Climate Science Glossary

Term Lookup

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


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

Term Lookup


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

Home Arguments Software Resources Comments The Consensus Project Translations About Support

Bluesky Facebook LinkedIn Mastodon MeWe

Twitter YouTube RSS Posts RSS Comments Email Subscribe

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

New? Register here
Forgot your password?

Latest Posts


Carbon Dioxide the Dominant Control on Global Temperature and Sea Level Over the Last 40 Million Years

Posted on 5 March 2013 by Rob Painting

Key Points:

  • Because the water contained in land-based ice sheets is ultimately derived from the ocean, over long (geological) timescales global sea level is largely determined by global temperature and, consequently, the temperature-dependent volume of ice stored on land.
  • Since the concentration of carbon dioxide in the atmosphere (The Greenhouse Effect) exerts such a powerful influence on global and polar temperature, it therefore follows that it should also exhibit a strong relationship with global sea level over geologic intervals of time.
  • Foster & Rohling (2013) examined time slices of paleo data covering the last 40 million years to uncover the details of this carbon dioxide-sea level relationship. Surprisingly, they found a consistent and robust relationship between carbon dioxide and sea level irrespective of other contributing factors.
  • Based on the concentration of carbon dioxide in the atmosphere as of 2011, the authors estimated that future sea level is committed to rise  24 metres (+7/-15 m) above present-day once the land-based ice sheets have fully responded to the warming and the Earth is once more in equilibrium. 
  • The authors estimated that this sea level rise will likely take place over many centuries, if not several thousand years, but it nevertheless represents the long-term consequences of human industrial activity, and is  further evidence that CO2 is the Earth's "main control knob" for global temperature. 

Figure 1(a) - Relationship between atmospheric CO2 (in parts per million) and global sea level (in metres) over the duration of the ice core record. The dashed horizontal line depicts the pre-industrial state for comparison (b) Shows the cross-plot of CO2 versus sea level rise over this period. CO2/Co on the horizontal axis is the CO2 ratio relative to the pre-industrial where 0.0 respresents the pre-industrial, with negative values (below) and postive values (above) relative to pre-industrial. From Foster & Rohling (2013) 

Background Context: Greenhouse Gases and Planetary Warming

Carbon dioxide is the most significant of the greenhouses gases, gases that trap heat in Earth's atmosphere and reduce the rate of heat loss to space in the upper layers of the atmosphere. As a result the temperature of the planet's surface and ocean is largely dependent upon the concentration of greenhouse gases in the atmosphere. Generally-speaking; increase the concentration of greenhouse gases and the planet traps more heat, reduce them and the planet cools. Because of this behaviour, renowned glaciologist, Richard Alley, has dubbed carbon dioxide Earth's biggest temperature control knob

The oceans are the Earth's largest heat reservoir and, in comparison to the atmosphere, they respond very slowly to warming because of their immense mass, vast heat capacity, and the length of time it takes for heat to be transported down to the deep ocean. The long-lived nature of CO2 therefore means that the oceans (and the atmosphere) will continue to warm, albeit at a comparatively much slower rate, even long after humans have ceased pumping planet-warming greenhouse gases into the atmosphere (Meehl [2012]). In other words, we have yet to see the land-based ice fully respond to human intervention in the Earth's  climate.

How High Will The Water be Mama?....

In examining ancient sea level it would, of course, be preferable to be able to match sea level at any given point in Earth's history with global temperature, especially as further warming is in the pipeline. However, a number of problematic assumptions have to be made in order to transform various paleo data into global temperature estimates over seperate geological periods. Foster & Rohling (2013), therefore, quantified the relationship between two entities that could be measured more directly - land ice volume/sea level and carbon dioxide levels. So, based on the carbon dioxide/sea level relationship, and the greenhouse gases humans have already emitted, how high is this sea level rise we are already committed to? The authors used high quality data from the Earth's geological past to answer this question.

Sea Level & the Ice Cores

Ice core records retrieved from drilling expeditions on the Antarctic ice sheet date back 800,000 years (Petit [1999]Luthi [2008]) and provide a high fidelity record of the Earth's atmospheric carbon dioxide concentration over this interval. This is due to the fact that the ice contains bubbles of air which became trapped in snowfall and were eventually sealed off from the atmosphere over time. Each air bubble, therefore, contains a tiny sample of carbon dioxide from the atmosphere at the time it was sealed off, and each successive layer of ice deposited enables a reconstruction of atmospheric carbon dioxide going back in time - as shown in Figure 1.

During the last few million years (the Pleistocene), the Earth has been in the grip of the ice ages - long cold intervals (glacials) interspersed with shorter warm periods such as today (interglacials)- which were driven by orbital forces, and reinforced by carbon dioxide acting as a feedback (although the details are sketchy). Even though the ice sheets only provide us with 800,000 years of this period, it is most probable that sea level rarely exceeded the pre-industrial level. This can be seen in Figure 1 for the last 550,000 years.

So, although useful in establishing the overall carbon dioxide/sea level relationship, the overwhelming majority of the ice core record is not greatly informative of the future because sea level and atmospheric carbon dioxide were generally much lower than present-day, and sea level change involved the growth or disentegration of the gigantic Laurentide and Fennoscandian ice sheets in the Northern Hemisphere - ice sheets which do not exist today. Furthermore, the  relationship of steadily increasing CO2 with rising planetary temperature  is not a linear one. So we can't simply extrapolate the past CO2/sea level relationship into the future with any confidence.

