Dr. Roy W. Spencer replies to “Tamino”‘s latest angry missive. As one commenter in my email list put it:
It is absolutely hilarious that Tamino’s lengthy, time-consuming, chest-puffing critique can be so comprehensively dismissed in a mere two sentences.
Here is what Dr. Spencer posted on his web page:
October 8, 2008: A Brief Comment on “Spencer’s Folly”
For anyone who has stumbled across a rather condescending critique of our latest research on feedback by someone who calls himself “Tamino”, I can only say that Tamino could have saved himself a lot of trouble if he would have noticed that all of my feedback work addresses TIME-VARYING radiative forcing (as occurs during natural climate variability), not CONSTANT radiative forcing (as is approximately the case with global warming). Tamino’s analytical solution does not exist in the time-varying case, and so his holier-than-thou critique is irrelevant to what I have presented.
Here is the original Spencer essay in PDF form, hosted on Roger Pielke’s website.
On the other hand, here is a recently published paper on climate sensitivity (PDF) that says the opposite. I’ll let the reader decide how well it defines the climate sensitivity, but I would note that since it uses GISTEMP data, which has a number of data problems that we’ve uncovered, for example here and here, the sensitivity may be overrated due to inflated trends in the GISTEMP database.
In the meantime, if you feel like supporting Dr. Spencer’s work, head on down to Barnes and Noble and get his latest book:
Spencer’s new book “Climate Confusion” is
now available at Amazon and Barnes & Noble.
(See book covers, and first page of each chapter.)
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
Awfully quiet over there; he must be breaking pencils.
==================================
Joel, Spencer challenges the “modelers”
“Now, if the modelers STILL insist that this short term (5-year) feedback behavior — even in the models — does not invalidate positive feedback for long-term global warming, I will respond: “OK, then adjust the models so they behave like the satellite observations on the short (5-year) time scale, and THEN show us how much global warming the models predict”.
So do you throw up your hands and say the “real data” only covers short time scales and ignore it? Or do you try to accept real data over models and try to fit “data” into the models? I go for the latter.
Thanks for your answer over there, Joel, but if the models work for monthly or yearly time scales, but not for shorter or longer periods, it still points to a problem in the models rather than in the data.
==============================
Joel (08:25:50) Hah, Neil’s hijack is hardly one. He restarted the conversation a couple of days ago and it generalizes with ours.
========================================
And now Tamino has retreated to blaming the data, calling it not ‘precise’ enough. Pitiful, just pitiful.
=============================
kim:
First, where did you get the statement that the models don’t work for shorter periods? I didn’t say or mean to imply that. I don’t know if they have been tested for shorter periods. However, since the convective processes that lead to the moist adiabatic lapse rate predictions operate on short timescales, there is every reason to believe that if these processes are dominating on the monthly-to-yearly timescale, they are dominating on the shorter timescales where they actually operate too.
Look, the take-home message is that the models predict convective processes to dominate and produce results close to moist adiabatic lapse rate predictions over all timescales (or at least all that I know about). The data shows this occurring except when one looks at the longest timescales in the data, which corresponds to then looking at the multidecadal temperature trend over the whole data set rather than fluctuations. For those longest timescales, the data deviates from the predictions and from its behavior at the shorter timescales…or, at least, some analyses of the data do (e.g., UAH does significantly but for RSS the deviation is not very significant). However, the data for the multidecadal trends has known problems (e.g., a cooling artifact for the radiosonde data due to better shielding of the sensors from the sun over time) and attempts to correct for that in various ways bring the data into much better alignment with the models and theory. Hence, lacking even any proposed mechanism that I know of for why the moist adiabatic lapse rate theory predictions should break down, it seems reasonable to think that the problem is most likely due to the known deficiencies in the data.
By the way, as a side comment, it is worth pointing out that this lapse rate effect provides a negative feedback in the climate models because the notion that the upper troposphere warms more than the surface means that the surface does not need to warm as much in order to warm the part of the atmosphere that is effectively emitting most of the radiation back out into space. So, at least naively, if this notion is wrong and the surface really warms as faster or faster than the upper troposphere then this would tend to imply less negative feedback and a higher climate sensitivity, all else being equal.
It’s just not going to work, claiming the data is bad. Sure, it’s possible, but the flaws in the models seem to coincide with just where they were expected from their poor performance. It’s a good deal more likely that is where the problem lies.
What about Spencer’s question about the inadequate cause and effect between clouds and temperature which damage the assumptions in the models?
================================================
There is no doubt whatsoever that the data is bad for the longtime trends in the tropical troposphere. First of all, the coverage of the radiosondes is very sparse, the problems with the changes in shielding and such over time are well-known, and the size of the estimated corrections that are needed is not small. As a result, there are a wide range of estimates from the analyses and re-analyses of the data. Second of all, for the satellite data, there are considerable differences between the different analyses…and again well-known problems with stitching together the results from the different satellites.
Okay, so can you explain to me how this mechanism could cause the moist adiabatic lapse rate predictions to break down on the very long time scales but to still hold for fluctuations on monthly to yearly timescales? It seems unlikely to me that this could be the case since I believe that Spencer it talking about processes that occur on shorter timescales than this.
I’m sorry, Joel, it is not convincing; the data, looked over by two independent labs, is wrong, and the models, which have failed, are right. Why not review the assumptions in the models about convection, water vapor, and clouds?
===================================
“So, in other words, you and the crowd at ClimateAudit have set yourself up as judge and jury and are now passing judgement on scientists.”
Yes. What’s your point?
“And yet, you have yet to get any impartial scientific authority to agree with you on these serious charges.”
Give me the name of an impartial scientific authority that has dealt with the subject of Mann refusing to update his series to the present. Give me the name of an impartial scientific authority that has even looked at the source and collection of the Graybill data and who has compared this with the Ababneh data.
“In fact, when the NSF was asked by McIntyre to weigh in on the issue of the release of information to him, they told him in no uncertain terms that Mann was not obligated to release to him the computer code that he was demanding.”
I don’t give a damn what NSF said. If he is developing computer code on the public dime, then the computer code belongs to the public. Beyond that, if you are trying to get the world to spend trillions of dollars based upon your computer code, then you have a moral obligation to let others look at the details of your gloom and doom.
Joel Shore,
You certainly have all the RealClimate Manifesto talking points down to a science.
What you fail to mention is satellite data can be analyzed and calibrated to known standards, then adjusted accordingly. This is standard procedure in industry.
When was the last time the surface station network was calibrated? To what standard?
You said:
“By the way, as a side comment, it is worth pointing out that this lapse rate effect provides a negative feedback in the climate models because the notion that the upper troposphere warms more than the surface means that the surface does not need to warm as much in order to warm the part of the atmosphere that is effectively emitting most of the radiation back out into space. So, at least naively, if this notion is wrong and the surface really warms as faster or faster than the upper troposphere then this would tend to imply less negative feedback and a higher climate sensitivity, all else being equal.”
Oh my, so no matter what happens, no component of the CO2 AGW hypothesis can be falsified. How convenient. How long can you get away with claiming the observational data is wrong?
Sorry Joel, it is written in stone:
Amplification of Surface Temperature Trends and Variability in the Tropical Atmosphere- Gavin Schmidt et al 2005
http://www.osti.gov/energycitations/servlets/purl/881407-xk2Sdg/881407.PDF
Is the following statement true or false?
Earth’s Energy Imbalance: Confirmation and Implications- Hansen/Schmidt et al 2005
http://www.sciencemag.org/cgi/content/abstract/1110252v2
“Our climate model, driven mainly by increasing human-made greenhouse gases and aerosols among other forcings, calculates that Earth is now absorbing 0.85 ± 0.15 W/m2 more energy from the Sun than it is emitting to space. This imbalance is confirmed by precise measurements of increasing ocean heat content over the past 10 years. Implications include: (i) expectation of additional global warming of about 0.6°C without further change of atmospheric composition; (ii) confirmation of the climate system’s lag in responding to forcings, implying the need for anticipatory actions to avoid any specified level of climate change; and (iii) likelihood of acceleration of ice sheet disintegration and sea level rise.”
By your logic, “precise measurements of increasing heat content over the past 10 years” would not extend beyond 2003 because observations since then disagree with the above statement and therefore must be flawed.
Where is the missing heat Joel?
Joel, climate science right now is a total disgrace. Just look at this little snippet:
http://www.climateaudit.org/?p=4064
There are trillions of dollars at stake and yet we get stonewalling from people put in charge that mostly work for us. I personally think Hansen should be fired.
As Anthony has put on his blog, climate science badly needs some sort of ISO 9000 type oversight.
As fas as Spencer’s data goes–I think you are stretching suggesting his data is wrong and the models are correct—-only in climate science does this type of joke happen.
Joel Shore, smooth talker..
Tilo, straight shooter…
Who would YOU buy a used car from?
Joel Shore
Thanks for contributing to discussion on this blog.
However in answer to Kim’s question: –
“What about Spencer’s question about the inadequate cause and effect between clouds and temperature which damage the assumptions in the models?”
Your answer: –
“Okay, so can you explain to me how this mechanism could cause the moist adiabatic lapse rate predictions to break down on the very long time scales but to still hold for fluctuations on monthly to yearly timescales? It seems unlikely to me that this could be the case since I believe that Spencer it talking about processes that occur on shorter timescales than this”.
dodges the question.
I believe Spencer is saying that the assumption that the observed cloud variability that was believed by some scientists to be the result of feedback , ignored the possibility that some, not fully understood or accepted phenomena, was partly responsible and this lead to the adoption of a climate sensitivity that is too high. In fact Spencer’s current work (see below) shows that, over the 5 year periods of his analysis, low clouds increased with warming i.e. negative feedback and that water vapor + lapse rate + high cloud feedback was weakly positive.
See his latest research update here: –
http://www.weatherquestions.com/Roy-Spencer-on-global-warming.htm
“October 8, 2008 RESEARCH UPDATE #2:
We have submitted a paper for publication to Geophysical Research Letters entitled, “Satellites Reveal a Climate System Less Sensitive than in Models”. This paper should answer the previous criticisms of our August 2007 GRL paper on negative feedback in the tropics that (1) it only applied to the tropics, and (2) that feedbacks diagnosed on short time scales might not apply to long-term global warming.
In the new paper we diagnose feedback parameters from 5 years of NASA Aqua satellite data over the global oceans AND perform exactly the same diagnoses on all possible 5-year periods in transient CO2 simulations from 18 IPCC climate models.
The results are, as you can see below, somewhat stunning…NONE of the five year periods from ANY of the IPCC climate models show the negative feedback behavior seen in the satellite data:
Contrary to my expectations, though, the negative feedback was not in the longwave (infrared); there was an excellent match between the models and satellite observations in that component, suggesting that the total [water vapor + lapse rate + high cloud] LW feedback was weakly positive.
Instead, the negative feedback was entirely in the reflected shortwave (solar)…suggesting that low clouds increase with warming. This is actually somewhat consistent with the IPCC AR4 report which admitted that feedbacks related to low cloud behavior were the most uncertain in the models. Since this is an apples-to-apples comparison between the models and the satellite observations, it will be difficult for the IPCC to ignore this kind of evidence.
The question of WHY the IPCC models would be so far off is, in my view, related to what I discussed above, in my Research Update #1: In previous analyses of natural co-variability between clouds and temperature, only feedback has been assumed to be operating, when in fact some of the variability is actually cloud fluctuations causing temperature change. In simple terms, there has been a mix-up between cause and effect, and that has led to climate models being built upon faulty assumptions.
Now, if the modelers STILL insist that this short term (5-year) feedback behavior — even in the models — does not invalidate positive feedback for long-term global warming, I will respond: “OK, then adjust the models so they behave like the satellite observations on the short (5-year) time scale, and THEN show us how much global warming the models predict”. ”
Touché!
It will be interesting to read this paper and the responses.
kim says:
What independent labs are you referring to? UAH and RSS? They show significant differences in the trends in the tropics. For the RAOBCORE radiosonde re-analysis project, there are dramatic differences just between different versions of their results…In fact, Douglass et al used an earlier version while later versions show significantly more warming in the upper troposphere.
I am not sure what you mean by “which have failed”. Who decided that the models have failed and why? The models are just that…Models of reality. They are good in some respects and less good in others.
Look, I don’t even know where to start here, and neither does anyone else judging from the distinct lack of even vague hypotheses that might explain how the models could be getting these assumptions wrong in a way that still causes them to do a very good job describing the tropospheric amplification for fluctuations on the monthly to yearly timescales but then somehow fails for the longer term trends. The processes that you mention all tend to occur on timescales of hours to days to weeks. How can one be getting these processes wrong in a way that still causes you to get good results on timescales one to three orders of magnitude longer but then to suddenly fail at timescales longer than this? Or, maybe there is some totally new process that operates on way longer timescales of, say, several years…but then what is such a process? Maybe there is someone out there who could provide at least some vague hypothesized mechanism of how this could happen but I have yet to meet such a person!
Tilo Reber say:
In a way, it is sort of cute how you guys believe that you set the laws…and organizations such as NSF are totally irrelevant. Apparently, you know better than them how one should balance the issues of intellectual property rights, scientific disclosure, and public access to federally-funded research. Furthermore, if it is not done just the way you want it, you don’t work to change the laws (a reasonable approach) but instead just insist that you are still in the right and start throwing around words like “fraud” to describe the actions of those who are doing things totally within the law (and, recently, going well, well beyond what the law demands in making essentially everything available).
And to extend that, if you’re confident in your conclusions and methods, you’d be happy to have anyone examine your data. It’s blatantly obvious that McIntyre is eminently more qualified at statistics that Mann and his team. And absolutely everything about the Hockey Stick is statistics, not “climate science”.
Apparently, you know better than them how one should balance the issues of intellectual property rights, scientific disclosure, and public access to federally-funded research.
Fine. Fine and dandy. All well and good.
But then it ain’t Science. It’s Alchemy.
The whole conversation with Joel Shore reminds me of a defense attorney who is paid to defend his client come hell or high water. If you can’t beat them with real data, then try to discredit the character of the prosecutor and the witnesses, and don’t overlook the power of trying to argue about the definition of tiny little words in order to distract attention from the main argument. Like Bill Clinton asked “What do you mean by “sex?””.
Joel, you are so obviously committed to AGW, it has become for you a marriage for better or for worse. You have now become irrelevant, and in my case-ignored.
I doubt there is any “law” involved. The fact of the matter is, McIntyre is a qualified peer reviewer. The IPCC saw fit to use him as an expert reviewer, so why does Mann see him as an amateur funded by oil companies? What does it matter who he’s funded by? Gore gets money from oil companies, so does Suzuki. Does that mean they can be safely ignored?
Joel (18:57:53) The satellites are more reliable than ground and probably more than you think.
Dontcha’ know the models have failed? Or been falsified? Or disconfirmed? See Lucia’s Blackboard and don’t try to give me any guff about projections not predictions.
I don’t think the models parameterize the convections, and water phase changes at the time spans that Spencer is measuring, and they fail at the long term.
We have descended to repetition in this dispute. You believe in the models, I don’t. They are not useful.
========================================
Joel Shore (18:57:53)
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
From: http://www.nasa.gov/offices/ogc/ip/1210.html
NASA’s ability to protect intellectual property is affected by the Freedom of Information Act (FOIA) [5 U.S.C. § 552]. The FOIA permits any person the right of access to Federal agency records unless specifically exempted by the Act. Proprietary documents containing private sector trade secrets and commercial or financial information, generated outside the Government, but in the control of NASA, are exempt from disclosure by FOIA exemption 4. Information disclosing inventions and information that is part of a patent application may be withheld under FOIA exemption 3.
When quoting or continuing, a reference post name & timestamp is helpful for us slower latecomers.
Bob B says:
First of all, you are phrasing this in a way that makes it sound like the data supports one thing and the models another (and partly it may be because you are mixing things up with the simultaneous discussion we are having in regards to the tropical tropospheric trends…which isn’t really related except quite tangentially). In fact, what I would say is that I believe that Spencer’s data analysis is likely wrong. And, the data analysis of many others, who have obtained estimates of climate sensitivity from observational data in various ways, such as Hansen’s method of looking at the last glacial maximum, are likely correct. Yes, these latter climate sensitivity estimates are also supported by the current sensitivity estimates from climate models…but this is not the only or even primary evidence for what the climate sensitivity is.
As for the second part of your comment, could you back this up by telling me what fields of physical science you have worked actively in to compare to?
Geoff Larsen says:
I understand what Spencer is arguing. The problem here is that there are two only tangentially-related discussions going on at the same time: one concerning Spencer’s work and one regarding the existence or non-existence of amplification of temperature fluctuations and trends as one goes up in the tropical troposphere. My point (making a vague connection between these two different discussions) is that even if Spencer is right about what he is talking about here, I don’t see how any sort of correction to the physical processes of the type that Spencer is talking about can solve the issue regarding the magnification of temperature fluctuations in the troposphere because he is talking about processes that operate on timescales shorter than timescales where the data show the magnification is in fact working in the way that the models predict. (To be fair, Spencer has not to my knowledge tried to claim that this work addresses the tropospheric amplification issue, but some people like kim seem to be implying that it could and I am just saying that I don’t see how it could.)
Jeff Alberts says:
Well, by that standard, Mann must be since his latest paper has the data and programs available freely on the web. Over at ClimateAudit, they seem to be having a field day looking at although so far they seem to be generating a lot more heat than light.
First of all, because it tries to be very open, the IPCC more-or-less allows anyone to become an “expert reviewer” who wants to. (Technically, you may have to be nominated by a government or an organization…but surely someone like McIntyre can find an organization skeptical of AGW who would be happy to nominate him.) Second of all, intellectual property rights are a matter of law. And, the NSF was very clear on what Mann’s obligations were to McIntyre or anyone else in the letter that they sent to McIntyre.
Dennis Sharp:
Do what you want. Sure, we all have our biases and I won’t deny that I have mine. But, perhaps you might examine your own biases and commitments too. And, I try hard to call them as I see them…and in this thread when Gavin Schmidt said something that I thought was incorrect (or at least partially so), I said so. I also have not dismissed Spencer’s work out-of-hand but I also don’t think Spencer’s short response to Tamino’s critique constitutes anything like a complete rebuttal.
Jojel (04:58:16) You were able to point out to me the confusion of the two arguments; I maintain that they are specific and general cases of the same argument, but I appreciate your criticism that they are only tangential. Nonetheless, most of what I’ve said about the data and the models still stands.
It is a tough nut to crack. It may be too complex for human understanding, yet.
===================================