Skip to content

The Met Office & The Pause

February 26, 2015

By Paul Homewood


h/t Green Sand




The Met Office have had another attempt at explaining the global temperature hiatus.

Using, you’ve guessed it, “models”, they conclude:


A new study, led by Met Office climate scientist Chris Roberts and colleagues, looks at this issue by using a huge archive of climate model simulations collected from a number of international research centres. These are used to study whether internal variability has the potential to offset the expected global surface warming rate of 0.2 °C per decade associated with human influences. 
This archive of 15,000 years of simulated climate represents a ‘laboratory’ in which to study the characteristics of the internal climate variability in the absence of changes in external forcing. These would be impossible to disentangle from the relatively short 150 years of the global observational record, where a number of external forcings come into play.
The researchers found that 20-year periods of global cooling in excess of 0.2 °C per decade can occur, as a result of internal variability alone, about once every hundred years.  Once cooling has been established for 15 years, there is also a high chance (up to one-in-four) that this will persist for a further five years.


One would have thought they might have put rather less emphasis on their models, given their abysmal failure so far.




They go on to say:



There has been a recent slow-down in the rate of global surface warming, despite ongoing increases in the emissions of greenhouse gases. This has raised the question of whether internal variability in the Earth’s climate system might be responsible. Here, internal variability refers to natural variations in regional climate, such as El Niño and the Pacific Decadal Oscillation, that arise in the absence of changes in forcing from volcanic aerosols, solar activity or human influences….

The new study also shows that periods of extended global cooling from internal variability alone are associated with more heat being taken up by the sub-surface ocean.



Surprisingly, they forget to mention that, while the PDO has turned negative, we are still in the middle of the AMO warm phase, which began in 1997.




This has the effect of  exaggerating global warming, as NOAA explain:



The solid blue curve shows the observed northern Hemisphere temperatures and the dashed blue curve is a smoothed version. The red curve is the temperature history for a model that responds to the external forcing of greenhouse gases and solar variability but not to natural climate variations. The blue alternations about the red curve represent the natural AMO oscillations. When the AMO decreases, as from 1950 to 1975, global warming may appear to be reversed. When the AMO increases, as from 1975 to the present, the global warming (red) is exaggerated.



When the AMO begins to turn cold, probably in the next decade, we will be likely to see NH temperatures drop as they did after 1940. If this happens, the 15 year pause may well turn into a 50 year one.

  1. February 26, 2015 5:54 pm

    “This archive of 15,000 years of simulated climate represents a ‘laboratory’ in which to study the characteristics of the internal climate variability in the absence of changes in external forcing. ”

    Surely, correct is:

    This archive of 15,000 years of simulated climate represents a ‘laboratory’ in which to study the characteristics of model output variability in the absence of changes in external forcing.

  2. Green Sand permalink
    February 26, 2015 6:03 pm

    Interesting look at the AMO:-

    “AMO by Month from 1909 as of Jan 2015.”

    “The AMO syncs up with climate change very nicely. This is just a view I like. The deepest blue sections are the 1970s.

    1978/79 was the coldest winter in US history.”


    “Sea Ice Extent – Day 42 – Fluctuations Getting Smaller – AMO Peaked?”

    “….The AMO is getting ready for the big plunge (see bottom graph’s) and is only staying high in just a few months – Aug/Sep/Oct.

    It may take decades for the AMO to hit bottom…..”

  3. February 26, 2015 6:14 pm

    Paul, I’ve been following up some of your sleuthing with more extensive data interrogation. You might find the post on Iceland particularly interesting since it shows how the adjustments are done. I’m really quite shocked. We have data deletion, data creation, data changed, superimposed upon wholesale data adjustments.

    Re-writing The Climate History of Iceland
    Temperature Adjustments in Australia

  4. February 26, 2015 7:14 pm

    One of the graphs says “Obsevational Estimates”. WTF?

    • Hector Pascal permalink
      February 27, 2015 3:59 am

      My reaction precisely. You beat me to the comment.

    • February 27, 2015 6:21 pm

      It’s science Jim but not as we know it. RIP Spock who died today.

  5. February 26, 2015 7:27 pm


    There’s a very important fact missing from the graph used by the Met Office headed “Global mean surface temperature anomalies”. There should be a vertical line added to the graph at 2005. Prior to this date the model simulations are simply ‘fitted’ to historical data. It is only post 2005 that the models are true forecasts. It makes it clear to those unfamiliar with the graph just how quickly and by how much the models and empirical data has diverged.

