Skip to content

Met Office Arctic Ice Forecast Hopelessly Wrong

October 2, 2014

By Paul Homewood




NSIDC have released the Arctic sea ice extent data for September, and the average for the month is 5.210 million sq km.

Each year the Sea Ice Prediction Network collect forecasts from ice extent from a variety of sources. The UK Met Office’s model based forecast was 4.1 million sq km, so they were only 27% out!

Their paper states:


Using the Met Office GloSea5 seasonal forecast systems we have generated a model based mean September sea ice extent outlook of 4.1 million sq km.


which should gives us all a huge amount of confidence in their models.


Last year, of course, their forecast was even further off line, 3.360 million sq km against an outturn of 5.244.


Figure 1 shows the recovery of sea ice in the last two years, both back close to 2005 levels.



Figure 1

  1. Joe Public permalink
    October 2, 2014 12:50 pm

    Just why is the UK taxpayer-funded Met Office wasting our money & their resources on trying to guess a physical event so far from our shores?

  2. myrightpenguin permalink
    October 2, 2014 1:07 pm

    Never mind, the mistress of bingo has got her damehood.

  3. October 2, 2014 2:57 pm

    So the story about the walruses gathering on shore (
    wasn’t true? It’s not melting ice that stranded them? There’s more ice than in the past? I am shocked! (/sarc)

  4. October 2, 2014 6:59 pm

    I seem to recall a wide margin of error in the MO figures but I can’t see the on the graph.

    • October 2, 2014 9:44 pm

      They give +/- 1.0 million sq km, i.e. nearly half the measurement.

      • Ben Vorlich permalink
        October 3, 2014 6:39 am

        I think that’s called guessing, although they nailed it with their top of the range guess

      • Allan M permalink
        October 3, 2014 8:24 am

        2 + 2 = 5±1; correct. 2 + 2 = 5±0.1; wrong. Neither is much use, though.

      • October 3, 2014 8:33 am

        Not quite nailed it, 4.1 + 1 = 5.1, actual 5.2.
        +/-1 1 million is suspiciously rounded.

  5. Anything is possible permalink
    October 2, 2014 7:42 pm

    Smart people voted for 5.2m in Anthony’s poll, without having a multi-million pound supercomputer to guide us.

    Co-incidence? I think not………..

    • John F. Hultquist permalink
      October 3, 2014 1:54 am

      Us = Smart
      Right. But then I wonder about the others there that pushed the estimate to 6.1.

      • Ben Vorlich permalink
        October 3, 2014 6:40 am

        Using the same margin of error as the MO they were OK too.

      • Kelvin Vaughan permalink
        October 6, 2014 3:31 pm

        Yipee I just found out I am a smart person.

        Actually only 8,8% voted over 6% the rest voted below.

  6. October 3, 2014 12:18 am

    There is more of a tale in there. Yes you are right but there is the mystery of producing late forecasts complete with update.

  7. Allan M permalink
    October 3, 2014 8:27 am

    Met Office Arctic Ice Forecast Hopelessly Wrong. Dog bites man.

  8. Kelvin Vaughan permalink
    October 6, 2014 3:17 pm

    Yipee I just found out I am a smart person.


  1. Met Office Arctic Ice Forecast Hopelessly Wrong | The Global Warming Policy Forum (GWPF)

Comments are closed.

%d bloggers like this: