Skip to content

Met Office Decadal Forecast Hopelessly Wrong

January 27, 2014

By Paul Homewood


h/t Paul Matthews


From the “Science is Settled Department”






Back in 2007, the Met Office published a glossy brochures, containing their projections and assessment of climate change. This included  their decadal forecast of global temperatures, which predicted that, by 2014, global temperatures would have “risen by around 0.30C compared to 2004”.




And the reality so far?




[Quick note – as the Met graph at the top makes clear, their forecast anomalies are relative to the period 1979-2001, whereas HADCRUT4 ones are based on 1961-90. As 1979-2001 was 0.194C warmer than 1961-90, the forecast numbers need to be increased by this amount, in order to compare with HADCRUT4]




It gets worse. The original paper by the Met Office’s Doug Smith, from which these predictions were drawn, states:

“with the year 2014 predicted to be 0.30° ± 0.21°C [5 to 95% confidence interval (CI)] warmer than the observed value for 2004.”

The temperature anomaly in 2004 was 0.45C, and rose by a tiny 0.04C in 2013, to reach 0.49C. If the 2014 outturn figure does not rise to at least 0.54C, the model will be outside its confidence intervals.






In their brochure, the Met were quite happy to brag off their scientific excellence.










Unfortunately, their work has been used in helping to formulate government policy regarding climate change. One wonders what difference it might have made if they had got their numbers right in the first place?

Next time someone tries to tell you that climate models have been reliable, remind them of this Met Office debacle.

  1. A C Osborn permalink
    January 27, 2014 7:31 pm

    Paul Hudson needs to be taken to task over these comments on the BBC web site on the Weather forecasting failures.
    The heat has gone in to the Arctic or the Oceans.

  2. January 27, 2014 7:40 pm

    So what will the impact of the oceans absorbing the heat?

  3. January 27, 2014 8:02 pm

    I’ve just noticed that the MO are forecasting the same temperature for 2014 as 2013.

    It’s also interesting that while their central estimate for 2013 after the October figures was more or less correct, it was within a huge range of 0.39c to 0.59c.

    This may not seem a big range, but it would have required a range of -0.05c to +1.15c for November and December.

    Hardly likely!

    • Green Sand permalink
      January 27, 2014 8:44 pm

      Hi QV, you maybe interested in the comment section of:-

      “HadCRUT 2013”

      Tim Osborn from CRU joins in referencing Folland et al 2013 on the “warm bias” in the MO’s annual “statistical model” and on how skilful the forecasts have been over the last 12 years

  4. Green Sand permalink
    January 27, 2014 8:33 pm

    Thanks Paul and make yourself ready!

    “Decadal forecast – Forecast issued December 2012. The forecast will be updated in January 2014.”

  5. Joe Public permalink
    January 27, 2014 9:01 pm

    Is the MO fit for purpose?

  6. mkelly permalink
    January 27, 2014 9:31 pm

    At least the stuff coming out of the smoke stacks looks like clouds instead of mud.


  1. Paleofantasies | Skeptical Swedish Scientists

Comments are closed.

%d bloggers like this: