RealClimate logo

Arctic Sea Ice decline in the 21st Century

Filed under: — group @ 12 January 2007 - (Français)

Guest Commentary by Cecilia Bitz, University of Washington

Last month a paper I co-authored received considerable media attention. Headlines read “Experts warn North Pole will be ‘ice free’ by 2040”, “The Big Melt: Loss of Sea Ice Snowballs“, and “Arctic Clear for Summer Sailing by 2040: Models Predict Rapid Decline of Sea Ice”. The story also reached NPR, BBC, CBC, the Discovery channel, and Fox News, among others. Dr. Marika Holland, the first author of the paper, was inundated with media attention. About a dozen journalists contacted me too. I was impressed by the questions they posed — questions that probably reflect what the public most wants to know. However, after giving lengthy interviews, I would read the resulting article and see my explanations boiled down to a few lines. In this essay, I’d like to explain the science in the paper and give my answers to the most often asked questions.

In our paper (with co-author Bruno Tremblay), we examined the September Arctic sea ice cover in the 20th and 21st centuries in climate models, and found occasional decades of very rapid retreat. The most extreme case was a decrease from 6 to 2 million square kilometers in a decade (see Fig 1). This is about 4 times faster than the decline that has been observed in the past decade.

Figure 1: (a) Northern Hemisphere sea ice extent in September from one integration of the Community Climate System Model version 3 (CCSM3) with observations from the satellite era shown in red. The light blue line is a 5-yr running mean. The three lower panels show the September ice concentration (ice floes are separated by open water) in three select decades.

It is common practice to run climate models multiple times with slight variations to the initial conditions. Because the system is chaotic, the natural variability in each run is random and uncorrelated from one run to the next. When an ensemble of runs is averaged, the natural variability is reduced in the ensemble mean, and it is easier to detect a significant trend.

An ensemble of runs offers an opportunity to evaluate rare events too, such as extreme sea ice decay. We were in search of evidence for “tipping points“, which several authors have speculated might exist in sea ice. RealClimate places sea ice in the category of systems with “known unknowns” with regard to tipping points. This means we know there are thresholds involving sea ice (e.g., it can cease to exist), but we don’t know when, or if, the climate will arrive at one.

Only one of seven ensemble members had an event as extreme as quoted above, and it resulted in near ice-free conditions for September by 2040 (see Fig 1d). (The sea ice grows back at least for some portion of winter for the duration of the 21st century.) However, every ensemble member had an event 5 years or longer at some time in the 21st century when the sea ice retreat was about 3 times faster than the observed retreat since 2001 (see Fig 2). These ensemble members took about 5–10 years longer to become nearly ice-free in September than the most extreme case.

As illustrated in Fig 1, the sea ice retreat accelerates during the 21st century as the ice decays and more sunlight is absorbed by the ocean (the positive ice-albedo feedback). Increasing ocean heat transport under the sea ice adds to the melt back. The retreat appears abrupt when natural variability in the ocean heat transport into the Arctic Ocean is anomalously high. We did not find clear evidence of a threshold, which can be difficult to identify given the variability and complexity of the climate system. Therefore we can neither verify or rule-out the existence of a tipping point. Regardless, the rapid declines seen in our runs are a serious concern.

Figure 2: Northern Hemisphere sea ice extent in September for all seven integration of the CCSM3 with observations from satellite era shown in black.

Most common questions asked by journalists

1) How does our model compare with the trend in the observed record?

The trends in the seven ensemble members for 1979-2006 span the trend in the observations: Some members retreat a little faster and some a little slower, as expected from the random natural variability in the runs (see Fig 2). The model also reproduces the mean and variance of the observations with good fidelity.

2) Other scientists are predicting an ice-free Arctic in September by the year 2060-2080, why is this model predicting it 20-40 years sooner?

First consider estimates based on extrapolation from the observational record. I’ve heard these numbers quoted in the media, but I have not seen a reference to a scientific paper that discusses the analysis in any detail. Figures 2 and 3 illustrate the danger of making an estimate of the future from the observational period. The future trend is not linear, the observational record is too short and the ice-free time is too far in the future to trust extrapolation. If one carries out such an exercise anyway, extrapolation from a linear fit to 1979–2006 gives a zero intersect (indicating the first ice-free year in the future) at about 2110 (see Fig 3). If instead one uses just the last decade, the extrapolation gives 2060. Both estimates are questionable, and so instead we turn to climate models.

Figure 3: Extrapolating into the future from the observational record.

3) Is sea ice in our model retreating faster than in other models?

Figure 4 shows September ice retreat in 16 models that were archived for the IPCC AR4. The most extreme predictions are from models that have too much or too little sea ice extent compared to observations, so it is important for a model to produce the correct sea ice coverage in the past. Some of the spread is expected from natural variability, but much depends on differing model sensitivity relating to the representation of sea ice, heat transport by the ocean, and cloud cover. It is not possible to identify the most accurate model prediction, although I think it is safe to rule out some of the outliers owing to their poor match to the observations.

About half of the models become ice-free in September during the 21st century. I included one ensemble member from our model, CCSM3, which is in the middle of the pack until about 2020. Our model run retreats faster than most after about 2020, but it isn’t radically different.

There is considerable uncertainty in future model projections, and Figs 2 and 4 illustrate why it would be better not to focus too much on the year 2040, which to our dismay was highly publicized. The more important message from models is that all but a few outliers predict enourmous sea ice retreat this century. At least a few respectable models predict a nearly ice-free Arctic by midcentury, with a retreat that may be punctuated by rapid events.

Figure 4: Northern Hemisphere sea ice extent in September from model integrations submitted to the IPCC AR4 with observations from satellite era shown in black.

4) Is it too late to save the sea ice?

The future emissions scenario discussed here is one that assumes modest increases in emissions. If humans can reduce the rise in emissions compared to this, then sea ice retreat would be slower and rapid events would be rarer, according to the IPCC AR4 models.

5) Have we crossed a tipping point?

I don’t think we have yet. If we fix the greenhouse gas and aerosol levels at year 2000 values and run the model into the 21st century, the sea ice retreats for only another decade or two and then levels off (some of the ensemble members even recover a little bit). So according to our model, the sea ice does not appear to have passed a threshold yet. We have not done an exhaustive study of any years beyond today, so unfortunately we cannot say with certainty that no tipping points exist. The bottom-line: The retreat can be surprisingly rapid even without clear evidence of a tipping point.

I thank Dr. Holland for valuable suggestions to improve this post and providing Fig 1. I thank Ian Eisenman for computing ice extent from the IPCC AR4 models shown in Fig 4. I look forward to reading your comments and questions.

127 Responses to “Arctic Sea Ice decline in the 21st Century”

  1. 1
    Todd Albert says:

    Great post. I look forward to reading the article, but I always love these clear explanatory summaries.

    The last two questions (too late, tipping point) are the types of questions I am asked most often by students, community members, peers, and others. I appreciate your hopeful answer and it lends credence to my typical answers.

  2. 2
    cat black says:

    re #1 [tipping] Seems to me that the most balanced answer would be that tipping points in Arctic ice melting may exist simply because they cannot be fully ruled out, and further more if they do exist then they may fall anywhere on the continuum of future data points, including the second Wednesday of any future (or even recently past) month you’d like to choose. One guess being more or less as good as another, given the current state of our knowledge.

    Not very satisfying, I suppose, for people who want “the answer”. But then again, allowing them to choose any second Wednesday for the point beyond which things are FUBAR might get across the larger issue, which is that they are, in some real way, making that choice every day in how they live their lives, and measures they are, or are not, taking to stave off this interesting little disaster looming on some random second Wednesday, in some random month, in some random year.

  3. 3
    Hank Roberts says:

    Dr. Bitz, have you heard from any researchers in other fields whose subjects have had the Arctic ice as a background assumption, who are now looking for what happens in their area of interest as this change happens?

  4. 4
    Andy Revkin says:

    Great followup, and much appreciated (from one who wrote about the paper here:
    Best journalism results from ongoing ‘conversations’ with scientists, which is facilitated by this kind of presentation. Worst journalism happens when everyone only focuses on an issue of this sort in the short span of hourse between when the press release is issued and the deadline looms.

  5. 5
    Phillip Shaw says:

    Just a note to say that the link to bitz_fig4.jpg may be broken. It just take me back to the home page.

    [Response: fixed. Thanks – gavin]

    On a more positive note, this site is a ‘must read’ for anyone interested in learning more about climate change. Your hard work makes it very readable to the typical layperson such as myself.

  6. 6
    Charles Muller says:

    Thanks for this interesting post.
    Do we know if Artic was free-ice in past climates ? I read that some peri-artic zones like North Siberia were probably warmer than present at some periods of the Holocene Climatic Optimum (8000-6000 BP), and that summer orbital forcing 60°N and poleward was much more important during the last interglacial (Eemian). Is there any proxy (or model reconstructions) for Arctic ice conditions during these times?

  7. 7
    Brian says:

    Can you someone point me to some quality reports regarding how life w/could be affected by zero sea ice?

  8. 8

    Dear Dr Bitz,

    Can I first thank you for putting your head above the parapet.

    The idea of a tipping point for the Arctic sea ice was first described in a now unfairly neglected book “Climate through the Ages” by C.E.P. Brooks, who was a leading British meteorologist sixty years ago. The point is that because the ice albedo effect is a positive feedback, in a warming world it will inevitably lead to a sudden collapse of the ice.

    As the Arctic ice sheet contracts the ice albedo decreases causing the melting to accelerate. It is this acceleration which ensures the rapid collapse of the ice sheet.

  9. 9
    John L. McCormick says:

    RE # 3

    Hank, I get the gist of your question.

    I want to see the National Academy of Science devote time and attention to the Western North American precipitation and temperature changes as the Arctic Ocean 90 percent albedo shifts to 40 percent absorbent.

    Certainly that dynamic deserves as much attention as was applied to the arbrupt climate change study. In fact, an ice-free Arctic by 2050 IS ABRUPT CLIMATE CHANGE.

    The convective currents from the Gulf and Southwest would IMHO be affected and likely cause planting and harvesting impacts in the world’s grain basket.

    Are we building a corn-to-ethanol industry in a box canyon?

  10. 10
    Richard Jones says:

    I enjoyed reading the essay, but isn’t A1B the most pessimistic of the IPCC scenarios, with CO2 increases > 1% pa (i.e. doubling the rate of historic growth)?

    What happens in mid-low case scarios?

  11. 11
    pete best says:

    Dear RC

    I bet it was the left wing media that reported the worse case year 2040 scenarios.

    Great work, once again we are shown that tipping points have not been demonstrated in the models are the alarmist 2040’ers are going a little over the top. It also gives us hope that by reducing emissions to year 2000 levels would be a good start and seems likely to happen if we weren’t for that fact that we are going to experience a third world population explosion by 2050.

  12. 12

    Dear Peter,

    Even if we reduce CO2 levels to those before the year 2000, say the year 1997, the ice will continue to melt at the rate it was melting then – 10cm per year, see

    “… the SCICEX data indicate that ice has continued to thin in some regions throughout the 1990s, at a rate of about 0.1 meter per year..”

    In 1997 it was 2m thick. If it has continued to melt at that rate, then it will be gone by 2017, not 2040. If the thinning has accelerated, and/or there is a minimum thickness for sea ice of say 0.5m, then it could disappear much sooner. With the effect of an El Nino it could go this summer!

  13. 13
    Eric Swanson says:

    Dr. Bitz,

    I was concerned about the strength of the sea-ice/ocean albedo feedback for many years. In 1992, I presented a paper on the subject and have tried to follow the science ever since.

    After reading your paper last month, I corresponded with Dr. Holland and also looked back at some of the work which was published after the SHEBA experiment. Also, I read some of the CCSM 3.0 manuals, seeking more information.

    After a brief look, I found that there appears to be no zenith angle dependence in the albedo for the ocean, but that a fixed value of 0.06 is used. I showed back in 1992 that this is value was incorrect for the direct beam component of the incident solar energy. This was found by Payne in 1972 (J. Atmos. Sci. 29, 959), who published results of a tower experiment over the open ocean. It is also quite likely that the albedo of sea-ice, especially when it is covered by melt ponds, could be a function of zenith angle. From what I’ve been able to learn of the SHEBA project, there was no measurement of the direct component of the insolation, only the hemispherical radiance on a horizontal surface. Lacking measurements of the direct component, it would appear to be difficult to quantify the effect of zenith angle on albedo.

    As an engineer who has studied solar energy systems, I know that the difference between the direct insolation and the diffuse energy can be very important. Dr. Holland agreed that the model could be improved if zenith angle effects were included. I wonder what you think about this question and why there has been so much trumpeting of your results, given the need for this obvious improvement in your models.

    Best Regards,

  14. 14
    Charles Muller says:

    A point I don’t understand in your GRL paper :
    “The model is run at a nominally 1-degree resolution with the north pole displaced into Greenland.”
    What is exactly meant by this “displacement” of North Pole to Greenland ?

  15. 15
    Pascal says:

    thanks for this article

    in fig1, I see a relative stability of sea-ice extent between 2005 and 2025 (about).
    How can you explain such a phenomenon?
    Is this the beginning of perennial ice (thicker) melting?

  16. 16
    Edward Greisch says:

    Thanks for actually showing multiple runs on the graphs.

  17. 17
    Jeff Weffer says:

    Won’t the ice just freeze back in the Winter?

    When the arctic circle goes into six months of darkness from September 21 to March 21 for example?

    The sea ice has already frozen back and now covers all of the Arctic. It won’t start melting again until July.

  18. 18
    crandles says:

    In its annual review of 2006,

    NCDC says:
    “Arctic sea ice conditions are inherently variable from year to year in response to wind, temperature and oceanic forcings. Quite often, a “low” ice year is followed by recovery the next year. But increasing surface temperatures in high latitudes have contributed to progressively more summer melt and less ice growth in the fall and winter. While natural variability is responsible for year-to-year variations in sea ice extent, three extreme minimum extent years along with evidence of thinning of the ice pack suggest that the sea ice system is experiencing changes which may not be solely related to natural variability.”

    Is it reasonable to ask how frequently in your simulations you see situations where there is such a sharp acceleration which lasts for at least three years like we have seen in the last three years? In what proportion of such similar situations in your simulations does the faster rate continue for some time?

    I created this:

    4.5 times faster may make it look like your extreme case of 4 times faster look quite conservative but it is important to realise you are looking at longer periods than three years. I could be jumping in and worrying too soon but it appears I am not the only one seeing a change in behaviour of the ice extent.

  19. 19
    Paul M says:

    Scientists like the ones on here have a peculiar job to do. They are born on the planet, nurtured by parents or adults who care obviously cared , and retain enough of their youthful curiosity to become members of the scientific community. Now they have to examine the ultimate demise of the very conditions that nurtured and supported life systems. Just like in the high school titration lab experiment, there will come a point (sooner than later) when BOOM, everything changes. Spring 2007 will be a time which will be studied millenia from now, if there are survivors to study it. Metaphorically speaking, we are all examinig why the tide is out like it is, not realising the tsunami that is on its way. I hate to be the gloom and doom guy, but get up, open your door, and look outside.

  20. 20
    Cecilia Bitz says:

    Thanks for everyones’ comments.

    Post 7 and 18, I’ll get back to you when I am in my office ASAP.

    Post 3 Have I heard from researchers with interest in the ice retreat? I am in contact with several biologists in my universtity and others that I have met in various meetings.

    Post 10 Isn’t A1B the most pessimistic SRES scenario? There is a good figure showing CO2 concentration levels in the A1B, A2, and B1 scenarios along with the full range of concentrations in all SRESscenarios at (scroll to bottom) Only the A1B, A2, and B1 SRES scenarios were completed by most modeling centers for intercomparison
    in the IPCC AR4. Of these three, A1B is in the middle, but the three only really diverge after 2050. So there is not much point in arguing about their relative levels prior to 2050. Furthermore, CO2 is not the only relevant greenhouse gas. The rapid events we found are more common in runs with higher GHG levels. There is a better chance of avoiding rapid events if we can reduce our GHG emissions.

    Post 13 Obvious need for improvement in models? Actually, the ocean surface albedo is not fixed at 0.06 in the model. It is a function of zenith angle and diffuse/direct beam. The details are in the CCSM3 coupler manual section 16.3.3 Ocean Surface Albedos

    The albedo is 0.06 only if there is no diurnal cycle in the model, which we sometimes use in test mode. However all the runs presented here and completed for IPCC AR4 have a diurnal cycle. The zenith angle dependence compared favorably to SHEBA data from Scott Pegau.

    Post 17, Doesn’t the ice freeze back in Winter? Yes it does, see my parenthetic note “(The sea ice grows back at least for some portion of winter for the duration of the 21st century.)” in paragraph 5. The link between sea ice and climate varies with season. The retreat is mainly in summer. The ice grows back in winter but remains thinner all year. In fact, Rothrock et al 1999 found the ice is 40% thinner in the central Arctic in the 1990s compared to submarine tracks taken prior to 1976. These submarine tracks were taken in later winter! The thinner ice then retreats faster in summer according to observations
    and models.

  21. 21
    Michael Salem says:

    Has commenter #11 misread the article? This person seems to think that ice stops retreating when emissions are reduced to yr 2000 levels, but I read the article to say that ice retreat stops when emissions are cut to zero and atmospheric greenhouse gas levels are reduced to yr 2000 levels. Which interpretation is correct?

    [Response: The latter (well, roughly). As long as emissions are non-zero, the rate of increase in greenhouse gas concentrations is positive, i.e. greenhouse gas concentrations continue to increase, and the enhanced (i.e., anthropogenic) greenhouse effect strengthens, temperatures further warm, and ice further melts. Decreasing the rate of emissions does not stabilize the greenhouse gas concentrations, it simply decreases the rate of their increase. -mike]

  22. 22
    Hank Roberts says:

    This is another paper mentioned here a while back; I’m curious whether the Navy data sets have any more information than is publicly available and whether their conclusion is much different.

    “… an analysis of the diminishing sea ice trend in the Arctic Ocean by examining the NPS 1/12-degree pan-Arctic coupled ice-ocean model. While many previous studies have analyzed changes in ice extent and concentration, this research focuses on ice thickness as it gives a better indication of ice volume variability. The skill of the model is examined by comparing its output to sea ice thickness data gathered during the last two decades. … The NPS model indicates an accelerated thinning trend in Arctic sea ice during the last decade. The validation of model output with submarine, EM and ICESat data supports this result. This lends credence to the postulation that the Arctic not only might, but is likely to be ice-free during the summer in the near future.”

    Also thanks for the response about interest from other researchers — I’m very curious what the biologists make of this change, in particular because I gather we’re still assuming that the biological recycling of carbon is going to stay effectively the same over the short term (as implied in Mike’s response above). I’ve wondered if the biologists will see any paths that include a significant fast change in biology, likely plankton species or abundance, that might alter how well the biosphere handles CO2 on an annual basis.

  23. 23
    Julie Brigham-Grette says:

    Post 6. Yes there is evidence for less sea ice during the Eemian. Some of us argue whether the entire Arctic was ice free in summers, so the answer is still out on that. But see this review and references there in for documentation on sea ice, marine temperatures, tree line movement etc.
    CAPE Last Interglacial Project Members, 2006, Last Interglacial Arctic Warmth Confirms Polar Amplification of Climate Change, Quaternary Science Reviews 25, 1383�1400

  24. 24
    Judith Curry says:

    Hi Cecilia,

    Very interesting paper and post. I would appreciate your assessment on the possible impact of missing or perhaps overly simplified physics from sea ice models that are used in climate models in your projection. I fully appreciate that the sea ice model that you are using is among the most sophisticated sea ice models currently being used in climate models. However, based upon my understanding of your sea ice model (and other sea ice models used in climate models), there are several processes that are highly simplified that would be expected to, with a less highly parameterized treatment, influence the ice albedo feedback (note some of these processes are mentioned in chapter 7 of IPCC TAR):

    – explicit melt ponds: albedo, latent heat, salinity effects
    – ice age effects on optics and thermodynamics, particularly for first year/thin ice and frazil ice
    – lead width distribution: impact on lateral melting and wintertime turbulent heat fluxes
    – treatment of fast ice
    – snow: nonlinear conduction, metamorphism, redistribution
    – ridged ice: including enhanced bottom melting

    While I understand that this blog is not the forum for detailed scientific explanations, I would appreciate a summary sense from you as to whether the incomplete treatments I’ve listed might collectively be likely to accelerate the melting or act to stabilize the ice. Do you see any possibilities among the list above that might provide a negative feedback that would help stabilize the ice? My basic concern is that the sea ice models that perform best in the current climate may have inadequate treatments of a warmer climate that is dominated by first year ice, very thick ridged ice and seasonal fast ice, and may have inadequate treatments of surface albedo and thermodynamic processes that would alter the transition from the current sea ice regime to the modified regime as a result of the projected warming. My own sense is that the loss of much multi-year ice could occur more rapidly than the models project, but the transition to a completely ice free arctic during summer would be slower than projected. Thanks for any insights that you can provide on this.


  25. 25
    Alexander Ac says:

    Re #22
    Considering the response of organisms to changes in the weather and climate is very complicated and in fact not a serious biologist would be able to predict the implications of certain future change. E.g. food chain and interaction between organisms are so complicated, that we can only “expect” some changes, but we to not know them for sure. What is more sure, that “in general” WE CAN say, that the more rapid change, the worse for organisms with low ecological valence and the better for organisms with high ecological valence (e.g. opportunistic organisms…). And of course from the past we know, what a rapid changes always brought – death to many, but life to others…
    As an starting ecophysiologist, I have a feeling, that the global primary productivity of terrestrial ecosystems (and hence the CO2 absorbing capacity) ecosystems will decrease with increasing rate of climate change and in fact, it already seems to be decreasing. For instance, during the El Nino years (especially 1998), there was a greatest interanual increase in the CO2 concetrations, mainly due to high “global stress” caused by adverse effects of El Nino… this year we should observe similar pattern.

  26. 26
    Steve Reynolds says:

    Re 21:
    Mike’s response> As long as emissions are non-zero, the rate of increase in greenhouse gas concentrations is positive…

    It is my understanding that both ocean and land are currently absorbing a large fraction of CO2 emissions, so at a rate that must be faster than the pre-industrial rate. Then emissions should not need to go to zero for CO2 concentration to decrease. Am I missing something?

    [Response: You misunderstand how these processes work. We are perturbing the carbon cycle from a pre-anthropogenic equilibrium. Carbon uptake by the ocean and terrestrial biosphere can only take up some finite positive fraction of the carbon produced by anthropogenic sources. The uptake processes simply act to slow the accumulation of greenhouse gases in the atmosphere relative to the rate at which it is being created by fossil fuel burning. There is no plausible mechanism that would allow natural uptake to produce a decrease in actual greenhouse gas concentrations in the face of positive anthropogenic emissions. As long as anthropogenic emissions are positive, greenhouse gas concentrations will increase. Interested readers are referred to the discussion in IPCC(2001). -mike]

    [Response: Actually, come to think of it, its a bit more complicated than that. It depends on the residence timescales of the various reservoirs. At sufficently low future emission rates we could eventually stabilize at concentrations lower than the current concentration, but not before reaching intermediate concentration levels higher than current. A nice comparison of CO2 emissions scenarios and associated future CO2 concentrations histories is available here. I’ll leave it to the carbon cycle specialists to comment further…I now see that David has already discussed this previously better than I can. I would refer folks there. -mike]

  27. 27
    Alan says:

    Re #11: “I bet it was the left wing media that reported the worse case year 2040 scenarios….if we weren’t for that fact that we are going to experience a third world population explosion by 2050″hasis)

    Hi Pete, a couple of questions. What “wing” of the media did you get your fact from? In the light of Cecilia’s lucid explaination, is it more politically correct to be an “alarmist”, an “ostrich” or a “deer in the headlights”?

    Thanks Cecilia for explaining the caveats to the “2040” headline, if more scientists made this kind of effort the internet would soon make bad science journalisim redundant.

    Perhaps the public’s preocupation with “tipping points” is the same phenomena seen when a mechanic advises that “you need new brake pads soon”, the first question many people ask is “how soon?”. To carry the anology a bit further, most will maintain their brakes as advised, some will wait until the pads are gone and the discs need machining, a minority will crash into a tree rendering the question moot.

  28. 28
    Edward Greisch says:

    I like the article and the downloaded paper. Could somebody please write a paper on when the famine will start in the USA? To most people, global warming is good. They don’t like winter and they are afraid of polar bears. [I’m afraid of bears too, but I like winter because there is no pollen in winter.] Most people would see the opening of the Northwest Passage as an economic benefit. As for the famine, most people would “let the grandchildren handle their own problems”. Most people also say: “We will adapt”, not knowing that adaptation is 99.999% death and extinction.
    Thanks for the opportunity to download papers from Dr. Bitz’s web site. Again, and knowing that you really can’t say and the problem is too hard, please write a paper on when the famine will start in the USA. AGW won’t be seen as bad until it is seen as bad for Human Americans.

  29. 29
    yves fouquart says:

    Hi cecilia and congratulations for this nice post that I found quite clear.
    Regarding to Judy’post, my question concerns the dynamics of high latitudes. Did you find any evidence for a significant change in dynamics and cloudiness in summertime in the Arctic regions in these simulations?

  30. 30
    Charles Muller says:

    #23 Thank you very much for the review, Julie. According to fig. 4, summer insolation forcing (compared to present) could have reached 40-75 W/m2 during the thermal maximum of the LIG – much more that any projection of GHGs forcing on 2050 or 2100 for the same melting season (JJA), I guess. But I suppose we must take account of winter negative anomaly for the same orbital forcing.

  31. 31
    Anonymous says:

    Dr. Bitz,

    Building on question #29, the link you referred to about cloud cover is a discussion about the paper put out by Francis & Hunter. I am confused about the use of Maximum Ice Retreat Anomalies (MIA) and the evidence that supports a negative correlation between MIA and DSF. A negative correlation? Does anyone know what this means for the ice-albedo feedback? Is DLF a prominent player in any of your ensembles?

  32. 32
    george naytowhowcon says:

    I have been monitoring Arctic Sea Ice minima / maxima extents for the last two years. I suggest that aerosols used for cloud forcing are creating a positive feedback of arctic surface temperatures during summer melts. The forced cloud cover acts as an insulating blanket trapping the earth’s radiant heat rather than allowing for cloudless nights to cool surface temps. Radiational cooling happens when there is little or no cloud cover.Thus increased summer melt. The remarkable recovery of the arctic sea ice is plainly visible. Currently, the Arctic Sea Ice is reaching out into the pacific from the berring streight, somewhat annomolous and demonstrates that it is indeed very cold there. The curiousity of the tropical jet of the eastern pacific crossing N. America and the Atlantic Ocean providing a warm winter to Moscow and Europe in general seems to have pushed Arctic air to South Asia. Are there any climate models showing the establishment of new path’s for mid lattitude jet streams consistent with Global Warming climate models?

  33. 33
    Charles Muller says:

    On the same subject (and also same journal and same publication month), I mention thereafter M. Winton’s analysis of some IPPC FAR models dealing with “tipping point” for summer Arctic sea-ice conditions.

    GEOPHYSICAL RESEARCH LETTERS, VOL. 33, L23504, doi:10.1029/2006GL028017, 2006

    Does the Arctic sea ice have a tipping point?

    Michael Winton
    Geophysical Fluid Dynamics Laboratory, NOAA, Princeton, New Jersey, USA

    Abstract – Two IPCC fourth assessment report climate models have Arctic Ocean simulations that become sea-ice-free year around in 1%/year CO2 increase to quadrupling experiments. These runs are examined for evidence of accelerated climate change associated with the removal of sea ice, particularly due to increasing surface albedo feedback. Both models become seasonally ice-free at an annual mean polar temperature of ~9°C without registering much impact on the surface albedo feedback or disturbing the linear relationship between Arctic Ocean climate change and that of the surrounding region. When the polar temperature rises above ~5°C, however, there is a sharp increase in the surface albedo feedback of one of the models, driving an abrupt elimination of Arctic ice and an increase in temperature above that expected from warming of the surrounding region. The transition to ice-free conditions is more linear in the other model, with ocean heat flux playing the primary driving role.

  34. 34
    William Astley says:

    In reply to comment 28 “…Could somebody please write a paper on when the famine will start in the USA? To most people, global warming is good. … As for the famine, most people would “let the grandchildren handle their own problems”. Most people also say: “We will adapt”, not knowing that adaptation is 99.999% death and extinction…”

    The above article does not state that warming of the arctic, is a catastrophic climate event.

