Here is the web page for Dr. Raymond S. Bradley who is listed as:
University Distinguished Professor in the Department of Geosciences and Director of the Climate System Research Center (http://www.paleoclimate.org).
Readers may also recognize Dr. Bradley from his co-authorship with Dr. Michael Mann in the famous MBH98 paper which produced the embattled “hockey stick” graph.
Dr. Bradley has also gained some recent notoriety with his accusations of plagiarism regarding the Wegman report to congress, by Dr. Edward Wegman of George Mason University, which was critical of MBH98’s statistical methods.
Here’s Dr. Bradley’s photo from his UMass web page:
Notice anything interesting? Here are some hints:
His graph for CO2 data titled “Greenhouse Gas Record from the Vostok Ice Core” shows a value around 360 ppm for CO2 at the “zero date” of the present history. The photo must be old, since the current value in the atmosphere from Mauna Loa is said to be around 390ppm currently.
So, it’s an old photo, what’s the problem you say?
For readers not familiar with the CO2 data from the Vostok Ice Core, you can find the official data set here from NOAA’s FTP servers:
CDIAC (Carbon Dioxide Information and Analysis Center)
ftp://cdiac.ornl.gov/pub/trends/co2/vostok.icecore.co2
NCDC (National Climatic Data Center)
ftp://ftp.ncdc.noaa.gov/pub/data/paleo/icecore/antarctica/vostok/co2nat.txt
NASA Goddard also offers access to the official Vostok data here:
http://gcmd.nasa.gov/records/GCMD_CDIAC_CO2_VOSTOK_ICECORE.html
…and they offer this helpful graph, which is time reversed from Dr. Bradley’s graph, with the present day on the left:
That’s odd, the Vostok CO2 data for the present is around 280ppm, way lower than the 360ppm shown on Dr. Bradley’s graph. Strange, but that NASA web page on Vostok Ice Core data shows the most recent update at:
So it must be current, right?
So let’s look at some other sources, maybe they are closer to Dr. Bradley’s value, surely there must be some update somewhere to this Vostok data that I’ve missed.
Let’s check Wikipedia, which always seems to be updated. Even though William Connelly doesn’t edit there anymore surely it’s been updated with this new data in the past year or so? Here’s the Wikipedia graph:
Source: http://en.wikipedia.org/wiki/File:Vostok_Petit_data.svg
That’s odd, the CO2 data there shows just over 280ppm of CO2 in the Vostok record. But they reference Petit, et al 1999 on that page. Hmmm, I went to find that paper, and was able to locate a PDF copy of it here: http://www.daycreek.com/dc/images/1999.pdf and I saved a local copy here Vostok_nature_1999 to prevent overloading that website with downloads. Here’s the title of that 1999 paper from Nature:
Climate and atmospheric history of the past 420,000 years from the Vostok ice core, Antarctica
J. R. Petit*, J. Jouzel†, D. Raynaud*, N. I. Barkov‡, J.-M. Barnola*, I. Basile*,M. Bender§, J. Chappellaz*,M. Davisk, G. Delaygue†, M. Delmotte*, V. M. Kotlyakov¶, M. Legrand*, V. Y. Lipenkov‡, C. Lorius*, L. Pe´ pin*, C. Ritz*, E. Saltzmank & M. Stievenard†
Oh, OK, that explains it, the CO2 levels in 1999 must have been 360ppm and that’s where that value on Dr. Bradley’s graph comes from. Let’s check the Mauna Loa record for 1999 here: ftp://ftp.cmdl.noaa.gov/ccg/co2/trends/co2_mm_mlo.txt
The values for 1999 are:
1999 3 1999.208 369.46 369.46 367.90 26 1999 4 1999.292 370.77 370.77 368.19 30 1999 5 1999.375 370.66 370.66 367.84 29 1999 6 1999.458 370.10 370.10 367.87 30 1999 7 1999.542 369.10 369.10 368.42 30 1999 8 1999.625 366.70 366.70 368.21 30 1999 9 1999.708 364.61 364.61 367.95 29 1999 10 1999.792 365.17 365.17 368.41 31 1999 11 1999.875 366.51 366.51 368.58 29 1999 12 1999.958 367.85 367.85 368.58 29
Well that explains it then right? The value of the CO2 atmosphere in 1999 was around 360 ppm, so that’s what Dr. Bradley was showing in that old photo. And the 1999 Nature paper from Petit et al must show the same value, right? Here it is:
Huh, that’s strange, it only shows around 280ppm of CO2 at the “present” of 1999 when this graph was published.
