Skip to content

Spot The Difference

January 10, 2013

By Paul Homewood







There has been much discussion in recent days about the new decadal forecast of global temperatures, sneaked out by the UK Met Office on Christmas Eve, and which shows flatlining temperatures instead of the the rapidly increasing ones previously forecast.

However, on studying the current forecast (at top), and comparing it with the previous version (below) issued in December 2011, I noticed a bit of jiggery pokery. Have a look and see if you can spot it.

The white curves are previous predictions, as the narrative explains. In the latest version, this line heads downwards from around 2005 to today’s levels. In other words, they seem to be giving the impression that previous predictions anticipated the drop in temperatures in the last couple of years.

Yet look at the Dec 2011 version, and you can see that this is absolutely not what they were predicting then. On the contrary, they were forecasting a significant increase in temperatures throughout the period.

It would appear that the Met have deliberately fabricated a new version of their Dec 2011 forecast, in order to avoid making the original version look too ridiculous.


Is this really what “science” has come down to?

  1. January 10, 2013 8:22 pm

    You have to watch the pea. They don’t have someone called John Cook working there, do they? If I knew they were up to this sort of thing I’d have sent them a bottle of Snopake well before the adjustment season started, though with our postal service it’d have arrived well after the 24th. One day kids just won’t know what Snopake is.

  2. Green Sand permalink
    January 10, 2013 9:17 pm

    Paul I am not sure about this in the past the “previous” or “retrospective” forecast have in fact been “hindcasts” it is an issue that has been taken up with Richard Betts.

    If they are hindcasts then it is possible they have been carried out by two different models the first – 2011 by HadCM3 and the 2012 using HadGEM3.

    More info here:-

    Sorry if this info turns out to be wrong but I believe the confusion is based on the misrepresentation of hindcasts. Please check it out, can’t do more at present I am under time constraint.

  3. Stephen Richards permalink
    January 10, 2013 9:30 pm

    I also am unconvinced that this is some new model. Ithink it possible that this is merely jigging about the input parameters to come up with an answer that suits their purposes and your observations appear to confirm that suspicion.

  4. January 10, 2013 9:48 pm

    Is this really what “science” has come down to? Yes, it is! They reshape the past to validate their future. They adjust the temperature record to validate their preposterous theory.

  5. January 10, 2013 9:56 pm

    Anybody who can superimpose the two images? The old one has quite large gaps no longer visible?

  6. January 10, 2013 10:47 pm

    The white line in the December 2012 wishcast would include the 2012 portion of the 2011 wishcast. They apparently simply connected the old previous wishcasts to the most recent wishcast and called that the previous prediction. I don’t think it really matters at this point as nobody is paying much attention to anything the Met Office says concerning events more than a week out. They have basically made themselves irrelevant by pretending to speak with authority and then badly missing the prediction. Now they are trying to pretend they were right on the money the whole time. Nobody believes that. It is pure desperation.

    Actually, whoever approved that graph for publication should be immediately sacked. It is just a flat out lie.

  7. tckev permalink
    January 10, 2013 11:07 pm

    This and NASA’s new view that the sun does more than they initially thought. (

    And as suspected CO2 is not THE culprit ( , and many, many more)

    So, when is the government going to roll back all the CAGW legislation?

  8. Paul M permalink
    January 11, 2013 12:08 am

    It was noted elsewhere that the “previous prediction” in the new version had “predicted” the eruption of Pinatubo. Clearly the Met office models are getting really good if they can predict volcanoes.

  9. lapogus permalink
    January 11, 2013 12:43 am

    Sadly this is what we have to expect nowadays from the alarmists in the Met Office and elsewhere. Regardless of the failure of the models to reliably cope with the non-linear chaotic climate system, the additional global warming [I’d suggest about 0.3C] they have been able to extrude from a combination of UHI, dubious station selection, homogenisation and bald adjustments to the instrumental data has been spent, and is now going to play against them, even if the actual global average temperatures flat-line.

  10. Steve permalink
    January 11, 2013 7:12 am

    There is an understandable misunderstanding here. “Previous predictions” really means “Hindcasts of previous periods”. Presenting the hindcasts alongside the forecast is, I assume, intended to give you another picture of how accurate/inaccurate they think the forecast is going to be (eg, to me it seems the hindcast clearly doesn’t pick up on some of the big ENSO events). To make the comparison appropriate, both the hindcast and the forecast have used the same model – i.e. the old graph uses the old model and the new graph uses the new model.

    Clearly as they thought that everyone probably knows that there were no decadal predictions in the 60s, 70s, 80s and 90s, the use of sloppy language was able to slip through.

  11. Steve permalink
    January 11, 2013 7:17 am

    Actually, if you go to the release of the research, the very first statement in the main text is:

    “During 2012 our decadal prediction system was upgraded to use the latest version of our coupled climate model. The forecasts and retrospective forecasts shown here have been updated to reflect this change.”

