Arctic Winter

It was pointed out in comments recently that Wang et al. (2012) found a cooling trend during the winter season in the Arctic (defined as the area from latitude 60N to the pole) from 1982 through 2004, using estimates of surface skin temperature from AVHRR (Advanced Very-High Resolution Radiometer) instruments aboard satellites.


In the last post we saw that none of three other data sets exhibit winter Arctic cooling. However, two of those data sets (satellite data from RSS and UAH) were measures of lower-troposphere rather than surface temperature, and for all three the time span studied started earlier than 1982 and ended later than 2004. Furthermore, the region studied wasn’t exactly the same, since I used NCAR/NCEP data from north of the arctic circle, and the satellite data sets from RSS and UAH miss the very-near polar region (although that’s only a small fraction of the Arctic). It may be fruitful to examine data for surface temperature only, for the same time span studied in Wang et al., for the same region.

To that end, we’ll examine the third data set used previously, the NCAR/NCEP reanalysis data, but this time we’ll use the data from latitude 60N to the pole. We’ll also look at the Arctic temperature estimates of zonal temperature from NASA. This covers the region from latitude 64N to the pole, and is only annual averages, but it won’t be the main source for comparison, I’ll just use it for a “reality check.”

Some people mistakenly regard “satellite” as a magic word, indicating that its data are unimpeachable, or at least clearly superior to other sources. One need only compare the data from RSS and UAH to dispel that notion. Despite the immense effort invested to correct for such issues as instrument calibration, sensor drift, and orbit variation, the two records — derived from exactly the same satellites — are in notable disagreement. It’s also well to bear in mind that AVHRR is not a satellite, it’s a type of instrument, one which has changed over the years, and that the AVHRR record is not from a single satellite but from instruments flown aboard 15 different satellites. The complications inherent in piecing together data from so many different satellites are enormous (as the RSS and UAH people can testify). And AVHRR doesn’t measure temperature, it measures reflectance in a variety of wide spectral bands; transforming that to surface temperature is a nontrivial effort. Its temperature estimates cannot be regarded as gospel, but only as estimates.

The NCAR/NCEP reanalysis data are the output of a computer model of the atmosphere, but one which is driven (and essentially controlled) by direct observations from surface stations, weather balloons, and other sources. Because it is driven by direct observations of relevant variables (like temperature measurements from thermometers), I regard it as at least as reliable as AVHRR data.

Here’s the data used in Wang et al. (temperature data in blue), for each season and annual averages, from 1982 through 2004 inclusive (winter is Dec-Jan-Feb, spring Mar-Apr-May, summer Jun-Jul-Aug, autumn Sep-Oct-Nov):

I digitized the graphs in order to obtain the numerical values used for temperature. I also computed seasonal and annual averages for the NCAR/NCEP data. First, here’s the comparison of all four seasons for the time span studied (NCAR in red, AVHRR in blue):

There’s quite a seasonal difference. The AVHRR data are considerably warmer in summer and colder in winter, hence show a much larger seasonal cycle. Clearly there’s notable disagreement between the two. In fact, here’s the difference between them:

Note the strong annual cycle in the difference, which has gotten larger over the years.

If we compare winter (DJF) season averages from the two we see that NCAR/NCEP shows a warming trend while AVHRR shows cooling:

In fact the difference has grown rather steadily:

In spring the two records are much closer to each other, not only their absolute temperature estimates but their trends as well:

In summer we find AVHRR hotter than NCAR/NCEP:

By autumn the two again show very different trends, with NCAR/NCEP indicating faster warming that AVHRR:

An interesting comparison is their estimated annual averages:

The AVHRR data show less, and less steady, warming. It also shows greater year-to-year variability. We can compare these annual averages to those from NASA GISS, which are based on actual temperature measurements but must interpolate over much of the Arctic (GISS data shown in black, offset to have the same mean value as the combined NCAR/AVHRR data):

The GISS data are in much better agreement with NCAR/NCEP than with AVHRR, especially the most recent years. I regard this as reason to put more confidence in the NCAR/NCEP data than the AVHRR data.

We can even extend that comparison to a longer time span, although I don’t have the AVHRR data outside the 1982-2004 interval:

The agreement between GISS and NCAR is outstanding, while that with AVHRR is poor. Again, this gives me much greater confidence in the NCAR/NCEP data than in the AVHRR data. It also indicates continued warming of the Arctic from 2004 to the present.

Considering this, and the results from both the RSS and UAH satellite data sets for lower-troposphere temperature, I regard the claim of winter cooling in the Arctic (60N to the pole) to be very implausible. I don’t regard it as impossible. There’s a chance — small I think, but not zero — that the limited amount of direct observations in the very northernmost part of the Arctic makes the interpolation/model simulation overestimate wintertime temperature dramatically. It’s also possible that “skin temperature” estimated by AVHRR is sufficiently different from surface air temperature (reported by NCAR/NCEP and NASA GISS) to account for the differences. But on the whole, when it comes to winter cooling in the Arctic I’d say I just don’t believe it.

Mention was also made of a paper by Cohen et al. titled Arctic warming, increasing snow cover and widespread boreal winter cooling, in the context of possible Arctic winter cooling. But Cohen et al. doesn’t really claim Arctic cooling in winter, they claim boreal (i.e., northern) winter cooling over land areas, mainly eastern North America and northern Eurasia. That’s a different subject, one which I may look into soon.

About these ads

74 responses to “Arctic Winter

  1. I think Watt’s has shown the urban-heat effect quite well. The cooling Arctic is clearly a result of previous weather stations having been sited near Santa’s workshop. Seems to have as much validity as some of the counter arguments anyway.

  2. The AVHSRR samples in two channels around 11 and 12 micrometers to detect skin temperature. At those wavelengths, the atmosphere is largely transparent (except for clouds) while the surface has high emissivities. There are differences in surface emissivities, however. Most notably, fine snow has emissivities of 0.992 and 0.99 for 11 and 12 micrometers respectively, while medium snow has 0.99 and 0.98, while coarse snow has 0.98 and 0.96 (figures eyeballed of a graph). That means a progressive shift from finer to coarser snow would be difficult to distinguish from a gradual cooling trend. Water as emissivities of 0.99 and 0.98, so a shift from coarse snow to open water would emulate a warming trend. The piece of the puzzle I am missing is whether warmer weather results in finer, or coarser snow. (Emissivities from Wilber, Kratz and Gupta, “Surface Emissivity Maps for Use in Satellite
    Retrievals of Longwave Radiation” (1999) which is available for download from NASA.)

    Having said that, it is not obvious to me that the enhanced greenhouse effect predicts winter warming in the extreme Arctic. Arrhenius calculated that the peak warming due to the Greenhouse effect was in the mid-latitudes, but that it moved further north with increasing CO2 concentrations. By his calculation, the maximum effect is at 40 degrees north with 0.67 or then CO2 levels, shifts to 60 north at two times CO2 (approx 600 ppmv). He suggested that the peak warming, including feedbacks, would be further north than his calculations suggested because of the effects of melting snow and ice. I do not know how well those predictions are confirmed by climate models.

    However, based on that information, in Winter in the very high Arctic there is no local feedback from melting snow or ice, and heat must be carried in by air or ocean currents. That from air currents will cool more efficiently in transit with higher CO2 and just possible could result in winter arctic cooling.

    I would be very interested on the opinion of somebody familiar with the models on these points.

    • “The piece of the puzzle I am missing is whether warmer weather results in finer, or coarser snow. ”

      It’s not a scientific result, but FWIW, the experience of my snow-country childhood and young adulthood says, unhesitatingly, “coarser.”

      • David B. Benson

        I agree, from a different snow country childhood and young adulthood.

      • Surface melting makes for coarser snow. Plus which we have the issue of black carbon contamination which will not as much effect the summer/fall temperature trends because much of the snow is melted.

      • Horatio Algeranon

        Well, “hoar frost” is a type of very coarse “snow” (ice crystals, actually) that can form on the surface on very cold nights (under the right conditions of atmospheric moisture).

        Horatio knows about this because he used to do a lot of back-country skiing (in Utah) and dug snow pits to look at the layers to decide if a slope was safe.

        When hoar frost gets buried, it can lead to avalanches because it acts like a layer of ball bearings that allows the snow above to slip (sometimes as a slab).

      • “Hoar frost”–I remember winter hiking in the North Carolina mountains a few years back; at mid altitudes there was the lightest of breezes, and the lee sides of every twig and branch had sprouted rime ‘feathers.’ It was breath-taking. (So, in a different sense, was the ferocious chilling wind at the top of the saddleback.)

        Nonetheless, if you are talking regular snow, warm temperatures–’warm’ as in ‘just below freezing’–can often give you big, fluffy photogenic flakes. (Also great packing snow.) Below zero (F, of course) you tend to get much smaller flakes. And they are dry–a mass of snow at these temps just won’t pack to form a snowball, unless maybe you warm it in your hands (not recommended.)

        And Eli is right–the south side of a March snowbank at 4 PM on a sunny day has snow that is strongly granular, even icy. Cf., “corn snow.”

        http://en.wikipedia.org/wiki/Types_of_snow

      • Slightly OT, but us who live in snowy climes, all ahve this tendency! I can tell not only the general level of humidity by the type of snow, but also the temperature, by walking in it: “Squooshy” = pretty much anything from OC to 5C; *packy-y” = -2C to -10C; “crunchy” = -15C to -20C, and then on down, it becomes various stages of “squeeky” snow. It’s said that the Inuit have many different names for different types of snow, but Coloradans rival that!..;)

      • Well, in the UK, we have snow depth grades..

        - 0 – 2mm -> ‘Train Stopping’ grade
        - 2 – 4mm -> ‘Road chaos’ grade
        - 4 – 8mm -> ‘All transport inoperative ‘ grade
        - >8 mm -> ‘Country shut down for the duration’ Grade

        (Although before you gloat, try surviving 8 weeks of continuous grey drizzle with your sanity intact. You may feel the urge to go and invade a country with a better climate)

  3. Philippe Chantreau

    I understand Norman’s wish. It would be so nice if the Arctic could be used as the planet’s AC unit. Wouldn’t have to worry about a thing. Ah, going back to blissful irresponsibility, instead of the guilt ladden version…

  4. There is a large misunderstanding of Arctic winter, as much as snow falling on its ocean with the flakes not melting because sea water is colder than 0 degrees C.

    With winter comes boundary layers, inversions, in darkness the Upper air immediately above the surface was usually warmer, not colder, than the surface. As AGW weakens and reduces winter, these boundary layers
    vanish, temperature profiles change from having a maxima above ground to maximum temperature on the ground The illusion of “cooling” in the lower atmosphere is demystified further by this example:

    October 16 1980 barrow Alaska sea ice a plenty:

    http://igloo.atmos.uiuc.edu/cgi-bin/test/print.sh?fm=10&fd=16&fy=1980&sm=10&sd=01&sy=1980

    Upper Air data (with inversion)
    Pressure altitude temperature
    989.0 4 -7.9
    850.0 1210 -1.9

    October 16 2012 Not much sea ice about (no inversion):
    Pressure altitude temperature
    1000.0 4 -2.9 -8.9
    850.0 1276 -13.9 -18.9

    The surface is warmer but the lower upper air is colder, satellite data must be complemented to establish a contrast.

  5. So it seems we have come full circle.

  6. Tamino, thank you for another great post, and thorough analysis of the various data sets out there regarding winter Arctic temperatures.

    Regarding winter cooling the Boreal forest area, this NCEP/NCAR re-analysis anomaly plot of the 2002-2012 winter temperature above 50Nin winter (DJF) indeed shows that Boreal area in Siberia experienced moderate cooling over the winter during the past decade, which is consistent with the assertions made by Cohen et al 2012 about changes in the AO and the Siberian High (pressure zone) during winter :

    http://web.mit.edu/jlcohen/www/papers/Cohenetal_ERL12.pdf

    Increase in the Siberian High they argue is the result of increased fall snow cover which is in itself caused by increased precipitable water in the warmer fall atmosphere due summer sea ice decline. Which was reported by Wang et al 2012 as well.

    Notably from the NCEP/NCAR plot above is also that anything over 70N is definitely warming in winter, and quite significantly so, as would be expected from an Arctic with thinner ice which is growing faster, thus shedding excess heat that it obtained from the albedo effect due to sharply reduced snow and ice cover experienced during the summer melting season.

    So it seems we have come full circle.

    • Rob Dekker,

      I used the same grapic from the NCEP/NCAR that you had posted. I changed your date to 1980 to 2004 (same date as Wang study) and made a new graphic. The arctic winter was indeed cold during this time frame very close to the Wang finding.

      [Response: I think you’re misunderstanding something.

      The map you refer to (his map, changed to 1980-2004) shows “cold” because it was colder than the 1980-2010 climatology (the baseline period). For winter 1982-2004 to be colder than winter 1980-2010, it must have gotten considerably warmer during the 2004-2010 period, which suggests that the Arctic is getting warmer during this season, exactly the opposite of what Wang claims. Also: a map of anomaly *during* the 1982-2004 period tells us nothing at all about the *trend* over the 1982-2004 period.

      I will also point out that those maps are for Dec. through Mar., whereas Wang et al. defines winter as Dec. through Feb.

      If you want to know what NCEP/NCAR data says about the *trend* in the Arctic winter during the 1982-2004 time span, read this post. I did that. It got warmer.]

      • Sorry about missing that I thought you had been using the other climate trends like GISS. On second look you did use the NCEP/NCAR data.

        My question is what is this data based upon? I was reading the page a bit and it seems it is a complex model based upon what has happened. Is it actual measurement?

        I have rread in places that Arctic temperatures are not eay to obtain directly because of limited measuring devices. I think now tere are floating bouys that collect data in real time.

      • Sorry Tamino, Norman, for entering the incorrect end month in the NCEP/NCAR plot.
        Here is the correct anomaly plot of the 2002-2012 winter (Dec-Jan-Feb) temperature above 50 N, by NCEP/NCAR.

        This shows the widespread winter warming that Tamino reports in this post, stretching not just over the entire Arctic ocean, but also deep into Northern Canada and even into the Northern US.

        Despite widespread warming in winter, the point made by Wang et al (as pointed out by Norman) is still valid : There is a distinct area over Central Siberia that shows a moderate cooling in winter. Cohen et al recognized that cooling of the Siberian Boreal forest area, and attributed it to increased snow cover in fall causing an increase in the Siberian High (pressure zone), which is in turn caused by reduced sea ice cover over the previous summer.

        The point I wanted to make with this plot is that overall the Arctic may be warming, as presented by Tamino here, and as one would expect with an increased GHG effect, as well as a winter warming caused by thinning Arctic sea ice pack shedding more heat, but there will always be regional effects that defy the overall trend.

  7. The world is round, isn't it?

    I agree that the cooling of Arctic winters is very implausible. The meting of sea ice and ice caps is evidence that the winters are warming and they are doing so at an accelerated pace. We are currently in the “icehouse’ phase, in an interglacial period. The natural variability between glacial and intergalcial cycles shows that we should soon be entering a glacial cycle. However, accelerated warming due to greenhouse gases is inhibiting this natural process. General trends show that warming will increase temperatures, enhancing the overall winter temperature in the Arctic.

  8. Michael Sweet

    The DMI graph of temperatures from 80 N to the pole are here. I gave them the old eyeball treatment and it does not appear that they show cooling from 1984 to 2004. There is a lot of noise that makes eyeballing hard to do in that time frame. From 2004 to the present the temperature in winter clearly increases, visible even over the noise. Are you aware of a graph of that data that shows what their result really is?

    [Response: One should be very cautious drawing climate-change conclusions from DMI temperature data, because it’s not just the output of a computer model, but of *different* models for different time periods, which are substantially different and not aligned to each other in order to eliminate bias.]

  9. Tamino,

    I read through your post too quickly the first time, I have reread it a couple times. I do like your in depth analysis of the topic you choose to investigate.

    You are comparing the NCAR/NCEP with the AVHRR and the GISS data set for Arctic temperature and remark how GISS and the NCAR/NCEP match.

    You point out that the NCAR/NCEP data is driven by direct observation of surface stations. Wouldn’t this data be the GISS data set so the two should be very close to each other if they are using the same source as your last graph of the OT shows?

    [Response: To a large degree yes. But NCAR also includes other data, including Arctic buoys, balloon-borne observations, etc., even the surface station data is not exactly the same, and assimilates them into a weather model. However the point is valid, that the data overlap will contribute to their agreement.]

  10. I was thinking of a simplistic analogy of why temperature alone may not tell the whole story. Even if the Arctic area has warmed in the winter months it doe not negate an overall cooling effect for Earth.

    If you have two houses in a cold winter. One has no insulation the other heavily insulated. If you heat the non insulated house to a temperature warmer than the insulated house you have a higher temperature in this house but it is losing a lot more heat than the insulated house.

    In the Arctic analogy. If warmer water or air moves into the region during winter, the temperature will rise but because of the radiation law of the 4th power emission based upon temperature, this will actually have a cooling effect than if the warmer water had not moved into the winter arctic.

    As Phil posted in the “Snow + Ice” thread, FYI that is thin will conduct a lot more energy than thick MYI. This conduction can warm the air but overall the system is losing more energy than if it had the well insulated thick ice cover keeping the energy of the ocean below in.

    I am still not sure of how an ice free Arctic would act overall for heat balance. In the summer if thick clouds form it would act similar to ice to reflect a large portion of the solar radiation so the heating would become limited at some point. Then in the winter all that open ocean would be radiator, radiating much more energy into space than if a thick ice insulating layer were present. How this balances out I do not know, but that is the question I have been asking.

    [Response: I don’t have the data to “do the math” on that one. But I do know that the sea ice loss in winter is less than in summer, and the research I’ve seen universally contradicts the idea that sea ice loss could have a net cooling effect. Until I see something other than vague speculation, I’m inclined to trust both the research of climate scientists and the implications of simple energy-balance models.

    You might get a much more informed opinion if you post your question at RealClimate — http://www.realclimate.org — post it on the latest “open thread.”]

  11. Norman said

    Even if the Arctic area has warmed in the winter months it doe not negate an overall cooling effect for Earth.

    Norman, nobody claimed that the Arctic in winter negates an overall cooling effect for Earth. In fact, nobody even claimed that the INCREASE in spring/summer/fall AND winter temperatures did not contribute to overall cooling of the planet. In fact, it most certainly did.

    With all due respect, but Tamino has now presented 4 posts which not just qualify, but actually quantify the effect of albedo changes due to ice and snow during Northern Hemisphere summer, as well as quantify the winter warming that follows.

    So, instead of putting up a strawman argument, the question you SHOULD be asking yourself is :

    If an more ice-free Arctic is warming in all four seasons, why do I think that it can cool the rest of the planet in ANY season ?

  12. To be more specific : If the coldest area of the planet warms, it will radiate more IR to space, and thus “contribute to cooling”.
    But it still makes the planet warmer.

  13. I used the ECMWF ERA-Interim reanalysis data (3-hourly data) to produce the same type of temperature data, going from 1979 to 2011. The results are:

    Year, DJF, MAM, JJA, SON, Year
    1979,-20.558,-11.602,5.43,-7.471,-8.973
    1980,-20.024,-10.623,5.601,-7.606,-8.306
    1981,-21.299,-10.34,5.776,-6.43,-7.546
    1982,-21.378,-11.351,5.36,-8.268,-8.88
    1983,-20.557,-11.125,5.595,-7.165,-8.38
    1984,-21.313,-11.01,5.824,-7.284,-8.282
    1985,-20.432,-11.463,5.619,-6.911,-8.302
    1986,-22.003,-11.006,5.444,-7.329,-8.455
    1987,-21.188,-11.45,5.431,-7.955,-8.952
    1988,-20.856,-10.539,5.971,-7.229,-8.033
    1989,-21.603,-10.333,5.899,-7.731,-8.227
    1990,-21.682,-8.726,6.197,-8.002,-8.108
    1991,-21.198,-10.545,6.265,-6.721,-8.003
    1992,-20.721,-10.839,5.195,-8.345,-8.731
    1993,-22.248,-10.467,5.879,-7.457,-8.226
    1994,-20.111,-10.636,5.879,-7.526,-8.493
    1995,-20.504,-9.433,6.069,-6.087,-7.295
    1996,-21.445,-9.666,5.471,-6.84,-7.763
    1997,-22.358,-9.672,5.748,-6.675,-8.043
    1998,-21.396,-10.229,6.316,-7.198,-8.157
    1999,-20.293,-10.891,5.671,-6.97,-8.318
    2000,-21.04,-9.845,5.923,-6.978,-7.831
    2001,-20.739,-10.955,6.17,-6.355,-7.792
    2002,-19.835,-9.643,6.078,-5.937,-7.493
    2003,-20.293,-9.22,6.299,-5.405,-6.959
    2004,-19.688,-10.137,5.924,-6.274,-7.865
    2005,-18.991,-8.93,6.118,-5.287,-6.592
    2006,-18.903,-9.358,6.134,-5.638,-6.877
    2007,-19.73,-8.997,6.408,-5.363,-6.696
    2008,-19.911,-9.645,5.988,-5.772,-7.299
    2009,-19.777,-10.216,6.113,-5.678,-7.366
    2010,-19.605,-8.533,6.435,-5.093,-6.703
    2011,-18.535,-8.945,6.324,-5.381,-6.706

    These values are all area-weighted (on 0.75deg. lat/lon pixels) from 60N to 90N. The yearly values are calendar years, for each year the January and February values used actually refer to the following year.

    Trends are (fitted with gnuplot, must really learn how to use R):

    DJF: 0.058deg.C/year +-0.014
    MAM: 0.064deg.C/year +-0.011
    JJA: 0.024deg.C/year +-0.0045
    SON: 0.076deg.C/year +-0.011
    Year: 0.056deg.C/year +-0.0088

    Errors are “Asymptotic Standard Error” values on the slope.

    I think these values are in quite good agreement with the NCAR/NCEP data. In particular, it is interesting that the smallest trend is in summer, which, as far as I can eyeball, is also the case in Tamino’s graphs.

  14. Horatio Algeranon

    For both the NCEP/NCAR and surface temp data, the “overall” warming rate (based on linear trend in annual anomalies) in the arctic since the early 80′s is roughly twice the global rate.

    But for the AVHRR data, the overall warming rate in the arctic is about the same as the global rate.

    See ANALYSIS OF LAND SKIN TEMPERATURE USING AVHRR OBSERVATIONS (BY MENGLIN JINl) for global trends (1982-1998) as well as a discussion of the potential problems/errors associated with AVHRR.

  15. Rob Dekker,

    That is what I was trying to find out. If a warmer arctic will actually make the planet warmer. If you note the data below your post it shows summer is not warming as rapidly as winter. That could be the cloud feedback limiting the summer heating.

    The warmer winter does not mean net global warming. The warmer winter will emit more IR than previous years. That is why a radiation balance is more important than a temperature trend. The Sahara desert may have warmed more at night than previous years yet it still remains a cooling area as it radiates away more energy than it received during the day.

    Thanks for all your input to my questions.

  16. Tamino will certainly find this paper interesting from a statistical perspective

    http://wattsupwiththat.com/2012/10/17/new-paper-cuts-recent-anthropogenic-warming-trend-in-half/

  17. Okay, I read the WUWT post and posted in reply: “Including the AMO as an explanatory variable is a mistake. My own work indicates that global temperature Granger-causes the AMO, not the other way around. Their regression is spurious.”

    Let’s see if they remove the comment, or just flood the page with idiot pseudoscience responses.

    • “Let’s see if they remove the comment, or just flood the page with idiot pseudoscience responses.”

      Alex, I’ll pick the latter, for $300…;)

    • Which version of the AMO did you use? I’m just curious whether you used versions meant to deal with contamination of the anthropogenic signal for the causality analysis. I don’t think using a global record to compare is really appropriate – rather Arctic or a specific portion of the North Atlantic.

      • On the issue of contamination, I see that Zhou and Tung used the linearly de-trended AMO SST as the regressor. The thing that bothers me a lot is that the nonlinear anthropogenic signal can still be embedded in the AMO index, and thus when you remove the influence of AMO you in fact remove the nonlinear signal (which perhaps explain why the residual compares poorly with the anthropogenic forcing.)

        I wonder if the analysis will look significantly different if they used an index that properly removes the global signal.

      • IanC, I just noticed that you found the same problem with Zhou et al as I did, and you found it before me. Could you please keep an eye on that “staatvanhetklimaat” post, in case people respond that misunderstand the issue ?

    • Barton,
      Thanks for you post !

      The origin of this article is from journalist Marcel Crok, here :

      http://www.staatvanhetklimaat.nl/2012/10/17/new-paper-cuts-recent-anthropogenic-warming-trend-in-half/

      Now I found the conclusion of the paper “for the past 100 years, the net anthropogenic trend has been steady at approximately 0.08 °C/decade” interesting, so I investigated it a bit.
      It seems to me that they are a bit sloppy regarding their attempts to remove the AMO (as an oscillation) from the temperature data, and concequently, I made a comment on the paper’s method in the link above, which goes like this :

      ——–
      I am not an expert in statistics, but something does not quite feel right with this paper.

      Here is the problem :
      The AMO itself is defined as the (Kaplan SST) of the Nothern Atlantic, with the LINEAR trend removed.

      This means that the AMO index not just contains any cyclical signal (if it’s there) but also the NON-LINEAR part of the underlying warming signal.

      So, if you then remove the AMO index from the global temperature record, by regression, you have just removed not just the syclical part you wanted to remove, but also the NON-LINEAR part of the (antropogenic) global warming signal that you were looking for in the first place !

      The method is designed to remove any non-linear component in underlying warming trend, by averaging it out to linear.

      No wonder they obtain a near-linear end result.
      ———

      Anyone : Please let me know where my line of thought is wrong, and if it is not, then do the conclusions of the paper still stand ?

    • I found it rather interesting that a grad student in Applied Mathematics (Zhou) apparently claims that AGW over the past couple of decades apparently is half the amount that climate scientists assert to be consistent with observations. All this based on AMO regressions.

      Tamino already suggested that one should be very careful using the AMO index for regressions

      http://tamino.wordpress.com/2011/01/30/amo/

      and the fIaws in Zhou and Tung were already pointed out by IanC above.

      Still, I dug a bit deeper, and found some interesting results, which I posted on the original “staatvanhetklimaat” site :

      http://www.staatvanhetklimaat.nl/2012/10/17/new-paper-cuts-recent-anthropogenic-warming-trend-in-half/#comment-2724

      —–
      The flaw I pointed out above in this paper by Zhou and Tung is pretty basic, and if valid, challenges the conclusions, and potentially nullifies them.
      Could you please ask Jiang Zhou for a response to my comment ?

      Also, could you please tell us a bit as to how this publication came about ?

      For example, which criteria did you use to pick this particular paper for your post ?
      And which scientists or other experts did you consulted on on Zhou & Tang statistical method before you decided to publish this piece ?

      By the way, isn’t the co-author (Prof. Tung) the Chief Editor of the journal in which this paper was published ?
      ——
      Comments ?

    • I posted this latest response on Zhou and Tung 2012 on both WUWT and on the original “staatvanhetklimaat.nl” site,
      ——
      Not sure WHO will have his knickers in a twist on this paper, but certainly not Tamino.
      This paper by Zhou and Tung is classic case of ‘extreme’ “contamination” in regression analysis.

      When you want to remove a known variation in a climate forcing from the global temperature record, you always have to be very careful not to mix cause and effect. After all, if your ‘forcing’ contains a global warming signal, then that too will be eliminated from the global warming record !

      In extreme cases, if your ‘forcing’ (cause) mainly consists of the global warming signal itself (effect), then you end up with a flat line. If you subtract a trend line from your forcing, you end up with a trend line. And that is exactly what happened in the case of Zhou and Tung.

      The global temperature signal is VERY strong in the AMO index they used :
      Check the Northern Atlantic SST record against the global temperature record here :

      These two records are clearly highly correlated (anyone want to determine the R^2 on this?).

      So, when Zhou and Tung ran their regressions to eliminate their AMO index, they eliminated not just some minor Atlantic SST anomalies, but also most of the global temperature record from the global temperature record. So, they were left over with mostly the trend line in the Northern Atlantic SST record. And remember that that is the trend line was removed BY DEFINITION to obtain their AMO index.

      Also it is not as if the trap (of contamination due to mixing up cause and effect) that Zhou and Tung fell right into is new or unknown in the field of climate science.

      In fact, there have been dozens of papers pointing out the problems with contamination in multi-regression analysis, and specifically to the AMO, attempt to avoid the problems, and including methods to avoid it date way back :
      For example, Mann and Park, 1994 discuss the issue and propose a multivariate signal detection procedures to tease oscillatory patterns apart from long-term (potentially non-linear) trends. Or Schlesinger and Ramankutty, 1994 who use climate model-based estimates of forced trends to estimate a possible residual oscillatory component instead of linear trends.

      And then Meehl et al., 2004; Barnett et al., 2005; Hansen et al., 2005 all mention that to properly deal with purely Atlantic variability, it is highly desirable to remove the larger-scale global signal that is associated with global processes, and is thus related to global warming in recent decades, from the AMO.

      Also, Trenberth and Shea, 2006 clearly explains the problem of global warming signal being present in the NA SSTs, and suggests a “alternative AMO” definition that at the very least subtracts the global SSTs from the Northern Atlantic SST record (similar to my Tamino’s suggestion to subtract GISS, but probably global SST is better, since it deals with ocean temperatures).

      Needless to say that Zhou and Tung’s conclusions that “There is no statistical evidence of a recent slow-down of global warming, nor is there evidence of accelerated warming since the mid-20th century.” as well as “the net anthropogenic trend has been steady at approximately 0.08 °C/decade”, as well as Marcel’s title of this post are incorrect and artificially created by Zhou and Tung defining the AMO as “linearly detrended”.

      You are simply looking at the 100 year linear trend line in the global temperature record, which is indeed something like 0.08 C/decade.

      Now, if I (as an amateur) can see that there is a problem with this paper in 10 minutes, and an hour to recognize what the problem is, and a couple of hours to find out the background and prior work on this issue, I wonder why was Dr. Tung, with 30 years of experience in this field, not able to do so ?
      —–
      I hope it makes a difference somewhere, sometime.

  18. snarkrates,

    I agree. The TOA balance for the Arctic would be the important factor to determine if losing sea ice will increase the cooling effect of the Arctic or lessen it. That is what I would like to know. Less ice in summer increases the energy content of the area (lower albedo). Will less insulation (less ice coverage and thinner ice that allows greater heat flow, warming the air but allowing more radiation loss) in the winter allow more energy to escape in winter than previous years?

    • The arctic must shed more energy as heat flows increase — otherwise you’re stuck predicting that the arctic will warm infinitely, which is nonsense. The heat loss from the arctic can be in various forms, including by evaporating ocean water and precipitating it in Britain, but globally it can only be by radiating out to space (plus a trivial amount in chemical processes).

  19. Norman,

    1) Although the Arctic is warming, it is still cooler than the rest of the planet (excluding Antarctica). That means a warming Arctic reduces the temperature differences between different regions of the planet. Because IR radiation increases with the fourth power of temperature, the more even in temperature the planet’s surface, the higher the global mean surface temperature must be to maintain an overall energy balance with incoming solar radiation.

    2) Even if ice cover is replaced by cloud cover thereby maintaining albedo, clouds have a significant greenhouse effect so that the result will be a net warming. Albedo will be constant (on this supposition), but the greenhouse effect will increase. Of course, the original supposition is unwarranted in any event. Cloud cover my increase, but not to 100% so that albedo will decline in any event.

  20. The tenacity of the participants here is appreciated. But, speaking from long experience at SkS, Norman will only see and understand that which accords with his preconceptions.

    • Daniel Baily,

      A nice putdown. I am asking a question to see if any have the answers. If I am too narrow minded to understand them, my loss not yours. Maybe with your many years of experience in the topic you have the answer or know of an article that does answer the basic question.

      Will less ice and snow in the Arctic radiate away more, less or the same energy than previous years with more ice and snow in that region?

      [Response: Three things:

      1. When sea ice gives way to open ocean, the TOA albedo changes by about 0.2. The insolation during summer is about 500 W/m^2. So, the absorbed energy *on that area* increases by about 100 W/m^2. I seriously doubt that the radiated energy during winter increases by such a large amount.

      2. Absorption happens throughout the atmosphere and on the ocean, so TOA albedo is the relevant quantity. But emission to space does not take place from the surface. It happens in the upper atmosphere. So the relevant temperature change which determines the change in emission is the change in the emitting layer of the atmosphere. By no means is it as simple as “ice radiates almost nothing while ocean radiates a lot.”

      3. You really should pose your question at RealClimate:

      http://www.realclimate.org/

      The moderators are professional climate scientists, and the readership is very knowledgeable.

      Finally: I agree that the criticism of your inquiry has been too severe. But you do seem to argue that “net cooling from ice loss” is not just plausible but likely, in spite of the fact that there’s no support for it in the literature and you haven’t done the math.]

  21. Hi Tom,

    Missed your intelligent and thoughtful replies to my posts on SkS.

    Your point 2 is not correct. Thick clouds warm at night but cool during the day. The Arctic in summer is under the sun 24/7. Cloud formation at this time will have overall cooling effect.

    http://scienceblog.com/1227/clouds-mitigate-effects-of-warming-on-arctic/

    http://isccp.giss.nasa.gov/role.html

    That was not my major point anyway. My question started when Tamino calculated the energy increase caused by the melting snow and ice in the Arctic. He calculated a 0.45 watt/meter increase in energy when spread over the globe. My question was what happens in the winter (zero solar radiation) when you remove what acts as an insulator (ice and snow). The thick ice prevents the warmer ocean below from exchanging energy with the surface and radiating it away. If you remove that thick ice will that area radiate away more, less or the same energy during the winter months?

  22. Tom Curtis said :

    Even if ice cover is replaced by cloud cover thereby maintaining albedo, clouds have a significant greenhouse effect so that the result will be a net warming.

    To which Norman said

    Thick clouds warm at night but cool during the day. The Arctic in summer is under the sun 24/7. Cloud formation at this time will have overall cooling effect.

    Tom’s point is that IF clouds replace open ocean, then the albedo will stay the same, but IR radiation to space will reduce, since top of clouds is typically colder that the surface.

    Norman’s point is that in general, clouds cause a cooling effect during the day, and because the Arctic in summer is under 24/7 sushine, clouds in summer will thus have an overall cooling effect.

    So who is right ?

    [Response: Note from Hudson’s graph that clouds over open ocean actually have *lower* TOA albedo than clear sky over sea ice, so ice loss causes albedo reduction even if 100% clear sky is replace by 100% cloud cover. Also, cloud altitude impacts their relative contributions to albedo change and IR absorption.

    I think you’d need a good forecast of cloud changes to know how that affects warming. But the sea ice loss all by itself seems to be clearly a warming influence.]

  23. Norman said :

    My question was what happens in the winter (zero solar radiation) when you remove what acts as an insulator (ice and snow).

    I’m confused, Norman. I thought that you were asking about the radiative balance in the Arctic. Not about an artificial removal of ice and snow in winter.

  24. Norman said

    If you remove that thick ice will that area radiate away more, less or the same energy during the winter months?

    Sorry, but I’m confused about this question as well.
    Is “If you remove that thick ice” an experiment ? Or an observation ?
    And when you ask “will that area radiate away more, less or the same energy during the winter months?” then compared to WHAT exactly ?

    There are many knowledgeble people on line here at Tamino’s, who can provide amazing insights and answers, as long as you can state your question unambiguously. Can you, Norman ?

  25. Rob Dekker,

    If you check out the first link in my reply to Tom Curtis, it shows that clouds in summer have slowed down the warming at a 2 to 4% increase in cloud cover.

    Tamino pointed out that 100% cloud cover in summer still has a lower albedo than ice so it will warm more relative to ice. But relative to no ice or clouds the clouds will considerablely slow the potential warming.

  26. Rob Dekker,

    I think is is self explanatory. If the multiyear ice melts away (no summer ice in the Arctic, which may happen) you start the winter months (no energy input from the sun) with the entire area with open water. This open water may then radiate more energy away to outer space than if there was a thick insulating ice layer above it. So will an ice free summer (which leads to an ice free winter for awhile) cause a much increased loss of stored ocean energy in the winter without the insulating property of water?

    I made and analogy of it in an earlier post. A house that has no insulation will lose energy much faster in winter than a well insulated huse.

    [Response: As I mentioned in a previous response, radiation from the ocean doesn’t make it to space. What radiates IR to space is the upper atmosphere. It seems to me that what counts for the outbound radiation change, is the temperature change in the radiating level of the atmosphere above the Arctic region.

    And as I’ve mentioned twice now: if you really want to know the impact of winter sea ice loss on outbound radiation balance, go ask at RealClimate:

    http://www.realclimate.org/

    Look for the “unforced variations” thread.

    I’ve also pointed out that I don’t have the data to “do the math” on that one, but I’ve seen nothing in the literature that supports the suggestion that sea ice loss could have a net cooling effect — quite the opposite — and the “back-of-the-envelope” numbers likewise don’t support the idea. If you can report some peer-reviewed literature, great — if you can do the math to support and argument, great — but honestly, simply repeating the same question over and over with the clear implication that “sea ice loss could have a net cooling effect” is a plausible idea, is getting rather tiresome.]

  27. I did post on Real Climate and did receive a response, thanks.

    The circumstantial evidence (premath) that seems it might be plausible (but other more complex actions may be taking place). A physical law is that in order for water to turn to ice it has to lose a precise amount of energy based upon the mass of the water and its temperature. If you look at the Arctic sea ice page the rate of ice growth is much more rapid this winter season than the long term average. This could imply that in order for this to take place that this season, the Arctic may be losing energy faster than the historical average in order for the ice to form a a much more rapid ratee.

    [Response: Which Arctic sea ice page? Is that faster rate of sea ice growth in *volume* (which is relevant) or area/extent? The only reliable pan-Arctic volume measure I know of is PIOMAS, and they only update their data monthly.

    I also recall having seen calculations that the energy stored in the formation/released in the melting of sea ice is a very tiny fraction of the Arctic energy budget. Anybody remember those calculations?

    Also, be careful about using the phrase “losing energy.” When ice forms or melts, energy is converted to/from latent heat but it is not *lost* to the earth. The same is true of energy which causes ocean water to evaporate.]

    • Salinity also counts and that is not such a uniform thing.

    • http://nsidc.org/arcticseaicenews/

      It would be in area and extent. If you observed two swimming pools of water. One’s surface completely freezes in a day, the other a week. What conclusion would you form from the observation? Would you assume the one that froze in a day was losing energy at a more rapid rate than the other (the water is losing energy somewhere).

      [Response: Take ten frozen pools of water. Case 1: three of them melt, then re-freeze. Case 2: all ten melt, then nine re-freeze. I think that gives us about as much insight as your example.]

      You made the point that radiation from the ocean does not make it into space directly. Chris Dudley on Real Climate in response to my post there said “With that background, we can give a simple answer. The thermal properties of the surface are unimportant, everything depends on how much sunlight is reflected.

      Now things are not that simple. We know that there is a direct infrared radiative connection between the surface and space, particularly in dry cloudless conditions.” The Arctic in Winter is generally dry and cloudless. HIs conclusion on this was “In the Arctic, while the infrared emissivity of soil, ice and water are about the same, there could be substantial differences in how summer stored heat comes out in the winter depending on how these surface elements change.”

      [Response: If you want to claim that all the emission of IR from the surface in the Arctic goes directly to space, you’ll have to provide some peer-reviewed science to back it up — because I don’t believe it. If you want to know what happens to the extra heat in the Arctic (because it really is hotter up there), don’t take IR absorption by the atmosphere, latent heat from evaporating water, and the reduced heat flux from lower latitudes, off the table.

      And even if it were true — so what? I don’t think anyone has disputed that reduction in winter sea ice cover will affect how heat moves around in the Arctic, between ocean and ice and atmosphere and space. What I haven’t yet heard from you is any argument, not even one, making it even plausible that increased winter heat loss could outweigh increased summer heat gain. No, I do not consider that claim at all plausible. Not even close.

      If you’re trying to satisfy your curiosity about how the numbers actually stack up, you’ll have to seek elsewhere because (as I said already) I don’t have the data to do the math on that one. If you’re just trying to convince me (or others) that it really is plausible, then vague assertions with no math and no scientific literature is a piss-poor way to do it.]

  28. Norman may have come across mention somewhere Curry and Schramm (1995) — that’s (for me right now) the top Google hit for a search on this subject. It’s apparently very popular somewhere with those who like this sort of thing.

    Doubt that because Google tracks users and slants search results to match readers to advertisers, so your search may differ.

    Those checking instead with Google Scholar have found there’s much more and more recent work on the subject, but aren’t apparently excited thereby.

  29. If the Arctic heats faster than mid latitudes, the decrease in temperature difference will slow circulation, the net effect of which is the damn fan in the air conditioner is broken and we are stuffed.

  30. Geo engineering is one of the solutions that is used by nations and scientist trying to find better solutions to reduce the amount of CO2 currently present in the atmosphere. Testing and studying the effects of ocean fertilization is very costly and time consuming studies. Therefore it is very likely that if someone else is willing to pump money into the study then the government is also very likely that they would turn a blind eye to it. When it comes to climate change profits simply out weigh the risks of climate change according to the stakeholders. It is pure human nature that as long as it is not affecting me and my pocket then I would continue the same lifestyle I have been living, and besides, the projections go as far 2040 as to that is when things would get serious. If you look at the age group almost the entire government panel would be not be here by that time and therefore they would not be here to experience the true effects of Global warming.

    As for the sea level rise on the eastern Cananda and Northern US as stated by Stephen Lacey’s post that with recent studies it has shown an increase. First of all, geographically speaking these are the closest places to the Arctic, as the Northern Hemisphere just came from summer, more and more square miles of the Arctic are melting away. Fresh water is less dense compared to sea water and therefore contributes dramatically to the sea level rise. Greenland is also melting at a very high rate which should be cause of concern for the government to spend more money towards Climate change and not be ignorant rather than investing billions of dollars to military projects.

  31. Here is an article about ice formation in the Arctic.

    http://www.arctic.noaa.gov/essay_wadhams.html

    Quote from article: “Cooling of the ocean surface by a cold atmosphere will therefore always make the surface water more dense and will continue to cause convection right down to the freezing point – which itself is depressed by the addition of salt to about -1.8°C for typical sea water. It may seem, then, that the whole water column in an ocean has to be cooled to the freezing point before freezing can begin at the surface, but in fact the Arctic Ocean is composed of layers of water with different properties, and at the base of the surface layer there is a big jump in density (known as a pycnocline), so convection only involves the surface layer down to that level (about 100-150 metres). Even so, it takes some time to cool a heated summer water mass down to the freezing point, and so new sea ice forms on a sea surface later in the autumn than does lake ice in similar climatic conditions.”

    From this article is would seem that a considerable amount of water has to cool before ice can begin to form. Ice thickness is much more complex. If it snows on top of ocean ice it greatly slows the rate ice will thicken (snow is a very good insulator).

    I may soon be able to submit primative math to the process of ice formation related to energy loss to see if the rapid growth of sea ice extent in 2007 shows a more rapid cooling in the Arctic.

  32. Philippe Chantreau

    By the same token, if wishful thinking could radiate energy to space, we’d all be freezing right now…

  33. Let me take a stab at the issue of radiative balance in the Arctic, and if an Arctic with reduced or removed summer ice cover would cause Arctic warming during winter (due to thinner winter ice cover) to increase so much that it would actually cool the rest of the planet or not (as per Norman’s inquiry).

    To answer to that question, I think we could start by checking how much MORE heat the Arctic takes up in summer due to albedo effect changes, and compare that to how much MORE heat the Arctic radiates away (in OLR (Outbound Longwave Radiation)) from the top of the atmosphere.

    For this calculation, I assume the ‘Arctic’ to be the area above 60N (which is 6.7% of the planet), which is where the bulk of snow/ice cover changes occurs.

    Tamino’s estimate for albedo (snow+ice) induced energy absorption due to snow/ice cover losses in summer is +0.45 W/m^2 globalized and annualized. Since that albedo effect is mostly concentrated in the area of 60N and above, in effect albedo change causes 0.45/0.067=6.7 W/m^2 forcing.

    In other words, OLR on an annualized basis, above 60N, should increase by some 6.7 W/m^2, just to “radiate away” the heat that albedo changes in summer caused.

    Now, PSD uses NOAA satellites to measure OLR, and they have been doing that since about 1975. Here is the timeseries interface :

    http://www.esrl.noaa.gov/psd/cgi-bin/data/timeseries/timeseries1.pl

    Fill in Latitude 90-60, fill in longitude 0 to 360, fill in “seasonal average”, and First month “Jan”, second month “Dec”, area weight grids “Yes” and “plot data” and this is what you get :

    http://www.esrl.noaa.gov/psd/cgi-bin/data/timeseries/timeseries.pl?ntype=3&lat1=90&lat2=60&lon1=0&lon2=360&iseas=1&mon1=0&mon2=11&iarea=1&typeout=2&Submit=Create+Timeseries

    That’s OLR (radiation leaving Earth to space) measured from the Top of Atmosphere above 60N.

    Note that over the past decade, there has been a significant increase in OLR, which roughly follows sea ice losses in summer. Eye-balling the change, some 4-5 W/m^2 increase in OLR has occurred, which suggests Arctic warming.

    Compare that 4-5 W/m^2 OLR “loss to space” to “increase in heat absorption” due to albedo effect, of 6.7 W/m^2. That suggests that most of the albedo heat DID make it out to space (of course after it warmed the NH for a while) but that something like 1-2 W/m^2 of heat is “missing”. It was added in summer due to albedo change, but did not get radiated away into space.

    Now, some of that “missing” heat could be absorbed by the (close to) 1000 Gton ice loss each year that PIOMAS is reporting. But even with that factored in, it seems that EVEN though the Arctic warmed up quite significantly over the past decade, it NOT YET shed more heat in OLR than it absorbed by albedo changes in summer. Which means that the Arctic may still be warming more, and thus we may still see more snow/ice cover loss in the years to come.

    Needless to say that observations suggest that the Arctic is still not in radiative “balance”, let alone contribute to “cooling” the rest of the planet, as Norman suggested.

    [Response: To get your graph, I think you have to start here:

    http://www.esrl.noaa.gov/psd/data/timeseries/

    and select “Interpolated OLR” as your dataset. If you go directly to the first URL you’ve posted it defaults to the “NCEP/NCAR Reanalysis monthly means” dataset (even though the URL is the same).

    Also, the graph your link (your 2nd URL) goes to is labeled “1000 mb” which is pretty close to the bottom of the atmosphere. However, the “NCEP/NCAR Reanalysis monthly means” dataset does includes OLR (select that as the variable), and even though one must select “surface” as the level because it’s not a level variable, the data file explicitly says it’s TOA. The graph looks like this:

    http://www.esrl.noaa.gov/psd/cgi-bin/data/timeseries/timeseries.pl?ntype=1&var=OLR&level=2000&lat1=90&lat2=60&lon1=0&lon2=360&iseas=1&mon1=0&mon2=11&iarea=1&typeout=2&Submit=Create+Timeseries

    Note that in the TOA graph, the increase in OLR since 1979 (when we have satellite data for sea ice and snow cover, the OLR graph starts in 1948) is only about 2 W/m^2 rather than 4-5. So, the energy imbalance is even greater (net warming) than you estimated.]

  34. From this evidence it would appear that the poles may radiate around 39 watts/meter directly into space. I would think more in winter months than summer (colder and drier). 10% of IR emitted by Earth goes directly to space (of 390 watts/meter flux). Most of the 10% occurs at the poles.

    http://oceanworld.tamu.edu/resources/oceanography-book/radiationbalance.htm

    My point was not really about if total loss of sea ice in summer would cool the Earth overall. I was just curious as to what the winter heat flux is for the Arctic without the insulation provided by snow and ice. I have seen many studies on the Arctic amplification, how melting ice will lower albedo and heat that area in the summer. It is not easy to find material on what occurs during the winter. That is why I have not linked to a peer reviewed article on it yet as I have not found one.

  35. Tamino,

    As you requested in your responses to my posts I did find a peer reviewed article that is stating the very thing I have been saying. The complete loss of sea ice will be balanced by a greater heat loss in winter.

    http://www.seas.harvard.edu/climate/seminars/…/Tietsche_GRL_2011.pdf

    [Response: No it doesn’t.

    What it says is that if ice is removed so that its concentration is out of equilibrium (which is what they did in their model runs), then local heat loss will cool the Arctic so that sea ice will recover to its equilibrium value. It does not say that a reduction in the equilibrium sea ice will have a net cooling effect on the planet.

    Their principal result is that anomalous ice loss will not lead to “hysteresis” (or a “tipping point”) through the albedo effect, addressing “the question of whether perturbations of sea!ice cover alone are able to trigger an irreversible climate change in the Arctic.” Their answer: “Arctic sea ice has a preferred equilibrium state that varies smoothly with the climatic forcing, and that there are recovery mechanisms that counteract the destabilizing ice–albedo effect after abrupt losses.”

    We’ve known about this paper since it was published. You have misrepresented it.

    And by the way, your link doesn’t work. The paper is here:

    http://www.seas.harvard.edu/climate/seminars/pdfs/Tietsche_GRL_2011.pdf

    ]

    • Tamino,

      The paper says summer ice can appear again in as short time frame as two years from a complete ice free summer conditions.

      Here is a quote from this article. It is pretty much what I had been posting on these threads. Less sea ice in the winter will lead to more heat loss during this season. Not the word “enhanced” heat loss.

      “Thus, sea!ice free summer conditions
      cause the ocean to gain excess heat through the surface
      during summer, but they also cause enhanced heat loss
      through the surface in the following autumn and winter,
      when the insulating sea!ice cover is anomalously thin.”

      The poles have always had a net cooling effect on the planet. The tropics absorb more heat then they give up and the poles emit more than they gain. Heat flow from the tropics to (eventually) the poles keeps the weather going.

      http://climateprediction.net/content/basic-climate-science

      Link to graph showing how poles cool the earth from above article.

      [Response: Nobody I know of has ever disputed that the poles cool the planet, they shed more heat than they take in. In fact it’s been referred to here as the earth’s “air conditioner.” But that’s not the issue. The issue is: if sea ice goes away, what difference does it make? Does it have a net warming or net cooling effect on the planet, compared to the previously existing state? Unless you have misrepresented your own questions and arguments, that has been the issue all along. Your graph and link are irrelevant to that question.

      Think carefully about what Tietsche et al. showed. In model runs they made sea ice “disappear” (as if by magic). In the first year after that, there was heat added in summer (from albedo loss) and heat lost in winter (by enhanced radiation). But the net effect after the first year was a lingering additional Arctic warming. This implies that the heat gain outweighs the heat loss — exactly the opposite of your belief.

      Tietsche et al. did show, however, that the lingering additional Arctic warming was not sufficient to sustain the sea ice decrease. Therefore sea ice returns to its equilibrium value determined by climate forcing, hence sudden sea ice loss cannot cause a “runaway” Arctic climate change. That’s the real point of their article. But at no point did they find any net cooling of the earth compared to its previous state, due to sea ice loss.]

      • And of course Tietsche et al did not include the (unfortunately highly probable) case of increasing GHG concentrations. It was meant to be a ‘surgical’ investigation, not a real-world simulation.

  36. I wonder if any satellite monitoring data from the cold war years was kept and has been declassified? Every 90 minutes, looking down, I’d imagine they’d take a background brightness measure at multiple wavelengths including the infrared. The Arctic should have been of special interest.

  37. The link is not working try again for anyone who is intersted. The paper is called “Recovery mechanism of Arctic summer sea ice” by Tietsche.

    http://www.seas.harvard.edu/climate/seminars/pdfs/Tietsche_GRL_2011.pdf

    Hope this link gets to the article.

  38. Tamino,

    Thank you for your replies to my posts. You provide thoughtful and intelligent answers to my questions. I will not waste your valuable time any further on this topic (Effect of ice loss on Arctic total energy flux) unless a new or conclusive research paper can redirect current opinion.

    [Response: I know it’s been a long and sometimes contentious exchange, but I think it’s been worthwhile. And, you have successfully emphasized that albedo change due to sea ice loss is far from the whole story. Perhaps advocates (like myself) are a bit too eager to tout the albedo effect.]

  39. Tamino, Norman,
    Regarding Tietsche et al :

    If you remove ice in summer it will absorb more heat due to albedo effect.
    But it will also cause open ocean in fall and thinner ice in winter, both of which will cause an increase in temperatures during freeze season, which will radiate more heat to space.

    What Tietsche et al showed is that the increased IR loss to space during winter exceeds the heat uptake due to albedo change in summer, and thus ice cover returns to the “mean” ice cover curve typical for GCMs.

    So, what they suggest is simple : that Arctic sea ice by itself is a stable system.

    That conclusion would be totally acceptable, if it were not for actual observed sea ice extent. which does NOT seem to return to a “mean”. In fact, it looks like it is in free fall, way below what GCMs are suggesting as the “mean” trend :

    http://neven1.typepad.com/.a/6a0133f03a1e37970b017744cf5360970d-pi

    If the Arctic is indeed a stable system than the only way to explain the remarkable difference between GCM runs and actually observed sea ice trend is that the “mean” is going down much faster than GCMs assume. That means there must be some source of heat that GCMs did not account for, or seriously underestimated.

    Now, remember that ice cover reduction so far only caused a 0.1 W/m^2 forcing (see Tamino’s “ice” post). And recall that snow cover reduction so far caused a 0.35 W/m^2 forcing, for a total of 0.45 W/m^2 globalized/annualized forcing (see Tamino’s “ice+snow” post).

    So could it be that Tietsche et al is right, and that ice cover will return to a “mean” curve, but that the “mean” curve depicted in the graph above is “pulled in” in time by snow albedo forcing, which is a factor 3.5 stronger than ice-albedo forcing ?

    Maybe the problem with GCMs is not their modeling of ice and ice dynamics etc, but it is the underestimation of snow-cover that makes them way too conservative.

    See also

    http://neven1.typepad.com/blog/2012/08/the-untold-drama-of-northern-snow-cover.html

  40. Tamino,

    Sorry I messed up the links to the OLR timeseries.
    Interesting to see that above 60N, it seems that increase in OLR is not enough to get rid of the increased heat accumulated due to albedo effect of snow and ice losses.
    That’s not a good sign, since it hints at an “instable” Arctic, self-amplifying its snow/ice losses.

    To be a bit more accurate, I did two more experiments :

    I took 70N (and above) as the limit where “ice albedo” changes operate.(which we estimated at +0.1 W/m^2 globalized forcing).
    If that albedo-induced heat stayed above 70N (which is 3% of the planet) OLR should increase to 0.1/0.03 = 3.3 W/m^2 to compensate.
    The OLR increase above 70N from NCEP/NCAP seems to have increase some 4 W/m^2 sincce 1990 :

    http://www.esrl.noaa.gov/psd/cgi-bin/data/timeseries/timeseries.pl?ntype=1&var=OLR&level=2000&lat1=90&lat2=70&lon1=0&lon2=360&iseas=1&mon1=0&mon2=11&iarea=1&typeout=2&Submit=Create+Timeseries

    so ice albedo effect may indeed have been compensated for by increased OLR to space. Any ice volume (and extent) must be (is probably) caused by heat moving in from below 70N rather than albedo changes above 70N itself.

    Then I looked at 50N, which is low enough to cover any snow+ice albedo effect (which we estimated at +0.45 W/m^2 globalized forcing).
    If that albedo-induced heat stayed above 50N (which is 12% of the planet) OLR should increase to 0.45/0.12 = 3.75 W/m^2 to compensate.
    The OLR increase above 50N from NCEP/NCAR seems to have increased only some 2 W/m^2 since 1990 :

    http://www.esrl.noaa.gov/psd/cgi-bin/data/timeseries/timeseries.pl?ntype=1&var=OLR&level=2000&lat1=90&lat2=50&lon1=0&lon2=360&iseas=1&mon1=0&mon2=11&iarea=1&typeout=2&Submit=Create+Timeseries

    This suggests that (at least compared to 1990) snow-albedo-induced heat between 50N and 70N transported to lower latitudes (thus warming lower latitudes, especially in summer), while also flowing up to above 70N, where it causes accellerated Arctic sea ice melt (which would explain a difference between GCM models and observed Arctic ice loss trends).

    Sorry for the long post, but I find this pretty interesting…

  41. Rob Dekker,

    I like the production of your graphs but you may want to reread Tamino’s post on the energy gain from lowered snow and ice in the Arctic. He does say it is just a rough estimate a ballpark figure. It would not be precise enough to compare the outgoing longwave radiation (measured by satellites that may not be totally accurate for what the real number is because of complexities in measuements) with Tamino’s rough calculation and make a determination of the net energy flow in the Arctic.

    Since you have talent at generating graphs perhaps if you put two plots on the same graph. Incoming shortwave radiation making it to the surface (not reflected by clouds or snow and ice) and the outgoing longwave radiation and compare the two to see what the net energy flow is in the region of study (arctic).

  42. Norman,

    He does say it is just a rough estimate a ballpark figure. It would not be precise enough to compare the outgoing longwave radiation (measured by satellites that may not be totally accurate for what the real number is because of complexities in measurements) with Tamino’s rough calculation and make a determination of the net energy flow in the Arctic.

    Norman, the warming above 50N due to (mostly snow) albedo is a factor 2 higher than the increase in cooling due to higher temperatures.

    Also, please note that this same data confirms your point : At least for the ice-albedo effect (above 70N) IR radiation seems to increase faster than the albedo effect in summer.
    Which also confirms Tietsche et al (that sea ice by itself will return to an “equilibrium” amount).

    If you truly believe that the satellites are not accurate enough and neither is Tamino’s estimate, then why don’t you present that plot you suggest yourself, and prove us all wrong ? And remember that inaccuracy goes two ways.

  43. Rob Dekker,

    I am not trying to prove anyone wrong. I did go to the NCAR/NCEP page and try to make a plot but all I got was a globe of anomalies over a 10 year period. I know you are good with the graphs and just requested a similar graph to your previous post with just an inclusion of shortwave radiation reaching the surface. It would be a more direct apple to apple (using the same data source) comparison. I am interested in this material as well.