Well OK, the archived NOAA data on the FTP server must be updated and have ~360ppm somewhere in the dataset, right? So I looked through it to be sure. Here’s the most recent data from: ftp://cdiac.ornl.gov/pub/trends/co2/vostok.icecore.co2
Hmmm, the most recent data is from 2342 yr BP (years before present) and shows 284.7. That can’t be right, because the distinguished Dr. Bradley shows the data at around 360ppm. Yet, the header shows the co-author names from the 1999 Nature paper on the Vostok ice core data analysis. Surely there must be an update to it?
Maybe the other NOAA data set from NCDC is what he used? at: ftp://ftp.ncdc.noaa.gov/pub/data/paleo/icecore/antarctica/vostok/co2nat.txt
Well, it agrees with the CDIAC data, but there’s still no ~360ppm of CO2 listed in the data for the most recent readings.
Well gosh, how can this be?
The answer is seems, is that there is no new data from the Vostok Ice core. It ended, and the official repositories of that data have no new data. The last CO2 value for the Vostok Ice Core dataset is listed as being 284.7ppm.
So how does Dr. Bradley get ~360ppm? Easy, I think he uses the same technique he and his co-authors learned when writing the famous MBH98 paper that made the hockey stick -splice the instrumental record onto the paleo record:
Graph above from Fred Pearce’s Feb 2010 article in the Guardian shows the instrumental record attached to the ice core record.
And here’s a later version from 2003 showing the same instrumental record splice along with paleo data (Figure 1. from Mann et al. EOS Forum 2003):
Source: http://www.ncdc.noaa.gov/paleo/pubs/mann2003b/mann2003b.html
So it seems rather apparent that Dr. Bradley (or whoever made the graph) simply took the Vostok Ice Core CO2 paleo data and “spliced” it with the instrumental record on the end. Or, as Joe Romm likes to say “make stuff up”.
The only problem is, as he presents it with the title of his graph: Greenhouse Gas Record from the Vostok Ice Core as shown below…
…it’s patently false in my opinion. Ditto for the red Methane line, but that’s another story.
Now here’s the problem. If you took surface temperature data from Antarctica, and spliced it with surface temperature data from Hawaii, and then presented it as the entire historical record from Antarctica, our friends would have a veritable “cow”.
Or, if you took stock performance data from poorly performing Company “A” and spliced on better performing stock data from Company “B”, and then made a new graph and used that graph to sell investors on Company “A”, the Securities and Exchange Commission (SEC) would have a veritable “cow” when they found out, wouldn’t they? People go to jail for such things.
But hey, this is Climate Science.
big h/t to WUWT reader Brian M. who sent the tip in via email.
Addendum: I should add that I have no evidence that this graph has been used in any scientific publications or professional presentations by Dr. Bradley, I’m only pointing out that for this photo, which appears to be staged, what is presented doesn’t match the actual Vostok data. Readers should not extrapolate anything beyond this scope until new examples are presented. – Anthony
Phil.
As I said, the day that Bradley begins to cooperate with Steve McIntyre is the day I will apologize to him. And to you, for that matter.
In the mean time, Bradley continues to ignore the scientific method, and you defend his deliberate anti-science. Where would we be if Albert Einstein had never published in Annalen der Physik without addressing the concerns of Wilhelm Wien, Max Planck and many others?
Bradley [along with Mann, Hughes and their clique with both front feet in the grant trough] refuses to share their methodologies and data with interested parties like McIntyre and McKittrick, making a mockery of testability – the basis for the scientific method. How can a conjecture like AGW be tested if the basis for it is kept as confidential as nuclear defense secrets? This is the weather we’re talking about.
So quit your hand-wringing over my holding Bradley’s feet to the fire. Either he plays the game fairly, or he is a scientific charlatan. If he – or you – want an apology, it is a simple matter for Bradley to contact Steve McIntyre and offer to cooperate with his requests for data, methods and metadata.
jimmi,
You misunderstood. I was not referring to you as a member of Mann’s clique – which is a specific term used in statistics. The Wegman Report to Congress explains the term. To understand the issues you need to read the report.
NO ONE responded to this intresting post!!!!
“timheyes says:
November 25, 2010 at 4:47 am
I think NASA have spliced proxy and direct measurements together on the CO2 graph on this page.
http://climate.nasa.gov/keyIndicators/index.cfm#SeaLevel
They claim it’s a proxy graph but curiously it matches the instrumental value at present day. Striking!”
Not proxy, but ice core vs direct atmosphere. Note how the NASA graph is described in their own words…” The chart on the left shows the CO2 levels in the Earth’s atmosphere during the last three glacial cycles, as reconstructed from ice cores. The chart on the right shows CO2 levels in recent years, corrected for average seasonal cycles.”
The chart on the left does show both methods combined, but does not state it. The biggest reason I find to question the ice core is, as has been pointed out, the very low levels are not backed by periodic plant die off such levels would cause. At any rate the chart on the left combines two methods, one that has been dispuded by legitmate scientist, and one that is not, thus it is bad science amd apparently done by NASA.
Mike Jonas says:
November 26, 2010 at 5:05 pm
About Jaworowski: I received a reaction of him some time ago about a different discussion, where I showed the combined ice cores CO2 graph for the past 1,000 years. His reaction was that there was no difference in age between the ice and the enclosed air, thus that the shift was arbitrarely to match the South Pole measurements.
I did write to him that Etheridge had measured the air composition of firn top down until closing depth and in the ice core, which shows that the gas composition is about the same as at the surface, but some 10 years older, while the ice at that depth was already 40 years. That confirmed the 30 years ice age – gas age difference as the theoretical migration speed calculations showed. The same for gravitational separation, as Oeschger found (thanks Eli, was not aware of that work).
I also included the fact that migration via cracks is always from high to low levels and that any remaining clathrates decompose very fast under vacuum at measurement time. I received no answer…
About the CO2 levels:
The problem is in the seasonal trend: if the data start in Februari, when seasonal CO2 levels at Barrow are at their maximum, but end in December, when the seasonal trend still is in its increasing part. That gives a begin-endpoint bias of about -4 ppmv. Mauna Loa picks up the seasonal changes a few months later and has a smaller seasonal amplitude, which gives that its begin-endpoint bias is about -2 ppmv. That makes that your trends have different biases… See:
http://www.ferdinand-engelbeen.be/klimaat/klim_img/month_2002_2004_4s.jpg
Better is to use yearly averages, as the effect of the seasonal variations is near zero after a full cycle. But all together, the differences all over the globe (except in the first few hundred meters over land) are small, compared to the huge exchanges that take place over the seasons. And these differences even are smoothed out in the ice core record. A lag of 1-2 years in the data (or about 4 ppmv) is not a real problem…
David says:
November 27, 2010 at 5:05 am
While I agree that the NASA graphs are at least sloppy and should show the overlap (the most recent ice cores reach 335 ppmv, not 380 ppmv), I disagree here:
The biggest reason I find to question the ice core is, as has been pointed out, the very low levels are not backed by periodic plant die off such levels would cause.
The equilibrium for C3 plants (mostly trees) is about 90 ppmv. At that level there is no growth anymore and CO2 uptake and respiration are in equilibrium (that is from experiments in greenhouses). But C3 plants start to have problems below 200 ppmv. Despite that the “background” CO2 levels, as measured in ice cores, were around 180 ppmv for long periods during the ice ages, the levels over land in general (especially in the morning) were high enough to allow a few hours of photosynthesis. Here what I wrote in another discussion (minus a few errors):
Here a few days of CO2 measurements at Giessen, mid-west Germany, compared to the same days at Mauna Loa, Barrow and the South Pole:
http://www.ferdinand-engelbeen.be/klimaat/klim_img/giessen_background.jpg
CO2 levels at Giessen increase enormously at night (little wind, inversion layer) and drop rapidely below background in daylight. With background levels at 180 ppmv, that would give a similar increase at night, but less drop in daylight as the photosynthesis gets limited by the lower CO2 levels.
And it seems that the absolute minimum for (C3?) photosynthesis is at 90 ppmv:
http://www.nature.com/nature/journal/v193/n4815/abs/193587a0.html
The balance for some more growth than decay (respiration at night) will be higher.
Ferdinand Engelbeen : I’ll deal with the 2nd issue first to get it out of the way – I realised after posting that I should have used whole years. I re-worked the numbers on whole years and it made virtually no difference (after all, only 1 month in ~30 years had been missing). I can supply the numbers if you like. My doubts remain.
On Jaworoski : This is an intriguing debate. On the one hand, we have a maverick scientist of high repute calling foul. On the other hand we have the orthodoxy closing ranks and presenting convincing-sounding arguments as to why they are right and the maverick is wrong. Is it AGW or stomach bacteria or Einstein or evolution or Galileo all over again? Or are the orthodoxy right, as I am sure they most often are?
I have found and read the Oeschger item that you referred to. I am really not in a position to comment on the technical stuff, but the issue that seems to be the biggest sticking point is the age of the air in the ice. The basic arguments seem to be that CO2 concentrations in 19thC ice are similar to today’s, therefore (a) Jaworowski : 19thC CO2 concentrations were similar to today’s, or (b) orthodoxy : the air in 19thC ice is today’s air.
To my simple mind, Jaworowski’s seems more reasonable. A lot of explaining has to be done to get (b), such as how air from above forces its way down through the ice into areas of much higher pressure in order to displace the air there. Now presumably that is not the mechanism, and any physical flow is actually the other way, but I think there is still some explaining to do.
Oeschger says “Fig. 5 b speaks for itself; why should there be such a drastic increase of CO2 and of CH4 (Fig. 5 a) in the middle of the 19th century?“. I don’t know where Fig. 5 b is, but presumably he is referrring to the same data as in the Jaworowski paper fig.10b.
http://www.greenworldtrust.org.uk/Science/Images/ice-HS/Fig-10.gif
The argument for (b) looks seductive, because it gets the graphs to dovetail neatly – but it is a bit too close to hockey-stick methods for my liking. As for Oeschger’s argument, I would say that firstly the further down the ice core one goes, the less able one is to relate the trapped air to short periods of time such as a few decades, and Hadcrut shows quite strong warming in the mid-19th century, peaking in 1878 then dropping all the way down again around 1910. Give or take a few years, this dovetails in quite nicely with Jaworowski’s 10b. NB. I think it very likely that the temperature/CO2 gradients shown over long timeframes in ice cores are not representative over short timescales such as a few decades.
thought on CO and the thermohaline current
http://upload.wikimedia.org/wikipedia/commons/thumb/a/a6/Conveyor_belt.svg/500px-Conveyor_belt.svg.png
Some of the increased CO2 gets dissolved in the water (making it less alkaline).
At the poles this gets pulled into the ocean depths by the THC.It then travels for around a 1000 years to reappear on the surface elsewhere (See map). After another 1000 years it is back where it started but with a higher concentration of CO2 dissolved.
Are we in a benign period where our pollution is being hidden from us by the THC? Will future generations then have to cope with our mess?
Will we be seeing the results of the 1st industrial revolution emerging from the depths soon?
walt man says:
November 28, 2010 at 5:35 am
No.
Mike Jonas says:
November 27, 2010 at 3:22 pm
I re-worked the numbers on whole years and it made virtually no difference (after all, only 1 month in ~30 years had been missing). I can supply the numbers if you like. My doubts remain.
I used the yearly averages, which show an increasing lag between altitudes and latitudes, but please send the data you have used, my email address can be found on the family web site.
To my simple mind, Jaworowski’s seems more reasonable. A lot of explaining has to be done to get (b), such as how air from above forces its way down through the ice into areas of much higher pressure in order to displace the air there.
