RealClimate logo

How long will global warming last?

Filed under: — david @ 15 March 2005 - (Français)

Guest commentary from David Archer (U. Chicago)

The notion is pervasive in the popular and scientific literature that the lifetime of anthropogenic CO2 released to the atmosphere is some fuzzy number measured most conveniently in decades or centuries. The reality is that the CO2 from a gallon out of every tank of gas will continue to affect climate for tens and even hundreds of thousands of years into the future.

The U.S. Environmental Protection Agency Inventory of U.S. Greenhouse Gas Emissions and Sinks (2005) has the CO2 lifetime listed as 5-200 years, for example [1]. I have seen “hundreds of years” in scientific manuscripts and in environmentalist literature. David Goodstein in his excellent book The End of the Age of Oil states, “If we were to suddenly stop burning fossil fuel, the natural carbon cycle would probably restore the previous concentration in a thousand years or so.” I assume that Goodstein is conservatively applying several century-long e-folding times to derive his thousand years, but he implicitly assumes that the CO2 will relax toward its 1750 concentration. The point is that it does not.

When you release a slug of new CO2 into the atmosphere, dissolution in the ocean gets rid of about three quarters of it, more or less, depending on how much is released. The rest has to await neutralization by reaction with CaCO3 or igneous rocks on land and in the ocean [2-6]. These rock reactions also restore the pH of the ocean from the CO2 acid spike. My model indicates that about 7% of carbon released today will still be in the atmosphere in 100,000 years [7]. I calculate a mean lifetime, from the sum of all the processes, of about 30,000 years. That’s a deceptive number, because it is so strongly influenced by the immense longevity of that long tail. If one is forced to simplify reality into a single number for popular discussion, several hundred years is a sensible number to choose, because it tells three-quarters of the story, and the part of the story which applies to our own lifetimes.

However, the long tail is a lot of baby to throw out in the name of bath-time simplicity. Major ice sheets, in particular in Greenland [8], ocean methane clathrate deposits [9], and future evolution of glacial/interglacial cycles [10] might be affected by that long tail. A better shorthand for public discussion might be that CO2 sticks around for hundreds of years, plus 25% that sticks around forever.

The sticking-around-forever idea is not new, and the picture has not changed by very much since the effect was first predicted back in 1992 [2]. You can estimate the magnitude of the effect pretty well just using CO2 thermodynamics and the back of an envelope. It could be argued (by someone with a cruel heart) that since we don’t understand why CO2 was lower during the last ice age, we ought not go around making forecasts for the future. Well, OK, but I would point out that CO2 in the past appears to act as an amplifier for orbitally forced climate change, so if anything, we might expect the carbon cycle in the future to amplify our own climate forcing, rather than counteract it. If the past is any guide, CO2 surprises in the future, in the long run, seem unlikely to help us out.

A long lifetime for CO2 adjustment is also consistent with an isotopic event in the deep sea sedimentary record from 55 million years ago, the Paleocene/Eocene Thermal Maximum event. The record tells the story of the sudden release of an isotopically light source of carbon, triggering a fast warming in the deep sea of about 5 degrees C. Both the carbon isotope signal and the temperature (inferred from oxygen isotopes) then relaxed back toward their initial values in about 100,000 years. If the released carbon were initially in the form of methane, it would have been oxidized to CO2 within a few decades, but as CO2 it apparently stuck around, warming the deep ocean, for a long time before it went away.

The shortest lifetime estimates, such as EPA’s 5-years, derive from the exchange flux of CO2 between the atmosphere and ocean, which is about 200 Gt C/year (1 Gt C is 1012 kg of carbon) in each direction. Because the exchange flux is back-and-forth, it has nothing to do with the net uptake by the ocean of new CO2 to the system, which relies on the imbalance between the upward and downward exchange fluxes. That imbalance is only about 2 Gt C/year.

Even the present-day net flux tends to underestimate the real lifetime of global warming. The atmosphere contains about 160 Gt more carbon than it did then. If we divide this number by the CO2 invasion flux into the ocean of 2 Gt C/year, we get an apparent uptake time scale of 80 years. This result is shorter than model air/water equilibration time scales by a factor of four or so. I believe the problem is with the simple calculation. The CO2 concentration of the atmosphere is going up continuously, and so it invades the ocean as it equilibrates with warm surface waters. If atmospheric CO2 were not going up, the warm surface waters would saturate in a year or two, the overall ocean invasion rate would decrease, and the lifetime estimates by this method would increase. Different parts of the ocean equilibrate with the atmosphere on different time scales, ranging from a year for the tropical surface ocean to a millennium for the deep sea. Overall, model experiments show a CO2 equilibration time of a few centuries [5, 6, 11, 12]. The other problem with both of these conceptions is that they implicitly assure us that the CO2 concentration is going back to its initial concentration, which it will not.

Another source of short-lifetime bias in the community probably comes from a calculation used to compare the greenhouse consequences of different gases, called the Global Warming Potential (GWP) [13]. Some trace gases such as methane have a stronger impact on the heat balance of the earth, per molecule, than CO2 does. However, to really compare them fairly one might want to factor in the fact that methane only lives about 10 years before it goes away (actually, it is oxidized to CO2, another greenhouse gas, but it is common to ignore that in GWP calculatons). Global warming potentials are calculated by integrating the radiative energy impact of a molecule of gas over its atmospheric lifetime. However, if the full lifetime of CO2 were considered, including that long tail, then methane would be by that calculation unimportant. On human time scales, methane is certainly an important greenhouse gas, and so what’s done is to arbitrarily limit the time horizon of the calculation to something like human timescales. Methane GWP is higher when considered on the 50-year time horizon than it is on the 500-year time horizon or it would be on a 500,000-year time horizon, if anyone bothered to do that calculation. Perhaps the adoption of time horizons for GWP calculations conditions scientists to believe that CO2 only persists for as long as this time horizon lasts. The table in the EPA document, for example, was associated with a discussion of global warming potentials.