Seeing Further Back in Time - Alkenones and Boron Isotopes

Throughout the last 40 million years the Earth has cooled in fits and starts from a very warm state with virtually no land-based ice, to the current glacial/interglacial period where monstrous ice sheets are capable of growing upon the land masses of the Northern Hemisphere. Given that Earth is headed toward a future where the polar ice sheets are greatly reduced, warmer and cooler periods  of ancient Earth might yield some insight. To examine these intervals it was necessary for Foster & Rohling (2013) to turn to other paleo data, such as alkenone and boron isotopes. These methods rely on the knowledge that carbon dioxide dissolved into the ocean, and therefore ocean pH, are dependent on the partial pressure of carbon dioxide in the atmosphere.

Alkenones are organic (natural) compounds produced by certain types of ocean surface-dwelling phytoplankton and, like other paleoclimate proxies, provide information about the state of Earth's climate at the time they were formed. Carbon dioxide from seawater is incorporated into living tissue through the process of photosynthesis, and alkenones are one of the by-products. As a result, alkenones contain information about the amount of carbon dioxide dissolved in seawater and therefore atmospheric CO2 concentrations - since they are derived from the well-mixed sea surface layers.  These organic compounds are extremely durable and are highly resistant (although not immune) to long-term degradation (diagenesis). Over the course of time the bodily remains of alkenone-bearing phytoplankton settle on the ocean floor and build up in  sediments. By examining sediments taken from cores drilled into the seafloor it is possible to use our knowledge of chemistry and biology to work backwards and reconstruct atmospheric CO2 at the time the alkenones were formed. See Pagani (2002) for a review of alkenones as proxies. 

Boron isotopes are another sediment-based method employed to determine ancient carbon dioxide concentrations. This technique differs somewhat from alkenones in that the chemical signatures are stored within the fossilized shells of foraminifera (forams), tiny marine life which live near the sea surface (planktic). Like alkenones, this surface-dwelling characteristic is important because the surface ocean layers are well-mixed with the atmosphere, and therefore ocean pH, once reconstructed, allows atmospheric CO2 to be calculated (albeit with some degree of uncertainty). Minerals such as boron dissolved in seawater are incorporated into the shells of forams and, because the boron isotope ratio is predominately a function of the ocean pH, they are recorders of ocean pH at the time when the shells were formed. See Foster (2008) for details. 

The S-Shaped Nature of the Sea Level/Carbon Dioxide Relationship

Foster and Rohling (2013) examined 5 time slices over the last 40 million years -the ice core record dating back 550,000 years, 2.7-3.1 million years ago during the Pliocene, 11-17 million years ago during the Miocene, 33-35 million years ago during the Eocene-Oligicene boundary, and 20-40 million years ago during Eocene-Oligicene. Sea level for the ice core record was based upon the Red Sea reconstruction (discussed in this SkS post). For the other time slices  estimates are based upon the oxygen isotope ratio (global ice volume signal) in fossilized forams (Waelbroeck [2002]), the magnesium/calcium content of foram fossils to isolate the oxygen isotope/global ice volume signal (Barker [2005]). And finally, analysis of the character of depositional features (lithofacies) and physical structure of ocean floor sediments (backstripping), which can be used to estimate global sea level volume.      


Figure 2(a) - cross-plot estimates of atmospheric CO2 and sea level. The data are labelled according to the time period and reconstruction technique employed. The dashed vertical and horizontal lines denote the pre-industrial CO2 and sea level respectively. The vertical and horizontal bars on each data point represents the uncertainty of the sea level and CO2 estimates respectively. 3(b) Probabilistic analysis that fully accounts for uncertainty in both sea level and CO2 data, with the bold line indicating the probability maximum . From Foster & Rohling (2013).  

Interestingly, data and subsequent atmospheric CO2 and sea level reconstructions, show very good agreement in spite of the various methods used. This can be seen especially over the ice core record, where data from the Miocene and Pliocene are consistent with the Red Sea reconstruction. This is surprising given the different continental configurations, oceanic pathways, and  mountain-building (orography) over this vast stretch of time might have been expected to yield varied estimates. This general agreement does, however, give some degree of confidence that the results are robust.    

The most notable feature of the reconstruction, however, is its s-shaped (sigmoidal) nature. Despite global temperature decreasing between 650 to 400 parts per million (ppm) of atmospheric CO2, the sea level response appears to taper off. It is obvious that this interval is very light on data points, but given the characteristic shape of all the data, it seems unlikely that the overall complexion will change once the interval is populated with more data. It should be noted, however, that this data (above 400ppm) is generally from intervals when land-based ice was growing and sea level falling as a consequence. In contrast, data from 180-400 ppm comes from both cooling and warming periods, which suggests that ice sheet growth and disintegration, between these CO2 concentrations, exhibit similar behaviour. 

That may not be the case with rising sea level at values above 400ppm, the system may exhibit hysteresis. In other words, above 400 ppm warming and sea level rise may not behave in the same manner as it did when the planet was cooling and sea level was falling. The ultimate fate of land-based ice above atmospheric CO2 concentrations of 400 ppm could be path-dependent. 

........24 Metres High and Rising? 

Atmospheric CO2 concentration as of 2011 was around 392 ppm, and based upon that and the carbon dioxide/sea level relationship revealed in their research, Foster & Rohling (2013) calculated that long-term sea level rise will reach 24 metres (+7/-15 metres at 68% confidence) above present-day sea level once the planet has fully responded to the warming. This is likely to be achieved through extensive disintegration of the West Antarctic and Greenland ice sheets, and a substantial part of the coastal sector of the East Antarctic ice sheet.

The existing land-based ice is equivalent to 60-70 metres, so the loss of 24 metres of sea level worth of ice suggests that over a third of the ice sheet volume may already be committed to disintegration. As noted in the introduction, this dramatic response represents the long-term consequences of human industrial activity and reinforces the concept that atmospheric CO2 is Earth's main temperature control knob.

For more information about this type of research go to:

0 0

Printable Version  |  Link to this page


1  2  Next

Comments 1 to 50 out of 91:

  1. Is fig 1a trying to show a cause and effect relationship between CO2 and sea level?  Or do they just happen to go up and down together, perhaps each caused by something else, temperature perhaps? 


    Because of this behaviour, renowned glaciologist, Richard Alley, has dubbed carbon dioxide Earth's biggest temperature control knob.

     Isn't this somewhat over-reaching?  There have been several ice-ages (as shown in above figure), none of these ice-ages was caused by CO2 levels, nor was the ending of these ages.  Clearly, something other than CO2 is the big control knob!


    0 0
    Moderator Response: (Rob P) Figure 1 shows what happened to sea level and atmospheric CO2. You have actually read the post you're commenting on right?
  2. Depends on how you look at it, Kevin.  Certainly the sun is the almost exclusive source of energy.  However, forcing from solar variation is not that strong (except in the very, very long range -- faint sun hypothesis, etc.).  It's just persistent.  Consider Milankovitch forcing.  It seems strong: it is responsible for major swings in global temp.  However, it's a pretty weak forcing.  It's just long-lasting enough to trigger a powerful, long-term ice-albedo process and a powerful, long-term change in ocean temp (leading to outgasing of CO2).  Because CO2 is not exclusively a feedback, it can and should be considered a powerful forcing--certainly enough to overwhelm orbital forcing (see Tzedakis et al. 2012).  As for CO2's dominance as a greenhouse gas, see Lacis et al. (2010).  

    0 0
  3. The Foster & Rohling link is a dead end.

    0 0
    Moderator Response: (Rob P) Fixed thanks.
  4. But the cached version does appear.

    0 0
  5. Kevin:

    Short answer: No, no overreaching. Richard Alley's characterization is correct. Watch his video, read the link in the OP on Milankovitch cycles, and read <a href="">this paper</a>.

    Long answer:

    WIth regards to the specific timeframe you mention, the Pleistocene, and its variation between glacial periods (massive ice sheets covering large segements of North America & Eurasia - the stereotypical Ice Age) and interglacials (ice largely confined to polar & alpine areas), orbital wobbles which alter the amount of summertime sunlight reaching mid-high latitudes in the Northern Hemisphere (around 65°N).

    When cooling into a glacial period, the wobbles act so as to reduce summertime solar insolation at this latitude, allowing snow to persist for longer and reflect more sunlight (snow having a greater reflectivity than the underlying terrain), which causes cooling. When warming up into an interglacial, the wobbles act so as to increase summertime solar insolation, starting the process of melting the continental ice sheets.

    What is critical to note is that the change in forcing in these orbital wobbles is, in and of itself, nowhere near significant enough to cause cooling sufficient to form continent-wide ice sheets, nor to cause warming sufficient to melt them.

    As noted in the OP, CO2 feedbacks then kick in, which themselves trigger additional feedbacks. The net effect of the feedbacks follows the direction of the initial change in forcing.

    Over the course of the Pleistocene, until the onset of the Industrial Revolution, CO2 concentration in the atmosphere can largely be attributed to this process, with some minor contribution from volcanism and sequestration from other geological processes.

    The critical point is that, as shown in the Lacis et al paper and the discussion of Pleistocene glaciation, CO2 is indispensible to the mediation of Earth system climate. Orbital wobbles have not always been an important forcing in climate (given the resolution studies of paleoclimate affords us), and solar forcing, while ultimately important, has to be mediated by greenhouse forcing in order for the Earth climate to support most life as we know it (some cold-resistant bacteria or other simple organisms notwithstanding).

    0 0
  6. Oops, forgot to work with WSIWYG when including links.

    Also, DSL basically posted a much more concise, therefore superior, version of what I did.

    0 0
  7. How about the first question?  Is fig 1a a cause and effect figure?

    0 0
  8. Kevin@7: "Foster & Rohling (2013) ... found a consistent and robust relationship between carbon dioxide and sea level"

    Correlation is not proof of causation.  The mechanism whereby high CO2 causes high sea level, however, is called the greenhouse effect, which neither this website nor any other should have to defend, at this point.

    0 0
  9. Key Point #4; ",... and the Earth is once more in equilibrium."

    At what point in graph 'A' was the earth in equilibrium?

    0 0
    Moderator Response: (Rob P)- The scale of the graph is not detailed enough to show this, but the 400-500 year-long Medieval Period is a good example. Sea level volume was static during Medieval time because it wasn't warm enough, in a global sense, to add noticeable glacial meltwater to the oceans, nor sufficiently cold enough to grow land-based ice.
  10. ubrew12,


    This is a peer reviewed paper showing that CO2 lags behind temp.

    We shouldn't have to argue that point!  So, your are saying that it is a correlation, not causation, but then you go on to mention that CO2 causes high sea level change via the greenhouse effect.  So, you are arguing that it is a causation relationship.

    That is complete nonsense!  The above paper shows that.


    0 0
  11. Kevin@10: Composer99@5 tried to inform you: through the greenhouse effect CO2 is a powerful feedback as well as forcer.  Claiming otherwise refutes the greenhouse effect.  We know that ice ages and interglacials are caused by orbital effects, but CO2 is a powerful feedback that enhances (i.e. causes) much of the enhanced response; the greenhouse effect demands it do so.  Hence, I can't claim that the historical rise and fall of sea level was caused by CO2, but I can claim that the AMOUNT of that response WAS caused by CO2 level fluctuations.  

    0 0
  12. ubrew12,

    but the paper I showed you shows that CO2 lags temp, that temp is the driver of CO2 levels, not the other way around.

    We know that ice ages and interglacials are caused by orbital effects,

    That was one of my first points, that it was somewhat ridiculous to call CO2 the big control knob, when the ice ages were caused by other factors.

    0 0
  13. A control knob is not a switch.  Alley is saying if you can control CO2, orbital effects will have little effect.  Go ahead and throw the switch, if the oven is turned down little is going to happen.  The mechanism whereby CO2 drives temperature can be proven in your garage.  The mechanism whereby temperature drives CO2 is speculative, involving ocean outgassing.  I believe it and so do you, but if you're going to get hung-up on causation, that should be your candidate.  Once CO2 levels increase, temp and sea level have little choice but to respond in kind, because of that thing you can prove in your garage.  So the lag you notice is evidence of a different switch, but the response has little choice but to reflect the control knob.  That the control knob can also be a switch shouldn't be hard to imagine.  In the last 150 years, thats been the case.

    0 0
  14. Kevin:

    CO2 lagging temperature applies where its emission by sources or absorption by sinks is temperature-dependent: CO2 emission from or absorption by the oceans, CO2 absorption by rock weathering, and similar processes.

    By contrast, CO2 leads temperature when its emission by sources is temperature-independent, such as emissions from volcanism (which has been a trigger for past warming, if not in the Pleistocene) and, since the start of the Industrial Revolution, the massive bolus of fossil carbon emitted by humans. (I am not personally aware of any temperature-independent carbon sinks, although perhaps the kinds of biological sinks that lead to the creation of fossil fuels count.)

    The correlation between sea level and CO2 is unsurprising because either the small orbital forcing (as in the Pleistocene) is magnified by the greenhouse effect, allowing sea levels to rise or fall in tandem with CO2 concentrations, or (as in the modern period, or perhaps warming out of Snowball Earth conditions when temperature-dependent absorption of CO2 is inhibited while vocanic emission continues apace) because CO2 emissions are driving both temperature changes and sea level changes.

    As far as I can see, the paper you link to explains in published form exactly what I was describing in #10. If you wish to pursue further discussion along the lines of "CO2 lags temperatures" I suggest this thread.

    0 0
  15. Should I repeat the concept for a fourth of fifth time?  CO2 is CO2.  It doesn't matter how it gets into the atmosphere.  If we put it there, it is considered a forcing.  If it gets there through ocean warming, it is considered a feedback.  That whole "CO2 lags temp" meme is one of the goofiest.

    And, again (because you did not read what I wrote the first time), orbital forcing starts the glaciation/deglaciation process.  Once the process begins, the initial orbital forcing is overwhelmed by feedbacks.  The primary feedbacks are, for glaciation, increasing ice/snow-albedo effect and decreasing ocean temp (increase in CO2 uptake).  The primary feedbacks for deglaciation are warming oceans (decrease in CO2 uptake) and decreasing ice/snow albedo effect.  Both processes are modified slightly by, respectively, associated changes in the biosphere (respectively, a decrease in biosphere uptake of CO2 and an increase in biosphere uptake of CO2).  Water vapor is a fast feedback and so doesn't enter into the conversation about climate-scale forcings and feedbacks. 

    0 0
  16. There's several points of confusion in the above comments, most of which miss the timescales being discussed in the paper, which (by and large) is a bit longer than orbital.  From a last ~60 million year perspective, CO2 is pretty clearly a big forcer of the gradual cooling we've seen over the Cenozoic and in the development of the great ice sheets.  You can say it's a "geologic feedback" in the sense that the CO2 responds to things like plate tectonics, mountain uplift, rate of organic carbon burial, etc  However you define a reference system, you can call something a feedback, but it doesn't change the physics that we don't have a way to get from the climate 60 million years ago to the climate today without changing the CO2.  On the glacial-interglacial timescales, you have to be careful, because the orbit is pretty clearly pacing the changes and the impact of CO2 becomes progressively more important as you move equatorward (e.g., in the CMIP5 models for example there is a more coherent negative correlation between model climate sensitivities and their tropical averaged LGM temperature anomalies, as opposed to the global mean doesn't really matter in this sense whether the orbit or the CO2 comes first, radiation still works...but the CO2 pretty clearly comes after Antarctic changes but before global-mean changes).  But the ice sheet changes can't be tied to just CO2, so I wouldn't out much stock in the paper on orbital timescales. 

    On the geologic timescales when you sample over lots of orbits, you still have to be careful applying the results from the paleo-record to the future, because there's almost certainly hysteresis in the ice sheets (e.g., whatever concentration of CO2 helped you glaciate Antarctica is almost certainly a different concentration of CO2 than you need to deglaciate Antarctica) and the rate of change probably matters too.  Some of these issues aren't well-sampled given the time resolution and (rather one-way) temperature evolution over the last 40 million years.

    0 0
  17. Kevin:


    "This is a peer reviewed paper showing that CO2 lags behind temp."

    And this, much more recent, just published in Science, shows that actually it hardly does at all:


    Parrenin, F. V. Masson-Delmotte, P. Köhler, D. Raynaud, D. Paillard, J. Schwander, C. Barbante, A. Landais, A. Wegner and J. Jouzel, 'Synchronous changes of atmospheric CO₂ and Antarctic temperature during the last deglacial warming', Science, 339, 1060-1063, 2013. 



    It would be a good paper for SkS to write about.


    Kevin, regarding the whole Milankovic cycles, CO2 feedback, etc, this is all so 1990s denialist argumentation that has been undermined by climate science for a very long time.  The paper I cite above compresses the time scale in which CO2 rises as a result of Milankovich-cycle forcing causing the earth to warm, which emphasizes that this rise leads to a feedback affect accelerating that warming.

    0 0
  18. The .pdf link is dead to me, and no cache is available. The supplemental info and data are available at the PNAS website.

    0 0
  19. First, I want to commend Rob Painting on an excellent article.  I particularly appreciated the clear statement of the basis for the conclusions, along with the relevant caveats.  Could a link to this article be placed at the bottom of the various sea level myth rebutals so that it is not lost in the continuous stream of posts at SkS, and left to languish in obscurity.

    In particular, a link would be appropriate on these pages:

    Sea level rise predictions;

    How much is sea level rising; and

    Why Greenland's ice loss matters.


    0 0
  20. What Tom Curtis said: Seconded. While it's good to debunk denialist articles individually too, the continous stream of crap, illogical presentations and outright lies by the usual suspects may effect also good sites. Luckily, the article refuting Goddard etc. types usually contain their name on the headline so I can skip the article. I mean, I have decided their originals (and comments therein) aren't worth the bandwith, but still someone might still believe then so the debunkings should continue, however frustrating it is. I'd like to call it 'The propaganda-machine of deniers', but I'm still not certain, they might just be assholes having fun at the expense of future generations. (oops, off topic, please delete.)

    0 0
  21. Kevin #12,

    The paper you cite was published in 1999.

    I suggest it has been superseded by these papers published much more recently:

    From the second:

     Here we propose a revised relative age scale for the concentration of atmospheric CO2 and Antarctic temperature for the last deglacial warming, using data from five Antarctic ice cores. We infer the phasing between CO2 concentration and Antarctic temperature at four times when their trends change abruptly. We find no significant asynchrony between them, indicating that Antarctic temperature did not begin to rise hundreds of years before the concentration of atmospheric CO2, as has been suggested by earlier studies.

    0 0
  22. A few points (in addition to seconding or thirding the Tom Curtis motion@19):

    1. I am unhappy with Ally's calling CO2 simply earth's main control knob. We encounter lots of control knobs in our everyday lives (or at least we used to in the predominantly analog era): volume control knob, screen brightness control knob, etc. These control knobs work virtually instantaneously. We have had very few control knobs that take significant time to see the effect of turning the knob. The most common examples (perhaps the only examples familiar to most people) is a thermostat: the thermostat for controlling home temperature, the thermostat in the refrigerator, the thermostat for the oven in the kitchen. I wish Ally had originally called CO2 earth's main thermostat.

    2. In the older Antarctic ice core data, while CO2 lagged temperature, CH4 was leading by more than CO2 was lagging. With the recent indication that the relative age of the air bubbles and the surrounding ice in the Antarctic core data need adjusting, the CH4 lead would increase as the CO2 lag decreases. As is known, CH4 is converted to CO2 and H2O in the atmosphere on the time scale of a decade. Both CH4 and H2O are greenhouse gases, but their residence times in the atmosphere are relatively short. The residence time of CO2 is on the order of a millenium, which is why it is the main greenhouse gas for long-term climate control.

    3. The atmospheric CO2 concentration does not directly control the sea level. Sea level is controlled by the amount of water in the oceans and the temperature of that water (water expands when it warms). When water is stored on land as ice, that water is not running quickly to the sea.. The causal chain is thus: high atmospheric CO2 results in increased temperature, which means warmer oceans and less land ice (more water ran into the sea), and consequently a higher sea level. On the other hand, low  atmospheric CO2 results in decreased temperature, which means cooler oceans and more land ice, and consequently a lower sea level.

    0 0
  23. BillEverett:

    With regards to your point (1), if memory serves (it has been some time since I have watched the video), I do not recall Alley referring to CO2 as a control knob in his speech (although it is referred to as such in the title of the presentation).

    He does, I am much more certain, note that the combination of volcanic CO2 emissions and rock weathering of CO2, taken together, regulate the climate in a manner similar to a thermostat (the word 'thermostat' or some synonym is used).

    Since the term 'control knob' is used prominently in the title of the presentation, your criticism still stands, as far as I can see.

    0 0
  24. I'll have a shot at explainging this to Kevin.  The climate has two positive feedback mechanisms of (usually, hopefully) limited range -- albedo, and greenhouse gasses.  Shrinking ice sheets increase heat absorption leads to more shrinking ice sheets leads to more heat absorption -- but they cannot shrink past zero.  There appears to be a mechanism where increase GHGs (CO2 and CH4) lead to increased warmth leads to increased GHGs -- perhaps because of reduced solubility in the ocean, perhaps because of methane hydrate releases.

    The initial kick that usually leads to change, either cooling or heating, does not need to be large -- if it is large enough to push either feedback mechanism into action, the feedback takes over, and then the change (ahem) snowballs.  This is why, in the case of the natural warming out of ice ages, you see the CO2 lagging the change -- the initial kick out of equilibrium is not large, and precedes the first increase in CO2.  But that first increase in CO2 leads to more warming -- which precedes the next increase in CO2.  And the next increase leads to more warming, etc.  This is consistent both with (1) the lag you see written about and (2) claims that CO2 is the cause of most of the warming.  You might argue that really this is all about ice caps and albedo, and that would be reasonable, except that we already know that CO2 is a greenhouse gas of no small power (else we would not be here to have this discussion, because the planet would be far colder, even colder than it is during ice ages).

    The reason, now, that we see CO2 rises not lagging temperature increases is because the CO2 increase is both artificial and rapid -- the earth (in particular, the oceans) are slow to warm, and we have also outrun any historically observed feedback effects.  

    0 0
  25. shoymore,

    There are others, and more recent, but it doesn't much matter.



    all well and good.  Except for the fact that the glaciation graphs tend to be saw toothed in shape, not sinusoidal. 

    1.  What causes this regularity?  With the built in pauses?

    2.  More importantly, what causes the sharp drop off?  Both CO2 and sea level drop rather drastically.  Apparently, this is natural, as it repeatedly happens again and again.  Why won't it happen again? 

    0 0
  26. I've been looking at sea level rise maps. At 24 meters Bangladesh and the Netherlands essentially cease to exist... along with various island nations, the U.S. state of Delaware, half of Florida, and most of the major cities on the planet. We'd better hope this study is wrong, actual rise is on the low end of the uncertainty range, or the timescale is in the thousands of years, because even a 'several hundred years' timescale for this kind of sea level rise could be economically devastating.

    0 0
  27. Kevin, the ice ages are generally understood to be initiated by changes in the Earth's orbit (i.e. Milankovic cycles), which explains the regularity (point 1).  Note very carefully that I said "initiated", rather than "caused".  This is because the changes in insolation caused by the variation in orbit is not sufficient to cause the observed change in the Earths temperature, so we know that there must be some feedback mechanism that amplifies the effect of the Milankovic cycles.

    One of these feedback mechanisms is the carbon cycle.  As the world warms up, the oceans release CO2 into the atmosphere, which is a greenhouse gas, which causes a bit more warming, so a bit more CO2 is released and so on.  Fortunately there are also negative feedbacks (principally the Stefan-Boltzman law), so this process doesn't continue forever.

    Now before we go onto more advanced issues, please can you explicitly state whether you agree with this, and if you don't agree, please explain why not.

    0 0
  28. Kevin, yes in the past CO2 levels have naturally gone up to ~280 ppm, dropped down to ~180 ppm, and repeated over and over again.

    You ask, "Why won't it happen again?"

    We're at ~400 ppm CO2 and climbing. Completely outside the bounds of that 'natural cycle'. The natural ~100 ppm CO2 drop from that cycle would normally occur over the course of next few tens of thousands of years... we have already increased CO2 by more than that 100 ppm and are still putting enough in to offset the entire decline every fifty years.

    So, basically... the phase of the Milankovitch cycle which in the past has caused CO2 levels to drop will 'happen again'. It will just be insignificant in comparison to the rate at which we are increasing CO2 levels.

    0 0
  29. BillEverett said... 'I am unhappy with Ally's calling CO2 simply earth's main control knob."

    Actually, Dr Alley calls CO2 the "Biggest Control Knob."  That would be very different than saying it's the "main" control knob.  

    I have somewhat of an enhanced version of this analogy.  You can maybe think of the whole system as a very complex and sloppy set of gears and pulleys.  A forcing is something like solar variation or orbital changes that are actively turning one or more of the gears, and those gears are in turn driving a broad series of other gears and pulleys.  Each of those turns at different rates and has a net effect on global temp.  CO2 happens to be just one of the cogs in the system that has the largest effect on global temp.  Even when a separate forcing is doing the work, that forcing is pushing the "big knob" (CO2) to have the largest effect.

    The difference today is, we are directly moving that CO2 knob ourselves.  It's still a sloppy system of gears and pulleys, and thus there is some delay in the net effect, as well as uncertainties with what the exact effect will be.  But there is no doubt, based on our understanding of how the system works, that we're fiddling with the big knob.

    0 0
  30. Curiously the British Court that defended Al Gore's film found the opposite to be true...

    Among the inaccuracies was this:The film suggests that evidence from ice cores proves that rising CO2 causes temperature increases over 650,000 years. The Court found that the film was misleading: over that period the rises in CO2 lagged behind the temperature rises by 800-2000 years. (among numerous citations!)

    0 0
  31. Yes, they did.  Now, what does it mean to you?

    0 0
  32. laughinchance - a lot of the change in temperature was caused by CO2 and CO2 changes lag temperature changes by 800+ years.  There is no contradiction there, it is just what you would expect if the natural carbon cycle provided positive feedback amplifying the effects of orbital variation.

    0 0
  33. LaughinChance, note Parrenin et al. (2013).  Abstract (my emphasis): 

    "Understanding the role of atmospheric CO2 during past climate changes requires clear knowledge of how it varies in time relative to temperature. Antarctic ice cores preserve highly resolved records of atmospheric CO2 and Antarctic temperature for the past 800,000 years. Here we propose a revised relative age scale for the concentration of atmospheric CO2 and Antarctic temperature for the last deglacial warming, using data from five Antarctic ice cores. We infer the phasing between CO2 concentration and Antarctic temperature at four times when their trends change abruptly. We find no significant asynchrony between them, indicating that Antarctic temperature did not begin to rise hundreds of years before the concentration of atmospheric CO2, as has been suggested by earlier studies."

    So who ya gonna believe?  Al Gore on video?  British court responding to a graph? Flying Spaghetti Monster?  Science?

    Or are you suggesting that CO2 can only be a feedback; when we put it into the atmosphere, it's radiatively inert?

    0 0
  34. What is sinusoidal is the game of whack-a-mole on rebunked talking points. 

    0 0
  35. LaughinChance @ 30...  It's interesting that your link goes to a website called "Global Warming Hoax."  And then when you click the link to locate the source of the statement, the link is dead.

    Not winning any style points on that one.

    0 0
  36. Kevin:

    With regards to your inquiries:

    [...] Except for the fact that the glaciation graphs tend to be saw toothed in shape, not sinusoidal.

    1. What causes this regularity? With the built in pauses?

    2. More importantly, what causes the sharp drop off? Both CO2 and sea level drop rather drastically. Apparently, this is natural, as it repeatedly happens again and again. Why won't it happen again?

    First, I should like to draw your attention to the fact that the graph in figure 1(a) goes backwards in time as you go from left to right. So what you are actually seeing is that CO2 and sea levels are dropping very gradually, with brief interludes and irregularities, and then rise very drastically.

    With that in mind, the answer to your question #1 is that it takes much longer to form the massive North Hemisphere ice sheets than it does to melt them.

    Another factor to consider is that there are several different sets of orbital changes (or wobbles as I have called them upthread) and these have different periodicity, which can accelerate or decelerate climate changes (when these are driven by orbital changes, unlike in the present).

    Your question #2 has already been responded to.

    0 0
  37. I'm surprised nobody has so far mentioned Shakun et al. (2012). This study explores the process by which increased summer insolation at high NH latitudes initates deglaciation and entrains powerful positive feedbacks that ultimately result in full glacial termination. There is an excellent article about S12 at RealClimate by Chris Colose (recently spotted in comments upthread) and also here at SkS.

    Very briefly, the process works like this:

    - By ~19ka, mid/high latitude NH temperature increase causes sufficient melt from NH ice sheets for freshwater flux to inhibit NADW formation and halt AMOC

    - NH *cools* as equatorial -> poleward heat transport stops

    - With the NH ‘heat sink’ turned off, the SH *warms*, as it must

    - Deep water warming in SH causes release of carbon from ocean sediments. This strongly positive feedback globalises and amplifies the warming

    - NH melt resumes, fully engaging strongly positive ice albedo feedback

    - Deglaciation accelerates until largely complete by ~11.5ka. Holocene interglacial begins

    The supposed lag between GAT and Antarctic CO2 disappears. Along with a pseudo-sceptic 'argument'.

    0 0
  38. Dikran,

    Now before we go onto more advanced issues, please can you explicitly state whether you agree with this, and if you don't agree, please explain why not.

    Yes and no.  Terrible answer, but I'll go on.  I agree with the initiation aspect, although I attribute more than just an initiation here, substantial "forcing" as well.  I even agree, up to a point, with the feedback mechanism.  The main problem I have is this:

    The temp leads the CO2, the CO2 generates more heat trapping, the temp goes up...etc.  Yes, I understand the Boltzman limitations, however, there is still a disconnect.  On another thread here, the feedback loop for watervapor is equal to the temp increase from CO2 - for every temp increase due to CO2, you can expect the same temp increase from watervapor, and maybe some more from the other feedbacks - CO2 climate sensetivity - and since these temps and CO2 concentrations (for the glaciation graphs) are all lower than in the present when this sensetivity relationships were "calculated", they should still all hold up (probably even increase).  None of this yet disputes your scenario. Yet.

    Why does it end?  The CO2 lags the temp.  What caused the temp to drop?  Another timely astronomical event?  It can't be from "diminishing returns" of the CO2 feedback loop, as I stated above, these are all at lower than present values.  Since these ice ages all end, and end with regularity (the same regularity that they start with) there must be some other, regular, forcing component to end them.

    Since there is another component that starts them, and another that ends them, I will gladly point out that CO2 really can't be the big control knob.  Since it really is something else that starts and stops them, it is not too much of a strech to find that this something else also controls or dictates the magnitude of these changes.


    Also, on another thread, it is stated that the removal of CO2 from the atmosphere is a worthwhile endeavor because the decline in temp is the same as the incline has been.  Looking at the graphs in fig 1a, I would argue that is not accurate.  If the rise and fall of the temps and CO2 is from the CO2 feedback loops, then why is the fall so different then the rise?  Clearly, there is a different mechanism at play.

    0 0
  39. Kevin, I don't see any dissconet there with regard to water vapour.  The principal difference is that CO2 is a long-lived GHG, whereas water vapour is a greenhouse gas, but its "adjustment time" in the atmosphere is a matter of days, so it only has a lasting effect on the atmosphere if it is amplifying something else that does (e.g. CO2).

    "What caused the temp to drop?" largely Milankovic cycles, as has already been pointed out.  Orbital cycles are sometimes favourable to initiating a glaciation, sometimes unfavourable.

    I would also agree that CO2 in paleoclimate has been mostly a feedback, hence it has been a thermostat, rather than a "control knob", although there have been events where it has initiated warming rather than merely amplifying it (e.g. end of "snowball Earth").

    However CO2 as well as being a feedback can also be a forcing (which is what anthropogenic emissions are doing), hence it can also be a "control knob" that we can use to set the thermostat.  However in my opinion, going on about knobs or thermostats is missing the point.  Paleoclimate shows that CO2 has a substantial effect on climate, hence if we double the atmospheric CO2 concentration (or more) we should expect it to have a substantial effect.

