RealClimate logo

Palin on Global Warming

Filed under: — group @ 5 October 2008 - (Italian)

Here at RealClimate we understandably have an intense interest in the positions of the Presidential and Vice-Presidential candidates regarding global warming and carbon emissions. What the stance bodes for future action on climate change is consequential in itself, but beyond that the ability to use sound science in this case serves as a bellweather for the candidates’ whole approach to science. Whatever else you can say about the candidates, it has been encouraging that both John McCain and Barack Obama favor mandatory action to reduce US carbon emissions.

But, enter Gov. Sarah Palin, McCain’s pick for VP. Palin’s position on global warming has been stated quite clearly in this recent interview with the publication Newsmax , where she says “A changing environment will affect Alaska more than any other state, because of our location. I’m not one though who would attribute it to being man-made.” How is this to be reconciled with McCain’s position? Do they just agree to differ? What does this bode for future actions if McCain were to win the election, especially in view of the fact that, in a Cheney-esque way, Palin is likely to be put in charge of energy policy? The recent vice-presidential debate sheds some light on the issue. A full transcript of the debate is here.

Palin seems to be attempting to defuse the whole issue by claiming the cause doesn’t matter. When the moderator asked her ” What is true and what is false about what we have heard, read, discussed, debated about the causes of climate change,” Palin responded as follows:

“PALIN: Yes. Well, as the nation’s only Arctic state and being the governor of that state, Alaska feels and sees impacts of climate change more so than any other state. And we know that it’s real.

I’m not one to attribute every man — activity of man to the changes in the climate. There is something to be said also for man’s activities, but also for the cyclical temperature changes on our planet.”

I’m pretty sure that that last statement is a garbled attempt to reiterate what she said in the Newsmax interview, but you be the judge. Unlike the previous quote, this one at least has a nod in the direction of acknowledging (tentatively) the possibility of a human influence. What’s important is what comes next:

“But there are real changes going on in our climate. And I don’t want to argue about the causes. What I want to argue about is, how are we going to get there to positively affect the impacts?”

Dare we say that it, in fact, very much makes a difference what is causing global warming? If CO2 really weren’t a major part of the cause, what in the world would be the point of John McCain’s (or anybody’s) stated policy of acting to reduce emissions? And even if you were of the school that says adaptation is better than mitigation, knowing the cause is an important part of knowing what kind of climate change you have to adapt to, how long it is likely to last, and how much worse it is likely to get in the future.

Biden’s answer, by comparison, was direct, straightforward, and simple:

“BIDEN: Well, I think it is manmade. I think it’s clearly manmade. And, look, this probably explains the biggest fundamental difference between John McCain and Barack Obama and Sarah Palin and Joe Biden — Governor Palin and Joe Biden.

If you don’t understand what the cause is, it’s virtually impossible to come up with a solution. We know what the cause is. The cause is manmade. That’s the cause. That’s why the polar icecap is melting.”

Well, maybe he left out the kind of caveats and qualifications you’d attach to the attribution of the recent loss of (North) polar sea ice if this were an AGU talk instead of a vice-presidential debate. Overall,though, the statement gets to the heart of the matter.

One can moreover doubt even Palin’s commitment to dealing with the consequences of climate change. Surely, that would include doing something to save the polar bears,yet the State of Alaska (against the advice of its own wildlife biologists) is suing the Interior department over its decision to list the polar bear as “threatened” — and this despite the fact that the Bush administration put so many qualifications on the listing as to make it essentially toothless. What’s even more telling is that the brief submitted to Interior drew heavily on a list of climate skeptics (including the Marshall Institute’s Willie Soon) that could easily have been culled from the infamous Inhofe 400. (see this article). Palin’s role in bringing this case has not been peripheral; she has been very much at the center of the effort, and has consistently questioned the causal link between CO2 and global warming in making the case. As early as Dec. 2006, she wrote to Secretary Kempthorne: “”When a species’ habitat (in this case, sea ice) is declining due to climate change, but there are no discrete human activities that can be regulated or modified to effect change, what do you do?” Further information about Palin’s long fight against the listing, and her view of the scientific issues involved, can be found here.