It could also be that we-who-only-live-to-be-77.2-years-old don’t want to worry about climate impacts from fossil fuel CO2 release 100,000 or even 1,000 years from now. That would be a perfectly rational position, and I have no argument with it. Climate change negotiations are grounded in IPCC projections and scenarios to the year 2100, a far cry from the year 100,000, but even 2100 seems almost unimaginably remote given the pace of social and technological change in the world today. On the other hand, nuclear waste lasts for millions of years for some isotopes such as iodine 129. The public seems to find this information relevant, so the true longevity of anthropogenic climate change might be considered by some to be relevant to here-and-now decisions as well. At any rate, the facts as reported ought to be accurate, rather than judging in advance that no one cares about climate impacts that last thousands of years and more into the future.

  1. (with the subscript “No single lifetime can be defined for CO2 because of the different rates of uptake by different removal processes”) (EPA (2005), Inventory of U.S. Greenhouse Gas Emissions and Sinks Draft Report: 1990 -2003, U.S. Environmental Protection Agency, Office of Atmospheric Programs
  2. Walker, J.C.G. and J.F. Kasting, Effects of fuel and forest conservation on future
    levels of atmospheric carbon dioxide. Palaeogeography, Palaeoclimatology, Palaeoecology (Global and Planetary Change Section), 1992. 97, 151-189.
  3. Joos, F., et al., An efficient and accurate representation of complex oceanic and biospheric models of anthropogenic carbon uptake. Tellus, Ser. B, 1996. 48: p. 397-416.
  4. Jain, A.K., et al., Distribution of radiocarbon as a test of global carbon cycle models. Global Biogeochem. Cycles, 1995. 9: p. 153-166.
  5. Archer, D., H. Kheshgi, and E. Maier-Riemer, Multiple timescales for neutralization of fossil fuel CO2. Geophys. Res. Letters, 1997. 24: p. 405-408.
  6. Archer, D., H. Kheshgi, and E. Maier-Reimer, Dynamics of fossil fuel CO2 neutralization by marine CaCO3. Global Biogeochem. Cycles, 1998. 12: p. 259-276.
  7. Archer, D., Fate of fossil-fuel CO2 in geologic time. J. Geophys. Res. Oceans, in press.
  8. Huybrechts, P. and J. De Wolde, The dynamic response of the Greenland and Antarctic ice sheets to multiple-centure climatic warming. J. Climate, 1999. 12: p. 2169-2188.
  9. Archer, D.E. and B. Buffett, Time-dependent response of the global ocean clathrate reservoir to climatic and anthropogenic forcing. Geochem., Geophys., Geosys., 2005. 6(3): p. doi: 10.1029/2004GC000854.
  10. Archer, D. and A. Ganapolski, A movable trigger: Fossil fuel CO2 and the onset of the next glaciation. Geochem., Geophys., Geosys., in press.
  11. Sarmiento, J.L., U. Siegenthaler, and J.C. Orr, A perturbation simulation of CO2 uptake in an ocean general circulation model. J. Geophys. Res., 1992. 97: p. 3621-3645.
  12. Sarmiento, J.L. and C.L. Quéré, Oceanic carbon dioxide uptake in a model of century-scale global warming. Science, 1996. 274: p. 1346-1350.
  13. Jain, A.K., et al., Radiative forcings and global warming potentials of 39 greenhouse gases. J. Geophysical Res., 2000. 105(D16): p. 20,773-20,790.

49 Responses to “How long will global warming last?”

  1. 1
    asdf says:

    I know it’s not as well understood — but what is the approximate time frame over which CO2 values returned to normal after the Permian/Triassic mass extinction event and the catastrophic deglaciations in the late Precambrian?

    I seem to remember estimates of around 3 million years for the Scythian period.

    Some folks at NASA recently discussed the use of a fluorochlorocarbons to “terraform” Mars. I assume the half lives of these compounds are closer to that of methane — and yet the effects are supposed to be ~ permanent. The message I take home is that some folks in the government believe in positive feedbacks and climate instability.

  2. 2
    Lynn Vincentnathan says:

    I do care about very long term effects, because my concern regarding GW from the very beginning was how I was impacting others, including those in the future, not what problems I may face personally within my lifetime. I do think this is the concern of most environmentalists, and most people with a developed and mature conscience. That seems to be the concern behind the nuclear waste, as well.

    Thanks for the information. It makes each little reduction of GHGs all that more important.

  3. 3
    John Davis says:

    My understanding was that about 60% of the additional carbon we pump into the atmosphere each year is immediately sequestered by natural processes, presumably largely biological. Intuitively this seems very hard to reconcile with the sort of geological timescales discussed above.

  4. 4

    A quick read of your paper (reference 7) brought up some questions.

    The statement is made that “Atmospheric pCO2 approaches equilibrium on a time scale of ~300yr”. In none of your model scenarios involving realistic near-term Carbon releases (1000 gigatonnes or less, based on the consideration that the 1990-1999 release is estimated in IPCC TAR to be 6.3 gigatonnes per year) does more than 20% of the injected CO2 remain in the atmosphere for 1000 years. It is based on these considerations that reasonable people consider the bulk of the CO2 to have a century scale life cycle in the atmosphere.