The explanation is quite simple: what is deposited is snow. As you know, that is very porous and air exchanges are quite easy, at the surface simply mechanical by wind and deeper by diffusion. That goes two-ways, always from the higher levels towards the lower levels (one of the other objections against what Jaworowski says).
There is increasing pressure on the snow by the subsequent deposit of other snow layers on top, which increases the density of the layers from snow on top to firn inbetween to ice at a certain depth.
That doesn’t give any pressure on the air in the still open pores, simply because all these pores still are in contact with each other until the surface, only reduces the porosity and thus reduces the diffusion speed.
There is a simple rule that theoretically (and measured practically) there is no diffusion anymore at a certain density of the firn. Below that, some of the pores are fully closed, others still are open and the number of closed bubbles increases with depth until all are closed.
Now, the main question: who is right?
Well, Etheridge did measure what happens in the deep: he measured CO2 levels in the firn top down at Law Dome. That showed that the levels at the start of the bubble closing depth (some 72 meters) were some 10 ppmv less than at the surface. As the measurements were done in 1993 and the South Pole air CO2 measurements go back to 1958, it was easy to know when the CO2 levels at the South Pole were 10 ppmv lower: that was in 1986, seven years earlier. Thus the composition (for CO2) of the air at 72 meter depth is only about 7 years older than at the surface.
At 72 meter depth there are already 40 years of snow deposit. These are easily counted, as winter and summer snow have a different density, which is visible as different layers. Thus with other words, at the same depth, the average gas age is 33 years younger than the ice age at Law Dome. See:
http://www.ferdinand-engelbeen.be/klimaat/klim_img/law_dome_overlap.jpg
Thus if Jaworoski says that there is no difference between the age of the ice layers and the enclosed air, then he is not only wrong, but denies what is measured in the field in 1993 and published in 1995/1996. So, I am pretty sure that Jaworowski is not the new Galileo in this case and that the “consensus” for once is right…
walt man says:
November 28, 2010 at 5:35 am
CO2 levels around a millennium ago were about 280 ppmv, during the LIA some 6 ppmv less for some 0.8°C less. Not really much difference to haunt us…
See:
http://www.ferdinand-engelbeen.be/klimaat/klim_img/law_dome_1000yr.jpg
walt man says:
November 28, 2010 at 5:35 am
Sorry, misinterpreted what you meant: from the current increasing levels by the emissions, a part is absorbed by the oceans and goes into the depth of the oceans. The deeper oceans already contain a tremendous amount of CO2 in the form of (bi)carbonates. Even with a doubling of the emissions, that increases the CO2 content of the deep oceans only with a few % and that is what comes back some 1,000 years later. Thus the resulting new equilibrium of CO2 in the atmosphere would be around a few % too.
Anthony,
You can use my full name. I always do on the internet. Thanks for doing the heavy lifting on this.
Ray Bradley,
“Where ice cores have been recovered from locations with very high accumulation rates a record of greenhouse gases can be obtained which extends in time up to a few decades ago, and these perfectly match the measured greenhouse gas values from remote locations around the world. Thus it is quite reasonable to plot the ice core greenhouse gas data with the instrumentally recorded data. This is well understood by students of paleoclimatology, but I can understand why it might not be so clear to those less familiar with the field.”
That’s a bad assumption. How do you know that ice cores being brought up from extreme pressure are not effected by the extraction process, or other effects. Surface ice is not very old or under very much pressure.
If I inflate a series of brand new bike tires over time to the same exact pressure the most recent ones will match the original measured pressure, but the older ones will not.
The fact that current periods match is no proof that the proxy doesn’t skew over time, and especially given that in this case depth increases with age.
Has anyone done comparisons between an area where snow accumulates more slowly with ones where it accumulates more quickly? This way one can compare the effects of depth vs. age. Also if this proxy is so accurate one would expect that ice cores worldwide would agree except for the effects of averaging due to speed of accumulation.
Sorry but I don’t trust climatologists because of all the wrong assumptions I’ve seen them make in the past. Like for instance assuming that increased melting of ice in the Himalayas would lead to reduced river flow. Of course, the opposite is true, a net increase in flow. The only way that wouldn’t happen is if it sublimates, and that is not the claim.
Ferdinand Engelbeen : The data is here:
http://members.westnet.com.au/jonas1/CO2Data_DataExtractedForFerdinandEngelbeen20101129.pdf
Text in the original download says “Source: R. F. Keeling, S. C. Piper, A. F. Bollenbacher and S. J. Walker Carbon Dioxide Research Group Scripps Institution of Oceanography (SIO) University of California La Jolla, California USA“. Barrow and S Pole data is dated May 2008, Mauna Loa February 2009.
I said in an earlier comment that “CO2 sloshes around the planet in ways that are not fully understood“. We have talked about the movement of CO2 around the atmosphere, but there are other movements too.
http://www.ldeo.columbia.edu/res/pi/CO2/carbondioxide/image/annfluxgmm2u2windmap.jpg
shows that CO2 is released by the oceans not only near the tropics, as would be expected, but also (rather unexpectedly??) in the far north and far south.
On Jaworowski and ice cores : You say “That doesn’t give any pressure on the air in the still open pores, simply because all these pores still are in contact with each other until the surface, only reduces the porosity and thus reduces the diffusion speed.“.
I find it extraordinarily unlikely that all the pores are open to each other at any depth. IOW when you say “the still open pores” you aren’t talking about all pores, you are probably talking about dramatically fewer pores as they are progressively buried.
But in any case, Jaworowski’s objection is to the assumption that “the age of the gases in the air bubbles is much lower than the age of the ice in which they are entrapped“, and cites Neftel claiming that “the air was about 90 years younger” than the ice. Jaworowski doesn’t claim that “there is no difference” (your words), but that it is “similar“. You are now talking about “the average gas age is 33 years younger than the ice age“. Seems we have come quite a long way towards Jaworoski’s position.
In that light, take another look at Jaworowski paper fig.10b.
http://www.greenworldtrust.org.uk/Science/Images/ice-HS/Fig-10.gif
A mere 33 years doesn’t get anywhere near closing that gap. The main thrust of Jaworowski’s argument seems to be standing up pretty well.
Mike Jonas says:
November 28, 2010 at 8:47 pm
Ferdinand Engelbeen : The data is here:
http://members.westnet.com.au/jonas1/CO2Data_DataExtractedForFerdinandEngelbeen20101129.pdf
I see the difference in interpretation now: indeed the difference between BRW and MLO is somewhat decreasing over the full period, but still MLO is lagging after BRW with several months. If the decrease in lag is substantial or temporarely should be seen in a trendline of the differences (right now I have not the time to plot it myself) between BRW and MLO, as BRW is far more variable than MLO…
That the oceans have different places where CO2 is upwelling and absorbed is known: Even in cold oceans, if there is upwelling if the deep oceans, that water is oversaturated with CO2 and there CO2 is released. That is part of the continuous flow of CO2 between upwelling and places of absorption. The mid-latitudes are intermittent: upwelling in summer, sinks in winter. But that all doesn’t influence the trend that much (except for overall temperature changes, which is minor), only the variability around the trend.
I find it extraordinarily unlikely that all the pores are open to each other at any depth. IOW when you say “the still open pores” you aren’t talking about all pores, you are probably talking about dramatically fewer pores as they are progressively buried.
Well it is measured: CO2 was measured in situ in still open pores, which was from the surface to where the first pores were closing: that is at 72 meters depth for Law Dome. Only 10 ppmv difference with CO2 at the surface, thus about 7 years older than at the surface.
CO2 was also measured in the already closed air bubbles at the same depth (72 meter) via the normal route for ice cores. Exact the same level of CO2 was found, thus again 7 years older than at the surface. Thus the average CO2 of all closed bubbles and the average CO2 of all open pores at the same depth is the same.
But in any case, Jaworowski’s objection is to the assumption that “the age of the gases in the air bubbles is much lower than the age of the ice in which they are entrapped“, and cites Neftel claiming that “the air was about 90 years younger” than the ice. Jaworowski doesn’t claim that “there is no difference” (your words), but that it is “similar“. You are now talking about “the average gas age is 33 years younger than the ice age“. Seems we have come quite a long way towards Jaworoski’s position.
Jaworowski states that there is no difference between the age of the ice layers and the age of the enclosed air (*). This is proven false, both theoretically calculated based on diffusion speed and pore diameter and measured in a few ice cores. There is nothing arbitrarely in using the average gas age and compare it with the atmospheric data.
The ice age – gas age difference depends of the accumulation speed, which is quite different for different ice cores. For the fast accumulating coastal ice cores at Law Dome it is about 33 years, for Siple Dome (Neftel) some 83 years and for far inland Vostok hundreds to thousands of years.
Literally from correspondention with him:
The air/ice age difference still remains as an unproved assumption, and the CO2 hokey curve from Law Dome, based on this assumption is false.
“The question remains; is it OK to splice data from two different sets, then present them as a single data set…”
The answer remains: absolutely not. Not EVER. Particularly when it is something as important as this…
Ferdinand Engelbeen : “CO2 was also measured in the already closed air bubbles at the same depth (72 meter) via the normal route for ice cores. Exact the same level of CO2 was found, thus again 7 years older than at the surface. Thus the average CO2 of all closed bubbles and the average CO2 of all open pores at the same depth is the same.”
Something’s wrong. CO2 is supposed not to travel through the ice, that is the whole basis of using ice cores to determine past CO2 levels. So the bubbles that had been closed for a few years should show lower CO2 levels than the ones that are still open. They don’t. This means that CO2 does travel through the ice, and very effectively.
One has to question what value the ice cores have.
Mike Jonas says:
November 29, 2010 at 11:47 am
Something’s wrong. CO2 is supposed not to travel through the ice, that is the whole basis of using ice cores to determine past CO2 levels. So the bubbles that had been closed for a few years should show lower CO2 levels than the ones that are still open. They don’t. This means that CO2 does travel through the ice, and very effectively.
One has to question what value the ice cores have.
Mike, as already said (and as described in the work of Etheridge of 1996, unfortunately behind a paywall):
CO2 was measured in situ in still open pores, which was from the surface to where the first pores were closing: that is at 72 meters depth for Law Dome.
and
CO2 was also measured in the already closed air bubbles at the same depth (72 meter) via the normal route for ice cores.
Indeed once the pores are closed, there is no migration anymore (or so little that it is undetectable). But as the closing process takes several meters from the first closed bubbles to where all pores are closed, therefore the first bubbles contain older air than the last closers. Depending on the accumulation rate, that takes about 8 years for Law Dome (2 cores), 21 years for the third core (taken at the slope with less accumulation per year), some 22 years for Siple Dome and some 560 years for Dome C.
Thus the average composition in the ice core at a certain depth is a mix of CO2 levels from different years, which makes that the resolution is worse, the less snow deposit there is over a year. But at the other side, the less deposit, the more layers one has at the same depth, thus the farther back in time one can measure the air composition from that period: from 150 years for Law Dome to 800,000 years for Dome C.
Ferdinand Engelbeen – I don’t want to flog a dead horse, but earlier you said “the average CO2 of all closed bubbles and the average CO2 of all open pores at the same depth is the same“, now you say “the average composition in the ice core at a certain depth is a mix of CO2 levels from different years“.
There seems to be a contradiction. Two bubbles at the same depth, one closed and one open, originally formed at the same time (because they are the same depth). The open one continues to change its CO2 content to match the outside world, whereas the closed bubble’s CO2 cannot move through the ice surrounding it and therefore matches what the outside world looked like at the time that the bubble closed. This means that the closed bubble and the open bubble will have different CO2 content, in line with your later statement. I don’t then see how your earlier statement can be correct.
So let’s work from your later statement. The argument for air age being much younger than ice age (shown in Jaworowski fig 10a) now depends on all pores remaining open until the last moment, at which time they all suddenly close. That seems absurd to me. Surely the reality is that many more pores will close in the early years than in the later years. This means that the average air age will not be all that different to the ice age, and certainly nowhere near enough to close the gap in Jaworowski fig 10b.