We will take this occasion to note also that Biden used the debate to reaffirm Obama’s long standing position in favor of “clean coal.” Whether this is a good or bad thing depends on the extent to which the candidates understand what should really be meant by this term. From the point of view of global warming, the only “clean” coal would be coal burned with 100% carbon capture and sequestration — certainly worthy of research and pilot implementation, but not by any means a technology that can be counted on at present to solve the problem. (And of course, the term “clean” is even then relative, since what mountain top removal mining does to the West Virginia hills and rivers is anything but “clean”).

So there you are. We report, you decide.

290 Responses to “Palin on Global Warming”

  1. 151
    Rod B says:

    Chris (134), I don’t think you stated the key questions correctly. The question is not if natural forcings overwhelms the greenhouse signal, but whether it has a significant (or the chances of) part. The physics of GHGs absorbing energy and keeping some within the biosphere is not at question. Plus it is possible, given the sink mechanisms, to add GHGs without seeing a rise in temperature, at least over the long-term. So? It seems like a no-op trivial question. Though your answers were strangely more to the point.

  2. 152
    Rod B says:

    Ray (139), if an action has a 90% confidence (probability) of being X, does it not simultaneously have a 10% confidence (probability) of being something other than X? If the IPCC believed with certainty that natural causes had zero part in the warming, they should have said so. But they didn’t. Now it seems you (guys) are trying to explain why they didn’t really mean what they said, or if one analyzes it with enough higher math it can be shown that “90%” really means 100%. Sounds confusing.

    [Response: No. All you can say is that IPCC thought that there might be up to 10% chance that the statement might be false. It doesn’t imply that there is any positive evidence for the contrary position. – gavin]

  3. 153
    Rod B says:

    re Lawerence (150): “…>90% == “Very likely”, >95% == “extremely likely” implies: 90.01 – 94.99 is in the range “very likely””.
    But isn’t also 90.01 – 90.02, 94.92, 93.8, 98.6, 99.99? If so, what’s the point?

  4. 154
    Rod B says:

    Gavin, but if a statement is false, doesn’t that mean it has to be something else — though not logically necessarily “contrary”?

  5. 155
    Jim Galasyn says:

    Gavin, thank you for the one-sentence summary. I’ll be using it elsewhere, for sure.

    All you can say is that IPCC thought that there might be up to 10% chance that the statement might be false; it doesn’t imply that there is any positive evidence for the contrary position.

  6. 156
    Ray Ladbury says:

    Rod B., Perhaps part of your difficulty in understanding has to do with the difference between probability and confidence. Probability has to do with the chances of an event occurring, while confidence has more to do with what we can infer given our current state of knowledge. You might look at confidence as how often we could be wrong if absolutely everything conspired against us.
    I think that my example of the urn full of balls (nominally black and white) is appropos. If we draw 22 balls from the urn and all are white, we can infer with 90% confidence that no more than 10% of the balls in the urn are black. The 90% confidence and 10% probability bound are inseparable. If we chose another confidence level, our bounding probability would be different. However, you can see that even though we have a bound of 10% at 90% confidence, you can see that you might not want to bet on black even with 10:1 odds, right? I mean, we don’t have any evidence yet that there are ANY black balls in the urn. If you took my proposition at 10:1 odds on black, I would have 90% confidence that the worst I could do was break even, while you have zero confidence that you have any probability of winning at all.
    Things get a bit more complicated when you bring in “Bayesian” vs. Frequentist probabilities, but the basics are the same. We have zero evidence that we can explain current warming etc. with “natural” causes (that is nobody has done it), and the fact that we can’t preclude with certainty that someone will be able to do so, doens’t mean we think it’s going to happen.

  7. 157
    Mark Bahner says:

    “The analogy would be a criminal trial in which the defence challenges every piece of the positive evidence indicating their client is guilty, but offers no evidence at all that anyone else committed the crime. The jury might well have some doubts remaining as to the defendent’s guilt, but there is still no evidence for anyone else’s. – gavin]”

    If the defendant has not committed the crime, then there is evidence that *someone* else has committed the crime.

    And in fact, the IPCC even names the someone else. They wrote:

    “Attribution studies show that it is very likely that these natural forcing factors alone cannot account for the observed warming.”

    By their definition of “very likely,” they are saying that there is a 90-95% chance that natural factors alone can’t account for the observed warming. That means there is a 5-10% chance that natural factors alone can account for the observed warming.

    [Response: In none of those attribution studies do the natural factors alone come anywhere close to explaining the recent trends, not even at the 5 to 10% level. The ‘very likely’ is a confidence statement, not a probability in the sense you assume. I challenge you to find one decent attribution study that shows the two-sigma attribution to natural factors anywhere close to the observed trend (for contrast see figure 9.9). The larger fraction of that 5 to 10% uncertainty is not associated with the expected influence of natural factors at all, but with the uncertainty that the attribution methodology is appropriate or that all the forcings are well described. – gavin]

  8. 158
    SecularAnimist says:

    Walt Bennett wrote: “If I was president, I’d want somebody like Sarah Palin asking certain questions, just to make sure somebody asked them.”

    Sarah Palin is not known for “asking questions”.

    [edit – this is not a forum for bashing political candidates on any thing other than their statements on climate change]

    She is known for rote regurgitation of fossil fuel industry propaganda that questions the role of CO2 emissions in causing global warming. She is known for disparaging and discouraging investment in solar and wind energy.


    If I were president — for that matter, as I am a US citizen and voter — I would not want Sarah Palin anywhere near any national office, let alone a “heartbeat away from the Presidency”.

  9. 159
    Ike Solem says:

    If Palin is going to rely on tobacco science experts like Soon and Baliunas as the basis of government policy decisions, then that is a fair topic for discussion – the core issue is still climate science.

    Michael Klare persists in misrepresenting the true purpose of the Alaska natural gas pipeline:

    The AGIA proposal, which has received more national attention, is intended to facilitate construction of a natural gas pipeline from Alaska’s North Slope to Canada and eventually the Lower 48.

    A detailed view on the Alaskan pipeline prospects and the Canadian tar sands is at Canada tar sands and the Alaskan gas pipeline blog, run by an Alaskan engineer.

    Of all fossil fuels, natural gas is the cleanest and delivers the most energy per CO2 released when burned, thanks to the four carbon-hydrogen bonds per carbon atom in methane. Gasoline has about 2 C-H bonds per C, and coal and tar sands and shale oil is the lowest. See for the basic overview.

    Thus, the order to phase out fossil fuel use would ideally be first to eliminate coal, tar sands and shales, followed by heavy sour crude, leaving you with the world’s limited reserves of low-sulfur light crude and natural gas.

    For decades, Alaskan natural gas has been viewed as “stranded asset”, because the markets are too far away.

    One option for moving Alaskan natural gas is in pressurized tankers – but few countries want to have a liquified natural gas terminal anywhere near a populated area, due to the catastrophic effects of an explosion.

    The other option is in pipelines, but natural gas pipelines must be pressurized using massive pumping stations. The energy cost varies with climate and terrain conditions, so going from the North Slope to Canada. For flat, mild conditions the loss is ~0.5% per 100 miles, and might be double that for the Alaskan pipeline.

    That’s why the pipeline terminates in Alberta, where it will be used to convert tar sands to synthetic crude for shipment abroad. The contract was given to Transcanada by Palin:

    TransCanada earns revenue by moving gas into the tar sands projects and by moving tar sand oil to the Lower 48 via projects like the Keystone Pipeline.

    For more on the Canadian tar sands see

    Another good discussion is at

    “With so much cash on hand, it’s time for the oil and gas companies to begin the transition into energy companies that recognize that if we’re to have a future at all, it will need to be powered by renewable sources.”

    Makes one wonder what the candidates would have to say about importing tar sand oil into the U.S. from Canada.