    Your model seems not to take into account the terrestrial sinks for CO2. When your model neglects the Calcium Carbonate processes and silicate weathering, it predicts that there will be *NO* change in atmospheric CO2 after the ~300y ocean uptake is complete. Is this a reasonable assumption?

    While it is a very important point for the lay person to know that the acidification of the ocean by CO2 (it combines with water to produce dilute Carbonic Acid) can reduce the effectiveness of the Calcium Carbonate processes at sequestering Carbon (and can even reverse it, by dissolving Calcium Carbonate), your model chemistry seems quite simplistic. No mention is made of the myriad of other processes (some of which operate on short time scales) whereby the acidity of the ocean is neutralized without involving CaCO3. How robust is your ocean chemistry model?

    Your finding of a “long tail” to the remaining 7 to 20% of the CO2 seems to hinge entirely on the ocean-acidity/CaCO3 argument and a neglect of other chemical and terrestrial sequestration processes which operate on short time scales.

    Finally I found a few typographical errors in the paper, which I hope your editors will find, e.g. “start contrast” (4th line from bottom, P. 11).

  5. 5
    Dian Deevey says:

    I am trying to critique various proposals for using CO2 uptake in forests to compensate for power-plant emissions of fossil-fuel-derived CO2 and other GHG. Some emerging compliance regulations incorporate the idea that such forests need survive only 100 years to be effective (for these compensation purposes). Archer’s work suggests that the 100-yr rule is an artefact of the 1997 IPCC decision to confine attention to scenarios that end in 2100. Does the 2100 scenario endpoint have some legal basis in the Kyoto Protocol itself?

    Archer’s information seems to imply that sequestration “offsets” in forests, soils, and the like are pretty pontless, and that we ought to focus on reducing emissions, period. (Including, of course, the kinds of reductions achievable via changes in land use–changes that probably would include preservation of existing forests, no-till agriculture, and the like.)

    I’d appreciate feedback on this interpretation.

  6. 6
    Dano says:

    I recently ran across this, Dian, while looking up something else:

    Martinelli N 2004. Climate from dendrochronology: latest developments and results. Global and Planetary Change 40:1-2 pp. 129-139. [On a global scale, this supposed positive influence on wood production and forest regeneration was thought to have the possibility of balancing the CO2 increase by carbon sequestering through photosynthesis. On the contrary, tree-ring records indicate that, under recent climate warming, drought may have been an important factor in limiting carbon uptake in a large portion of the boreal forest, one of the planet’s major potential carbon sinks. If this limitation in growth due to drought stress is sustained, the future capacity of northern latitudes to sequester carbon may be less than currently expected.]

    It’s a conclusion they make as they studied why some researchers are having trouble using recent tree rings from certain spp. and microclimates in some areas.



  7. 7
    dave says:

    Very interesting. Please relate your post to this source The Oceanic Sink for Anthropogenic CO2 (Science, Vol. 305, 16 July 2004). The bottom line is — sorry for the long quote, but this is worth reading — that

    There is a potential for both positive and negative feedbacks between the ocean and atmosphere, including changes in both the physics (e.g., circulation, stratification) and biology (e.g., export production, calcification) of the ocean. These processes are still not well understood. On the time scales of decades to centuries, however, most of the known chemical feedbacks are positive. If the surface ocean pCO2 concentrations continue to increase in proportion with the atmospheric CO2 increase, a doubling of atmospheric CO2 from preindustrial levels will result in a 30% decrease in carbonate ion concentration and a 60% increase in hydrogen ion concentration. As the carbonate ion concentration decreases, the Revelle factor increases and the ocean’s ability to absorb more CO2 from the atmosphere is diminished. The impact of this acidification can already be observed today and could have ramifications for the biological feedbacks in the future. If indeed the net feedbacks are primarily positive, the required socioeconomic strategies to stabilize CO2 in the future will be much more stringent than in the absence of such feedbacks. Future studies of the carbon system in the oceans should be designed to identify and quantitatively assess these feedback mechanisms to provide input to models that will determine the ocean’s future role as a sink for anthropogenic CO2. [Emphasis added]

    In addition, in Table 1 of the referenced source, only about 50% of the estimated emissions from fossil fuels over the nearly 200 year period from 1800 to 1994 (given in petragrams Pg) is taken up by the oceans, disregarding estimated and highly uncertain source emissions from land use. If the low source estimate of these land use changes over that period is taken into account, the percentage taken up by the oceans decreases. Furthermore, the data shown in the same source show that ocean CO2 uptake as a percentage for the period 1980-1999 appears to be decreasing. Since the message I’m getting from the Archer post is that about 75% of atmospheric CO2 will be taken up by the oceans over an approximately 300 year period, given some baseline stabilization of pCO2 levels, I find it hard to reconcile his conclusions on these “short” time scales with the historical data.

  8. 8
    Ned Ford says:

    Several concepts may clear up some of the uncertainty. Science 4 JULY 2003 VOL 301 “Increase in the Export of Alkalinity from North America’s Largest River” Peter A. Raymond and Jonathan J. Cole offers a strong candidate for the “missing sink” in the form of surprisingly large riverine export of carbon, mostly mineralized. This article challenges the assumption that returning agricultural land to forest is consistently prudent. In some important cases the agricultural use may be sequestering more carbon than forest growth would. Most of this carbon, being mineralized, does not directly interact with the two-way exchange of carbon across the air/ocean surface. It may add as much as ten percent to the total of anthropogenic emissions that wind up in the ocean in the short term, and may join the total pool of mobilized carbon if ocean chemistry is sufficiently disrupted.

    In another article, probably in Science in the last three years, but I can’t place the reference at the moment, the rate of permanent fossilization of land-based plant material is defined as about 1% of anthropogenic emissions. This rises to about 2%, if the rate of fossilization of charcoal from forest fires is considered, and both of these rates are supported by the fossil record. Thus land-based plant sequestration should not be regarded as permanent. There is strong evidence that the net movement of carbon in and out of lang-based plants since the beginning of the industrial era has been into the atmosphere.

    Carbon sequestration through tree planting and better agricultural soil management practices has a large number of arguments in favor of it. It should be viewed as replacing carbon lost through previous harvest or soil mismanagement, and should NEVER be used as an argument to burn more fossil fuels. At this point we have absolutely zero evidence that any public policies have the capacity to produce a net gain of land based carbon biomass. The only evidence that leans that way, North American and European forest growth, is well within the limits of the known cyclical aspect of the global land-based carbon inventory.

    So it looks like about half of our fossil emissions are in the ocean and half are in the atmosphere. I’m basically in agreement with Dr. Archer’s assessment, although I think the greatest complication in his long-term view is whether we saturate the ocean to the point that when surface waters from the fossil era return to the surface some thousand years from now, they become a source, or will they still be a sink. That depends on how much we emit in the next several centuries, and exactly what happens to them in the depths of the ocean. If the answer to the second question is well understood I’d appreciate some direction to the science.

    I do also think that the emerging understanding of the effects of acidifying the ocean means that we really ought to try to avoid finding out.

  9. 9
    Thor Olson says:

    I don’t think humpty dumpty is going to be so easy to put back together again. Consider how fast China is building carbon burning power plants. How likely is it we have seen the peak of carbon emissions?

    It sure looks like few people can afford to care about climate impacts thousands of years and more in the future. Most people are too busy going about their lives and will wait until the wolf is at the door. In the next fifty years alone you will see if a lot of people can retire without the world economy falling apart. In a hundred years, the ethnic population of Italy and much of Europe may have shrunk dramatically due to low birth rates.

    Unless we have some really compelling economic solution, why wouldn’t the world community permit things to get so bad that the ice caps to melt and the oceans rise?

  10. 10
    Hans Erren says:

    As a backgrounder it is useful to look at the actual data for the last 45 years. Considering the observed growth of emisions and the observed growth in atmospheric co2 concentration, a sink saturation has not yet been observed, quite opposite the sink has increased from 1000 MtC/y in 1960 to 3000 MtC/y in 2000.

  11. 11
    Lee A. Arnold says:

    This is also interesting from the viewpoint of mental models and perceptual frameworks. Browsing the cognitive science journals online, I found a study suggesting that people regularly misapprehend the nature and effects of perturbation: Peter A. White, “Naive analysis of food web dynamics: a study of causal judgement about complex physical systems,” in Cognitive Science; abstract at:

  12. 12
    Ferdinand Engelbeen says:

    Another point: during the Cretaceous the level of CO2 in the atmosphere was many times current and temperatures were far higher. But that was the time that the forefathers of the current sea species (mainly planktonic coccoliths) made the white cliffs of Dover and thick carbonate deposits in many places. If (and only if) these species didn’t lose their ability to grow in warmer and more acidic, CO2-rich waters, would that not speedup carbonate deposit?

  13. 13
    David Archer says:

    In response to comment #1.

    It took on the order of 100 kyr for the oxygen and carbon isotopic signals in the ocean to return to their original steady-state values after the PETM. The response time of atmospheric pCO2 ought to be faster than the response time of carbon isotopes, but then it is unclear to me what would be keeping the temperature up, if not higher CO2. Gavin Schmidt pointed me to a source (Bains et al., Science, 1999) where he sees different recovery times between 13-C and 18-O. I think the data plotted in Zachos et al. (Science 2001) is clearer, and I don’t see any difference between the recovery times. Still some loose ends.

  14. 14
    David Archer says:

    In response to comment #3.

    You’re correct that natural processes take up half or more of the carbon we release each year. The ocean takes up some, with no biology involved, just chemistry, and forests apparently take up the rest. The other nearly-half, however, still accumulates.

    What if you were spending money from your checking account twice as fast as you earned it? Can you calculate how long it will take to get out of debt based on the ratio between income and outflow? Not really; they are different questions.

    We could cut emissions by half, and the CO2 would stop going up. That’s what IPCC stabilization scenario model results essentially showed, and it demonstrates how Kyoto was just a baby first step in what would have to be done to prevent CO2 from rising beyond some level. If this were done pCO2 in the atmosphere would sort of float at the present higher-than-natural level. We would still be adding carbon to the atmosphere / ocean system, and it would still theoretically take 10^5 years to get pCO2 back down to a natural interglacial value of 280 ppm.

  15. 15
    David Archer says:

    In response to Comment #4

    None of the IPCC projections thought that emissions would remain at 6 Gtons C / year. The business-as-usual predicts a release of 1600 Gton C by 2100. For the 300 Gton model run, 17% remained after 1 kyr, and for 1000 Gton, it was 19%. After 100,000 years, the atmospheric fraction was 7%. Your “reasonable people” are right, the bulk of the carbon goes away in centuries. But there is that long tail.

    The way the terrestrial biosphere fits into this is that, to bring CO2 back down to its preanthropogenic value, the biosphere would have to take up an amount equal to what we have emitted. The terrestrial biosphere is about 500 Gtons C, with soils another 1500 Gton C or so. It is reasonable to hope that the terrestrial biosphere will take up some hundreds of Gton C, but the land carbon reservoir would have to double if we wanted it to take up thousands of Gton C. That’s asking a lot, especially on a heavily populated planet. There are several mechanisms that govern terrestrial uptake of carbon. One is CO2 fertilization, another is the longer growing season. On the other hand, soil respiration rates increase as temperatures rise. Tropical soils don’t have much carbon in them for this reason.

