Yucatan Coral dataset oddity

Posted: October 23, 2012 by tchannon in Analysis, Cycles

Image

Citation: Vásquez-Bedoya, L. F., A. L. Cohen, D. W. Oppo, and P. Blanchon (2012), Corals record persistent multidecadal SST variability in the Atlantic Warm Pool since 1775AD, Paleoceanography, doi:10.1029/2012PA002313, in press.
http://www.agu.org/pubs/crossref/pip/2012PA002313.shtml

[updated post 27th Oct, clarification of plot, see comments, was not intended as a formal work]
Read a little on the net, noticing Rog had posted a link to GWPF giving a link to a dataset
https://tallbloke.wordpress.com/2012/10/21/roger-andrews-how-nasa-giss-manufactures-warming-in-the-antarctic/#comment-33579

On looking GWPF, broken link, figured and reported it, pulled the small dataset. Odd moment thing is tell software here to go do a quick look, only takes a few seconds, and see what gives.

The results tweaked my interest when I looked at the structure. I decided to try a couple of daft things, never know what insight will fall out. This was a case of keep on simplifying.

Second try… definite boggle at the result, never seen this before and is curious enough to warrant an article, quicker anyway that than trying to put plots in a comment.

(note: I am not saying this is a fixed entity, this is pondering, something like this might appear often)

The match is no big deal except for one thing, this has mimicked what seems to be a ~55y wave, perhaps akin to the much vaunted 60y and yes it matches that more or less.

A twist, you are seeing pure modulation products. It is actually a pair of closely spaced sines producing a product. This is what the software did when told to use just one generator but allowed to use modulation. There is no 55 year at all.

It is a pair of ~110 year producing sum and difference product, one being 55y. The other one? This will create some interest

Pasting, 54.9 and 8229.9, comes from 110.53 and 109.07

Maybe, needs some digging

“The ~3 m cycle corresponds to individual paleosols, with a period of 7–8 k.y. This period is similar to millennial-scale sub-Milankovitch cycles found in marine and lacustrine successions of Pliocene–Pleistocene age.”
http://geology.gsapubs.org/content/36/7/531.abstract

Whether such a thing could be found in such a tiny dataset (in time) seems unlikely. Might just be noise playing games.

We know there is circa 100 year in many datasets so that is not a great surprise however it is worth keeping in mind deVries is a loose spectral construct out to very long periods. A commonly mentioned one is ~200y so these ~110y could be to do with the second harmonic.

There seem to be many of these weak-strong-weak-strong waveforms around so having one produced by modulation might be an area ripe for something to have a good think and look. Keep in mind though there has to be mechanism, can’t have effect without a rational way it could happen.

zip archive contain an .ods spreadsheet is here Yucatan-coral (model is live and can be extended to see the longer waveform, change output sample rate or extend by copy/fill)

Data links
ftp://ftp.ncdc.noaa.gov/pub/data/paleo/coral/caribbean/vasquez-bedoya2012.txt
ftp://ftp.ncdc.noaa.gov/pub/data/paleo/coral/caribbean/vasquez-bedoya2012.xls

Comments
  1. To tchannon: you can clearly see the 59,57 yr. Jup/Sat Scafetta cycle…here for
    200 years. We can go further back for 10,000 years all along the Holocene, and
    you may detect the corresponding 59,57 year peaks as mini-spikes in the GISP2
    data set……..Here proven in Yucatan corals…. and there exist paleo-AMO-studies,
    proving the same 60 year cycle…. JS

  2. In our recent paper identified as one macro-forcing mechanism, macro signifies
    that it is clearly visible in Holocene proxies….

    [Moderation note] J Siefert’s self published paper has been discussed here and at WUWT. We are awaiting proofs for various aspects.

  3. vukcevic says:

    Hi Tim

    I wouldn’t think it is unusual. 55ish year period is common in the N. Hemisphere.
    http://www.vukcevic.talktalk.net/Arctic.htm
    About a year I discussed privately with Dr. J Curry link between Arctic atmospheric pressure and the N. Atlantic’s subequatorial ocean temperatures (source of hurricanes)
    http://www.vukcevic.talktalk.net/AHA.htm (with 15 year delay, finally worked out the delay)
    I suggest not a fix, just Natural Variability.
    All coming from the Earth’s core 55 and 83 year periods, number of times mentioned on WUWT, Dr. S says : NO.
    You may hear about it again elsewhere soon, but vukcevick found it out some while ago.

  4. vukcevic says:

    On the other hand there is ‘solar’ cross-modulation with about 105 year period.
    http://www.vukcevic.talktalk.net/LFC4.htm

  5. tallbloke says:

    Excellent work Tim. Roger Andrews found a ~110 year signal in the interplay of SST and SAT. Anthony and Basil found 54.56 years in their study. Getting De Vries from 110.53 and 109.07 is tricky though. It is shorter than double those periods. Unless the shorter periods are being modulated by something else?

    Let’s keep digging. 🙂

    Vuk: Fascinating, you are getting there.

  6. Ray Tomes says:

    Rog, the long Be10 and C14 series both show cycles of around 208 years and 104 years and several cycles around 50-60 years. With about 260 years data, you can determine the period of a cycle around 55 years with an accuracy of typically +/-1 year. You cannot determine two cycles of 110.53 and 109.07 years. You cannot detect a modulation of 8000 years.

  7. tallbloke says:

    Hi Ray,

    Yes, I recognise that the exact periods and beat periods Tim came up with are speculative. We’re just cruising around the permutations to see if anything crops up which works on ‘several different lines of evidence’.

    When we did the cycles analysis on the Lean’s TSI reconstruction we came up with a periods of 57.2 and 112.5 years among others, including one at 11.51 and another at 11.09. https://tallbloke.wordpress.com/2011/02/21/tallbloke-and-tim-channon-a-cycles-analysis-approach-to-predicting-solar-activity/

    11.09 is close to J-E-V at 11.07, and 11.51 is near the average of 11.07 and Jupiter’s orbital at 11.86

    11.08+11.86/(11.86-11.08)= 29.41, which is very close to Saturn’s orbital period of 29.45 years.
    Earth’s orbital period 1 divided by Saturn’s orbital period 29.45 is 0.033. Multiply this by Earth’s orbital period in days and you get 12.4 days, which is half the Sun’s equatorial spin rate and a third of it’s polar spin rate. Another of those curious things which shouldn’t work.

    More similar curiosities here.
    https://tallbloke.wordpress.com/2011/11/19/solar-planetary-spin-orbit-coupling-more-evidence/

    I’m still thinking there will be a fundamental physical relationship which will explain this. I just need to work out what it is. On the face of it it doesn’t seem unreasonable to me that there would be a link between the rate the Sun spins at and the rate the planets orbit around it. After all, we know there was a strong spin-orbit coupling in the early solar system. Something is maintaining it, and I’m going to find out what it is. 🙂

  8. mmhhh….

    “….~55y wave…..”

    http://www.ncgt.org/newsletter.php?action=download&id=113

    9/56 YEAR CYCLE: WORLD MEGA VOLCANIC ERUPTIONS

    David McMINN
    Independent Cycle Researcher

    Abstract: A 9/56 year cycle was found to be applicable in seismic timing for various regions and countries around the world. Additionally, a 9-45 year cycle was established for major world earthquakes (M => 8.5) since 1900. Such cycles were also hypothesised to arise in the timing of mega volcanic eruptions over the past few centuries. A 9/56 year grid was confirmed for more moderate world eruptions (VEI = 4). However, this was not observed for larger events (VEI
    => 5) and a 9-27/56 year grid was found to be more relevant for these mega events. Strangely, eruptions causing major loss of life could be correlated with a 9/56 year grid, which seemed unusual. It was speculated that patterns based on multiples of 9 and 56 years were caused by Moon-Sun tidal harmonics triggering mega seismic and eruptive events.

    🙂

  9. tallbloke says:

    Michele, yes, 3 times the Lunar Nodal Cycle of 18.61 years is 55.83 years
    45 years is the return period of the inner solar system planets.

  10. tchannon says:

    This is a machine least squares fit to an ambiguous problem.

  11. Henry Pool says:

    Henry@tchannon
    this is not a coincidence. There is some lag with my own curve here,
    http://blogs.24.com/henryp/2012/10/02/best-sine-wave-fit-for-the-drop-in-global-maximum-temperatures/
    but I would expect that as I am looking at energy-in whilst you are looking more at energy out.

    The 88 year energy-in is the important one to keep an eye on. I found a clear correlation with the Nile flooding as well, which also follows on a 100 year cycle.

    This worth looking into.
    As to the mechanism:
    So far, I do not exclude a gravitational or electromagnetic swing/switch that changes the UV coming into earth. In turn this seems to change the chemical reactions of certain chemicals reacting to the UV lying on top of the atmosphere. This change in concentration of chemicals lying on top of us, in turn causes more back radiation (when there is more), hence we are now cooling whilst ozone & others are increasing.

  12. suricat says:

    Can anyone tell me what the blue and red traces in tim’s graph represent?

    In the meantime, here’s a paper on the tourism forcing to ‘Yucatan coral’:

    Click to access Bozec_2008_ICRS.pdf

    Best regards, Ray.

  13. tchannon says:

    Straight plot of the dataset and a plot of a very simple model derived by software from the dataset.

    Didn’t bother doing the niceties on a quick show but everything is in the spreadsheet inside the zip, including the dataset reference.

    I sometimes take a quick look at datasets as they come to my attention.

  14. suricat says:

    Thanx for your reference tim, but columns “F, G & H” are not disclosed for their origins and are also obscured for rows “72, through 86” by your ‘graph’ (I’m guessing that the L/H scale of the graph is for temp anomaly [or is this growth and for which colour?]). How can this be resolved?

    Best regards, Ray.

  15. tchannon says:

    Try and have a look tomorrow.

  16. tchannon says:

    Edited the article. I hope that is clearer.

  17. suricat says:

    No problem tim. 🙂 I didn’t realise that the download wasn’t a ‘locked’ spreadsheet and I can alter the format at my end.

    I’m not used to free access on the net. Thanks. 🙂

    Best regards, Ray.