What do you make of the new skeptical global warming paper by Christy and Singer et al?

Question

I can only find the full paper downloadable for $25, and I'm not willing to pay that, but here it is if anyone wants to read it:

http://www3.interscience.wiley.com/cgi-bin/abstract/117857349/ABSTRACT?CRETRY=1&SRETRY=0

Abstract: "We examine tropospheric temperature trends of 67 runs from 22 'Climate of the 20th Century' model simulations and try to reconcile them with the best available updated observations (in the tropics during the satellite era). Model results and observed temperature trends are in disagreement in most of the tropical troposphere, being separated by more than twice the uncertainty of the model mean. In layers near 5 km, the modelled trend is 100 to 300% higher than observed, and, above 8 km, modelled and observed trends have opposite signs. These conclusions contrast strongly with those of recent publications based on essentially the same data."

Do you think their argument has merit? Why?

No political rhetoric please - just scientific arguments.

The report is published in the December 2007 issue of the International Journal of Climatology of the Royal Meteorological Society [DOI: 10.1002/joc.1651]. The authors are Prof. David H. Douglass (Univ. of Rochester), Prof. John R. Christy (Univ. of Alabama), Benjamin D. Pearson (graduate student), and Prof. S. Fred Singer (Univ. of Virginia).

Title: "A comparison of tropical temperature trends with model predictions"

Larry - does that link work for you? I've tried it at least a half dozen times over the past 2 days and it's never worked for me.

Hard to read, but you can view it as HTML here:

http://216.109.125.130/search/cache?ei=UTF-8&p=International+Journal+of+Climatology+of+the+Royal+Meteorological+Society+DOI%3A+10.1002%2Fjoc.1651&fr=yfp-t-501&u=icecap.us/images/uploads/DOUGLASPAPER.pdf&w=international+journal+climatology+royal+meteorological+society+doi+10+1002+joc+1651&d=SKTl5rXiP7hn&icp=1&.intl=us

Marc - another link would be appreciated if you can find one. The ICECAP link still hasn't worked for me after a dozen tries.

gcnp gets at the heart of my question here. I'm wondering if this paper says anything new, because it seems to be rehashing arguments that have already been made regarding the tropical troposphere measurements vs. models.

Have Christy et al. improved the tropospheric analysis, or are they simply summarizing what we already know?

As gcnp points out, one flaw in the model does not mean the whole model is flawed.

Answer 1

I don't like the model ensemble approach taken by the authors. I think that a more valid comparison would be between mulltiple simulations of each model and the observed data. A better approach would be to run each model enough times with random initial conditions from the population of observed initial conditions to have statistical certainty > 95% that the range of possible outcomes from that model have been captured. In the study the number or runs is 1 - 9. In my branch of physics 100 is a small number of simulations. The agreement with observations could then be calculated on a model by model basis.

The model ensemble approach is open to manipulation by assigning more weight to models that either fit the observations well or do not fit the observations well. It is also possible to select particular runs of a simulation that empasisze some effect. (If any of this is new to Exxon, I hope that my cheque is in the mail).

I would rather discount the possiblity that any bias was deliberately added. Suppose instead that that the author's conclusion that the models do not accurately predict atmospheric temperature profiles is correct. That would mean that model(s) need to be improved in a manner that preserves the good fit at the polar regions and low altitudes.

One final note. The degree of precission achievable by modelling varies from one dicipline to another. A theorist working on a quantum many body problem would be estatic to achieve the degree of agreement achieved by the climate models. The therory of quantum mechanics is not in any imminent danger.

Edit: Ben O makes a good point about the standard deviation being meaningless. As the number of runs for a model increases the mean should tend to its true value roughly as sqrt(N). Increasing the number of runs also gives a better estimate of the model variance. If model A agrees with observations within 1 SD and model B agrees within 2 SD they may both be deemed to agree with the observations by the criteria used, but model A is the better model. Using a small number of runs is just cherry picking whichever conclusion you want.

Answer 2

It's easy to critise other peoples work and I can see a few things that don't look scientific.

The author has produced a data set of model runs. This data set deviates from a normal distribution so much that the value of standard deviation is meaningless. The data is not normally distributed - sometimes the model gets a 'wild' value which drowns out all of the other data values for calculating the mean and standard deviation. The authors' claim that the real world data is outside 2 standard deviations of the mean is not meaningful.

According to the paper, previous researchers have used the total spread of model results as their uncertainty - this is not ideal as the more runs there are, the bigger the uncertainty. You could fit any real world data with enough model runs. The uncertainty of the model apears to have been artifically increased for the purposes of getting the real world data to 'fit' the model.

The issue I see is that there is so much uncertainty in the model that researchers can find real world data that goes in the opposite direction to the trend predicted by the model, but still within the range of uncertainty and claim it fits. It does fit, but so does a data set full of zeros. This point doesn't prove much.

If the authors' were to increase the size of their data set and do some meaningful statistical analysis, they could probably demonstrate that the real world data deviated from the model by a quantifiable amount. This isn't surprising as every mathematical model of a real world phenomenon has some variation and bias associated with it - climate science shouldn't be any different.

Answer 3

This is *so* detailed it isn't possible to give a sensible answer in less than a week, which is my estimate for how long it would take to fully review the technical content of the manuscript (there are ten empirical data sets and 22 model outputs). If I were going to undertake this project, I would start by figuring out if the data in this report:

http://www.climatescience.gov/Library/sap/sap1-1/finalreport/sap1-1-final-execsum.pdf
(see Figure 1)

matched the data trends in the Douglas et al. manuscript from Int. J. Clim. Furthermore, I would carefully go through the report above and the conclusions from the manuscript to see if there was any new real information in the manuscript. Chances are there is not, it's just been repackaged so as to be more sensational.

The problem of climate models doing a bad job reproducing tropical temperature lapse rates is not news. We've even discussed it here at length before, and one that is the subject of a lot of research effort. Figuring out why there is a problem here is important, but focusing on the tropical lapse rate issue and ignoring the other places where the models do work (e.g., arctic amplification) suggests a biased approach to the problem. It does not call into question the underlying science, but represents the common contrarian argument that the effects of global warming are greatly overestimated. However, this contention is refuted by the direct empirical evidence of *accelerated* warming in the cryosphere.

Answer 4

http://216.109.125.130/search/cache?ei=UTF-8&p=International+Journal+of+Climatology+of+the+Royal+Meteorological+Society+DOI%3A+10.1002%2Fjoc.1651&fr=yfp-t-501&u=icecap.us/images/uploads/DOUGLASPAPER.pdf&w=international+journal+climatology+royal+meteorological+society+doi+10+1002+joc+1651&d=SKTl5rXiP7hn&icp=1&.intl=us

"If these results continue to be supported, then
future projections of temperature change, as depicted in
the present suite of climate models, are likely too high."

Let's hope so! They do stop short of saying global warming isn't happening.

"The last 25 years constitute a period of more complete and accurate observations and more realistic modelling
efforts. Yet the models are seen to disagree with the
observations. We suggest, therefore, that projections of
future climate based on these models be viewed with
much caution."

This should get a lot of attention. If it holds up, hopefully it can be replicated, extended, and better models can be developed.

Then again, it needs to be reconciled against
observations that have seemed to indicate accelerating warming.

I can't wait to see how it's received.

Then again, it seems identical to what Douglass and Singer summarized here in 2004, so what's new?:
http://www.heartland.org/Article.cfm?artId=15727

Does it really take 3 years to get completed research published, or is that the case only if it's difficult to find someone who reviews it and then considers it worth publishing? What was the response in the atmospheric science community at that time? Why wasn't the model updated in the meantime? I don't see any exodus of scientists from IPCC, and if there were people changing their minds, ExxonMobil's PR folks would be all over it.

Why weren't scientists convinced that this report had an impact, then or in the 3 years since? If scientists aren't impressed, why is is presented to the public as if it were significant?

Dr S Fred Singer is a co-author of the paper, and to facilitate discussion on it, here's a link to his press release on the paper you're referring to:
http://science-sepp.blogspot.com/2007/12/press-release-dec-10-2007.html

"Climate skeptic S. Fred Singer founded Science and Environmental Policy Project in 1990. SEPP's mission is 'to clarify the diverse problems facing the planet and, where necessary, arrive at effective, cost-conscious solutions.'"
http://www.exxonsecrets.org/html/orgfactsheet.php?id=65
Dr S Fred Singer's Science and Environmental Policy Project has received $20,000 from ExxonMobil.

"SEPP was the author of the Leipzig Declaration, which was supposedly based on the 'scientfic' conclusions drawn from a November 1995 conference in Leipzig, Germany, which SEPP organized with the European Academy for Environmental Affairs. SEPP publicly used the Declaration to suggest there is little scientific consensus on global warming. According to P.R. Watch, news reporters discovered that in the end, twenty-five of the signers were TV weathermen - a profession that requires no in-depth knowledge of climate research. Other signers included a dentist, a medical laboratory researcher, a civil engineer, and an amateur meteorologist. Of the 33 European signers, four of them could not be located, 12 denied ever having signed, and some had not even heard of the Leipzig Declaration."
http://www.exxonsecrets.org/html/orgfactsheet.php?id=65

In a February 2001 letter to the Washington Post, Singer denied receiving funding from the oil industry, except for consulting work some 20 years prior. His organization SEPP, however, received multiple grants from ExxonMobil, including 1998 and 2000.
http://www.exxonsecrets.org/html/personfactsheet.php?id=1

So are we to believe that this guy who has set up his own organization to work for the past 16 years to discredit global warming and allegedly has multiple misleading acts to his credit is now dealing in the truth?

John Christy allegedly has a history of changing his mind on global warming:
"While he now acknowledges that global warming is real and the human contribution is significant, Christy has been a long-time skeptic who previously argued that satellite climate data do not show a trend toward global warming, and even show cooling in some areas. His findings have been widely disputed."
So global warming was not real, then it was real and the human contribution was significant, now his latest paper (not reflected in the excerpt above) says it's not? Which way is it this week, John?
http://www.exxonsecrets.org/html/personfactsheet.php?id=903

Of all of them John Christy is the most interesting to me. He's a professor and Director, Atmospheric Science Department, University of Alabama at Huntsville. He's in his field and I actually like that he seems to change his mind a lot since that indicates that he may be less attached than some to a particular answer.

Atmospheric science is extremely complicated and the models do need to be improved. I'm curious to see how this paper is received and where it might lead.

On the other hand, why do you suppose a scientist would veer sharply away from the scientific study to make the specific point that Singer makes on his blog: "attempts to control CO2 emissions are ineffective and pointless. – but very costly"? Suddenly he assumes the role of policymaker and economist? How much of that was present as intent as he conducted his analysis? Given that slip and his clear ties to industry opposition, I have to conclude that Singer seems to be acting squarely in his SEPP advocacy role, so until his claims convince a healthy percentage of the people working in this area I'm back to square one, questioning the context and intent of his research and result. To me it appears to have about as much credibility behind it as the Leipzig Declaration that Singer authored in 1995. He sure has been at this detractor business a long time. Can't the GW opposition find anyone without this kind of history to publish something?

Answer 5

Dana-->

your link does not work, could you post the reference. I might be able toget around the fee.

EDIT:

Larry's link works for me. ICECAP seems to have some server problems on occasion. I can't always access it when I want to.

If you still need another link, let me know and I'll see what I can do via the local university library.

EDIT 2:

http://www3.interscience.wiley.com/cgi-bin/fulltext/117857349/PDFSTART

Hopefully this will work for you. I accessed it via the university library, so I am not sure that you will also be able to get it.

Answer 6

I think that this point has been brought up numerous times, that tropospheric temperatures have been warming slower than the surface record. Which indicates that the surface warming is not being caused by the atmosphere. There are other points that have been brought up as well. They will eventually all make their way through the peer review process, so I think this is just the first in many of such papers that will cast doubt on the possibility of atmospheric trace gasses altering climate, when they clearly never have done so in the past.


http://canadafreepress.com/index.php/article/908

http://www.care2.com/news/member/101945465/566559

EDIT:

One flaw in the model?

The South Pole is not warming, and it is not because of Ozone, to indicate such would mean that you do not understand the absorptivity spectrum of O3.

.
.

Answer 7

Since warming of the troposphere does not adhere to the current models, that indicates that the models are flawed. I think it provides evidence that skepticm of the models is justified in my opinion. What is left for the global warming doom and gloomers except reliance on their unreliable models.

Answer 8

Link is broken without a cookie. How about a real citation?

Answer 9

They are credible. I haven't read the paper in question but I'm familiar with some of the work leading up to it.

Answer 10

Here is a working link:
http://icecap.us/images/uploads/DOUGLASPAPER.pdf

Dana, yes, I have it up on another tab now.