My simple solar-planetary energy model

Posted: January 5, 2010 by Rog Tallbloke in climate, solar system dynamics
Tags: , , , , , , ,

A lot of people are puzzled by the current El Niño. Global average Sea Surface Temperature (SST) has been high, but we don’t seem to have the balmy winters of ten years ago. My simple model explains why.

Temperature reconstructed from solar and planetary motion

The graph compares sea surface temperature HADsst2GL (red curve), with curves generated from solar and planetary data.

The black curve uses a combination of Length of Day (LOD) data and sunspot number data. The monthly sunspot number values are added cumulatively as positive or negative values departing from my estimated ocean equilibrium value of ~40SSN. The LOD values are added via a simple best fit scaling technique using a hghly sensitive piece of equipment called tallbloke’s eyeball.

The yellow curve uses the sunspot numbers again, but instead of LOD data, I use the fact that LOD variation approximately correlates with variation in the distance of the solar system centre of mass in the ‘z’-axis from the solar equatorial plane (SSB-z) and substitute in those values instead as a scaled LOD proxy.

The green curve goes the whole hog. Since the SSB-z data can also be used as a proxy for sunspot numbers (on a different smoothing and lag value to the LOD proxy), it is used both for sunspot proxy and LOD proxy. This enables me to reconstruct past and predict future planetary surface temperatures, to a limited degree of accuracy.

There are a couple of obvious problems. The method does not capture individual El Niño events well. Nor does it predict individual big volcanos, although the volcanic explosivity index does correlate well with the motion of the planets, as I will show in a future post. One further problem is that the technique does not capture the collapse in solar activity which seems to occur when Uranus and Neptune are in conjunction, as at 1800-1840 during the Dalton Minimum, and during the Maunder minimum in the 1630′s . Whether we will see a similar deep solar minimum now following the conjunction of these two planets in 1993 remains to be seen.

The large departure of my reconstruction from the SST data around the WWII years is I believe due to well known issues with the switchover from bucket and thermometer measurements to ship engine cooling intake sensors on military vessels.

So, the basic premise of my model, is that a cumulative count of sunspots above and below the ocean equilibrium value I have determined will mimic the retention and release of energy from the ocean. At the same time, multi-decadal changes in Earth’s length of day which also correlate with the timings and sign of the major oceanic periodicities (PDO, AMO) add detail to the picture.

The high SSN of the late C20th means according to my model, that a lot of heat got absorbed into the ocean. Now the sunspot numbers are falling, that heat is being released again by El Niño’s and the temperature is dropping because that heat is escaping to space and not being replaced by solar energy into the oceans at the rate it was in the ’80′s and ’90′s. I have done calcs on this to support my theory and I will present them soon.

Comments please.

  1. Dave E says:

    You have my email Bob.

    I can put you in touch with someone who has used Leifs TSI reconstruction in a simulated electronic/electrical model & got a graph similar to the stated temp increases.


  2. Dave E says:

    Oops. Don’t know why I put Bob there. Senior moment perhaps.


  3. tallbloke says:

    Thanks Dave, I’ll drop you a mail later.



  4. Paul Vaughan says:

    I recommend seriously rethinking this:

    “The large departure of my reconstruction from the SST data around the WWII years is I believe due to well known issues with the switchover from bucket and thermometer measurements to ship engine cooling intake sensors on military vessels.”

    Look at NPI, ALPI, PDO, EOP, & the works of Zolotova & Ponyavin.

  5. tallbloke says:

    Hi Paul,
    I’ll take a look and a think, but I see Jeff Id’s reworking of the land ocean data via the MET supplied perl scripts is looking flatter around then too.

  6. Paul Vaughan says:

    I will be very suspicious of any “reworking” of temperature series that “flattens” ’40s oscillations. I encourage you to consider that you may be overlooking informative variables (possibly because you are [at least in part] limiting focus to a z-axis in some of your analyses…)

  7. tallbloke says:

    I used the HADsst2GL product for this plot, but in one of the emails leaked out of the CRU, Phil Jones is quite blatant about the 40′s SST’s not being right.

    I do think there was some considerable el nino activity which my method failed to capture, but unless you can actually offer something for me to look at, I can’t go on a variable treasure hunt right now.

  8. Paul Vaughan says:

    I’m in no way convinced that Jones is sufficiently aware of patterns in ’40s EOP.
    I’d like to convey my insights to you, but I can think of no way to do so succinctly at this stage. I’ll drop notes here & there when I can. It seems you agree that something we might tentatively call an ‘unusual’ oscillation/redistribution caused some kind of a ‘hiccup’ (between basins, if my sense is right, based on my analyses) between the late ’30s & ’50 — I’d be content with such a tentative mutual understanding.
    I’d provide Barkin links/notes, but since I found *so many articles, I opted to speed-read-for-gist and omit note-taking (which I normally do). I don’t regret the choice because I’m buried in competing pursuits.
    I do hope someone else will take up the matter of summarizing Barkin, because I have wavelet software to write & wavelet analyses to run (and I’m not convinced that there are a lot of folks in the community equipped to productively work that front, which I consider to be of paramount importance at this stage, since it yields such precise insights into co-phasing [as opposed to co-amplitude] patterns). Maybe someone will volunteer — if so, I’m prepared to provide feedback (as time permits).
    Best Regards to All.

  9. tallbloke says:

    As you can see from this graph of mass distibution vs solar activity
    there is a rapid swing from ~1910 1950, which nicely brackets the mid thirties ‘shift’ we have both noticed from our independent analyses. You may well be right that some of the temperature oscillation of the 40′s is linked to ‘basin shift’ due to changes in Earth Orientation Parameters, somewhat confounded by a shift of which boats were in which basins taking measurements during those years too.

    Each to their own strengths, and I do appreciate the quality of the stats analysis you have undertaken. I promised to read some Barkin, and haven’t had time. He is still on my list though, through your recommendation. I’m happy for this blog to run at a measured pace, it’s not a gish gallop to get bums on seats. I’m very pleased to have had some great posts from people actually working on this stuff, and I hope it continues at whatever pace suits the contributors.

  10. Paul Vaughan says:

    Note the low ~1900 in r”:…png
    Note aberrations in the phase (red) that show a flip in phase-relation with polar motion group-waves (blue):
    (Bear in mind the issues with the early polar motion record – details another day.)
    Look at the geomagnetic pattern culminating in the ’40s:
    Compare with:,CanadaPrecipitationTimePlot.PNG
    Keep in mind that regional precipitation is spatially correlated over very wide regions.
    See what happened either side here:
    (pattern relates to Australian drought too)
    The cumulative sums of NAO & SOI reveal *unique phase-relations of north-south & east-west oscillations in the late ’30s to early ’40s. And remember the ~’20-’40 Arctic warming & Southern Ocean cooling.
    There’s a lot going on there. I have some ideas….

  11. tallbloke says:

    Forgive my Acronym failure, what is PPT?

  12. Paul Vaughan says:

    PPT is short for precipitation in some research fields.

    There are plenty of other ’20-’40 ‘oddities’ – fundamental constraints flipped on (or off) & then off (or on) during that interval — I’m in no way convinced the ‘experts’ have learned all there is from the records we have (despite all the whining from some that we don’t have enough data). I outline a short selection of the patterns, providing links to journal articles, under the main article here:
    (…but that’s by no means exhaustive).

  13. Paul Vaughan says:

    …btw: I’ve found an article that suggests I was right to suspect nutation to be a factor related to the late ’80s / ’90 reversal here:

    Nutation longitude appears to have passed the main north-south axis of NAO around that time. It appears that it should be possible to greatly improve on the work that is being done on links between NAO & polar motion. There is clearly a need for interdisciplinary collaboration. The Earth orientation experts know a lot of physics & differential equations etc., but they appear to have weaknesses analyzing time series to learn more from data. This is actually quite a serious impediment to efficient progress.

    The pattern to which I’ve linked here can probably be cleaned up substantially – that was a quick ‘back of the envelope’ calculation. I know of at least 5 really simple parameters that I didn’t even bother to optimize, so that already-robust relationship is screaming out something that has somehow gotten overlooked by folks who I would expect should notice it.

  14. tallbloke says:

    Hi Paul, you seem to have linked one of your intriguing graphs rather than the article, but there’s plenty to talk about in that anyway. I also found LOD needed to be shifted around four years to get my ‘best fit by eyeball’ (lol, I’m so amatuerish).

    Your combination of sun to solar-barycentre motion acceleration and the Lunar nodal cycle with polar motion velocity(?) is very interesting. Please tell us a bit more about that.

  15. Paul Vaughan says:
    A complex wavelet lets you extract phase. You get x & y wavelet-transforms from which you can construct radius & angle (using a transformation from Cartesian to polar coordinates).
    The red curve in this graph is a dead-simple combination of phases:
    The curve is isolated by time-integration of the phase-combination over dominant harmonics.
    Note that it is not LOD in the graph, but rather the rate of change of LOD (which leads LOD by a quarter-cycle).
    The lag suggests there is something else to learn about the JN/4 timescale (which is consistent with my findings from time-integrated cross-correlation analyses).
    It is the magnitude of the year-over-year differences in aa index that underpin the blue curve — time-integration over dominant harmonics isolates the decadal pattern.
    Correction: Above I should have said there are at least 8 simple parameters (not to be confused with variables) that can be optimized to fine-tune the already-robust fit.

    Interpretation: Terrestrial oscillations are in resonance with processes at JN, SN, & UN timescales.

    The QBO signal fits this picture like a glove.

  16. Paul Vaughan says:

    correction: black curve – not red curve (red is -LOD’)

  17. tallbloke says:

    “Terrestrial oscillations are in resonance with processes at JN, SN, & UN timescales.”

    For the uninitiated, these initials refer to planetary pairs and the resonances they create via their orbital periods. Jupiter-Neptune, Saturn-Neptune, and Uranus-Neptune.

    This is fantastic work you have been doing Paul, and I look forward to your summary so I can give it the cohesion it deserves on a separate thread.

  18. Paul Vaughan says:

    FYI, I don’t plan to formalize communications until there is funding.

    Reply: Understandable. Since you consider yourself an unpaid researcher, I’ll consider myself your unpaid assistant and I’ll just do the best I can with what we’ve got. I bet you won’t be able to resist correcting me where I get it wrong though. You never know, with the huge exposure you get here, it may attract the funding you need. ;)

  19. Paul Vaughan says:

    I’ll make one last note here for anyone exploring the calculations:
    It’s the decadal (Schwabe cycle timescale) rate of change of this curve that you should be looking at:
    Don’t be distracted by the spiky features at interannual timescales when working on the decadal problem – (they can be cleaned up later using nonstationary-period variable-bandwidth moving time-integration). Just difference the curve & repeat-smooth with a simple boxcar-kernel at 11.1-year-bandwidth to achieve a 1/4 cycle shift (corresponding with rate of change of the magnitude of year-over-year differences averaged across Hale cycles) during preliminary analyses to get the gist of decadal dynamics.

  20. Paul Vaughan says:

    tallbloke “You never know, with the huge exposure you get here, it may attract the funding you need.”

    I drop notes online in case I don’t get funding. I should clarify that while I am not interested in having a post devoted to my work at this time, I don’t regret the effort I’ve invested in dropping some notes in threads here —- your “mixing bowl” is valuable tallbloke, particularly since no time is wasted on inefficient exchanges the way you are moderating comments here. All the best.

  21. Paul Vaughan says:

    Forgot to point out that the split in nutation obliquity…
    … matches PDO:

    AMO pierces the split:
    …so maybe there was a ‘spill’ between the basins via the Arctic (via ocean or atmosphere-teleconnection or whatever – due to whatever circulation pattern disruption or whatever).

    Note also …
    …which show patterns related to the polar motion phase-relations-shift from cophasing related to r” (or something loosely phase-confounded with it) to cophasing related to plain JN, which ‘snapped into gear’ (very firmly) coincident with the regime-break pointed out by Zolotova & Ponyavin – links to their articles here:
    It’s complicated enough that there’s *plenty* of work here (due to the complexity) to go around for *lots* of researchers. Cheers.

  22. P.G. Sharrow says:

    Examined the above graph, very interesting. the nominclature is not very clear. I hope the yearly is all correct. You seem to show, from 1910 to 2010, a fairly consistant increase in energy input into the oceans.
    The solar output is now the lowest measured and the atmosphere is cooling and shrinking/ losing energy.

    This could mean we now have warm/energized ocean under a cool/low energy atmosphere. A serious change in climate conditions. Perhaps wetter weather, polar weather at lower latitudes due to weaker circumpolar winds. I think whiteman will need larger wood pile.:-(

  23. tallbloke says:

    Hi P.G.
    Good summary. If my theory is anywhere near the mark, the ocean retains energy to a much greater extent than currently supposed by the mainstream. Because they can’t believe their own data, they’ve been assuming that some of the ARGO buoys are faulty. No doubt some are, but I think they have probably overcompensated because an ocean losing overall ebergy since 2003 doesn’t fit with their co2 driven climate theory.

    My understanding from a simple logic point of view, is that the ocean must have shifted energy downwards during the run of high amplitude-short minimum cycles in the late C20th. Now the sun is quiet, some of that energy is rising to the surface again and coming out of the ocean all over the globe.

    Skeptics are resistant to this idea of stored heat ‘in the pipeline’, but what they need to realise is that it was put there by the sun, not co2! My calculations show that the oceans absorbed energy at an excess rate of around 4W/m^2 between 1993-2003. This is far higher than the alleged co2 forcing, and can only be due to less cloud and higher solar activity.

    But the air as you note, is cooler and less humid, so the heat coming out is not being trapped in by water vapour over the east Pacific like it was in the ’98 el nino, so much of the heat energy is escaping straight to space. This means a loss of energy from the overall Earth climate system and so, colder continental winters, drier, clearer summers, and warm sea surface temperatures until the modoki el nino subsides later this year.

    I must get around to writing a separate post on my ocean hypothesis.

  24. P.G. Sharrow says:

    A cooling atmosphere over a warm ocean will be much more cloudy and tend to hold the warmth near the surface. Situation we have now in warm ocean air temperature.

    One thing I see is that very few people can see beyond a thermometer as an indicator of energy content. The energy storage in the ocean is also in chemical exchanges and the biosphere. The ocean is not just water, it is also a soup of various activities. A giant energy battery.
    Sorry to make your devinations more complex.

  25. tallbloke says:

    Quite right. This is partly why I started looking at other indices of energy input to the Earth climate system. Oceanic biota don’t affect sunspot numbers or planetary orbits.

    I will take note of your comment when I write my ocean heat content post though. I re-read James Lovelock’s original ‘Gaia: A new look at life on Earth’ book recently, and he has some great insights into chemical and biological processes in the oceans. Well worth a read.

  26. P.G. Sharrow says:

    It certinly is possiable that changes in sunspot activity does not cause weather changes on Earth. It may well be that the thing that causes changes in sunspot activity also effects Earth weather systems.
    One could posit that sun spots and earth cyclones are thermo-dyanamicly similar. Thunderstorms create near space EMF sprites and flares that look very much like sunspot activity. Just add super heated plasma and multiply by about 20,000 times. :-q
    High sunspot activity, at the same time high hurricane – cyclone activity. Smaller cyclones at mid-latitudes, smaller sunspots at mid-latitudes. Speckles on the sun, thunderstorms on the earth.
    Sun and earth connected but not directly.

  27. tallbloke says:

    Hi P.G.
    High sunspot numbers are the symptom of high solar activity. Would you agree that faster solar wind and the buffeting rapid fluctuations in it bring higher storm activity to the Earth?

  28. P.G. Sharrow says:

    When you measure the radiant temperature of a simmering pot it’s 212f and if it is boiling briskly 212F. The rapidly boiling pot moves a great deal more energy to it’s surroundings. A faster fluid flow will transfer greater energy amounts then a slower one. Plain old applied science.

    Higher sunspot activity is indicative of higher EMF activity and therefor greater energy transfer due to better “Q” coupling of the sun earth fields. Electronic engineering.

    Storms are caused by thermal energy transfer from warm damp atmosphere to cool dry atmosphere, and winds from high pressure to low pressure areas.

    Energy moving from high to low potential.

    It has often seemed to me that some spots on the earth spawn pressure cells that are stuck in place for some time before moving off on the trade winds.
    At times an area that has spawned high pressure cells, spawns low ones, and the jet winds loop around them like water moving around obstructions in a stream. Purhaps gravity or magnetic pumping. Astromachanics ?

    The short answer is yes.
    I just created these word pictures to help think out the answer.

  29. tallbloke says:

    I like it, keep it coming.

  30. Stephen Wilde says:

    Hi tallbloke,

    Can your LOD influence also accommodate a 1000 year cycle such as would include MWP, LIA and Modern Maximum ?


  31. tallbloke says:

    Hi Stephen! Thanks for dropping by.
    The LOD data doesn’t go back far enough, but the graph of the solar system angular momentum cycle produced by Semi certainly shows bumps in all the right places.

    You can download the full pdf here:

    The thread is here:

  32. Ulric Lyons says:

    P.G. Sharrow

    “It may well be that the thing that causes changes in sunspot activity also effects Earth weather systems.”

    I would second that point of view. The well predictable solar triggers for cyclones, or day to day weather events, are not allways directly associated with sunspot activity.

  33. tallbloke says:

    I agree too. It could be that sunspot numbers are just a useful proxy for other electromagnetic phenomena which have terrestrial effects.

  34. P. Solar says:

    I would not be at all surprised to find there is signal from the position of the planets in the SSN (likely to be reflected in Earth climate). Also directly seismic events and volcano activity (and hence climate).

    However, I think Semi is going off on a limb. It’s well known that with enough terms you approximate any series with anything. What he’s doing with is a kind of ad hoc Fourrier series. When you start introducing the smoothed, delayed cosine of the semi angle between Venus and Mars for no objective reason , you’ve lost the plot. Sorry, hardly Kepler.

    Your LOD + SSN looks quite convincing. Have you posted details anywhere yet?

    Try plotting against ICOADS with -0.24 correction for ((date>1941.71) && (date<1946.12))

    may fit better that frigged hadSST data.

  35. tallbloke says:

    Hi P. Solar and thanks for the input. A copy of your monthly ICOADS series with the 0.24C drop through the war years would be most welcome. I’ll plug it into the model and see what we get.

    The reason I haven’t published details yet is I’m still working on trying to find the algorithm which will work to match and predict the collapses in solar activity which seem to happen every ~180 years +/- 30. Once that is in place in the model I think I’ll be able to generate curves which match cosmogenic isotope records better and provide a more convincing result people will take more notice of.

    The inner planetary angles seem to relate more to individual solar seismic events such as big flares. Ching Cheh Hung successfully predicted those using a system which included adding a ~30 degree angle to the planetary alignments with the sun as one of the trigger points.

    I’ve also discovered a better relationship between JEV and solar activity levels by considering the alignments along the curve of the Parker Spiral rather than in straight lines.
    This finding seems to add weight to the hypothesis that we are looking at an electromagnetic underlying mechanism rather than gravitational, though both are operative.

  36. P. Solar says:

    ICOADS monthly can be downloaded here:

    I just used gnuplot to make the 0.24 adjustment directly when plotting. The formula is trivial enough that you should be able to do , even by hand, if needs be.

    plot s=0 \
    , icoads_m  using 1:( (($1>1941.71) && ($1<1946.12)) ? ($2) : NaN ) tit "icoads WWII un-adj "  w l linec rgb "green"\
    , icoads_m  using 1:( (($1>1941.71) && ($1<1946.12)) ? ($2-0.24) : $2 ) tit "icoads WWII-adj -0.24C "  w l linec rgb "red"\
    , SST_m  using 1:($2) tit "hadSST2 "  w l linec rgb "blue"\

    If you have trouble doing that, email me and I’ll send the data.

  37. olsonjs444 says:

    Thanks for your efforts on this topic, Tallbloke. Make sure you read Richard C. Willson’s work (ACRIM Science team at NASA) available at the following links:
    (slides 11, 12 … pages 10, 11 in the .pdf… are very appropriate to this discussion; watch out for slide 23… page 22 — it is very misleading, no doubt due to Hansen’s influence, because no one knows what TSI was in the late 19th or early 20th centuries)
    (skip to page 4 of this .pdf — the figures shown there should end all debate regarding the validity of AGW theory — or rather, the lack thereof, and any alleged warming from CO2)