  12. January 11, 2013 9:34 am

    I don’t think that they are “previous predictions” as in predictions made years ago. I think that they are model runs started in past years to validate the model. If you look carefully at the graphs, both sets of white lines look to be different on both graphs. This suggests that they came from different models.
    The latest graph suggests that the new model has more consistently predicted lower levels of warming which is in line with the latest decadal forecast from the MO.

  13. Barry Woods permalink
    January 11, 2013 9:34 am

    ” comments I think worth highlighting from Bishop Hill
    (Richard Betts = Met Office, Head of climate impacts)

    Yes, “retrospective forecast” means hindcast here.

    Jan 10, 2013 at 11:11 PM | Richard Betts


    Paul Mathews is a Reader of Mathematics at Nottingham University.

    Richard, if the white line is a hindcast, why is it described as a “previous predction”?

    For the umpteenth time, the cynical attitudes expressed here are a direct result of this type of behaviour from the met office.

    Jan 10, 2013 at 11:43 PM | Paul Matthews

    lots of similar thoughts in the comments

  14. Marc Fearby permalink
    January 11, 2013 9:40 am

    What sort of moronic scientist considers white a good colour for a series in a graph with a white background? These chaps are indeed bonkers!

  15. Chris permalink
    January 11, 2013 9:55 am

    It’s those magic 5-day forecasts again. They are probably updating their forecasts and only publishing the last ones. That’s how many financial analysts get it right every time too.

  16. Dave permalink
    January 11, 2013 10:09 am

    This difference between the two figures is because when you work with models you start the model running at some date in the past and see how well it “predicts” events that have already happened – this allows you to assess the accuracy of the model. If it works well you can then run it into the future to make a reasonable guess at what will happen.

    The top figure shows that the new model better “predicts” past events than the old one. This is why the Met Office have updated their model.

    The Met Office explain this on their website:

  17. Barry Woods permalink
    January 11, 2013 10:16 am

    New Graph, implies they accurately forecasted the eruption of a volcano Mt Pinutobo

    (before it happened) and the also forecast resulting temp change very accurately (compared to prev graph)

    quick somebody tell the volcano experts the Met Office can predict major volcanic eruptions!!

    I’m beyon being a climate sceptic now – climate cynic?

  18. dearieme permalink
    January 11, 2013 11:18 am

    Remember Paxman’s question: why are these lying liars lying to me?

  19. January 11, 2013 12:04 pm

    From the New Scientist:

    “Even so, the fundamental physics about how greenhouse gases trap heat in the atmosphere hasn’t changed. And we know that, even as atmospheric warming stalls, the oceans are continuing to warm. That may could explain why Arctic sea ice melted so dramatically last summer, even though air temperatures were not exceptional.”

    Yet when I plot the SST figures from HadSST2 for 1998 to 2012, the linear trendlines give:

    NH SST y = 0.0044x + 0.3421

    SH SST y = -0.005x + 0.3299

    Global SST y = -0.0003x + 0.3359

    So how can we say that the oceans are still warming, surely this shows they are now in cooling phase?

    • Craig King permalink
      January 11, 2013 2:15 pm

      Yes indeed Adrian, how very can they?

      Barry is right, climate cynic it is.

  20. January 11, 2013 12:52 pm

    This article appears to have misunderstood what the two graphs are showing, as some of the other comments have already noted.

    To explain: The red shaded area and the white line in the middle on the 2011 graph (shown bottom in your article) shows hindcasts from the old HadCM3 model – we run the model for previous years to see what it would have predicted and then we see how closely it matches the real-world observations. This gives a measure of the skill of the model. For the 2012 forecast (shown top in your article), we upgraded to the HadGEM3 model, and as explained before, we tested this thoroughly using the hindcast from that model – I’m sure you’ll agree it wouldn’t make sense to give a forecast with one model, and a hindcast using another. What you can actually see from the new shaded red area and white line in the 2012 forecast is that our new model HadGEM3’s hindcast much more closely represents real-world observations – indicating a higher level of skill. So this article from is really based on the fact that the author has noticed our new model has increased skill.

    • January 11, 2013 1:06 pm

      Thanks, Dave.

      I think in that case you need to alter the narrative to the graph, which says

      Previous predictions starting from June 1960, 1965, …, 2005 are shown as white curves,

      as this is a highly misleading statement.

      I think it would also be a good idea to issue a separate graph, comnparing earlier forecasts with the current one and actual temperatures.

      I would be happy to publish this if you could arrange something.

    • Paul M permalink
      January 11, 2013 3:32 pm

      Dave Britton, are you being deliberately obtuse? Surely you are aware that the problem arises from the false description of hindcasts as “previous predictions”.

    • MikeC permalink
      January 11, 2013 4:49 pm


      You say Figure 1 shows the hindcast from the new model (yes, that explains the difference obviously with the previous Figure 1).

      But under the a subsequent section call “Verification” Figure 3 shows a comparison between (a) Observations November 2007 to October 2012 and (b) Five year mean forecast from November 2007.

      The “previous predictions” in Figure 1 are not from 2007 as you have just stated they are from the new model. What am I missing here?

