Phil Jones: “We don’t know what natural variability is doing”

Posted: October 17, 2012 by tallbloke in climate, media, methodology, Natural Variation, Politics

The Daily Mail article by David Rose which has caused a bit of a flurry at the Met Office and elsewhere has been the subject of several blog posts recently. I think it’s worth highlighting this quote from Prof. Phil Jones, head of the CRU at the UEA:

We don’t fully understand how to input things like changes in the oceans, and because we don’t fully understand it you could say that natural variability is now working to suppress the warming. We don’t know what natural variability is doing.

Prof Jones admitted that he and his colleagues did not understand the impact of ‘natural variability’ – factors such as long-term ocean temperature cycles and changes in the output of the sun. However, he said he was still convinced that the current decade would end up significantly warmer than the previous two.

The data does suggest a plateau, he admitted, and without a major El Nino event – the sudden, dramatic warming of the southern Pacific which takes place unpredictably and always has a huge effect on global weather – ‘it could go on for a while’.

So, what “convinces” Phil Jones that the “current decade would end up significantly warmer than the previous two”.

Is science about being convinced?

Where’s the beef?

I notice Evan Jones of the Watts et al team got a comment in on the Mail article:

“Our research shows that the data itself is inaccurate. The adjusted Tmean trend for all USHCN stations is 90% higher than the raw trend of well sited rural, non-airport stations, using Leroy (2010) methodology, with all TOBS-biased stations removed from the sample (and with MMTS conversion accounted for). It is 50% higher for the sample of all stations, including airports, semi-urban, and urban stations (again, with TOBS bias removed and MMTS conversion factored in). Since the same boyz who do USHCN also do GHCN, the problem is not confined to US data, but is a worldwide error in methodology. The warming trend, since 1979, at least, is spuriously near-doubled.”

You have to wonder when the temperature record is in the  hands of the strongest advocates of the CAGW hypothesis, both in Britain and  the U.S.

My own investigations show that you can reconstruct the HADsst3 dataset very accurately – correlation R^2=0.9 for monthly data from 1952, 0.87 from 1876) from the naturally varying AMO (Atlantic), SOI(Pacific), SSN (Sunspot number) and a smattering of CO2 to keep the warmies happy (around 0.75C per doubling). The forward projection shows a moderate cooling to 2050.

I’ll put up a post going into more detail about this model soon.

Comments
  1. jazznick says:

    Another thing Jones was ‘convinced’ about in the Climategate e-mails (3259), this one to Tom Wigley.

    Bottom line – their is no way the MWP (whenever it was) was as warm globally as the
    last 20 years. There is also no way a whole decade in the LIA period was more than 1 deg C
    on a global basis cooler than the 1961-90 mean. This is all gut feeling, no science, but
    years of experience of dealing with global scales and varaibility.

  2. tallbloke says:

    The ‘whenever it was’ jibe about the MWP could equally well be leveled at the ‘modern warm period’ where 30% of the country by country records show flat or cooling temperatures in the later C20th…

  3. Joe's World(progressive evolution) says:

    TB,

    The only concern that climate science has is on temperature data and ways to manipulate it.

    They really are NOT interested in actually studying the planet…just the data to try and generate a projection.

  4. tallbloke says:

    I’m going to disagree with Joe here. There is a majority of scientists who are interested in how climate systems work, and a small minority who have been manipulating data to get the co2-centric models to work. Lots of people with many different approaches and motivations can’t be lumped under a single label.

  5. Joe's World(progressive evolution) says:

    TB,

    Your right. I should not have classed every climate scientist with the same brush even though they are still under the umbrella of the “consensus”.

    Judith and others have been trying to open the doors that the majority do not want open to be more open and more honest with scientists and their field of study.

  6. Ray Tomes says:

    There is a clear 208 year cycle in many proxies, solar and climate, going back for millenia. This natural variability was rising for entire 20th century and falling for 21st century. We don’t know know why it happens but it does. It seems to me that AMO and SOI etc are not useful for predictions because we cannot predict them very well (yes I know there is a 60 year cycle) and they are closely related to temperature anyway.

  7. Ray Tomes says:

    Someone needs to write a book about “temperature adjustments”. Given the urban heat islands and such the adjustments should be mainly downwards, but they are mainly upwards. If you search for these terms you will find that data is being fiddled in USA, UK, NZ, Australia, and possibly elsewhere. If you don’t conclude that it is a conspiracy, then you didn’t get the facts right.

  8. tallbloke says:

    Hi Ray.
    I know the 208 yr ‘de Vries’ cycle shows up as a reasonably sharp peak in specral analysis of many different proxies. Is it possible to get an idea of it’s range of variability or is the data too noisy?

  9. Ray Tomes says:

    Hi Rog, give me a temperature series and I will give you the coefficient for the 208 year cycle. I don’t know what temperature series to trust. 🙂

  10. Tim Cullen says:

    “We don’t know what natural variability is doing.”

    Does that mean they can’t read a thermometer?
    Does that mean their science is worst than originally thought?

  11. tallbloke says:

    Ray, try the c14 ice core proxy.

