This paper is to be published on-line on Friday in Physics Letters A Dr. Douglas graciously sent me an advance copy, of which I’m printing some excerpts. Douglas and Knox show some correlations between Top-of-atmosphere radiation imbalance and the Pacific Decadal Oscillation (PDO). The authors credit Dr. Roger Pielke Sr. with reviving interest on the subject due to his discussions on using ocean heat content as a metric for climate change.
Abstract
Ocean heat content and Earth’s radiation imbalance
D.H. Douglass and R, S, Knox
Dept. of Physics and Astronomy, University of Rochester, PO Box 270171, Rochester, NY 14627-0171, USA
Earth’s radiation imbalance is determined from ocean heat content data and compared with results of direct measurements. Distinct time intervals of alternating positive and negative values are found: 1960–mid-1970s (−0.15), mid-1970s–2000 (+0.15), 2001–present (−0.2 W/m2), and are consistent with prior reports. These climate shifts limit climate predictability.
Introduction:
A strong connection between Earth’s radiative imbalance and the heat content of the oceans has been known for some time (see, e.g., Peixoto and Oort [1]). The heat content has played an important role in recent discussions of climate change, and Pielke [2] has revived interest in its relationship with radiation. Many previous papers have emphasized the importance of heat content of the ocean, particularly the upper ocean, as a diagnostic for changes in the climate system [3–7]. In this work we analyze recent heat content data sets, compare them with corresponding data on radiative imbalance, and point out certain irregularities that can be associated with climate shifts. In Section 2 the conservation of energy is applied to the climate system and the approximations involved in making the radiationheat content connection are discussed. In Section 3 data sources are enumerated. Section 4 gives the radiation imbalance for the Earth’s climate system. In Section 5, climate shifts, radiative imbalances and other climate parameters are discussed. A summary is in Section 6.
Discussion:
…
What is the cause of these climate shifts? We suggest that the low frequency component of the Pacific Decade Oscillation (PDO) may be involved. The PDO index changes from positive to negative near 1960; it remains negative until the mid-1970s where it
becomes positive; then it becomes negative again at about 2000. This mimics the FTOA data. The PDO index is one of the inputs in the synchronization analysis of Swanson and Tsonis [43]. One would like to be able to predict future climate. Such predictions are based upon the present initial conditions and some expectation that changes in the climate state are continuous. However, if there are abrupt changes such as reported by Swanson and Tsonis then this is not possible. These abrupt changes presumably
occur because the existing state is no longer stable and there is a transition to a new stable state.
Summary:
We determine Earth’s radiation imbalance by analyzing three recent independent observational ocean heat content determinations for the period 1950 to 2008 and compare the results with direct measurements by satellites. A large annual term is found in both the implied radiation imbalance and the direct measurements. Its magnitude and phase confirm earlier observations that delivery of the energy to the ocean is rapid, thus eliminating the possibility of long time constants associated with the bulk of the heat transferred. Longer-term averages of the observed imbalance are not only many-fold smaller than theoretically derived values, but also oscillate in sign. These facts are not found among the theoretical
predictions.
Three distinct time intervals of alternating positive and negative imbalance are found: 1960 to the mid 1970s, the mid 1970s to
2000 and 2001 to present. The respective mean values of radiation imbalance are −0.15, +0.15, and −0.2 to −0.3. These observations are consistent with the occurrence of climate shifts at 1960, the mid-1970s, and early 2001 identified by Swanson and Tsonis. Knowledge of the complex atmospheric-ocean physical processes is not involved or required in making these findings. Global surface temperatures as a function of time are also not required to be known.
~snip~ Off topic.
Sandy (05:32:53) :
“Seawater is saline, so the effect of temperature is slightly different. The density keeps right on increasing as you approach the (lowered) freezing point.”
No it doesn’t. If it did sheer pressure would cause sea-water to solidify at depth and an ice with a higher density than water would stay on the bottom.
No you were wrong when you said this a couple of days ago and you’re wrong now. I refer you to the explanation I gave then.
No, look at the phase diagram of water, water has an anomalous property that its solidus slopes towards lower temperature at higher pressure, that is why water doesn’t freeze at the bottom of the ocean it has nothing to do with a maximum density. The first poster is correct sea water has a maximum density at its freezing point.
http://www.its.caltech.edu/~atomic/snowcrystals/ice/h2ophase.gif
********
Leif Svalgaard (09:36:11):
Regardless of the discussion about whether it is heat or energy, let us call it H. Then you can calculate H for a temperature of 274K and for 290K and you’ll find that H[274] = 274/290 * H[290] or only 5% smaller, so water at 274K [deep ocean] holds 95% of the energy [heat] of water at 290K [surface], so there is about ten times as much heat below 300 meters than above.
What is of interest for us living in the air just above the surface, is not really how much heat [H] the oceans hold, but how much of that comes from ocean water warmer than the air temperature or some other reference temperature. In studies of hurricanes and tropical storms, a reference temperature of 26C = 299K is often used, so that dH = H-Ho [where Ho is that for 26C] can be positive or negative. One can also use a point in time [say 1970] as a reference to get dH = H(t) – H(1970).
*********
Leif, I agree w/your statements. On an absolute scale, the deep water obviously holds more total heat (joules) than the shallow warm water because it’s much more massive. But as you say & I thought had pointed out was that deep water temps relative to land/ocean surface temps is the point. Since they are colder on average compared to land/ocean surface temps, the deepwater actually stores cold (lack of heat), not warmth. Of course, this “cold” is shielded/insulated from the air by the warm surface waters, and can only cool the weather where upwelling forces it to the surface.
Do you disagree w/my point that the present, stratified state of the world’s oceans (alot of cold bottomwater) cannot have “heat in the pipeline” that would make a significant difference to climate decades later?
frederic (10:46:33):
I’ve read this in several papers, but a quick google search didn’t find them. I’ll search a bit more.
Leif,
The wavelength where temperature is maximum would do just fine. Shorter wavelengths carry more energy than longer wavelengths so it is the variability of the total energy content that I am after.
Is it seen to vary in each of the situations specified and if so is anyone measuring those variations ?
For example the sun provides energy to us at a whole spectrum of wavelengths so what I want to know is how much variability there is in the distribution of wavelengths within the total energy supplied to us.
Then the same for energy entering the oceans, passing from the oceans to the air and passing from the air to space
Leif,
I am puzzled by your implied suggestion that a body at a specific temperature necessarily radiates at a single specific wavelength.
The sun would seem to be an exception in that we get radiation from a very wide spectrum from a single body at a specific temperature.
beng (10:01:01) :
Do you disagree w/my point that the present, stratified state of the world’s oceans (alot of cold bottomwater) cannot have “heat in the pipeline” that would make a significant difference to climate decades later?
I think the whole point of the article in the topic was “Its magnitude and phase confirm earlier observations that delivery of the energy to the ocean is rapid, thus eliminating the possibility of long time constants associated with the bulk of the heat transferred.”, so we agree.
Stephen Wilde (10:38:14) :
the variability of the total energy content that I am after.
That is given by TSI.
For example the sun provides energy to us at a whole spectrum of wavelengths so what I want to know is how much variability there is in the distribution of wavelengths within the total energy supplied to us.
This is given by something called the Spectral Irradiance, e.g. http://lasp.colorado.edu/sorce/data/ssi_data.htm
Then the same for energy entering the oceans, passing from the oceans to the air and passing from the air to space
I think the Outgoing Radiation is also measured at several wavelengths, but don’t have any handy links. Should be easy to find on the net.
Leif,
Can you describe something for me ?
You are expert in solar issues and it would take me much time to check out for myself matters which you could address immediately.
The thing I am curious about is the relationship between TSI and the Spectral Irradiance.
It seems to me that if one were to get more of one wavelength from the sun and less of another then the TSI could stay the same notwithstanding that variation.
However different wavelengths are processed differently by the Earth’s oceans and air so one could presumably get a different climate response from that change in the Spectral Irradiance even though TSI remains pretty much the same.
So how much change do we see in the Spectral Irradiance, how often and in which wavelengths ?
Stephen Wilde (10:43:00) :
I am puzzled by your implied suggestion that a body at a specific temperature necessarily radiates at a single specific wavelength.
No, the spectrum peaks at a wavelength given by the temperature [for black body radiation – Wien’s law]
Stephen Wilde (11:20:21) :
However different wavelengths are processed differently by the Earth’s oceans and air so one could presumably get a different climate response from that change in the Spectral Irradiance even though TSI remains pretty much the same.
But the wavelengths outside of the visual contribute small percentage of the total.
So how much change do we see in the Spectral Irradiance, how often and in which wavelengths ?
The link I gave you has a lot about that. One thing that often surprises people is that some of those variations go in the opposite direction of what they think. For example the near ultraviolet from 242-310 nm varies opposite to the solar cycle, more UV at minimum, e.g. http://www.leif.org/research/Erl70.png
So there is no simple answer to your question, it all depends on what wave length interval and time scale, etc. I can’t give any ‘sweeping’ answer [except my boilerplate one: the variations don’t matter as they are so small 🙂 ]. We can research this piece by piece, but there must be a clear goal or intent, rather than a fishing expedition. And lots of people have already looked at that.
Thanks Leif. it is getting clearer so I’ll go to the point.
Does Solar Spectral Irradiance shift over time between the shortwave and longwave ends of the spectrum and if so by how much and is there any link to levels of solar activity ?
If there is more UV at minimum would there be more shortwave at maximum ?
The reason I ask is that it is only solar shortwave radiation that penetrates relatively deeply into the oceans so an increase in shortwave will add more energy to the oceans and a decrease in shortwave will add less energy to the oceans even if there is little or no change in TSI.
The effect of only a tiny shift towards the shortwave end of the spectrum could make a profound difference to the amount of energy absorbed by the oceans especially if continued for decades or centuries.
Essentially, could shifts in Solar Spectral Irradiance be the (or a) climate smoking gun ?
>> oms (22:31:07) :
Seawater is saline, so the effect of temperature is slightly different. The density keeps right on increasing as you approach the (lowered) freezing point. <<
If the bottom of the ocean had a temperature of 3C, then I would guess that the maximum density of sea water would be at 3C. It’s interesting what Encyclopedia Britannica says–not that anyone considers them to have any expertise (I think there was paper published in Nature or Science that claims Encyclopedia Britannica was no more accurate that Wikipedia).
Jim
And a tiny shift of the WHOLE spectrum towards the shortwave or longwave ends of the spectrum would presumably affect the energy content of ALL the radiation reaching the Earth (even if TSI remained the same) so the effect on the oceans would be much greater than would be expected from a change in a single wavelength or a limited group of wavelengths.
EVERY wavelength more energetic than a particular level would bear an increased capacity to penetrate the ocean surface.
Jim Masterson (12:24:42) :
Jim, the table you linked looks correct. The densest seawater (especially near the bottom) is both salty AND cold, and water near the saltier end of the table is densest near 0 C.
Stephen Wilde (12:26:55) :
This seems true, but then you’d also expect to see a corresponding change in TSI.
Sort of, but it is not quite true that the shortest wavelengths penetrate the sea surface most effectively.
Wozniak and Dera, Light absorption in sea water:
Spectra of light absorption in the visible range measured in different sea waters
oms (13:09:40)
There would not be a corresponding rise in TSI if the increase in energy value of one wavelength was offset by a decrease in energy value of another. For example a large reduction in longwave could be offset by a small increase in shortwave but more energy would enter the oceans for the same value of TSI.
I’m waiting for Leif to tell me whether that is possible.
It does not matter whether the shortest wavelengths penetrate the sea surface most effectively. What matters is the degree of variability in the supply of those wavelengths which do penetrate the sea surface most effectively.
I hope Leif can shed light on that too.
Stephen Wilde (12:26:55) :
There would not be a corresponding rise in TSI if the increase in energy value of one wavelength was offset by a decrease in energy value of another.
I fail to see the relevance of all this. The variation of the energetic UV is less than that of TSI, measured in W/m2, and almost all that UV is absorbed high in the atmosphere anyway, so the variation of the amount that reaches the ocean is minute compared to that of TSI.
Stephen Wilde (12:26:55) :
There would not be a corresponding rise in TSI if the increase in energy value of one wavelength was offset by a decrease in energy value of another.
Too many questions all at once. Let us take one at a time and be very specific: which wavelength interval, for example, and received where. [only one interval and one location to begin with].
NOAA Paleoclimatology What’s New.mht:
“Loulergue et al. Nature
Vol. 453, No. 7193, pp. 383-386, 15 May 2008. doi: 10.1038/nature06950
Atmospheric methane is an important greenhouse gas and a sensitive indicator of climate change and millennial-scale temperature variability. Its concentrations over the past 650,000 years have varied between ~350 and ~800 parts per 109 by volume (p.p.b.v.) during glacial and interglacial periods, respectively. In comparison, present-day methane levels of ~1,770 p.p.b.v. have been reported. Insights into the external forcing factors and internal feedbacks controlling atmospheric methane are essential for predicting the methane budget in a warmer world. Here we present a detailed atmospheric methane record from the EPICA Dome C ice core that extends the history of this greenhouse gas to 800,000 yr before present. The average time resolution of the new data is ~380 yr and permits the identification of orbital and millennial-scale features. Spectral analyses indicate that the long-term variability in atmospheric methane levels is dominated by ~100,000 yr glacial-interglacial cycles up to ~400,000 yr ago with an increasing contribution of the precessional component during the four more recent climatic cycles. We suggest that changes in the strength of tropical methane sources and sinks (wetlands, atmospheric oxidation), possibly influenced by changes in monsoon systems and the position of the intertropical convergence zone, controlled the atmospheric methane budget, with an additional source input during major terminations as the retreat of the northern ice sheet allowed higher methane emissions from extending periglacial wetlands. Millennial-scale changes in methane levels identified in our record as being associated with Antarctic isotope maxima events are indicative of ubiquitous millennial-scale temperature variability during the past eight glacial cycles.”
”
Ahn and Brook Science
Vol. 322, No. 5898, pp. 83-85, 3 October 2008, doi:10.1126/science.1160832.
Reconstructions of ancient atmospheric carbon dioxide (CO2) variations help us better understand how the global carbon cycle and climate are linked. We compared CO2 variations on millennial time scales between 20,000 and 90,000 years ago with an Antarctic temperature proxy and records of abrupt climate change in the Northern Hemisphere. CO2 concentration and Antarctic temperature were positively correlated over millennial-scale climate cycles, implying a strong connection to Southern Ocean processes. Evidence from marine sediment proxies indicates that CO2 concentration rose most rapidly when North Atlantic Deep Water shoaled and stratification in the Southern Ocean was reduced. These increases in CO2 concentration occurred during stadial (cold) periods in the Northern Hemisphere, several thousand years before abrupt warming events in Greenland.”
Leif Svalgaard (21:37:38) :
Stephen Wilde (12:26:55) :
“There would not be a corresponding rise in TSI if the increase in energy value of one wavelength was offset by a decrease in energy value of another.”
Too many questions all at once. Let us take one at a time and be very specific: which wavelength interval, for example, and received where. [only one interval and one location to begin with].
Over in the Livingston thread I posted this useful link:
http://solar.physics.montana.edu/SVECSE2008/pdf/floyd_svecse.pdf
An extract from Leif’s link:
“Solar Ultraviolet (UV & EUV) Irradiation
Interesting Questions for Further Research :
1) What are the detailed mechanisms of solar UV irradiance
variation?
2) What is the connection between magnetic activity and UV
irradiance variations?
3) What is the contribution of UV variation to that of the TSI?
4) How much does the solar UV vary over time periods longer
than the solar activity cycle?
5) What was the solar UV irradiance during the Maunder
Minimum?
6) How well does the Mg II index describe relative irradiance
variations from the EUV to the visible?”
My comments:
Now the interesting thing is that those issues need further research i.e. we do not know the answers.
I consider that without the answers we cannot assess how much variability there is in the amount of energy from shortwave radiation from the sun and which can penetrate the ocean surface deeply enough to overcome the evaporation barrier and thereby add to the total oceanic energy store.
Note that such variations can occur without a significant change in TSI because we are talking about variation of the distribution of specific wavelengths within the spectrum and not the total power or energy value of all the energy delivered.
If over time the Earths receives a different proportion of the energy from the sun in the form of wavelengths that are more or less efficient at penetrating the ocean surface then the oceanic energy store will change up or down without a corresponding change in TSI.
Thus the climate smoking gun is not TSI after all but rather changes in the Solar Irradiance Spectrum (SIS) and according to Leif’s link the scale and effect of changes in the SIS are currently unknown with further research needed.
There is the most likely explanation for the observed fact that climate changes are apparently too large to be explained by TSI changes. Just switch the wavelengths received around a bit and the necessary changes in solar energy flow into the oceans can be explained without a proportionate change in TSI. Voila.
I feel a new article coming on (The Climate Smoking Gun – The Solar Irradiance Spectrum).
Stephen Wilde (23:34:28) :
Now the interesting thing is that those issues need further research i.e. we do not know the answers.
I think that is a misrepresentation. What is meant was that we do not know as much as we would like, not that we don’t know anything.
And you logic is faulty, it does not follow that because we don’t know anything that that which we don’t know is the cause of something we do know. This is the classic Al Gore Argument: “If you don’t know anything, everything is possible”.
We do know quite a lot as the link explains. And as the article of this thread explains direct measurements “eliminated the possibility of long time constants associated with the bulk of the heat transferred”, so no long-term storage in the Oceans. So we may have a smoking gun, but no dead body 🙂
“Leif Svalgaard (00:00:36) :
Stephen Wilde (23:34:28) :
Now the interesting thing is that those issues need further research i.e. we do not know the answers.
I think that is a misrepresentation. What is meant was that we do not know as much as we would like, not that we don’t know anything.
And you logic is faulty, it does not follow that because we don’t know anything that that which we don’t know is the cause of something we do know. This is the classic Al Gore Argument: “If you don’t know anything, everything is possible”.
We do know quite a lot as the link explains. And as the article of this thread explains direct measurements “eliminated the possibility of long time constants associated with the bulk of the heat transferred”, so no long-term storage in the Oceans. So we may have a smoking gun, but no dead body :-)”
Reply:
Not knowing as much as we would like is quite sufficient for current purposes. We do not know how much the energy flows into and out of the oceans change independently of TSI variation. Enough said.
If the smoking gun is variations in the SIS then the dead body is the periodic change in the rate of energy flow from the oceans. One does not need long term storage, merely 30 years as per the observed PDO phase shifts.
It would seem that it takes 30 years or so for the changes forced by variations in the SIS to overcome other oceanic variables and drive an oceanic phase change.
A far simpler and more elegant solution than I ever expected to find.
Longer term changes due to TSI variability over centuries would still be in the background but multidecadal changes would be down to the SIS changing over time and building up an effect within the oceans.
Tallbloke has pointed out that the oceanic phase shifts in the Pacific seem to occur every third solar cycle and usually around solar minimum. That would fit the approximate 30 year cycle and the reason for it occurring at minimum may be linked to your comment that counterintuitively UV (shortwave) is greater at minimum and so would more often cause (or fail to prevent) a phase shift at solar minimum.
Stephen Wilde (00:23:03) :
We do not know how much the energy flows into and out of the oceans change independently of TSI variation. Enough said.
Therefore you can say nothing about this. You can surmise, guess, postulate, posit, assume, presume, etc, but it is all just hand waving and no substance, because you cannot explain an effect if the input is unknown. ‘Nuff said.
Stephen Wilde (00:23:03) :
We do not know how much the energy flows into and out of the oceans change independently of TSI variation.
And what we do know about what causes TSI, it is not the case that a given change in UV will have the opposite change in TSI. TSI has a base cause: the photosphere is hot. On top of that the magnetic field effects add to TSI and to the UV [which is a part of TSI], so both go up and down in concert. Within the UV regime there is some variation: some wave lengths vary opposite to others [because of the many absorption lines present in those regions], but UV as a whole vary with TSI. And in any case these variations are very small when measured in W/m2.
So to have a smoking gun you must be specific: which wave length region? what does it do to the oceans? etc. The old saw ‘isn’t it conceivable that…’ won’t cut it in my book.
Stephen and Leif, as I just wrote on the P&L thread, I continue to be intrigued by the idea that the variation in the shape of the peaks of the cosmic rays each solar cycle alternates and that two of each type of cosmic ray peak and one of the other type are in each of the alternating warming and cooling phases of the PDO. It is so, that the difference of distribution of the cosmic rays around the poles of the sun is small, but if the phenomenon is linked to the behaviour of the PDO then a mechanism may be suggested.
========================================
Stephen Wilde (00:23:03) :
This has been, and continues to be, a subject of rather intensive observational programs.
How much do you expect us to know about it?
steve (06:24:12) :
And if you take that 1.2mm and subtract the upper range of ground water addition to sea level
Why would I want to do that? There is a cycle. Water evaporates from the oceans. It rains on the land and oceans. It fills the water table under the land and runs back to the oceans. If you extract it from the water table and then let the waste flow to the sea overland you are short circuiting the process, but not fundamentally changing it.
Expanding Earth: I think the sea level datum is relative to the land surface, not the earth’s centre. So this is probably already accounted for in the satellite altimetry, but needs checking out.
The bottom line is, sea surface temperatures have risen. As far as we can make out from the XBT units, ocean heat content has risen. Water expands when you warm it.
The temperature range across the top 700m of the ocean is consistent with a fairly linear drop from surface to thermocline. The 0.3C rise in SST 1993-2003 is consistent with a 0.15C rise in the average temperature of the top 700m.
This is consistent with a rise in OHC of around 14×10^22J, which is consistent with a sea level rise of around 16mm. Which is consistent of IPCC estimates of the total sea level rise being roughly half due to thermal expansion, and the rest due to meltoff and the other factors. It’s all debatable, but the estimates seem to fit the theory reasonably well, notwithstanding the puzzle of how the heat is mixed down.