More on the Arctic

By Michael Mann & Phil Jones (guest)

Svalbard, an Arctic island in the Northern North Atlantic, is predicted to warm considerably more than most of the rest of the earth in many model-based scenarios. See for example the figure to the right, which represents a relatively high-end IPCC Third Assessment Report scenario for the projected surface temperature difference between the period 2071 -2100 and 1961-1990. Svalbard is the island north of Norway at about 80N between 15-30E.

The enhanced warming in this region is related to the issue of polar amplification that we have discussed previously on RC. It also happens that the Svalbard meteorological station is the 2nd station in the World Meteorological Organization (WMO) meteorological station list. This means that it tends to get noticed. The Climatic Research Unit (CRU) of the University of East Anglia maintains one prominent version of the global surface temperature data set and as part of its routine quality control, CRU flags any unusual (anomalous warm or cold) new measurements that come in. Svalbard has now been flagged consistently over the past several months, but the values have been confirmed as accurate by the Norwegian Met Service, which operates the Svalbard station.

Here are the recent Svalbard monthly surface temperature measurements, the long-term (1961-1990) means (“ybar”) and standard deviations (“sd”), and associated anomalies i.e., departure from average (“delta”) for Dec 2005 through April 2006 (all in degrees C):

Month Value ybar sd Delta

Dec 05 -3.8 -13.3 4.4 +9.5

Jan 06 -2.7 -15.3 4.7 +12.6

Feb 06 -9.8 -16.3 3.7 +6.5

Mar 06 -13.1 -15.8 3.7 +2.7

Apr 06 0.0 -12.4 2.7 +12.4

The numbers are fairly remarkable. April’06 was warmer than any previously recorded May, and January ’06 was warmer than any previously recorded April. The previously warmest April was -7.0C (1996) -4.3C (2004). There is currently an absence of sea ice off much of the coast of Svalbard, which is also unprecedented for so early in the year.

The April mean temperature is almost 5 standard deviations above the mean, a “5 sigma event” in statistical parlance. Under the assumption of stationary ‘normal’ statistics, such an event is considered astronomically improbable (< 1 in 106), and, like the summer heat wave in Europe in 2003 (which was a 5 sigma event in Switzerland, 3 sigma over Europe as a whole), deserves special attention. As we have nonetheless remarked before on RC, particular events, even seasonally-persistent anomalies as unusual as these, do not “prove” anthropogenic warming. But in a statistical sense, large outliers like this make it more probable that the underlying distributions are shifting and give us a glimpse into the types of anomalies we might expect to become more common in the decades ahead.

Correction and update: (1 June) Micheal Shouler points out that we misread the previous April record (corrected above). And now that the May 06 data has come in at a record-breaking +0.9 C, our statement that April 06 was warmer than all previously recorded Mays is still true – but only just! Things move fast in this field…

88 comments on this post.
  1. Adam:

    Quick point, the link “absence of sea ice” gives “404 not found”.

    [Response: Thanks for pointing this out. Apparently that file was moved. I’ve replaced with an alternative image at the same site. – mike]

  2. Hank Roberts:

    That link works for me. Note the target is a .pdf file.

  3. tom brogle:

    How is the Antarctic sea ice behaving ? Is it also showing Polar Ampification?

    [Response: See our previous article “Thank you…” for a link to an extended discussion of Antarctic trennds. – mike]

  4. stephan harrison:

    Is this pattern linked in some way to the very cold winter that central and eastern Europe experienced this winter?

    [Response: This pattern appears far more persistent, as it continues well into the present season. In part, this may relate to the longer timescale of response of certain components such as sea ice to warm atmospheric conditions (such as were found for Svalbard this winter). – mike]

  5. Karen Street:

    Mike, please explain again your answer to question 4. I too am interested in the answer.

    [Response: This is an example of a more general phenomenon. The climate system represents an interaction between components that respond on fairly fast timescales (e.g. the atmosphere) and components that respond on longer timescales (e.g. the oceans, ice sheets, and to some extent the vegetation and land surface). Sea ice falls into a middle category, but it is certainly a slower response component than the atmosphere. This means that if the atmosphere tends to lock into a particular pattern in a given winter leading to e.g. a colder central europe but a relatively warm North America and a warm Eurasian arctic such as was observed this winter, then the oceans and the sea ice, both may show the effects of this atmospheric pattern well into the spring and summer, even if the atmospheric pattern itself broke down by the end of the winter. So the warm temperatures this winter over much of North America and much of the Euroasian Arctic are relevant to understanding why sea surface temperatures over the Gulf of Mexico are unusually warm this spring, and why there is so little sea ice bordering Svalbard right now. – mike]

  6. Terry Haran:

    I think you meant to say that “The *January* mean temperature is almost 5 standard deviations above the mean…”, not April. It’s also worth noting that apparently the variability in March mean temperatures must be much lower than the variablility in April mean temperatures since a March delta of only +2.7 produced an sd of 3.7, while an April delta of +12.4 produced an sd of only 2.7.

  7. Alex R:

    Is there any simple explanation for why models show so much more warming in the northern hemisphere than in the southern?

    (This, of course, has policy consequences as well, considering that the vast majority of the world’s highly developed economies are in the northern hemisphere…)

    [Response: Largely that oceans warm slower than land, and there is more land in the NH – William]

  8. Alastair McDonald:

    At the end of 1999 it was reported that the Arctic sea ice had thinned by 40% during the years preceding 1997 [DA Rothrock, Y. Yu, and GA Maykut, “Thinning of the Arctic Sea-Ice Cover” (Geophysical Research Letters, vol. 26, no. 23, December 1, 1999)].
    Since then I have been monitoring the Arctic sea ice on my web pages at The maps there are produced by Robert Grumbine of NOAA, and they are not altered by me in any way. They show that the lack of sea ice around Svalbard has been happening for over four weeks, but it also shows that the extent of sea ice in the whole of Barents Sea is unusually low. Moreover, the three maps for the same day of the year, today, three years ago and six years ago show that the concentration of the ice has steadily decreased. That indicates to me that the thinning of the ice has continued, and if another 40% has melted in the last ten years it may have reached a state where it is no longer sustainable.

    This early collapse of the Arctic sea ice has not been predicted by the computer models, and it seems obvious to me that therefore the models must be wrong. I very much doubt that they are correctly calculating the effect that will occur when the sea ice suddenly disappears and the planetary albedo is permanently altered.

  9. tom brogle:

    Of course the fact that much of Europe has had a colder winter is no evidence that GW is slowing. The fact that a small part of the arctic is warmer than it ought to be, is conclusive of the opposite.

    [Response: See response to comment #5 above. Parts of Europe were cold, but much of North American and the Arctic were anomalously warm. While the European winter cold was not a statistical outlier, the Svalbard temperatures were indeed a statistical outlier, as shown above. – mike]

  10. Terry Haran:

    Ooops! Please disregard my previous comment. Obviously sd is the variability in degrees C, so the number of standard deviations that the delta value represents is delta / sd; hence April is about 12.4 / 2.7 =~ 4.6 or almost 5 standard deviations above the mean, and March has a somewhat higher variablility than April (3.7 vs. 2.7). Sorry for the confusion on my part.

  11. Steve Latham:

    I believe that a warming skeptic would argue against presenting the mean and variance for the area 1960-1990 as being ‘average’ or ‘normal’ and therefore a good basis for comparison.

    [Response: The anomalies reported are dramatic record breakers for the entire record back to 1911. The reference period was only used to define a representative mean and variance (neither of which are all that sensitive to which period is used to define the long-term statistics). It doesn’t influence the conclusion that the e.g. January was warmer than any previous April. – mike]

    We already know the Earth has warmed since 1975, so it shouldn’t be very surprising that some places have warmed a lot.

    [Response: You’ve sort of missed the point here. What you are stating, in different language, is what is implicit here. That the ‘null hypothesis’ of no warming looks increasingly less tenable. i.e., that events such as this reinforce the evidence that the statistics are changing in a direction consistent with what is predicted by anthropogenic climate change. – mike]

    Next, a skeptic might make point #9 that a focus on one place is unwarranted. My lack of Bayesian training/understanding leaves me by default a frequentist, and the frequentist inside me cries out for some kind of Bonferroni correction or Fisher’s method of combining probabilities over time and/or area to evaluate the result in context. Of course, neither approach is appropriate because results aren’t independent in time or space.

    [Response:Your two statements above are not really consistent. On the one hand you’re arguing that this isn’t representative because it only looks at one station (as we point out, there are good a priori reasons to look at this particular station) and yet on the other hand, that the trend here is likely indicative of larger-scale trends because of spatial correlation. The latter statement is true, and, of course, it is implicit in why we look at Svalbard, which is one of the few long-term available meteorological stations which fall within the broad region where models predict the greatest warming (e.g. as indicated by the A2 scenario simulations results shown) – mike]

    But there is something to be said for putting the result in context! Perhaps a more complete post would include evaluation of what happened at Svalbard in other recent years

    [Response: If you look at the table, its clear that what has happened in this year is remarkably different from what has happened in any past year (the record goes back to 1911), including recent years. – mike]

    , or, since the main point seems to be about variance in sensitivities among places, a comparison to results at other places would be in order (e.g. somewhere with a WMO station that models suggest will not warm very much). Sorry for whining — I’ll check the links to see if I can find the context.

  12. Caspar Henderson:

    In May 2003, a group of us with CapeFarewell saw a polar bear swimming offshore at Bear Island, some distance S of Svalbard. The captain and his crew said this was a freakishly rare event these days. See here

    A trivial and thoroughly unscientific anecdote; but at an intuitive level it struck home.

  13. pete best:

    Here in the UK the cool recent winter was explained by the North Atlantic Oscillation being weaker between the front in the Azores and the front off Iceland. Is there any evidence of the NAO being effected by GW?

    [Response: The NAO is predicted to tend increasingly towards the positive phase in some simulations of the response of the atmospheric circulation to anthropogenic forcing. See e.g. this Nature article by Shindell et al (1999). – mike]

  14. Dan Hughes:

    The figure shown in the introduction for this thread are calculated results for the period 2071 – 2100 using the IPCC TAR A2 scenario.

    What are the calculated results for December 2005 – April 2006 using the historical record as the scenario?

  15. Chip Knappenberger:

    Dear RC,

    Using maps of climate model projections of patterns of warming as a priori evidence of the temperature behavior in specific locations is an acceptable practice? Can this technique only be used in the positive sense? Or, since SRES A2 doesn’t show a relatively high warming rate for the Antarctic Peninsula (relative to the rest on Antarctica), a location often referred to as having one of the greatest warming rates on the planet, does the behavior of the Peninsula serve as evidence counter to our understanding?

    Also, I am confused about comparing Switzerland and Svalbard. How do I interpret the fact that the SRES A2 projection that you showed indicates that Switzerland should only be warming at about half the rate of Svalbard and yet both places recently experienced 5 sigma events? Do the projected patterns of warming matter? Is it the warming rate that matters or the variances or both? Based on the recent events in Switzerland and Svalbard, and using the a priori information from the A2 depiction of projected climate changes that you provided, where are some other places that we should expect to see large positive temperature events? Or, perhaps more importantly, where are some places that we should NOT expect to see them?


    -Chip Knappenberger
    (disclosure: I have, to some degree, been funded by the fossil fuel industry since 1992)

    [Response: You raise some good quesitons. With respect to your first point, note that this post did not make any claim that the data validate the model, or vice versa. There happens to be qualitative agreement, but Mike and Phil weren’t making the claim that the observations prove that the models are accurate. The lack of warming respresented in (most) models for the Antarctic Peninsula is indeed an important issue. However, modelers are well aware that they don’t get sea ice, ocean heat advection, or stratosphere-troposphere coupling right in the models in the Antarctic at the moment — largely due to lack of resolution. All of these are implicated in the particular pattern of warming and cooling in Antarctica. It comes as no surprise, then, that we don’t get the Antarctic warming right. We don’t really understand it yet, in my view. Having said that, model runs I’ve looked at do indeed show more warming over the Peninsula that anywhere else in the Southern Hemisphere — but they don’t show as much as observed. -eric]

  16. Steve Latham:

    Hi Mike,
    Re #11, I followed the global surface temperature data set link and tried to download some temperature data (first time I’ve ever tried) and got a bunch of missing data (-9999) and a lot of zeros. I wanted to look at the table of data back to 1911, but I’m embarrassed to admit that I didn’t figure out how to do it.

    I confused myself by trying to come at this like a GW skeptic (which I am not), and it wouldn’t surprise me if I sort of missed the point. I thought you were saying that Svalbard 2006 was special. I was saying that to evaluate that special-ness, there should be more context — the number of 4 and 5 sigma events (compared to ’61-’91 means) since 1992, say, over the number of such events possible is the kind of thing I would search out for confirmation. And I was thinking in terms of “On record breaking events” by Rasmus:
    (which is missing the helpful graphs, btw).

    [Response:You can see an illustration for the April month here, based on station data from Svalbard airport (other months do not look as spectacular…). The data is also available on-line from eKlima, but this web-site is unfortunately only in Norwegian (and one needs to registrer before accessing data). -rasmus]

  17. Alastair McDonald:

    Here’s the real truth about what is happening in Svalbard. It is not an isolated incident. We have passed the tipping point in the Arctic. see:

    Meltdown fear as Arctic ice cover falls to record winter low

    David Adam, environment correspondent
    Monday May 15, 2006
    The Guardian,,1774815,00.html

  18. Steve Sadlov:

    This year, the North Pacific is colder than normal and the North Atlantic / low longitude Arctic is higher than normal. How long is the Atlantic Cycle, and I don’t ask this from sheer ignorance. I ask it due to the fact that while ENSO is comparatively well studied, the Atlantic version is less studied. For example, what if there is a 40 plus year supercycle in the Atlantic? There has been lots of debate on this, but no bullet proof conclusions.

  19. llewelly:

    Related data point: In this graph one can see that over the arctic as a whole, sea ice extent is currently lower than it was at this time last year. By a related graph one can see that this has been the case for all ~4.5 months of 2006. A similar graph for the Barents Sea is available on the same site.

    I’m wondering how arctic climatologists make seasonal forecasts of sea ice extent, and where such forecasts are announced.

  20. Steve Latham:

    Dear Chip (#15), I’m going to try a quick answer before somebody more knowledgable steps in. If you are trying to question whether model projections have more often been supported or refuted by subsequent data, you will no doubt find that the answer is the former (overwhelmingly). If you want to examine the accuracy of model outputs for specific geographic locations in 2071-2100, you’ll have to wait for average temperatures to be recorded in the future. I admit that I didn’t fully understand this post, but surely you are tilting at windmills. I think the takehome message here is that as average temperatures in time and space rise, some locales (in both time and space) are likely to reach very extreme values. What do your benefactors want you to say with regard to costs borne by those forced to endure those extremes?

  21. Gareth:

    I posted this elsewhere (in the wrong place – now the right place comes along!):
    What are the model sensitivities to Arctic sea-ice melt? We’ve recently had an interesting post on the Greenland and Antarctic ice sheets, but I can’t recall much on sea ice. Specifically, what do the current models suggest is a reasonable time frame for Arctic sea ice reduction (timing and extent), and are we “on target”? I get the sense that we’re into positive feedback here, and that there’s some surprise at how fast things are happening, but perhaps I’m only reacting to alarmist media coverage… ;-)

  22. greeny:

    How is the Antarctic sea ice behaving ? Is it also showing Polar Ampification?

    [Response: See our previous article “Thank you…” for a link to an extended discussion of Antarctic trennds. – mike]

    Comment by tom brogle â?? 22 May 2006 @ 9:51 am

    could you please add a hyperlink to the “Thank you…” article
    because I cannot find it in your archives or with a search.

    [Response: I meant our most recent previous post (May 18) “Thank you for Emitting” just below this one. It has links to our previous discussions of the Antarctic. – mike]

  23. Hank Roberts:

    #20, Steve L., you’ve run together the statement in #14 (Dan Hughes) — referring to dates in the future — with the following post #15, written by Chip K.

  24. Steve Bloom:

    Re #21: The difficulty with abrupt feedbacks is that they’re a little tough to incorporate into a model before there are any observations. We saw the same problem with dynamical ice loss. There are plenty of other examples, another Arctic one being methane loss from tundra. It’s this sort of thing that makes a lot of people (including most climate scientists, I think) very nervous about even the lower-end IPCC scenarios. But regarding the sea ice, read the story linked in #17 and in particular the last paragraph. Eek, if so.

    If there were a sea ice modeler among the RC authors, perhaps we could even hope for a post on all of this. :)

  25. Dan Hughes:

    The processes that can lead to reduction in the mass of ice compared to a reference period include: (1) air temperature above the freezing temperature for longer periods of time, (2) increased water temperature for the case of floating ice, (3) increased rates of sublimation, (4) increases in the ice-air and ice-water surface area (smaller chunks of ice have more surface area than a single chunk having the same mass), (5) decreased precipitation under conditions suitable for freezing, (6) increases in the absorbed radiative flux compared to the reference period, and (7) increases in precipitation under conditions suitable for melting. There are most likely others.

    Which of these has been identified by measured data to be the dominant factor in the observed reductions in ice mass.

  26. Gareth:

    SB: “Eek, if so.”


  27. Nigel Williams:

    As we fear about sea level rise, what sort of loss of water do the models see from the global atmosphere to the vacuum of space? Is there actually significantly less available water (as surface water or ice) on the earth now than during past events, and if so would that lower the high-tide mark for a given temperature rise over that at previous events?

  28. Hank Roberts:

    Wow. From the Guardian article Alastair suggested in #17:
    “… Experts at the US Naval Postgraduate School in California think the situation could be even worse. They are about to publish the results of computer simulations that show the current rate of melting, combined with increased access for warmer Pacific water, could make the summertime Arctic ice-free within a decade….”

    Any submariners (US, ex-USSR, or others) reading and able to comment here?

  29. Hank Roberts:

    Here’s a place to look for more info — theses by graduates of the US Navy Postgraduate School, url for search is:

    I tried “sea ice” and: search found 63 titles.

    The first one is about computer modeling of arctic sea ice.

  30. Gareth:

    Further to “Eek”, if you put an ice-free Arctic into models with roughly today’s level of CO2 (or slightly higher, say ten years worth – what, around 400ppm?), what does that do to the rest of the world?

  31. Steve Latham:

    Hi Hank,
    No, I was writing to Chip (#15) about his questions. He was wondering about recent high temperatures in Switzerland and the Antarctic Penninsula, and whether they should undermine our confidence in model results. I was trying to point out that the model results are for average temperatures a long time into the future. Dan Hughes (#14) asked a more sophisticated question than did Chip (IMO), and I hope somebody will soon be able to provide a retrospective evaluation of near term GCM predictions at local scales. My limited understanding, though, is that precision in the model results is low and that only coarse regional predictions would be precise enough to allow meaningful comparisons to observations.

  32. Steve Bloom:

    Re #19: The National Snow and Ice Data Center at is a much better site for all of this, as UIUC (which is a one-person operation) has data but little explanatory material. For the Antarctic, see the British Antarctic Survey at (although the site seems to down just now). The latter is rumored to employ a part-time sea ice modeler who might be able to contribute a sea ice update post here.

  33. Steve Bloom:

    Re #28: It was the thought of this potential for cascading positive feedbacks that led to my pithy expletive. Fortunately at a certain point negative feedbacks kick in, or we’d all have to learn to speak Venusian. :)

  34. Gareth:

    Re #30: And there’s more… Reuters UK are currently reporting two papers from GRL’s next issue:

    “It means the warming is happening faster, each decade is actually warming faster than it would have,” [Margaret] Torn [Lawrence Berkeley Nat Lab] said in a telephone interview. “It’s the pace of change that will be one of the big problems. It’s how humans adapt and the cost that will depend on the rate of change of climate.”

    The problem with the negative feedbacks is that if we don’t impose them on ourselves, they may only happen long after our civilisation has gone the way of the Maya.

  35. Steve Bloom:

    Re my #33: What was #28 became #30, in case that’s not obvious.

    Re #31 (Steve L.): Just to make sure you know, Chip wasn’t wondering about any of this. He already knows the answers. Some might call it trolling…

  36. Coby:


    If there were a sea ice modeler among the RC authors, perhaps we could even hope for a post on all of this

    None of the good sea ice modelers have time to blog!
    :-) :-) :-) :-) :-) :-)

  37. Hank Roberts:

    At least they write!

    This may be the basis of the recent press report, though I’m guessing; the author’s conclusion is about summertime ice, not year-round ice conditions.

    McNamara, Terry P. Determination of changes in the state of the Arctic ice pack using the NPS Pan-Arctic Coupled ice-ocean model.

    This thesis provides 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 first dataset used is the collection of draft measurements conducted by U.S. Navy submarines between 1986 and 1999. The second is electromagnetic (EM) induction ice thickness measurements gathered using a helicopter by the Alfred Wegener Institute in April 2003. Last, model output is compared with data collected by NASA�s ICESat program using a laser altimeter mounted on a satellite of the same name.

    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.

  38. Hank Roberts:

    P.S. — why do we have that Navy submarine data available for research? The area first declassified is called the “Gore box” — the thesis noted above mentions it without explanation, and says subsequently a somewhat larger area was allowed declassified.

    For an explanation we can turn to, of all places, the explanation here:

    “not all the Arctic was analysed in this way. For security reasons relating to the Cold War, only the `Gore Box’ was involved, a roughly rectangular region of the central Arctic which the then Vice-President Gore moved to have de-classified for purposes of sea ice data analysis. The criteria for determining the boundaries of the `Gore Box’ is not known …”

  39. Paul Biggs:

    Most of the glaciers in Svalbard are of the ‘surge’ type. It is therefore difficult to use the front position of a single glacier as a climate indicator, because the front will shrink and retreat in periods between surges. The front position therefore gives little information on whether the ice mass is growing or shrinking. The dynamics of surging and the relationship of surging to climatic factors are poorly understood.

    I assume Professor Ole Humlum of the Norwegian research centre on Svalbard is real. He pointed out that glaciers there typically experience a rapid advance lasting 5 to 7 years, then retreat slowly for the next 80 to 100 years; an entirely natural phenomenon.

    I did a quick search a few weeks ago:

    “The overall total net balance is slightly negative, -4.5 ± 1 km3 yr^-1, giving a specific net balance of ca. -120 ± 30 mm yr^-1 over the archipelago. The contribution of ice caps and glaciers on Svalbard to global sea-level change is, therefore, close to 0.01 mm yr^-1 as an average value over the last 30 years, which is less negative than former estimates.”

    “For the largest icecap in Svalbard (Austfonna) the upper accumulation area shows a remarkable thickening, which, when converted to an ice equivalent value, represents a 36% increase in accumulation rate.”

    “The project also aims to improve current knowledge on Svalbard polythermal glaciers: their hydrothermal properties, drainage and behaviour in the changing climate.

    About 60% of Svalbard (36 600 km2) is covered by glaciers of various types. Although this is less than 1% of the total glaciated area of earth’s land surface, globally it is an important area to study glacier behaviour because of the short response times of the glaciers to changes in climate compared to the ice-sheets of Antarctica and Greenland. Svalbard glaciers are classified as sub-polar or polythermal in respect of their thermal structure, typically containing layers of ice that are both at and below the freezing point. This means that their structure is different and more complex than the well-studied mountain glaciers in temperate regions, and therefore their response to changes in climate is also different and more complex and needs careful studying.”

  40. Peter Hearnden:

    According to GISS April 2004 had a mean of -4.3? March 1996 was -7C?

    Whatever, they’re still just amazing figures and, fwiw, I think the analysis right.

  41. Paul Biggs:

    Here’s one I missed:

    Two ice-cored 18O records from Svalbard illustrating climate and sea-ice variability over the last 400 years
    Authors: Isaksson, Elisabeth1; Kohler, Jack1; Pohjola, Veijo2; Moore, John3; Igarashi, Makoto4; Karlof, Lars1; Martma, Tanu5; Meijer, Harro6; Motoyama, Hideaki4; Vaikmae, Rein5; van de Wal, Roderik S.W.7

    Source: The Holocene, Volume 15, Number 4, May 2005, pp. 501-509(9);jsessionid=wn42i7at0pet.victoria

  42. jhm:

    This must have been covered elsewhere, but what are the factors (as far as we think that we know them) that cool the planet after periods of warming? A related question: how possible is a ‘point of no return’ Venus scenario? and what do we think about the conditions that need to be met?

  43. PHEaston:

    Global warming risk “much higher” than already expected:

    The research will be published in the journal Geophysical Research Letters.

    Note that “The studies challenge the consensus view of the Intergovernmental Panel on Climate Change (IPCC)”, so how much should they be taken seriously?

  44. Mauri Pelto:

    It will be interesting to see the impact of this warmth on glacier mass balance on Svalbard. Two glaciers Midre LovenBreen and Austre Broggerbreen have mass balance records since 1968 and 1967 respectively. The mean annual balance for the first 25 years of the record was -0.32 m/a and -0.4 m/a. From 1993-2004 the balance has been -0.51 m/a and -0.6 m/a respectively. Averaging the loss of a half meter of ice thickness across the entire glacier each year is not a small change. Is it even going to get worse? Fortuitously a project called SLICES is underway (Changes in Glacier Geometry and Extent in Svalbard: Implications for Sea-Level Rise during the 20th and 21st Centuries) undertaken by the University of Wales, Swansea.

  45. Paul Biggs:


    Climate modeller James Annan:

    [full text replaced with link]

    [Response: You are confusing two issues. These new papers (which we’ll have a post on shortly) are not about climate sensitivity in the sense discussed by me and Annan, but about other feedbacks (specifically carbon cycle issues) that are not part of the ‘climate sensitivity’ question, but that are clearly relevant to what will happen in the future. This difference was explained in my last post on the subject: -gavin]

  46. Hank Roberts:

    The BBC article names the several research teams involved.

    As I understand the last round of IPCC models don’t incorporate (because they couldn’t quantify) feedbacks from soils releasing added CO2 (nor feedbacks from burps of methane, which no one has modeled yet, right?).
    Models incorporating these feedbacks would necessarily challenge the 2001 IPCC conclusions, eh? Presumably the 2006 IPCC conclusions will anyhow, whether they include these particular feedbacks or not.

    The whole question of how modelers could possibly include feedbacks we know exist but can’t quantify — like CO2 from massive erosion or dieoffs of plants, or methane from clathrates burping — is a bit of a nightmare, I suspect.

  47. Hugh:

    A little supplement to #43:

    which includes…

    “Both researchers emphasize that the large temperature range they predict – 1.6 to 6 degrees Celsius – does not mean that we have an equal chance of ending up with less warming as with greater warming. In other words, it doesn’t mean that the uncertainties are symmetric about an average increase of 3.8ºC.

    “People see this uncertainty and think that we have an equal probability of dodging a bullet as catching it. That is a fallacy,” Torn said.

    “By giving the appearance of symmetric feedback, people have an excuse to say, ‘Maybe we don’t have to worry so much,'” Harte said. “But while there are uncertainties in the feedbacks, all the major feedbacks are positive, meaning they would increase warming, and we know of no significant negative feedbacks that would slow warming.”

  48. wayne davidson:

    Whoaa! And I thought we warmed up (Canadian Arctic), between +3 to +6 C anomalies all winter, with the spring early by a month, and ice thinning and cracking open earlier as well. Swalbard had it much warmer. Hard to imagine, but so. The big question remains: how fast is this global warming? From a perspective of the NH Polar regions, it is hitting the North at a strong pace, agreeing with those who say that IPCC is too conservative, but the key issue of certainty will not be achieved, given complexities related to this Polar warming, it is therefore imperative to explain these related anomalies, ie Russia winter of 2005-06 without confusing some that GW is nevertheless gaining ground. For those who think that the Polar regions are small, think again, they always had a huge impact on the weather at lower latitudes.

  49. Paul Biggs:

    Re; 45

    Sorry, I was linking the 2 issues – the estimate of climate sensitivity and the added CO2 released from ecosystems. If the temperature response to CO2 is logarithmic, adding more CO2 will have diminishing effect. I also wonder if the infra-red absorption bands can become saturated?

    [Response: The logarithmic effect is already incorporated in radiative transfer models. And a “diminishing effect” does not necesessarily mean a “small effect,” if that’s what you’re getting at. Look at it this way: we all agree that the effect of doubling CO2 will be substantial, and any feedback that accelerates CO2 release will make the day of doubling come sooner, giving ecosystems and society less time to adapt, and requiring tighter emissions targets if one is to avoid doubling. The answer to your question at the end is “no.” At least not in any sense that should concern us in the least. The CO2 bands are still not completely saturated even for Venus, insofar as the effect on the top of atmosphere radiation budget goes. There are still bands that are relatively transparent to infrared and can be strongly affected by adding more CO2. In fact, the question of saturation doesn’t even completely make sense as applied to greenhouse warming, because even in an optically thick atmosphere the radiation that escapes to space always comes from the high (and generally cold) tenuous part of the atmosphere, where the bands are by definition not saturated. In that sense you never reach a point where adding more CO2 stops increasing surface temperature. –raypierre]

  50. Hank Roberts:

    Gavin, may I suggest a new term — “climate reactivity” — for effects of those factors not well understood, not incorporated into models, and not currently part of climate sensitivity calculation.

    The term has analogous uses in chemistry, nuclear physics, and even psychology/psychiatry.

    It always means those factors that may not be well understood and can cause unexpected trouble.