by Judith Curry
There is much hype and debate surrounding Spencer and Bradwell’s new paper “On the misdiagnosis of surface temperature feedbacks from variations in earth’s radiant energy balance.” So lets sort through all this.
On the misdiagnosis of surface temperature feedbacks from variations in earth’s radiant energy balance
Roy Spencer and William Braswell
Abstract: The sensitivity of the climate system to an imposed radiative imbalance remains the largest source of uncertainty in projections of future anthropogenic climate change. Here we present further evidence that this uncertainty from an observational perspective is largely due to the masking of the radiative feedback signal by internal radiative forcing, probably due to natural cloud variations. That these internal radiative forcings exist and likely corrupt feedback diagnosis is demonstrated with lag regression analysis of satellite and coupled climate model data, interpreted with a simple forcing-feedback model. While the satellite-based metrics for the period 2000–2010 depart substantially in the direction of lower climate sensitivity from those similarly computed from coupled climate models, we find that, with traditional methods, it is not possible to accurately quantify this discrepancy in terms of the feedbacks which determine climate sensitivity. It is concluded that atmospheric feedback diagnosis of the climate system remains an unsolved problem, due primarily to the inability to distinguish between radiative forcing and radiative feedback in satellite radiative budget observations.
Published in Remote Sensing, link to paper [here].
Spencer discusses the paper on his blog.
UAH Press Release
Pielke Sr has posted the UAH press release in its entirety at WUWT. The title of the press release is “Climate models get energy balance wrong, make too hot forecasts of global temperature.” Apart from the title, the press release doesn’t go too much further than the paper. The key quote in the press release IMO is this one:
“The main finding from this research is that there is no solution to the problem of measuring atmospheric feedback, due mostly to our inability to distinguish between radiative forcing and radiative feedback in our observations.”
I agree with this statement. However, if there is no solution to measuring feedback, I would say that SB are concluding too much from their analysis about feedback, sensitivity, and the performance of models.
Forbes.com
The article at Forbes.com is entitled “New NASA Data Blow Gaping Hole in Global Warming Alarmism.”. The word “alarmist computer models” is used 8 times in this article, if I have counted correctly. Dudes, it may be appropriate to use the word “alarmist” in some circumstances, but not as an adjective to describe a computer model. The piece concludes with:
When objective NASA satellite data, reported in a peer-reviewed scientific journal, show a “huge discrepancy” between alarmist climate models and real-world facts, climate scientists, the media and our elected officials would be wise to take notice. Whether or not they do so will tell us a great deal about how honest the purveyors of global warming alarmism truly are.
The author of the piece, James Taylor is a lawyer at the Heartland Institute. Dudes, it may be appropriate to use the word “alarmist” in some circumstances, but not as an adjective to describe a computer model. This does not help the Heartland Institute to be taken seriously in the climate debate, even by skeptics.
Climate change debunked? Not so fast
Livescience.com has an article entitled “Climate change debunked? Not so fast.” Some excerpts:
New research suggesting that cloud cover, not carbon dioxide, causes global warming is getting buzz in climate skeptic circles. But mainstream climate scientists dismissed the research as unrealistic and politically motivated.
“He’s taken an incorrect model, he’s tweaked it to match observations, but the conclusions you get from that are not correct,” Andrew Dessler, a professor of atmospheric sciences at Texas A&M University, said of Spencer’s new study.
The study finds a mismatch between the month-to-month variations in temperature and cloud cover in models versus the real world over the past 10 years, said Gavin Schmidt, a NASA Goddard climatologist. “What this mismatch is due to — data processing, errors in the data or real problems in the models — is completely unclear.”
“I cannot believe it got published,” said Kevin Trenberth, a senior scientist at the National Center for Atmospheric Research.
Several researchers expressed frustration that the study was attracting media attention.
“If you want to do a story then write one pointing to the ridiculousness of people jumping onto every random press release as if well-established science gets dismissed on a dime,” Schmidt said. “Climate sensitivity is not constrained by the last two decades of imperfect satellite data, but rather the paleoclimate record.”
“It makes the skeptics feel good, it irritates the mainstream climate science community, but by this point, the debate over climate policy has nothing to do with science,” Dessler said. “It’s essentially a debate over the role of government,” surrounding issues of freedom versus regulation.
Spencer himself is up front about the politics surrounding his work. In July, he wrote on his blog that his job “has helped save our economy from the economic ravages of out-of-control environmental extremism,” and said he viewed his role as protecting “the interests of the taxpayer.” When asked why his work failed to gain mainstream acceptance, Spencer cited funding as a motivation for climate change researchers to find problems with the environment.
Spencer strikes back
Spencer strikes back with this post on WUWT. A few excerpts:
Kevin Trenberth’s response to our paper, rather predictably, was: “I cannot believe it got published” Which when translated from IPCC-speak actually means, “Why didn’t I get the chance to deep-six Spencer’s paper, just like I’ve done with his other papers?”
Finally Gavin Schmidt claims that it’s the paleoclimate record that tells us how sensitive the climate system is, not the current satellite data. Oh, really? Then why have so many papers been published over the years trying to figure out how sensitive today’s climate system is? When scientists appeal to unfalsifiable theories of ancient events which we have virtually no data on, and ignore many years of detailed global satellite observations of today’s climate system, *I* think they are giving science a bad name.
Small cloud study renews climate rancor
Mail.com gets it right with this headline: “Small cloud study renews climate rancor.” Some excerpts:
Several mainstream climate scientists call the study’s conclusions off-base and overstated. Climate change skeptics, most of whom are not scientists, are touting the study, saying it blasts gaping holes in global warming theory and shows that future warming will be less than feared. The study in the journal Remote Sensing questions the accuracy of climate computer models and got attention when a lawyer for the conservative Heartland Institute wrote an opinion piece on it.
The author of the scientific study is Roy Spencer of the University of Alabama Huntsville, a prominent climate skeptic. But even he says some bloggers are overstating what the research found. Spencer’s study is based on satellite data from 2000 to 2010 and is one of a handful of studies he’s done that are part of an ongoing debate among a few scientists.
At least 10 climate scientists reached by The Associated Press found technical or theoretical faults with Spencer’s study or its conclusions. They criticized the short time period he studied and his failure to consider the effects of the ocean and other factors. They also note that the paper appears in a journal that mostly deals with the nuts-and-bolts of satellite data and not interpreting the climate.
“This is a very bad paper and is demonstrably wrong,” said Richard Somerville, a scientist at the Scripps Institution of Oceanography at the University of California San Diego. “It is getting a lot of attention only because of noise in the blogosphere.”
Kerry Emanuel of MIT, one of two scientists who said the study was good, said bloggers and others are misstating what Spencer found. Emanuel said this work was cautious and limited mostly to pointing out problems with forecasting heat feedback. He said what’s being written about Spencer’s study by nonscientists “has no basis in reality.”
Trenberth and Fasullo’s analysis
Trenberth and Fasullo were quick off the block with a post at RealClimate. Some of their main points:
The paper has been published in a journal called Remote sensing which is a fine journal for geographers, but it does not deal with atmospheric and climate science, and it is evident that this paper did not get an adequate peer review. It should not have been published.
The paper’s title “On the Misdiagnosis of Surface Temperature Feedbacks from Variations in Earth’s Radiant Energy Balance” is provocative and should have raised red flags with the editors. The basic material in the paper has very basic shortcomings because no statistical significance of results, error bars or uncertainties are given either in the figures or discussed in the text. Moreover the description of methods of what was done is not sufficient to be able to replicate results. As a first step, some quick checks have been made to see whether results can be replicated and we find some points of contention.
The main part of the TF analysis relates to the comparison of models and observations of the lagged relationship between monthly surface temperature anomalies and net radiative flux anomalies. TF attempt to reproduce S&B’s results, and make a key point that the comparison depends critically on which climate models you select. If you select climate models that do not simulate El Nino’s very well, you will get a poor comparison, whereas you get a good comparison using the best models.
To help interpret the results, Spencer uses a simple model. But the simple model used by Spencer is too simple. We have already rebutted Lindzen’s work on exactly this point. The clouds respond to ENSO, not the other way round [see: Trenberth, K. E., J. T. Fasullo, C. O’Dell, and T. Wong, 2010: Relationships between tropical sea surface temperatures and top-of-atmosphere radiation. Geophys. Res. Lett., 37, L03702, doi:10.1029/2009GL042314.] During ENSO there is a major uptake of heat by the ocean during the La Niña phase and the heat is moved around and stored in the ocean in the tropical western Pacific, setting the stage for the next El Niño, as which point it is redistributed across the tropical Pacific. The ocean cools as the atmosphere responds with characteristic El Niño weather patterns forced from the region that influence weather patterns world wide. Ocean dynamics play a major role in moving heat around, and atmosphere-ocean interaction is a key to the ENSO cycle. None of those processes are included in the Spencer model.
Even so, the Spencer interpretation has no merit. The interannual global temperature variations were not radiatively forced, as claimed for the 2000s, and therefore cannot be used to say anything about climate sensitivity.
JC comments
I was one of the scientists contacted by the AP, they caught me when I was busy so I did a quick read, here was my response:
The paper itself makes the following point, which is reiterated in the press release: “The main finding from this research is that there is no solution to the problem of measuring atmospheric feedback, due mostly to our inability to distinguish between radiative forcing and radiative feedback in our observations.” This point has been made by others, including myself. Our understanding of feedbacks comes primarily from theory and models; diagnosing feedbacks from observations requires many simplifying assumptions.
The paper is of interest primarily in the context of comparing models with observations in terms of certain metrics, such as the lagged relationship between monthly surface temperature anomalies and net radiative flux anomalies. The models clearly show discrepancies with the observed relationships.
Also, it needs to be understood that given the short period of their data set, Spencer and Braswell are looking only at fast feedback processes associated with clouds (not the longer feedbacks associated with oceans and ice sheets). How to translate all of this into a conclusion that climate models are producing incorrect sensitivity to greenhouse warming is not at all clear.
The paper makes a useful contribution, but in the end they make the same error in interpretation that they accuse others of making. In my opinion it is not correct to infer from their analysis that global temperature variations were largely radiatively forced.
The complexity of the interaction between natural internal variability, surface temperature, clouds and radiative forcing are not adequately sorted out in terms of causal mechanisms to justify such a conclusion, in my opinion.
JC conclusions
Challenging climate models with observations is extremely important. I disagree with Gavin Schmidt that paloeclimate data is more robust than the satellite data., but both data sets are essential in this endeavor. So the line of research undertaken by Spencer and Braswell and Lindzen and Choi is an important one, and the issue of better understanding the role of clouds in climate is one of the highest priorities in climate research. But drawing inferences from such studies regarding feedbacks and sensitivity invariably leads to disputes because of the simplicity of the models and assumptions that are used and the fundamental fact that diagnosing feedbacks in the complex climate system can’t really be done. At best you can identify metrics such as the change of radiative flux with surface temperature, and compare observations and models.
TF’s analysis points out some significant flaws in S&B’s paper, and my analysis puts into perspective the relatively limited kinds of conclusions that can be drawn from such an analysis.
So should the paper have been published? I would say yes, although the reviewers and editors should have insisted on more information regarding the climate model simulations that were actually used in their analysis. Was the journal Remote Sensing remiss here? Well no more so than PNAS has been in some recent publications. Remote Sensing is a new open access journal; the only climatologist that I spotted on their editorial board is Toby Carlson. Remote Sensing is a plausible journal to have published this paper, and it seems that Spencer wanted to avoid the possibility of reviews by Dressler and Trenberth. If Roy Spencer didn’t make provocative political statements, this paper would not receive MSM attention and Dessler and Trenberth would probably be less motivated to spend time criticizing his research and wouldn’t be invited by the media to comment on it.
Trying to keep papers from being published isn’t useful (although a good editorial process is extremely useful), and on this particular topic (clouds and climate, comparing models with observations) we need more papers, not fewer. Science proceeds by putting ideas and analyses out there for other scientists to consider and rebut. Add a dose of politics into this, and you exacerbate scientific rivalries into media flame wars. So lets douse the flames and discuss the science.
Moderation note: this is a technical thread and comments will be moderated for relevance. Your general comments should be made on the week in review thread.
Not being a modeler, I wasn’t qualified to test the models to see if they justified the Spencer/Braswell (SB-11) assertion that the models seriously underestimate the heat loss to space their paper describes. However, Trenberth and Fasullo ran several models (see their RC post), and found that some of them yielded results similar to SB-11 while others were discrepant. A model that was skillful at simulating ENSO performed better than one with less skill. It was also characterized by high climate sensitivity to CO2.
I came across those model runs after already having read the paper to make my own judgments. The first problem I encountered was the lack of detail in describing the methodology. Many previous climate papers have been criticized for lack of full detail, but this paper is skimpier than even those. We really don’t know how much variability exists around the data points they describe nor does the paper describe in detail exactly what statistical methods they used to derive their correlations and with what confidence. If you look at Figure 2b, which is at the heart of the paper, it’s clear that even small inaccuracies in timing would lead to large differences in the estimated temperature/flux relationships. Their model is also arbitrary in assumptions about mixed layer depth, which is an important element in evaluating the rate at which energy enters the oceans. These flaws, of course, don’t tell us whether the SB-11 interpretations are right or wrong, but only that they have not been substantiated by the paper.
The uncertainties extend beyond the lack of detail, however. The authors examined global or semi-global temperatures in relation to what were primarily ENSO changes arising regionally in the tropical Pacific. Because of large scale lags in heat redistribution during ENSO fluctuations, it is hard to know how to interpret temporal relationships between flux changes arising in the tropics and the timing of temperature changes elsewhere. Indeed, ENSO phenomena during the evaluated interval included multiple El Nino and La Nina events. Each of these involves a complex chronology of energy exchanges. For example, during an El Nino, the ocean gives up heat to the atmosphere, the surface warms, and as a result of the Planck Response (via the Stefan-Boltzmann relationship), the atmosphere loses heat to space. However, the warmed surface and atmosphere can also respond with an increase in water vapor through the Clausius-Clapeyron relationship, leading to atmospheric heat gain via the greenhouse effect of the water vapor – a positive feedback. In turn, an El Nino-associated increase in regional cloud cover can increase planetary albedo with a consequent heat loss – a negative feedback. Because the timing of each of these responses will differ, conclusions about net atmospheric heat loss or gain are critically dependent on the interval separating observed temperature and flux changes. Furthermore, if a La Nina has followed an EI Nino, is a increase in negative heat flux (i.e., loss to space) a negative feedback consequence of the El Nino or a positive feedback on La Nina cooling? It would require sophisticated modeling and statistical analysis to distinguish these possibilities and the very different implications they involve. Appropriately, SB-11 hedge on the interpretation of their data as a quantitative measure of feedback strength.
The problem appears to be compounded by the magnitude of anomalies depicted in SB-11 Figure 2. Net flux anomalies in general vary over a range of about 1 W/m^2, with an occasional further 0.5 W/m^2 deviation. However, the surface temperature anomalies vary over a range of about 0.4 C. It is hard to see how the flux changes of that modest magnitude could be translated almost immediately into such large global surface temperature changes. This strongly implies that the flux changes are unimportant as causes of the ENSO-related temperature changes and instead reinforces generally accepted interpretation of the latter as internally generated rather than forced. Flux changes as a consequence of the temperature change are more plausible, and would lend support to a positive feedback scenario.
Having said all this, I can’t exclude the possibility that radiative losses to space are fairly high during the fluctuations observed, as suggested by SB-11. What to make of this is a different matter. As described above, a climate model with high CO2 climate sensitivity can duplicate these results. In addition, however, it’s important to realize that there is no necessary relationship between short term fluctuations, arising regionally, and involving (as ENSO does) warming or cooling originating in the ocean and imposed on the atmosphere, as opposed to long term changes distributed globally, and imposing a temperature change originating in the atmosphere on a previously unwarmed or uncooled ocean, as occurs with CO2 forcing. Convection is known to be different, changes in relative humidity and cloud cover are known to differ, and so the forcing/feedback relationship cannot possibly be the same The only question is whether the differences are large or small, and the current paper can’t tell us that. It therefore is uninformative regarding climate sensitivity to changes in CO2.
Finally, the paper is already being strongly criticized for requiring ENSO changes in the ocean to be significantly mediated by antecedent cloud variations with no obvious mechanism for the variations – the implication being that cloud fluctuations help cause ENSO. I agree that this would appear to conflict with much of what we understand about ENSO mechanisms, and instead substitute a cloud-based initiation process of unknown cause. It also appears inconsistent with the mismatch between the magnitude of the temperature change and a flux change that seems far too small to be the principal cause. On the other hand, I would like to remain open-minded about this possibility, however remote it seems. If correct, it would help to justify what Spencer and Braswell have previously asserted to be a negative feedback for ENSO events. Cloud-initiated ENSO would not, however, affect estimates of multidecadal climate sensitivity to CO2.
Thanks Fred, your comments provide a good start for this thread.
I notice that below, there is some extensive discussion of the effects of ENSO on cloud type and amount that goes beyond the simplistic example in my above comment. While there is agreement that clouds probably do not drive ENSO, their responses appear to be complex and heterogeneous, precluding a very precise estimate of their net feedback potential (although with suggestions of a net positive feedback). An informative analysis of the heterogeneity, based on multiple ENSO events, from 1955 through 1996 appears in a J. Climate paper by Park and Leovy on Marine Low Cloud Anomalies Associated with ENSO.
If we are speaking about ENSO – it is more precise to say that low cloud cover in the eastern Pacific is negatively correlated with sea surface temperature.
The change in cloud radiative forcing was ~ 2W/m^2 increase in SW up after 1998 – from ‘Earthshine’ and satellite observations. From COADS observations there was a similar decrease in the 1970’s at the time of the 1976/77 “Great Pacific Climate Shift’ to El Niño dominant conditions.
From the 1980’s to the 1990’s there was a decrease in SW up of 2.4 W/m^2 in ERBE and 2.1 W/m^2 in ISCCP-FD
CERES shows interannular change that is up to 2 W/m^2 – a significant variation and one that, ‘it is uncontroversial to say’, is poorly modelled.
ENSO cloud radiative feedback dwarfs the effects of greenhouse gases over the recent past. This is the mechanism that replaces CO2 as the major cause of recent warming.
“ENSO cloud radiative feedback dwarfs the effects of greenhouse gases over the recent past. This is the mechanism that replaces CO2 as the major cause of recent warming.”
Robert – I can’t disagree that during strong ENSO events, the radiative and temperature effects exceed those of long term forcing from rising CO2. However, ENSO changes tend to move in both directions, whereas responses to CO2 will always be in the direction of the CO2 concentration change. During El Ninos, the warming will “dwarf” GHG warming – temporarily, and during La Ninas, the cooling will outweigh GHG warming, but again temporarily, and so the long term average effects won’t reflect what is happening during one of these events, nor can they be said to dwarf long term CO2 effects. In that regard, what are you referring to when you mention “recent warming”? The most recent major ENSO event has been a La Nina, from which we appear to be emerging. That will have its effects, but can we really say that they “replace” the effects of CO2, as though to imply that CO2 has somehow stopped exerting greenhouse effects? Would the term “mask” be a better term than “replace” to describe the effects of ENSO on CO2-mediated trends?
Recent warming refers – as it always has – to the last few decades of the 20th century.
The actual period of warming coincides with the last warm phase of the Pacific Decadal Variation – 1976 to 1998. It is a period with a cool PDO and a higher frequency and intensity of El NIno.
You can see this visually in Claus Wolter’s Multi-variate ENSO index – http://www.esrl.noaa.gov/psd/enso/mei/
ENSO varies on these 20 to 40 year periodicities in frequency and intensity as well as between states on a 2 to 7 year periodicity. We have data and supporting observations – which I have gone into above. So if use the ISCCP-FD – the global dataset between the 1980’s and 1990’s – we are looking at a net 2.0 W/m^2 warming. It is possible that about 0.6 W/m^2 was the anthropogenic CO2 etc component. So ENSO cloud feedback as the major cause of warming – it replaces CO2 in the sense of taking the top spot.
It presumes that the numbers are right – after much fiddling with the data. But there are a couple of sources of satellite data – there is ‘Earthshine’ data and there are surface observations – all very consistent.
CERES confirms the scale of ENSO variability – but as I say ENSO varies over decades as well. ENSO is not a simple oscillation – technically it is non-stationary and non-Gaussian. This simple way of understanding that is that the frequency and intensity of ENSO is modulated at all timescales – and with little certainty of its future evolution. .
There is an 11,000 year reconstruction at Fig 5 here – http://www.clim-past.net/6/525/2010/cp-6-525-2010.pdf
‘The actual period of warming coincides with the last warm phase of the Pacific Decadal Variation – 1976 to 1998. It is a period with a cool PDO and a higher frequency and intensity of El NIno.’
It is a period with a warm PDO…
“However, ENSO changes tend to move in both directions”
The idea that La Nina is simply the opposite to El Nino or that ENSO indice describe a single phenomenon seems flawed.e.g.
http://staff.washington.edu/chiodi/posters/ChiodiHarrisonAGUFall2008Poster.pdf
There are several other independant approaches to the study of ENSO that seem to be coming to a simolar conclusion.
HR – I don’t think the issue is whether El Nino and La Nina are exact opposites – they are not – but whether their net effects average out globally over the long term. In the absence of forcing from some other source, that is probably the case, or at least there is neither good evidence nor a principle of physics that would suggest the opposite.
For some long term numbers,see Figure 1.3 (page S24) of the BAMS document on State of the Climate in 2010 – particularly the NINO3.4 index.
Fred,
thanks for parsing through some of the details of this issue.
One thing that Spencer has been harping on is the fact that BOTH models with high and low sensitivity can qualitatively reproduce past warming, so the interpretation of that fact as a metric can be confusing.
Having thought about what they’re analysis really shows, however, I think the actual consistent conclusion worth noting is that while there may be a long term trend in the data, short time fluctuations can dominate the signal for a time period between a few to 10’s of years. Spencer hopes this means that the last 30-40 years of satellite observed warming are due to these fluctuations. Unfortunately, it ALSO means that the downturn in temperatures in the last 10-15 years could be a strong downward fluctuation on top of an increasing long term trend.
As Judith has pointed out above, to only consider one of these possibilities and not the other is not good science at this point. If we can take away the point that the comparison between observations and models is difficult, then the most we can say now is that we have observations.
maxwell – if the IPCC takes away that point it might as well close up shop.
Maxwell- As I interpret the paper, “short term” refers to months, not years. The data involved values recorded over course of a decade, but the correlations related fluctuations in flux to those in temperature on a scale of months..
Fred,
it very well may be the case that paper explicitly explores the ‘short time’ fluctuations in one decade of data because that’s really the only satellite Spencer will use at this point as he knows the weaknesses of past instruments quite well.
His larger narrative (for the past few years) concerns longer time fluctuations like the PDO. The time period of PDO oscillations is on the order of many years to a decade. My guess is that Spencer will use the ‘shorter time’ fluctuations of this and more recent papers and expand the argument over the coming years to PDO and other longer time period ‘internal’ processes as the satellite data sets grow.
I was more speaking of his larger narrative, as I have a tall stack of other papers that are occupying all of my time at the moment. Sorry for any confusion concerning that point and thanks again for expounding on the main points of this most recent paper.
‘Much work has been done on ENSO over the past few decades and
pretty much everyone agrees that it’s a stochastically triggered, coupled
dynamic mode of the atmosphere, ocean system. I’ve never seen any
suggestion that it’s triggered by clouds. To the extent that clouds amplify
ENSO, that’s the cloud feedback,and that is what I measure in my paper.’
Andy Dessler in an email correspondence with Roy Spencer
Although it is mistakenly suggested at realclimate that S&B(11) imply this.
‘It is not controversial to state that climate models are deficient in terms of tropical variability in the atmosphere on many timescales [Lin et al., 2006; Lin, 2007] and a more realistic simulation of ENSO events in coupled simulations remains a high priority for model developers. During El Niño,
the warming of the tropical eastern Pacific and associated changes in the Walker circulation, atmospheric stability, and winds lead to decreases in stratocumulus clouds, increased solar radiation at the surface, and an enhanced warming..’
http://www.cgd.ucar.edu/cas/Trenberth/trenberth.papers/TFOW_LC_GRL2010_GL042314.pdf
‘Even so, the Spencer interpretation has no merit. The interannual global temperature variations were not radiatively forced, as claimed for the 2000s, and therefore cannot be used to say anything about climate sensitivity. Clouds are not a forcing of the climate system (except for the small portion related to human related aerosol effects, which have a small effect on clouds). Clouds mainly occur because of weather systems (e.g., warm air rises and produces convection, and so on); they do not cause the weather systems. Clouds may provide feedbacks on the weather systems. Spencer has made this error of confounding forcing and feedback before and it leads to a misinterpretation of his results.’ T & F realclimate
Trenberth and Fasula are contradicting themselves. It is sometimes a little difficult to believe that science is not being corrupted in this debate.
Nice catch!
Robert – I find no contradiction in the quoted sections from TF. If the clouds are responding to changes originating in oceans (e.g., reduced low clouds in the Eastern tropical Pacific, they are not a forcing and the temperature variations are therefore not radiatively forced, as stated by TF.
Indeed, if the responses of clouds to ENSO were a uniform phenomenon, the above would indicate a significant positive cloud feedback for ENSO – e.g., El Nino warming causes reduced low clouds, causing reduced albedo and even more warming. In fact, the cloud responses tend to vary by area and probably by the phase of either an El Nino or La Nina, and so the feedback estimate may be more complicated – for example, increased low clouds have been observed in the western Pacific during El Nino, although their quantitative importance is uncertain. On balance, a net overall positive feedback seems plausible, which would contradict Spencer/Braswell conclusions.
Cloud feedback estimates for CO2 forcing differ from those associated with ENSO, but in the case of CO2, all or almost all the models estimate a positive cloud feedback. In each case, the LW feedback is positive. The models differ, however, regarding SW feedback, which some find negative and others positive. In the latter case, the estimates are consistent with observational data showing a trend, over more than a decade, of reduced low cloud cover and an increased high cloud/low cloud ratio. The data also show significant shorter term fluctuations of uncertain relevance to long term forcings.
‘…and winds lead to decreases in stratocumulus clouds, increased solar radiation at the surface, and an enhanced warming..’
They are an ENSO feedback and as much a radiative forcing as water vapour.
As for the rest – show me the numbers…
Fred: If the clouds are responding to changes originating in oceans (e.g., reduced low clouds in the Eastern tropical Pacific, they are not a forcing and the temperature variations are therefore not radiatively forced, as stated by TF
I think one of the problems is in the language choices of climate scientists. They use terms like “radiative forcing” in a non-standard and not necessarily consistent fashion. However, it’s interesting that this nit pick on TF stands in opposition to common language use within that community (and is still at odds with its use in system theory)… just google “cloud forcings” if you want to argue whether climatologists regard this as a feedback or a forcing.
Anyway, if the cloud are responding to changes originating in ocean, and if the cloud’s change the Earth’s effective albedo as a result, there would still be a net modulation of solar forcing. In this case, the changes in the cloud patterns would be a form of parametric forcing, however, it would be one associated with internal variability of the system.
On a side note, whether something is internal or external is merely a question of where you draw the boundary for internal versus external, and it usually isn’t done arbitrarily…for example you wouldn’t want to include solar physics in your Earth-based climate model, you’d generally separate it into its own “external” domain. In a similar vein, economic responses to climate change wouldn’t normally be included in the climate model, even if they were parametrically forcing climate change through increased CO2.
You could, in that case, use the changes in solar forcing associated with the change in cloud patterns to deduce something about climate sensitivity, and TF are just wrong in this case. Such an estimate can also be done without resorting to a full scale climate model, you’d just use measured cloud patterns (from e.g. satellites) to estimate changes in solar forcing, which should be usable in conjunction with changes in surface temperature to infer something about climate sensitivity.
Alex – I believe “unforced” generally refers to a temperature change not due to an external perturbation of the radiative balance. An increase in surface temperature due to an El Nino that shifts ocean heat to the surface and atmosphere is unforced in that sense, as far as we know, because it is internally generated and occurs without a prior change in the radiative balance at the TOA. I think it is equivalent to your use of “non-radiative”. It is at the crux of the argument made by SB-11, who suggest that much of it is instead “radiative”. If the temperature changes they document during ENSO events are not initiated by radiative flux changes, then the flux changes are feedback responses, and feedback is stronger (more positive) than if much of the flux change is an initial cause of temperature change. Some of this is spelled out in their equation 1, where a significant value for their term N(t) signifies a low sensitivity. They confuse the issue, however, by calling N(t) an “internal radiative forcing”, their own idiosyncratic terminology, but they mean a flux perturbation as an initiating event rather than a feedback, and suggest that random variations in clouds could cause that initiating perturbation.
My earlier point was that even if an initial temperature change is not caused by a flux change, flux changes may precede the bulk of temperature changes when most of the temperature change is part of the feedback response rather than the initiating event. To illustrate with an impossibly extreme example, consider an SST increase of 0.01 C occurring without a TOA flux change, but then mediating positive feedbacks that greatly reduce net outgoing flux so that the temperature rises an additional 0.5 C (50 times the original change). Here, the data will show that almost all the temperature change lagged rather than led the flux change, even though no external flux change had been imposed on the system. This should not be interpreted to mean that the flux responses to temperature are small, as would be true for a low climate sensitivity value., since in this example, the value of sensitivity would actually be infinity.
Carrick – My apologies. My response was intended for Alex Harvey below. I will reply to you shortly.
Carrick – Sorry for the switched replies. This one is a response to your comments.
I think you’re right that “radiative forcing” is a term not always used consistently, but I don’t think that’s a problem here. In this case, it refers to an imposed perturbation of the radiative balance at the top of the atmosphere (TOA) that invokes a temperature response. This is the conventional use that is almost always meant. “Cloud forcing” is one of the rare exceptions, but it refers to the difference between a climate without or without a specific set of clouds and is not used to imply that clouds initiate a TOA flux perturbation. Cloud feedback is not the same as cloud forcing, although even the IPCC has at times confused the two.
You refer to cloud “modulation” of solar forcing, which may be legitimate, but clouds can’t “change” solar forcing because the term “forcing” specifically refers to changes in the TOA flux balance that occur before the climate system has had a chance to respond (except for stratospheric adjustments). Changes that occur in response to the imposed perturbation are feedbacks. A cloud change in response to changes in the ocean is also a feedback and can’t be called forcing in any commonly understood sense of the word within climatology.
TF use these terms in the commonly accepted sense, and I can see nothing wrong with their interpretations. Semantics aside, I see no problem with their understanding of the physics.
ENSO Cloud feedbacks result in a change in flux at TOA. Semantics in a narrative framework.
The discussion since your response Fred has started to confuse me a bit. SB’s simple model just leads to a rough estimate that 70% of the observation data response is due to “forcing” and 30% is due to “unforced variation”. I don’t see that as earth shattering as a skeptic, but modelers that assume less than 10% of the observation data change is due to “unforced variations” may.
Trenberth’s claim that models with more ENSO parameters better match the SB results is puzzling. SB list the six IPCC models they used pretty clearly and from what I have read so far I have no clue which models Trenberth is referring to, are they one in the middle so they are not part of the three most/least sensitive?
The SB paper is a rebuttal not an attempt to develop any theory other than Dressler missed something, “unforced variations”. The conversation on this blog turns to whether ENSO changes clouds or clouds may change ENSO when I think the real issue should be whether a climate oscillation other than ENSO changed clouds which changed ENSO, a significant forcing/feedback caused by internal variability, which is not included in the models. A pretty valid point since most of the models I have seen tend to start missing something between 2000 and 2005.
Dallas – The good match was with the MPI-Echam5 model, which has a high climate sensitivity.
I don’t believe SB produced evidence that 70% of the data represented a response to forcing. In fact, as I argued above, the magnitude of the temperature change would appear to be much too high for more than a small fraction to be a forcing response.
Forcing (at a minor level) is a possibility, but the SB data are also consistent with no forcing at all – at least from ENSO. Their main argument is a “lead/lag” principle – if a change in radiative flux precedes rather than follows a temperature change, then the temperature change must be a forced response. However, they haven’t shown such a lead/lag response. Rather, they show (or purport to show since we don’t have the statistics) that a significant flux change precedes the maximum temperature response. “Maximum temperature response”, however, is not the same as “temperature response”, and tells us little about whether flux changes preceded or followed initial temperature responses. It is plausible that the process was initiated by the beginnings of an unforced ENSO event (e.g., an El Nino warming ocean). Under such a scenario, the warming in turn will be followed by positive feedbacks that reduce net outward radiative flux. The temperature will rise globally to a maximum. If this maximum occurs after the positive feedback is already declining from its maximum (not implausible), the result will be the reported data even though no forcing was involved.
Dallas – Regarding a separate point, the modelers estimate all or nearly all the ENSO variability to be unforced.
Thanks Fred for your discussion.
RC notes:
Yet I find a bibliography by Paul D. Farrar with numerous papers on Galactic Cosmic Rays and ENSO:
Bibliography relating to cosmic rays – El Niño – clouds Some indicate that Solar variations modulate GCRs which influence clouds which impact albedo and climate.
While RC shows some evidence to counter Spencer & Brasswell, why should we not see RC’s claims as: begging the question, in light of such papers proc/con ENSO and galactic cosmic rays?
Spencer posts: Rise of the 1st Law Deniers
Trenberth & Fasullo object that another model “The MPI-Echam5 model replicates the observations very well.”
That appears to confirm Spencer’s objections that almost all IPCC models do NOT model ENSO well.
Trenberth & Fasullo summarize: “The bottom line is that there is NO merit whatsoever in this paper.” That appears to be rhetorical excess trying to divert attention from IPCC failures, and directly denying the temperature lag to radiative forcing that is an essential consequence of the First Law.
Fred,
Yeah, right.
Dr. Curry,
At least you are dealing with the paper seriously and professionally.
The reason, I guess the Forbes writer used the ‘alarmist’ language is to make it clear that skeptics are friggin’ sick and tired of AGW true believers denigrating and belittling those who dare question the consensus.
‘Alarmist’ is not an inaccurate way to describe much of what the public square has been subjected to the last 20+ years. That alarmism is nearly always wrong in any venue, and has been the cause of wasting tens of billions in climate science, needs to be addressed.
rather than “alarmist computer models”, “alarmists’ computer models”. perhaps it was just a typo.
In modern idiom, ‘alarmist computer models’ works just fine.
But if you like ‘alarmist’s computer models’, I am OK with that as well.
Are there any computer models which aren’t alarmist? If so, the adjective is useful in distinguishing which models he is referencing. If not, then the adjective may be redundant, but descriptive. In any event, I don’t see any way it hurts his credibility (especially in an area such as climate science where credibility wrecks are so common).
I think JC’s issue with describing a computer model as ‘alarmist’ is that it is a subjective quality, where computer models should only be considered to offer objective data.
What I think the author means to indicate by the use of the term is models that show large feedback, and thus produce a temperature estimate that a person would consider ‘alarming’.
No no no no no. Computer models do not offer objective data. Computer models are written by human beings. These people choose what to include in the model and what to ignore. They choose the values of the parameters that make up the model. They choose the initial conditions to start the model off.
As omnologos and hunter point out, the only problem with term “alarmist computer model” is a missing ‘s.
A short comment. SB11 shows El Ninos precede radiative cooling, as expected, but also that radiative warming precedes El Nino, somewhat unexpectedly. This is likely a flaw in the analysis of lead/lag correlations where they subtract an average radiative forcing to do the correlation. A situation where periodic El Nino warming was followed by radiative cooling would look like radiative warming leads El Nino even if the precedent radiative effect was zero, merely by subtracting the average. I would be skeptical.
It seems to me that Roy Spencer is playing by the “rules”; that is, he is trying to carry on a serious scientific discussion through the peer reviewed literature. This is one of a series of papers of his on this same subject. IIRC, it took Roy 2 years to get the first paper published. I think any comments such as we have seen very rapidly, are useful, but surely what is needed is for someone to debate this paper in the peer reviewed literature. Let us see what a peer reviewed rebuttal looks like.
I would also like to point out that there is very little, when it comes to climate sensitivity, which relates to observed data. So far as I can tell, the bulk of the “data” purporting to show that climate sensitivity is high, is based on the output of non-validated models. So, any attempt by Roy Spencer to try and take actual observed data, and show how it relates to climate sensitivity, ought to be welcomed by any true scientist. As Judith has, indeed, done. But the predictable knee-jerk reaction of the proponents of CAGW, is, to me, little more than hand-waving.
“Let us see what a peer reviewed rebuttal looks like.”
You don’t need a climate model to see how the rebuttals are going to go, do you really think the Team or the consensus is going second guess the settled nature of the science (their conclusion)? They can barely hold themselves back from ad hom attacks and heading for the margin of error of the tools and data at hand.
People looking to win with hard science evidence are sadly mistaken, the warmists knew this long ago and set about enforcing conclusions about “might” rather that explaining nuances. This is how it has survived vast model failures as well reinventing itself time and again when the last claim fails.
cwon14 writes “People looking to win with hard science evidence are sadly mistaken, the warmists knew this long ago and set about enforcing conclusions about “might” rather that explaining nuances. This is how it has survived vast model failures as well reinventing itself time and again when the last claim fails.”
I have no reason at all to doubt that you are absolutely correct. But if you are right, then it brings up the most important issue. Are we discussing science or politics? I believe we ought to be discussing science. And if so, then the attitude of the proponents of CAGW are absolutely reprehensible. This is where Dr. Judith Curry ought to be absolutely outraged at the lack of respect for science shown in this despicable attack on Dr. Roy Spencer’s paper.
This is why some of us despair about Dr. Curry’s attitude. She will come to the brink of behaving like a true scientist, and then pulls back when she needs most to condemn the actions of the proponents of CAGW.
Jim, I share your sadness. I’m back to my Albert Spear anology on the prior thread. Go take a look (and I realize it can painful) to a Hoi Polli left wing sites like the Huffington Post, Dr. Curry is all but a hated figure in the left’s lexicon. This is what happens when movements crumble, the mob hanging on want to kill not only opponents but any perceived weak link in the “cause”. So my inclination is to cut her slack but again is she a hero of reason? She only moderated in public from 2006, how reasoned could that be judged? The science (it’s close to a null theory at any point) with peers were being blacklisted, defunded, slandered and the rest in plain sight. She closed ranks with her base political culture. Given the special knowledge of the topic how can anyone support cap and trade yet she gave money to Obama. As far as I’m concerned that was like supporting the funding of trains to Dachua while claiming you wouldn’t know what was going to happen there.
Regardless, your point is well made while I’m sure mine will be considered rude. What are we to do with the science enclave that is ready to jump the shark politically at any point over a wide range to topics? AGW is in the Hall of Fame of political mania with Eugenics and Lysenkoism are likely only chapters along the way. If we want deeper reforms we should look at the linkage of Keynesianism to the current government funded science cartels;
http://joannenova.com.au/2009/07/massive-climate-funding-exposed/
Since the science was always so weak the level of social insecurity is only that much greater. You don’t need a science background to figure this out. AGW without policy actions would have remained just a research and academic football, one of many. When they got onthe road to carbon rationing and taxes it became primetime. This could never make it on primetime. I’d like to see Dr. Curry evolve further if she takes the issues seriously she can already see what a full recant would mean;
http://www.nytimes.com/1996/06/18/science/scientist-work-richard-s-lindzen-skeptic-asks-it-getting-hotter-it-just-computer.html?pagewanted=all
Look at how soft the wording was then, soon would come death threats from mobs. Luke warm has social benefits and both sides will court her opinions all the more.
Both the arguments presented on this paper and the http://www.knmi.nl/cms/content/99641/tracing_the_upper_oceans_missing_heat writings of Katsman and van Oldenborgh bring up issues that should be answerable better using more detailed data from most relevant geographic areas and from different depths in the oceans. The arguments rely presently on models at a level of detail that should not be beyond reach empirically. That approach might lead to improved understanding much faster than waiting for long enough periods of global data.
That’s at least the impression I got from reading these two paper today.
“The main finding from this research is that there is no solution to the problem of measuring atmospheric feedback, due mostly to our inability to distinguish between radiative forcing and radiative feedback in our observations.”
Right. IPCC ‘GHG’ scenarios are laughable.
Is the forcing/feedback nomenclature standard in (atmospheric) physics?
Imagine if the realclimate boys worked that hard to check everyone’s work. It might even resemble realscience.
stan,
I always wonder that! Where’s the critical and scientific scrutiny?
It makes perfect sense for SB to publish in Remote Sensing. These scientists study satellite data, it’s what they do, they think about it every day, they analyze remote sensing data. Their satellites look at temperatures and energy flows and SB analyses them. Therefore it stands to reason that they have a very different perspective on climate. Climate research and forecasting is a highly multi-disciplinary field, and SB as remote sensing experts contributor as they can. Dr. Curry’s professional approach to the SB paper should be the norm, as opposed to the kind of coverage it attracted at SlashDot, or the kind of thinking that insists that this paper should never have been published. Keeping opposing views out of the scientific literature was exposed by ClimateGate, and SB has experience it first hand.
Technically, how are we to interpret this comment, JC?
“…the fundamental fact that diagnosing feedbacks in the complex climate system can’t really be done.”
She’s right.
Diagnosing feedbacks in a relatively simple system is usually impossible if you have to treat the system as a black box. This is because you can get the same system response from both feedback and non-feedback systems.
Look up indeterminate or indeterminate systems
“The main finding from this research is that there is no solution to the problem of measuring atmospheric feedback, due mostly to our inability to distinguish between radiative forcing and radiative feedback in our observations”
I think this is correct if SB-11 are referring to direct measurements. Ironically, as far as I know, they and Lindzen have been the main participants in attempts to do that. Estimates from most other sources utilize indirect methods. For example, Hansen attempts to determine the temperature response to a quantified forcing (i.e., climate sensitivity), which yields feedbacks by comparison with the “no-feedback” (Planck Response only) temperature response. The model-based methods in AR4 Chapter 9 similarly attempt to estimate climate sensitivity rather than feedbacks per se. The methods described in Chapter 8 attempt to estimate individual feedbacks (water vapor, lapse rate, ice-albedo, and clouds) and then combine their interactions to estimate net feedback. All of these methods are characterized by substantial uncertainty, but it would be inaccurate to describe them as providing “no solution”; rather they offer approximate solutions based on indirect methods rather than direct measurement.
Is that ironical, or rather an indirect acceptance that the approach that Spencer has been studying for long has turned out to be too difficult, at least for now?
He has been a creative scientist, but the results have often not been as strong as he has first presented. He may have been specifically looking for evidence supporting low climate sensitivity, but it’s good that someone is trying also that.
He has published his results and tried to defend them. He is entitled to do that and so are others to criticize his work. His papers may get more publicity than they reserve, but so do some other papers as well.
The model-based methods in AR4 Chapter 9 similarly attempt to estimate climate sensitivity rather than feedbacks per se. The methods described in Chapter 8 attempt to estimate individual feedbacks (water vapor, lapse rate, ice-albedo, and clouds) and then combine their interactions to estimate net feedback. All of these methods are characterized by substantial uncertainty, but it would be inaccurate to describe them as providing “no solution”; rather they offer approximate solutions based on indirect methods rather than direct measurement.
Indeed no solution is not the answer,multiple ie infinite solutions to the same problem exist.
This is well known as we are in the realm of arbitary axioms ie Diophantine equations where as has been proven the solutions breed like rabbits (exponentially) ie Fibonacci numbers.
The problem is well known eg Hilberts 10th. As the positive affirmation is impossible we can only resort to negative conclusions (negative proofs are very dooable) ie they are unsolvable.
http://www.scholarpedia.org/article/Matiyasevich_theorem
If what is meant is an infinite range of solutions, I would disagree. I don’t believe that the existence of an infinite multiplicity of solutions is compatible with climate observations and basic physics principles that include, as one example, the predicted and observed increase in atmospheric water vapor in a warmer atmosphere. Once physics is involved, and not exclusively mathematics, the solutions are still bounded by relatively large margins, but they are legitimate solutions within those margins.. In fact, some of that remaining uncertainty is simply measurement uncertainty and is reducible by more powerful and accurate measurement techniques.
‘Atmospheric and oceanic computational simulation models often successfully depict chaotic space–time patterns, flow phenomena, dynamical balances, and equilibrium distributions that mimic nature. This success is accomplished through necessary but nonunique choices for discrete algorithms, parameterizations, and coupled contributing processes that introduce structural instability into the model. Therefore, we should expect a degree of irreducible imprecision in quantitative correspondences with nature, even with plausibly formulated models and careful calibration (tuning) to several empirical measures. Where precision is an issue (e.g., in a climate forecast), only simulation ensembles made across systematically designed model families allow an estimate of the level of relevant irreducible imprecision.’
Until we see the results of ‘systematically designed model families’ we have no idea what the level of imprecision in models is.
‘Sensitive dependence and structural instability are humbling twin properties for chaotic dynamical systems, indicating limits about which kinds of questions are theoretically answerable. They echo other famous limitations on scientist’s expectations, namely the undecidability of some propositions within axiomatic mathematical systems (Gödel’s theorem) and the uncomputability of some algorithms due to excessive size of the calculation’
McWilliams 2007 – Irreducible Imprecision in Atmospheric and Oceanic Simulations – http://www.pnas.org/content/104/21/8709.full.pdf+html
So Fred, you obviously live on a planet where there’s no such thing as “weather” and “weathermen”. Which planet do you live on?
“Alarmist models” ie models that give alarming projections of future climate. So it is a perfectly good phrase to describe the class of models that this paper challenges with real data.
As time passes and we get more empirical data from satellites measuring energy in and out of the earth system so the boundaries of AGW will become clearer.
At present combating the extreme claims of warmist climate commentators seems to becoming like shooting fish in a barrel.
Since Prof Trenberth expressed his concern at the travesty relating to missing ocean energy the traffic all seems to be one way. Lindzen and Choi , Loehle and Scafetta, Spencer, continued flat lining of global average temperature, still no tropical tropospheric hot spot, all empirical and all indicating that the alarmist models are plain wrong.
With all the money going into warmist research we are not seeing much return of empirical substance when it comes to keeping the scare agenda moving forward.
Ditto Ormand.
Why can’t we all agree that the science and models ‘concerning the assumed feedbacks’ are still in their infancy.
JC?
Maybe it is not the models that are in their infancy, but the modellers?
As was put previously, maybe homo-sapiens has reached adolescence and therefore, obviously, we know everything?
Both models and modelers are in their infancy. There are millions of factors involved in climate, a few thhousands of factors in the models and in the modelers minds making them unreliable results.
Trying to model the climate on a modern computer is kind of like trying to build a skyscraper with stone tools.
Spencer and Brasswell (2011) report that sensitivity is not calculable from CERES and surface temperature data because of ENSO cloud radiative feedbacks. A null result. Since ‘much of the temperature variability during 2000–2010 was due to ENSO (Dessler 2010), we conclude that ENSO-related temperature variations are partly radiatively forced. We hypothesize that changes in the coupled ocean-atmosphere circulation during the El Niño and La Niña phases of ENSO cause differing changes in cloud cover, which then modulate the radiative balance of the climate system.’
The hypothesis is not original. It goes back to Zhu et al (2007), Burgmann et al (2008) and Clement et al (2009). The latter two using COADS surface as well as satellite observations – and focusing on decadal Pacific variation. Low level cloud forms over a cold ocean in a La Niña and dissipates over a warm ocean in an El Nino. The ENSO cloud feedback radiative effects are an order of magnitude greater then CO2 radiative forcing on a decadal basis. The planet is cooling for another decade or three as upwelling of frigid and nutrient rich water intensifies in the eastern Pacific. This is the mechanism that supplants CO2 as the major cause of recent global warming.
If the climate models are so unassailable – could one of the modeling experts succinctly explain how well the models describe the little ice age, and according to those models what key factors caused that event?
http://www.pages-igbp.org/download/docs/Gonzales-Rouco%20et%20al._2011-1(7-8).pdf
Thanks – this says that the models have major discrepancies with respect to the reconstructions, e.g. ‘The results presented here highlight major discrepancies between millennium simulations and reconstructions.’
Lots of chicken and egg queries raised here especially as to whether ENSO variability causes cloud changes or whether cloud changes cause ENSO variability.
The late 20th century was characterised by a run of strong El Ninos yet cloud cover DECREASED. Now despite the beginning of a negative PDO cloud cover is INCREASING. So that is the global scenario whatever happens locally or regionally.
El Nino increases surface air temperature to enhance convection. That enhanced convection then descends in the sub tropics to strengthen and widen the sub tropical high pressure cells for LESS clouds and MORE energy into the oceans.
In the process the mid latitude jets get pushed poleward and/or become more zonal for reduced total cloud cover globally.
However that is only half the story because at the same time the level of solar activity is affecting the vertical temperature profile of the atmosphere too but from the poles and can either supplement or offset the effect of ENSO on the size of the sub tropical high pressure cells and the latitudinal position of the jets.
In view of the data how can some keep asserting that El Nino produces more clouds and La Nina produces less clouds?
In my opinion the heat energy for El Nino accumulates initially because the ITCZ is always north of the equator (due to the current global landmass distribution) which creates an imbalance of solar shortwave input either side of the equator.
Periodically the imbalance becomes large enough to provoke an El Nino event as the system tries to move back towards thermal equilibrium either side of the equator.
In theory the release of ocean energy during an El nino should reduce ocean heat content for a net cooling of the entire system. However during the late 20th century ocean heat content rose despite the run of powerful El Ninos. Now, with more La Ninas the ocean heat content should in theory be recharging but it is not.
The reason is that El Nino actually increases the width of the subtropical high pressure cells pushing the jets poleward and allowing MORE sunlight into the oceans.
However that is not always enough to offset the energy release. Another factor is required to achieve that.
That other factor is the level of solar activity. When the sun is active the polar vortex becomes more positive pulling the jets poleward which supplements the effect of the EL NIno allowing even more sunlight into the oceans. So it was during the late 20th century. Strong El Ninos combined with an active sun allowed the surface pressure distribution to change enough to allow sufficient shortwave into the oceans such that ocean heat content rose despite the strong El Ninos.
That is the only scenario whereby all the observations can be made to work in a coherent fashion and it would appear to be consistent with Dr. Spencer’s findings.
Mine is the only hypothesis that accommodates all that plus a cooling stratosphere when the sun is active and a warming stratosphere when the sun is inactive.
During the late 20th century warming trend the stratosphere was observed to cool and that was also supposed to be in accordance with AGW. However since the 90s that cooling has ceased and the stratospheric temperature trend is now one of slight warming:
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.”
ENSO has its origins in upwelling in the eastern Pacific – there is no suggestion anywhere that clouds cause ENSO.
Someone suggested it earlier in the thread.
As explained above ENSO arises because the ITCZ is always north of the equator then when an El Nino occurs the increased evaporation causes more convection and cloud in the ITCZ but the increase in descending air in the subtropics widens the high pressure cells pushing the jets poleward for reduced cloudiness globally.
However in my view that process is then modified by solar effects from above which change cloudiness first thereby altering sunshine into the oceans to change the energy available for ENSO.
Found it, Fred said this:
“Finally, the paper is already being strongly criticized for requiring ENSO changes in the ocean to be significantly mediated by antecedent cloud variations with no obvious mechanism for the variations – the implication being that cloud fluctuations help cause ENSO. I agree that this would appear to conflict with much of what we understand about ENSO mechanisms, and instead substitute a cloud-based initiation process of unknown cause. It also appears inconsistent with the mismatch between the magnitude of the temperature change and a flux change that seems far too small to be the principal cause. On the other hand, I would like to remain open-minded about this possibility, however remote it seems. If correct, it would help to justify what Spencer and Braswell have previously asserted to be a negative feedback for ENSO events. Cloud-initiated ENSO would not, however, affect estimates of multidecadal climate sensitivity to CO2.”
Now that would fit with the top down solar part of the equation because an active sun pulling the jets poleward would reduce global cloudiness for more energy into the oceans to enhance ENSO. So my proposal is that both processes apply and the outcome in terms of climate depends on the balance between them.
I think the top down solar effect is a sound mechanism for cloudiness changes initiating ENSO changes.
I have been into it in some detail here:
http://www.irishweatheronline.com/features-2/wilde-weather/the-sun-could-control-earths-temperature/290.html
Tropical Pacific sea level pressure, cloud cover, sea air temperature, sea surface temperature, meridional and zonal surface winds, thermocline depth, North and South Equatorial current strength, Equatorial Countercurrent strength, etc., are all closely coupled by the process of ENSO. I have never read a study that says cloud cover changes causes ENSO; they are simply an integral part of it.
I didn’t say cloud cover changes cause ENSO. I said that the cause of ENSO is the fact that the ITCZ is always north of the equator causing an input imbalance as explained above.
However if solar effects can change cloudiness that will influence the amount of energy entering the oceans and that would affect the energy available for the ENSO process.
Sorry, Stephen. My reply should have followed your July 30, 2011 at 5:32 pm comment. I was supplementing your opening “chicken-and-egg” paragraph, nothing more.
You wrote, “I said that the cause of ENSO is the fact that the ITCZ is always north of the equator causing an input imbalance as explained above.”
It’s not always north of the equator. There’s a seasonal component to the ITCZ location. It wanders north and south of the equator over the course of a year, accompanying the warmest waters.
ITCZ January – July:
http://en.wikipedia.org/wiki/File:ITCZ_january-july.png
Equator and Tropics:
http://ro.wikipedia.org/wiki/Fi%C8%99ier:Wendekreise.png
Sorry chaps, I should have said that the mean position of the ITCZ is always north of the equator:
http://www.theweatherprediction.com/habyhints2/453/
Chief Hydrologist is right to say that there is no suggestion that clouds cause ENSO. So far as I know, the specific cause of ENSO is unknown. It cannot be confidently concluded that some kind of cloud driver is not the specific initial cause of ENSO, but that is not the point. The point is not as to the specific initial cause of ENSO.
The point is that clouds form part of the mechanism of the effect of ENSO and as such they are part of the effect of the unknown initial specific dynamical driver of ENSO, quite likely to do with the motions of the solar system objects, and thus the relevant cloud effects are quite likely due entirely to external dynamical drivers, negligibly affected by feedbacks. This point seems to be missed by Chief Hydrologist. Christopher Game
The literature I have quoted suggests a cloud feedback negatively correlated to SST.
Although if you read carefully – it is suggested that it is linked through SAM and NAM to solar UV variability.
.
Chief Hydrologist says: “ENSO has its origins in upwelling in the eastern Pacific – there is no suggestion anywhere that clouds cause ENSO.”
Origins? I’m not sure I follow. The upwelling in the eastern equatorial Pacific is one of the sources of the waters that are carried west by the trade wind-driven North and South Equatorial Currents, warmed additionally by decreases in cloud cover and increases in Downward Shortwave Radiation during a La Niña event, “piled up” against Indonesia in the Pacific Warm Pool, and eventually serve as fuel for the next El Nino, but don’t forget the Humboldt and California Currents. They also feed the tropical Pacific.
But I agree with the latter part. Cloud cover is an integral part of ENSO, not a cause.
The problem with blogs – as I said elsewhere the Humboldt Current (which involves frigid and nutrient rich upwelling in the eastern Pacific) is the source of the cold tongue in ENSO. But for the Pacific decadal variation – we need to add upwelling in the north Pacific in the region of the PDO – that results in the typical V of the cool mode that we are in at the moment.
http://www.osdpd.noaa.gov/data/sst/anomaly/2011/anomnight.7.28.2011.gif
Stephen Wilde | July 30, 2011 at 5:32 pm | “[…] how can some keep asserting that El Nino produces more clouds and La Nina produces less clouds?”
In order to efficiently understand, just ask them “where?” if they have not been explicit.
During El Niño, the warming of the tropical eastern Pacific and associated changes in the Walker circulation, atmospheric stability, and winds lead to decreases in stratocumulus clouds, increased solar radiation at the surface, and an enhanced warming.’ Trenberth et al 2009
I heard not heard anyone saying otherwise.
“increased solar radiation at the surface”
Isn’t that true for most of the 20th century … there was more bright sunshine … hence global brightening/dimming/brightening again in synch with PDO.
Stephen Wilde says: “In view of the data how can some keep asserting that El Nino produces more clouds and La Nina produces less clouds?”
I will assume the “some” refers to my descriptions of ENSO. As Paul Vaughan noted, it depends on where we’re discussing cloud cover. Trenberth and Fasulo (2009) were discussing cloud cover in the tropics. My descriptions as part of my discussions of ENSO are typically of eastern tropical Pacific cloud cover.
http://i52.tinypic.com/9uq0zo.jpg
Thanks Bob but I was referring to Trenberth et al who actually talk about the regional cloud cover changes but seem to imply that the effect is global. I know you limit yourself to the ENSO region.
They often say that the increase in that regional cloudcover increases global albedo but in fact global albedo declines and I think the reason is that the poleward shifting of the mid latitude jets (or increased zonality) decreases global cloud cover more than it is increased by changes in the ENSO region.
Stephen, your regular exposition on this would benefit tremendously from the addition of latitude & season as factors. Anomalies are hydrologically inadequate (since hydrology is a function of absolutes). Tip: Variance should not be overlooked, as often happens in these discussions where there tends be excessive fixation on averages; flow is a function of spatiotemporal gradients, regardless of quantities like “global average”. Leroux (1993) puts it in simple terms (literally makes it as dead-simple as comparing pictures to “spot the differences”). Sidorenkov (2005) gives meaty expositions that might help some comparatively interpret Leroux’s (1993) pictures. [Links available here: http://wattsupwiththat.com/2011/04/10/solar-terrestrial-lunisolar-components-of-rate-of-change-of-length-of-day/ .]
Thanks Paul but I am considering changes beyond the normal range of seasonal variability. What interests me is the background trend that leads to longer term changes such as from MWP to LIA to date because that is what led us to the late 20th century warm scenario without the need for any CO2 involvement.
Also interesting is how that background trend affects or is affected by ENSO variability and the Pacific Multidecadal Oscillation (not PDO as Bob Tisdale points out).
Of course seasonal effects are relevant on short timescales and latitudinal changes are a consequence of the underlying trends but they fall into place automatically if the overview of the mechanics is correct.
All observed climate change is just the consequence of a temporary reordering of the surface air pressure distribution with a consequent shifting of the established climate zones.
The important thing is to nail exactly how and why the surface air pressure reordering occurs and I don’t think that has ever been given proper attention in the models or anywhere else.
It should be possible to diagnose a global warming or a global cooling trend from simple observations of the behaviour of the surface air pressure distribution because that always reflects the speed of the water cycle and the rate of energy transfer (via latent heat) from surface to space at any given time.
I keep repeating my exposition from different perspectives wherever I see that a point under discussion can fit my scenario. In practice that is turning out to be rather frequent with little or nothing as yet delivering a convincing counter proposal.
Stephen, I loudly applaud your cognizance of (& reiteration of) the importance of pressure distribution, but one cannot sensibly ignore the fact that hydrology is a function of absolutes. Water can cause a layer of rock to exfoliate from a rock face. This is not because of the “average” properties of the water over long periods of time; rather it is due to the VARIANCE (HAMMERING) of the properties of the water. Earth Orientation Parameters (EOP) make it crystal clear that solar & lunisolar signals hammer the variance. Averages tell something about north-south terrestrial asymmetry and about interannual variability, but multidecadal averages are a function of decadal loosening & tightening (via solar cycle acceleration/deceleration) of SEASONAL cluster widths. Ignorance of the seasons limits conceptualization to factors such as ENSO and PRECLUDES the possibility of sensible conceptualization of multidecadal terrestrial variability (probably why we have these abstractly-“imaginative” theories about AMOC that are not sensibly grounded in observation). I suppose you could argue that your interest is limited to centennial &/or millennial timescales! (for which we have no quality EOP data). Best Regards.
Yes I can see that in the short term there would be an element of climate ‘hammering’ by solar and lunar influences but it would be very difficult to separate that from chaotic internal variability.
In any event it could be subsumed into the category of weather rather than climate.
Clearly there is a point at which the short term chaotic variabiliy and/or solar and lunar ‘hammering’ are overridden by the longer term centennial and millennial signals but as far as the late 20th century warmth is concerned I see the longer term cycling as more significant because it is approximately 1000 years from the MWP and a little under 500 years from the depths of the LIA.
I’m happy to leave the short term detailing to you and others. My concern is to establish a useful mechanism for the longer term into which the shorter term variations can be slotted later.
Having started from first principles and having treated the observations like a jigsaw I’m pretty sure I am getting very close.
Stephen Wilde | July 31, 2011 at 2:42 pm | “Yes I can see that in the short term there would be an element of climate ‘hammering’ by solar and lunar influences but it would be very difficult to separate that from chaotic internal variability.”
At interannual timescales there’s topography-related turbulence (spatiotemporal chaos) that blinds LINEAR exploration, but even linear methods do fine at multidecadal timescales for those whose conceptualization adequately extends beyond eye-catching amplitudes & averages to (solar cycle acceleration/deceleration-related) changing decadal clustering of seasonal hammering. It’s no different than anchoring 2 ends of an accordion along a graduated yardstick and distorting the uniform spacing of the folds relative to the tick-marks. I shared a skeletal overview here: http://judithcurry.com/2011/07/14/time-varying-trend-in-global-mean-surface-temperature/#comment-87370 . North-south asymmetry plays a leading role in short-term variation. Maritime-continent contrasts are associated with the longer-term patterns. I appreciate your interest in the nature of the spatiotemporal variability.
I’ll have to associate myself with some of TF’s comments.
1. As always I’d like to see data and code for the paper. I don’t tire of asking for this.
2. Error bars please.
3. The question about which models capture the observations best needs to be answered. and then those that dont need to be fixed or junked.
So fully agreed, Steven. Don’t ever tire of asking these questions.
Without these answers, all conclusions are politically subjective, personal interests or pseudoscience. Untrustworthy to spend a dime of tax money.
Dr Curry’s initial posting reads: ““The main finding from this research is that there is no solution to the problem of measuring atmospheric feedback, due mostly to our inability to distinguish between radiative forcing and radiative feedback in our observations.”
I agree with this statement.”
The import of this is to support the usual empirical scientific rule for investigating dynamical systems of unknown structure, that one should seek dynamic external drivers of the system, drivers that are known to suffer negligibly small feedback from the interior of the system. This rule is generally known as the experimental method. For systems such as the earth’s energy transport process, the dynamic external drivers must be supplied by nature, such as by the orbiting of the earth, and usually cannot be supplied by the actions of the ‘experimenter’ who in this case is merely an empirical investigator.
The “forcings and feedbacks” formalism of the IPCC is an attempt to deny this basic empirical rule, and to pretend, for the sake of deception, that a CO2 doubling is, or can be treated as, a dynamic external driver, when in logic it is a change in an internal parameter. The reason for the attempt at deception is to exploit the propaganda value of the ideas of “amplfication” and “positive feedback”. Christopher Game
“to pretend … that a CO2 doubling is, or can be treated as, a dynamic external driver, when in logic it is a change in an internal parameter”. The consequential difference being? (Not challenging, just seeking understanding).
The problem is what observations are available say for cloud forcing,can we exact sufficient quality information,and what are the effects from changes in SW radiation causal,feedbacks, chance, etc.
There is no validation of cloud albedo and GCM handle clouds (read observations) poorly ie they underestimate cloud cover significtantly without peturbation.eg Ramanthan
It is remarkable that general circulation climate models (GCMs) are able to explain the observed temperature variations during the last century solely through variations in greenhouse gases, volcanoes and solar constant. This implies that the cloud contribution to the planetary albedo due to feedbacks with natural and forced climate changes has not changed during the last 100 years by more than ±0.3%; i.e, the cloud forcing has remained constant within ±1 Wm–2. If indeed, the global cloud properties and their influence on the albedo are this stable (as asserted by GCMs), scientists need to validate this prediction and develop a theory to account for the stability
Downward SW seems to anticorrelate with cloud observations eg Pavlakis fig 6
http://i255.photobucket.com/albums/hh133/mataraka/dsrregions.jpg
http://www.atmos-chem-phys.org/8/5565/2008/acp-8-5565-2008.html
This seems to support the decreasing cloudiness up to 2000 and increasing since. See fig 1
http://bbso.njit.edu/Research/EarthShine/literature/Palle_etal_2006_EOS.pdf
It does – as does the studies by Clement et al (2009) – http://www.sciencemag.org/content/325/5939/460.short – and Burgmann et al 2008 – http://circulaciongeneral.at.fcen.uba.ar/material/seminarios09/Burgman_etal_2008.pdf
And NASA – http://isccp.giss.nasa.gov/zFD/an2020_LWup_toa.gif
‘The overall slow decrease of upwelling SW flux from the mid-1980’s until the end of the 1990’s and subsequent increase from 2000 onwards appear to caused, primarily, by changes in global cloud cover (although there is a small increase of cloud optical thickness after 2000) and is confirmed by the ERBS measurements.’
But if we are looking at the thermal origins of ENSO – it is in the regions of strongly upwelling in the north Pacific and the Humboldt Current. If we are looking at why this this changed at decadal and longer timescales – the decadal variability of SAM and NAM emerge as physical systems systems affecting climate and ocean currents in this region.
SAM and NAM are influenced by solar UV warming in the middle atmosphere – a lsolar link that is being newly explored.
Chief, you can’t leave out NPI. See Table 3 & Figure 3 here:
Trenberth, K.E.; Stepaniak, D.P.; & Smith, L. (2005). Interannual variability of patterns of atmospheric mass distribution. Journal of Climate 18, 2812-2825.
http://www.cgd.ucar.edu/cas/Trenberth/trenberth.papers/massEteleconnJC.pdf
Nor, given your lucidly astute comments which I enthusiastically quoted here [ http://judithcurry.com/2011/07/14/time-varying-trend-in-global-mean-surface-temperature/#comment-87429 ], can you deny the inadequacies of these indices.
They are mere linear transects, the trajectory of which we have not yet unraveled, across a spatiotemporal attractor. This certainly constitutes VERY useful information, but we have a responsibility to duly temper our interpretations.
Well yes – there is one world climate system with standing waves – in a hydraulic analogy – in atmosphere and oceans. ENSO, PDO, NAO, IOD, THC(?), and the Pacific – North American Pattern.
We have evidence of the attractors in physical systems – but the thresholds of bifurcation involve solar UV forcing of SAM and NAM.
Consider replacing PDO & PNA with NPI and completely dropping THC (orders of magnitude smaller than what surface winds do, so dealing with that has to wait — cart-before-horse thing).
As for your ideas about solar UV forcing SAM & NAM: It might be interesting to see you try to CONCISELY put that into a simple & convincing nutshell.
Paging Erl Happ.
===============
The related north Pacific indices? Showing decadal variation? You are quibbling about nothing. The PDO is an oceanic index – the PNA and PNI atmospheric – but they are integrated in a single Earth System with – to use a hydraulic analogy – persistent standing waves seen in diverse indices.
I rely on observation and not mad theories – http://www.whoi.edu/page.do?pid=12455&tid=282&cid=54347 – the wind changes are NAM driven.
Talk about mad theories – http://www.realclimate.org/index.php/archives/2006/10/carl-wunsch-the-economist-and-the-gulf-stream/ – the MOC modulates north Atlantic circulation – so it is not an either/or situation but as usual complex and uncertain.
As for UV – it is well known that UV is absorbed by ozone.
‘During the descent into the recent exceptionally low solar minimum, observations have revealed a larger change in solar UV emissions than seen at the same phase of previous solar cycles. This is particularly true at wavelengths responsible for stratospheric ozone production and heating. This implies that ‘top-down’ solar modulation could be a larger factor in long-term tropospheric change than previously believed, many climate models allowing only for the ‘bottom-up’ effect of the less-variable visible and infrared solar emissions. We present evidence for long-term drift in solar UV irradiance, which is not found in its commonly used proxies.’ Lockwood 2010
http://iopscience.iop.org/1748-9326/5/3/034008/pdf/1748-9326_5_3_034008.pdf
Judith Lean (2008) commented that ‘ongoing studies are beginning to decipher the empirical Sun-climate connections as a combination of responses to direct solar heating of the surface and lower atmosphere, and indirect heating via solar UV irradiance impacts on the ozone layer and middle atmospheric, with subsequent communication to the surface and climate. The associated physical pathways appear to involve the modulation of existing dynamical and circulation atmosphere-ocean couplings, including the ENSO and the Quasi-Biennial Oscillation. Comparisons of the empirical results with model simulations suggest that models are deficient in accounting for these pathways.’
Lockwood, M., Bell, C., Woollings, T., Harrison, R., Gray. L. and Haigh, J. (2010), Top-down solar modulation of climate: evidence for centennial-scale change, Environ. Res. Lett. 5 (July-September 2010) 034008 doi:10.1088/174
Lean, J., (2008) How Variable Is the Sun, and What Are the Links Between This Variability and Climate?, Search and Discovery Article #110055
Cheif, in your musings about solar factors, SAM, & NAM, I advise maintaining vigilant cognizance of the synchronicity of the interannual components of NPI & NAM:
1902-1954: http://wattsupwiththat.files.wordpress.com/2011/05/vaughn_npp_image6.png
1954-2006: http://wattsupwiththat.files.wordpress.com/2011/05/vaughn_npp_image7.png
(You will note 2 phase-relation reversals, which show up as changes in color [from yellow-flames to ice-cool-blue] on multiscale complex correlation plots.)
Also see:
http://wattsupwiththat.files.wordpress.com/2011/05/vaughn_npp_image5.png
Caution: Perhaps largely due to the nature of its construction, people are misinterpreting PDO (particularly the early part). I can suggest that one sensible option to consider is adoption of a related index that is reliably straightforward & simple to interpret.
http://wattsupwiththat.files.wordpress.com/2011/05/vaughn_npp_image2.png [where D-T = -SOI]
Still, which of these indices can possibly have a reliably straightforward & simple linear relationship with hydrology across all seasons & regions? None. It’s physically impossible. Even in the best case scenarios, there WILL be phase-relation reversals due to spatial nonstationarity and hydrology’s dependence on absolutes. I advise the mainstream to get off the hydrologically-seriously-misleading anomalies & linear EOFs.
The anti-correlation applies to surface flux in the central Pacific – not hugely relevant to regions of the eastern Pacific where ENSO cloud feedbacks occur.
‘ Most reduction in low cloud amount related to the 1997-1998 El Nino
occurs in the eastern tropical Pacific associated with an upward large-scale motion and a weak atmospheric stability measured by the 500 hPa vertical velocity and the potential temperature difference between 700 hPa and the surface, and is negatively correlated to the local SST anomaly. In addition to the other mechanisms suggested by the previous studies, our analyses based on the ISCCP observations indicate that the change in atmospheric
convective activities in these regions is one of the reasons responsible for the change in low cloud amount.’
https://www.cfa.harvard.edu/~wsoon/EarlyEarth07-d/ZhuHackKiehlB07.pdf
To imagine that clouds don’t exhibit secular change is convenient.
http://thinkprogress.org/romm/2011/07/29/282584/climate-scienists-debunk-latest-bunk-by-denier-roy-spencer/
No science here.
so predictable
So shallow and low.
I tried to read Romm, I swear. But in just 3 paragraphs my IQ went down and my belligerence index went up.
I had to stop to protect myself and my family.
Well !
Hi
I read all comments and theres a lot of different aspects of uncertanty. But reading them all R Spencers conclusion i is still valid. And no one so far has tries to explain the difference between the models and the actual meashurements. Ther is a lot of comments dicussing irrelevant topics as smoke screens away from the central issue?
It s obvious as always when the “Team ” is challanged that there is much at stake and the agression has to be a proof of that they are not in anyway intrested in other aspects or conklusions that challenges thier own.
Its so sad that the scientific issue is lost. Its turned out to be a dog fight instead. The models are wrong its obvious and no news.
And many of the commentaters can not have read the papaer at all.
‘The canonical way to think about clouds is that they are a feedback—as the climate warms, clouds will change in response and either amplify, (positive cloud feedback) or ameliorate (negative cloud feedback) the initial change.
What this new paper is arguing is that clouds are forcing the climate, rather than the more traditional way of thinking of them as a feedback.’ Dessler today
‘Much work has been done on ENSO over the past few decades and pretty much everyone agrees that it’s a stochastically triggered, coupled dynamic mode of the atmosphere, ocean system. I’ve never seen any suggestion that it’s triggered by clouds. To the extent that clouds amplify ENSO, that’s the cloud feedback, and that is what I measure in my paper.’ Dessler yesterday
‘Finally, since much of the temperature variability during 2000–2010 was due to ENSO (Dessler 2010) we conclude that ENSO-related temperature variations are partly radiatively forced. We hypothesize that changes in the coupled ocean-atmosphere circulation during the El Niño and La Niña phases of ENSO cause differing changes in cloud cover, which then modulate the radiative balance of the climate system.’ Spencer and Braswell 2011
‘It is not controversial to state that climate models are deficient in terms of tropical variability in the atmosphere on many timescales [Lin et al., 2006; Lin, 2007] and a more realistic simulation of ENSO events in coupled simulations remains a high priority for model developers. During El Niño, the warming of the tropical eastern Pacific and associated changes in the Walker circulation, atmospheric stability, and winds lead to decreases in stratocumulus clouds, increased solar radiation at the surface, and an enhanced warming.’ Trenberth et al 2009
Veracity in communicating science is a casualty of the climate wars?
Chief Hydrologist,
Terrific comment. I would love to see what Dr. Curry says about this.
Perhaps Dessler never got around to reading Trenberth’s paper? Or maybe he just disagrees?
Caveat: I’m a chemist, not a climate scientist (whatever discipline that might encompass). Actual climate scientists are encouraged to explain if and how I’ve got it wrong.
As I understand it, the key contention between catastrophists and lukewarmers concerns the size and direction (positive or negative) of the net feedbacks from CO2 increases; direct warming effects are modest, on the order of 0.6 – 1.2 degrees C per CO2 doubling (correct?) In turn, these feedbacks are mostly theoretical estimates rather than direct empirical measurements, since no unambiguous method of direct measurement is generally accepted. These theoretical feedback estimates are incorporated into various climate models.
The present S&B publication uses an empirical satellite-based measurement of net radiative energy flux (which, if accurate, necessarily must reflect whether the earth is cooling or warming, independent of any model). These empirical data are compared with expected radiative values from a relatively simple climate model. The discrepancies between measured net energy flux and the flux expected from the model suggest, but cannot prove, that the model is wrong because it exaggerates the positive feedback component of climate sensitivity.
Trenberth et al. respond that more sophisticated models can account for the observed net fluxes; therefore the presumed feedbacks and concomitant climate sensitivity are not in question.
Does this summary fairly represent the matter at issue?
If so, here’s my question: can the empirical net flux data,or any other data for that matter, be used to estimate climate sensitivity independently from the assumptions embedded in any particular climate model?
Another way of posing the same question: is “climate sensitivity” a fundamental measurable physical quantity like heat capacity, or is it a conceptual artifice which arises as a convenience for ease of calculation within our primitive climate models?
My take is that climate sensitivity is very dependent on the time scale you measure it, and the only true value is the equilibrium response (for infinite time scale). I don’t think studies like this get a good value because of the short periods of the forcing that limit the response to a relatively shallow layer of the ocean (25 m in the paper). For me, the more credible studies look at warming over multiple decades and the corresponding forcing change to produce a sensitivity. I think this paper also comes to the conclusion that shorter period fluctuations are too noisy to get anything useful out because the radiative forcing and response can’t be separated easily without some gross assumption, like the forcing change being entirely due to CO2, for example. I agree most with Kerry Emanuel’s view that there are useful things in this study, primarily in the call for caution in using short-term data this way.
Barry – You ask very perceptive questions, and you are right that the climate sensitivity issue is at the heart of questions surrounding the magnitude of CO2-mediated global warming. I’ll try to respond to some of them ( in italics):
As I understand it, the key contention between catastrophists and lukewarmers concerns the size and direction (positive or negative) of the net feedbacks from CO2 increases; direct warming effects are modest, on the order of 0.6 – 1.2 degrees C per CO2 doubling (correct?)
Direct calculations made by differentiating the Stefan-Boltzmann equation yield a value of 1 deg C. More complicated models incorporating seasonal and latitudinal heterogeneity yield values that all approximate 1.2 C. There is no good evidence for less than 1 C.
In turn, these feedbacks are mostly theoretical estimates rather than direct empirical measurements, since no unambiguous method of direct measurement is generally accepted. These theoretical feedback estimates are incorporated into various climate models.
Feedbacks can’t be measured directly, but they can be estimated from a combination of fundamental principles and observational data, as I mentioned in one of my above comments. As just one example, Hansen has utilized estimates of climate forcing from paleoclimatologic measurements and compared them with temperature responses. This yields a direct value for climate sensitivity (temperature response to a forcing), and the feedbacks can be calculated from this value. The uncertainties relate to the accuracy of the historical data and their relevance to current climate conditions, but the calculated values are not theoretical, and in Hansen’s case, are not model based. Other approaches utilize models to match observations.
The present S&B publication uses an empirical satellite-based measurement of net radiative energy flux (which, if accurate, necessarily must reflect whether the earth is cooling or warming, independent of any model). These empirical data are compared with expected radiative values from a relatively simple climate model. The discrepancies between measured net energy flux and the flux expected from the model suggest, but cannot prove, that the model is wrong because it exaggerates the positive feedback component of climate sensitivity.
The satellites measure flux anomalies (changes from a baseline) and not net radiative energy flux. They are not accurate enough to determine whether the earth is cooling or warming, but rather whether it is changing in one direction or another. For example, an increase in net downward flux tells us either that the earth is either warming more or cooling less than previously, but not whether it is warming or cooling. S&B compare their results with an array of complex climate models. Contrary to their claim, some of the models do yield similar results to the ones they observed, while others do not. A model which matches their results is characterized by a high climate sensitivity to CO2 involving strong positive feedback, reinforcing the principle that the data they derive from changes in ENSO (El Nino and La Nina) tell us little about sensitivity to CO2.
Trenberth et al. respond that more sophisticated models can account for the observed net fluxes; therefore the presumed feedbacks and concomitant climate sensitivity are not in question.
Does this summary fairly represent the matter at issue?
Not exactly, Barry. The models run by Trenberth are all sophisticated. It is the S&B model that is very simple.
If so, here’s my question: can the empirical net flux data,or any other data for that matter, be used to estimate climate sensitivity independently from the assumptions embedded in any particular climate model?
Another way of posing the same question: is “climate sensitivity” a fundamental measurable physical quantity like heat capacity, or is it a conceptual artifice which arises as a convenience for ease of calculation within our primitive climate models?
All estimates involve assumptions, some very reliable, others less so. Climate sensitivity to CO2 has been approached by a multiplicity of approaches utilizing different aspects of modeling (and in Hansen’s case, independent of models). Short term sensitivity to the type of ENSO variations addressed by S&B may be quite different because the climate response to ENSO differs from the longer term response to CO2. The estimates for CO2 tend to converge on a range of values, but the range is fairly broad (2 to 4.5 C per CO2 doubling), with some outliers. To answer your final question, I don’t believe climate sensitivity can be called “fundamental” because it varies with circumstances and might be different in a climate very different from our current one. Going forward, It will certainly be difficult to measure it directly with confidence in the accuracy of our results, because we will rarely be confident that we know the exact values of the forcings we are matching a temperature response to, but it is not an artifice. Given a specific forcing (a change in the radiative balance) with no change in other factors, the climate system will inevitably respond with a given temperature change over a long term average (although perhaps punctuated by short term fluctuations reflecting chaotic behavior), and so “sensitivity” is a very real property of climate. The challenge will be to narrow the range within which climate sensitivity to CO2 is likely to fall, but we are still far from a very precise figure.
There is no good evidence for less than 1 C.
Lindzen & Choi appear to estimate 0.7 C: (yes?)
http://www.masterresource.org/wp-content/uploads/2011/06/Lindzen_Choi_APJAS_final.pdf
Is this evidence considered “not good”?
(note that I am a VERY much a layman)
Ah – sorry – L&C’s 0.7C is with feedback – sorry about that. You were referring to NON feedback sensitivity. I see that L&C give a non-feedback sensitivity of 1.1 C. ;^)
That is not evidence for the Planck Response (“no feedback”) sensitivity, which is somewhere between 1C (if calculated roughly) and 1.2 C (if calculated based on various regional and seasonal heterogeneities). Lindzen/Choi has other problems, but it is irrelevant to the topic of the direct “no feedback” response.
Fred,
Obviously Lindzen/Choi overestimated 0.7 C and your 1.2 C is beyond reality, i.e. non-sense and you are still insisting. Wonder people queried which planet you are living.
Fred Moolton-
Thanks for your thoughtful reply. I appreciate your clarification concerning the magnitude of the direct warming effect from CO2 doubling,
Your point that satellite data reflect relative rather than absolute radiative intensities is very useful. However, in the lab we often used relative radiative intensity measurements for convenience, but could convert them into absolute intensities by calibration. Isn’t a calibration protocol possible with the satellite data? Are these data too imprecise to be worth the effort? The incoming solar radiation at the top of the atmosphere apparently is measured routinely to high accuracy; I am baffled as to why the outgoing radiation cannot be measured with similar precision and accuracy, The effort to do so would seem to be well worth the trouble and expense, since it would provide direct measurement of earth’s temperature balance: exactly the quantity over which the contending sides are in such dispute,
In discussing S&B’s satellite flux anomaly data versus climate models, you make a fascinating statement:
“A model [ advanced by Trenberth] which matches their [S&B satellite data] results is characterized by a high climate sensitivity to CO2 involving strong positive feedback, reinforcing the principle that the data they [S&B] derive from changes in ENSO … tell us little about sensitivity to CO2.”
This statement gets to the heart of my questions concerning the validity of the whole concept of “climate sensitivity.” It isn’t apparent to me that the earth should respond identically to an increment of heat energy, regardless of whether the heat change comes in the form of added shortwave solar radiation at the top of the atmosphere, or reflected shortwave energy from near-surface clouds, or absorbed long wave radiation integrated over the troposphere. Each of these energetic alterations might result in different proportions of that energy ultimately absorbed or radiated by the system.
Yet, to the trifling extent of my familiarity with climate models, these different energy changes are treated as though they are equivalent, e;g;, climate sensitivity parameters are constants within a particular model. Is this in fact the case?
And if this is not the case, then shouldn’t we dispense with the whole notion of “climate sensitivity” in favor of a complicated multidimensional factor which reflects the reality embedded within these models?
And this leads finally to a last and crucial point. You state that:
” Hansen has utilized estimates of climate forcing from paleoclimatological measurements and compared them with temperature responses. This yields a direct value for climate sensitivity ( temperature response to a forcing), and the feedbacks can be calculated from this value. The uncertainties relate to the accuracy of the historical data and their relevance to current climate conditions, but the calculated values are not theoretical, and in Hansen’s case are not model based.”
But this raises two questions. For one, if the response function varies depending upon the particular source of the forcing, then how can the paleological record be helpful to the present situation, given that the large climate variations of the holocene were not accompanied by significant changes in CO2?
For another, how in principle can the sparsely quantified paleoclimate record be used to calculate a model-independent climate response function, if the data-rich current climate record cannot? Is this merely because the current record covers too short a time scale to permit robust extrapolation? What am I missing here?
Anyway, Fred, I appreciate your comments and hope you can further enlighten me.
The general properties of the atmosphere are influenced mostly by dynamics of air as an essentially ideal gas and by the dynamic behavior of water (evaporation and condensation). The differences between forcings due to increased CHG concentrations or due to a change in solar irradiance are largely canceled automatically by a compensating change in convection.
The energy fluxes of the atmosphere are driven by the heating of the surface and the loss of heat from the top of troposphere to the space. The direct effect of additional radiative forcing may be divided differently between different altitudes, but when the compensating effects are added, the overall forcing determines almost totally the outcome. Some minor differences will certainly persist, but they are irrelevant when compared to all other uncertainties in the understanding of the climate.
These compensating changes in the convection are in a sense also feedbacks, but they are not handled as such, because it’s easier to determine the reaction of the atmosphere to the forcing including changes in convection as part of the primary effect, than to study what happens to the atmosphere excluding them. This is related to the central role of adiabatic lapse rate in the understanding of the atmosphere. The compensation occurs in the processes that return the lapse rate to values allowed by convection.
The incoming radiation was recently adjusted downward by about 5 W/m2 – the problem is calibration. What do you calibrate against?
The anomalies are much more accurate and are very useful. The Earth is always in a radiative disequilibrium – leading to warming or cooling.
Ein/s – Eout/s = d(GES)/dt
Where Ein/s and Eout/s are the average radiative flux x 1 second (unit energy) for a period and d(GES)/dt is the rate of change of global energy storage. If we know the planet is warming from ocean and atmospheric date then Ein > Eout and vice versa.
If we look at actual data for 1985 to 1998 – we know the planet was warming so Ein > Eout.
http://isccp.giss.nasa.gov/projects/browse_fc.html – look at near global radiative flux at TOA. Net flux positive shows the planet warming by convention.
There is little trend from 1985 to 1998 for Ein – see the graph here http://www.agci.org/docs/lean.pdf This paper is easy to read and quite informative – I found – but she doesn’t have ENSO cloud feedbacks so not quite right.
What changed was Eout – there was less light being reflected out at the end of the period and more IR out as a result of less cloud.
‘It is not controversial to state that climate models are deficient in terms of tropical variability in the atmosphere on many timescales [Lin et al., 2006; Lin, 2007] and a more realistic simulation of ENSO events in coupled simulations remains a high priority for model developers. During El Niño, the warming of the tropical eastern Pacific and associated changes in the Walker circulation, atmospheric stability, and winds lead to decreases in stratocumulus clouds, increased solar radiation at the surface, and an enhanced warming.’ Trenberth et al 2009
It seems suddenly to be controversial.
Spencer and Brasswell find that they could not calculate sensitivity – a null result.
I think that the idea of a constant climate sensitivity is nonsense. In some cases climate will behave one way – and a different way in another. Climate is a complex and dynamic system with negative and positive feedbacks and uncertain thresholds.
Sensitivity is defined by increasing CO2 in models to twice pre-industrial and allowing the models to equilibriate. We get a different answer from each model – and remain uninformed of the ‘imprecision’ of any particular model.
‘Atmospheric and oceanic computational simulation models often successfully depict chaotic space–time patterns, flow phenomena, dynamical balances, and equilibrium distributions that mimic nature. This success is accomplished through necessary but nonunique choices for discrete algorithms, parameterizations, and coupled contributing processes that introduce structural instability into the model. Therefore, we should expect a degree of irreducible imprecision in quantitative correspondences with nature, even with plausibly formulated models and careful calibration (tuning) to several empirical measures. Where precision is an issue (e.g., in a climate forecast), only simulation ensembles made across systematically designed model families allow an estimate of the level of relevant irreducible imprecision.’
Until we see the results of ‘systematically designed model families’ we have no idea what the level of imprecision in models is.
‘Sensitive dependence and structural instability are humbling twin properties for chaotic dynamical systems, indicating limits about which kinds of questions are theoretically answerable. They echo other famous limitations on scientist’s expectations, namely the undecidability of some propositions within axiomatic mathematical systems (Gödel’s theorem) and the uncomputability of some algorithms due to excessive size of the calculation’
McWilliams 2007 – Irreducible Imprecision in Atmospheric and Oceanic Simulations –http://www.pnas.org/content/104/21/8709.full.pdf+html
Sensitivity is meant to be constrained by paleoclimatic data. Paleoclimatic data is described by the NAS committee on abrupt climate change as feeling about in a dark room.
http://www.nap.edu/openbook.php?isbn=0309074347
http://www.whoi.edu/page.do?pid=12455
These are quite good sources on an alternative way of looking at climate – which I think you are well on the track off.
Without the big policy disagreement essentially every scientist would agree without hesitation that models are severely deficient. On the other hand most of them would also state that even so deficient models tell a lot. When we have this controversy, many scientists think that saying openly that models are deficient would be interpreted widely to mean that they have no value. Therefore they formulate their views in so many different and sometimes contradictory ways.
Bingo, Pekka. You’ve put your finger right on the corruption. Now address it, please.
=============================
That’s your interpretation, not mine. I consider it a problem, but not a crime.
Many people think too much, how they will be interpreted, believing that they can get their message through with less bias, by being less than fully open, but that doesn’t work.
At the same time very many people are fully convinced that what has been told is much more seriously biased than it really is, but that cannot be corrected by increasing the bias, because that will soon be noticed, and the message will be reinterpreted based on that. At the same time also the most honest and objective data will be assumed to be highly biased. That is a sure way of increasing skepticism, and even more erroneous views on the best scientific understanding.
Being fully open and aiming at maximal objectivity will not lead to unbiased public understanding, but it’s still the best choice available. That’s true even, when some opponents do their best to increase skepticism.
So you think ‘corruption’ implies a crime? Try the dictionary.
=========================================
Not only is it probably not criminal, also it is not just ‘a problem’ but rather ‘the problem’. Now fix it Dear Liza, Dear Liza, you’ve the insight to lead the charge.
============================
Widespread recognition of the inadequacies of the models would go a long way toward fixing this whole mess. So how to get this done? I’m certainly open to suggestions.
====================================
Fixing it is not on Liza’s agenda. Destroying the American economics is on her agenda. Wasting the American tax with climate researches is on the Administration agenda.
Barry – Others have addressed some of your latest questions. The only thing I’ll add is an example of how different climates can be characterized by different climate sensitivities.
One of the positive feedbacks on any warming process is the ice-albedo effect, which refers to the principle that a warming climate melts highly reflective snow and ice, exposing the underlying water or ground that is less reflective. This reduces the Earth’s albedo (fraction of reflected sunlight) and allows more solar energy to be absorbed, amplifying the original warming signal.
The magnitude of this effect depends on the extent of snow and ice. In a glaciated climate such as existed 25,000 years ago, warming operated on a much larger ice area than exists currently, and so the amplification (positive feedback) was greater. If the current climate warmed in the future to the point that all snow and ice disappeared, that particular feedback would also disappear and so climate sensitivity would be less.
Despite all this, our current interglacial interval (the “Holocene”) of the past 11,000 years has been stable enough so that huge variations of the sort I described are unlikely. Climate sensitivity to a particular forcing (e.g.,CO2 increases) will not necessarily remain constant, but it will not vary as much as would be the case in comparing a glacial with an interglacial interval or in comparing current climate with one vastly different in some other sense.
Fred,
“Not exactly, Barry. The models run by Trenberth are all sophisticated. It is the S&B model that is very simple.”
Can you elaborate ‘sophisticated’ and ‘very simple’? I tend to believe in simple models that matching satellite data than sophisticated models do not matching the satellite data though all of these models are likely fraud to represent the Earth complicated energy system.
Sam – The GCMs Trenberth refers to were not constructed to simulate the specific fluctuations evaluated in the SB-11 paper, and so when one or more of them fit well, that tells us something about their skill. The very simple model SB-11 used (neglecting critical ocean variables, for example) was designed with the specific interval in mind, and so could be made to fit the data. That is less informative because it doesn’t tell us whether some very different model also designed to fit the data wouldn’t yield a very different interpretation.
Fred,
OK so SB-11 did not take into account of ocean variables. So what else that make GCM so sophisticated and SB-11 so simple? Or what are the vaiables that GCM models had put in that the SB-11 did not have. My second question is then are all those additional variables in the sophisticated GCM departed from the satellite measurements?
Fred Moolten
I have taken the liberty of extracting four very pertinent sentences in your response to Barry Elledge, which would summarize for Barry and others how the S&B findings fit in with earlier model-derived information on climate sensitivity.
Indeed, Fred. That appears to me to be the main problems with the interpretation of paleo-climate data: the data themselves are dicey enough and the interpretation is rather subjective. As a result, these data are far less meaningful than real-time physical observations.
Essentially it is a model-derived estimate, only as good as the assumptions fed into the models. The inherent problems with the paleo-climate interpretations are mentioned above. The real weakness is that there are no empirical data (i.e. from actual physical observations or reproducible experimentation) supporting the climate sensitivity estimates.
I would agree, Fred. But you could replace the word “precise” with “meaningful” and I would still agree.
This gives credence to the findings of the S&B study:
And to JC comment to this conclusion:
And, finally to this statement in JC conclusion:
Amen!
Max
Max – Thanks for your input. I think we agree on main points, but I believe we have more observational confirmation and less model dependence than you state. Even so, there is a clear need to constrain climate sensitivity better. This particularly applies to long term global climate responses to persistent forcing from CO2 or persistent solar changes, because any attempt to extrapolate from short term fluctuations – particularly regional ENSO-based fluctuations – is likely to give a false picture.
Fred,
Why are we still waiting on you to provide examples of mitigation strategies that work?
Speaking of pictures, do you think claiming that the slight warming from the 1970’s until the late 1990’s was long enough or great enough to declare a worldwide climate crisis?
Barry
Excellent question. For some clues, see previous posts: judithcurry.com “Climate sensitivity” e.g. <a href=http://judithcurry.com/2011/07/07/climate-sensitivity-follow-up/<Climate Sensitivity Followup, including links to Idso etc on climate sensitivity e.g.
Idso, S.B. 1998. CO2-induced global warming: a skeptic’s view of potential climate change. Climate Research 10: 69-82.
Note the bimodal distribution of sensitivities: with IPCC models tending high, and skeptics finding low sensitivities, especially from methods outside of global climate models.
Errata: Climate Sensitivity Followup,
Spencer has the best global temperature record in the world for the last 31 years but lacks the courage to say so. To me this paper is about a peripheral issue. The real issue is this: why do ground-based global temperature records show warming for periods of decades when the satellite temperatures cannot see this warming? Both cannot be right. Either warming exists in the eighties and nineties or it does not. I have concluded that it does not and the ground-based records are cooked. As in falsified. You will find proof of this in my book “What Warming?” If you want to dispute this conclusion you will have to explain away figures 24, 27, and 29 in the book. This is an empirical observation and it cuts the legs right out from under Hansen’s 1988 testimony that warming had started.
The satellite records show warming in the 80s and 90s
lol, Arno, I hate to say this, but I’m probably not going to see your figures 24,27 or 29. Sorry. :-( Perhaps if you describe them for me, I can give you an explanation. As lolwot points out, the sat data did show warming in the 80s and 90s. Further, Dr. Spencer has been described by many people with many words……..”lacking courage” isn’t a trait I would attach to Dr. Spencer.
Here is both UAH’s and RSS’ temps graphed. I’ve trended the 80s and 90s for you. http://www.woodfortrees.org/plot/uah/from:1979/plot/uah/from:1980/to:2000/trend/plot/rss/from:1979/plot/rss/from:1980/to:2000/trend
Arno, I’m as big of skeptic as anyone, but, you’re gonna have to come up with something else to plug a book.
http://www.woodfortrees.org/plot/uah/from:1979/plot/uah/from:1980/to:1997/trend/plot/rss/from:1979/plot/rss/from:1980/to:1997/trend
If rapid flux response to heating dominates, there is no energy surplus to drive climate sensitivity. That is, quick negative feedback trumps slow positive feedback. That’s my takeaway from this study’s possible implications.
Thanx Brian,
That’s what i thought was the gist of the paper, from studying Enso events and the sattelite records, it appears according to this paper, that energy escapes to space more readily during a warming then the models predict.
Therefore, if correct, as the CO2 effect progresses, the rate of energy escape to space will actually increase in a non-linear way, thus constraining the warming of the planet to the low end of the predictions.
Was kinda wondering why most of the discussion has been about Enso events, their origin, whether clouds cause Enso or vice-versa, etc.
I thought it was a simple statement, based upon real observations, that our climate system has a way to deal with excess energy that the climate models have not figured out yet.
Why do people feel satisfied in offering a defense by pointing to others and saying “But, but, they’re a bigger one?”
My parents rejected such logic as childish when I offered it in my pre-teens.