    There is a general misconception concerning the relative affects and danger of planetary cooling as opposed to warming.

    Famine has occurred in the past, due to cooling of the planet, not warming. The planet is at the end of the interglacial period and has started to cool. The warmest planetary temperature in this interglacial period occurred roughly 8500 years ago. At that time the tree line, in the arctic, was roughly 200 miles north of its current location. (Lamb has a picture of the arctic tree stumps in his book “Climate, Past, Present, and Future, and notes that the stumps, indicate that the tree line in the arctic was roughly 200 miles north of its current location.) The planet was roughly 2C warmer, 8500 years ago, as compared to current temperatures.

    As the planet is 70% covered with water, when it is warmer there is more rain. For example, during the Holocene climatic optimum (warmest period during this interglacial), the region that is now the Sahara desert was a savana, that supported Elephants, Rhinos, and so forth.

    The planetary climate is unstable and drier during the colder glacial period. For example, there was an 800 times increase in dust in the Greenland ice sheet proxy data during the glacial period as compared to current conditions. There was an order of magnitude increase in sea salt in the ice sheet data, which indicates a massive increase in storms, during the colder glacial period as compared to the current warmer interglacial period.

    Life is difficult during the colder glacial period. For example, the early human population was reduced to less than 20,000 as a result of the climatic affects of the Toba eruption, which occurred roughly 70,000 years ago, during the glacial period.

  35. 35
    Ike Solem says:

    Thanks for the informative post. The 1999 paper, “Thinning of the Arctic Sea Ice Cover” can be found at

    “In summary, ice draft in the 1990s is over a meter thinner
    than two to four decades earlier. The mean draft has
    decreased from over 3 m to under 2 m, and volume is down
    by some 40%. The thinning is remarkable in that it has
    occurred in a major portion of the perennially ice-covered
    Arctic Ocean. This is not a case of thicker ice appearing
    in one region simultaneously with thinner ice appearing in
    another, induced perhaps by a change in surface winds and
    ice advection.”

    The combination of decreasing thickness and extent, as well as less recovery of winter ice in each successive season, indicates that the trend is not an anomaly. There is more data on decreased winter ice recovery.

    Keep in mind that at the same time there is an ongoing trend of melting Arctic permafrost. Since ~30% of the global soil carbon is locked up in the permafrost, it is possible that this could impact the global carbon cycle. There is also the issue of methane clathrates (RC)

    For more on the permafrost issue, see:
    â��Boreholes in Svalbard, Norway, for example, indicate that ground temperatures rose 0.4C over the past decade, four times faster than they did in the previous century…”
    “For the high-emission scenario, the area with permafrost in any of these layers shrinks from 4 million to just over 1 million square miles by the year 2050 and decreases further to about 400,000 square miles (1 million square kilometers) by 2100. In the low-emission scenario, which assumes major advances in conservation and alternative energy, the permafrost area shrinks to about 1.5 million square miles by 2100.â��

    Given all this information, perhaps Roger Pielke Sr. should revist his 2005 statement: “Our conclusion is that the Arctic Systems Science report, which received so much media attention, significantly overstated the actual trends of Arctic sea-ice coverage.”

    This report, available at, should be required reading for any journalist working ‘the climate beat’. (at least read the conclusions!). Journalists should also stop relying on a small clique of contrarians (who have been proved wrong time and time again) for sound bites. I suspect that what happens is that editors direct journalists to ‘appropriate experts’ as deadlines loom – but that’s poor journalism, and also results in a public perception that there is still any significant debate on the basic issue. Spinning articles using terms like “emotional debate”, “climate heretics”, and so on is also a disservice to the public; the issues involved are scientific in nature, not emotional or religious.

  36. 36
    Henry says:

    I don�t suppose the corporate stool pigeon who wrote that idiotic piece on the arctic being free of ice for white boys in sailboats considered what the fate of polar bears and other life beyond the cushy confines of pseudoscientific journalism in London will be? Or where all this melted ice is going to go, maybe in New York, bigger hurricanes?

  37. 37
    Charles Muller says:

    I suggest everybody should read ALL conclusions of ALL climate scientists, not just those which comfort his/her own’s opinion.

    For example, I don’t think current models and observations of permafrost enable us to say if thawed zones will be source or sink in the carbon cycle of the coming decades. As far as there are preliminary but contradictory results (eg Payette 2005 or Christensen 2004), we should be more careful before any statement on positive feedbacks. Broader assessments (like PErmafrost and Carbon Emissions – PEACE) will be useful.

    (For methane hydrates, I didn’t read any alarming projections since Kennett et al. 2003’s piece – in my opinion, Schmidt et Shindell 2003, Sowers 2006 or McDonald 2006 conclusions didn’t really favour the worst scenario hypothesis).

  38. 38
    pete best says:

    Dear RC

    Figure 4 possibly shows the current arctic ice line (red) on the lower side of the mean model which you can see from figure 2 to. Is 2040 the earliest that the the artic sea ice will be clear in the summer ? Could it not be even earlier or do energy constraints or ice albedo feedback be a limit on how early it can be.

  39. 39
    Mark Hadfield says:

    Great post Cecilia.

    Would you care to take a look at this article on John Fleck’s blog:

    He says

    “In the news release, only the ‘ice-free in 2040’ model run is discussed. It is the model run featured in the animation included with the news release. Is it any surprise that this is the scenario journalists chose to focus on, given that this is what the scientists, through their news release, told them was the most important?”

    Would you like to comment on this?

  40. 40
    Grant says:

    Re: #34 (William Astley)

    You say

    The planet was roughly 2C warmer, 8500 years ago, as compared to current temperatures.

    What’s your source for this claim? Graphs published by James Hansen indicate that we’re soon going to cross the temperatures in altithermal (the 8500 yr ago period you refer to) and Eemian (the previous interglacial) times. This graph also casts your claim in doubt:

  41. 41
    Hank Roberts says:

    David Archer has a recent summary article:

  42. 42
    Ike Solem says:


    The permafrost has been an inert member of the global carbon cycle; one could get at the age of the carbon locked up in the permafrost by looking at the 14C radioactive isotope content (and having just checked, there is indeed a substantial literature on this very topic) – for example, see While there is no evidence that any person has ever eaten the flesh of frozen mammoths, microbes will not be so particular. To quote from the cited abstract, “Our results indicate the potential for ancient C to fuel microbial respiration and C release once permafrost Yedoma soils are thawed”.

    It is highly unlikely that the permafrost will become a sink for global CO2 as warming proceeds. For a ScienceDaily news report, see Greenhouse Gas Bubbling From Melting Permafrost Feeds Climate Warming” Whether it is methane or CO2 that is released will probably depend mainly on the local hydrology; wet conditions favor more anoxic conditions which favors methane release.

    A good article on the potential effect of increased CO2 on carbon sinks (photosynthesis) can be found here. Note that plants are also very sensitive to extreme weather events (a searing but short-lived heat wave can have a dramatic effect on the yearly agricultural yield, for example – something that is hidden in yearly avg. temps). There is really no evidence that a “CO2 fertilization effect” will mitigate atmospheric CO2 levels or lead to increased agricultural productivity, despite what some fossil fuel-funded ‘research centers’ would have you believe.

    On the issue of methane clathrates, it might be worth considering where the heat to melt the ice is coming from. In the 1999 Rothrock et al paper, they point to ocean currents, atmospheric poleward heat transfer, and downwelling radiation as the possible conduits. A warming Arctic ocean might lead to methane clathrate release, though the response time is probably over a century or more (see this 2004 report on ocean temps and global biogeochemistry ) This is why we need to take action now, not at some unknown point in the future.

    Finally, you can always find scientists who will uphold the contrarian position in any scientific debate – as a good example, one can find PhD’s who will tell you that AIDS has nothing to do with the HIV virus – but it would be a serious mistake to base government policy on such claims. When we read media articles on the ongoing AIDS epidemic, no journalist feels the need to include statements from ‘the other side’. The relationship between increased temperature trends, climate change and human carbon dioxide and other greenhouse gas emissions is of a similar certainty.

  43. 43
    Nigel Williams says:

    I note that all the models show a decline in ice extent, with the middle order models dropping below 50% (6 million sq km) by 2050.

    By this stage (50%) I would think that the entire ice cap has become detatched from tera firma and is floating around at the mercy of wind and currents.

    Depending on surface circulation one can imagine the residual ice cap sort of sliding off the pole and heading into the Atlantic…?

    How do the models view the effects of the remaining ice cover floating away from the polar regions into more southerly regions that favour melting, rather than them staying obligingly close to the pole and away from higher solar insolation and warmer current areas?

  44. 44
    Hank Roberts says:

    >34, 40
    Lamb? Tree stumps?
    He may have been looking at something like the Medieval Warming, a local not a global condition; the more modern work tries to account for anecdotes in the larger context.

    Lamb,H.H., 1972 and 1977, Climate: past, present and future: London, Methuen, volume I 613 pages, volume II 835 pages.

    Lamb,H.H., 1982, Climate, history and the modern world: London, Methuen, 387 pages.

    Someone at CRU may be able to put that into context.

    “Back in the 1960s, all seemed stable with the world’s climate, and no-one had heard of ‘global warming’, until pioneering climatologist Hubert Lamb began to establish climate change as a serious research project. ”

  45. 45
    Nigel Williams says:

    …and is the loss of ice cover likely to lead to any significant change in ocean circulation through the polar regions, or is the water depth below the existing ice cover such that circulation will continue virtually unaltered?

  46. 46
    Charles Muller says:

    #42 Ike, permafrost is not the central subject of this discussion, so I won’t insist on it. But I think you should balance more carefully what is “highly unlikely” or not. Recent papers like Zimov et al. perspectives in Science 2006 are still a bit speculativeâ��and carbon cycle models still in their infancy.

    #40 For Western Artic, some recent works suggested a warmer than present condition. See for example Kaufman et al review in QSR :
    That’s why I ask in #6 if we have any information about summer sea-ice extent for this period (and LIG, July answer in #23). More broadly, I’d like to know if sea-ice reacts in the same way to orbital and GHGs forcings in current models and if not, what are the main differences in ice behavior and/or the main drivers in present and future ice melting (see also Judy questions in #24 for a much more precise context).

  47. 47
    Eric Swanson says:

    Re: #20

    Dr. Bitz, thanks for the pointer to the CCSM3 documentation. I had looked at it, but not gone back and carefully studied Briegleb, et al., 1986.

    I had taken a look at Scott Pegau’s data, plotting each day’s worth, however, I have not seen his paper in which he documents the results. I will need to find a copy to see whether he compared direct and diffuse radiances. His archived data shows that he made his measurements at times near local noon in mid-summer, when the sun was rather high in the sky. There was only one day with a full 24 hours of measurements. Also, he noted that there were problems with ice formation on the instruments, as there were with the other station instruments. Lastly, he used different types of instruments for measuring the downwelling and upwelling energy, which could be a further source of errors. Since SHEBA, there is new information about the Eppley PSP type pyranometers, pointing to a previously unknown temperature impact on their output.

    Re-reading Dr. Holland’s e-mail, I note that what she suggested was that the sea-ice models were in need of improvements, as Judith Curry also pointed out in #24 above.

  48. 48
    Ike Solem says:

    RE#46, #40, #34

    Charles – the statement being questioned was “the planet was actually 2C warmer 8500 years ago” and the paper you reference is a review of a temperature maximum in the western Arctic, robustly supported by a lot of paleoclimate data. The western Arctic is not the planet! Having actually flipped through the paper, I was struck by their final statement:

    “Unlike early Holocene warming, however, future warming will not be counterbalanced by the cooling effects of a residual, decaying North American ice sheet.”

    That ice sheet was the Laurentide ice sheet, a remnant of the glacial past – and one might also add that the Greenland Ice Sheet looks to go the way of the Laurentide in the near future. This is typical abuse of scientific publications by climate contrarians.

    In any case, warming the Arctic will affect both permafrost and sea ice – and then there was Ward Hunt Ice Shelf – just because you see icebergs doesn’t mean the planet isn’t warming.

    Regarding permafrost, there are many publications indicating that the newly exposed Siberian regions were a net carbon sink in the early Holocene, as peatlands rapidly developed – but that was between 11.5 and 9 thousand years ago as the world was coming out of a glacial period. For example, see here and links contained – but you’re not going to claim that’s evidence that the permafrost will act as a carbon sink in the modern world?

    In case you were going to, here’s the current situation (1993 Nature paper): Recent change of Arctic tundra ecosystems from a net carbon dioxide sink to a source. Note – Siberian lakes are disappearing.

  49. 49
    Cecilia Bitz says:

    Sorry I can’t answer every question. I appreciate the scientists who
    have shared their expertise in this discussion and answered many
    questions far better than I can.

    Comment 12 Will sea ice continue to thin at 10cm per year even if the
    CO2 levels are held fixed at 2000? No, assuming there are no
    unforeseen tipping points, the thinning rate would slow down due to
    stabilizing effects. In fact, the thinning rate slows down in CCSM’s
    A1B scenario, even though the CO2 level continues to rise. Retreat
    accelerates, but thinning decelerates in our A1B scenario.

    Comment 14 How is the North Pole displaced into Greenland? Just as
    longitudes (or meridians) converge at the real NP, the width of grid
    cells converge in a model that is square with real longitudes and
    latitudes. This is undesirable, so this model’s NP is displaced into
    Greenland, and the grid cells are NOT square with real latitudes and

    Comment 24 Judy – Thanks for your comments. Oversimplified is not the
    same as missing though. The items you list that contribute to
    ice-albedo feedback are parameterized in the model. I am not sure that
    more sophisticated physics would yield higher sensitivety. Among the
    items you list that are totally absent from models, I don’t think any
    are strongly amplifying or stabilizing. Nonethesless, we are already
    working hard to improve the model physics for IPCC AR5, so one day we
    should know for sure. I am curious why you have a hunch that the
    transition to an ice free arctic during summer might be slowed?

    Comment 39. Why did our paper focus on an extreme case? In a paper
    _about_ rapid ice retreat, it is not unreasonable to highlight the run
    with the most extreme reteat. I am grateful to have this opportunity
    (in addition to the paper) to discuss other model results
    and the uncertainty in the magnitude and frequency of rapid ice
    retreat. I don’t think it should be surprising that a press release
    from NCAR would focus on the NCAR CCSM results, the model Dr Holland
    and I helped develop over the past decade, rather than other IPCC AR4

    Comment 43. Can sea ice slip into the North Atlantic as the
    concentration declines? Probably not catastrophically. About 10% of
    the ice volume is exported into the North Atlantic every year
    today. The rate depends on the ice velocity and the thickness. The ice
    can move faster as it thins, but there is an upper limit set mostly by
    the wind speed. Most models compute ice dynamics today and they
    account for variations in the export rate. Frankly, I don’t remember
    whether the export increases or decreases in the future in CCSM. Dr
    Holland would know.

    Comment 45. Does the Arctic Ocean circulation change? It does in
    CCSM. Relatively warm water circulates into the Arctic Ocean from the
    North Atlantic. At present this current travels near the surface until
    it splits around Svalbard (the islands between the northern tip of
    Greenland and Norway) and then subducts below the surface, remaining
    well isolated from the sea ice in most of the Arctic Ocean. In our
    model this warm current stays nearer to the surface a little further
    past Svalbard on the eastern branch in the future. There are other
    more changes, but I mention this one because it affects ice

  50. 50
    William Astley says:

    In reply to comment 48: Muller’s referenced paper supports Lamb’s text book which states that the warmest period in the arctic occurred roughly 8500 years ago, at which time the average temperature was 2C warmer than the current climate (As noted Lamb’s text book includes a picture of tree stumps that are 200 miles north of the limit of the current arctic tree line.) The following is a quote from the paper in Muller’s comment (I include a link to the Holocene maximum temperature paper below).

    “The timing of the Holocene Temperature Maximum varied spatially, but the increase in temperature… At the 16 terrestrial sites were quantitive estimates have been reported (mainly summer estimates) were 1.6C +/- 0.8C higher during the HTM than present.”

    As to the quote “Unlike early Holocene warming, however, future warming will not be counterbalanced by the cooling effects of a residual, decaying North American ice sheet.”

    Unfortunately its appears planetary warming is about to end and the planet will suddenly cool which seems ironical as all of the discussion in the forum centers on catastrophic warming.

    I must be the only person in the forum who has read the GCR papers, read the solar papers that discuss the current southern coronal asymmetry of the sun, who has looked at solar data that supports the fact that the sun is at its highest activity level in 8000 years, who looks at the current solar data (see coronal hole CH255 and data Dec. 16 in the attached link) or who has examine paleoclimatic data outside of “recorded times”. “Recorded times” Gavin notes is the last 150 years.

    I must also be the only one who has investigated the data concerning geomagnetic field changes (cyclic 60 degree changes in inclination and dipole magnitude, starting about 800kyr ago, which coincidentally coincides with the end of the 41kyr glacial cycle and the start of the 100 kyr glacial cycle.).

    I am now looking into the question as to what is the source of geomagnetic field and why would the geomagnetic field suddenly and periodically change. I have unfortunately found the solution to that problem. Unfortunate as it supports the sudden cooling statement.

    Solar Terrestrial Activity