    I don’t know of a myriad of processes that can neutralize CO2 in the ocean without CaCO3. Surface ocean pH values are going down, that’s a fact. The model neglects corals and shallow-water CaCO3 accumulation, so the time scale for neutralization could be faster. The final atmosphere / CaCO3 buffered ocean equilibrium value of 7% remaining in the atmosphere, however, is simple thermodynamics, unaffected by uncertainties in the kinetics of neutralization.

  16. 16
    David Archer says:

    In response to comment #5

    One hundred years makes perfect sense on a societal level. I don’t know anything about the legality of the 100 year time window for Kyoto negotiations, though.

    Growing trees that then release their carbon 100 years from now might reduce the size of the transient atmospheric high level, without changing the long climate tail. It is interesting to note that some of the carbon sequestration proposals are not much better in this regard. Pumping CO2 into the deep ocean lowers the atmospheric transient but not the long tail. I’ve heard a leakage rate of 0.1% per year tossed around as acceptable for terrestrial carbon sequestration; pumping CO2 down into the ground. This would give the storage a 1000 year lifetime, again without changing the long tail.

  17. 17
    David Archer says:

    In response to comment #10

    The sink has gone up because the atmospheric pCO2 has gone up; it’s farther from the equilibrium value with respect to the ocean. The same models that get the ocean sink right today also predict that not all of the carbon will go into the ocean (if the researcher runs the model that long). The existence of a significant sink today does not affect the conclusion of the long tail, nor even does the observation that the sink today is growing.

  18. 18
    David Archer says:

    In response to comment #7

    The ocean model I used does carbon chemistry, with plankton and gas exchange and all that. It does a pretty good job of mimicing the distributions of lots of geochemical and physical tracers. I expect that it should do a pretty good job of mimicing the “Revelle factor” change as CO2 goes up.

    The circulation field of that model is “frozen”, that is, it does not respond to changes in climate. It’s missing a potential positive feedback there, as the ocean might be stagnant for awhile, slowing CO2 invasion.

    I drive the mean ocean temperature using changes in atmopsheric CO2 with a time lag of 1000 years. I’m using a deep ocean “climate sensitivity” of 3 degrees C for doubling CO2, from Stouffer’s recent work. If that’s wrong in one direction or the other, the temperature feedback would be wrong. As is, it increases the long time-during fraction by about 20%.

    The rate of CaCO3 production by plankton is taken to be fixed in time, as is the rate of weathering on land and the rate of shallow-water deposition. These would probably be negative feedbacks, making the relaxation toward equilibrium somewhat faster than I got. This wouldn’t change the amount of the long-duration fraction.

    I didn’t model terrestrial carbon at all, but rather lumped it together, including terrestrial uptake, into net emissions. There is more fossil fuel carbon available than the size of the terrestrial biosphere, including soil carbon, so I didn’t bother with it. Small potatoes.

  19. 19
    Hans Erren says:

    Does anybody know where to find updates of global CO2 emissions?

    The official source stops in 2000…

  20. 20
    Dave Dardinger says:

    The article states that the up-down exchange fluxes only differ by about 2 gt/yr. But this exchange is primarily from a bulk flow of water. Thus if the carbonate concentration in the upper ocean water increases, the downward movement of carbon will increase and the difference will increase. Do you have any historical data which would indicate how the fluxes have changed over time? I know that figure 4 in climate change 1994 showed that there was actually a net upward movement of CO2 (100 GtC/yr up and 91.6 down) from water movement. It was only the net 10 GtC/yr downward from marine biota which produced a net downward flow of carbon. And if the surface concentration of carbonate increased by exchange with the atmosphere there’s lots of room in the ocean depths to store it before it would cause much difference in the concentration of carbonate in upward moving water. So I don’t think it’s fair to assume that the 2 GtC/yr will not change. It most likely will increase in step with increased carbonate concentration in the upper ocean.

  21. 21
    dave says:

    Re: #18

    Thanks for your response, David. You said

    The rate of CaCO3 production by plankton is taken to be fixed in time, as is the rate of weathering on land and the rate of shallow-water deposition. These would probably be negative feedbacks, making the relaxation toward equilibrium somewhat faster than I got. This wouldn’t change the amount of the long-duration fraction.

    In the Feely et. al. paper “Impact of Anthropogenic CO2 on the CaCO3 System in the Oceans” (Science 2004 305: 362-366), the conclusion is that CaCO3 production is decreasing as surface seawater pCO2 (acidity) rises. Although, as you state, the feedback appears to be negative, there is also some question raised in the Feely paper of what the effects will be on overall phytoplankton primary production. I wonder what the effects will be on this production on various timescales and what will happen to deep ocean overturning and the so-called “biological pump” on long timescales.

    By the way, there are two articles “The Climate Change Commitment” by Tom Wigley and “How Much More Global Warming and Sea Level Rise?” by Gerald Meehl et. al. in this week’s Science 18 March 2005. Here’s a summary from the magazine New Scientist Ocean heat store makes climate change inevitable. Both studies (which I have not seen) focus on warming in the pipeline and no doubt involve modelling of future ocean warming.

  22. 22

    Is anyone accounting for greenhouse gas emissions from volcanism? Will volcanism increase as global climate continues to warm? That will influence how long global warming lasts. It took the climate about 50 million years to cool after the PETM all things considered… except human activity.

  23. 23
    Jim Dukelow says:

    David Archer wrote:
    “On the other hand, nuclear waste lasts for millions of years for some isotopes such as iodine 129. The public seems to find this information relevant, so the true longevity of anthropogenic climate change might be considered relevant to here-and-nov decisions as well.”

    This throw-away comment reflects a mass-media conventional wisdom about radiation and associated health and safety issues. The flip side of a long half-life is that not many I-129 atoms are decaying right now. When I arrived at the Hanford Site 25 years ago, part of the folklore (expressed as a joke) was that the carcinogenic dose of I-129 was 28 pounds of it in the thyroid. This folklore is supported by research reported in grey literature and the peer-reviewed literature by Joe Soldat and his colleagues at the Pacific Northwest National Laboratory in the 60s and 70s. Basically the issues are that I-129 has a relatively “soft” beta decay (75 kev), a relative short biological half-life in the body, and an extremely low specific activity (that is, number of decay events per second per unit mass).

    The specific activity of I-129 is so low that it is relatively difficult to measure it in the environment (or was — 15-20 years ago the method of choice was to zap the sample with neutrons so as to transmute the I-129 into something that could be more easily measured.). I-129 does have significant value as a tracer of movement of fission products through the environment and the biosphere, since any of it you find was created in the last 60 years and the quantity is essentially unchanged in that time.

    Jim Dukelow

  24. 24
    Hans Erren says:

    The global anthropogenic emission data set ends in 2000.

    Does anybody now about recent updates?

  25. 25
    Rick Watkins says:

    Re.# 9
    Pertinent questions – perceptive comments. Surely global emissions can only rise as the ‘slash and burn’ economic model develops in the east. There is a lot of quibbling amongst countries as to who can/should/does emit what – the bottom line is global.
    I live in Denmark – a country with a carefully manicured image of environmental friendliness, accepted here and abroad. The country is, however, in almost complete denial on the subject of global warming – not only whether something should be done, but whether humans are involved at all. This scepticism permeates not only the population, but the media and institutions too. When a report on the subject does slip through it’s usually with a sceptical tone….

  26. 26
    David Archer says:

    In response to comment #12

    If we raise the CO2 concentration in the atmosphere to Cretaceous levels and hold it there for 10,000 years or so, the CaCO3 cycle in the ocean will restore the carbonate ion concentration back toward CaCO3 saturation. So the CaCO3-secreting critters that comprise the white cliffs of Dover were not living in an acid world at all. The acidity comes in the transient.

    Tangentially, Caldeira (Nature, 2003) points out that the change in atmospheric CO2 now and in the near future is much faster than natural changes from the ice core record. The CaCO3 cycle was able to keep up with the 10-kyr CO2 rise as the ice sheets melted, but it won’t be able to keep up with the 100-yr anthropogenic CO2 rise. So the ocean will be much more acidified than occured naturally within the ice core record (now 800 kyr).

  27. 27
    David Archer says:

    In response to comment #20

    The 2 Gton / yr net CO2 invasion into the ocean is driven by the rising atmospheric concentration, not by water flow or heat fluxes or anything like that. The preanthropogenic atmospheric CO2 concentration was stable, indicating that at that time the fluxes into and out of the ocean balanced. There were some locations, first among them the equatorial Pacific ocean, where CO2 outgassed, and other locations where CO2 invaded the oceans. The net flux, driven by the rising atmospheric CO2 concentration, may rise for a while, but it could also fall as the buffer capacity of the seawater is exhausted (see comment #7, for example.

  28. 28
    David Archer says:

    In response to comment #22

    Volcanoes are generally thought to be imperturbable outside drivers of climate; I don’t know of any possible feedback that could make volcanic emissions respond to changing climate.

    Volcanoes release CO2, both from metamorphic cooking of CaCO3 back into igneous rocks, and from juvenile carbon (mostly released at mid-ocean ridges). In the natural steady state, this release is thought to be balanced by weathering of igneous rocks (essentially the reaction looks like CaO + CO2 => CaCO3). Because the weathering of igneous rocks could very easily be climate-responsive (there is no weathering in the Antarctic dry valleys, for example), these processes combine to stabilize CO2 concentration and climate on time scales of hundreds of thousands of years. It is called the Silicate weathering thermostat, and has been extensively modeled by Berner and colleagues. Berner has a new book about this topic, pricey but elegant.

  29. 29
    Jeffrey Davis says:

    (not for publication)

    If you would have an acronym dictionary on the site, it would be great for your simply-interested/non-technical readers. Like me.


    [Response: See our glossary. -mike]

  30. 30
    Johan Svensson says:

    Skiping ahead, what if emissions are brought down to whatever it takes to keep it steady.
    What is the limit of global release to keep the balance at 280 ppm?
    Is it even possible with todays usage of fossile fuel versus energy needed?

  31. 31
    Hans Erren says:

    re: #19 and #24

    I received from Tom Boden a global emission update through 2002

    graphs here:

  32. 32
    David Archer says:

    In response to comment #30

    We’ve left 280 ppm far behind. 280 ppm is the preanthropogenic concentration, from 1750 or earlier. The CO2 concentration today is 365 or something like that, going up at about 1.5 ppm per year. In my model, that is to say neglecting surprises but just considering the atmosphere / ocean / CaCO3 system, if we stopped releasing CO2 today and closed the terrestrial biosphere to either releases or uptake of carbon, just closed the system, CO2 would relax down to some value higher than today. We’ve released about 300 Gton C so far (unless Ruddiman’s right and we started 8000 years ago). From a table in reference #7, 16.8% of a 300 Gton release will still be in the air in 1000 years. A convenient conversion factor is 1 ppm about equals 2 Gt C in CO2 in the atmosphere. So 17% of 300 Gt C is about 25 ppm higher than preanthropogenic, or about 305 ppm in the year 3000.

    Surprises that could change this include going into a glacial climate state, which had the mysterious ability to draw down CO2 in the past, or uptake by the terrestrial biosphere. The total biosphere including soil carbon is about 2000 Gt C, so it’s not absurd to imagine growing it to 2300 Gt C. But it would have to take up the entire 300 we emitted, is the point.

  33. 33
    Hans Erren says:


    I don’t think it is justified, given the noisy data to assume a mechanism other than a first order Fick’s diffusion.

    Using a simple first order Fick’s law diffusion fit on the emissions and concentrations
    1958-2002 yieds – with an equilibrium value of 280 ppm – a diffusion constant of 0.98135.

    The 2002 level is 373 ppm (excess of 93 ppm above 280 ppm)

    1) Stopping emissions completely in 2002 gives values of 320 ppm in 2050 and 297 in 2100, which is a decimation of already 82% in 98 years.
    2) reducing emissions with 2% per annum gives a peak of 391 ppm in 2028 and a reduction to 348 ppm in 2100
    3) holding emissions at 2002 level gives a growth to 428 ppm in 2050 and 453 in 2100
    4) increasing emissions annually with 1% gives 463 ppm in 2050 and 596 ppm in 2100

    The emission trend of the last 55 years shows an unintentional decreasing emission growth. No SRES scenario assumes a continuation of the current decreasing trend.


  34. 34
    Dave Dardinger says:

    re comment 27: Your point is that the amount the oceans can absorb from the atmosphere is determined /limited by the condition of the surface waters. But that implies that the amount of carbon being carried from the surface to deeper waters is fixed, or nearly so. The point of my previous message was to say that it isn’t.

    Further, if in the past we had roughly 90 GtC moving in both directions between ocean and atmosphere, it would be absurd to claim that this value was somehow fixed and that changes on the order of a few percent in either direction would totally change things. Most negative feedbacks require a fairly substantial disequilibrium before they are very noticable. This is why I ask if there is data about the flows between surface and deep waters in the past. It’s possible there’s an upper limit to flows based on pH, but it’s more likely that instead we’ve been seeing the lower limit on flows in the recent past as the biosphere draws down the CO2 level as far as it can and still operate. Increased atmospheric levels of CO2 up to several doublings will likely just increasingly rev up the biosphere. Since it’s unlikely we’ll do much more than double the prehistoric CO2 level before using up cheap fossil fuel, I’d be very surprised if anything untoward occurs.

  35. 35
    Ivan Doumenc says:

    Is there a delay between the release of a tonne of CO2 in the atmosphere and the moment that this tonne starts to impact the earth’s climate?

    I have heard theories that what we are seeing today is the result of C02 released 100-120 years ago, and that today’s releases will start to have their effect on our climate in 2120.

    I would like to know if this theory has a scientific basis.

    Thank you,

  36. 36
    David Archer says:

    In response to comment #35

    The CO2 appears in the atmosphere immediately, of course, at which point it begins to wind its way into the oceans and the biosphere. Wigley made a controversial but I think correct point a few years back that if the real concern of Kyoto and IPCC is the really severe climate change at the end of the century, the cutting emissions now will have less effect on that than cutting emissions closer to that time. Emissions now will have had some time to soak into the ocean.

    There is however a time lag for the warming, which is what you’re probably thinking of. It takes a long time, decades, to warm the interior of the ocean. This means it takes a long time to ramp up to the full steady-state climate impact of a slug of CO2.

  37. 37
    Chris Randles says:

    Regarding post 32

    Posting 87 includes “The roughly 500 billion metric tons of carbon we have produced is enough to have raised the atmospheric concentration of CO2 to nearly 500 ppm.” So 500 billion tons would be enough to raise from 280ppm to nearly 500 ppm which looks more like 1 ppm equates to 2.5 billion tons. On this basis a 100ppm increase (from 280 to 379 (not 365)) would be more like 250 Gton of carbon instead of 300 Gton.

    I think it would be sensible to sort out these numbers.

  38. 38

    In JGR Atmospheres, February 15, 2005 there is a paper by a Stanford prof. which claims “lifetime of CO2 ranges from 30-95 yr although a more likely upper limit may be 50 or 60 yr.” See section 1 in:
    I have no expertise in this area but I would like to understand why there is such a difference is estimates of CO2 lifetime. Any thoughts ?

  39. 39
    David Archer says:

    In response to comment 34

    Several sources list the mass of the atmosphere as 5.3E18 kg. Assuming an average molecular weight of 28.8 (80% N2, 20% O2) gives us 2.21 Gton C per ppm of CO2. Sabine et al (Science 305: 367, 2004) seem to be using 2.1 Gton C per ppm (consistent with an atmospheric mass of 5.0E18 kg). Let’s assume that Sabine is right and I’m wrong.

    Sabine has budgets of emissions and current perturbed inventories as well. They have emissions to 1994 of 244 Gton C. CDIAC (link in comment #24) says 283 Gton C to the year 2000. The ocean inbentory (the whole point of the paper) was 118 Gton C. Sabine figures a net release from the terrestrial biosphere of 40 Gton C between 1800 and 1994. So if we put together a combined net source, it would be close to 320 Gton C, which could have raised pCO2 by 150 ppm, to about 430 ppm.

    500 Gtons of emissions seems out of the ballpark; could it have been Gton of CO2, rather than Gton of C? Mass units for chemistry are endlessly confusing.