Aristotle would be so proud textbook examples of logical fallacies are still being produced.
“The basic material in the paper has very basic shortcomings because no statistical significance of results, error bars or uncertainties are given either in the figures or discussed in the text. Moreover the description of methods of what was done is not sufficient to be able to replicate results.”
_______
Just wait until McIntyre gets ahold of this study.
This is hilarious! I just skimmed through post and comments, but if this is repetitive……get over it. It needs stated over and over again.
“The study in the journal Remote Sensing questions the accuracy of climate computer models ….
Holy crap!!! Where have people been? Or, rather, what orifice were their heads in? Questions the accuracy of models? Have we found one that could be called almost accurate? Just one? None of the models are correct. Show me one that predicted the last decade. Any one! You’d thought as many monkeys we had at the keyboard one would have accidentally made one that was accurate…… but they didn’t. Jeez, the laws of random don’t even apply!! Why? Because the models aren’t random, they’re wrong. They are wrong on the very base assumptions. Every one of them!!! Which brings me to my other point. Dr. Curry, it is clear that describing models as alarmist models is accurate. I’ve never once seen a discussion of models predicting a flat line or cooling period, and yet, here we are, going into the second decade of no temp rise.
If there is any discussion to be had about Dr. Spencer’s paper, it should be, not what can we do to cast doubt about it, rather it should be to state, yes, he’s correct, the models are crap. They are devoid of value. They should thank Dr. Spencer for being the millionth person to point out that the models are wrong and have been wrong since their use. They need scrapped and forget everything we thought we knew and start anew. Then, play with the cloud feed-backs….. It doesn’t really matter about ENSO this or La Nina that or anything else. The heat is going up, up, and away. Model something that accounts for that. You would have thought the Kev would have been happy someone finally ended his travesty! But, no, the ingrate wants to whine.
Lastly, to the whinebags about the sensationalizing the media did with this……sucks don’t it? Is that the only sensationalized story you guys can recall regarding our climate? Turns out, the damned thing is a double edged sword!! AHAHAHAHAHAHAHAHAHHHAA!!!!!
James
Precisely. There is a divergence problem. The models don’t fit reality. There are some really clever people here with a major reality problem.
Dr. Curry,
You write “I disagree with Gavin Schmidt that paloeclimate data is more robust than the satellite data., but both data sets are essential in this endeavor.”
I have to disagree with this statement. Paleoclimate data is far too uncertain to provide any value at all to the discussion of climate sensitivity. Even if it were possible to get a viable global temperature record from tree rings or ice cores or sediments, we still know nothing about the level of global cloud cover or cosmic rays or … you get the idea. Gavin Schmidt is clearly grasping at straws in a doomed effort to prop up a crumbling theory.
Paleoclimate data adds nothing to the discussion of sensitivity. To pretend it does is a tremendous disservice to science.
I would add to the problems with knowing the forcings and the questionable proxy data in general, that there is also the question of how stable a constant climate sensitivty is. I see no reason to assume it does not vary depending on surface conditions.
The point is you cannot calculate climate sensitivity unless you know a great deal about natural climate variability. Paleoclimate data cannot come anywhere close to providing us with enough information to determine if changes in the temperature were caused by atmospheric CO2, changes in clouds from cosmic rays, solar variation or any number of other things.
Gavin Schmidt’s remark would be true only if CO2 was the only driver of climate. It isn’t. Schmidt’s view the climate is driven only by greenhouse gases is a fantasy world created by Jim Hansen and the IPCC.
On many occasions I have seen Dr. Curry point out that uncertainties are higher than the IPCC likes to admit regarding attribution of climate change. If we cannot tell how much of climate change is natural and how much is anthropogenic now, we certainly cannot say anything specific about changes 1,000 to 2,000 years ago.
“Even if it were possible to get a viable global temperature record from tree rings or ice cores or sediments, we still know nothing about the level of global cloud cover or cosmic rays or … you get the idea.Clouds would be a feedback so….
We don’t need a viable global temperature record we just need to know roughly how much temperature has changed between two states. We know the glacial period was about 5C cooler globally than the present state and we know the state millions of years back was several degrees warmer than present.
In light of a low climate sensitivity like 0.2C/wm-2, that would require 25wm-2 forcing to explain a 5C warming. 25wm-2! That’s massive. Do we really believe there is some unknown forcing out there that dwarfs anything known? It’s kind of infeasible that noone would have found the big un after all these decades.
Furthermore the implications of such a large unknown forcing out there would mean one forcing dominates temperature over time. In which case global temperature should closely correlate with that one forcing. That should make it easy to identify – if it exists, so again we wonder why it hasn’t been found.
Instead the global temperature record doesn’t appear to correlate exactly with any particular forcing. It looks instead to be the sum of small and different forcings over time amplified by high climate sensitivity.
lolwot,
You write “Instead the global temperature record doesn’t appear to correlate exactly with any particular forcing. It looks instead to be the sum of small and different forcings over time …”
i can agree with this portion of your comment. However, it is impossible to know anything about climate sensitivity because we don’t know which different forces were at work over time or how great those forces were. This is what makes Gavin Schmidt’s comment so ridiculous.
Schmidt’s comment is as mockable as anything written by the Sky Dragon theorists.
We need more and better paleoclimate data; what we have right now is not very robust
Gavin’s argument is a vestigial argument. It was first formed when climate scientists thought CO2 was the only driver of the climate. If CO2 was the only driver, then Gavin’s comment might make sense. We now know there are many things which affect the climate and we have zero data on some of them. Gavin’s comment is laughable and should be laughed at.
When did climate scientists think that CO2 was the only driver of climate?
Since Tyndall thought there were other greenhouse gases other than CO2, I think the true answer is never.
bob,
I used CO2 as a place keeper for greenhouse gases. Climate scientists used to think GHGs were the only variable, and CO2 the most important among them. They denied, as some today maintain today, that solar variability plays a significant role in climate change. Water vapor and clouds changes were considered a response to changes in GHGs, not something which may respond to PDO, solar variability or perhaps other unknown mechanisms (there are several competing theories on clouds these days).
The point I am making stands. Paleoclimate data is not up to the task of telling us anything about climate sensitivity. We do not have a global cloud cover time series for the 20th century much less one going back 2,000 years or more. It is not that we lack good data. We have no data at all for some of the most important issues. Without understanding the role of natural climate variability, we can never come to any conclusions about climate sensitivity to CO2.
Jim-
Thanks for your response. I can well believe that climate response is time-variant as well as dependent on myriad other variables: No theoretical reason exists for presuming the response function is constant with respect to any variable.
Time-dependent response to increasing CO2 logically would seem to exhibit two types of terms: one type which reflect stored but unmeasured heat, possibly in ocean depths or ice melt (Trenberth’s “missing heat”); and another possible type which reflect nonthermal changes in the earth, such as changes in vegetation or other biological processes which affect energy storage in the longer term.
Yes, there may be many responses to a forcing, but the ones that matter are those that cancel the forcing. For example if the sun became 1% hotter, the earth would have to warm or increase albedo to stay in equilibrium. Doubling CO2 is equivalent to the sun becoming 1% hotter.
Doubling of Co2 supposedly is 2.7W/sqm
Bright sunshine can be 1000W/sqm versus 300W/sqm or so when cloudy.
UK has 100 more hours of bright sunshine per year compared to 1929 – as measured by the Met.
Doubling CO2 is 3.7 W/m2 at the top of the troposphere. The sun’s average global input at the top of the troposphere is about 340 W/m2.
I doubt Phoenix gets a lot of sunshine at night. However, it does get an average of 8.17kWh/sqm of sunshine in June.
1.1% more sunshine in June can match all the extra energy supposed caused by a doubling of CO2 in June (and Co2 has not doubled).
A very small increase in bright sunshine can explain any warming.
“Doubling CO2 is equivalent to the sun becoming 1% hotter”. Interesting. The sun is 3% closer to the earth at perihelion than aphelion. That would make the sun apparently ?% hotter at perihelion. Can you work it out? Is it negligible?
Good question Simon. For a point source the square law works but for the sun…dear me.
The square law still applies, so it is about a 6% range, maybe 7%.
If it does then it’s 9% innit?
What is 1.03 squared?
Oops my bad. Apols.
“Doubling CO2 is equivalent to the sun becoming 1% hotter.” Most unsubstantiated statement of this topic? Equivalent to ice-cream heating up the Sun statement?
Jim D,
Hypothetical question. If the sun becomes 1% hotter, what increase in CO2 is the 1% hotter sun (and warmer Earth) equivalent to? In equilibrium.
Doubling CO2 is equivalent to the sun becoming 1% hotter.
Less clouds or less aerosols or a little more bright sunshine (or a combo) can do it to. The sun doesn’t need to change.
I just gave an equivalence. It is all hypothetical. You could reduce albedo by 1% (equivalent to a cloud cover reduction of maybe 6%) and get the same heating.
http://hockeyschtick.blogspot.com/2011/05/study-finds-global-warming-from-natural.html
“The paper finds the decrease in reflectance from clouds (albedo) over only the past 21 years has accounted for a change in solar energy delivered to the Earth surface of ~ 7W/m2, whereas all greenhouse gases are claimed to only account for (assuming you believe the IPCC) a ~ 2.4 W/m2 change over a much longer 104 year period”
However, you will notice that the net albedo change over the period is approximately zero despite the large fluctuations, so it doesn’t explain any of the net warming between 1983 and 2003.
7W/sqm … equivalent to a 2% increase in TSI.
It explains pre-1998 warming very well.
Did you understand my question? How will the atmospheric CO2 response when the sun becomes 1% hotter? No anthropogenic CO2.
The CO2 wouldn’t respond much, maybe 10 ppm due to the warmer ocean (from Henry’s Law), but the water vapor would have a positive feedback to the warmer temperature.
How warm will Earth become after the sun becomes 1% hotter? After the feedbacks.
The climate sensitivity argument would be similar to that for doubling CO2, so about 3 degrees. But we don’t know if there might be a positive GCR feedback that some suggest. In the 11-year cycle, the sun changes by 0.07% and it seems to produce a cycle up to 0.2 degrees.
actually doubling CO2 is equivalent to the sun becoming 1.5% hotter.
Doubling of CO2 is roughly a 3.7wm-2 forcing.
Sunlight aborbed by the Earth is about 239wm-2
doubling over what range?
http://en.wikipedia.org/wiki/Radiative_forcing
I don’t understand the question. However long it takes?
The top of atmosphere number is what I would compare with.
It does not snow much when the Arctic Ocean is cold and Frozen. Glaciers do not advance in cold frozen Arctic Ocean times. Glaciers do advance in warm thawed Arctic Ocean times. If the sun becomes hotter, it melts more Arctic Sea Ice and it snows more. Pay attention to how much it snows as a function of Low Arctic Sea Ice Extent.
This extra snow in warm times does increase Albedo.
This lack of snow in cold times does decrease Albedo.
This thread, and most of the other threads, lack this Albedo feedback.
A fool or a naive person might expect a global temperature forecast to be right on target. Some critics of forecasts appeal to the naive and foolish by falsely implying anything less than “on target ” is a useless forecast. A useful forecast is one that is more right than it is wrong.
Were expecting another decade or 3 of global cooling as upwelling in the eastern Pacific intensifies in the cool phase of the Pacific decadal variation?
Cooler SST and more cloud?
Oh well – you will find out.
Well, I certainly hope I will find out, but a decade is a long time, and I may be long gone by then. Do you think I could find out in heaven?
I don’t know – but I hope you aren’t ill.
I hope we don’t need to wait a decade – there will be a time when the anomalies of natural variation (er.. ENSO cloud feedback) become so great that the new paradigm becomes inevitable.
If you actually start to ask yourself why it isn’t warming – I am sure you can do it sooner.
The “lack” of warming since about 2002 happens to cover a period that starts with El Ninos (2002-2007 was almost completely el ninos) and ends with majority La Ninas. This period also covers descent from solar maximum to solar minimum.
Both those things have a marked impact on a period of time only about ~9 year long. But they become irrelevant over a timespan of say 20 or 30 years.
Correct as regard single solar cycles but the background solar signal arises from slow changes in the average level of multiple solar cycles for up to 500 years at a time (MWP to LIA to date).
By affecting cloudiness and albedo the sun changes energy input to the oceans which over time skews the relative strengths of El Nino and La Nina from above for long term net warming or net cooling.
Additionally the oceans seem to have their own internal variability on the 60 year timescale which also skews the El Nino / La Nina balance from below.
So, limiting the relevant time period to ~9 years is erroneous.
Where ~9 years could be significant is as regards the time it takes for the energy from ENSO events to filter to the Arctic Circle. I say that because the highest peak of El Nino warming during the late 20th century was 1997/8 and Arctic ice melt reached a maximum by 2007.
You are still ignoring the frequency and intensity changes in ENSO on decadal scales in a cool Pacific Decadal Variation Mode.
No I’m not. That is covered by the internal oceanic variability on the 60 year timescale.
Assuming that’s aimed at me and not Stephen Wilde, I am not ignoring the frequency and intensity of ENSO changes. We are already near maximum of La Ninas. We’ve gone from near maximum El Nino intensity and duration to near maximum La Nina intensity and duration in a decade. That has had a cooling effect over the period (about 0.1C cooling) on hadcrut, helping along with the Sun, to keep hadcrut flat over the period
Will ENSO decrease more this decade and cause another 0.1C cooling? I very much doubt it. If the past is any guide we are near as low as it can get. For it to keep dropping would imply by end of the decade we’d have to be pretty much in permanent La Nina conditions.
And overriding all of this is that eventually ENSO is going to rise again, adding all that warming back, potentially resulting to a 10 year period of 0.3C warming.
Hello lolwot,
Given that the negative PDO is 30 years long and that the switch is supposed to have been around 2003 I think you are a little out. Furthermore the fastest rate of cooling is in the second period of fifteen years. We haven’t seen anything yet.
Note too that we no longer have significantly increasing ocean heat content and global albedo has increased to reduce solar shortwave into the oceans so future El Ninos are going to get weaker.
Since the LIA there has been a background warming from one period of positive PDO to the next. That ran in parallel (with some aberrations) with the increase in solar activity since around 1600.
The recent solar minimum has been so deep and so extended that we could now be in the process of switching to 400 or 500 years of weakening El Ninos compared to La Ninas. So yes, by the end of a decade or two we could well be in a period of heavily dominant La Nina.which could go on for hundreds of years.
Invest in blankets.
Hello lolwot:
PDO shifted around 2003 so we have a way to go. We haven’t seen anything yet. It is a cumulative effect with sizeable lags so we are just on the turn.
Ocean heat content is no longer rising and global albedo is higher so El Ninos will keep getting weaker.
The recent low level of solar activity suggests a longer term decline in El Nino intensity possibly equivalent to the decline from MWP to LIA. Potentially 400 to 500 years of weaker El Ninos to come.
Verdon and Franks (2006) used ‘proxy climate records derived from paleoclimate data to investigate the long-term behaviour of the PDO and ENSO. During the past 400 years, climate shifts associated with changes in the PDO are shown to have occurred with a similar frequency to those documented in the 20th Century. Importantly, phase changes in the PDO have a propensity to coincide with changes in the relative frequency of ENSO events, where the positive phase of the PDO is associated with an enhanced frequency of El Niño events, while the negative phase is shown to be more favourable for the development of La Niña events.’
We are talking here of 20 to 40 year ENSO modes – cool and warm Pacific Decadal Variation modes. The PVD added to warming in 1976 to 1998 and produced non-warming since. Indeed most of the warming happened in 1976/1977 and 1997/1998 – it is easy enough to verify for yourself. As we are in a cool phase since 1998 – a decade or 3 of cooling is on the cards. You are neglecting the scope of ENSO variability. You are also neglecting ENSO cloud feedback as a radiative forcing – not merely an exchange of energy between oceans and armosphere..
There is an 11,000 year reconstruction of ENSO here in Fig 5 – http://www.clim-past.net/6/525/2010/cp-6-525-2010.html
We don’t know where the Pacific variability is going.
PDO has been getting weaker for 30 years
http://www.woodfortrees.org/plot/jisao-pdo/plot/jisao-pdo/mean:60
PDO is now very close to the minimum of the last 1950-1970 cool cycle, it simply isn’t going to get much lower at all. This is reflected in recent ENSO frequency. We are not going to reach a point of a permanent La Nina – that didn’t happen in the past. Recent La Nina frequency has been quite high. I don’t see it getting much higher.
As an indicator of ENSO frequency it’s impact on global temperature is immediate. Meaning there is no lag time. That means the bulk of cooling from ENSO frequency changes associated with the PDO has effectively already taken place.
Sure, I’ll play your strawman argument. For the record, I stated, “Have we found one that could be called almost accurate? Just one?”
Notice how I didn’t state, “right on target.” or any other words that carried that meaning?
Now, instead of casting dispersions my way with your innuendo, you could have produced a reasonably good model to show just how “foolish” and “naive” I was. But you didn’t. But, that’s only because you can’t. Show me one that predicted the decade of flat-line temps. Heck, show me one that doesn’t predict thermarmageddon! Show me one that has provided any utility to mankind other than showing us how not to model climate!
You know what’s foolish and naive? It is believing we can model something we clearly don’t understand. What is disingenuous is making other people believe we can. Which one are you?
The average annual global land-ocean temperature anomaly(base 1951-1980) was -0.27 in 1988, the year Hansen made his forecasts, and 0.63 in 2010, the most recent year for which an annual average is available.
http://data.giss.nasa.gov/gistemp/graphs/Fig.A2.txt
Sounds like you believe no forecast at all is best. What might that be? No long-term change in global temperature?
You want just one “almost accurate” forecast? OK, lets look at Hansen’s three forecast of average global temperature for 1988-2020. I sure wish I had stock market forecast as accurate as his temperature forecasts have been so far.
Hansen’s three forecasts of the temperature anomaly for 2010 were: 1.1 for Scenario A, 1.0 for Scenario B, and 0.6 for scenario C. (See Figure 3 in the following link.)
http://www.skepticalscience.com/A-detailed-look-at-Hansens-1988-projections.html
The forecast for Scenario C is on-target, but the Scenario has been criticized for being based on an unrealistically low assumption about future CO2 levels.
The forecasts for Scenarios A and B are higher than the observed temperature anomaly for 2010. The observed increase in the anomaly over the 1988-2010 period was .90 (-0.27 to 0.63 = .90), while the forecasted increases were 1.37 for A and 1.27 for B.
Are forecasts that overstated the observed 12 -year gain in the temperature anomaly by one-third to one-half good forecasts? I think they would be good compared to a forecast of no change in the anomaly over the12 years, and good compared to a forecast that understated the gain by more than one-half or overstated it by more than one-half.
The target year for Hansen’s forecast is 2020. So we shall see how the scenarios turn out in the end.
suyts, in my previous post the first paragraph and it’s link should be the third paragraph. I apologize for the mistake. I’m tired and it’s past my bedtime, so I wont be responding further soon.
lol, is ok M. I responded without having my coffee quota so it may look a little disjointed to you. That said, what I stated below is still valid. If I had a model that stated there would be no warming….. in other words a temp anomaly of zero…..I’d be just as correct as Hansen…… unless you do the rationalization wizardry you’ve produced. But, in your apology for Hansen, you nailed it……..
“The forecast for Scenario C is on-target, but the Scenario has been criticized for being based on an unrealistically low assumption about future CO2 levels. “ ———- agreed. Why is that do you suppose? Did we quit emitting CO2 or other GHGs? No, they continued to rise in a exponential manner….albeit, almost lineally.
What’s odd about you guys trotting this study out and stating it was essentially correct is that Hansen has already listed his thoughts as to why it was so wrong…….. (China’s cooling coal and aerosols.) You’re defending the work of a guy that no longer defends this particular effort. Funny stuff.
There are two errors in my post: (1) As Bruce pointed out elsewhere in this thread, the 1988 global temperature anomaly is 0.31, not – 0.27(which is for 1888). (2) Because Hansen developed his forecast scenarios before the 1988 anomaly was available, the base year for his forecasts is 1987,which had an anomaly of 0.27. I will describe the corrections I have made.
The observed temperature anomaly rose from 0.27 in 1987 to 0.63 in 2010, a gain of 0.36.
http://data.giss.nasa.gov/gistemp/graphs/Fig.A2.txt
Scenario A forecast the temperature anomaly would rise from 0.27 in 1987 to 1.10 in 2010, a gain of 0.83, which is more than double the observed gain of 0.36. Scenario B forecast a rise to 1.00 in 2010, an increase of 0.73, or about double the observed gain. Scenario C forecast a rise to 0.60 in 2010, which is slightly below the observed 0.63.
http://pubs.giss.nasa.gov/docs/1988/1988_Hansen_etal.pdf
You correctly note that a no-change extrapolation of the 1987 anomaly would have been just as accurate as the Scenario B forecast, since it would have understated the observed gain by 0.36 while the forecast overstated it by 0.37, and of course more accurate than Scenario A which overstated the observed gain by 0.37.
No-change extrapolations, however, may not be very useful for decision making. Consider a stock price forecast like Hansen’s A, which predicted the price of the stock would rise from $27 a share in 1987 to $110 in 2010, compared to a forecast of no change in the price, and the price actually rose to $63 a share. From a decision making standpoint, which forecast would you find the most useful?
Hansen’s Scenario C was predicated on “drastically reduces trace gas growth between 1990 and 2000 such that the greenhouse climate forcing ceases to increase after 2000”.
No such thing happened.
Depends on how one bets it. If most believed it would get to $110, then they would have held too long, and would take a bath. Sorry M. but that’s a horrible analogy! Luckily, I’m here to help you refine your arguments! Typically, no one makes money on stocks until they cash in.(Dividends notwithstanding.) So, under your analogy, we’ll have people with money tied up in stocks until hell freezes over because they’re betting on $110.
On an aside, the market doesn’t change, anymore. And it won’t until something significantly changes. I won’t go into the dynamics here, but here’s a tip…for example, the NYSE,…when you see it dipping below or near 10,000…..jump into a mutual fund betting across the board. Wait for it to move to about 12,000….. move to something safe…… wait for it to dip again….. rinse and repeat as necessary. Keep an eye on inflation as it may raise the bar a bit, but not the dynamic.
So where does it go now?
http://www.realclimate.org/index.php/archives/2009/07/warminginterrupted-much-ado-about-natural-variability/
0.4 degrees C difference by 2015? You don’t have all that long to wait – hope you make it.
M.carey,
You completely baffle me. You do a bit of mathematical jiggery pokery, and at the same time point us to Gavin’s graph that clearly shows that actual temps are running below Hansen’s scenario C. Scenario C “assumed a rapid decline in greenhouse gas emissions around the year 2000”. So, temps are even below what Hansen predicted for a massive reduction of emissions – a reduction that clearly hasn’t happened.
And you’re using this as an example of a good model prediction. Baffling.
M. Carey this is a fruitless discussion that has been played out countless times. The problem is alarmists fail to address reality. And, BTW, a skepticalscience analysis holds no weight for me.
Let’s not play with his manipulated graphs, lets just go to the source…… I believe you can get a copy here…..http://pubs.giss.nasa.gov/docs/1988/1988_Hansen_etal.pdf
I think we can agree scenario B will most closely resemble our GHG forcings we have today. Going with B, we see figure 3a clearly shows the anomaly to be right at 1.1 degrees C…… Yet, reality shows us that we have a current anomaly at 0.5 deg. C. Using GISS’ own data.(Mind you that’s only using GISS’ data, the other temp databases are in disagreement with GISS, and using HadCrut or satellites we would be shown to be in even more error.) He was off only by over 100%!!!! Take that back to skepticalscience.
So, yes, in this case, I would say it would have been better not to have made a predictions that to have exposed his lack of understanding. We passed laws based on a projection that was 100% off. Very nice. I prefer we base our laws on a thing called reality.
That said, I find it entirely amusing that you’d trot out this little example of
errantalarmist science when asked to produce a decent model. M. do you ever wonder why modeling of climate science hasn’t advanced in over 20 years? Because the base assumptions are in error!!! You can’t produce a model more accurate than Hansen’s antiquated musings? As shown, that isn’t a testament to Hansen’s prognosticating abilities, it is an indictment of the models. We’re being asked to raise the debt ceiling? How much have we paid scientists and modelers not to advance our current understanding? Ok, so maybe being unable to produce a better model than Hansen’s isn’t all that amusing after all……….James
M Carey: “The average annual global land-ocean temperature anomaly(base 1951-1980) was -0.27 in 1988, the year Hansen made his forecasts, and 0.63 in 2010, the most recent year for which an annual average is available.”
Actually, it was:
.31 in 1988
.56 1998
.63 2005
.63 2010
http://data.giss.nasa.gov/gistemp/graphs/Fig.A2.txt
So that’s .22C over 22 years and 0 over the last 5.
UHI in China dwarfs those changes.
http://wattsupwiththat.com/2011/07/28/new-paper-uhi-alive-and-well-in-china/
And thanks to NASA we now know that UHI can be 7 – 9C.
http://www.nasa.gov/topics/earth/features/heat-island-sprawl.html
Considering UHI … it is probably cooling.
No. its .32 over 22 years.
Where is the hockey stick, mm? In view of the lack of warming over the last 22 years and the tax money spent in climate, will the Administration investigate to the hockey stick like the drowned polar bears? Billions of dollars were wasted in funding these climate models. The Administration is accountable for every dollar wasted.
M. Carey
Contrary to your assertion, scenarios A, B and C are not examples of forecasts but rather are examples of projections.
Terry, you are correct, but the difference between a forecast and a projection is not a distinction many people appreciate. Obviously, all three of those projections can’t be the best forecast. One will be better than the others, unless there’s a tie.
lol….. I didn’t forecast that someone would say this, but I did predict it…………’cause a projection is way different than a forecast…… Its like a forecast with your fingers crossed…..no?
If it has to do with tempertures (or related by talking points) that oscillate isn’t there good chance of mutually exclusive results? If I tell you the stock market will rise Monday based on a model is my model proven valid if it does rise?
Your view of forecasts seems equally simple minded.
If you tell me the stock market won’t change over the next 20 years, and it doesn’t change Monday, will that validate your method or lack of method
My previous post was intended as a reply to suyts.
Anyway, to get back on track.
The models do not include a specific negative response, namely a change in the speed of the water cycle as manifested by a latitudinal shift in ALL the surface air pressure systems.
Such shifts respond to system imbalances caused by solar variations affecting the vertical temperature profile of the atmosphere from above AND oceanic variations affecting the vertical temperature profile from below.
Shifts of the surface pressure distribution involve changes in all the features of climate that are involved in the speed of energy transfer from surface to space including cloudiness and albedo which in turn affect the amount of solar shortwave that penetrates the oceans to power the system
More CO2 just speeds up the rate of energy transfer to space and in doing so causes a miniscule change in surface pressure distribution for no discernible effect on total system energy content.
Described in tiresome detail here:
http://www.irishweatheronline.com/features-2/wilde-weather/setting-and-maintaining-of-earth%e2%80%99s-equilibrium-temperature/18931.html
The scenario I describe fits with the observations reported by Dr. Spencer in his paper but not with the model expectations. That scenario provides a plausible mechanism for the observations reported by Dr. Spencer.
I should have made it clear that I am describing latitudinal shifts beyond normal seasonal variability.
Hi Stephen
Sorry for this late comment but I have been on holiday and have been trying to catch up on the many posts I missed.
Have you been over to Dr Spencer’s blog and run your ideas past him? He is usually quite approachable. Well explained hypothesis BTW. Cheers,
Rob.
Thanks Rob,
I’ve linked to it on his site and it has been set out in quite a few locations but as yet no convincing reason has been put forward which leads me to believe it is defective.
One of the reasons I started posting here was because of the vast experience and intellectual firepower of the contributors but as yet none are biting the bullet.
It just needs one person in a position of authority to pick it up and run with or tell me of any fatal flaw.
M Carey per your link giss.nasa:
1988 .31 .25
2008 .44 .55
2010 .63 __
That is about .32 or about .30 for twenty years.
From your Hansen BAU scenario (note to readers from Hansen 1988, not the misleading work linked):
Hansen states: “Scenario A assumes growth rates of trace gases typical of the 1970’s and 1980’s…assumed annual growth of about 1.5% of current emissions. ”
The rate for the world according to the Earth Policy Institute from 1990 to 1999 was about 0.8% per annum, and from 2000 onwards about 2.4 per annum. This is equalivalent to 1.7% per annum.
If you wish to claim their is something wrong with his modelling then do so. But, don’t try to confuse people. Our CO2 actual emissions are greater than what Hansen used. Note Hansen used the wording growth of 1.5% current emissions,
Even if you knew exactly what the forcings were at the end of the last glacial maximum, you still could not say with any certainty that this can be correlated to today’s sensitivity. You would have to determine how regions changed over time and determine how those changes affected the regional climate. Looking at irrigation studies indicated that not only does adding water to the system change the temperature but it can change it in opposite directions depending on the latitude. Is this a function of temperature? Does this change over time? How do melting glaciers affect the regional climate in each region and how are the regions affecting once the glaciers are completely gone? This isn’t even going into what the possible differences in ocean and air circulations were. I do recall reading some time back Pielke Sr arguing that we have no idea what a kilometer high wall of ice does to the boundry layer. Paleoclimate is interesting but it is a leap of faith to assume we can make judgements on todays climate from it.
One takeaway from the paper is that climate models show a lag between the initial heating and the increase in LW outgoing. (If I am not understanding the paper, let me know.) Spencer maintains that LW should increase simultaneously with the atmospheric temperature.
It seems he has a good point here. Radiation moves at the speed of light. Why would there be any delay in an atmospheric temperature increase and outgoing LW? This deserves an explanation from Spencer’s detractors.
As far as clouds go, I’ve viewed the video made by the CLOUD experiment researchers and seen the correlation between cosmic rays and temperature. But if you look at a 10 or 11 year moving average of the temp and neutron count, a proxy for cosmic rays, there is not correlation. The 10 year moving average, in effect, show long term, low-frequency effects. So, based on that, it appears the moderation of climate by cosmic rays, and therefore clouds, just isn’t happening.
For long-term, 10 year moving averages of these, see:
http://ossfoundation.us/projects/environment/global-warming/myths/henrik-svensmark/image/image_view_fullscreen
The 10 year moving average, in effect, show long term, low-frequency effects. So, based on that, it appears the moderation of climate by cosmic rays, and therefore clouds, just isn’t happening.
Jim – have you seen this article: http://climaterealists.com/attachments/ftp/Climatechangeisdominatedbynaturalphenomena.pdf showing an apparent close correlation, over the last 100+ years, between the time-integral of the sun-spot number, the effective SST, and temperature? What do you think about this work?
I have more questions than answers and, unfortunately, right now I don’t have time to give your link the attention it deserves. At the moment, my only point is that I don’t understand why there isn’t a better wiggle match between the 11-year moving averages of neutron count and global temperature.
I’d guess that short term solar effects are dwarfed by ENSO variability. The solar influence is most clear on the timescale of MWP to LIA to date.
Although solar variations seem to affect cloudiness which then affects the energy available for ENSO the process is very slow. We are ten years into the quieter sun and negative PDO yet the North Atlantic remains warm and the expected cooldown has not really dug in yet.
ENSO and the Pacific Multidecadal Oscillation (60 years or so) seem to drift in and out of phase with solar effects hence the problems of correlation (or frequent absences of correlation).
WuWT has a very relevant new post, on a new GRL paper entitled
The importance of the deep ocean for estimating decadal variations in the Earth’s radiation balance
http://www.agu.org/pubs/crossref/2011/2011GL047835.shtml
http://wattsupwiththat.com/2011/07/31/what-do-sea-measurements-reveal-about-earths-temperature-trend/#more-44407
Sure the deep ocean can hide a lot of heat, but heat can’t be transported there quickly. Trenberth was right years ago in that famous NPR interview: the missing heat has been re-radiated to space.
=================================
More likely it never got into the oceans because albedo increased when cloud quantities started to increase again around 2000.
Yep.
===
How much is presently known of the changes in heat content of various parts of the oceans? Changes in the total heat content are discussed in many papers, but what is now known about its distribution by geographic area and depth?
Some results from Levitus 2006 are shown in AR4 WG1 Figure 5.3, but we should know now a bit more. ARGO has collected data down to 2000 m. That should already add to the knowledge.
Many of the papers discussed here recently contain arguments that should be at least partially confirmed or disconfirmed by such empirical data. Comparing the model results with any empirical data from the deeper ocean should tell more about the validity of various hypotheses on the role of the deep oceans than data that is related to the issue only through overall balance. What is seen above -2000 m should tell also about the possible role of greater depths.
I haven’t been following the literature enough to know, what has been done with the data, but the recent papers have made me wonder.
And Argo data doesn’t show deep transport of the heat. I believe the geographical distribution of Argo is broad enough to have found that transport if it existed.
==================================
Technically, that would be counterproductive. Admitting the truth is totally un-scientific. The truth must be proved. Repeatedly! No matter the expense… or the seeming insanity of those involved in the process.
PS: Besides the “truth” issue, we can NEVER have too much data.
My last intended to be in reply to EJ | July 30, 2011 at 4:37 pm
Your post is excellent, Professor Curry. Those who are relatively new to the debate could do no better than taking your post as their starting point. Could you please expand one point, if you have not done so above? You write:
“But drawing inferences from such studies regarding feedbacks and sensitivity invariably leads to disputes because of the simplicity of the models and assumptions that are used and the fundamental fact that diagnosing feedbacks in the complex climate system can’t really be done.”
Do you mean that it cannot be done in principle or that it cannot be done now?
Would you please provide a reference or two to articles or posts by you on the topic of models and the sort of empirical work discussed here?
Theo, here are a few links on this subject:
http://judithcurry.com/2010/12/29/climate-feedbacks-part-i/
http://climateaudit.org/2010/01/18/curry-reviews-lindzen-and-choi/
http://judithcurry.com/2011/02/03/nonlinearities-feedbacks-and-critical-thresholds/
Many thanks to you, Saint Judith.
Dessler (and now Trenberth) keeps coming up with the line that “clouds don’t cause ENSO” as a rebuttal when nobody actually said such a thing. It’s a base strawman attack. When two things correlate, one thing could cause the other or a third thing could cause both. In this case the third thing is natural variation. We sure as shooting know that CO2 doesn’t cause ENSO so afaik it’s perfectly acceptable to accept natural variation and deeply simplistic to dismiss all this hard work by a totally unscientific one liner.
All this kefuffle about a observational finding. Yet why does the real mainstream not correct Trenberth when he tells us that a 4% increase in water vapour causes droughts, floods and tornadoes and presumably Jet Stream shifts? On the face of it that is a much less supported idea (not even worthy of the title hypothesis) than the idea that the climate models are unfit for purpose. Yet he gets away with it time after time! Neither he, Schmidt or Dessler are mainstream as they claim but rather they are part of a radical climate clique who ritually imbue CO2 as the main climate control despite an ever increasing pile of evidence to the contrary.
And of course the last line of defense, in this case by both Schmidt and Trenberth, is to claim that the data are wrong and the models are right. Well as they are both climate modelers why wouldn’t they say that? It’s only arrogant faith in their own shamanism and is a mockery of proper modeling endeavour, which MUST be validated by hard data.
Finally, as is well known, the paleo data don’t tell you anything about climate sensitivity without using many unsupported assumptions and cherrypicking of the dataset you prefer.
Your entire post is well worth reading. You write:
“Dessler (and now Trenberth) keeps coming up with the line that “clouds don’t cause ENSO” as a rebuttal when nobody actually said such a thing. It’s a base strawman attack.”
In addition, no one knows the relationships between cloud behavior and ENSO. As for the Gaia Modelers, they posit only processes of radiation and treat natural processes such as ENSO as “emergent phenomena.”
More sunlight into the oceans provides more energy for ENSO. Less energy in then less for ENSO.
Therefore the strength of El Nino events is directly affected by global albedo and cloudiness.
Clouds do not cause ENSO but they affect the amount of energy available to it.
ENSO is caused by the fact that the mean position of the ITCZ is north of the equator. That causes an imbalance of solar shortwave into the oceans either side of the equator and periodically the imbalance becomes large enough to fuel an El Nino event.
Less clouds weaker El Ninos, more clouds stronger El Ninos. However it is global not regional cloudiness that counts. El Ninos give more clouds in the ENSO region but less clouds globally and vice versa for La Nina.
Then there is the top down solar effect too.
When the Southern Annular Mode is positive – storms spinning off the polar front push further in higher latitudes. The circumpolar flow through Drakes Passage is reduced. Cold southern ocean water piles up off the coast of South America and upwelling in the region of the Humboldt Current intensifies – initiating La Nina. When SAM goes negative – the cold tongue slows, temps in the ocean surface increase and the trade winds falter – initiating an El Nino in which warm water piled up against Australia and Indonesia flow eastward across the Pacific.
Solar UV heats ozone in the middle atmosphere and is significant source of warmth and variability above the Antarctic (and Arctic).
http://www.osdpd.noaa.gov/data/sst/anomaly/2011/anomnight.7.28.2011.gif
You can see in the map cold upwelling still in the south east Pacific. The conditions for a La Nina are still in progress. Once the ITCZ moves to the north in the SH spring – La Nina is odds on of reforming stronger than last year. ENSO is a SH summer phenomenon.
There is a similar process in the NH – with a different oceanic morphology. I would link it physically to the PDO.
ENSO is a southern hemisphere summer phenomenon. SAM responds to temperature in the middle and upper atmosphere. Colder and air densities increase. So it tends to be more positive in winter – which is when the polar fronts most commonly have landfall in southern Australia and Africa, New Zealand and South America.
Complex system I believe – where solar UV pushes systems past La Nina threshold and then feedbacks of wind, cloud, currents and waves kick in. Upwelling in the Humboldt Current is the ENSO switch.
In a La NIna – clouds form over cool seas and are carried on the trade winds to create rain in Australia, Asia and Africa. In an El Nino cloud dissipates over warm water – but the sub-tropical jet streams carry high cloud and moisture over north and south America.
La Nina result as a global average in increased reflected shortwave – strongly in the tropical Pacific – and vice versa for El Nino.
Thanks CH, If there is a substantive flaw in what I said then I’d like to know but you need to be a bit more specific because I can’t see the point you are driving at.
Is your description inconsistent with what I said and if so how?
Do you have a specific question? I provide a fuller description of the physical aspects – not merely waffle about clouds.
My ‘waffle’ was a cut down explanation as to how changes in cloudiness could affect the energy available to the ENSO process.
How does your ‘waffle’ address the point?
I was also suggesting that the sign of the global effect of El Nino on cloud quantities is opposite to the regional effect because El Nino drives the mid latitude jets poleward.
Is there anything in your ‘fuller description’ that is inconsistent with what I said?
I think that all the events you describe could well follow on from the forced changes in surface air pressure distribution caused by the interaction of top down solar and bottom up oceanic influences.
You have theories with no physical basis – no mechanism – no observations. Nothing to connect a narrative that appears full grown like Athena from the mind of Zeus.
Cloudiness is an ENSO feedback – observed from changes in the eastern Pacific. The satellite data shows decreases in cloud to the late 1990’s and an increase since – corresponding to the decadal shift in Pacific climate that occurred in the late 1990’s.
What do you mean by ‘top down’ modulation? It is just empty words.
Steven,
“I think that all the events you describe could well follow on from the forced changes in surface air pressure distribution caused by the interaction of top down solar and bottom up oceanic influences.”
Land mass bottom up is very significant and may well be greater than oceanic influence during the daytime. Land mass is usually at a higher temperature than ocean water during summer and in winter ocean water dominated the bottom up. Specific heat capacity of water, is huge compared with any substantial materials on land, rules the winter bottom up LW radiation.
That’s all nicely said James. The NYT’s keeps quoting Trenberth as if he’s solidly in the mainstream. It’s infuriating to see his wild attributive claims passed on to unsuspecting readers as near gospel.
It’s becoming increasingly clear as the drumbeat of legit scientific doubts continues to intensify, that this is really not about the science within the liberal media. This greatest of “pseudoscientific frauds” in the words of the late Harold Lewis will only be exposed when enough courageous scientists like Dr. C. stand up and say “Wait a minute. There are problems here.”
I have no doubt it will happen, especially as the climate continues to cool over the next decade. At a certain point, some of these people will have to realize that they’d better get on the right side of history before it’s too late. In the next few years hopefully, there’s be a sudden rush for the exits which the NYT’s is going to have to report on whether they like it or not.
Meanwhile, this ongoing food fight between the alarmists and skeptics will get us exactly nowhere. We yell at them, they yell at us, and we only become more entrenched, more self-righteously angry.
Without my and others seeing the data it is difficult to judge whether the value that is sought is clearly encoded in the data.
In particular it would be useful to see the covariance (dispersion) matrix for the observables T and Net Flux (SW+LW).
If it is clearly encoded then the assumption of the simple model they pick should allow the conductance (dF/dT) that they seek to be read from the matrix directly, it is not necessary to pick a mixed layer depth.
Simaraly the question of whether the variations in the net flux are driven primarily by fluctations in non-radiative or radiative processes is decidable according to whether the covariance of lagged flux against temperaure and lagged temperature against flux are positively or negatively correlated. A strong correlation either way would add confidence in the result.
Similarly the required thermal constant (effective thermal mass of the ocean) would be encoded in the time constant required to give rise to the covariance of lagged temperature against temperature. This could be decoded by multiplication by the conductance.
There are circumstances that would favour these parameters being clearly discernable from the data. If they are lucky this may be the case but in many cases they will not be clearly present in such a short histort.
Alex
Alex – How would the principle you describe address positive feedback? Specifically, if an unforced temperature change induced feedbacks that induced a flux change that then resulted in further temperature change, and if much of the total temperature change was a consequence of that flux change, then temperature would lag flux change even if no forcing were involved? Can this be distinguished from a scenario initiated by a flux change (i.e. a forcing)?
Hi Fred,
I am not really sure what you mean by unforced change in temperature as I think that enrgy must be required. I think that radiative (cloud, RH, albedo, etc.) and non-radiative (release of energy from the oceans or similar) covers most eventualities. If you mean a change in temperature by some arbitary realese of energy not due to an initial change in the radiative balance that would be covered by the non-radiative class.
That energy and its resultant temperature response and that effects of that, the short term responses (the paper uses monthly averages so that may bundle up initial forcing and RH and cloud changes feedbacks into the average) would not be distinquishable, only the effective conductance into space once these short term effects have played out would be captured. That assumes that these responses or feedback effects have timescales smaller than a month.
I think the only figure that one might hope to pull from the data would be the combined effect of the initial forcing plus the feedbacks arising.
I think that in general all perturbations are born equal in that scheme of things, with the exception that one might get a much tidier data set if the forcing were primarily non-radiative. When wishing to know something about the radiative aspect it seems easier if that is not the part of the system that is generating the purturbations.
If one required the separation of the initiating flux from the fedback response would would need to be averaging over perhaps an hour not month.
I am not sure if that helps you as it is not what you asked. I cannot see how one can change the temperature without a flux of energy from somewhere.
I think it implies that the system jumped suddenly to a far from equilibrium state. That would I think be equivalent to jumping far from equilibrium as achieved by a sudden change in the albedo for instance but that would be the same as any other radiative forcing. I think that in general all bases are covered and the system plays itself out in the same way irrespective of the reason for a lost of equilibrium.
BTW, the last from Isaac Held was well worth the read as was the background from a paper by Gritsun and Branstator. Given my choice of how to comment here, it is not coincidental that it deals with a lot of the issues raised, e.g. dispersion and response matrices and the fluctuation dissipation theorem that underpins all this.
Theirs and Held’s were about the most interesting contributions that I have read in a while. Climate paper’s differ tremendously, particularly in terms of depth and required background. A consideration for which thankful.
FWIW I have a suspicion that the data used in the S&B paper is pretty unhelpful and that the nice behaviour required to determine the parameters is not present or someone would have done it.
Alex
Alex – By mistake, I responded to you as a “reply” to Carrick above, at Comment 92429. If you want to respond, you could do it there, but perhaps here would be just as good, since the above link will let readers see what I wrote. Sorry for the confusion.
Fred,
By mistake …
See here for my reply:
http://judithcurry.com/2011/07/30/spencer-braswells-new-paper/#comment-92488
Fred,
Thanks, I am inconsistent when using these terms but then so many are.
Let me call these wiggles about a mean that do not move the mean fluctuations and things that would move the mean forcings.
I really do not think it matters much how the fluctuations occur provided that both T and net flux are undergoing perturbation and relaxation. Providing one looks in the right place for the result.
The issue comes about if one looks at ratios in the non-lagged covariance where one would need to know which caused the other.
If one looks at the lagged covariance the cause and effect issue is long gone and one is looking at the way the equilibrium is restored, the relaxation.
The relaxation is not so concerned about it got itself to where is. So one can take the ratios on the lagged data. At least with a simple system, like they have modelled, the lagged covariance is blind to the original cause.
That said, their model is simplistic and has one important quality in that the state is totally determined by just the current values and a simple time constant type relaxation. It has no true “memory” in that the process of getting to the current state is unimportant only that it got to that state is. I suspect that a more realistic model, one that more accurately rendered its long term persistence would complicate things somewhat but I believe that only in the lagged data is there hope of bypassing the cause/effect issue.
I think that arguing over cause and effect in cases like these is just that, an argument. As I mentioned before, the analysis should indicate whether the system was driven primarily by radiative or non-radiative perturbations and should be decidable and not particularly relevant.
I also suspect that there may be issues due to a real ocean having an “in phase” component to its flux, i.e. the flux and temperate not being precisely 1/4 cycle out of phase which is the case for the slab ocean. This would show up as an additional conductance parallel to the conductance into space and could result in over estimating the conductance (under estimating the sensitivity). I would have to think about it a bit more but right now it is a concern I could not dismiss.
Alex
My:
“This would show up as an additional conductance parallel to the conductance into space and could result in over estimating the conductance (under estimating the sensitivity). I would have to think about it a bit more but right now it is a concern I could not dismiss.”
I have had that think, and it does not seem to be an issue in this case. It does effect the form of the response function but it should not affect the ratio that gives the conductance.
Alex
Alex – Thanks. I’ll have to ponder your comments further, along with some effort to sort out in my own mind how to resolve the difference between on one hand, the relationship of a forcing (imposed flux perturbation) to a temperature change, and on the other, the relationship between a total flux change and a temperature change when the total flux change also includes feedback effects.
Fred,
This is not for you but encouraged by you.
It is my view that the feedback metaphor like the greenhouse metaphor has taken on a life of its own.
This becomes apparent to me when positive feedback is discussed or thought about.
There are numerous real world occurrences where positive feedback occurs, where one of the results of a stimulation is a fresh occurance of that stimulus, the first punch in a public disturbance.
There are other processes whereby a system tends towards a condition that helps promote just those aspects of the system that promote that tendency. The outcome is accentuated but by an adaptation in the system that acts to better “match” the response to the stimulus. An adaptive process of dynamically adjustment towards a better match (positive) or a weaker match (negative).
In the brawl, the originator can quietly leave and the mayhem will continue. There is a positive feedback effect. The original stimulus has been produced as part of the response which goes on to produce a further response, etc..
In other cases the perpetuation of the originating stimulus is absolutely required. The system is not self sustaining. Such a system may evolve to better match that stimulus, positive matching, which enhances the response, or might tend towards a more mismatched state, a negative matching which lessens the response.
This matching response is perhaps the better metaphor for the “feedbacks” in the climate system. One might better think on it as positive and negative matching responses.
Such matching responses cannot give rise to true runaway conditions, they are not self sustaining as they do not reproduce the original stimulus they just match to it in a way that maximises the response. The original stimulus is an absolute requirement.
In our case the original stimulus is commonly the sun and the positive matching response might be the increase in water vapour, a greenhouse gas, with temperature. This common “feedback” is perhaps better seen as a dynamic positive matching response of a passive but adaptive system. It improves the match between the system and the stimulus. The original stimulus is not reproduced, as in the brawl, nor amplified, nor modified in any way. Nothing has actually been fed back. The system has simply moved into a state whereby its repsonse is more sensitive to its stimulus.
The mathematics are ambiguous on this point. Or rather the are open to both metaphors. They can be seen as representatvie of a feedback loop, which has tranfer functions (amplifiers) and connections between outputs and inputs. Or they can be seen as representing a passive but adaptive network where the impedences of the components vary according to the state of the system.
In many cases how one reasons about a system is deeply coloured by a personal choice of metaphor. I prefer the response matching metaphor but I may be almost alone in this. Although joined by a mathematically commonality they are seperated symbolically. I have no need to look for “amplifiers” and their inter-connections. I need not fear self sustaining behaviour. That said the commonality does run deep and various instabilities are common to both systems but the difference between the active feedback metaphor and the essentially passive adaptive response matching metaphor does materially effect reasoning. I can rest assured that should the sun go out the system’s response will be to cool even if just prior to that moment its development resembled a runaway condition.
I like my metaphor, it is just that a metaphor, but I find it useful.
I contrast the chain of events whereby the sun gets brighter, a greater energy flux impinges on the surface of the oceans,the oceans warm up, the atmosphere gets more humid, this increases the flux impinging on the ocean, etc., with my notion that warming increases the impedence (reduces the conductance) from the surface via the atmosphere into space. This is just as true, quicker to say, and has no notion of amplification or interconnection, there is no invocation of some metaphoric wiring diagram. Significantly it is what actually occurs, an observable (at least in theory) the impedance is modified. The system its observables adapt in a passive way, this is not really the case when it is viewed using the feedback metaphor.
I have tried to encourage people to see it this way but I have found a great deal of resistance particularly from professional climate scientists who seem to be deeply wedded to the feedback metaphor. Perhaps I have failed to get across the fact that the mathematics is the same. It is not that they expressed a preference but that they insist I be wrong headed.
The significance of ones choice of metaphor particularly in the context of what is observable should be neither exaggerated nor downplayed. In one view the conductance, an observable, is a product of a complex feedback network, in the other it is a simple physical property of the atmosphere that is determined by its constituents. They amount to the same thing, but in the first case the meaning of the observable is bound up in a network of theoretical relationship, in the other its just the physics of the gas in the column and retains its meaning even if that column were isolated from the system. So it comes down to whether one wishes to think of observables as a property of some particular reference system, or as something that is meaningful in its own right, irrespective of whether or not they are considered to be part of a system.
I may have overstated the differences by way of emphasis, but they do exist. The feedback metaphor does lead people to expect to find amplifiers, power supplies, and loops in which energy or temperature flows. These are ghosts conjured by the metaphor, so perhaps I merely choose to exchange a prefered set of apparitions for a deprecated set.
Alex
Alex – As always, your comments are thoughtful, provocative, and well-informed. I’m inclined to like the feedback concept as a climatologic principle because it allows specific parameters to tell us how the climate will respond to a stimulus, and because I can visualize the concatenation of events that follow such stimulus. In particular, I find a feedback factor f useful in describing feedbacks in accordance with a Taylor series in which a stimulus is followed by a fractional response f, which induces a response to the response, f^2, and so on iteratively, so that the original “no-feedback (or Planck only) response is multiplied by 1 + f + f^2 + … f^n…., yielding the parameter 1/1- f. As long as f 0, it is characterized by amplification. I find this not hard to intuit.
The concept also allows us to see an unstable “runaway” climate as a circumstance where f can exceed unity. In the case of increased solar irradiance, this can reach a value exceeding the Kombayashi-Ingersoll limit beyond which the Planck Response can’t prevent the interminable amplification of warming until all liquid water evaporates (as perhaps occurred on Venus).
My biggest problem with the concept, as commonly employed, is that it seems to imply that a “positive feedback” must inevitably lead to a runaway. This misimpression arises because the Planck Response is not incorporated explicitly into the listed feedbacks in many descriptions but is implicit only in the assignment of a value to f below unity. If the Planck (Stefan-Boltzmann) response to temperature change is also visualized as participating, it is clear that the temperature response can converge to a limit. I would like to see the Planck Response routinely listed as a negative feedback, in which case our current and foreseeable future climates will almost certainly be characterized by net negative feedback even in the presence of an amplified temperature response to CO2 forcing.
In the above, the vagaries of html eliminated the “greater than” sign, for the statement that was intended to mean “as long as f is greater than 0”.
The confusion is that many in the engineering community regard a positive feedback as influencing the original forcing which can lead to a runaway. In the climate system this would occur if adding CO2 somehow leads to an equal or greater amount of CO2 coming in via a feedback (like the microphone-speaker sound feedback). Note there is an element of this direct feedback through Henry’s Law where a warmer ocean will release more CO2, but that amount turns out to be much less than what was put in to create the warming (I figure 10 ppm per degree). The CO2 feedback is therefore less than 0.1 C per degree warming making it much smaller than the H2O vapor feedback.
So many positive feedbacks (CO2, H2O, albedo…). What are the negative ones? What is the net feedback?
This CO2 warming hypothesis will fail spectacularly. I wonder what the believers will say then about the confusion.
As Fred says, the warming itself can be considered a negative feedback as it increases outward IR enough to cancel these other effects.
That’s not a valid argument. Stability depends on the sum of all feedbacks that increase the radiative imbalance. If warming by one unit leads to additional radiative imbalance that causes additional warming by more than one unit, the result is runaway warming, which stops when negative nonlinearities are strong enough to force the feedback strength again below the stability limit.
The same condition can be expressed also in terms of the radiative imbalances: If the warming caused by one unit of radiative imbalance causes more than on unit of additional radiative imbalance, then we have the runaway situation.
To be clear, I was distinguishing between direct and indirect feedbacks, because I know many are only familiar with direct feedbacks. Of course, I agree that both can have runaway effects.
How do you differentiate direct and indirect feedbacks?
In my view the feedback through CO2 is not more direct than any of the other commonly mentioned feedbacks.
Pekka, a direct feedback is one that changes the CO2 level in this case, which is not what climate feedback does (primarily). An indirect feedback is one that amplifies the response to the increase in CO2 without amplifying the CO2 increase directly, which is a description of the climate feedback.
That I thought as your message, and that’s, what I consider totally misleading. It’s relevant only, if our ultimate worry is CO2, not warming.
Fred,
Thanks I pleased that they are provocative. I pay back the compliment in that I think it worthwhile.
Most of what I was saying relates to the metaphor which colours the mental model that we hold. The mathematics are the same but the thinking is directed differently.
A lot of this comes down to equivalence. That a system modified by feedbacks has an equivalent system that does not have feedbacks but has modified parameters.
In the world of control systems or op amps, the real system has clunky great components and feedback loops but viewed as a black box it has an equivalent virtual system that behaves in the manner of the equivalent system.
In the case of a system like the climate this equivalence is still useful as a concept but the roles are reversed. It is the virtual system that has the active units and feedback loops and the equivalent system that is real. The real system is essentially passive (has no active components equivalent to op amps) but capable of a complex response.
A thing to note is that this is an equivalence, a thought process that works in the feedback view works in the non-feedback view. The choice is one of convenience and to swap from one to the other may be insightful.
I am not suggesting people abandon the 1/(1-f) stuff but look at its equivalent to see if that helps.
Looking at 1/(1-f), the first thing that strikes one is the (-) sign which comes about from the need to think that the effect is “positive”.
In origin we have T/F = R (temperatue flux and thermal resistance)
R = 1/G (thermal conductance)
T/F = 1/G but G is additive so G = G1+G2+G3
T/F = 1/(G1+G2+G3) which is nicely symmetric (no (-) signs)
We can choose to break this symmetry by singling out one of the Gs
T/F = (1/G1)/(1+G2/G1+G3/G1)
T/F = R1/(1+A2+A3) An= Gn/G1
so the fs are negated ratios of conductances. It is just that the conductance representing “positive feedback” has a negative conductance.
fn = -An
Now in the real world negative thermal conductances would break a certain fundamental law and heat would flow from cold to hot. This is why the real system cannot runaway unboundedly. So what is going on:
These are not real Gs but the value of the differential dFn/dT which can under some circumstances be negative over a certain domain where more than just F and T are varying.
So the same system can be viewed as if there were active components and feedback loops, (the equivalent virtual system) , or as a system with passive componts components with quirky values. The quirkiness of the values reminding one that they are limited to a certain domain in the system space.
I think that the T/F = 1/(G1+G2+G3) is an equivalent form worth remembering. None of the Gs is special were I to pick G2 which has a negative value (all the values being the same as before).
T/F = R2/(1+A1+A3) the denominator would be negative as R2 is negative and A3 would have changed sign (if R1 was positive in the previous example.
so positive value for f3 = -A3 would now represent a negetive feedback in the formula
T/F = R2/(1-f1-f3)
That is trivial but people here and elsewhere do query why it is the Planck conductance that is singled out and it is just a convenient convention with no special physical significance but much practical benefit.
Returning to the black box. Lets say they have invented a black box and inside it is equivalent to a load of special gubbings arranged in an active feedback loop.
Once connected to an electrical supply it senses the amount of energy being dissipated inside the box and uses this to produce an output that is fed back to the input terminals where the supply is connected in a negative sense and acts to regulate the dissipation so that it tends to counteract any tendency to over heat. So it embodies a negetive feedback loop stabilising the dissipation.
I buy one and connect it to a supply and all is well for a while but it starts to get a bit hot and then runs awy and bursts into flames.
They ask me what I did and it turns out that rather than connect it to a source of constant voltage I connected it to a source of constant current.
They inform me that it was functioning normally but the feedback loop was now positive which is why it ran away. I informed them that a device which was reversed its its inclination was not fit for sale.
I returned home not best pleased. I break open the remains of the box and I find a incandescent lightbulb.
The same component due to its non ohmic performance is equivalent to either a virtual system with negative feedback or positive feedback. In the real world it is just a lightbulb but in the virtual world it is a seemingly well defined feedback network but the sense (sign) of the feedback is undetermined.
For W = V^2/R and W = I^2 · R is the variation of R with tmeperature doesn’t alone tell us the sign of the feedback or the usefulness of the virtual model.
Does this mean that there is no active feedback equivalent? No, it is the “seemingly well defined” bit that was the problem.
Those people who dislike, distrust or think my alternative metaphor is a deceit have possibly failed to read the chapters on the equivalence of feedback and non feedback networks and of active and passive circuits.
By way of an example every potential divider is an amplifier whose gain equals B/(A+B) where resistor A is connected to the input, B is connected to ground and their junction is the output. The trick is to make A negative if one wants a gain greater than unity or B negative if one wants a gain less than zero.
Similarly one can simulate a negative conductance by active components in a feedback loop. One does so when one constructs an oscillator and fools a passive circuit into behaving as if it had negative damping.
Regarding the feedback equation:
It needs to be remember that
F = G · T { · is the multiplication operator}
is a very special idealised case of
F = G * T { * is the convolution operator}
So T = H * F = F * H {where H * G = unity or rather the delta distribution}
We are dealing with response functions and neat tricks like forming
T/F = R1/(1-f2-f3) is not on, one has to find the deconvolution of G e.g the deconvolution of (G1+G2+G3) which does not preserve the individuality of the components.
I tend to use both metaphors but mostly not the feedback one as I do not find it useful when working with response functions which is by default approach.
Your:
“This misimpression arises because the Planck Response is not incorporated explicitly into the listed feedbacks in many descriptions but is implicit only in the assignment of a value to f below unity. If the Planck (Stefan-Boltzmann) response to temperature change is also visualized as participating, it is clear that the temperature response can converge to a limit.”
If I may be so bold. This I should say is a use of the alternative metaphor seeing the components as fully equivalent components of the conductance.
This has been hopefully provocative if not immediately useful or even particularly helpful.
Alex
Thanks, Alex. It’s always helpful to see something familiar in a fresh perspective. It also reminds me how little I remember of the principles of electrical circuits.
One of the biggest problems in using the concept of feedback in climate studies is that the most common feedback formulas are one-dimensional. For a multidimensional problem of a couple of independent variables the simple formula 1/(1-f) becomes a matrix formula. Instead of on parameter of f that leads to instability when f = 1 of > 1, we have to look at the eigenvalues of a matrix. The Earth system has an infinite and continuous set of variables, which is more complex even in the stationary case and may lead to totally differently behavior, when the temporal dynamics is considered.
The concept of feedbacks is likely to be useful to a point, but trying to draw too many conclusions from it, is certain to lead to errors. There is absolutely no guarantee that the approach of SB11 will give valid results until the whole dynamics has been checked and found to behave in a way that allows for this description. The issue can be studied for any complex model, but even after that it remains open, how well it can be applied to the real Earth system.
We have had threads on feedbacks on this site, and many good observations on the role of feedbacks have been presented in those threads, but separating the pearls from the background noise is certainly difficult also in case of those threads.
Schmidt
“Climate sensitivity is not constrained by the last two decades of imperfect satellite data, but rather the paleoclimate record.”
Given the above I wonder if Gavin could explain the apparent contradiction between this statement and his historic and continued support for Mann’s splicing of the paleo record with the modern instrumental record in order to ‘create’ the hockey stick?
The paleoclimatic record he is referring to is not that of the last 2000 years, but that of much longer periods. Mann is not really a relevant reference here, but the presentations of Richard Alley tell more on what the supposed evidence is. This is one of his presentations
http://www.agu.org/meetings/fm09/lectures/lecture_videos/A23A.shtml
I share with many others the doubt that we know far too little on many important factors to justify everything that Alley tells, but there’s also much to ponder.
Dudette,
it is perfectly acceptable to use alarmist as an adjective to describe computer models that have been tuned to scare the cash and freedom out of people.
The paper discussed here by Spencer and Braswell, (SB11) is mainly a rebuttal to Dessler. Dessler has stated in his paper in Science 2010 that radiative forcing during the decade 2000 can be neglected due to the dominating role of ENSO and that feedback may be diagnosed by regressing TOA net radiation versus temperature anomaly.
In mathematical terms both SB and Dessler assume that TOA net radiation is a linear function of the temperature anomaly:
R = F – lam*T + eps
where R is he TOA net radiation, F is the radiative forcing, lam is the radiative feedback parameter, T is the temperature anomaly and eps accounts for internal variability-
Dessler states that F is a constant during the studied decade and that the variation of eps over the decade could be treated as a random variable. Hence regressing changes in R against changes in T over the decade should give lam.
SB disagree and state that F is not a constant over the decade.
Spencer and Braswell have shown already in their 2010 paper (SB10) ”On the diagnosis of radiative feedback in the presence of unknown radiative forcing” how a phase space plot of R versus T may give information on the variation of the sum F + eps (according to Dessler’s assumption this sum averaged over a number of months should be constant since F is constant and eps is random).
In a phase space plot in which F+ eps is a constant all the points would be on a straight line with the slope = lam. This is not the case according to figure 4a in SB10. This supports SB:s claim that Dessler’s assumption of a constant forcing is not valid.
In SB11 the authors support the same claim by showing that there is a time lag between radiative forcing and temperature response both in the satellite observations and according to the climate models. Trenberth and Fasullo (TF) confirm this time lag in their comment on Real Climate but disagree that this means that F has varied over the decade.
However, in SB11 the authors have proven their point by simulations with a simple energy balance model. Although TF claim that SB do not consider the effect of ENSO, SB:s model (equation 1 in SB11) includes a non-radiative forcing S that may well represent the forcing from ENSO on the mixed layer of the ocean (equation 1 in SB11 models the atmosphere and the mixed layer).
In SB11 it is shown that if the only forcing is non-radiative, S, then there will be no time lag. Only if also the radiative forcing F (N in SB11) varies there will be a time lag between TOA net radiation and the temperature response.
Hence, the fact that satellite observations as well as climate models show a time lag between TOA net radiation and the temperature response is another support for SB:s statement that the radiative forcing has varied during the decade 2000 contrary to Dessler’s assumption.
Pehr – What I (and others) perceive to be the errors in the above interpretation is described in some detail in my first comment (the first one in the thread) and the subsequent exchanges of comments. It is almost impossible physically for there to have been significant net forcing variations during the decade described sufficient to account for the magnitude of the temperature fluctuations over intervals measured in months rather than decades . The timing of the flux/maximum temperature relationships tells us very little in a system where major ENSO-related oscillations are changing energy flows in one part of the globe out of phase with other regions, and where a temperature maximum may reflect climate phenomena coincident with a decline in feedbacks that are nevertheless positive. (Note that the paper, in Figure 3, doesn’t describe the timing of flux/temperature relationships overall but rather relative only to a temperature maximum, which can be very misleading) Furthermore, models that can simulate ENSO well can duplicate the results reported in the paper, and at the same time exhibit a high climate sensitivity to CO2. The paper tells us very little if anything about the latter.
Other deficiencies are also described above, but they relate more to inadequate description of methods and statistics than to the value of the conclusions. Even so, I believe the journal referees should have insisted on more details before the paper was accepted for publication, because the Figure 2 data, while not able to confirm the authors’ conclusions, are nevertheless interesting.
I have a slight variation on Fred’s argument. While everyone can agree that after an El Nino max we expect to see radiated energy leaving the atmosphere over a period of time, the controversial part is whether there is actually an accumulation of radiative energy preceding the max. I would say that SB11 has interpreted a negative anomaly in radiative flux as a heating term, when in fact it could be zero in absolute terms, and only appears negative because he is looking at anomalies, which requires removal of a mean value (as I posted before). He has to demonstrate that it really is a warming term of some significant magnitude, which he has not.
Jim – Although I had the same thought initially as you describe, I don’t think it matters whether a flux change preceding temperature maximum tells us that the climate system was accumulating energy or merely losing less. Even a reduced cooling should change temperature in an upward direction from where it was – if it was previously stable, it should rise. Since we know that temperatures do in fact rise during an El Nino, it would be reasonable to ask whether this was a consequence of a preceding flux change in the warming (or reduced cooling) direction. Unfortunately, the paper doesn’t really answer the question, and the correct answer is probably “no” for all the reasons discussed previously here and elsewhere.. Certainly, the “magnitude” you mention wasn’t there.
Yes, reduced cooling does not always lead to warming, but we don’t have sufficient information about how La Ninas fit into SB11’s analysis, which would help us understand these graphs in a more absolute, rather than relative, sense. Did he consider the whole period and minima, and how those could have influenced the flux prior to maxima? That is, are his negative radiative forcings really just a response to previous minima, rather than a lead-in to the next maximum?
Looking again at SB11, he does consider minima too, but my comment still stands that when looking at maxima in isolation, how can he be sure it is not a response to the previous minimum that precedes the maximum. The time lag of 12 months may be suggestive of it depending on the ENSO frequency rather than causing the maximum in some way.
I still think everyone is trying to read too much into the paper. It poses more questions than it attempts to answer. ENSO, as Fred has said is assumed to be an unforced variation, which a good deal of it probably is, but ENSO has variation due to something that also appears to be internal, PDO probably, during the period in question.
One side of that ENSO variation appears to be cloud cover changes. The other side tends to wander, making it hard to nail down. The cloud change has a radiative impact, the less well defined other side likely does as well. The models that include ENSO and have a high sensitivity tend to match the SB results better, but change in forcing assumed to be CO2 sensitivity could very well be forcing changes due to internal variation linked with oscillation which influence ENSO.
So since the models with high sensitivity which Trenberth says match better probably have too high a sensitivity for a longer period, the difference between a best guess sensitivity and the high sensitivity of the Trenberth mentioned models should be in the ballpark of the elusive natural variability that may produce longer term trends.
Fred,
As a modeler myself, although in chemical engineering, I appreciate the clear mathematical logic I find in SB10 and SB11.
The core of their logic is that they express the TOA net radiation as a linear function of the temperature anomaly. This follows from a Taylor expansion and should be valid for small temperature changes.
Then the radiative forcing F must be found in the temperature independent part of the expression. If the internal variability, eps, cancels due to being random the radiative forcing is the only part left.
From the measurements of the TOA net radiation SB10 observe, by using a phase plane plot, that the radiative forcing is not constant over the studied decade in so far the internal variability term really cancels in their averaging of the radiation data.
Note that there is nothing said here about the cause of the changing radiative forcing. That the radiative forcing is changing is solely a consequence of the evaluation of the radiation data with SB:s method of using phase space plots.
I also appreciate the mathematical logic in SB11. They assume that the climate system consisting of the atmosphere and the mixed layer may be forced both by radiative forcing and by forcing from below, non-radiative forcing, which would be the forcing from ENSO.
The model tells them that a pure non-radiative forcing would not cause a time lag between TOA net radiation and temperature response. Only if there is also a radiative forcing there could be such a time lag. Observations show that such a time lag exists, and this has been confirmed by Trenberth and Fasullo.
These are facts telling us that the radiative forcing has changed during the studied decade. In SB10 this depends on the assumption that the internal variability term, eps, cancels on averaging the data. In SB11 it depends on the suitability of the simple energy balance model of the atmosphere and the mixed layer used.
But the same conclusion about a varying radiative forcing has been found with two different methods in SB10 and in SB11.
I think this concerns the main theme of SB11, which is a rebuttal to Dessler’s claim that the radiative forcing has been constant during the decade. This assumption is vital for Dessler’s diagnosing of the feedback from clouds in his Science 2010 paper.
How to explain the fact of a varying radiative forcing is another issue. Is this result reasonable considering what is known about the ENSO? You describe many processes that occur in that context which should influence the atmosphere and the mixed layer in many different ways. Those processes seem to be very complex and interdependent.
Perhaps the complexity of this system with all this interdependence is so high that taking all this together there will appear both a non-radiative forcing acting on the mixed layer and a radiative forcing originating from the ENSO but then also feeding back on it. Perhaps the whole problem is so complex that it is not easily possible to understand how all this works?
“The model tells them that a pure non-radiative forcing would not cause a time lag between TOA net radiation and temperature response. Only if there is also a radiative forcing there could be such a time lag. Observations show that such a time lag exists, and this has been confirmed by Trenberth and Fasullo.”
Pehr -I interpret you to mean a lag during which a TOA net energy gain (or reduced loss) is followed after an interval by a temperature rise. However, I don’t see data in SB-11 to tell us whether a rise in temperature was preceded by a TOA energy gain, because they only relate TOA fluxes to “maximum temperature”. It is possible, even probable, that the initiating event was a rise in sea surface temperature without a preceding flux change, and that the latter only followed as part of the feedback.
Additionally, there appears to be a problem attributing temperature changes to flux changes in a quantitative sense. For most of the interval, net fluxes varied within a range of 1 W/m^2, with only an occasional interval involving an additional 0.5 W/m^2. Temperature varied over a range of 0.4 C. The question then arises – how much warming (cooling) is possible with a rise (fall) of 1 W/m^2? Based on the Stefan-Boltzmann equation, this translates to a temperature change of about 0.3 C, but such a change occurs very gradually, to approach the 0.3 C asymptotically over many centuries, and in a climate system with an ocean (even an extremely shallow one) could not result in a temperature change over monthly intervals much exceeding about 0.01 C – certainly nowhere near 0.4 C. I concluded that almost all of the 0.4 C variation must be unrelated to TOA radiative forcing even if a very small amount of such forcing were operating. If there is some way to reconcile these differences, I’m unaware of it.
Nothing we know about ENSO suggests that is anything but an internal oscillation redistributing ocean heat rather than responding to an external forcing. Perhaps that is wrong, but the data in the paper fall short of demonstrating a flaw in that understanding. In my view, the data are inconsistent with ENSO as a response to a strong forcing. It is also conceivable that there is a very minor anthropogenic forcing component to ENSO itself (a small one), and that would reconcile a putative minor forcing with the basically internally generated nature of ENSO. That is very speculative.
Fred,
However, I don’t see data in SB-11 to tell us whether a rise in temperature was preceded by a TOA energy gain, because they only relate TOA fluxes to “maximum temperature”.
My interpretation of the SB11 data is that a rise in temperature in fact was preceded by a TOA energy gain. Otherwise the data does not make sense and if not Trenberth and Fasullo would not have missed to point that out in their Real Climate blog post. TF have thoroughly analyzed SB:s data and write:
As a first step, some quick checks have been made to see whether results can be replicated and we find some points of contention.
The basic observational result seems to be similar to what we can produce but use of slightly different datasets, such as the EBAF CERES dataset, changes the results to be somewhat less in magnitude. And some parts of the results do appear to be significant.
You further wrote:
Temperature varied over a range of 0.4 C. The question then arises – how much warming (cooling) is possible with a rise (fall) of 1 W/m^2?
The TOA radiation depends on the temperature anomaly according to the equation:
R=F-lam*T+eps
Consequently the TOA net radiation varies both with varying forcing and varying temperature. The combinations of R and T that we see from SB:s data do not appear unlikely to me. Compare figure 4a in SB10 where you can see both changes in R up to almost 2 W/m^2 combined with 0.4 C but also changes with practically no variation of R combined with 0.4 C. The processes are complex in this respect.
Pehr – I believe TF have rejected the radiative forcing claimed by SB-11 as being responsible for ENSO. Some models that simulate ENSO well as an unforced phenomenon appear capable of yielding data similar to SB-11, and so I don’t think TF are saying that they agree with the “radiative forcing” conclusions in SB-11.
I haven’t revisited SB-10 recently, but I believe it primarily addressed the question of feedback. In my view, if it implied that a 1 W/m^2 (or even a 2 W/m^2) flux change could cause global temperature fluctuations of 0.4 C within a few months, that would be hard to justify.
To quantify, I agree with Fred, 1 W/m2 can only warm 25 m of water by .02 degrees per month. We see that 1 W/m2 is not sustained in SB11, so this is the largest magnitude we can expect. These amounts of radiation fluctuation cannot influence the ocean heat content in any significant way, and any causation implied by this study is from a poor interpretation of the statistics in a cycling ENSO system.
Jim D,
”To quantify, I agree with Fred, 1 W/m2 can only warm 25 m of water by .02 degrees per month”.
That is correct.
However, according to SB:s model the mixed layer can also be warmed by forcing from below, by the heat from the ENSO. This forcing is not a radiative forcing and does not show up in the measurements of the TOA net radiation.
In fact, this non-radiative forcing from below is a central part of SB:s reasoning about diagnosing of feedbacks.
Pehr, SB11 have said that in the presence of radiative forcing with frequencies comparable with the temperature variation frequencies it is not possible to separate out a climate sensitivity, and I would agree with that part. However, they have not shown that radiative forcing can independently have these frequencies, and that it is not entirely a response to the temperature. Forcing here is a strict term for external effects not caused by surface temperature variations.
Fred,
”…I don’t think TF are saying that they agree with the “radiative forcing” conclusions in SB-11”.
We agree on this point and I have tried to tell this in my previous comments. However, TF have confirmed that SB have made a correct calculation of the time lag:
”The basic observational result seems to be similar to what we can produce but use of slightly different datasets, such as the EBAF CERES dataset, changes the results to be somewhat less in magnitude. And some parts of the results do appear to be significant”.
TF agrees with SB:s calculation of the time lag but they do not agree with SB:s conclusions from the time lag.
Fred,
Sorry, I failed to use quotes. The following paragraphs are quoting you.
—–
“However, I don’t see data in SB-11 to tell us whether a rise in temperature was preceded by a TOA energy gain, because they only relate TOA fluxes to “maximum temperature””.
——
“Temperature varied over a range of 0.4 C. The question then arises – how much warming (cooling) is possible with a rise (fall) of 1 W/m^2?”
To me the whole chain of articles SB10 -> Dressler -> SB11 gives the following impression:
1. SB10 made estimates on the feedbacks telling that they don’t observe strong positive feedbacks.
2. Dressler wrote a paper to tell that SB10 results have very limited value and that they cannot be used to draw conclusions about the feedbacks. Dressler’s paper appeared to be rather a rebuttal of some claims done based on SB10 than an independent contribution.
3. SB11 tells that it’s indeed not possible to draw conclusions on the feedbacks, not even at the very limited level that Dressler indicated.
Thus everybody should now agree that it’s not possible to say much about the feedbacks based on the analysis that S&B and Dressler have done in those papers. Short term variations in temperatures and radiative imbalances can be explained in too many different ways to have any unique relationship to feedbacks even, when the lags are considered in the analysis of the relevant time series. Surface temperatures do not vary uniformly over the globe or even over the oceans. Cloudiness and albedo are not more uniform.
From the multitude of many chickens and many eggs very many narratives can be developed and it’s almost certain that several of them agree equally well with this limited set of observations.
To me, SB10 was also about the uncertainty of obtaining feedback by this method. I read between the lines at the time that it was a criticism of Lindzen and Choi, among others who have been trying to do this. If you read just the abstract, it is only about this issue, not about the feedbacks for which the paper is often quoted. They raise a good issue, that when forcing is not constant or not changing in a known way, the method has problems. Aerosol changes, or other long-term albedo changes such as unaccounted for volcanic effects, for sure, mess these analyses up. Forster and Gregory may also be affected by this as they explicitly ignore albedo changes as a forcing term in the 90’s decade when (I figure) any correlation of albedo and surface temperature would impact their sensitivity. (This was the study in the Nick Lewis thread).
Finally we end in the dilemma that it’s impossible to interpret much of the data except with the help of a large GCM as so many different factors contribute that only a detailed enough model can help in getting some order to the matter. On the first round a limited number of features of the model are significant, but digging deeper more and more of the model gets involved.
I used the word dilemma because the dependence on a complex model means on the other hand that estimating the accuracy and reliability of the conclusions becomes almost impossible, when the role of the model and of the data cannot be separated.
I think that the problem with these satellite-based studies is that the time scale is not long enough. I base my confidence in AGW on the fact that the theory fits the 30-year warming rate predicted from the CO2 increase with a sensitivity between 2 and 3 deg C per doubling. This only requires looking at the trend and not extracting a signal from a lot of noise. The trend argues against low sensitivity, and as it continues another decade or so it will be possible to get an even more accurate sensitivity out of it. Obviously, this isn’t the only evidence I use, because paleo climates also offer supporting evidence.
The difficulty is indeed that no single piece of data provides conclusive evidence. Most comprehensive and accurate data (like satellite and ARGO data) is available only over very short periods and even the best data has problems of quality and coverage. All evidence from both observations and models confirm that there is much variability on different time scales from annual to multidecadal (and little evidence to tell, whether that extends to centuries). The spatial variability is similarly large and combined in a complex way with temporal variations. It will take really long before the basic properties of the variability can be extracted directly from the empirical data.
Using the data maximally taking all temporal and spatial details into account provides much more information and may, in principle, give the possibility of developing and validating large AOGCMs more rapidly. The relationships between temperatures and clouds are causal on local level, while they are only statistical on global level (or over large regions). The causal explanations depend of course on many less local phenomena, but the physical effects occur certainly locally. Because conditions are very different in different areas, one must not assume that the correlations found in one area can be generalized to other areas. Thus some effects related locally to ENSO may be totally different of the simultaneous variations of the global average.
Once more: The global average of one variable is not causally related to the global average of another variable. The local variables may have causal relationships and they influence the global averages, but that is not the same thing as a direct causal relationship.
I wrote before that the dependence on large models is a dilemma, but it may be dilemma that can be resolved only by developing the models further, and perhaps even more importantly developing the approaches used in validating the models, and in making the methods of validation transparent enough to people who are not directly involved in the modeling activities.
I have stated on many occasions that I consider the WG1 report a mostly valid description of the existing knowledge. There are clear weaknesses in the way uncertainties are discussed, and that makes it also difficult to judge the validity of some conclusions dependent on the level and nature of the uncertainties.
Very much depends finally on two classes of evidence: model based analyses and paleoclimatic analyses. Both involve in an essential way subjective judgment. The scientists of both fields know that and try their best to improve objectivity, but they cannot get rid of subjective judgments and of the influence of the history of research in their fields.
Models are built gradually improving on older models, but that approach may involve earlier erroneous choices. Further development can compensate in many respects for such an error, but the final outcome may be seriously flawed in some ways, and correcting for that might require a major rewrite of the model.
Similarly paloeclimatic interpretations are built on narratives that have been developed by the earlier research. The data is too sparse and indirect to tell a full picture, and thus the role of narratives is important in putting the data in context. Some specific errors in that narrative may be compensated in various ways, when new information is added, but that may lead to wrong conclusions.
The arguments for being skeptic (in the traditional meaning of the word) on the reliability of conclusions are strong. When scientists work with problems as difficult as modeling the Earth system or interpreting paleo proxies, they make errors. Understanding the level of reliability and accuracy of the results requires deep knowledge on the work, and all best experts have their own biases related to their involvement in the research. When these experts can build and present their narratives uncontested, they may appear much more convincing and self-critical than they really are.
Accepting all that, I still have a fair amount of trust in the conclusions of the climate scientists, but I would certainly like to have better evidence. I would like to see more scientists discussing openly their problems rather than presenting well prepared narratives (like the lively lectures of Richard Alley).
The original publications have in part the same problem as the well prepared presentations. For an outsider they don’t really tell, where the problems are. The argumentation of the type that we have seen around the papers of S&B is welcome in that respect, but similar discussion would be needed much more widely – and on the most important issues rather than some details of little relevance that happen to be the easiest to argue on, and therefore so strongly dominating in the blogosphere.
Pekka,
‘The relationships between temperatures and clouds are causal on local level, while they are only statistical on global level (or over large regions). The causal explanations depend of course on many less local phenomena, but the physical effects occur certainly locally. Because conditions are very different in different areas, one must not assume that the correlations found in one area can be generalized to other areas. Thus some effects related locally to ENSO may be totally different of the simultaneous variations of the global average.’
The data used in S&B, Dessler etc is globally averaged TOA flux – and indeed global average temperature anomalies. So it uses measures of global changes and focuses in on ENSO only because that is the source of the major changes.
The interpretation of satellite radiant flux, temperature and ARGO ocean heat content data is very simple – and tells all sorts of things directly about warming and cooling, clouds, SST, etc. Data is valuable – GCM are rubbish. Climate is orders of magnitude more complex than models.
Cheers
Rob,
As long as the relationships are between global averages at time scales, which are not very long as well, there are too many alternative explanations for most effects, and that’s is exactly the conclusion of the paper: It’s not possible to determine the strength of the feedback from the data, because there are too many explanations even in the very limited set that the paper takes into account.
Even if the models are rubbish in some sense, they are often good enough for figuring out some of the processes that may result from the interconnected dynamics. If two models that contain a reasonably correct description of physical processes related to the observations lead to different interpretation of some empirical observations, then the observations alone cannot tell, which interpretation is correct. The models may be very revealing in this way even, when they cannot be taken as more generally valid descriptions of the real Earth system.
For many restricted issues, it’s possible to build into a model essentially everything that we know to be related to that issue. If that can be done in several different ways and further conclusions are different, that’s direct evidence on the limits of our knowledge. Improving on such models is one good way of communication between scientists, and also a way of keeping earlier knowledge while adding new. Models are very often really rather tools in learning than tools in making forecasts or projections. Making projections to compare with data in the future or to compare with other models may be a part of that learning process.
You are not precisely right Pekka – a warming feedback could not be discerned as a result of other factors. Prominently ENSO radaitive and non-radiative feedback.
subscribe
@Judy
No, this is not a technical thread. I read your analysis of Dr. Spencer’s paper with great disappointment. Though I am not a scientist, I have known for years more radiation is escaping into outer space than the IPCC has stated. Here is how I figured this out. The global warming scientists have high balled every single estimate and they want us to believe there is a giant dome around the earth, trapping radiation. Fortunately, I have a brain and knew there was no such dome, and thus the radiation was escaping into outer space. What is your problem, Judy? You talk down to everyone. It’s becoming increasingly obvious that climatology is a gimmick science. You don’t know as much about the earth as you pretend.
Chief Hydrologist | August 1, 2011 at 4:40 pm |
“You have theories with no physical basis – no mechanism – no observations. Nothing to connect a narrative that appears full grown like Athena from the mind of Zeus. ”
Au contraire, lots of mechanisms and lots of observations. Lots of work too and a lifetime of observational experience and study albeit as an enthusiastic amateur from long before the whole subject became so fashionable.
See here:
http://www.irishweatheronline.com/category/features-2/wilde-weather
and here:
http://www.irishweatheronline.com/features-2/wilde-weather/setting-and-maintaining-of-earth%e2%80%99s-equilibrium-temperature/18931.html
In fact there is nothing else that fits the observations and basic physics so comprehensively.
Whoops, correct first link here:
http://www.irishweatheronline.com/features-2/wilde-weather/the-sun-could-control-earths-temperature/290.html
Chief Hydrologist said:
“Cloudiness is an ENSO feedback – observed from changes in the eastern Pacific. The satellite data shows decreases in cloud to the late 1990′s and an increase since – corresponding to the decadal shift in Pacific climate that occurred in the late 1990′s.”
Exactly, but the consensus says that El Nino produces more evaporation and more clouds. Like you I noticed that that may be so in certain regions of increased convection but elsewhere under the subtropical high pressure cells there is increased downwelling of air that dissipates clouds. At the same time those stronger cells widen the tropical air masses pushing the mid latitude jets poleward for even less clouds. So the global effect of El Nino is less clouds.
You are right too about the increase in cloudiness from the late 90s but in fact that preceded the decadal shift in Pacific climate to a negative PDO which I read has been placed at around 2003.
Chief Hydrologist asked:
“What do you mean by ‘top down’ modulation? It is just empty words.”
Read my above link about how the sun could control Earth’s temperature. The decline of solar activity from the late 90s onward is what most likely caused the increase in cloudiness to begin before the start of the negative PDO.
A theory from a lawyer with no observations – no references – no mechanisms. It is all the wildest flight of fantasy.
Indeed whenever people start talking about new theories of the universe – usually the bullshit is flying.
There are a lot of observations and references in those articles. I have to assume you have some sort of mental block on the issue.
You’re right – I can’t be bothered with nonsense.
Chief
Cheer up! It will be summer down there before too long! I, for one, would be interested to hear a technical explanation for why you disagree with Stephen’s ideas if you have the time. Cheers
Rob
‘This article will try to show that that basic assumption which has been incorporated into all current climate models and theories may be wrong.’ .
Higher solar activity puts energy into the tropical oceans and drives the sub-tropical jets poleward – changing convection patterns in undefined ways and alternately increasing and decreasing cloud? ENSO is a bit of an afterthought.
It is vague speculation with no recognition in the literature – with which I can’t be bothered. Long ago and far away I was simply building a larger picture of the ENSO processes.
ENSO causes cloud changes on inter-annual to decadal timescales – shown in CERES data. As in the S&B paper etc.
http://www.earthandocean.robertellison.com.au/
Solar energy is needed to fuel ENSO in the first place and due to albedo changes the amount of ‘fuel ‘available for ENSO must therefore change. ENSO is not an afterthought but it may well be a secondary process.
ENSO is probably only one part of the story because it does not account for the long term changes in the positions of the jets as from MWP to LIA to date and the long slow global temperature swings that lie beneath ENSO and the Pacific Multidecadal Oscillations.
For the other part we have to look to the solar influence.
I agree that there is no recognition in the literature for the precise combination and sequrence of events that I describe but it is not mere speculation.
Lots of new observations have been coming to light over recent years and the literature does not provide a scenario that fits them all.
Accordingly I set out to create a scenario that complied with basic physics and as many of those observations as possible.
My scenario can be falsified or will need substantial revision if the stratosphere resumes significant cooling whilst the sun remains less active. However if the stratosphere fails to resume cooling or more especially if it starts to warm then everything else falls into place and my scenario would then be the only one that fits the facts.
By all means ignore it if you wish but if the stratosphere behaves as I think it will (it has already begun to do so) then my hypothesis will not be ignored. Indeed it would then be the only hypothesis left standing.
CH,
“A theory from a lawyer with no observations – no references – no mechanisms. It is all the wildest flight of fantasy.”
As opposed to a bunch of faux scientists with twisted observations, self referenced, and made up mechanisms.
I think Stephen has a slight edge.
“Indeed whenever people start talking about new theories of the universe – usually the bullshit is flying.”
I see, you prefer the good old Phlogistone theories with long and hallowed histories even though they have ultimately been proven wrong?
By the way, when you can start making sense with all the “information” you inundate us with we might pay attention to you.
Thanks kuhnkat.
Only a ‘slight’ edge ? :)
Suit yourself.
The following is only tangentially relevant to the specific topic of this thread, but I don’t know of any better place to link to it. It’s a comprehensive BAMS document on The State of the Climate – 2010 that addresses some of the phenomena covered here as well as many more. Parts deal with 2010 exclusively and other parts with recent climate multidecadal history. I think it’s worth the time spent to go through it.
Not much related to anything but The Times published its first weather forecast 150 years ago today (1st August). Apparently the forecaster, Admiral Robert FitzRoy, got that one right but things neither went smoothly nor well for him after that.
Alex
My favorite weather forecast came quite a while back – I think from radio/TV comedian Henry Morgan –
Tomorrow will be muggy – followed by tueggy, weggy, and thurgy.
Based on the discussion here this is an attempt to summarize Spencer’s and Braswell’s view on the climate system, some important results of their research and the recent criticism of their work.
SB consider a conceptual model consisting of the troposphere and the mixed layer with radiative forcing at the top and non-radiative forcing at the bottom. This is also applied as a simple energy balance mathematical model of the climate system that may be solved numerically as an ordinary differential equation.
They assume that the TOA net radiation is the sum of a temperature independent forcing and a temperature feedback term, proportional to the temperature change, and a random internal variability:
R = F – lam*T + eps
If the temperature changes due to a non-radiative forcing with the radiative forcing constant, for example caused by the ENSO, the change in TOA net radiation R will be proportional to the temperature change T with the slope = lam = the feedback parameter.
In SB10 it was shown that when such events happen the points in the phase plane plot of R vs. T will give at straight line with the slope lam. Straight lines in the phase plane plots in SB10 (fig. 4) that may correspond to such events show a slope of lam = the feedback parameter = 6 W/(m^2 K) (this is a rather high value suggesting an insensitive climate, at least in the short term).
In the case of a pure non-radiative forcing there is obviously no time lag between the TOA net radiation R and the temperature change (assuming that the atmosphere responds swiftly with changing radiation on a changing temperature). However, as shown in SB11, in case of a contribution from a radiative forcing there will be a time lag because of the slow heating of the mixed layer.
Both a phase plane plot in SB10 (fig. 4a) and the time lag analysis in SB11 support that both non-radiative forcing and radiative forcing have been important during the decade 2000. This is contrary to the assumption that the radiative forcing may be neglected in the cloud feedback study by Dessler in Science 2010 (Dessler in response to SB11 has criticized their conclusions as incorrect).
Both in SB10 and in SB11 the authors show many examples that advanced climate models do not agree with satellite observations of TOA radiation data (that is not unexpected).
Trenberth and Fasullo have criticized SB:s work on the ground that their model is too simple and cannot account for the effect ENSO. However, both in SB10 and SB11 the authors obviously use a non-radiative forcing in their model which is suitable for representing forcings by the ENSO on the mixed layer. Hence the criticism by TF is questionable.
Another criticism is that the range of changes in the TOA net radiation (typically 1 W/m^2) is not sufficient to explain the wide range of the temperature changes (up to 0.4 C) considering the heat capacity of the mixed layer. However, most of the temperature changes are caused by non-radiative forcing as an effect of the ENSO. The radiative forcing is much smaller although it cannot be neglected according to SB11. Considering the effect of the total forcing there is no such mismatch between changes in the TOA net radiation and the temperature as stated according to that criticism.
SB:s comparison of satellite data with their simple model for the climate systems gives some interesting patterns and explanations which also agree with the results of Lindzen and Choi. While some criticism that SB on some issues have jumped to conclusions may be justified, other parts of the recent criticism of their most important results based on fundamental issues is questionable . Occasionally also the motive for some of the criticism may be doubted.
Hi Professor Curry…
When I read SB11 I was somewhat confused by their use of the word “forcing” for clouds, as I would have considered it a feedback. Then I read your chapter on feedbacks:
Presumably if it’s a common misnomer, people in the field are used to translating.
It did occur to me, though, that if changes in cloud parameters or characteristics were being caused by large (2-3 orders of magnitude) changes in bacterial or fungal sporulation, this would probably count as forcing in a climate-only model. My question: has there been any work to identify the possible climate sensitivity to very large fungal or bacterial blooms?
Given that growth conditions for these organisms are strongly dependent on annual weather conditions, it could potentially be an important feedback in the combined bio-climate system, wherever it fits into the models (if at all). The first thing that came to my mind was the possibility of a feedback loop between ENSO and spore production in the Amazon basin, or perhaps the Western Pacific equatorial rain forests.
Could you point me to any open access work modeling possible sensitivity of this type? Thanks.
P.S. I found the studies below, while Googling for papers. I scanned the abstracts, but couldn’t find anything regarding the sort of feedbacks I was looking for.
Hoose, Corinna, Jón Egill Kristjánsson, Jen-Ping Chen, Anupam Hazra, 2010: A Classical-Theory-Based Parameterization of Heterogeneous Ice Nucleation by Mineral Dust, Soot, and Biological Particles in a Global Climate Model J. Atmos. Sci., 67, 2483–2503. doi: 10.1175/2010JAS3425.1
V. T. J. Phillips, C. Andronache, B. Christner, C. E. Morris, D. C. Sands, A. Bansemer, A. Lauer, C. McNaughton, and C. Seman Potential impacts from biological aerosols on ensembles of continental clouds simulated numerically Biogeosciences, 6, 987–1014, 2009
R. Iannone, D. I. Chernoff, A. Pringle, S. T. Martin, and A. K. Bertram The ice nucleation ability of one of the most abundant types of fungal spores found in the atmosphere Atmos. Chem. Phys., 11, 1191–1201, 2011 doi:10.5194/acp-11-1191-2011
C. E. Morris, D. C. Sands, M. Bardin, R. Jaenicke, B. Vogel, C. Leyronas, P. A. Ariya, and R. Psenner Microbiology and atmospheric processes: research challenges concerning the impact of airborne micro-organisms on the atmosphere and climate Biogeosciences, 8, 17–25, 2011 doi:10.5194/bg-8-17-2011
D. G. Georgakopoulos, V. Despres, J. Frohlich-Nowoisky, R. Psenner, P. A. Ariya, M. Posfai, H. E. Ahern, B. F. Moffett, and T. C. J. Hill Microbiology and atmospheric processes: biological, physical and chemical characterization of aerosol particles Biogeosciences, 6, 721–737, 2009
Heald, C. L., and D. V. Spracklen (2009) Atmospheric budget of primary biological aerosol particles from fungal spores Geophys. Res. Lett., 36, L09806, doi:10.1029/2009GL037493
Heidi Bauer, Fabio L.T. Goncalves, Elisabeth Schueller, and Hans Puxbaum Fungal spores as potential ice nuclei in fog/cloud water and snow
C Hoose, J E Kristjansson, and S M Burrows How important is biological ice nucleation in clouds on a global scale? doi:10.1088/1748-9326/5/2/024009
U. Pöschl, S. T. Martin, B. Sinha, Q. Chen, S. S. Gunthe, J. A. Huffman, S. Borrmann, D. K. Farmer, R. M. Garland, G. Helas, J. L. Jimenez, S. M. King, A. Manzi, E. Mikhailov, T. Pauliquevis, M. D. Petters, A. J. Prenni, P. Roldin, D. Rose, J. Schneider, H. Su, S. R. Zorn, P. Artaxo, M. O. Andreae1 Rainforest Aerosols as Biogenic Nuclei of Clouds and Precipitation in the Amazon SCIENCE VOL 329 17 SEPTEMBER 2010 DOI:10.1126/science.1191056
F D Pope Pollen grains are efficient cloud condensation nuclei Environ. Res. Lett. 5 044015 doi: 10.1088/1748-9326/5/4/044015
W. Winiwarter, H. Bauer, A. Caseiro, H. Puxbaum Quantifying emissions of primary biological aerosol particle mass in Europe doi:10.1016/j.atmosenv.2008.01.037
“Spencer & Braswells new paper | Climate Etc.”
was a delightful post, can’t wait to look over alot
more of your articles. Time to waste numerous time on the
net lol. Thanks for your time -Graciela