Pinatubo and the Albedo Thermostat

Guest Post by Willis Eschenbach

I got to thinking that the eruption of Mount Pinatubo should provide a good test case for my theory that changes in albedo help regulate the temperature and keep it within a narrow range. When a big volcano erupts, it throws both black and reflective particles and aerosols high into the stratosphere. As a result, the albedo (reflectivity) and the shortwave absorption of the stratosphere both increase, available solar energy decreases, and the planet cools. By that theory, available solar energy (after albedo) should show a dip after the eruption followed by a recovery to pre-eruption values.

I have demonstrated in the post cited above and elsewhere that cloud albedo varies with temperature in order to prevent excessive heating or cooling. As a result, if my hypothesis is correct, we should see a more complex and different reaction to the eruption. We should see a planetary albedo response to Pinatubo that restores the temperature. If there is a temperature governing mechanism in play, we would expect the change in albedo to provide extra energy to counteract the cooling effect of the eruption. So we’d expect the available energy to overshoot, providing more energy than normal, until the balance is restored.

So I plotted up the change in available solar energy, which I calculated as total solar irradiance TSI * (1 – albedo), for the period 1984-1998. I also plotted the global average temperature for the period, converted to a blackbody equivalent temperature in watts per square metre. The results are shown in Figure 1.

Figure 1. Anomaly in solar energy available after albedo reflection (yellow), and global  equivalent blackbody surface temperature anomaly (dark red). Date of the Pinatubo eruption is shown in green.

How do these observations support the idea that there is an albedo-based temperature regulating mechanism?

Look at what actually occurred after the eruption. The amount of solar energy began rising, and continued to rise in opposition to the falling temperatures. It remained high, significantly higher than most of the record, until the temperatures started to rise in 1994. Only after temperatures had returned to pre-eruption levels did the amount of solar energy drop back down.

This is precisely the response that would be expected from a governing mechanism involving the albedo. It would let more energy into the system to counteract the volcanic cooling, and would continue doing so until the temperature was restored.

So I hold that this provides clear visual evidence that an albedo-based temperature compensating mechanism does exist. I also say that this mechanism explains the poor performance of the GISS climate model, whose predictions of Pinatubo I discussed here and here. Because it does not include any albedo-based compensatory mechanism, the GISS model predicted a much larger temperature drop than actually occurred.

w.

DATA NOTES:

The albedo data is from Long-term global distribution of Earth’s shortwave radiation budget at the top of atmosphere (PDF), N. Hatzianastassiou et al., Figure 8(c). Temperature data is from the HadCRUT3 anomaly dataset and the HadCRUT absolute temperature dataset. Absolute temperatures plus anomalies were first converted to the equivalent blackbody temperature in watts per square metre using the Stefan-Boltzmann equation. Both datasets then had their monthly averages removed before graphing.

Advertisements

36 thoughts on “Pinatubo and the Albedo Thermostat

  1. Willis –
    is this a new factor worth considering?
    21 Aug: National Science Foundation: Newly Discovered Icelandic Current Could Change Climate Picture
    Current called North Icelandic Jet contributes to key component of ocean circulation
    If you’d like to cool off fast in hot summer weather, take a dip in a newly discovered ocean current called the North Icelandic Jet (NIJ)…
    Now, in a paper published in the August 21st online issue of the journal Nature Geoscience, the team of researchers–including the two Icelanders who discovered the current–has confirmed that the Icelandic Jet is not only a major contributor to the DSOW but “is the primary source of the densest overflow water.”
    “We present the first comprehensive measurements of the NIJ,” said Robert Pickart of the Woods Hole Oceanographic Instititution in Massachusetts, one of the co-authors of the paper…
    “We’ve identified a new paradigm,” Pickart said, likely a new, overturning loop of warm to cold water.
    The results, Pickart says, have “important ramifications” for ocean circulation’s impact on climate.
    Scientists have been concerned that this overturning loop–some call it a conveyor belt–is slowing down due to a rise in global temperatures…
    Eventually, this could lead to a colder climate in the northern hemisphere…
    http://www.nsf.gov/news/news_summ.jsp?cntn_id=121415&org=NSF&from=news

  2. Yes, Roy Spencer has done some work on radiation budget around Mt P. eruption. Find it on his site. there is about a 15 to 18 month lag in the temperature change wrt the drop in radiation budget. You can see an indication of this in your plot. .
    This is a key issue since Hansen’s (mis-)use of an exaggerated volcanic cooling in the 1960-1995 period is key to the GISS model’s excuse for pumping up the calculated CO2 warming by about a factor of three.
    If the climate is auto-correcting to a large extent the net effect of volcanoes is very much smaller any they can’t get away with frigging the climate sensitivity to “prove” CO2 AGW.
    I have been trying to get some direct evidence of this volcanic rebound looking at rate of change of temperature in the whole thermometer period. It seems to be there but I have not been able to produce concrete evidence that would stand up to determined criticism. Phase relationships are key here.
    (BTW it’s not an equivalent blackbody *temperature* if it’s measured in in watts per square metre 😉 )

  3. Looking at that paper it seems that the albedo data is purely from a rather crude computer model projecting from climatic data. It was “verified” against the usual IPCC suspects.
    It may be more enlightening to use real data: a) because it may actually prove something if you show a link and b) you may find a clearer correlation using real data rather than make believe data.

  4. The[] delay between the eruption and the peak in available solar energy anomaly is longer than I would expect.
    Dr. Roy Spencer has an interesting article on using Pinatubo to calculate sensitivity.
    http://www.drroyspencer.com/2010/06/revisiting-the-pinatubo-eruption-as-a-test-of-climate-sensitivity/
    He shows a graph of the net radiative flux anomaly and the corresponding graph of optical density. http://www.drroyspencer.com/wp-content/uploads/Pinatubo-revisited-SAGE-tau-and-ERBE.gif
    The net flux anomaly had decline to zero when the optical density was still about 50% of the peak. The net flux anomaly is the sum of the forcing + feedbacks. He calculated a slightly negative feedback, that would correspond to a bit less than 1 degree C rise for a doubling of CO2
    (assuming that doubling of CO2 is a radiative forcing of 3.7 watts/m^2, which Willis’s recent article shows might be more speculative and unsure than the IPCC would have us believe).

  5. Good article Willis. You will be able to prove your point with the data – it matches up every time you or someone else looks at a particular aspect of it.

  6. It’s interesting but the general level of random variation makes me uncomfortable with it.
    I also noticed that both temperature and albedo have an apparent upward trend, so is this confirmation of your idea? Or is it a refutation?

  7. Albedo is the name of the game. The variation in Earth’s albedo is poorly understood. But, this is the single most influential variable. And these “Scientists” ignore this?

  8. Have you looked at the Lake Toba megavolcano 75 Kya on Sumatra? A 100×50 km. caldera. Prof. John Westgate (UoT) and others identified it in the ’90s. It was about 1000X Mt. St. Helen’s size, and dropped sea temps by about 6°C for a millennium. A huge sulphur compound cloud that circled the Earth was associated with it.
    About the time of the population “bottleneck” that isolated and selected the forebears of humanity on the E. Africa Coast.
    There are no coincidences.

  9. P.S. The “1000X” is a fairly direct measure of the ash generated; about 1 km^3 of rock was put into the air by MSH, and a full 1000 km^3 by Toba.
    And the ocean cooling lasted a few millennia, IIRC.

  10. That dip in the black body temp anomaly circa 1984- could that have been caused by albedo-based temperature compensation resulting from the Mt. Saint Helen’s 1980 eruption?

  11. Hi Willis,
    One comment about “overshoot” in your posts about Pinatubo and climate models.
    You do not need a “governor” type control for a system to overshoot, or even to oscillate continuously. (For non-engineering types: a governor is proportional plus integral, PI control, or proportional plus integral plus derivative, PID control, as opposed to simple negative feed-back control or proportional only, P control.) You need only have some negative feed-back combined with sufficient “time lag” in the negative feed-back mechanism for there to be overshoot. Damped or un-damped oscillation are very common behaviors in simple proportional-only control systems (that is, those with only negative feedback). You are of course correct that negative feedback alone can’t completely compensate for an external applied “forcing”; in any such system, there must be some net shift in response to an applied forcing. However, depending on the effective size of the feedback, the net shift (or “off-set”) can be quite small.
    The time required for transport of heat from the tropics to higher latitudes might provide sufficient time lag for significant overshoot to happen in response to a volcanic eruption, even if the response is in fact only negative feed-back.

  12. Very interesting Willis, and certainly worthy of looking at a bit closer perhaps, but a few observations of other factors that may be involved in one way or another:
    1991 was also a near the 2nd double-peak of solar cycle 22
    , and total solar irradiance from the sun was high (meaning GCR’s were low).
    See:
    http://www.climate4you.com/images/SolarIrradianceAndSunspots.gif
    http://www.climate4you.com/images/CosmicRaysAndSunspotsMonthlySince195801.gif
    Also, ENSO activity was favoring longer EL Nino period during this time, and had actually transitioned from a La Nina to El Nino prior to the eruption. Just wondering how much these may be relevant to your supposition.

  13. Willis, if your theory is correct, then albedo should increase as global mean temperature anomaly increases. In Figure 1 above, one can see that the blackbody anomaly increases (with some fluctuation) from 1984 to 1997. However, Figure 5b of the Hatzianastassiou paper that you cite shows the albedo decreasing over the same time interval. Don’t those data contradict your theory?

  14. Like CO2 in a glacial termination the lag in albedo change mean that it follows temperature change not precedes it and therefore cannot be invoked as a cause of temperature change!
    More seriously, the figures you use for change in albedo only cover the change in albedo with reference to Short wave radiation from the SUn. It does not include changes in the long-wave albedo caused by alterations in the emission and absorption within the atmosphere by optically active sulphur compounds and particulates.
    Without dealing with the TOTAL energy flux in the system it is ‘premature’ to identify a homeostatic or stabilising feedback in the system.

  15. Willis, would you say that there was a net warming or a net cooling of the stratosphere following the Pinatubo eruption ?

  16. I’m not comfortable with assigning any significance given the level of noise in the graph.
    To put it another way, if anyone was provided with “Figure 1”, with the green line and the dates removed, would anyone be able to draw in the green line, and say with any confidence whatsoever “look – a major volcano must have erupted just HERE”.
    I think not. And without that, can this theory be justified?

  17. “Long-term global distribution of Earth’s shortwave radiation budget at the top of atmosphere.” by
    N. Hatzianastassiou et al.
    “…On a mean annual and global basis, the model is in very good agreement with ERBE, overestimating the outgoing SW radiation at TOA (OSR) by 0.93 Wm−2 (or by 0.92%), within the ERBE uncertainties. At pixel level, the OSR differences between model and ERBE are mostly within ±10 Wm−2, with ±5 Wm−2 over extended regions, while there exist some geographic areas
    with differences of up to 40Wm−2, associated with uncertainties in cloud properties and surface albedo…”
    Willis, much of the data used to measure radiative changes seems to be based on modelled data, rather than observation. In view of this, is the data quality/accuracy sufficient to support or refute your hypothesis?

  18. Interesting – I hadn’t realised how much albedo was derived rather than measured. I’m not entirely convinced by the correlations, but it’s an intriguing theory. Presumably you’re saying albedo could essentially drive climate. As always clouds will complicate that. Have you looked at the ICTZ and the albedo changes in African/ Australian savannah’s? It’s amazing how quickly black burnt landscape greens – that must be an awesome albedo change.
    It makes me wonder if someone has tried to do something along the lines of Lovelock’s daisy world model but replacing the daisys with albedo from approximations to best-guess land use changes and using real solar radiation. I wonder if it would self-stabilise?
    Start cold: with high global albedo because of snow/ice. Tropical deserts become cooler and more vegetated which would then warm the tropics, decreasing albedo. Circulation would then begin to melt ice/snow to give more oceans which would warm to give more wild fires, probably increased precipitation and ultimately expand tropical deserts giving lower albedo to the tropics. temps begin to fall which triggers fresh snow at high alts and the polar regions. Albedo increases and the tropics begin to cool together with increased precipitation from warm oceans and the tropical forests expand while deserts shrink.
    I know the Sahara was vegitated not so long ago. Has that been cyclical, or does the soil get so denuded once desertified that there’s no way back, even with cooler temps and increased rainfall?
    Does climate variation need CO2 at all?
    Too many questions, not enough comment. Sorry!

  19. I’m afraid I’m a bit perplexed here. How can the action of increased albedo following an eruption result in compensating increased solar irradiance after the sky has been cleared? What is the supposed mechanism? Or have I totally misunderstood the concept.

  20. Gary Pearse says:
    August 22, 2011 at 9:18 am (Edit)

    I’m afraid I’m a bit perplexed here. How can the action of increased albedo following an eruption result in compensating increased solar irradiance after the sky has been cleared? What is the supposed mechanism? Or have I totally misunderstood the concept.

    Gary, the mechanism is that the clouds are related to the temperature of the earth, which stayed low even after the sky had cleared, because the world had cooled. As a result, there was less clouds, leading to more available solar energy to warm the planet.
    w.

  21. Willis,
    You have an interesting idea. The atmosphere isn’t warming enough for a GHG effect, but it is warming correctly for an albedo effect. I’ve been intrigued about the lower stratosphere cooling that appears in the satellite record.
    It appears that only certain volcanic eruptions can cause an effect. The St. Helens eruption doesn’t even show, but both El Chichon and Pinatubo do. Also there’s a step down cooling that doesn’t seem to recover. The cooling of the stratosphere can be caused by GHGs, but other things can cause a similar effect–like semi-permanent albedo changes.
    It would be nice if climate scientists were actually doing their jobs and investigating things like this. Instead they only have one tool–a hammer (CO2) and that makes everything else a nail (AGW). It’s too bad because they now seem more worried about aliens attacking instead of doing better science.
    Jim

  22. The reason I asked about the effect of Pinatubo on stratospheric temperatures is that stratospheric cooling results in more poleward (or more zonal) jets for reduced global cloudiness and with more energy entering the oceans for overall system warming whereas stratospheric warming results in more equatorward (or more meridional) jets for increased global cloudiness with less energy entering the oceans for overall system cooling.
    I would like to know whether that proposition is consistent with observations following the Pinatubo eruption.

  23. The impression I have is that volcanic eruptions give a rapid stratospheric warming which soon dissipates but that there is a separate longer term underlying trend in stratospheric temperatures that may be solar induced.
    That is consistent with the chart kindly supplied by Jim Masterson.
    Thus whilst the sun was more active the stratosphere generally cooled with more poleward/zonal jets but now that the sun is less active we seem to be seeing the stratosphere no longer cooling and perhaps beginning to warm with more equatorward/meridional jets.
    The volcanic effects being mere short term interruptions in a longer term solar induced trend.

  24. @- Stephen Wilde says:
    “Thus whilst the sun was more active the stratosphere generally cooled with more poleward/zonal jets but now that the sun is less active we seem to be seeing the stratosphere no longer cooling and perhaps beginning to warm with more equatorward/meridional jets.”
    The long-term cooling observed in the stratosphere is not obviously correlated with any solar changes, which have been too small to have such an effect. Neither is there clear evidence of stratospheric warming – unless you have a link that shows otherwise!
    The decades long trend in stratospheric cooling is a product of the change in two key components of the stratosphere with major thermal effects. The increase in CO2 increases cooling and the decrease in ozone reduces warming.
    Both of course are anthropogenic changes.

  25. Thanks Willis for the explanation of the phenomenon: from volcanic plume -increased albedo-cooling-reduces the cloud cover after for a period after the sky has cleared permitting more sunshine in.
    Jim Masterson says:
    August 22, 2011 at 2:46 pm
    Why some volcanoes have the cooling effect, others not. It takes an explosive event strong enough and voluminous enough to project the plume up into the stratosphere. If it doesn’t reach the stratosphere, it is dispersed in the lower-intermediate atmosphere and settles to the surface more quickly. Mount St Helens had a strong local effect to be sure but not a significant global one.

  26. One of these days we’re going to get a volcanic eruption that cools the earth enough to spell the end of the Holocene interglacial. The conditions are ripe. Interglacial is already 2000 years older than average of 10,000 years. Axial precession is halfway to its least amount of tilt and orbital precession has closest approach to sun just three weeks short of middle of winter. Both parameters are and will continue for several thousand years moving towards point of least NH difference between summer and winter temperatures which is what gives ice & snow the upper hand.
    We’re sitting ducks. And environmental whackos want to help it along by slowing down anthropogenic warming and thus decreasing whatever safety margin we have against a volcanic eruption ending the interglacial period. Lunatics…
    .

  27. Stephen Wilde says:
    August 22, 2011 at 3:13 pm

    The reason I asked about the effect of Pinatubo on stratospheric temperatures is that stratospheric cooling results in more poleward (or more zonal) jets for reduced global cloudiness and with more energy entering the oceans for overall system warming whereas stratospheric warming results in more equatorward (or more meridional) jets for increased global cloudiness with less energy entering the oceans for overall system cooling.
    I would like to know whether that proposition is consistent with observations following the Pinatubo eruption.

    Stephen, that’s an intriguing hypothesis. Do you have any citations or further reading on that?
    w.

  28. izen says:
    August 23, 2011 at 2:46 am

    … The long-term cooling observed in the stratosphere is not obviously correlated with any solar changes, which have been too small to have such an effect. Neither is there clear evidence of stratospheric warming – unless you have a link that shows otherwise!

    Thanks, izen. I agree with you on both counts. As the satellite shows, what we have is a very odd looking record for which I’ve never heard an explanation.

    The decades long trend in stratospheric cooling is a product of the change in two key components of the stratosphere with major thermal effects. The increase in CO2 increases cooling and the decrease in ozone reduces warming.
    Both of course are anthropogenic changes.

    The odd thing about the stratosphere is that it is a series of stable plateaus, punctuated by volcanic eruptions. After each eruption, a new, lower plateau has started up.
    If the cause of the decline were in fact the increase in CO2 and the decrease in ozone as you say, then what is the explanation for the level plateaus between the eruptions?
    Alternatively, why would the effects of the CO2 and ozone only affect the temperature immediately after the eruptions?
    I don’t have answers to those questions. Until I do, I’m unwilling to ascribe causality unless it includes an explanation for the odd looking historical record.
    Regards,
    w.

  29. “The long-term cooling observed in the stratosphere is not obviously correlated with any solar changes, which have been too small to have such an effect. Neither is there clear evidence of stratospheric warming – unless you have a link that shows otherwise!”
    Your wish is my command:
    http://www.jstage.jst.go.jp/article/sola/5/0/53/_pdf
    “The evidence for the cooling trend in the stratosphere may need to be revisited. This study presents evidence that the stratosphere has been slightly warming since 1996.”

  30. “The odd thing about the stratosphere is that it is a series of stable plateaus, punctuated by volcanic eruptions. After each eruption, a new, lower plateau has started up”
    That’s not the way it looks to me. I see a downward trend up to about 1996 disrupted by two volcanic eruptions along the way and a cessation of cooling from the mid 90s when solar cycle 23 started to fade from its peak and now early signs of a recovery.

  31. Stephen Wilde says:
    August 26, 2011 at 12:59 pm

    “The odd thing about the stratosphere is that it is a series of stable plateaus, punctuated by volcanic eruptions. After each eruption, a new, lower plateau has started up”
    That’s not the way it looks to me. I see a downward trend up to about 1996 disrupted by two volcanic eruptions along the way and a cessation of cooling from the mid 90s when solar cycle 23 started to fade from its peak and now early signs of a recovery.

    Thanks, Stephen. That’s one of the problems with only having such a short satellite record of the temperatures. When you remove the volcanic sections, we’re left with three short records. None of the three have a trend which is significantly different from zero, with the first one way too short to signify anything.

    However, as you point out, it could also be interpreted as a decline from the start of the record to about the middle of the record, followed by the second half of the record which has no trend at all.
    Either one is odd. And a vague reference to solar cycles doesn’t explain what you describe as a fifteen-year steady decrease (with volcanic punctuation), followed by fifteen years of dead-level no-trend temperatures.
    So under each interpretation (either steady decline or plateaus for half the record, followed by level temperatures for the second half of the record) the stratospheric temperature record does not have an obvious explanation.
    All the best,
    w.

  32. I think we need to look at possible alternative explanations for a decline for 15 years and there are two competing scenarios:
    i) Human emissions of CO2 causing a cooling of the stratosphere by reducing upward energy loss as proposed by AGW theory.
    ii) An active sun causing a cooler stratosphere rather than a warmer stratosphere which is the opposite of the conventional wisdom as proposed by me alone as far as I know (before Joanna Haigh said we might have to reverse the usual sign of solar effects on the atmosphere).
    However i) does not explain the plateau whereas ii) does because the cessation of stratospheric cooling occurred as the level of solar activity declined after the peak of cycle 23.
    The shortness of the record is indeed a problem but it may well be that so far the quiet sun has only had long enough to stop the stratospheric cooling trend with a warming trend still to become established if the sun remains quiet long enough.Looking at single cycles is not enough. Looking at groups of successive cycles is better. Best is looking at solar variability from LIA to date (1600 to 2000).
    The flatness of the trend in the period between the two eruptions is odd but may simply be a side effect of the two eruptions occurring so close together.It may be the case that the natural trend has to play catchup for a while after the effects of such eruptions.
    The test will be what happens to stratospheric temperatures over the next few years if the sun stays quiet.
    The importance of stratospheric temperatures is that they control the height of the tropopause which in turn affects the surface pressure distribution globally. I suspect that a warming troposphere at a time of active sun has a cooler stratosphere and more poleward/zonal mid latitude jets letting more energy into the oceans.A cooling troposphere at a time of less active sun has a warmer stratosphere and more equatorward/meridional jets letting less energy into the oceans.
    I know that is contrary to existing climatology but it is the only scenario that fits observations without the need to invoke a human influence from more CO2. That anthropogenic scenario is dead because the stratosphere stopped cooling despite continuing increases in emissions and global CO2 levels.
    I have the only hypothesis currently available which fits the observations so well.

  33. Stephen;
    iii) the H0 is that the surface record is fudged, and the actuality matches the stratospheric pattern, not the accepted curves (especially since The Great Dying of the Thermometers circa 1990.)

Comments are closed.