  40. 40
    David Archer says:

    In response to comment #36

    Wow, this was interesting, thanks! The problems with Jacobson’s calculation are (1) the assumption that the atmospheric pCO2 is relaxing toward preanthropogenic in Jacobson’s equation 1, and (2) the wide range of uptake time scales for different parts of the ocean, which interact with the accelerated rise in atmospheric pCO2. I wrote up a simple model in excel which puts CO2 into the atmosphere and takes it into the ocean according to various reservoir sizes and ventilation times. The estimated uptake timescales are within the range he reports for his data-driven calculation, 50 years or so, even though the mean uptake time of the ocean reservoirs in that model, weighted by their sizes, is 600 years. I have emailed my spreadsheet, and a mention of, to Dr. Jacobson.

  41. 41
    Pascal says:

    It is obvious that we’ll not avoid the anthropogenic greenhouse effect.
    What are you thinking about a fight against this effect by , for example ,the injection in the higher tropsphere of very tiny (1 to 5 microns)aerosols?
    These aerosols must be chosen transparent to the IR and opaque to the visible radiation.
    In another way what are you thinking about the artificial injection of micronised CaCO3 or MgCO3 in the oceans to neutralize atmospheric CO2?

    thank you to apologize my bad english.

  42. 42
    David Archer says:

    In response to comment #33

    The ocean uptake physics are governed by ocean circulation, which we know a lot about by measuring the carbon-14 distribution in the ocean. This is not noisy data. The constants governing the carbon chemistry in the ocean are known to many significant digits, from nice well-behaved laboratory data. The thermodynamic formalism of how one does carbon chemistry calculations are almost fundamental laws of science. The rate of gas exchange, I’ll grant you, is a noisy function of wind speed, but the long-term uptake is not terribly sensitive to gas exchange; it’s circulation that matters primarily in the long run. .

    The equilibrium pCO2 is not 280 ppm. This is an assumption you are putting into your calculation, not a conclusion you are taking out.

  43. 43
    David Archer says:

    In response to comment #41

    With regards the idea of decreasing earth’s temperature by injecting aerosols in the stratosphere: I am not an expert on this, but I haven’t heard any reasons why it wouldn’t work. The climate cooled down significantly, for several years, after Pinatubo.

    With regards neutralization with CaCO3 or MgCO3: It might be easier to do the chemical reaction in high-CO2 flue gas from power plants, rather than try to make it work at seawater temperature and pH. I would think that neutralized CO2, in the form of bicarbonate at seawater pH, would be much less harmful to ocean biota than would unneutralized CO2, either directly injected or by passively letting it invade the surface ocean from the atmosphere.

  44. 44
    Hans Erren says:

    Re: # 42


    Considering the vostok Ice core the equilibruim CO2 level is a fairly linear function of temperature (9.8 ppm/K)
    I agree that the pre-industrial equilibrium of 280 ppm should now be 287 ppm, but that increase is still small compared to the anthropogenic signal of 93 ppm.

    Secondly the complete atmospheric reservoir of 730 GtC is negligable compared to the oceanic reservoir of 38000 GtC using first order principles.

  45. 45
    Pascal says:

    In response to #43

    thanks for your response

    I’m agree with you for the neutralization of high-CO2 flue gas from power plants.
    I had the idea to put micronized CaCO3 on the surface of the oceans primarily to increase the global albedo of the Earth .
    It was the same idea that the aerosols.
    I thought that CaCO3 was a good candidate for this purpose .
    It was a natural and very abundant element ,it could give a very little “milky” aspect to the oceans ,maybe sufficiently to increase the albedo.The increasing of the pH of the ocean was a little plus.
    But it is an evidence that it should’nt be a low-cost operation.

  46. 46
    Eli Rabett says:

    WRT # 43 and 41: Good luck finding an aerosol that does not absorb in the IR:). Almost by definition an oxymoron because of the relationship between vibrational motions and IR absorptions.

  47. 47
    Hans Erren says:

    re #43


    I am aware of the temperature sensitivity of the CO2 equilibrium of the atmospheric CO2 level: the vostok ice core record yields a sensitivity of 9.8 ppm/K. therefore the current equilibrium should be 288 ppm (0.8 K higher as in preindustrial times).

    Secondly the total atmospheric CO2 reservoir is negligable compared to the oceanic reservoir (730 GtC vs 38000 GtC) in first order diffusion effects.

  48. 48
    ed says:

    How long will global warming last?

    My perdiction: 500 years

    My Opinion: The icecap over Greenland will melt over the next 500 years. At that point, the Atlantic Ocean current will reverse and initiating a cooling period (

  49. 49
    John Dodds says:

    Re #48 & since this question/topic invited rampant speculation by its choice of a title:
    My prediction is closer to 10-20,000 years.
    Since this is when the orbital fluctuations that supposedly caused the last ice ages willl reverse and we will START to get colder because there will be less energy coming into the earth from the sun.

    My problem with the 500 year scenario is that CO2 induced warming increases the amount of energy staying on earth, thus melting (some of?)the Greenland ice. BUT Reversing the Atlantic ocean current due to fresh water ice melt, is a local phenomenon, not global AND it does little to reduce the slow steady heat/energy buildup globally – so warming will continue.

    As for the CO2 increase, if the price of oil keeps going up at its current rate, $20/barrel in the 90s, $58 today, some broker estimated $105 in the near future, THEN solar energy becomes very affordable very quickly. The “dreaded” oil companies will rush to make a solar profit. CO2 goes down. CO2 induced warming goes down, BUT the Milankovitch orbital mechanics still dictate that in 10,000 years we will be closer to the sun, & probably warmer.

    Speculating is such fun. You can make a case for just about anything!!
    Time to close this topic guys!