence time of 5 to 15 years (Essenhigh, 2009, Segalstad, 1998, Segalstad, 1982, Houghton et al., 1990, Stumm & Morgan, 1970, Revelle & Suess, 1957, etc.).
David, like so many before you, you are confusing residence time (which is how much CO2 is exchanged between the different reservoirs each year) with excess CO2 decay time. The former is about 150 GtC/800 GtC present in the atmosphere. That doesn’t change the total amount of CO2 in the atmosphere with one gram. The excess decay rate is what matters: 60% of the emissions (including land use changes) is only 5 GtC/year real reduction of the total amount of CO2 emitted (some 9 GtC/year), thus the real excess decay rate is about 40 years half life time if we should stop all emissions today. Not 5-15 years. And as the difference still is 4 GtC/year, that is what shows up as extra increase in the atmosphere, whatever the origin of the molecules of CO2 which were captured or released by the CO2 exchanges with other reservoirs.

Stephen Rasey says:
December 27, 2010 at 11:40 pm
At first glance, it might seem that the trapped air may be richer in CO2 than in the atmosphere at deposition because the CO2 molecule is larger and heavier than N2 and O2. But with a variable pressure and diurnal temperature, as that constricting bubble breathes, once a CO2 escapes, it is much less likely to reenter than an O2 or N2.
The other point I did not see mentioned here is the relative solubility of the air gases in Ice and water. When dealing with ppm’s, is the solubility of CO2 on the surface of an ice crystal inconsequential?
Indeed there is some fractionation of the heavier molecules and isotopes in the lower firn layers. This is compensated for by taking into account the fractionation of 15N/14N with depth. See:
http://www.pnas.org/content/94/16/8343.full
Further, the smallest molecules (Ne, Ar, O2) show a deficit in the resultant air bubbles, because these may still escape during the closing process below a certain porosity. The vibrational diameter of CO2 is somewhat above this porosity diameter, thus doesn’t show such a deficit. See:
http://icebubbles.ucsd.edu/Publications/closeoff_EPSL.pdf
Water is less of a problem, as the waterlike layer at the ice – air surface is about 5 atoms thick and even less waterlike at the intercrystalline borders. Migration of CO2 is extremely low, even for the higher temperature (-20°C) ice cores. And any CO2 dissolved in some water is effectively removed as the measurements are made under vacuum, removing any water over a cold trap and destructing any remaining clathrates. Comparison of this method with another method, where all ice is sublimated and cryogenically separated (mostly done for more accurate mass spectrometer analyses of isotopes) shows the same results…

Joel Shore says:
December 26, 2010 at 7:33 pm
“Exactly! Why is it not obvious to all scientists that CO2 levels before 1959 danced around wildly before 1960, but in a very careful choreographed manner so that the level remained within about 10ppm when effectively low-pass filtered in the ice core data?”
Aye, yi, yi. I just saw this comment. Joel might as well have just broadcast “I know nothing about filtering theory or how ARMA processes work.” Because, that is what this statement reveals. This is precisely what a low pass filter does. It’s not magic. It’s math. It’s what is done every single day in real time as you tune a radio or TV station, or make a call on your cell phone.
Ferdinand Engelbeen says:
December 31, 2010 at 1:34 pm

Water is less of a problem, as the waterlike layer at the ice – air surface is about 5 atoms thick and even less waterlike at the intercrystalline borders. Migration of CO2 is extremely low…

When, oh when, will you learn the difference between conjecture and proven facts? Maybe in the new year? This is so depressing…