  21. John Blake permalink
    January 11, 2013 1:44 pm

    More Slingo-slinging: No doubt the Madame’s bonus evaluation will cite “performance beyond the call of duty.” Claptrap amounting to fraudulent misrepresentation is the Met Office’s –and indeed the Cameron Administration’s– entire stock-in-trade.

  22. Michael Martin-Smith permalink
    January 11, 2013 3:20 pm

    It appears that predictions of Climate Changes are best left to the Retrospectroscope. We should certainly NOT be using them to inform taxation, energy or regulatory decisions.
    The real world economy has faced and continues to face recession and unemployment in all meaningful activities apart from officious regulation, debased money, and taxation to little benefit for real people..

    The Climate Change and ludicrous Green Energy industries now gross hundreds of billions of pounds, and are now seen to be largely fraudulent seedbeds of bureaucracy and unnecessary costs for hardpressed consumers.

    Take, for example, the polluting low energy lightbulbs made compulsory by Chancellor Angela Merkel ( now known in my home as “Merkel” bulbs). Only a politician could have dreamt that one up, and then belatedly told us that we could not put them in household rubbish bins because of their environmental toxicity ( yes, really!). Costly, polluting, and inefficient – just like a politician, in fact..
    .LEDs are dear but at least light up fully, at once,and have a 10+ year lifespan. Why didn’t they wait until LEDs became available at 100 watts equivalent and at a quarter of their present price before legislating at all?

    Perhaps unemployment and recession are NOT the real problem- but their distribution is.

    Can we look forward to taxmen,, and the ubiquitous Climate Change advisory industry regulators crowding out JobCentres? No, for they would be unemployable in a healthy society.

    If this continues, lamp-posts will soon acquire another function…

    • John permalink
      January 11, 2013 5:09 pm

      Philips Lighting has been the most profitable part of Philips for ~ hundred years.

      My bet is that Philips successfully lobbied for the light-bulb change, so low-tech, low cost light bulbs could be replaced by fancy, expensive, new CFLs protected by all kinds of patents.

      Now, how did the lobby cash get into politicians pockets?

  23. January 11, 2013 5:10 pm

    New old joke

    Q- How’s the Met Office like certain members of the Royal Family?

    A-No matter what they do it ends up in controversy, misunderstanding and a lot of raised eyebrows.

  24. January 11, 2013 5:17 pm

    According to the Wiktionary “previous” is synonymous of “former”.

    IOW “what was but is no more”

    Nobody who’s interpreted the latest brouhaha as a rewriting of history should feel compelled to apologise about what’s been a natural and logical reaction based on standard English usage.

  25. January 11, 2013 10:06 pm

    The only good thing about all this is that it is becoming increasingly obvious to everyone that the alarmists lie and lie again. I’ve never known such a group to say one thing, have it proved wrong, then come back and say they never said that. They smudge over (re-write) the past and hope nobody notices? Incredible!

  26. prairledogged permalink
    January 13, 2013 9:54 am

    “Day by day and almost minute by minute the past was brought up to date. In this way every prediction made by the Party could be shown by documentary evidence to have been correct; nor was any item of news, or any expression of opinion, which conflicted with the needs of the moment, ever allowed to remain on record. All history was a palimpsest, scraped clean and reinscribed exactly as often as was necessary.”

    ― George Orwell, 1984

    • GrumpyDenier permalink
      January 14, 2013 2:16 am

      Thanks for that. I’ll have to go back and read that again. It was a long time ago and I’ve forgotten the lessons. I’ve nicked that for future consumption.

  27. David Peltz permalink
    January 13, 2013 10:10 am

    If those graphs (the vertical axis of which appears to be clearly mislabelled) are accurate indicators of average annual global temperatures one can only draw a few conclusions………. 1) Over any decade average temperatures vary in a pretty consistent pattern of a maximum of 0.3 degrees approximately. 2)from 1974 approx to 1996 approx, there was an increase of approx 0.4 in the mean. 3) Since 1996 approx the mean is unaltered. I don’t have access to graphs for levels of Industrial/Transport Activity but if the increase in this period isn’t showing a pretty unbroken sharp increase then I’d be surprised, notwithstanding the GFC. Motives, viewpoints, camps, conspiracies etc are irrelevant.Global warming as a result of increased activity seems unlikely unless it is masking/masked by other factors with greater primacy. No, just making everything stink, using it all up, covering it with ugly concrete just because families are no longer fashionable are far greater reasons to re-cycle, control building, than a theory of cataclysm that might be possible but pretty unlikely.


  1. We apologise for the late arrival of Global Warming....
  2. Did Met Office Manipulate Forecast Data? | The Global Warming Policy Foundation (GWPF)
  3. Eye on Britain (2)
  4. Preliminary Tornado Report for 2012: a low year | Watts Up With That?
  5. More global warming silliness | Western Free Press
  6. Met Office Decadal Forecast – 2007 Version | The Global Warming Policy Foundation (GWPF)

Comments are closed.

%d bloggers like this: