What caused global warming in the late C20th?

Posted: June 2, 2011 by tallbloke in climate, Ocean dynamics, Solar physics, solar system dynamics

Southern Hemisphere Sea Surface Temperature Anomaly

We can see that the decadal regression from the trough of the La Nina following the 1998 super El Nino to the peak of the 2010 El nino is negative in southern hemisphere sea surface temperature. Eyballing the data, you would have thought that it would be a positive warming slope. It isn’t. I’m sure you’ll agree that this tells us something about the underlying trend since the start of the C21st.

This is in contrast to the effect of the El Nino Southern Oscillation (ENSO) during the warming period from the mid ’70’s to the turn of the millenium, when El Nino was stronger than La Nina, and the released oceanic heat was resulting in upward step changes in global temperature, identified by Bob Tisdale.[1]

I believe there is an underlying reason for that, and it has less to do with co2 than with the solar contribution to climate change. In a nutshell (heh!), here’s my hypothesis. I hope you find the time to plough through it:

1) Downwelling longwave radiation from greenhouse gases (mostly water vapour, plus co2) can’t penetrate the ocean surface beyond it’s own wavelength. This is well known physics. But the assumption has been that the warming of the surface by this longwave ‘back radiation’ is ‘mixed down’ into the ocean. This is incorrect. When pushed, the physics people supporting the idea of co2 driven warming say that rather then ‘back radiation’ warming the ocean, the effect of additional co2 is to raise the altitude at which the bulk of radiation to space takes place at. This will have the effect of ‘thickening the blanket’ and hence causing the ocean to cool at a slower rate than the rate energy from the sun enters it, causing warming. However, I’ve done the calcs, and the extra height comes to about 150-200m. This isn’t going to make a big enough difference to account for the amount that the ocean warmed from 1980 to 2003.

2) According to data gathered from weather satellites, and measurements of how much of the sunlight hitting Earth gets reflected onto the moon, cloud cover, particularly in the tropics, reduced from 1980 to 1998, and then started increasing again. This allowed more sunlight to get to the ocean (and land) surfaces. Unlike ‘back radiation’ from greenhouse gases, the radiative energies in sunshine penetrate deep into the ocean (up to 150m) and transfer their energy into the seawater. At those depths, wave action, internal currents and tidal flows mix the energy well down into the briny, creating a relatively linear dropoff in temperature from the near surface waters down to the thermocline, between 35m in the tropics, down to 1000m in the higher latitudes. it is this increase in ocean temperature which has caused around half the sea level rise since 1980, due to thermal expansion. The surface temperature of the ocean (70% of Earth’s surface) has risen in proportion to the increased ocean heat content (OHC) in the upper 700m, maybe deeper.

3) The surface of the ocean is on average several degrees warmer than the near surface atmosphere. On average, a net amount of about 65 Watts of energy per square meter is being transferred from the ocean to the atmosphere. The sea surface temperature (SST) thus drives the atmospheric temperature, notwithstanding the relatively small amount of solar radiation absorbed directly by the water vapour and co2. The dominance of the ocean is easily demonstrated by comparing time series of SST and near surface air temperature. The changes in air temperature lag behind changes in SST by several months globally. The ocean contains as much energy in the top two meters as the entire atmosphere above it. This is because the specific heat capacity of water is much, MUCH higher than air. The ocean drives the atmosphere, the tail does not wag the dog.

4) The big uncertainty in climate science is the overall feedback from clouds. High cloud holds heat in at night and makes things warmer. Low cloud blocks sunlight out and makes things colder during the day. Proponents of the anthropogenic global warming hypothesis (AGW) are convinced the overall cloud feedback is positive, and that warmer temperatures mean more evaporation and clouds, trapping heat in – a positive feedback. This is the only way they can get the co2 driven model to work. Co2 on its own isn’t enough, there has to be a positive feedback from water vapour and clouds. Empirical work done by renegade climate scientists such as Dr Roy Spencer[2] and Professor Dick Lindzen empirically find the opposite from satellite data – that cloud feedback is negative, and predominantly cools the Earth down by reflecting sunlight out. Logically, if cloud and water vapour feedback was strongly positive, the Earth’s climate system would be inherently unstable, and the oceans would have boiled 550 million years ago when co2 levels were twenty times higher than now.

5) The AGW proponents say the Sun can’t be responsible for late C20th global warming because the amplitude of the solar cycles has been diminishing since the late 1950’s. There are several problems with this view. Firstly, although the peaks of the solar cycles have been getting lower, the cycles were shorter than average (around 10 years rather than the 11 year average) and the upramps and downramps were steep, and the minima between them brief. This means the average sunspot number over the period about 70 was well above the longterm average of around 40. What I have discovered, is that by making empirical comparisons between SST and the sunspot number (SSN), we find that there is a consistent relationship between the sunspot number and ocean surface temperature when the effects of the oceanic oscillations are accounted for. This means two things: Firstly, there is a relationship between solar activity levels and cloud cover, since small changes in cloud cover make a much bigger difference to the amount of sunlight hitting the ocean than solar variation does. Secondly, there must be a level of solar activity, as indicated by the sunspot number, at which the ocean neither cools nor warms. I have empirically determined this to be around the same value as the long term average sunspot number, about 40SSN. Another problem with the ‘solar cycles diminishing since the ’50’s argument is that Dr Leif Svalgaard (NASA)[3] has used solar magnetic records (derived from geomagnetic records) to determine that Waldmeier, who was in charge of counting the sunspots from 1945 until the mid ’80’s was overcounting by around 20%. Correcting this flattens the previously apparent drop in solar activity a lot.

6) Recently, some experiments have been done at Aarhus in Denmark[4], and at CERN in Switzerland, which support Henrik Svensmark’s hypothesis that solar activity levels affects the number of cosmic rays getting into Earth’s lower atmosphere and seeding clouds. More active sun, stronger solar wind, less cosmic rays reaching Earth. I have discovered that there is a close correlation between solar activity levels and the specific humidity level up near the tropopause. This will have a bigger effect on the altitude at which radiation of heat to space from Earth occurs at than changes of the atmospheric concentration of co2 from 0.027% to 0.039% will. High altitude specific humidity has been falling since the late 50’s in proportion to the drop in the peak amplitude of the solar cycles. Hungarian physicist Ferenc Miscolzci[5] has determined that the overall optical depth of the atmosphere has stayed pretty much constant, and so the reduction in specific humidity at higher altitude has been matched by an increase at lower altitudes – more lower cloud now the sun has gone quiet. All this means that the solar signal is amplified by terrestrial mechanisms, and indeed this has been shown empirically by Professor Nir Shaviv[6], who used the oceans as a calorimeter to determine the size of this amplification. However, this amplification is masked in the surface temperature record because of the phasing of ENSO, El Nino’s build during solar cycles and release their energy into the atmosphere when the sun is at minimum, La Nina’s often occur near solar maximum. There has been a big El Nino following solar minimum for the last 5 cycles at least. This anti=phase masking has helped lead to the facile (and incorrect) conclusion that the Sun’s variation doesn’t affect climate much.

7) All the foregoing leads me to two principal conclusions:

c1) Late C20th Global warming was caused predominantly by the Sun, not human emitted co2.

c2) It’s going to get colder, since the sun entered a long minimum in 2005 and shows no signs of getting lively again yet. However, the oceans will continue to kick out heat into the atmosphere for quite a while yet, because a lot of additional heat was stored in them while the sun was more active than average in the latter half of the C20th. This will happen in the form of El Nino’s such as the one we had last year, but rather than the El Nino’s causing an upward step change in global temperature which remains after the El Nino’s have finished, as happened in the 1975-2003 period while the sun was more active than average, they will be followed by La Nina’s which take the global temperature lower, as ocean heat content diminishes.

To test my hypothesis, I’ll make the following prediction:

The current small recovery in temperature following La Nina will be short lived, and global surface temperature will fall again, dropping to below January 2008 levels sometime in the next 6-10 months. This is risky, because ocean dynamics are poorly understood, and no-one knows how much or how fast the accumulated heat in them will be released. So, we will see. If the prediction fails, it isn’t a fatal blow to my hypothesis, but a rethink on temperature stratification in the deep will be in order. Another risk is that a couple of big volcanic eruptions going stratospheric might affect short term temperatures too. Volcanos and Earthquakes are more frequent at times of solar minimum, as we’ll see in another post.







  1. tchannon says:

    Twiddle, does this add some context?

    Whole HADSST timespan with filtered
    (for full size, right click and view locally)

    Usual cautions.
    Do I trust SST? No! There are serious data source problems common with land data.

  2. Malaga View says:

    Great posting….
    So basically: Climate and Volcanism are controlled by the SUN… give or take the odd Lunar / Planetary assist.

    The ocean contains as much energy in the top two meters as the entire atmosphere above it…..El Nino’s build during solar cycles and release their energy into the atmosphere when the sun is at minimum, La Nina’s often occur near solar maximum.

    So the oceans are THE heat pump moderating climate.
    Storing heat and releasing CO2 at solar input maxima.
    Releasing heat and storing CO2 at solar input minima.
    So we have a mix of daily, annual and solar cycles driving Weather / Climate.
    A huge surprise [not] for those who learnt about the Gulf Stream / Atlantic Drift.
    So after the AGW logic excursions we are back to good old fashioned Science.

    Phew… that’s a relief!

    Now why don’t I believe the official records – especially CO2!

  3. tallbloke says:

    Tim, that’s a beautifully produced graph. Yes, it adds a lot of context! Right smack on the peak of the curve.

  4. tallbloke says:

    MV: Yes. What I forgot to say was that as El Nino builds, the warm water which drives it is below the surface, hidden from the SST record. ARGO should have revealed this. When they finally release the data, I’ll be making some close studies of the temperature/depth profiles in the Pacific.

  5. Roger Andrews says:


    “It’s going to get colder”. Well, according to my recent predictions the oceans are indeed going to get colder (https://tallbloke.files.wordpress.com/2011/05/ra-fig6.png?w=600&h=434) but the air will continue to warm, although not that much. (https://tallbloke.files.wordpress.com/2011/05/ra-fig7.png?w=600&h=434)

    The hole in these predictions, as we discussed, was that I wasn’t able to model sea-air transfer. However, relative to your comments on El Niños I’m e-mailing you some cartoons I put together a few months ago showing how sea, surface air and troposphere temperatures changed in the central Pacific during the 1997-1999 El Niño/La Niña events. I don’t know exactly what they mean, but they’re interesting to look at.

  6. Chuckles says:

    ‘What caused Global Warming in the late c20th?’

    Surely the answer has to be – ‘The existence of the IPCC and UNFCCC, and the availability of entirely too much grant money?’

  7. tallbloke says:

    Roger, thanks, I’ll take a look at those.

    One thing to consider: If air temp went up relative to SST, then the oceans would warm. Can’t see it myself, so we really are going to have to update your post with a coupled model at some point. The ocean drives air temp. If the air gets warmer but the ocean isn’t backing it up, then the extra warmth will simply dissipate to space.

  8. Roger Andrews says:

    Tim C:

    “Do I trust SST? No! There are serious data source problems ….”

    The main problem with the SST record isn’t data quality, it’s the “bias corrections” that get applied to it, which in the case of HadSST2 are designed purely to make the SST record fit the surface air temperature record as closely as possible even though there’s no basis for assuming that SSTs should track SATs or even that the SST record is seriously distorted by instrumental biases. However, after the bias “corrections” are applied the record does get seriously distorted. The abrupt 0.4C drop in your graph in 1946 is, for example, an uncorrected bias effect (see http://www.atmos.colostate.edu/ao/ThompsonPapers/Thompson_etal_Nature2008.pdf)

    But how do I know that the uncorrected (ICOADS) SST record is OK? Well, I don’t, but it does match TSI and other solar series quite well, which we would expect given that the sun heats the oceans, and it shows a well-defined cyclic relationship with SAT that also appears to be solar-generated. It also correlates remarkably closely with a relative sea level rise series I recently put together from tide gauge records (maybe worth a thread, TB). I don’t think we would see correlations like this if the ICOADS SST record was complete garbage.

  9. Tim Channon says:

    Does this relate to solar?

    Yes a little if a not infrequent comment by solar folks is correct, every other cycle which suggests magnetics. Also known as the even and odd cycle problem.

    19, 21, 23 seem to matter

    18, 20, 22

    Crazy observation which might make little sense to most readers, could the 1951 and 1990 solar barycentre events be switch points: note that both are during even numbered cycles. Taken further, could these be ramp break points?

  10. lgl says:

    c1) Late C20th Global warming was caused predominantly by the Sun, not human emitted co2.

  11. Stephen Wilde says:

    I agree with the general thrust of the above but there are a couple of areas where the detail doesn’t match my understanding. They are not critical to the overall hypothesis though.

    As Rog knows I am thinking along the same lines but am trying to create a synthesis that combines all the relevant observations with a physically plausible overview.

    Here is detail on the solar induced top down mechanisms:


    which fits in with Rog’s general thesis quite nicely.

    As regards the Svensmark idea I think that the shifting surface air pressure distribution is a more likely cause of changes in cloudiness and albedo than cosmic ray quantities but we shall see.

    I will shortly be publishing a detailed account of the bottom up oceanic part of the same process.

    At that point I should have a reasonably complete climate description for comparison against future climate events.

  12. tallbloke says:

    Tim, hang fire for my next sunspot asymmetry update. I found ‘switch points’ in the data (which depend on smoothing choices). I’d welcome discussion with you on how best to work with the data.

    lgl, thanks for stopping by. No doubt Dr Svalgaard would have a thing or two to say about the solar reconstructions both Mann and Stienhilber used. 🙂
    Are they both using Lean 2000 or something else?

    Stephen, yes, I can see our work integrating well. I agree we might not need Svensmark. My discovery of solar correlation with specific humidity could mesh with your surface pressure/low cloud ideas. Let’s take that forward over a beer sometime this year.

  13. Stephen Wilde says:

    “Let’s take that forward over a beer sometime this year.”

    That would be nice. Send me an email to make the arrangements. Can you get it from your site data?

    When my oceanic ‘bottom up’ article is published perhaps you could use it here as a topic ?

  14. lgl says:


    The Mann graph is his EIV temperature reconstruction. I don’t think he used any solar at all.
    Steinhilber is mostly based on 10Be.

  15. tallbloke says:

    lgl: Thanks. 10Be has problems according to both Vukcevic and Svalgaard. I’m inclined to rank it in there with other proxies but don’t set too much store by it. What interests me is the extent to which glacial advance and retreat occurs on the basis of quite small temperature changes. Though those changes might be larger in localised areas, so we need to be cautious of both extrapolating local conditions from global eastimates, as well of being careful about extrapolating global estimates from local proxies.

  16. tallbloke says:

    Stephen, yes, email coming your way. And yes, I’d be very happy to republish your paper as a guest posting, thanks for the offer.

  17. lgl says:

    I haven’t seen any problem with 10Be on this time scale, and even Leif is happy with Steinhilber, showing only 0.9 W since Maunder.

  18. tallbloke says:

    lgl: Fair enough. I’m more than happy with 0.9W, given terrestrial amplification. 😉

    By the way, that was a cool 3 line argument you presented Leif with concerning gravity. I’m still pondering that one.

  19. Roy Weiler says:

    I am not taking issue with your presentation, but I do have a question. Anthony Watts recently introduced his ENSO page. While perusing it I noticed, for the first time, the steady increase in ENSO 4 region.

    Nino 4

    I suspect it is supportive of your thesis, but I am not quite sure how.

    Can you comment?

  20. tallbloke says:

    Hi Roy and welcome to the talkshop.
    Yes, you can see the Nino 4 region ‘went positive’ around the time of the great climate shift in the mid ’70’s. El Nino started to dominate then. I think we’re now just going over into a shift towards La Nina dominance. The ocean ‘cycles’ of around 60 years apparent during the late C19th and C20th seem to be more or less in tune with solar activity rising and falling over the decades. As the sun puts heat into the oceans which builds up when the sun is more active than average, so the ocean finds a way of getting rid of some of it when the sun is temporarily less active between solar cycles. The eruptions of energy which come out in El Nino events are huge energy releases, capable of warming the atmosphere worldwide. Roger Andrews sent me some graphics illustrating this earlier today. I’ll make a new article with them soon.

  21. Roy Weiler says:

    So you are saying, the continued increase in the ENSO 4 region is the delayed response factor you imply in your thesis? And further, this is what explains why the last El Nino came in much lower the 1998?
    In essence, this represents the stored heat in the system, and will be ‘used up’ during the next cycle of stronger La Ninas?


  22. gnomish says:

    nice work.

  23. tallbloke says:


    I don’t think your graph (which I’ve now included inline in your previous comment) shows a “contunued increase”. Your eye is maybe decieved by the high amplitude of the 2010 El Nino, but look at the strengthening negative periods below the median line.

    And that El Nino was not “much lower” than the ’98 event, but around the same. I think we reached the top of the curve in 2005 and the two big El Nino’s are almost a mirror image of each other. The difference is that the ’98 El nino occured just after the minimum at the start of a strong solar cycle, which kept the ocean heat content on the rise, despite the enormous release of energy into the atmosphere, whereas the 2010 event happened in a big solar minimum from which the sun hasn’t really recovered. So the ocean heat content is falling. There is no extra solar heat going into the ocean, so El Nino depletes the store of energy.

  24. tallbloke says:

    It’s a stunning day here. I’m going to ride up to the mountains and stay out overnight.

  25. lgl says:

    A bit OT but since you started it …
    Leif has a problem with that gravity argument because years ago, when discussing the spin-orbit coupling, he agreed if putting a huge rocket on the Moon, accelerating it, the rotation of the Earth would change. So why wouldn’t the same mechanism apply when Ju is accelerating the inner planets.

  26. Michele says:


    you read my e-mail..

  27. Roy Weiler says:

    “Your eye is maybe decieved by the high amplitude of the 2010 El Nino, but look at the strengthening negative periods below the median line.”
    I suspect you are right on this.

    The ARGO network seems to support the idea of no longer increasing heat content in the oceans.

    I will watch your work with interest. I suspect climate scientists have jumped ahead of what is actually known at this point. Another 30 years of observations may begin to help in understanding how the climate system works.

  28. Doug Proctor says:

    Taken in isolation, tchannon’s graph of SH SSTs looks like a 120 year sine function, in which we are at the 60-year mindpoint, with start rise in 1948 and high point in 2007/2008. Dropping now ….

    Now that is taken in isolation from the longer term records and, though a valid impression, has little meaning in itself. Though, that said, is the warmist camp not basing all of its catastrophic predictions on the post-1965 correlation of the global/other temps with their modelled predictions?

    It seems as though the warmists are using the 1850 to 2011 history as justification for the 2011 – 2100 period, but the signal was not supposed to be “visible” until about 1980, and the past is not considered, in the IPCC theory, to be a predictor of the future. I just saw some CO2 records that showed a larger than expected surge of CO2 production now that the recession is over in the lesser developed world. And that the CO2 increases as per 2010 data push the IPCC into the 4C by 2100 range. This is a fabulous development for those who disagree with the IPCC mantra. The further down the IPCC rabbit hole we are being led, the further short-term observation is going to differ from prediction.

    It is funny, though, in an odd way, that so many different and reasonable-looking patterns can be drawn from the same data. Makes you wonder how much filtering of acceptable patterns occurs prior to the search for warming patterns, and how valid a lot of statistical analyses are. Other than confirming what you’ve already determined by eyeball and belief. The other, less controversial way of thinking about it, is to ask how robust the data is if patterns within it can be found to serve such different masters. How much is data, then, and how much is noise? And are there patterns to the noise?

  29. lateintheday says:

    Very nice post. I can actually get my head around this one which makes a nice change.
    A couple of things I’d like to throw in both as questions and also (if answered in the positive) as additions to your ideas.
    First, some months ago Hansen stated that we should be in a cooling phase according to Milankovic theory. This was rebutted by a comment (at WUWT) stating that the Earth-Sun distance was far less significant than the angle/tilt. In which case, the Northern hemisphere is actually receiving sunlight more directly at present and hence we are in a warming phase hemispherically speaking. I’ve no idea who is right on this.
    Second, which may have been part of the same comment, it was stated that the NH summer is now around 7 or 8 days longer than the SH. They also suggested that NH summer is now significantly longer than 10k years ago. Again – I’ve no idea.
    Finally, you don’t specifically mention the net effect from polar warming. From bits that I’ve read, it seems entirely possible that the heat lost to space in the arctic area (through lower sea ice extents) outweighs the energy gain from loss of albedo. I’d liken it to a very slow response automatic window. Too slow to keep temps perfectly steady, but in the long term a significant control nevertheless.
    Oh, one more thing. Does the energy uptake of the oceans remain constant? I’m wondering if a warm sea surface is less able to absorb more energy because of other physical mechanisms such as evaporation.
    Keep up the good work.

  30. tallbloke says:

    Hi and welcome to the talkshop,

    All good questions and observations. I have become so confused by reading several different accounts of how the Milankovitch cycles affect Earth that I’m starting my own investigation from scratch. This post is about climate in the fairly narrow timeframe, and I cast it that way to avoid as many areas of great uncertainty as possible.

    New post on orbital eccentricity almost ready to go.

    Don’t know wenough about polar ice and arctic sea currents to comment on the albedo loss/oceanic emission question, but my guess is the ocean will lose a heck of a lot more than the ice reflects. This would create a self correcting feedback tending to homeostasis.

  31. Ulric Lyons says:

    “What caused global warming in the late C20th?”

    = changes in the frequency AND timing of peaks in solar wind speed.
    Average global temperatures show the largest rises with higher solar wind speeds during N. Hemisphere winters.

  32. Murray says:

    Tallbloke, glacier advances/retreats are more due to cloud cover (insolation) and precipitation than to temperature. Thus apparently small temperature changes may correlate with glacier changes, but are not likely to be the main driver. Small average temperature changes my be correlated with quite large insolation/precipitation changes at high altitudes/latitudes.

  33. Eddy says:

    Are you saying that the sun does not warm the oceans, or only that certain wave lengths do not warm it? You might want to look at the ‘afternoon effect’ a well known solar phenomenon.

  34. tallbloke says:

    Hi Eddy, No. I’m saying the Sun does warm the oceans. It is ‘back radiation’ which does not warm the oceans. The greenhouse effect slows down the rate at which the oceans can cool.

  35. Stephen Wilde says:

    “The greenhouse effect slows down the rate at which the oceans can cool.”

    I don’t think it does that either because it just accelerates evaporation for a zero net effect on the energy flow from water to air. I went into that in some detail here:


    All that the greenhouse effect does is slow down radiative loss of energy from air to space but that slowing down in the air alone then seems to be offset by a faster water cycle cancelling it out again.

    Would you like to use that article as a guest post here?

    It is a companion piece to my article as to how the solar effect could work. Taken together the two articles cover both the top down solar effects and the bottom up oceanic effects.

    The next step is to analyse how the two processes interact to shift the climate zones to and fro.

  36. tallbloke says:

    Stephen, happy to repost your article.

  37. Ulric Lyons says:

    A thought on clouds: If we see an increase in low cloud and a simultaneous decrease in mid level cloud accompany a temperature drop, in the absence of any perceived external driver, one could easily conclude that increased low cloud causes cooling, and maybe even that increasing mid level cloud causes warming !! (the latter being pretty ridiculous as the mid level clouds are colder)
    While given the situation where externally (solar) forced temperature changes are driving the cloud cover levels, higher levels of low cloud gives more means to distribute ocean heat from clouds formed above them, to maintain higher night-time/winter warmth in colder land regions.

  38. […] about the decrease in cloudiness over the last 50 years while the Sun’s activity levels were well above the long term average. More […]