Bart says:
December 31, 2010 at 8:22 pm
When, oh when, will you learn the difference between conjecture and proven facts? Maybe in the new year? This is so depressing…
Sometimes it helps if you read the existing literature…
The migration of CO2 in ice cores was theoretically calculated by looking at the migration in remelted layers of a “warm” (-24°C) ice core, Siple Dome:
http://catalogue.nla.gov.au/Record/3773250
The extra smoothing of the gas age averaging caused by migration is about 2 orders of magnitude less than the averaging itself. For the Siple Dome ice core, the average gas resolution is about 22 years. Migration increases that with about 0.2 years.
At lowest depth, the layers become thinner and migration may relatively increase and give a doubling of the resolution to some 40 years.
The Vostok and Dome C ice cores are much older at depth, but also much colder (-40°C), which means less water (no water layer at all at -32°C, as long as no salt inclusions are involved). But there is a simple proof that migration doesn’t play a role at Vostok and Dome C:
The Vostok (and Dome C) ice cores show a quite nice relationship between the temperature proxy (dD and d18O) and CO2 levels, be it with a lag. Nevertheless, the ratio between CO2 and temperature is about 8 ppmv/°C between interglacials and glacial periods over the past 420,000 years (recently conformed for the full 800,000 years in Dome C). If there was some migration, the ratio would fade over time for each 100,000 years period further back. That is not the case. Thus there is little migration of CO2 in the Vostok and Dome C ice cores…

Bart says:

Aye, yi, yi. I just saw this comment. Joel might as well have just broadcast “I know nothing about filtering theory or how ARMA processes work.” Because, that is what this statement reveals. This is precisely what a low pass filter does. It’s not magic. It’s math. It’s what is done every single day in real time as you tune a radio or TV station, or make a call on your cell phone.

A low-pass filter would only work this way if all of the frequencies in the signal (with sufficient amplitude) were much greater than the cutoff frequency of the filter. This would seem highly unlikely on the face of it but now that we have about 50 years of steadily-rising data from Mauna Loa, I think we can say with confidence that there is already a low enough frequency in that data that this statement cannot be true (at least for the faster accumulating ice cores). [I haven’t tried fitting sine curves to the Mauna Loa data but I imagine that if you did then you would be hard-pressed to get this 50 years as even representing a quarter of a cycle, which means that the period would be at least 200 years.]

Ferdinand Engelbeen says:
January 1, 2011 at 2:04 am
“That is what the mass balance tells us.”
It really doesn’t tell us that at all, Ferdinand. I do not know why you cannot grasp your fallacy here. I only know that you cannot, because we have been over and over it, and others have tried to explain it to you, too. It is painful to watch.

Bart says:
January 1, 2011 at 10:05 am
“Thus you are looking at the cause of the noise around the trend, not the trend itself.”
No. It is a consistent, decades long, coherent signal. It is not noise.
I am not sure where you are talking about. The trend of accumulation of the emissions and the trend of accumulation in the atmosphere show an incredible good correlation. While correlation is not proof of causation, I don’t think that in this case the increase in the atmosphere caused the emissions. Or that any natural cause can show such a perfect match over time with the emissions. Not only over the MLO period, but even so back to 1900, based on ice cores with a resolution of about a decade.
If you only look at the year by year emissions and increase in the atmosphere, you are looking at the first derivative of the trend, not the trend itself, which is influenced by (at least) two variables: the emissions and the variability in sink rate, which is mainly influenced by the temperature variability, but the latter is less important for changes over the past century, as the temperature only shows a small change over the total period.

Bart says:
January 1, 2011 at 10:19 am
It really doesn’t tell us that at all, Ferdinand. I do not know why you cannot grasp your fallacy here. I only know that you cannot, because we have been over and over it, and others have tried to explain it to you, too. It is painful to watch.
I still wonder why so many brilliant persons have such a problem with the basic logic of a mass balance, while most housewives know that from their own budget:
If you add 100 euro’s (dollars,…) every morning in your wallet and end the day with 50 dollar more than at the end of the previous day and you repeat that for weeks, then the increase of money in your wallet is entirely caused by the daily addition of 100 euro’s. No matter how much ins and outs you had of your wallet over the rest of the day. The housewive knows that, besides the daily addition, the rest of the day there was a net expense of 50 euro’s, so there was no net gain from “something else”.