    In most cases where this graph has been printed recently this line has been added with an appropriate legend added, but this is the Met Office we are dealing with here, where obfuscation seems to be the default setting, so it’s no surprise to see the disingenuous approach to presentation.

    • February 27, 2015 10:25 am

      I have been trying to make this point myself for several years.
      Hindcasts are no guide to the future accuracy of models.

      • Green Sand permalink
        February 27, 2015 10:41 am

        QV, what do you think is going on with the thickness of the red “Observational estimates” line?

        If it is to represent “uncertainty” as is seems to be, tighter during the base period 1981 – 2000, why does there appear to be so much “uncertainty” in our present day “Observational estimates”?

      • AndyG55 permalink
        February 27, 2015 10:55 am

        “Hindcasts are no guide to the future accuracy of models’

        Particularly when they hindcast to mal-adjusted data.

        If the did a hindcast to data like it was before it got GISSised, they would have a rather better chance of being somewhere near correct in their projections.

        But if they did that, CO2 sensitivity values would be basically zero, like in reality.

      • February 27, 2015 11:11 am

        At first glance I assumed it was the range of observations, i,e HadCRUT, GISS, NOAA etc, but i see it only mentions HadCRUT on the graph, so it must be the error range.
        I notice the all seem to be below the CMIP5 models, although it doesn’t say which RCP they are from.

      • February 27, 2015 6:17 pm


        Don’t worry about the thickness of the red line (observations). The only red pen they could find in the Met Office was the really thick felt one they use to colour in all their ‘hot’ maps.

      • Brian H permalink
        February 28, 2015 7:09 am

        A true hindcast uses a different starting date and no overlap with the ‘training period’ which is pre-fitted. It’s a ‘hindcast’ because that starting point is prior to the training.

      • February 28, 2015 10:40 am

        I’ve been trying to work out what the “observational estimates” lines represent and firstly, they seem much wider, even than the total uncertainties in the HC4 data file. It may be something to do with the fact that I am calculating the anomalies of the uncertainties, which tend to reduce them.
        Also the figures only seem to go to 2013 and that year is lower than 2012 when I would have expected it to be higher, which makes it appear lower that all of the models.

  6. February 26, 2015 7:59 pm

    Woah Nelly!

    “Once cooling has been established for 15 years, there is also a high chance (up to one-in-four) that this will persist for a further five years.”

    25% is a HIGH CHANCE??? Since when? Is that a “scientific” scale that has been kept secret?

  7. February 26, 2015 8:07 pm

    This is what I wrote to the “Climategate” inquiry of the House of Commons and is now on the UK Parliament website:

    THE PAUSE: “at the end of the 20th century we saw a run of three decades of warming (1970s, 1980s, 1990s) followed by a decade of pause.”

    “Unless the warming in these decades far exceeded the normal inter-decadal change in global temperatures, the mere fact of three decades of warming is in itself entirely consistent with the null hypothesis of natural variation. Indeed a similar warming period occurred pre WWII.”

    “The Null Hypothesis (Natural Variation) is Consistent with Global Temperatures”

    “Natural climatic variation is easily confused with long term change”.

  8. john cooknell permalink
    February 26, 2015 8:47 pm

    My own thoughts are that the Met Office et al will stick with their climate models, I see where they are heading, there is precedent for what they are doing.

    Then, in the not too distant future, they will use the same models to show how bad things would have been had the human world not turned away from adding CO2 to the atmosphere.

    My perception is the UNEP Ozone secretariat is already doing just that. Climate Model Ozone predictions being proved correct by Climate Model verification. Reality need not intervene and confuse things.

    Or perhaps my eyes are blinded by cynicism and doubt!

    • February 27, 2015 10:29 am

      Even if CO2 is not reduced.
      Merely the intent to reduce is good enough.
      Or miniscule reductions in, say the US and Europe while China and India continue to increase.

  9. February 26, 2015 9:22 pm

    Reblogged this on the WeatherAction News Blog and commented:
    “20-year periods of global cooling in excess of 0.2 °C per decade can occur, as a result of internal variability alone, “
    As TB pointed out this explains the warming circa 76-98. That is the death knell for AGW and highlights Lamb’s concerns about the “psychological reactions—even in the influential research community—to the variations towards greater or less warmth as and when they occur.”
    Co2 is busted.

  10. AndyG55 permalink
    February 26, 2015 10:38 pm

    OT, but an interesting article

  11. KTM permalink
    February 26, 2015 11:22 pm

    Just eyeballing it, the actual temperature shift since ~1965 has been roughly 0.3C while the average model has predicted warming of ~1.2C. That isn’t just a miss, it’s a spectacular failure.

    It’s stunning to think that with that minimal predictive ability, the rationale in a nutshell for CO2 being the globe’s thermostat is that ‘we understand all the natural climactic wiggles so anything we can’t explain must be man-made, with CO2 as the prime suspect’.

    They don’t know much of anything, as far as I can tell.

  12. February 27, 2015 2:39 am

    Thanks, Paul.
    Trying not to be unkind, I’ll concede the IPCC models got the sign correctly up to 1998, but that was only in “hindcast”.

    • February 27, 2015 6:09 pm

      Since all model output prior to 2005 is curve fitting/hind casts I think we should all be appalled that the modellers could only get it roughly right up to 1998. A bit negligent you might say.

      • February 27, 2015 6:16 pm

        Actually if you look carefully, for the first 20 yrs or so, the observations are bouncing around the top part of the range, so arguably, even that part of the hindcast overstated warming

  13. tom0mason permalink
    February 27, 2015 2:51 am

    “…Once cooling has been established for 15 years, there is also a high chance (up to one-in-four) that this will persist for a further five years.”

    As usual the Met Office is just unbelievable! They haven’t seen the cooling, they didn’t see the slowdown in warming (aka The Pause or The Hiatus) coming.
    How could the Met Office (of all people) possibly know what will happen in a year or two, or five — they can’t forecast the weather with good accuracy for the next 15 days.

    Barbecue summers, worst floods on record (till you actually look at the records), and children not knowing what snow is. And the long list of unforseen this, that and … etc, etc, etc.

  14. February 27, 2015 6:21 am

    In my opinion the Met study suffers from the same problem as Steinman et al 2015.

    The methodology reflects a profound misunderstanding of how to do science. It is not proper to confound observations and model outputs and then compare the result with other observations.

    What they should have done was to subtract the effects of oceanic oscillations (observations) from observed warming (B-A). Then they would have got C, a vector of values derived only from observations.

    C would be an estimate of observed climate change not caused by whichever combination of ocean oscillations they have selected.

    C is what they should have compared with each model output and the model mean to determine which models get closest to the observations and if the model mean can match the estimate of climate change derived from observations.

    They way they have done the study, model outputs are so entangled with observations we cannot know whether the results depend on the models or on the observations or some undefined mixture of both.

  15. February 27, 2015 10:32 am

    Do they actually believe that studying climate models is the same as studying climate?

  16. Tim Hammond permalink
    February 27, 2015 12:31 pm

    Amidst all the other garbage, is 1 in 4 (and then only “up to”) really a “high chance”?

    Are these people utterly uncritical in everything they write?

  17. February 28, 2015 3:06 pm

    All the media publish any drivel from the AGW lobby ensuring their on-going brainwashing. The only places that I know where any other views are published are Sunday Telegraph, The Engineer, Process Engineering. The Institution of Chemical Engineers has board and editorial policies refusing to publish any criticism of the global heating “proven science” and I was told this by the editor! The I.Mech. Eng. seems the same: where is the debate occurring other than on the excellent blogs like this one? The “I” also does not publish criticisms and the Guardian……

    I have always been suspicious of conspiracy theories, but in this case I am now convinced that the establishment has set-out to stop opposition and free debate. The UKIP policy on energy (virtually the Christopher Booker line) looked promising as a debate point, but has not happened.

    Does anyone have suggestions as to how to raise the debate to its proper place?

  18. March 3, 2015 5:38 pm

    I asked the MO how the red “observaional estimate” line was calculated and they send the following reply:

    “Observed temperature trends. We use the following observational data sets to estimate GMST
    trends: (i) 100 realizations of HadCRUT431 available from
    (ii) Two versions of HadCRUT4 in which unobserved grid boxes are filled using either optimal
    interpolation or a hybrid method that incorporates satellite temperature data19 available from kdc3/papers/coverage2013/series.html. (iii) GISTEMP32 available
    from (iv) NOAA Merged Air Land and SST Anomalies data33
    available from Observed SST trends in the Ni˜no 3.4 region are calculated
    using the Hadley Centre Sea Ice and Sea Surface Temperature gridded data set34 available

    It also states in the paper that the CMIP5 model simulations shown on the graph are generated using RCP 4.5.”

    So at first glance, it appears that the lines are not just HadCRUT4 data as stated in the graph.
    I don’t know if this explains the fall in 2013 but they have also sent me a copy of the paper which includes a slightly different graph which may explain this but I haven’t had a chance to study it yet.


  1. These items caught my eye – 27 February 2015 | grumpydenier
  2. Better a pleb than a denier! | Wolsten

Comments are closed.

%d bloggers like this: