DMI Arctic Data Death Spiral

DMI shows Arctic sea ice extent rapidly declining, and far below last year.

icecover_current_new (3)

But their maps show essentially no change over the last three days.



And DMI maps show that there is slightly more ice than last year.

   OSI_HL_SAF_201505241200_pal 2016-05-25-04-10-07

This entry was posted in Uncategorized. Bookmark the permalink.

38 Responses to DMI Arctic Data Death Spiral

  1. Sarmange says:

    As we know the DMI has been using two models with different resolution in parallel; let us call them M15a and M30.
    M15a stands for a resolution of 15% and the coastal mask that was used until summer 2015. (*)
    The M30 stands for 30 % resolution and a coastal mask which may or may not be the same as for M15a.
    In summer 2015 the coastal mask was changed (refined) for the 15% model and we got what we can call a model M15b – a 15% model with a new coastal mask. (*)
    The M30 was abandoned, leaving us with the 15b model only.
    The new coastal mask may have influenced the measured data in a significant way. We can’t compare the former M15a with the M15b plots, much less the M30 ones
    DMI describing the reasons for the change:
    (*) “In summer 2015 this mask was refined and the masked region was subsequently smaller, thus leaving more area for classified sea ice and open water. “

  2. Gail Combs says:

    Tony do you use a pixel count to look at the red vs green ratio?

    • Caleb says:

      Gail, he did a pixel count in a prior post, and on May 23 it showed that 2016 had 245 more pixels than 2015. However the DMI graph shows 2016 has less ice than 2015. It seems to me they “have got some ‘splainin’ t’do”.

      I’m hoping I’ve made a mistake, and I am not seeing what seems to be seen, because DMI’s mistake would be pretty glaring. Maybe I will stand corrected tomorrow.

      • Neal S says:

        DMI has on their website “An appearent divergence between ice extent algorithms is an artifact -” and their explanation is …

        Which I can summarize as they claim their coastal region mask is much better now. But the obvious question this raises in my mind, is why they don’t run all the old data through the new mask, so that you can really get a good apples to apples comparison and not apples to oranges.

        But this is totally independent of their showing rapid decline in the sea ice extent chart, and yet this not being readily visible in the polar images over that same time.

        • dave1billion says:

          I may be wrong but I believe that message refers to the previous 30% ice extent map that they have discontinued.

          The explanation of the discrepancy made sense for the deprecated 30% map. They used a different “filter” for the coastal regions and the new data set contained areas that were not included in the previous 30% data.

          So as the ice cover grew as winter came on it showed an increase in the 30% ice extent over previous years that was an artifact of the fact that they were adding ice in areas that were previously excluded.

          This only became apparent as the 30% ice extent grew and entered these new areas.

          They decided to trash the 30% ice extent report completely. I would have preferred that they just showed it without the previous data but nobody asked my opinion.

      • Caleb says:

        If the narrowing of the “coastal mask” made such a big difference, there should be a glaring circle of green in our host’s map. It should amount to more than 245 pixels.

        Maybe I need new glasses, but I don’t see it, when I squint at the above map. In fact there are even some red pixels along the East Siberian coast. I have the feeling that the difference between the 2015 “mask” and the 2016 “mask” is so narrow it is less than a pixel wide and doesn’t show up.

        If I am wrong, I won’t mind being shown I am wrong, but at this point I have a sense the people fussing about the change in the DMI “coastal mask” are “straining at a gnat while swallowing a camel.”

        And what is the camel they swallow? It is exactly what our host demonstrates: The maps show more ice while the graphs show less ice .

      • Gail Combs says:

        Thanks REN,

        It should make for an interesting summer.

        So far we have been FREEZING! Actually it was really close to freezing on the 16th at 37F (2.7C ) in May in the sunny south???

      • ren says:

        I wonder what they will say in winter: La Niña on the way, AMO is falling, solar activity a minimum.

        • Gail Combs says:

          I am hoping for massive snowstorms and exceptionally cold weather especially in the mid-Atlantic states during the last two weeks of October.

          Why? The US election is November 8, 2016.

          • ren says:

            Gail look at this:
            Is El Niño occurs in the middle of the lunar cycle?
            “With the culmination of the 18.6-year cycle of the Moon in 2006 and again in 2024-25, also called the Major Lunar Standstill, we are afforded the unique opportunity to observe the monthly, annual, and 18.6-year wanderings of the Moon. The 18.6-year cycle is caused by the precession of the plane of the lunar orbit, while this orbit maintains a 5° tilt relative to the ecliptic. At the peak of this cycle, the Moon’s declination swings from -28.8° to +28.8° each month. What this means is that each month for the years 2005-2007 and also 2023-2026, the Moon can be seen rising and setting more northerly and also more southerly than the solar extremes, and will transit monthly with altitudes which are higher in the sky than the summer Sun and lower in the sky than the winter Sun.”
            2006-9=1997 and 2006+9=2015

    • tonyheller says:

      Yes I do!

  3. GW says:

    I thought DMI was the reliable, trustworthy one all these years ? Have they finally succumbed and gone to the dark side ?

    • Ben Vorlich says:

      Probably more a case of older experienced and worldly wise Science Graduates retiring and being replaced by young Climate Science graduates than a change of policy.

  4. Colorado Wellington says:

    Only they know what they’ve done and why but we must never dismiss the Willie Stark axiom:

    “Man is conceived in sin and born in corruption and he passeth from the stink of the didie to the stench of the shroud. There is always something.”

    • gregole says:


      Say, with the low ice extant and all (\sarc!), is there any news I may have missed of some intrepid adventurers planning to row through the Northwest Passage? Seems like this is the year!

  5. Sparks says:

    Why are they comparing 2016 with a Mean of 1979 – 2000?

    • AndyG55 says:

      Because 2016 is basically right at the top of the AMO cycle, and 1979 was the bottom.

      The period 1979-2000 covers basically the whole upward leg of the AMO..

      That means that the current Arctic sea ice level is pretty much EXACTLY WHERE IT SHOULD BE for the phase of the AMO.

      It also means that we can expect the sea ice level to start climbing again , probably even coming year, now that the AMO has started to head downwards, and the El Nino is gone.

  6. Scott Scarborough says:

    “Sunshine Hours” has JAXA sea ice and it shows the same thing… an unprecedented drop in sea ice. Could they all be getting their information from the same source?

    • AndyG55 says:

      “an unprecedented drop in sea ice”

      There’s that ignorant wording again.

      The current Arctic sea ice level is NOT unprecedented, except over a very short time period.

      During the first 3/4 of the Holocene there was regularly ZERO summer sea ice.

      The current level is, in fact, ANOMOLOUSLY HIGH as it struggles to recover from the effects of the Little Ice Age which was the COLDEST period in the last 10,000 years.

  7. AndyG55 says:

    Iceland sea ice index

  8. Caleb says:

    I admire our host’s common sense, as he simply compares last year’s maps with this year’s maps, counts the pixels, and demonstrates the maps show there was more ice last year than this year, and therefore there must be some glitch involved, for this year’s DMI graph to show so much more ice last year than this year.

    I figure glitches are to be expected, with satellites malfunctioning, as they recently have done. What we should do is discuss the glitch like mature adults, and seek a remedy. On my obscure site I brought up the glitch, seeking sensible answers.

    A visitor then commented that the maps our host uses are “ice edge” maps, and he should have used “concentration” maps, leaving me with this link.

    I myself do not have the ability to tell the difference between the so-called “ice edge” maps and the “concentration” maps. My eyeballs lack the ability our host has to see differences in the amount of pixels.

    Therefore I informed the person he should not disparage our host’s choice of maps on an obscure site, where our host could not respond, but should come to this site. Hopefully he will soon appear and we can converse with him.

    I myself confess I am unsure of the difference between the two types of maps, and would appreciate being educated.

    By the way, the blogger called him-or-herself “cdqgfj “.

    • Colorado Wellington says:

      What, cdqgfj doesn’t like to come here anymore? Did he have to retreat to a safe space? That’s Andy’s fault!

      • AndyG55 says:

        Thanks :-)

      • Gail Combs says:

        Hey Colorado,

        I did my bit to help Andy. Of course so did you and gator.

        • AndyG55 says:

          You sure did.

          Love the way I could make a short comment, and then you would flood it with links and information.

          Where is gator, we need him to help fight the increasing stream of alarmist trolls.

          I am running short of time due to work (necessary for income), and play, to keep up with things.

        • AndyG55 says:

          typo correction…

          ” the increasing stream of DESERATE alarmist trolls.

        • Colorado Wellington says:


          You can’t just post what you want. There are the rules!

          How am I supposed to pick Andy as the target, freeze him, personalize him, and polarize him if I have to list everyone else including myself? Andy was expected to hurt and be cut off from his support network. He should get no sympathy. He was to recoil in shame instead of thanking me. That’s not how it’s supposed to work!

          This radical thing is hard for the self-taught. No wonder our young people have to go to universities, study hard under experienced 60s professors and do years of practice on campus.

          I’d pick you next time but Alinsky says to go after people and you are an institution!

          Alinsky’s Rules for Radicals:
          13.“Pick the target, freeze it, personalize it, and polarize it.” Cut off the support network and isolate the target from sympathy. Go after people and not institutions; people hurt faster than institutions.

          • Sparks says:

            Institutions?? What the Heller radicals??

            You cant hurt people who are grounded in reality, that’s why the global boredom of man made planetary warming crumbles, it is a failed ideology that has built it’s foundations on shaky ground.

          • AndyG55 says:

            “How am I supposed to pick Andy as the target, freeze him, personalize him, and polarize him ”

            now I’m hurt. :-(


          • Sparks says:

            My dog once spent hours barking up the wrong tree, it was so funny when I seen him realize the cat moving in the other tree.

          • Sparks says:


            If anyone tries to target, freeze, personalize, and polarize you in an unkind way, let me know.

            I love bullies, when they runaway, it’s just so cute, I even love the ones who stay, I get to wave as they are taken away in an ambulance…

          • AndyG55 says:

            I think I can probably look after myself….

            .. even from the barbs and arrows of the grate CW ;-)

    • Peter Ellis says:

      Well, really you shouldn’t use the images at all, because they’ve been through a whole bunch of processing and compression steps to put them up for display. Look at all the jpg artifacts on Steven’s overlays, for example.

      This is causing errors – for example the overlay of the last couple of days has green pixels (“increased ice”) in the Kara sea, but when you look at the individual day data, the Kara sea is 100% full in both – white and grey pixels should both count for extent. Something’s going wrong with Steve’s colour replacement process, likely because the white and grey pixels have been dithered into each other by the jpg compression.

      The underlying data is available in three different formats (HDF5, NetCDF, GRIB), and Steve should be using one of those.

      Details for how to work with HDF5 format files are here:

  9. Jerry Gorline says:

    I used to go there a lot, DMI. Guess they got money under the table to post garbage, too bad. Obviously not correct any more.


Leave a Reply to Jerry Gorline Cancel reply

Your email address will not be published. Required fields are marked *