Unauthorized notes, ~ Ben Santer ~ The General Public: Why Such Resistance?

Today I was asked if there's a transcript available of the Ben Santer talk. To the best of my knowledge there is none. However, when I hear an especially good lecture, I'll return to it and do some detailed note taking along with transcribing specially valuable sections... I find it an excellent learning exercise.

So below are my notes, transcribed sections are in "quotes," the rest is paraphrased, with any personal commentary in {brackets}.

It isn't as polished as it should be and unfortunately right now I don't have the time to invest, however at some point I will return and clean this up...


~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
YouTube Uploaded by StanfordUniversity on May 13, 2010
PhD. Ben Santer and the climate debate -
“The General Public: Why Such Resistance?”
http://www.youtube.com/watch?v=OTsc3jV1Otw
{1hr 45min}

(February 25, 2010) Ben Santer, a research scientist from Lawrence Livermore National Laboratory, discusses the recent problems with the use of the freedom of information act for non-US citizens to demand complete records, including emails, on scientific research projects. Santer posits that this is a dangerous dilemma that will ultimately inhibit scientific research.

This course was originally presented in Stanford's Continuing Studies program.

Stanford University:
http://www.stanford.edu/



{ Schneider - Gives a little background to “the IPCC’s first ugly scene”
Ben Santer’s Report on the “Finger Prints” of human activity on climate.}

1:30 "Basically those finger prints pointed to a discernible impact of human activities on climate. This triggered a massively ugly scene, at the, it was the first ugly scene at the IPCC history at a plenary session. Which is where the hundreds of politicians get together and have to agree to the IPCC Summary For Policy Makers word for word. And the guy being attacked was him (Santer). So it was because of this chapter saying there was a “discernible impact”, though that word came out later ...

"The chief Saudi negotiator basically said this was bad science, tried to drive a wedge between north and south countries, was very effectively doing it. And there was what is called a Contact Group, which is where when you can not agree on the plenary, you go off and negotiate. We negotiated like all day to get this language right."

2:30 "Now the Saudis who made all the fuss and the Kuwaities never sent anybody. And they add many, many delegates. And this one guy from Kenya came, and he had actually proposed dropping the entire chapter because he believed the Saudi.... this was the southern solidarity. Now this guy, he was meteorologist, not a famous science star, came and he watch the entire day and changed his mind and decided the process was fair and open.

"So we go back and the normal practice in IPCC after a Contact Groups where the group agrees, is that the text is put up there on a screen. And it is pro-forma accepted because if you started fighting over it again you’d never get out.

"Well of course the Saudi’s immediately raise their hands and Al-Saban starts in: this is unacceptable to us."

3:15 "So Ben had the temerity, this mere scientist to say: “But, Sir your delegation made the most noise and you did not even have anyone at the group.
And El-Saban slams his fist on the table and (exclaims): I’m a representative of a sovereign country, you’re just a scientist, you can not talk to me like that and we’re a small delegation, we didn’t have time!"

3:35 "The Kenyan guy raises his flag, my stomach is in knots, and he get and says, I’m a small delegation, I’m it. But I was convinced by the Saudi’s that this was really important, so I went. I’m now satisfied the lead authors are correct and I withdraw my objection and urge everyone to vote for it. And it passed."

3:55 "Then the Wall Street Journal started. . . scientific cleansing
4:00 "Because in an IPCC report, every single meeting is about revising language. So at the direction of the plenary he (Dr. Santer) corrected the language, to which these guys at the Wall Street Journal redacted the components... that were caveats, therefore this was a distortion of that."

4:25 "So there’s this mister mild mannered Ben Santer the statistician, quiet guy on the block. All of a sudden now out there as this object of derision by these manufactured, trumped up charges of the ideology of no government control and the fossil fuel industry."

"Unfortunately for them, Ben very shortly thereafter got a MacArthur Genius Award... because this was the first scientific study really showing that there was a discernible impact of human activities on climate by good statistical testing. So Ben has now worked in many other areas. He’s shown why the satellite data measurements were valid, but not the way the original guys showed it. Why a series of recent papers now which purport to show why the mainstream is wrong should never have gotten through peer review, and so forth, for which he has been FOIA attacked, he’ll tell you all about that.
"So here mister mild manner will come up and tell you about life in the scientific fast lane is like when you hurt the interests of trillion dollar industries."

Ben Santer’s talk:
5:45 ...spending past four months defending friends and self
6:05 I going to tell you what has caused me the most personal concern... my concerns and personal difficulties
6:15 2008 paper...
6:20 sort of a morality... story about science, non-science, and non-sense
6:40 Satellite data... 1990... folks who did the original research suggest that the atmosphere was cooling... that was a huge problem, ... models... 7:10 how could the surface be warming and the air above it cooling?
7:30 This was a fundamental problem to the theory
7:40 Break history of the scientific problem
A brief history of the MSU debate (2000-2006)
MSU revisited: the claims by Douglass et al. (2007) and Singer et al. (2008)
Statistical issues
Douglass et al. and Santer et al. significance tests
Audits, FOIAs, and lessons learned

8:40 review of the statistical issue involved
“inconvenient observations” (James Schlesinger quote at 8:48 - claim no atmospheric warming proves no global warming.)
9:45 explaining the physics
10:15 University at Huntsville (UAH) observations and the great satellite debate
10:40 Constructing climate quality temperature records from satellite data... most difficult task
initially UAH - John Cristy and Roy Spencer late 90s another team started process this data.
Remote Sensing Systems (RSS) - a private company
Frank Wentz and Carl Mears
10:50 UAHs initial records indicated no tropospheric warming.
12:40 graph (1979-1904) of UAH satellite dataset implied that the troposphere cooled as the tropical surface warmed
13:10 a lots of squiggles - we knew why a bunch of ups and downs happened, what we didn’t understand was why the ground temperature record trended up, while the satellite troposphere measurements trended down.
13:30 RSS then looked at the same raw satellite data but produced a strikingly different graph, where both surface and troposphere temps trended up
14:15 Why did they get different results from identical data?
14:25 .. satellite orbital drift (14:30 graphic)
15:10 UAH was aware of this drift and has
This systematic drift in the time a satellite passes over a specific location needs to be accounted for, which the UAH did since the drift rate was known.

15:25 UAH “recognized that, that this was a problem, that they had to account for this drift effect, but they got the sign wrong, they literally got the sign wrong. They made a sign error in correcting for the effects of satellite orbital drift.”
15:40 Three papers published in Science partially resolved the “great MSU debate.”
“The Effect of Diurnal Correction on Satellite-Derived Lower Tropospheric Temperature”
“Radiosonde Daytime Biases and Late-20th Century Warming”
“Amplification of Surface Temperature Trends and Variability in the Tropical Atmosphere”
~~~
“Alabama group argued there was a conspiracy to publish these papers at the same time. There was no such conspiracy. These were independent efforts here.”

16:00 First paper was by the Santa Rosa (RSS) team explaining the UAH data processing error.
16:35 UAH claims they can’t be wrong because they are in agreement with weather balloon data.
Santer explains balloon data gathering... thermometers have thermo shielding to protect against the heating of direct sunlight... over time the shielding has become better and better thus introducing some non-climatic bias into the weather balloon data.
17:40 The second paper was by Yale University and the Geophysical Fluid Dynamics Lab at Princeton.
Balloon data is taken twice a daily, mid-day and mid-night. This paper examined these two set independently and found the night time temps were increasing, were as mid-day measurements didn’t.
19:00 The third paper was Santer’s team. Our paper showed that when you looked at the adjusted data, when you didn’t have this problem with the change sign error, when you look at adjusted weather data sets for the thermo shields, there was no disagreement between climate models and what they told us about the expected changes in atmospheric temperatures between observation and basic physical theory. We thought we had resolved the problem.
19:40 Temperature Trends in the Lower Atmosphere - Steps for Understanding and Reconciling Differences. A report to congress (2 dozen) that involved all the major protagonists UAH, RSS, weather ballon experts, surface temp experts, climate modelers.
The report included dissonant voices.
20:20 The reports conclusion statement: “Previously reported discrepancies between the amount of warming near the surface and higher in the atmosphere have been used to challenge the reliability of climate models and the reality of human induced global warming. This significant discrepancy no longer exists.”
So we thought that was the end of the story... not quite

20:45 In 2007 a paper was published
~ Douglass et al reached conclusion very different from those of the first US CCSP Report.
IJC 2007 ~
Douglass and astronomer
Christy who was involved in the UAH data set
“Using the same data that was used to reach the CSSP report’s conclusions they reached a fundamentally different conclusion”
22:10 Fred Singer one of the authors of report based on single study Douglass et al claimed: “Nature, Not Human Activity, Rules the Climate” and it became the center piece for the: “Summary for Policymakers of the Report of the Nongovernmental International Panel on Climate Change”

22:40 Singer claimed: Douglass et al. findings are “an inconvenient truth”, and prove that “Nature rules the climate: Human-produced greenhouse gases are not responsible for global warming.” US National Press Club 2008
“These were extraordinary claims, to say that one paper can literally turn the world of science on its head and somehow magically counteracted literally hundreds of scientific papers that showed that there was a profound human effect on climate.”

23:00 “How could they make this claim? OK, in order to look at how they could make that kind of claim, we need to get into some statistical issues.”
23:40 Signal and noise in model data. . .

{~ ~ ~ education time ~ ~ ~}

31:15 “Ok, so what did Douglass et al. do. . . “
What are the elements of the Douglass et al. “robust consistency test”?
Explaining, “took 19 models combined and took the average trend, subtract form that the single observed trend and then we’ll divide by some term and called the result the estimate of uncertainty in the multi-model trend.”
31:50 “That estimate of uncertainly is something statisticians call the “standard error of the mean.” It’s an estimate of how well you know some mean quantity from a finite sample of results.”
32:20 “Let me show you what that looked like...” {explaining graph}
33:15 Getting to the nitty gritty of the Douglass at all “element III of the Douglass et al. “consistency test”: observed temperature trend.”
33:50 But, you can also see if you... if you applied their test to each of the individual models, then 11 of the 19 would be deemed not models. They’d be deemed inconsistent with the model average. Any test that says that half the things that you use to perform the test are actually inconsistent with the average is a really, really weird test. This is not even an undergraduate type of error that you would make in the first year of statistics. ... it doesn’t pass any of the most basic sanity tests.”
34:45 “What did they do wrong? First of all they assumed that trends were perfectly known, both models and observations. That’s wrong as I’ve showed you noise is an important factor and you have to incorporate that uncertainty in estimating trends as part of your test. That’s really important and they didn’t do that.”
~~~~~

What were the principle errors in the Douglass et al. statistical test?
~ They ignored the influence of interannual variability on the observed trend.
~ They assume, incorrectly, that the “signal” component of temperature change in observations is perfectly know.
~ Their use of “sigma-S-E” term is incorrect.
~ sigma-S-E is an appropriate measure of how well the multi-model mean trend can be estimated from a finite sample of model results.
~ sigma-S-E (standard error of the mean) is NOT appropriate for determining whether the multi-model trend is consistent with a single (uncertain observed) trend!

35:45 When Douglass et al. was released it receive extraordinary media coverage...
36:20 “.. it was so clear that something was wrong in this (Douglass) paper and that needed to be brought out into bright daylight.”
37:00 Douglass et al. with error bars includes... {revealing}
... there is no statistical difference between the models and the observations...
37:55 “OK, but people might say: well your test is flawed too. I don’t believe that the test you’ve come up with that incorporates this uncertainty in the observations is a meaningful one, or appropriate test. One way of addressing that kind of criticism is by performing tests with random data...”
39:10 Here you see the results of such tests...
40:25 this is how the Douglass test behaved...
40:35 “... What happened is that more than 80% of the time with randomly generated data their test would conclude that the randomly data sets were significantly different. It was absurd, completely absurd, demonstrably wrong, we are not talking about shades of gray difference here about opinion
here between reasonable scientists, we are talking about something that was simply incorrect.”

41:15 poster
What were the bottom-line conclusions of the Santer et al. paper?
~ Our results contradict Douglass et al.’s claim that all simulated temperature trends in the tropical troposphere and in tropical lapse rates are inconsistent with observations.
~ This claim was based on use of older radiosonde and satellite datasets, and on two methodological errors.
~~ Neglect of observational trend uncertainties introduced by interannual variablity
~~ Application of an inappropriate statistical “consistency test”
~ We find that there is no longer a serious discrepancy between modeled and observed trends in tropical lapse rates. {published Oct 2008}

41:20 “The Douglass et al. claim was based on older data sets. They had actually ignored data sets that showed warming of the tropical atmosphere even though they had those data sets in their possession. But more crucially their claim was wrong because it was based on two serious methodological errors. They assumed that the trends and the observations and the models were perfectly known. No uncertainty which was clearly wrong. And they applied this inappropriate statistical consistency test which they had devised themselves.

Now you would think if you come to a conclusion that turns the world of science on its head and claims to undercut the findings of IPCC Fourth Assessment report, claims to undercut the finding of the US climate change science report, you would ask why. Why did we get a completely different result that these other guys who were using the same data as they did.”

42:20 “They did not ask that question. They got the result that they liked and they published that result.”
42:30 “Nature wrote a little thing saying this is now the end of the story the riddle is resolved. We thought so too.”
~~~~~~~~~~~~~~

43:00 Santer’s paper was published October 8th, 2008. Steven McIntyre requests “the model monthly data (49 series) used for statistical analysis in Santer et al. 2008 while on travel in Hawaii and in the UK at the time McIntyre’s request arrives, when I got back to my office...
43:30 “I basically told McIntyre replied if you wanted to audit our findings you are free to do so, the climate model data that we use are freely available, in fact, three and a half thousand researchers around the world use the climate model data that we have archived at Livermore, it’s an open data base... Mr. McIntyre had in his possession, or had the ability to access exactly the same data that we have used in our study. And indeed I should point out that the Douglass et al. paper used the same archive...”

44:20 The freedom of information requests
44:50 “He (McIntyre) then began to express moral outrage on his blog, ‘Santer was stonewalling him’. I began to get hate mail. I began to get threatening letters... etc...
45:30 ... I released all the intermediate data on PCMDI’s website January 13rd, 2009. Again the raw climate data that I used was freely available to anyone in the world, even Mr. McIntyre...
46:00 some excepts from McIntyre blog revealing...
46:40 “This did not give me a warm and fuzzy feeling that this was a scientific peer who was truly interested in scientific discovery and understanding the nature and causes of climate change...”
47:05 “so I think this all boils down to a couple of issues this freedom of information request stuff. And as you can see this is at the heart of much of the climategate debate...”
47:20 I believe Phil Jones got into the difficulties he did because of McIntyre... over sixty FOIA requests in a short time...
47:35 “I believe the purpose of those requests was not valid scientific discovery, but harassment and that’s very troubling to me.”

47:45 “What questions did McIntyre’s freedom of information requests raise in our case? First of all what he wanted was the intermediate calculations that we had done from the raw climate model data. So a reasonable questions is, did we do those intermediate calculations correctly? Now, we had already been audited, Douglass et al. to their credit had gone back to the raw data, they had already found very similar intermediate results to the results that we had. The difference between the Douglass et al. and the Santer et al. paper was never in the intermediate calculations. It was in the statistical tests that the two groups had used. So McIntyre’s request was completely frivolous, because we had already been audited effectively on the first part of the problem, did we do the intermediate calculations correctly.
Did we apply the appropriate statistical tests? Again, as I’ve showed you one doesn’t have to stand up there and say trust me... there are ways of adjudicating, you can look at how these test perform with randomly generated data.”

49:00 “Clearly as I’ve shown you the Douglass et al. test fails. It does not perform according to expectations, it does not perform as a reasonable statistical test should perform...”
... and ours performed to theoretical expectations.
49:30 “So question two, Yes, we had examined how these two sets of tests performed, ours performed according to theory, Douglass et al. test did not perform according to theoretical expectations. “
49:50 “Question three: And this is the more difficult one” What is our legal responsibility as scientists? Are we required to provide self appointed auditors who have no training in climate science everything they ask for? Every single little thing, email correspondence, intermediate calculations, explanations of how to use computer codes. These are all things that have been requested in the past. Do we have that requirement?
“And if we don’t give some self appointed auditor everything they requests can they make our lives miserable?...

50:40 “This is why I am in trouble now I’m in trouble for setting the record straight on a paper that should not have been published.
“I don’t think we have that responsibility. I think we have a reasonable responsibility as scientist to insure that others can replicate what we have done. That they have in their possession all the data that we used at the outset of our study, that they can see whether we did all calculations correctly, and whether we drew appropriate conclusions based on those calculation that we did...”

51:35 Back to science. . .
52:00 the non-science is what Douglass et al. did. They did not behave like normal scientists would do. They used the same data that others had used, models and observations. They got fundamentally different results after devising their own statistical tests.
“And rather than probing, and then asking is our test appropriate, are the inferences that we drew appropriate. They just ran with the result that they got.

52:45 “And the non-sense I think is what Mr. McIntyre has done.
All of this expression of moral outrage on his blog and playing this game where you literally insight others people to hatred, and get them to do all sorts of things that have nothing to do with science whatsoever. And nothing to do with the search for scientific truth.
Again from my perspective Mr. McIntyre had access to all the information he needed, both model and observational data to test our findings and determine whether we are right or we are wrong.

I think he’s using freedom of information requests not only in my case but in many other cases as a mean for harassing and intimidating other scientists. I think there is a calculated attempt to mandate the scientific equivalent of what people used to call “no-go areas” in Northern Ireland. You go there and come up with results that we don’t like. We’ll be down on you like a ton of bricks. And paleoclimate is one of those areas now and also the micro-wave sounding satellite data that’s another one of those “no-go areas.

“I don’t think we can let this stuff stand. Now more than at any other time we urgently need people looking at these key areas of climate science and not be intimidated by bullies.”

54:10 “One of the really troubling things is the asymmetry in this auditing. Mr. McIntyre purports to have significant statistical training, yet he did not audit the Douglass et al. paper which had a grievous and very obvious statistical error. If someone really where interested in dispassionately getting to the bottom of things and why two papers reach fundamentally different conclusions they would not behave in the way the Mr. McIntyre has behaved.”
54:50 Climategate tells us that there are folks out there, forces of unreason who are not really interested in improving our understanding of the science. And the very sad thing to me is that even after publication of our paper Douglass et al. have not admitted any error in what they did.
55:15 Timeline of key events related to the publication of the Douglass et al. and Santer et al. papers.
55:30 Based on their selective analysis of the hacked emails of the Climatic Research Unit David Douglass and John Christy published something at AmericanThinker blog, in which they accused me of serious professional misconduct. They did not admit any error in their statistics, but they accuse me of conspiring to delay publication of the print version of their paper...

... Inhofe and Barton are now investigated Santer ...a
57:00 It is kind of a morality play for me, I guess to be filed under the category of no good deed goes unpunished...
57:15 Science is not self correcting, there is no magical correcting out there. When a paper has been published, and extraordinary and wrong claims are made on the basis of that paper, people have to do the work of exposing the errors, remedying these errors, trying to explain what went wrong, trying to do it right there is cost involved in trying to do that.
But, that does not absolve us from the responsibility of trying”

~ ~ ~ ~ ~ ~ ~
Q&A


1:10:45 “... as with this latest Inhofe thing. Talking of criminal prosecution... I believe it is McCarthyism... targeting of individuals because of the work they do, and because of what that work shows. It’s as simple as that. It is chilling I think for young people, but I guess the encouragement I would like to give people in the audience who are thinking about working on these issues: don’t give up, work on this stuff, it’s critically important. You can’t be silenced by the like of these guys at all. One cannot leave the playing field to these sorts of customers.
1:16:00 ... his request that Santer et al. 2008 be published as a stand alone paper and not as a commentary...
That has now email has been portrayed “as me exerting evil influence on the journal to do my bidding, I believe psychologists call that kind of behavior “Projection” when you actually project onto others your own values and beliefs and behavioral patterns.
You know, from my point of view we did absolutely nothing wrong here. We did the science in order to address a flawed paper, we published that. It was entirely appropriate to ask whether our paper would be considered as a stand-alone contribution. It was and it should have been and I had no control over the decision to hold back the printed version of their paper.
Again, the printed version wasn’t the one that got the attention. It was the on-line (Douglass et al.) paper that had been printed ten months before ours...” Fox & Press Club attention.

1:19:10 “... it’s part of being a scientist. We have the responsibility not only to do research but also to set the scientific record straight, to point out abuses when extraordinary and wrong claims are made on the basis of demonstrably flawed stuff. You can’t ignore it, you can’t let it slide. If you do that contributes to public misunderstanding of the most basic elements of climate science.
People say: ‘well yea, I don’t believe any of this stuff because satellite data, these guys showed that the atmosphere is cooling”... Many people remember that the atmosphere is cooling claim very few people remember that claim was made of the basis of a sign error, the guys got the sign wrong.
Many people believe and remember the Douglass et al. claim, ‘didn’t these guys show that models and observations are just at odds and their really statistically different?’ That got on Fox news. No body remembers that another paper came along and showed that these guys had screwed up the most basic statistical test. So it is incumbent on all of us to set the record straight.”
1:23:55 “The bottom line is that no matter how loudly these guys shout and scream you can’t change the science. Ultimately scientific truth will prevail...”
1:26:50 “... I think there’s a game here that this self appointed auditor can request whatever he wants from climate scientists he can use the power of his website to stir people up into some frenzy. Get them all riled up, ‘so and so has not given me exactly what I wanted.’ Do something about it...
1:27:30 “ ... in part it’s based on this intemperate language that goes on at some of these websites. In part because of the intemperate language of people like Senator Inhofe, who refers to us now as criminals, climate science as a hoax and fraudulent. Those kinds of words are not conducive to rational debate, rational public discourse at all.”

1:35:50 Q: What are the journal standards in terms of providing data to reviewers.
{Interesting in depth look at a scientists source codes and whether or not that should be made available.}
1:38:55 Schneider: “... the process of science is much better served by having independent codes... ”
1:38:30 Schneider: “... so my journal... decided that the data had to be there, the methods had to be very explicit, no hidden how you did the equations, and that algorithms that you used were there, but that the codes were not to be provided and in fact the National Science Foundation has very recently agreed that its investigators are not required to turn over personal source code to any request...”

1:40:20 “Just following up on what Steven said, the key thing here is not only were the primary model data available to Mr. McIntyre, the algorithms that we used to do the intermediate calculations had been published, were available in the peer review literature. Our statistical tests were spelled out in excruciating detail in our paper. There was never any question on the reproducibility issue. This man had access to all the information that he needed in order to try and replicate our calculations and figure out whether we had reached appropriate conclusions. As I mentioned Douglass et al. had already sort of audited the intermediate calculations. McIntyre was unwilling to do that.”
1:42:00 “I think this latest swing of the pendulum which was mentioned today, criminal prosecution or things like that being threatened. May hopefully get some attention.
The only crime I’m guilty of, in my mind, is trying to set the scientific record straight, I don’t think that’s a crime yet...”

1:42:30 “People in the end will realize that this is not a political issue, this is not an ideological issue, it should not be. It’s a science issue and it behooves all of us to understand the science, to take the decisions on what to do about it on the basis of the best available science. Clearly the Douglass et al. paper was not the best available science. Extraordinary and false claims were made on the basis of that. It’s sad how those claims are now being accepted as the truth and now being woven into the fabric of these kinds of criminal prosecution things. But, in the end, the climate system, no matter how loudly these people can scream the climate system is not going to be changed by this, it’s not going to be phased by this. It will continue to warm. The sad thing is that action on addressing that problem will likely be delayed because of this information.


~ ~ ~ open, honest and eloquent... worth taking the time to listen to the entire talk in person. ~ ~ ~

Ben Santer PhD.
The General Public: Why Such Resistance?
http://www.youtube.com/watch?v=OTsc3jV1Otw