    The difference in slope between the start of  a glaciation and the end has already been explained by composer99 above.

    0 0
  40. BTW Kevin, if you want to get to the bottom of something, the fastest way to do so is to discuss one issue at a time, or ask one question, or make one point.  If you branch off into three or four points in every post it more or less guarantees that many of them won't get answered.  Indeed a common tactic in rehtorical debate is to do just this precisely in order to prevent getting to the bottom of the key issue.  If someone is genuinely interested in scientific truth, a good indication is generally that they will keep to the point and narrow the discussion to ever more focussed questions that identify the key issues.

    0 0
  41. @1.Kevin " they just happen to go up and down together, perhaps each caused by something else..". It's known as hard physics that CO2x2 will force back ~3.5Wm-2 (skeptic Lindzen says 3.4). By simple math that power would raise ocean temp. to full depth by 1 C in 147 years with no other factors. Ocean mixing to depth takes ~1,000 years so it will be less than 147 years for top, say, 1000m. So you ask whether something known to be enough to melt massive ice was simply a coincidence in time with some mysterious other thing never seen (that you cannot suggest yourself) that does precisely the same. I'm sure not. What think you ? 

    0 0
  42. Kevin...  Just to be completely clear here, even if CO2 does lag (noting the most recent research suggesting there is no lag), that doesn't in any way cast doubt on the role of CO2 as a greenhouse gases, nor does it cast doubt on the warming effects of man-made CO2 from burning FF's.  It's a complete non-issue with regards to published research.

    0 0
  43. Composer99,

    First, I should like to draw your attention to the fact that the graph in figure 1(a) goes backwards in time as you go from left to right. So what you are actually seeing is that CO2 and sea levels are dropping very gradually, with brief interludes and irregularities, and then rise very drastically.

    Thanks, I am used to seeing this the other way around, and did think it looked odd.  But that doesn't counter my points.

    Carbon Dioxide the Dominant Control on Global Temperature and Sea Level Over the Last 40 Million Years

    Just pointing out that the title of this article can't be accurate.  As I pointed out, and everybody agrees, CO2 didn't start or end any of the ice ages.  I also pointed out 1 paper for, and yes, someone pointed out one against, temp leading CO2.  If temp leads CO2, clearly, CO2 can't be dominate control of temp!

    Sorry about the multiple points, but they are all so closely related.

    0 0
    Moderator Response: [RH] Fixed image width.
  44. Kevin, there have been other occasions where CO2 has led temperature changes (e.g. uplift of Appalacians leading to increased chemical weathering, reduction of atmospheric CO2 and cooling temperatures), and there is also the fact that if CO2 is released by anthropogenic emissions then obviously any temperature change will lag emissions.  Sorry, but you are just engaging in pedantry as far as I can see.  Viewing CO2 as a thermostat is misleading as it can also be a forcing, calling it a control knob likewise doesn't tell the whole story as it often acts as a feedback.  This pedantry seems to be aimed at avoiding the key point of the story, which is that our anthropogenic emissions are twisting the control knob on the thermostat, and paleoclimate tells us that there is good reason to expect that to have a significant effect.

    0 0
  45. Kevin said... " If temp leads CO2, clearly, CO2 can't be dominate control of temp!"

    This is absolutely wrong.  I think you're getting twisted up in the semantics of the analogies and thus missing the point of the actual published research.

    Think of it as gears.  CO2 is the cog in the system that causes the greatest amplification on a given forcing.  It doesn't matter if there is a delay in the response or not.    

    What we are doing by burning fossil fuels is taking what is normally a CO2 response, and we're turning into a forcing.

    0 0
  46. Sorry, but you are just engaging in pedantry as far as I can see.


    ( snipped)

    I did not make any claims as to the last 100 years.  This article is talking about the past 40 million, so any discussion about anthropogenic CO2 is off topic.

    0 0
    Moderator Response: (Rob P) - sloganeering snipped.
  47. Kevin @ 43

    It is widely accepted that the gradual reduction in CO2 levels over the ~50Ma since the Eocene Optimum caused a very gradual cooling *within which* many different climate states occurred (Zachos et al. 2008).

    This is how CO2 and sea levels can broadly correlate over the last ~40Ma.


    0 0
  48. Kevin @ 38

    Looking at the graphs in fig 1a, I would argue that is not accurate.  If the rise and fall of the temps and CO2 is from the CO2 feedback loops, then why is the fall so different then the rise?  Clearly, there is a different mechanism at play.

    Please see 37. Perhaps slowing down and reading some of the responses and links might be a good idea at this point.

    0 0
  49. Kevin @ 46...  This is exactly what I'm talking about with regards to the semantics.  You are interpretting this to mean that CO2 has been a forcing over the past 40my.  No one is making that claim.  

    What is being said is that CO2 is the biggest "control knob."  That doesn't mean that CO2 is controling global temperature.  It means that it is the mechanism that causes the biggest amplification on forcing.

    0 0
  50. Kevin @ 38

    Looking at the graphs in fig 1a, I would argue that is not accurate.  If the rise and fall of the temps and CO2 is from the CO2 feedback loops, then why is the fall so different then the rise?  Clearly, there is a different mechanism at play.

    Please see 37. Perhaps slowing down and reading some of the responses and links might be a good idea at this point.

    0 0

1  2  Next

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

The Consensus Project Website


(free to republish)

© Copyright 2024 John Cook
Home | Translations | About Us | Privacy | Contact Us