And the AGW case depends on that gap being closed.
Mike Jonas says:
November 29, 2010 at 5:10 pm
Ferdinand Engelbeen – I don’t want to flog a dead horse, but earlier you said “the average CO2 of all closed bubbles and the average CO2 of all open pores at the same depth is the same“, now you say “the average composition in the ice core at a certain depth is a mix of CO2 levels from different years“.
OK, I think that I need to be more clear:
For the Law Dome ice core (other depths and years applicable for different ice cores):
– underway down to about 72 meter: only open firn (compressed snow), no closed bubbles. In situ measured CO2 gradient about 10 ppmv top-down. Near top equal to above surface, at 72 m depth about 7 years older.
– at closing depth (72 meter): in situ measured open pores CO2 10 ppmv below surface. Thus 7 years older than surface. Ice layers counted: 40 ice layers, ice is 40 years old. Ice core extracted and ice with first closing bubbles crushed and air from the bubbles measured: the same CO2 levels is found.
– at final closing depth (about 85 meters): ice 48 layers, 48 years old. CO2 in last open cores: about 18 ppmv less than surface, or about 13 years older. CO2 in ice at same depth: about 20 ppmv less than surface, or about 14 years older. Hardly a difference.
See: http://www.ferdinand-engelbeen.be/klimaat/klim_img/law_dome_firn.jpg
Why so little difference? With depth, the diffusion speed of any molecules becomes more and more difficult, due to smaller and smaller pores. At a certain depth (72 m for Law Dome), some ice layers are sealed (vertically, but not horizontally) mainly for winter deposits. Thus the still open pores below first sealing depth have no exchange with the atmosphere anymore. For Law Dome, first closing depth and sealing depth are equal, but some ice cores (Siple Dome) have sealing layers above first closing depth.
Etheridge says the following about the enclosing:
Air becomes sealed in the ice sheet and is subsequently enclosed in bubbles as the open porosity of the firm layer decreases with depth. The enclosed air at any depth in the ice has a mean age a(a) that is younger than the age of the host ice layer a(i), from which the air is extacted. This difference, da, equals the time T(s) for the layer to reach a depth d(s), where air becomes sealed in the pore space, minus the mean time T(d) for the air to mix down to that depth.
Mixing of air from the ice sheet surface to the sealing depth is primarly by molecular diffusion.
Thus my interpretation was wrong. The gas phase averaging happens already above sealing depth and the mixed air is the same in still open pores and already closed bubbles below any depth below the sealing depth. Thus you were right to point to the discrepancy. Not so for what Jaworowski says.
So let’s work from your later statement. The argument for air age being much younger than ice age (shown in Jaworowski fig 10a) now depends on all pores remaining open until the last moment, at which time they all suddenly close. That seems absurd to me. Surely the reality is that many more pores will close in the early years than in the later years. This means that the average air age will not be all that different to the ice age, and certainly nowhere near enough to close the gap in Jaworowski fig 10b.
And the AGW case depends on that gap being closed.
As the measurements show:
At bubble closing start, the average (from about 10 years) gas age at 72 meter at Law Dome is some 7 years younger than at the surface. The ice age at the same depth is 40 years. Difference 33 years.
At end bubble closing, the average (from about 10 years) gas age at 85 meter at Law Dome is some 14 years younger than at the surface. The ice age at the same depth is 48 years. Difference 34 years.
Thus there is a gas age-ice age difference of about 33 years in the Law Dome which depends of deposition speed (which may be different if the climate changes, even for the same core).
No matter if all bubbles closed at once or several meters lower at once, the ice age – gas age difference remains, That means that there is no gap between the gas age estimates and the direct measurements of the atmosphere, even an overlap of some 20 years:
http://www.ferdinand-engelbeen.be/klimaat/klim_img/law_dome_sp_co2.jpg
Thus figure 10a of Jaworowski is proven right (as measured for Law Dome, calculated for Siple Dome), figure 10b of Jaworowski is based on the assumption that there is no difference between the age of the ice layers and the enclosed air, which is proven wrong.