Abrupt Cooling at the Eocene-Oligocene boundary

The Earth has gone through significant climate change over the last 100 million years going from the mid-Cretaceous “greenhouse” to the late Cenozoic icehouse. Roughly 33.7 million years ago, at the Eocene-Oligocene boundary (EOB), the world would begin to develop permanent ice sheets, particularly the East Antarctic. This accompanied the so-called Oi-superglacial driven by a reduction in atmosphere CO2 concentrations.

deep_time_climate
IPCC AR4 Chapter 6, Figure 6.1

In paleoclimate studies using stable isotopes, oxygen is of primary interest, along with deuterium in ice core research. The oxygen isotopic composition of seawater is the most commonly used metric of deep-time paleotemperatures. Oxygen isotopes from benthic (bottom-dwelling) foraminiferal calcite obtained from deep-sea cores generally provide good representations of global variations as they are isolated from geographic, and seasonal variations, unlike surface records. The oxygen isotopic composition of a sample is often expressed as a departure of the 18O/16 ratio from an arbitrary standard.

S= 103*[(18O/16O)sample – (18O/16O)standard]/18O/16Ostandard

In which resulting values are in “per mil” (%o) units. Positive values represent higher ratios in the sample (more 18O) and negative values represent lower ratios (and therefore isotopically lighter). During glacial times, the preferential removal of isotopically light oxygen to form ice sheets leads to an increase in the 18O/16O ratio.

In pre-Quaternary climates, benthic (bottom-dwelling) 18O/16O records do not discriminate between ice volume and cold temperature. Despite this climate transition at the EOB, this dilemma of cooling vs. ice-sheet growth and sea-level fall reconstructed from oxygen isotope values jhas not been resolved in any satisfying manner. Some studies have showed ice sheet growth at the EOB unaccompanied by cooling, something that is certainly counter-intuitive.

In a recent paper in Science, the authors Liu et al. use two independent temperature-proxies (called the alkenone unsaturation index (UK’37) and tetrather index (TEX86) from 11 different locations in the ocean to determine what part of the oxygen isotope changes is from ice-volume and what part is from temperature. They can then subtract their results from the 18O/16O record to resolve the component that has to do with ice sheet growth. Further insight on this is given in the Perspective article by Penn State’s Lee Kump.

Both temperature proxies show high-latitude cooling a little over 33 million years ago. Model simulations run by the authors show ~5°C of high-latitude cooling consistent with their proxy data, global-mean surface cooling of 4.4°C, with slightly less ocean floor cooling than surface changes. The authors warn these values may be underestimates (for the model, the inability to account for the full range of ice sheet feedbacks, and for the proxies, low resolution data, sediment gaps, and poor core recovery.)

There’s large uncertanties, as the authors show Antarctica became roughly 40 to 120% of its modern size during the EOB transition, and and 30 to 100% of Antarctic ice volume during the LGM. There was no significant Northern Hemispheric glaciation at this time. The transition is consistent with reductions of greenhouse gas concentrations because of the globally synchronous changes at the time. In their model, a range of atmosphere CO2 concentrations appropriate for the Late Eocene and Early Oligocene conditions was applied. The reduction of atmospheric CO2 levels necessary to explain the full range was more than CO2 proxy estimates suggest, which Lee Kump interprets as meaning we’re not including some important variable effecting long-term sensitivity estimates (which I discussed in The Uncloudy Cretaceous).

7 responses to “Abrupt Cooling at the Eocene-Oligocene boundary

  1. Just dropping by.Btw, you website have great content!

    ______________________________
    Who Else Wants To Discover A Rebel Psychiatrist’s Amazing Secret That Lets You Put People Under Your Control Quickly & Easily and Get Them to Do Anything You Want?

  2. Chris,

    I see you’ve read this article as well.

    I would like someone to tell me, so what caused the CO2 levels to fall suddenly?

    That’s the question one would surely have to answer when proposing a bizarre theory like this, but the researchers of Dr. Liu’s 10-author team(!) do not even trouble themselves to ask it.

    The actual peer-reviewed Science article, is Liu, Z, M Pagani, D Zinniker, R DeConto, M Huber, H Brinkhuis, R Sunita, R Shah, M Leckie, and A Pearson 2009: Global Cooling During the Eocene-Oligocene Climate Transition, Science 27(323), 1187-1190.

    I note that M. Pagani is a frequent alarmist co-author with James Hansen in a number of other Hansen et al. papers.

    The CO2 ‘hypothesis’ is not mentioned in the abstract, and in the conclusion they merely claim that their results are ‘…consistent with a scenario of global cooling forced by a reduction in greenhouse gas concentration rather than the more regionalized effects of ocean gateways.’

    So again, why are they proposing CO2 as the cause in the first place? It’s fine to show that ‘if CO2 changed, then we might see this.’ Surely, you’d have to firstly argue the case for why the CO2 changed?

    In their introduction it seems to me that much better theories are just brushed aside in passing:

    Proposed causes for this fundamental change in Earth’s climate state [viz. the large drop in temperature Eocene->Oligocene] include changes in ocean circulation due to the opening of Southern Ocean gateways [the standard theory as I understand it], a decrease in atmospheric CO2, and a minimum in solar insolation.

    Their CO2 hypothesis is again just introduced arbitrarily on p. 1189.

    “To evaluate how surface temperature change was translated into the deep ocean, we ran coupled atmosphere-ocean simulations with Eocene boundary conditions and CO2 levels representative of pre- and posttransition atmospheric mixing ratios.”

    Yet it is well-known that CO2 levels always fall following decreases in temperatures; they do not drive the changes.

    The next thing we know, voila, headlines have appeared all around the world:

    ‘Prehistoric global cooling caused by CO2’!

    So tell me why I shouldn’t regard this as peer-reviewed pseudoscience, with no purpose other than to generate more unfounded alarm … ?

    Response– Are you still on this “CO2 lags temperature” nonsense? You’re about a decade behind everyone who is taking this subject seriously. Your “Yet it is well-known that CO2 levels always fall following decreases in temperatures; they do not drive the changes” is a straight out misrepresentation of peer-reviewed research.

    By “suddenly” we’re talking about millions of years, with long-term CO2 controls involving geological and biological boundary condition changes which are irrelevant over decades. Standard explanations are quite easy to find. And the CO2 is necessary to explain the full range and global extent of the long-term changes which cannot be explained by ocean circulation changes (this term more or less provides a redistribution of heat rather than significant global temperature anomaly).

    I appreciate your input but I’m not interested in further discussion unless you retract the quoted statement that you probably know is just bogus. Your conspiracy theories are also nice, but not desired. — chris

  3. Hi Chris,

    To begin with, I should make it clear that I am a curious, concerned layperson and not a trained scientist. My background is in computer science & history & philosophy of science. Thus, it’s quite likely I make mistakes from time to time.

    That said, I am nonetheless somewhat surprised to hear you say that it is ‘nonsense’ that CO2 lags temperature. Indeed, the following post at RealClimate.org by Eric Steig acknowledges this:

    Steig, E 2007: The lag between temperature and CO2. (Gore’s got it right.), at http://www.realclimate.org/index.php/archives/2007/04/the-lag-between-temp-and-co2/

    Steig quotes Lorius et al. 1990 as saying “…changes in the CO2 and CH4 content have played a significant part in the glacial-interglacial climate changes by amplifying, together with the growth and decay of the Northern Hemisphere ice sheets, the relatively weak orbital forcing…”

    He goes on to say that “CO2 acts as an amplifier.”

    The driver of the temperature changes (in the glacial-interglacial cycle anyway) is said to be orbital changes as proposed by Milankovich.

    Now I am aware that people have proposed that CO2 decreases from reforestation caused the Little Ice Age but at least in that case, a cause for the decreased CO2 was proposed.

    There can surely be no way of knowing for certain just from the raw data as to whether these temperature changes preceded or followed the CO2 level changes.

    Thus, surely I am right to ask, why have they not discussed what might have caused CO2 levels to fall so far?

    Response– The punching bag statement that “CO2 lags temperature” is generally confined to the start of glacial-interglacial cycles. It doesn’t include the full range of those cycles, and it doesn’t include other deep-time paleoclimatic events (e.g., the PETM, the Cretaceous, when CO2 definitely mattered), and CO2 is most certainly not a temperature feedback today. It becomes a chicken and the egg problem. It should not be surprising that changing the surface temperature affects the atmospheric chemsitry (through changes in the biosphere, changes in gas solubility in the water, etc) and on the flip side, changes in atmospheric chemistry affect temperature through radiative perturbations. And yes, this had been predicted long before it was observed in the ice core record. You can see for yourself one of the papers cited by the “Swindle Documentary” to make their “CO2 lags temperature” argument, and if you read the body (or just the conclusion) you’ll find they left some information out.

    Individual peer-reviewed articles are not assessment reports, and they generally have a very narrowed question that they want to ask, and they address it. As such, they usually ignore many other interesting questions that can come out of their study. This is why science builds on previous work and is a self-evolving and self-correcting process. As I’ve said, long-term removal of CO2 involve tectonic movements (e.g., the lifting of the Tibetan plateau, see Raymo and Ruddiman 1992) and changes in biomass. CO2 concentrations over millions of years are constrained by volcanic outgassing and the chemical weathering of rocks.– chris

  4. I recently posted this on CA
    http://www.climateaudit.org/?p=4804#comment-329759

    The plots are from EPICA data, mainly not VOSTOK
    The time scales are reversed (years before present)
    The plots show temperature rise pretty much co-incident with CO2 rise. (samples are often too widly separated to safely say which came first!

    ————————
    Does temperature precede GHG change?
    Some plots
    CO2 in most cases rises at the same time as temperature. CH4 seems to terminate the warm period in many cases. Data is from EPICA core as this is more detailed than vostok. BUT core dates can still be spaced at over 2k years per sample in some periods. N2O and O3 have not been plotted.
    Where is the data that shows temperature rise precedes CO2?
    0 to 40,000 years. GISP2 and EPICA temperatures plotted on this graph. Co2 steady rise is simultaneous with temperature @17500ybp
    note that only greenland gisp2 temperature shows a definite younger dryas – the antarctic EPICA data shows a flattening only.The EPICA CH4 data shows a misplaced drop around the younger dryas. Note the dust levels during the low temperature portion.

    40k to 100k years Note the dust levels are non zero during this period and high during the low temperature portion.

    100k to 200k years Co2 rises simulaneously with temperature @136kybp. Note the dust levels are high during the low temperature portion. CH4 termination of warm period

    180k to 260k years Co2 rises simulaneously with temperature @252kybp. the 220kybp is less defined. Note the dust levels are high during the low temperature portion. CH4 and CO2 termination of warm periods

    280k to 360k years Co2 rises simulaneously with temperature @341kybp. Note the dust levels are high during the low temperature portion. CH4 termination of warm periods

    360k to 460k years Co2 rises simulaneously with temperature @432kybp. Note the dust levels are high during the low temperature portion. CH4 termination of warm period

    460k to 560k years Co2 rises simulaneously with temperature @532kybp. Note the dust levels are high during the low temperature portion. CH4 termination of warm period

    560k to 650k years Co2 rises simulaneously with temperature @629.5kybp. Note the dust levels are high during the low temperature portion. CO2,CH4 termination of warm period

    650k to 760k years Co2 rises simulaneously with temperature @740.5kybp. Note the dust levels are high during the low temperature portion. CO2,CH4 termination of warm period @694kybp. Note dip at 722kybp has no CH4/Co2 driving. It is possible that dust level rises at this time but granularity of dust data is not sufficiently small to line up.

    750k to 800k years Co2 rises simulaneously with temperature @796kybp. Note the dust levels are high during the low temperature portion. CO2 termination of warm period

    Methane data is from:

    Loulergue, L., et al.. 2008.
    EPICA Dome C Ice Core 800KYr Methane Data.
    IGBP PAGES/World Data Center for Paleoclimatology
    Data Contribution Series # 2008-054.
    NOAA/NCDC Paleoclimatology Program, Boulder CO, USA.

    Age scale is gas age

    CO2 data is from
    0-22 kyr BP: Dome C (Monnin et al. 2001) measured at University of Bern
    22-393 kyr BP: Vostok (Petit et al. 1999; Pepin et al. 2001; Raynaud et al. 2005) measured at LGGE in Grenoble
    393-416 kyr BP: Dome C (Siegenthaler et al. 2005) measured at LGGE in Grenoble
    416-664 kyr BP: Dome C (Siegenthaler et al. 2005) measured at University of Bern
    664-800 kyr BP: Dome C (Luethi et al. (sub)) measured at University of Bern

    Age scale is gas age

    I assume the gas age takes into account the delay in trapping?

    The age used is EDC3 and a comparison between dome fuji and vostok is here
    The EDC3 chronology for the EPICA Dome C ice core
    http://www.clim-past.net/3/485/2007/cp-3-485-2007.pdf

    gas to ice age 0-41k
    http://www.clim-past.net/3/527/2007/cp-3-527-2007.pdf
    “Although the exact causes of the 1arge
    overestimate remain unknown, our work implies that the suggested
    lag of CO2 on Antarctic temperature at the start of the
    last deglaciation has probably been overestimated.”

  5. Chris,

    My main issue with the CO2 driver hypothesis that appears in the Liu et al 2009 paper is that it seems to be superfluous, post hoc to the paper’s thesis, which ostensibly concerns measurement of temperature over the Eocene-Oligocene boundary. The hypothesis just appears out of nowhere on p. 1189 and it is not well-motivated at all. I cannot see that the paper is attempting to support the CO2 hypothesis, and that the conclusion that the results ‘are not inconsistent with’ the hypothesis is completely irrelevant. The results are not inconsistent with many things, so why mention them unless they provide positive evidence & argument?

    In any case, you would surely have to concede that the headlines that appeared within days in many newspapers around the world that ‘CO2 caused ancient global cooling’ go way beyond the paper’s actual conclusions? You would surely agree that nothing at all has been ‘proven’?

    On CO2 lagging temperature, I would be interested to know if you have read the following paper:

    Roe, G. (2006) In defense of Milankovitch. Geophys. Res. Ltrs., 33, L24703, doi:10.1029/2006GL027817

    http://earthweb.ess.washington.edu/lnk/gerard/Publications/Roe_Milankovitch_GRL06.pdf

    This paper seems to show that any role for a CO2 driver in the ice age cycle is limited at best.

    Response– There are a couple of schools of thought here. One is that CO2 forces ice volume (slow ice changes and corresponding lags of thousands of years). Another view holds that CO2 acts as a positive feedback on ice-sheet mass balance. Simple time-series analysis cannot resolve this and so a variety of views exist. The answer also depends on where in the glacial-interglacial cycles you look at (Termination 1, 2, 3 ??) since they don’t all behave the same way. See Ruddiman 2006 for some discussion. In regard to Termination 3 (~240 kyr before present) the Caillon paper I cited states “Second, the CO2 increase clearly precedes the Northern Hemisphere deglaciation.”

    As for the Roe paper, you also need to look at the geographic location of interest, and whether he’s just talking about ice volume or temperature. In fact the primary influence of Milankovitch is its solar insulation impact above ~60 N which constrains NH ice volume, since mass balance of ice sheets is sensitive to summertime temperature. There’s very little global insulation change. The author notes “This certainly does not rule out CO2 as a primary cause of tropical or other climate variations, or of the apparent synchronization of the ice-age signal between hemispheres.”

    The radiative forcing into the LGM from CO2 is quite large, with that and ice sheet forcing the largest two components (see this diagram) and so your interpretation of its minor role is not correct (this is not an argument made in any of the cited studies). Howver, its relationship with global ice volume is still not well understood

    The conclusion that long-term temperature decline out of the Eocene-Oligocene (and a transition into a world favorable to ice sheets) is caused by declines in CO2 is not confined to this paper. Headlines that you describe are accurate but not “news.” I don’t think this is a point of emphasis by Liu et al. as much as a consistent explanation. The more interesting conclusion is the partitioning between ice sheet growth or cooler SST’s as interpreted from the proxy data.– chris

  6. Chris,

    Hold on, it is quite clear that the authors of the Liu et al. study are positively claiming that the CO2 hypothesis is not proven: “Proposed causes for this fundamental change in Earth’s climate state include changes in ocean circulation due to the opening of Southern Ocean gateways (4), a decrease in atmospheric CO2 (5–8), and a minimum in solar insolation (2)” and “These results are consistent with a scenario of global cooling forced by a reduction
    in greenhouse gas concentration rather than the more regionalized effects of ocean gateways (5–8).” I’m sorry, but it says what it says: the causes of the E-O cooling remain hypothetical, and our paper provides no evidence that can be used to settle the matter one way or another. If they were claiming otherwise, they would state “Our results SHOW…” or “This study has DEMONSTRATED…” The conclusion, “these results ARE NOT INCONSISTENT WITH” is irrelevant, and unscientific.

    The same goes with your Roe quoation. He says, “This certainly does not rule out CO2 as a primary cause of tropical or other climate variations, or of the apparent synchronization of the ice-age signal between hemispheres.” But if it doesn’t rule it IN, then it is a completely irrelevant statement. You know, it doesn’t rule out the possibility that Caesar winked as he crossed the Rubicon either. Wouldn’t you agree?

    These sorts of remarks are just not scientific. Why mention the CO2 hypothesis at all if the intention is not to provide a positive motivation for it?

    Response– My general remarks about the CO2 connection were not meant to be confined to this paper specifically, but rather the literature as a whole on deep-time paleotemperatures and CO2 connection. I didn’t say Liu et al. “proved” anything (I generally don’t use such language because it’s hard to “prove” anything in science). However, they did suggest that models with CO2 reduction are a consistent explanation given the spatial extent of the cooling. They ran simulations with CO2 levels representative of pre- and post transition atmospheric concentrations, compared simulations with proxy data, and so this part of their paper was not just something “thrown in.” T

    What is a better question (and what is more uncertain) is how that forcing compared globally and regionally with other things. CO2 had to play at least some role from a purely physics-based approach, and it is the key climate driver over geologic time, as well as the standard starting point for a transition from a hothouse to an icehouse world. There is plenty of other literature (some referenced by them) on this subject so I don’t really see what you’re arguing against. Dr. Robert Anderson has some perspective on CO2 feedback mechanisms in glacial times in his comment inside my other thread “new hypothesis for deglacial CO2 rise.” — chris

  7. I am very enjoy your blog, your blog is very true of the bar, hoping to see you more exciting content! I wish you have a happy day!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s