  12. donald penman says:

    “We don’t know what naural variability is doing”, which is a good reason for having more research into natural variation rather than manmade variation and co2.

  13. Tenuc says:

    Meanwhile, over on Lord Lawson’s Global Warming Policy Foundation web site we have a very damning article which refutes the CAGW mantra that current warming is exceptional…

    The title of the paper is, “The extra-tropical Northern Hemisphere temperature in the last two millennia: reconstructions of low-frequency variability,” by B Christiansen of the Danish Meteorological Institute and F C Ljungqvist of Stockholm University.


    “Reconstruction of the extra-tropical NH mean temperature (C) based on the gray-shaded proxies in Table 1 reaching back to at least 300 AD. Calibration period 1880–1960AD. Only proxies with positive correlations and a p-value less than 0.01 are used. The included proxies are given in the legend. Thin curves are annual values; thick curves are 50-yr smoothed.

    Red curves show bias and confidence intervals for the 50-yr smoothed values. From ensemble pseudo-proxy studies mimicking the reconstructions, we have calculated the distribution of 50-yr smoothed differences between reconstructions and target. The biases and the upper and lower 2.5% quantiles are calculated from these distributions. In the figure the biases (full red curves) have been added to the real-world reconstructions. Likewise, the upper and lower quantiles have been added to the real-world reconstructions (dashed red curves).

    The green curve shows the observed extra-tropical (>30 N) annual mean temperature. The yellow curve show the temperature average over grid-cells with accepted proxies. Both curves have been centered to zero in 1880–1960 AD.”

    Article here…
    http://www.thegwpf.org/climate-warmer-1000-years-ago/

    Full paper (free) here…
    http://www.clim-past.net/8/765/2012/cp-8-765-2012.pdf

    Makes one suspect that Prof. Phil Jones and the rest of the IPCC cabal of CAGW cargo-cult climatologists are being very economical with the truth.

  14. Michael Hart says:

    Phil Jones: “We don’t know what natural variability is doing”

    Sid Vicious asked Lemmy, the bassist of Motörhead, to teach him how to play bass with the words, “I can’t play bass,” to which Lemmy replied “I know.”

  15. Brian H says:

    donald penman says:
    October 17, 2012 at 4:52 pm

    “We don’t know what naural variability is doing”, which is a good reason for having more research into natural variation rather than manmade variation and co2.

    Remember how CO2 was justified? Adding it to non-functional models of natural variation made them work (hind-casting). But if the plug no longer works …

    LOL

  16. tallbloke says:

    I can get as good a correlation between my model and HADsst3 whether I give more weighting to the solar component than co2 within the range of uncertainty of the effect of either, and the same vise versa with more weighting given to co2 than solar.

    This doesn’t prove that the climate is insensitive to both, because both the cumulative sunspot count and co2 rose over the latter C20th, along with global temperature.

    But now the sun has gone quiet and the cumulative count running total is falling, and co2 is continuing to rise, and the temperature has flatlined for 15 years. This means nature is now ready to perform the crucial experiment and show us whether it was solar or co2 which was responsible for the C20th warming. Will temperature go up? Or down? or stay level?

    Of course, we have to remember that if we subtract the AMO and SOI from both the model and the SST curve, the whole thing looks a lot flatter anyway, and the timescale we have to wait for a definitive answer stretches out.

    But if the co2 adherents claim that a flatlining or fall in temperature is due mainly to the oceanic oscillations and the Sun has little effect, then they have to accept that a good proportion of the rise in temperature from 1975-2003 was due to them also.

  17. tallbloke says:

    Bill Illis has also produced a model similar to mine which he tests against HADcrut4.
    He also includes volcanic and SAMO and gets a correlation of R^2 = 0.8
    Bill includes a ‘warming trend’ of 0.045C per decade
    He says he can get a closer fit if he uses a ln(co2) value of 2 which is around 1.3C per doubling.

  18. oldbrew says:

    Re the HADCRUT 4 graph:We could be essentially repeating the 1945-1975 sequence, just from a slightly higher temperature starting point? Climate science orthodoxy might not want to hear that.

  19. oldbrew says:

    Again re HADCRUT 4: 1880-1911(-ish) appears to follow the same pattern. I know it’s not news, but just to re-inforce the point.

    NB slightly OT but Steve McIntyre has a great comment at climateaudit talking about ‘social priming’.

    Quote: ‘As partial support for the concept of “social priming”, it seems to me that there is a statistically significant increase in the incidence of drivel in writings by activists after being primed with words that relate to “climate skeptics”. This hypothesis will be more difficult to test among authors where the incidence of drivel is already high, even without social priming.’

    http://climateaudit.org/2012/10/14/lewandowsky-and-hide-the-decline/

  20. Ray Tomes says:

    Hi Rog

    The C14 cycles analysis including 208 year cycle is shown at http://cyclesresearchinstitute.wordpress.com/2010/06/27/cycles-in-sunspot-number-reconstruction-for-11000-years/

    There is an apparent modulation (perhaps 2300 years) which is the beats of the 224 and 208 year components. There are good planetary reasons for a 2300 year modulation, but no obvious reason for the 208 year cycle.

    The Be10 analysis at http://cyclesresearchinstitute.wordpress.com/2011/07/13/analysis-of-be10-records-as-a-solar-irradiance-proxy/
    also shows 207 year cycle also with a modulation. Both have the same phase for 208 year cycle (maximum about 2000 AD) and same phase for the ~2300 year modulation which is maximum amplitude around 3100 BC, 600 BC and 1500 AD.

    Of course these reconstructions are for solar output / sunspots numbers not for temperature. In terms of most reasonable temperature reconstructions, the amplitude of the 200 year component is likely to be just 0.1 or 0.2 C (i.e. range 0.2 to 0.4 C).

    However it is my conviction that all proxy series underestimate temperature fluctuations substantially. I would say that Mann and Jones reconstructions for the period before thermometers need the variations to be multiplied by 4 to get realistic temperature fluctuations. This is a natural consequence of the mathematics of making proxies with rather low correlations and then averaging them together.

    Regards
    Ray

  21. Ray Tomes says:

    BTW, if this multiplication by 4 for pre-instrumental proxies is done, then the hockey stick turns into something more reasonable.

  22. oldbrew says:

    TB says: “But if the co2 adherents claim that a flatlining or fall in temperature is due mainly to the oceanic oscillations and the Sun has little effect, then they have to accept that a good proportion of the rise in temperature from 1975-2003 was due to them also.”

    Exactly.

  23. tchannon says:

    Tenuc,
    I’d be interested in seeing what Steve Mcintyre has to say about Christiansen and Ljungqvist

  24. Ray Tomes says:

    Regarding fitting data with cycles and making the elephant wave his trunk …

    There is a test called Bartell’s test for significance of a cycle period in a set of data. It is based on the consistency of a cycle over a period of time – is the phase reasonably consistent every cycle so that when phase and amplitude of each wave of the cycle plotted on a plane give a scatter that is not explicable by expected null hypothesis. When a cycle has p<0.05 we might find it interesting, but when it has p<0.0001 we can really get excited. There are many cycles in solar activity in its various forms that have p much less than 0.0001.

    I highly recommend that people working with cycles and regressions do look at the statistics. Cycles Research Institute offers a free package (CATS = Cycles Analysis Timeseries Software) which includes the ability to look for cycles (with Bartels's test), perform regressions etc and be sure about how meaningful results are. See http://www.cyclesresearchinstitute.org/cats/index.shtml

  25. Paul Vaughan says:

    Universal (for planet-star pairs):


    +

    =

    (slow animation of preceding pair)

  26. oldbrew says:

    Phil Jones: ‘We don’t know what natural variability is doing.’

    Here’s a clue perhaps. NASA says ‘current estimates indicate that the Sun is probably now as active as or more active than it has ever been during the past 8,000 years.’

    http://earthobservatory.nasa.gov/blogs/climateqa/has-the-sun-been-more-active-in-recent-decades-and-could-it-be-responsible-for-some-global-warming/

  27. tallbloke says:

    The 8000 years quote is from Sami Solanki, chief solar physicist at the Max Planck Institute in Zurich. Whether or not that is true, it’s clear that the Sun has become generally more active since the Maunder minimum at the depth of the little ice age. The effects of raised solar activity levels play out over long periods of time, because the ocean has a lot of inertia and so the heating of it by raised activity levels takes a long time. That’s why arguments that the Sun can’t be responsible for the late C20th warming because the peak amplitudes of the cycles declined after 1960 are spurious. The key point is that solar activity remained at levels well above the long term average, and above the level whjere the oceans gain heat from solar radiation faster than they lose it, for a long time in the C20th, and all the way to 2003.

    To simulate the effect of solar radiation on ocean heat content, I integrated the sunspot number as a running total departing from the ocean equilibrium value (40SSN). That’s the yellow curve in the components plot in the headline post.

  28. tallbloke says:

    Ray, many thanks for the links to the CATS software. There is so much to be discovered that has been ignored by mainstream climate science!

  29. tallbloke says:

    Paul, Interesting plots. How do you define ‘solar cycle deceleration’ ?
    Clearly it’ll involve solar cycle lengths, but more specifically.

    Thanks.

  30. Paul Vaughan says:

    TB, I see oldbrew has addressed your SCD question.

    The same method applied to semi-annual LOD gives a curve that matches ENSO indices.

    The method derives universal constraints for systems with memory & nonstationary drivers.

    (Someone with the time could find engineering data to illustrate application of the method in familiar mechanical settings.)

    I advise all sensible parties to completely extinguish their tolerance for “sun has no effect” dark ignorance &/or dark deception artistry.

  31. the1pag says:

    Can undoubtably cooler Atlantic Ocean water to be encountered by hurricane Sandy as it moves northward perhaps temper its ferocity beyond that being hyped by East Coast media, particularly where i live here in NJ where NOAA is showing it may come ashore?

  32. tallbloke says:

    Hi Bob, and welcome. I don’t know how it’ll play out. I hope you and yours stay safe, warm and dry, whatever happens. Our ability to predict the magnitude and direction of storms is better than it was decades ago but is still a long way from reliable.

    Take care.