MET- Office: New four year ‘decadal’ forecast spaghetti

Posted: January 30, 2014 by tallbloke in Accountability, Forecasting, humour

Ed Hawkins tweeted up  the latest offering from the MET-Office this morning. It’s a “Decadal forecast”, which runs from now to the beginning (not the end, Ed) of 2018. Stop tittering at the back there! But compounding matters, the ‘forecast’ is a spaghetti of similarly coloured lines. I said STOP LAUGHING! 🙂


Here’s the twitter exchange I had with Ed:


  1. Johnbuk says:

    This is what you get when you leave a ball of wool on the floor and think the kitten is asleep. Usually takes less than 5 years though.

  2. tallbloke says:

    Great write-up by Simon Carr on the HoC Climate committee meet with Lindzen, Lafroimboise and Nic Lewis:

    I live tweeted this event yesterday. What a laugh. Gavin trying to wriggle meanings as usual. Check out hashtag #IPCCreview here

  3. A C Osborn says:

    What I find amazing is that they admit there is a “Pause”, even admit that it may last another 5 to 10 years and then put out a crap graph showing mostly warming. Talk about inconsistent.
    Are they totally stupid or what?

  4. Rog

    Some links for the forecast.

    Click to access decadal_forecast_2014-2018_jan2014.pdf

    The blog page has turned the blue pencil tool into the blue spray can tool, week two in the MS paint course, they must have finished the Excel course by now.

  5. Jaime Jessop says:

    That’s a lot of taxpayers money for grown men and women to scribble down with a blue crayon that which they really, really, REALLY want for the next four Christmases!

  6. Ed Caryl says:

    Rejoice! It will only take five years to falsify these. Reality will be just below the green area.

  7. tchannon says:

    Sorry Rog, I have written something as you asked but on logging in I discover you posted first.

    Prior to final edit this is what I was going to post, usually toned down drafts, I’ll leave babbling.

    Met Office Decadal Forecast: 2014-2018
    Computed November 2013, issued January 2014

    This is so bad I am having great trouble comprehending what the Met Office is babbling about.

    I am not showing a plot because the graphic figure it has published does not show the forecast figure. The forecast is two spot numbers for a non-specified date. What I take to be a “global” estimate of historic temperature does not seem to reference a source, perhaps I missed this.

    No historic information is provided comparable in kind with the forecast, that is, data points at the same sample rate as the forecast, or at least not that I noticed.

    The figure shown commences 1960, perhaps notably omitting what I thought were higher temperatures prior to 1960.

    As I read it decadal (10 years) does not mean decadal as in a forecast for 10 years ahead but means filtered with a cutoff at 10 years and data points at the irrational rate of one every 5 years, the impossible to achieve Nyquist limit. Hence there is one data point 5 years after the date of forecast. This seems to be untrue since the date must be the mid point of the time span which is stated 2014-2018, four years. So if I take that to mean to November 2018 the mid point is summer 2016. No doubt if I have this wrong many will tell me.

    If they really meant 10 years, figure they show does not reach 10 years in advance, perhaps it could have been clearer and they should not have written averaged over 5 year period, which is not decadal, must be over 10+ years.

    Averaged over the 5-year period 2014-2018, global average temperature is … [forecast to] be between 0.17°C and 0.43°C above the … [reference period] (1981–2010) … .

    Met Office text edited for clarity.

    Or put another way mid 2016 +0.3C above reference, +- 0.13C

    I expect I will get flak over that.

  8. oldbrew says:

    95% confidence says the straight grey reference line will turn out to be more accurate than the blue spaghetti.

  9. DB says:

    Their Figure 3
    includes previous forecasts. However, the shape and range indicated for the 2005 forecast does not match what they published in Science at the time (Figure 4).

  10. michael hart says:

    The most obvious conclusion is that the graphic is intended to neither convey any meaningful information, nor engage the viewers concentrated attention. One of Sir Humphrey’s finest creations.

  11. tom0masont says:

    And still the monks of the climate monastary try to save their belief system by publishing complete blue crayon marks on a graph. Do they understand the underlying processes – of course they don’t.

  12. Joachim says:

    The GREEN 1960-1990 (Models) is below the black line (why not accurate?) and
    from 2010 on, above the black line….is this the global warming? Why was GREEN 1960-90
    .set as much too low?

  13. […] Tallbloke has the spaghetti graph, which is the basis for this, and as he rightly comments, it looks like kiddie’s scribble! […]

  14. Stephen Richards says:

    This is so bad I am having great trouble comprehending what the Met Office is babbling about

    That is the point Tim. This is UK Met Off speak. It is a new language for expressing scientific conclusions, derived from corrupted or inadequate data coming from computer programmes run on the new PS4. It’s a mixture of gobbledigoop and merde du taureau.

  15. DirkH says:

    CO2AGW theory states that the future always warms, and the past always cools. GCM’s do the future warming, homogenization the cooling of the past. At any given moment, no matter what the temperature is, the future is always warmer, the past always cooler. It takes thousands of scientists to constantly correct the numbers. It’s a little bit like the moon landing; only difference is it’s a hoax.

  16. craigm350 says:

    OB that was my initial thought too when I perused this morning 😀

  17. Doug Proctor says:

    I’m still holding for 2015 when the observations fall below the bottom of all the predictions/projections/forecasts/model runs/things that we might have said but didn’t really mean to be taken literally or even qualitatively because we don’t have good enough computers and …. say, is that a squirrel over there (let’s get out of here) ….

  18. Brian H says:

    Another deposit from the Mutt Orifice.

  19. oldbrew says:

    Is there a place for it here?

    Select 1976 in the ‘April Fool Archive’ for a good one by Patrick Moore.
    Or click here:

  20. […] Tallbloke has the spaghetti graph, which is the basis for this, and as he rightly comments, it looks like kiddie’s scribble! […]

  21. Berényi Péter says:

    I think they are telling us that it may get cooler in the meantime, but there is no way it will be colder by the beginning of 2018 than it is now. If that’s the proposition, we can assign a definite truth value to it in 4 years time. Be patient.

    The latest decadal forecast, issued in January 2014, shows that global temperatures are expected to maintain the record warmth that has been observed over the last decade, and furthermore that it is possible that new record global temperatures may be reached in the next 5 years.

  22. Me_Again says:

    Surely the % confidence is ass about face? I’d have thought at the extremities the confidence would be lower. Difficult to see how you can have a 90% confidence at the top and bottom of the graph.

    But then I’m not a mathematician and there’s no doubt a way of having 90% confidence at both ends of the probability………..

  23. […] MetOffice’s 2014 Decadal Forecast over at Tallbloke’s Talkshop in an article entitled MET- Office: New four year ‘decadal’ forecast spaghetti.  This is what fellow WordPresser Tallbloke had to […]

  24. MET forecast against mine. See here

    Note that in the MET figure the green area representing the CMIP5 projection has been shifted down to make the illusion that the current temperature is within it.

  25. tallbloke says:

    Nicola: Can you link the figure? I can’t see any new plot on your page. Thanks
    I did recently post your IPCC vs planetary models here

  26. My figure is here updated to Dec/2013:

  27. old44 says:

    Last time I saw a graph like that was after I dropped a plate of spaghetti.

  28. Berényi Péter says:

    Here is their Decadal Forecast for the next 4 years issued in January 2014. The funny thing is they have also made a pamphlet about the “pause” in July 2013.

    The recent pause in global warming: What are the implications for projections of future warming?

    In this paper Figure 3 has HadCRUT4 data up to 2010 along with an “observationally constrained range” beyond that date. I have copied that over to Figure 3 found at their Decadal Forecast page.

    Here is the result. It can clearly be seen that their current forecast does not fit into the range of their “observational constraints”, which is kinda weird.

    Of course, the in the paper explaining the pause away the relation between CIMP5 projections and observationally constrained range is different, the latter is a bit lower there than the former. It is the the other way around after copying over, but that’s because it was fitted to observations, not CIMP5 projections. Which means the relation between CIMP5 range and observations is different in the current paper, the former is shifted downward now by about 0.3K, just as Nicola Scafetta says.

    Which is exactly the kind of game which only brings shame to players in science.

  29. tallbloke says:

    Peter B: The MET-O have moved to a later baseline for this latest effort.

  30. Berényi Péter says:

    @tallbloke – No, it has not. Check texts.

  31. Robert says:

    JusGish Gallop

    Named for the debate tactic created by creationist shill Duane Gish, a Gish Gallop involves spewing so much bullshit in such a short span on that your opponent can’t address let alone counter all of it. To make matters worse a Gish Gallop will often have one or more ‘talking points’ that has a tiny core of truth to it, making the person rebutting it spend even more time debunking it in order to explain that, yes, it’s not totally false but the Galloper is distorting/misusing/misstating the actual situation. A true Gish Gallop generally has two traits.

    1) The factual and logical content of the Gish Gallop is pure bullshit and anybody knowledgeable and informed on the subject would recognize it as such almost instantly. That is, the Gish Gallop is designed to appeal to and deceive precisely those sorts of people who are most in need of honest factual education.

    2) The points are all ones that the Galloper either knows, or damn well should know, are totally bullshit. With the slimier users of the Gish Gallop, like Gish himself, its a near certainty that the points are chosen not just because the Galloper knows that they’re bullshit, but because the Galloper is deliberately trying to shovel as much bullshit into as small a space as possible in order to overwhelm his opponent with sheer volume and bamboozle any audience members with a facade of scholarly acumen and factual knowledge.

    In a debate on the morality of America’s Founding Fathers, a Gish Gallop might look like this:

    “Sure we think that they were good folks, but did you know that Washington not only had more than 100,000 slaves, but he also staged gladiatorial games and made them fight to the death? He also ran a network of opium dens and used his gladiators as couriers to deliver opium all over the 52 states. In fact Washington’s opium smuggling got so bad that the British had to step in which caused the Opium War that led to the Revolutionary War and John Locke’s famous statement that he had to be given the liberty to smoke opium, or he’d prefer death. That also points out another problem, in that most of the Founding Fathers were part of Washington’s opium cult and Ben Franklin’s most harmful invention was actually a process to purify the active ingredient in opium and inject it. That’s right, Ben Franklin invented heroin! What’s more, by the time Andrew Jackson was president the US government was so full of drug addicts that they created a soft drink that was just a way to get cocaine into their systems. Don’t believe me? It was called Coca Cola because it was a cola with cocaine in it. Go look it up and you’ll find I’m right, coca cola really did contain cocaine!”

    by FinnAgain January 07, 2012 t discovered the meaning of