Home/BlogThis is an update of my last post, where I described the results of a Forcing-Feedback-Diffusion (FFD) model of ocean temperature variations to 2,000 meters deep.
The model assumes the GISS-assumed forcings since 1880, including greenhouse gases, volcanoes, and manmade aerosol pollution. To those, I added a forcing term proportional to El Nino/La Nina activity, equal to 0.9 x MEI (Multivariate ENSO Index), in Watts per sq. meter. I adjusted ocean turbulent heat diffusion coefficients and the El Nino term until I got a correlation of 0.95 between the model temperature variations for the surface-50 meter ocean layer for the period 1955 through 2010. It also matches the warming trend in the 650-700 m layer during 1955-2010, which is the deepest layer for which we have long-term data to compare to.
I’ve now extended the model simulation back to 1880, since GISS forcings go back that far, and NOAA has an MEI reconstruction of El Nino/La Nina activity that goes back even before that. The optimum climate sensitivity was 1.1 deg. C for a doubling of atmospheric CO2, a climate sensitivity so low that the IPCC considers it very unlikely.
The results look like this (click on image for full size version):
Note I have also added the HadSST2 sea surface temperatures, scaled to match the variability of the Levitus 0-50 meter layer during 1955-2010.
The match is pretty good, except the model does not capture the exceptionally cool conditions during 1900-1935. Since this was a period of low sunspot activity, it could be this is a cosmic ray effect on global cloud cover.
The important message here is that this simulation was done with a very low climate sensitivity, corresponding to only about 1/3 the warming rate the IPCC projects for the future in response to increasing atmospheric CO2.
A Note on Deep Ocean Heat Storage
For those interested in the deep-ocean heat storage issue, the assumed model diffusivities I used (which average 3.7 x 10-4 m2/sec down to 2,000 meters depth) are considerably larger than what is usually assumed in climate model simulations. If I use a diffusivity much closer to what is traditionally used (1.2 x 10-4 m2/sec), then I have to reduce the model sensitivity to 0.9 deg. C for 2XCO2 in order to still match the near-surface warming trend, otherwise the model warms too much compared to observations.
Roy, I have more or less an undergraduate level of understanding of this subject matter. But my question on this is: how reliable are the ocean and land temps globally in the pre-satellite measurement era (about 1979 from what I understand)? Just asking for my own education. Thanks for the interesting model.
Kind regards,
Sam
No one knows. I’m just assuming the SSTs are “truth”, and seeing what kinds of physical processes can explain them.
This is very impressive result.
I have taken a look at the GISS data and they seem to be a bit scarce at the end, some series are missing completely and some are suspiciously constant near the end – did you use some kind of extrapolations/fixes to these data or did you use another more complete source?
I’d also like to ask whether you think that the source data can be extrapolated somehow and the model ran a few years to the future. For the GISS data some series seem to be unpredictable (especially StratAer) but some are pretty smooth and might bear a bit of extrapolation. I am completely unsure about MEI data though.
I extrapolated the GISS total forcing into the future.
I do think it would be possible to make some predictions out a few years, given certain assumptions. Been thinking about that.
The fit in the 50’s and sixties will be better, I think, with the new HADSST3. It’s available currently up to 2006 at the KNMI climate explorer. I find it very impressive that the model is able to capture so many inter-annual fluctuations of temperatures so well, even in the early period. It’s not bad for a simple model.
Yes…it makes me wonder whether the MEI reconstruction from the early 20th Century is actually based upon global average temperatures in some way. If so, those early fluctuations that seem to agree so well are not independent.
Hello Roy, you said this
“Since this was a period of low sunspot activity, it could be this is a cosmic ray effect on global cloud cover.”
which suggests that you are pretty much sold on the Svensmark hypothesis.
I’m in agreement that it seems to get cloudier when the sun is less active (more cosmic rays) but personally I prefer the hypothesis that when the sun is less active the mid latitude jets move towards the equator and/or become more meridional due to a surface expansion of the polar air masses under more negative polar vortices thereby causing longer air mass boundaries, more air mass mixing and more clouds.
Do you not think that that is a more plausible explanation than simply more cosmic rays?
The thing is that to achieve such a shift in the surface air pressure distribution one needs an adjustment in the atmospheric vertical temperature profile but the Svensmark idea does not seem able to provide that.
What do you think?
If as you suggest some cooling influence causes a cloud change which amplifies the cooling, that is positive feedback. So, warming from more CO2 would then cause fewer clouds, amplifying the warming (which is what the IPCC believes).
But I *do* agree that changes in the general circulation of the atmosphere can cause global average cloud changes, which in turn can cause warming or cooling. Chaos in the climate system. Instead, the IPCC insists on such changes being the fault of small human-induced forcings, which would require positive feedbacks to cause the magnitude of the observed changes.
It seems that this model has possible time delays at many ocean depths. This makes it not far from a model with the form of a partial differential equation simultaneous with an ordinary differential equation. The order classification of ordinary differential equations does not naturally apply to such models. But if one were willing to take the crude view of the ocean as having dynamics with a single time constant or rate coefficient, one might view this model as a quasi-second order order linear model, externally driven. It seems to me that what Dr Spencer calls the “internal forcing” by ENSO would more conventionally be called ‘external forcing by ENSO’; for it is imposed from an outside dataset, not generated by the internal dynamics of the model.
In the corrupt language of the IPCC’s “forcings and feedbacks” formalism, a “climate sensitivity” of 1.1C for a CO2 doubling would be regarded as nearly a “no-feedback” sensitivity. In that corrupt language, a “climate sensitivity” of 0.9C for a CO2 doubling would be regarded as showing a mild degree of “negative feedback”. (A “climate sensitivity” of 1.4C for a CO2 doubling would then be said to show a mild degree of “positive feedback”.)
Because Dr Spencer’s present model has, on the granting of the quasi-second order linear viewpoint, two rate coefficients with negative feedback, I suppose it is capable of showing overshoot in time in response to some driver time-courses.
According to the viewpoint of the authors of the IPCC “forcings and feedbacks” formalism, properly speaking, the “climate sensitivity” is not a contributory moiety component of an ocean-atmosphere model, as it here appears to be, but instead is an overall quantity comprising at once both the ocean and the atmospheric responsiveness; this is a part of the expression of the IPCC “positive feedback” story. Christopher Game
Dr Spencer writes of “Chaos in the climate system.” Mathematically, that would require at least third order non-linear dynamics. Reasonable, but very hard to identify precisely from empirical data. Christopher Game
By “chaos” I mean the ability of a nonlinear dynamical system to enter different states with a small change in initial conditions. We already know this happens in the atmosphere on time scale of days to weeks.
But the ocean has nonlinear dynamical processes occurring on years to decades, if not centuries. If an ocean circulation change causes a 1% change in cloud cover, you have climate change.
“Chaos” does not necessarily imply a short time scale.
In climate modelling, there certainly seems to be alot missing for inputs and the weighting applied must be a tough challenge.
For instance, imagine a 105 day dual planetary rift (one at sea one on land or any combo)and lets say we lose 9 % of the tropospheric oxygen in the 1st 70 days and the after sunlight reappears following 600 days, we find snow cover where it usually isn’t (powerfull feedback)and luckily the rifting only lasted 105 days, because planetary “burps” and such might just bring on a good dose of glaciation.
Anywhoo, it seems our understanding of climate changing factors is quite limited and the possibilities seems very large, including changes in the tropospheric heights throughout the history of our planet and the dynamic circulations evident in the air and sea.
We have a long way to go before claiming indisputable causes for changes that we see and have seen.
Cheers:)
Roy-You mention this concern:
“it makes me wonder whether the MEI reconstruction from the early 20th Century is actually based upon global average temperatures in some way. If so, those early fluctuations that seem to agree so well are not independent.”
Well, not global average temps, but tropical pacific sea surface temperatures and sea level pressures are the main components used. This is because we have records of these variables going back longer than the cloud etc. data used for the regular MEI. Basically the assumption, I think, is that the relationships between these variables is at least quasi-stationary over time. This may not be entirely adequate, but it’s hard to say.
To the extent that global average temps depend on the temperatures of the small area associated with the tropical pacific, there is some lack of independence. I think this is a minor issue, however, compared with the more difficult question of whether any variable can be reliably and consistently linked to ENSO over long time scales. SOI and SST seem to de-couple from one another in the past, but this could easily be due to issues with nineteenth century barometers or a buckets issue or something. One of the advantages of the traditional MEI is that it uses several variables. The extended version basically reconstructs this with just two variables. It’s probably the best we can get, though.
For mathematicians, deterministic chaos means that the system (usually a system of ordinary differential equations) has a dynamical regime which never settles towards a fixed point, never settles towards a limit cycle, but traces out an endlessly varying path within a finite region. Such a trajectory often passes very close to paths about which indefinitely small displacements can lead to large ones, but such paths can easily exist in non-chaotic systems. Deterministic chaos mathematically requires at least third order non-linear dynamics and is unremarkable, to be expected as likely enough, amongst systems with third and higher order non-linear dynamics. Deterministic chaos is mathematically excluded for a second or first order system.
In an isolated system it is very hard to identify precisely the chaotic dynamics from merely empirical data; in a system driven by an external driver with chaotic dynamics of its own, such as the motions of the sun, planets, and moons according to Newton’s laws, precise identification of the internal chaotic dynamics will be even more difficult. That is not to suggest it will not be there; just hard to identify precisely. In practice, one must come up with sound general physical principles by which to understand systems like this. Christopher Game
“But I *do* agree that changes in the general circulation of the atmosphere can cause global average cloud changes, which in turn can cause warming or cooling.”
Which option do you prefer from the following?
i) As per Svensmark a quiet sun leads to more cosmic rays to produce more clouds and thus amplify cooling
ii) As per me a quiet sun leads to more negative polar vortices to produce more clouds and thus amplify cooling.
ii) Neither of the above, in which case what do you think happens?
I understand your idea, and I think either one is plausible. I suspect there are other possibilities as well.
“If as you suggest some cooling influence causes a cloud change which amplifies the cooling, that is positive feedback. So, warming from more CO2 would then cause fewer clouds, amplifying the warming (which is what the IPCC believes).”
If the sun is to be the cause of anything significant as regards a change in surface pressure distribution an amplification process is necessary because changes in TSI are so small. Clouds are likely integral to the amplification process.
However once the amplification has
occurred the system response is then highly negative and in my view operates by speeding up or slowing down the water cycle.
So if the sun causes an initial change that leads to positive feedback (whether towards warming or cooling)from cloudiness changes the system as a whole then applies a negative feedback to neutralise the effect as quickly as possible.
A warming effect is confronted by a faster water cycle sending energy faster to space.
A cooling effect is confronted by a slower water cycle sending energy slower to space.
The surface pressure disatribution is the visible manifestation of the speed of the water cycle at any given time.
Let me say what has been said by both Stephen and Dr. Spencer is right on.
But It is a combination of items from the quiet sun leading to more negative polar vortices which result in more clouds, to an increase in cosmic rays probably contributing to an increase of clouds, not to mention a study that was just released that has found that sulphuric acid particles from increase volcanic activity if they become large enough could behave as seeds for cloud formation.
Past history shows an increase of volcanic activity is associated with prolong solar minimum periods with active spurts from time to time, within the minimum.
Past history shows an increase of volcanic activity is associated with prolong solar minimum periods with active spurts from time to time, within the minimum.
any idea for references about that?
I wanted to add that I think the main reason for changes in cloud cover is due to the hypothesis which Steve just posted which is,that when the sun is less active the mid latitude jets move towards the equator and become more meridional. This then results in a more -AO ,which then results in addition to more clouds ,an increase in snow cover and precipitation ,which sets up a positive feedback for N.H. cooling, due to these items casuing earth’s albedo to increase.
Remember a slight change in albedo has a big impact on temperatures.
Another point Steve has brought up is at times the oceans can offset the solar effects. I think this is especially true at the start of a prolong solar minimum after an active solar period. The oceans prior to the solar minimum were receiving energy from a more active sun for many many decades, the oceans respond with large lag times to external changes of energy they receive. This means it takes time for the excess heat content of the oceans to subside once a change to the amount of energy the oceans is receiving changes.It does not happen over a short period of time.
Also geological activity was less during th active sun period, when a prolong solar minimum starts geological activity prior to the minimum is low thus a lag time will be required for the increae in geological activity through putting more so2 particles among other items into the stratopshere to have a more long lasting significant effect.
However over time these items instead of offsetting the solar effect will first beomce neutral and then supplement the solar effects.
I also think where the cloud cover increases during a prolong solar minimum is very important to what effects the change in cloud cover will have on earth’s climatic system.
“If I use a diffusivity much closer to what is traditionally used (1.2 x 10-4 m2/sec), then I have to reduce the model sensitivity to 0.9 deg. C for 2XCO2 in order to still match the near-surface warming trend, otherwise the model warms too much compared to observations.”
Roy, have you actually run your model with a sensitivity of 0.9 deg C for 2xCO2 to see how it might affect your fit? The time frame of particular interest is, of course, 1895 to 1935.
This looks like overfitting. Have you tried doing a match for just a portion of the data, 1880-1950, and then see how well this predicts the later portion?
Other considerations also need attention/studying, such as the ebb & flow of the Vanallen radiation belts which are fed from the electron, proton, gamma ray etc.. solar influences (primarily Earth-bound CME’s) and the declination/inclination of the planets magnetospheric distribution/position. The chemical changes/distribution along with the always changing stripping/recombination of particle constituents are not mapped very well and also in the infancy for studying.
The upper troposhpere and spheres beyond are can only be investigated from satellite sensors and through ground based sounding techniques and have just touched upon the science.
All The Best!
What I meant to suggest was to run your model with a climate sensitivity of 0.9 deg C for 2xCO2 along with the traditional diffusity of 1.2×10^-4 m^2/s in order to retain the near-surface characteristics while perhaps changing the forcing-feedback-diffusion model fit to HadSST2. I’m not sure the fit would be affected at all by this, but if it is improved, this would argue for the lower climate sensitivity. If it is a worse fit, this would be evidence to support a (very slightly) higher climate sensitivity of 1.1 deg C for 2xCO2.
Roy: I suspect you might get a better match in the early part of the long-term data if you were to use HADISST for your Global and NINO3.4 SST anomaly data.
Certainly the impetus for cloud seeding is important or maybe as important as the color of the earths surface (white,green,brown,black etc..) over time periods where the sun is free to shine. The high altitudes, ice-caps and mountain ranges in the northern/southern hemispheres undergo dramatic changes over time with varying amounts of snow cover and tinting. There’s something to be said of the reflective and absorbtive properties and the influences that bring snow or lack thereof. We periodically see walls of water (lakesize) break from glaciers, with the most recent ocurring last January 6th from the Gakona glacier. It was quite interesting that a steady drip-drip-drip within the glacier exploded when mother nature provided a chilly -57 F, which through expansion blew the ice holding wall apart. Anyway, here in the upper Copper River Basin of Alaska, where we’re 1st to see the warmup effects ocurring in the northern hemisphere and last to see the cooldowns, life is interesting.
Also, Roy you have a great website – Keep up the “Top Notch” work.
Best Regards!
I honestly don’t know why you waste time with those GISS fudge factors.
Dr Spencer writes: “The important message here is that this simulation was done with a very low climate sensitivity, corresponding to only about 1/3 the warming rate the IPCC projects for the future in response to increasing atmospheric CO2.”
Above I wrote, but saw no reply to, the following: “According to the viewpoint of the authors of the IPCC “forcings and feedbacks” formalism, properly speaking, the “climate sensitivity” is not a contributory moiety component of an ocean-atmosphere model, as it here appears to be, but instead is an overall quantity comprising at once both the ocean and the atmospheric responsiveness; this is a part of the expression of the IPCC “positive feedback” story.”
How are we defining the “climate sensitivity here”?
Are we using the same definition as does the IPCC?
The IPCC definition includes the long-term effect, if any, of putative ocean heat storage; that putative heat storage might take many years to occur and is a main reason for the delay in reaching the long-term steady-state “equilibrium” “climate sensitivity”, the main element of the IPCC’s beloved “positive feedback”, the main characteristic of which is gradual accumulation of heat, which the IPCC loves to call “amplification”, contrary to the usual scientific meaning of the word.
Perhaps I misread Dr Spencer here, but as I read him, his definition of “climate sensitivity” does not include the putative long-term effect of ocean heat storage; rather, his “climate sensitivity” is an ingredient contributory component moiety used in the construction of what the IPCC would call “climate sensitivity”.
Does it make sense to ask Dr Spencer’s model for its long-term ocean-included “climate sensitivity” as per the IPCC definition? Christopher Game
SOLAR LAG TIMES
They are always different with each particular solar minimum and the lag times vary for different items.
Case in hand ,I think the atmospheric response to changes in solar activity, that is the atmospheric circulation becoming more meridional versus zonal, has a rather short lag time, in contrast to oceanic temperature changes that have a long lag time.
The oceanic temperature lag times, is what is working agaist the low solar activity as of now due to the accumulation of energy the oceans received prior to 2005. This will take time to work through the system.
The geological activity through more volcanic activity will evenually have a bigger impact on the climate as more so2 becomes concentrated in the atmosphere over time.
Solar changes versus PDO due to the sun’s motion about the center of mass of the solar system caused by the Jovian planets effect on the sun, which in turn could influence earth’s rotational rate if true, probably have a moderate lag time. Ian Mason has done much research in this area.
In summary the solar minimum is at work but it will take time due to the oceans slow response to all play out, but it will, and is playing out as I write this.
climate sensitivity (to CO2) is zero or negative according to peer-reviewed “model-independent estimate of climate response”.
Warmism and Luke-Warmism are massively holed, deep below the waterline. Soon may they sink!
Brian H-Your reading of that paper is wrong. They find a positive trend associated with increased CO2, according to their statistical model. This would correspond to a small but positive sensitivity to CO2, which also makes sense physically. They found that there is little to negative feedback, but that is related to sensitivity as 1/1-f and for all values that make physical sense it is positive. It also isn’t proof of much of anything, really. But contrary to what you say, there is no “massive hole” in lukewarmer’s views (I think “they” as a “group” are poorly defined so it is hard to speak of “their” views) there are however serious flaws in AGW alarmism and many so-called lukewarmers have played a role in identifying them and calling them out.
Dr. Spencer,
How do you feel about the “temperature errors” the ARGO
system experienced in 2005-2006? Have the oceans been
cooling or warming?
If any of you are interested, I just posted a critique of Roy’s latest modeling efforts, such as the one in this post.
http://bbickmore.wordpress.com/2011/07/26/just-put-the-model-down-roy/
Bickmore will never give up on his quest to prove human global man made warming is for real, despite all the evidence that says otherwise.
The blind leading the blind.