  10. 160
    Hank Roberts says:

    > a confidence statement, not a probability

    This is a really important difference, and one I doubt I understand well enough. I know that the difference and the problem understanding it has come up in discussions of medical statistics, epidemiology, and many other areas.

    If anyone has a pointer to a good clear teaching site or lesson source, please suggest it.

    Best I’ve found is the BMJ’s page here:

    “… There is much confusion over the interpretation of the probability attached to confidence intervals. To understand it we have to resort to the concept of repeated sampling. Imagine taking repeated samples of the same size from the same population. For each sample calculate a 95% confidence interval. Since the samples are different, so are the confidence intervals. We know that 95% of these intervals will include the population parameter. However, without any additional information we cannot say which ones! Thus with only one sample, and no other information about the population parameter, we can say there is a 95% chance of including the parameter in our interval. Note that this does not mean that we would expect with 95% probability that the mean from another sample is in this interval….”

  11. 161
    Mark says:


    A level maths-and-stats covered this, so this was a long time ago for me.

    But the confidence level has its basis in sampling error. If you took the average height of 100 people in a town of 10,000 residents, you would have a very precise average for those people and statistics would give you a range. So the probability of someone taken at random from that sample being within a certain height would be worked out from that set. Now you could take that probability and expect it to apply to the entire town (or the planet). But you could have selected all the short people.

    Randomly, you can’t tell.

    As a Dilbert cartoon had it, the random number generator was a troll saying “nine nine nine…” and when asked if he was really a random number generator, the other troll answered “You never can tell with random”.

    So you take your numbers and work out the probability that your sample was biased. That gives you your confidence level in your distribution inferred from the smaller sample you took.

    This does NOT mean that your sample isn’t all the short people. It’s just saying how likely it is that your statistics are wrong when applied to a group that wasn’t in the sample.

    Confidence in your prediction.

    In this case, with the IPCC report, what makes it harder is that the scientists wanted it in the 95%+ confidence limit and the politicals wouldn’t accept it (guess who were the loudest contrarians..!). So it was downgraded to 90%+ confidence.

    The confidence here doesn’t fit too well, but it could be argued as necessary because it could be a result of

    1) The FSM turning up the heat because we’re killing off the pirates
    2) Errors in measuring may be “the cause” of most of the warming
    3) Our instruments are merely getting more accurate (see the value of the electron charge as given by experiments which showed a line going up)

    With #1 it may be astronomically unlikely but that ISN’T ZERO. Add up all the nearly infinite whacko ideas out there and you could be called out for ignoring them when you cannot *PROVE* they cannot be true.

    The other two are admitting it is possible that the “real” warming is only 0.5C, not 0.8C and so the biggest part of that 0.8C shift is not man made.

    They are unlikely but they aren’t open to assessment because we can’t go back in time and use new instruments in the past. So figure in “&lt5% and who can complain we’re ignoring it” and you get the 95% confidence.

    And 5% of that is dropped because the politicians have no confidence in being able to say “95%”.

    Bringing us down to 90-95% confidence limit. 5% of which is “politicians didn’t think they could say 95%”.

    ***Apologies for posting with a less than sign***

  12. 162
    RichardC says:

    120 Roger said, “Richard C- I don’t expect the RC guys to correct either your math”

    Fortunately, the math is right here:
    Odds on percentage black in Ray’s urn are about:
    0% 20%
    1% 17%
    2% 13% note that 50% of the time there are 0-2% black!
    3% 10%
    4% 9%
    5% 7%
    6% 5%… Since you implied that my math was incorrect, please correct it, Roger.

    137 Steve, if one doesn’t know the cause, then your only choice is to do nothing about the cause – at least now, and at least by “us”, which is the goal of Palin’s “Drill baby, drill” movement. Neither mitigation nor adaptation are everything or nothing questions, but one can’t choose the best balance of mitigating cause and adapting to harm if one doesn’t have a clue about cause. And without knowing cause, there is no point in planning for future adaptation. Toss out GHG as a known cause and there is no reason to believe that we won’t enter a new ice age. So which adaptation do we choose, to hot or cold? We’re stuck with adapting to whatever is here and now. Your stance inevitably results in reactive policy.

    138 Rod, I accepted Roger’s contention that 95-100% was out of bounds for the sake of discussion. I agree with you that Roger was even more wrong than I said. Gavin’s comment that there was probably debate as to which phrase to use nailed it to 90-100%. Mark’s explanation of what the debate actually was set the nailhead. The answer seems to be: “durn near 100%” with a reduction to 95-100% to account for mankind’s lack of omniscience, and a further reduction for political reasons, which can and should be ignored.

    143 Ike said, “I just read Roger Pielke Jr.’s post. [self-edit].” Ike’s a wiser man than me.

    The Trans-Canada pipeline you speak of doesn’t reach the USA, but a second pipeline might subsequently be built linking Alberta to Chicago, so a bit of gas might make it to the USA by 2030 or so. The competition for the pipeline is CANDU nuclear reactors. Should the tar sand business go nuclear, then you’re right, there would be no economic value in the pipeline. Another pipeline, the All-Alaska pipeline could substitute. It would follow the oil pipeline route, bringing gas to Fairbanks and Anchorage, and load the remainder onto tankers. Given the timeframes, perhaps they should just build a port on the north slope, taking advantage of the ever-lengthening shipping season. By 2030, keeping the lanes open shouldn’t be an issue. First year ice isn’t terribly dangerous. As long as it all melts each summer, tankers can cruise to the north slope almost year round. Plow the lanes as needed and drill, baby, drill.

    145 Geoff, this thread is obviously an attempt to link current events to climate science. It obviously is sequestered from the rest of the RC site. The RC staff obviously have limited their comments to be as close to their vision as possible. Some of us who have posted here have not met those same standards (I haven’t, for example). So were they wrong to allow these comments to post? It depends on your reaction to Mayberry’s Goober yelling, “Citizen’s arrest!” while chasing jaywalkers.

  13. 163
    Ray Ladbury says:

    Probability vs. confidence. Mark is basically right–confidence has more to do with sampling errors. In essence, the question is if nature became very perverse and gave you a very improbable sample given your parent probability distribution, how would it make you draw the wrong conclusion. Note that a confidence level does not necessarily imply that there is ANY evidence of a biased sample. It merely says that the evidence available to date allows us to eliminate the possibility that we’re wrong for all but the 10% nutty samples that could occur. And given that we know CO2 is causing at least some warming, while we have zero evidence of natural factors conspiring, this is a very different thing that saying we have 10% probability.

  14. 164
    Mark Bahner says:

    Gavin Schmidt writes, “The ‘very likely’ is a confidence statement, not a probability in the sense you assume.”

    So you say. But that’s not what the IPCC says. Go to page 23 of the Technical Summary (TS). You’ll see:

    “The standard terms used in this report to define the likelihood of an outcome or result where this can be estimated probabilistically are:

    Likelihood Terminology Likelihood of the occurrence/ outcome

    Virtually certain > 99% probability
    Extremely likely > 95% probability
    Very likely > 90% probability”

    On page 22, they explained what their wording was when the wanted to express confidence. For example, “Very High Confidence” is explained by the IPCC as meaning, “At least a 9 out of 10 chance of being correct.”

    Now perhaps you think the statement that the IPCC made was wrong. Perhaps you think they should have written something like, “Based on attribution studies, we are highly confident that these natural forcing factors alone cannot account for the observed warming.”

    But that’s not what they wrote.

    He continues, “I challenge you to find one decent attribution study that shows the two-sigma attribution to natural factors anywhere close to the observed trend…”

    Hey, if you don’t like what the IPCC wrote, talk to them. They wrote the report, not me. (If *I* had written the report, I would have at least had the technical ability and honesty to come up with falsifiable predictions of future climate forcings and global temperature changes. But I digress.)

  15. 165
    Rod B says:

    Ray (156), you seem to have a point. I can see the semantic/definitional difference. It’s still a bit fuzzy what the IPCC truly meant; I gots to go back and look at it.

  16. 166
    Mark Bahner says:

    Oops. I wrote, “Based on attribution studies, we are highly confident that these natural forcing factors alone cannot account for the observed warming.”

    But a better way of putting it would have been, “We are highly confident that attribution studies show natural factors alone cannot account for the observed warming.”

  17. 167

    I do find the IPCC statement surprising. (“It is very likely that anthropogenic greenhouse gas increases caused most of the observed increase in global average temperatures since the mid-20th century.”).

    To me, these words mean that there is a less than 10% chance that GHG did not cause most of the warming. Based on the scientific literature, and on this and other websites, I think that one of the two stronger qualifications would have been better suited in this case.

    Regarding Gavin’s inline response, I associate the word “likely” more so with a level of probability than with a level of confidence. If the latter was meant, they should have used the word “confidence” (as Mark Bahner also noted in 164). In its current form it’s ambiguous.

    I appreciate that the attribution methods are not foolproof, and that it’s more of a statistical outcome similar to Ray’s example of the white balls. I’m aware that there’s no evidence for natural forcings having acted to warm the planet in the last 50 years. But when I read the words, I (as would many others I’m sure) come to the above conclusion as to what is meant.

    Could it be that scientific reticence had something to do with the qualification chosen?

  18. 168
    John Mashey says:

    re: #159 Ike
    What? The complete paragraph from which you quoted the first sentence is:

    “The AGIA proposal, which has received more national attention, is intended to facilitate construction of a natural gas pipeline from Alaska’s North Slope to Canada and eventually the Lower 48. Under the bill, Alaska will provide incentives, including a $500 million handout, to any company willing to build the $40 billion-plus conduit. (The $500 million will be used to help defray the costs of gaining regulatory approval, clearing environmental hurdles and so forth.) In August the Alaska Senate approved a state license for TransCanada Corporation of Calgary to pursue federal certification for construction of a 1,715-mile pipeline from the North Slope to Canada’s Alberta Gas Hub. Palin said at the GOP convention that the pipeline will “help lead America to energy independence,” but it’s clear from her advocacy of the AGIA and TransCanada’s application that her principal goal was to increase Alaska’s income from gas production. (The fact that the proposed pipeline will end in Canada, not the United States, does not seem to have attracted any notice.)”

    How exactly is that “persists in misrepresenting”?

    a) From that, I’d think you and he would agree.
    b) Moreover, from happening to hear him talk and talking to him yesterday, I’d think likewise.

    I certainly got the idea that Palin wanted the gas to go to the Alberta Hub, picked the TransCanada proposal (over the BP+ConocoPhilips “Denali” one, which at least had a “maybe” extension to the Lower 48). Masochists interested in the convolutions of AGIA might want to peruse big AK AGIA website, including the public comments [which range from “keep it all inside AK for us to use”, “get it to lower-48”, “LNG to China”, “sell to Canada and get money quick”], and the brain-hurting spaghetti of AK and US Federal laws, and a raft of conflicting agendas and intentions.

    My bottom line: I don’t think this will get built soon.

  19. 169
    Fred Staples says:

    For the sake of the argument (132), Ray Pierre, let us assume that (with sufficient notice) anyone with a Physics degree can calculate the atmospheric lapse rate from the ideal gas laws, the force of gravity, and the specific heat of the atmosphere.

    Let us also assume that we do understand the “higher is colder” AGW theory – increasing the CO2 concentration in the upper atmosphere will slow the radiation to space and increase the altitude of the effective radiation level, so reducing the effective radiation temperature. From the Stefan-Bolzmann radiation Law, the lower temperature will reduce the radiant energy output below the incoming solar energy, and the entire system will heat up.

    If we attribute the marginal lower atmosphere heating to this effect, we must expect to find at least comparable heating in the mid and lower troposphere, which we do not.

    You explain the absence of mid troposphere heating by the corrupting effect on the measurements of low stratospheric temperatures “seen” by the satellites. Would we not expect that effect to have been reasonably constant from 1978 to date? It was not.

    From a visual inspection of the UAH charts, there is no significant atmospheric warming at either level from 1978 to 1996. The famous 1998 peak was reversed on both charts before the end of 1999. Both charts experienced the relatively sharp increase from 1999 to 2001, but the upper troposphere increase was much lower, and more rapidly reversed – hence the difference between the long term trends.

    Currently we are back to 1978 levels on both charts.

    I accept that we are stretching statistical significance to breaking point with this analysis, because there is so much variability and so little atmospheric warming (could you measure the temperature of your house, month in month out, to this degree of accuracy), but the AGW theory is simply not consistent with this data.

    [Response: Repeating the same incorrect logic and ignoring the uncertainties in both your understanding and the data does not move anything forward. – gavin]

  20. 170
    Ray Ladbury says:

    OK, so now we’ve shifted our misunderstanding of concepts from confidence to likelihood. It’s still a misunderstanding, but this is progress. Likelihood is also not probability. Rather it is a functional over a probability model and the data available. Maximum likelihood is probably the most well known use of this functional. Here you determine the most “likely” parameters of the model as those which maximize the probability of realizing the observed data given our model (e.g. Normal, Lognormal…). However, Likelihood methods are much more powerful than that. Because errors on the parametric estimates from likelihood models tend to be normally distributed, the ratio of logarithms of likelihoods for different parametric values tend to follow Chi^2 distributions with number of degrees of freedom equal to the number of parameters in the model. Thus we can determine confidence intervals as well as best estimates for the parameters. Likelihood also plays a critical role in Bayesian as well as Frequentist statistics.

    OK, so much for defining likelihood. The relevant issue here is that likelihood depends on both the data and the model. As such, the value of the likelihood depends on how well our model fits AND on how representative our data are. In this sense, likelihood tends to depend on sampling errors–like confidence–and on systematic errors in our model. However for a given model, the systematic errors will tend to be independent of the data, so likelihood is much closer to a “confidence” measure than a probability measure (note that the integral of likelihood over all parametric space is not normalized, so it can’t be a measure of probability).

  21. 171
    Hank Roberts says:

    Yep. I don’t see the logic in Mr. Bahner’s

    “If the defendant has not committed the crime, then there is evidence that *someone* else has committed the crime.”

    Does not compute.

  22. 172

    Returning to the earlier comments regarding the “reality-based” community, I think that the current economic crisis illustrates what eventually happens. The Bush model of “creating one’s own reality” (echoed in Palin’s “idea” that you can fight GW without knowing what causes it) leads to, shall we say, a divergence between reality and proclamation. You can see that clearly with Bush’s current attempts to rally the economic troops: the papers dutifully report what he says, and few even bother to roll their eyes. I almost feel sorry for the man, but the wounds to his credibility are self-inflicted. Similarly for talk radio and the right-wing blogosphere, in varying degrees–they’ve told us black is white a few too many times.

    Most relevant for our purposes here is that the same will happen with the denialist crowd. They are now at the point, at least in some cases, of actually denying that surface measurements have any validity (they always want to use UAH troposphere figures, of course.) (To which I replied recently, “You do live at the surface, don’t you?”) This degree of nuttiness can’t be sustained long.

    (reCaptcha says “trial cases,” but I’m going to request another as there is a bit following which may be signal or may be noise–and lo, I get “securities no.” Definitely some kind of oracle–!)

  23. 173
    Ike Solem says:

    John, the quote is “In August the Alaska Senate approved a state license for TransCanada Corporation of Calgary to pursue federal certification for construction of a 1,715-mile pipeline from the North Slope to Canada’s Alberta Gas Hub.”

    Well, Alberta’s gas hub feeds into tar sand production –

    Production of Canadian Tar Sands requires 1,200 SCF (standard cubic feet) of natural gas per barrel. The 4.5 BCFD of Alaskan gas can be used to support 3.75 MMBPD of tar sand oil production.. this analysis show just how important Alaskan Gas is to North American oil production.

    Second, we have the retooling of U.S. refineries to accept tar sands oil:


    The refinery upgrades… also will allow the refinery to thermally convert and upgrade heavy Canadian crude oil into higher quality products such as gasoline, diesel and petroleum coke.

    2) “US Refinery Investments Align With Oil Sands Supplies to 2015, 18 September 2008”

    So, any rational analysis of the real position of the global oil industry that ignores this fundamental issue is something of a misrepresentation.

    The underlying issue here is the lack of a free market in energy, largely due to control of the political situation by fossil fuel interests. This is why you get $18 billion in federal credit subsidies for a pipeline that links BP, Exxon and Conoco gas fields in Alaska (which will supply Transcanada) to BP, Exxon and Conoco tar sands projects in Alberta. There is also Sarah Palin’s direct $500 million gift to Transcanada. Transcanada’s business model involves moving gas to tar sand fields and then moving heavy crude to U.S. refineries.

    Before it was Transcanada, it was Mid-American Energy, but they decided to withdraw from the project due to political concerns over the corrupt Alaskan contracting process. Then, Sarah Palin was elected, and she managed to get Transcanada the contract. The Washington Post (at MSNBC) ran a story on this today, actually, sticking to the misrepresentation:

    Much of [Palin’s] time was devoted to discussions of a proposed 1,700-mile pipeline that would deliver natural gas from the North Slope of Alaska to the lower 48 states… Her first contact with Washington came on Jan. 17, 2007, when the vice president called her to discuss the project, the calendar shows.

    In early January 2007, Palin met with Marathon Oil executives, and the next month, while attending a meeting of the National Governors Association in Washington, she met privately with Exxon Mobil executives..

    Notice the story above about Marathon’s refinery upgrades to handle more Canadian tar sand oil? The WP story, sticking to course, makes zero mention of the real agenda: converting tar sands to heavy crude using Alaskan natural gas.

    Anyone who really thinks that global warming is a problem needs to understand that the tar sands cannot be developed. There needs to be a public discussion, and there should also be a national ban on importing tar sand oil. We can’t go around exporting our fossil carbon emissions to other countries while claiming we are “cleaning up our act.”

    We do that in California enough – our recent global warming law ignores the fact that a whole lot of coal-fired electricity is imported into California. Not only that, Bay Area and Los Angeles refineries have also been getting ready for more tar sand oil:

    SEVERAL OIL GIANTS have plunked down a collective billion-dollar- bet on East Bay refinery upgrades… Some refineries will gain more capabilities to process “heavy” crude oils and crude laced with sulfur.

    Finally, the world’s biggest investors are backing this process. Here is an example: Warren Buffet controls Mid-American Energy (the original pipeline contractor that Transcanada replaced). Here is he is congratulating Palin on her victory: . Buffet also has large stakes in Conoco and others.

    Then, we see Obama saying that he is considering making Buffet head of Treasury, and he has also said that the Alaskan gas pipeline “is a priority.” Is he even aware of what the gas will be used for? Now… look. This is just ridiculous. Why can’t the press cover this basic issue?

  24. 174
    JCH says:

    At the rate the economy is slowing down, the country will be in compliance with Kyoto by the middle of last week.

    Ike, Buffett is an advisor to Obama, and a supporter. Buffett would never take the treasury job, nor is it going to be offered to him. Buffett likes Hank Paulson, and will probably encourage Obama to retain him in the job.

  25. 175
    Mark says:

    Mark #164

    They don’t say that anything else caused it either. Since you take merely what they say at face value, then all they are saying is that there is a less than (did it againn…) 10% chance that there is no warming.

    But if you have a 1% chance of being cured of cancer with a new drug costing £10,000 would YOU Demand that you get it? Or would you go “nah, it’s not gonna happen”? In the UK, look at the lawsuit to make herceptin available to the NHS patient even for cancers it has not been proven to work against.

    A lawsuit, I must add, financed by the company making the drug.

  26. 176
    paulina says:

    Re Palin’s pipeline to the Alberta Tar Sands, these may also be of interest:
    ALERT: Palin’s pipeline is a climate crisis acceleration machine
    and this follow-up by the same author.

  27. 177
    John Mashey says:

    re: #173 Ike
    My problem is that “persists in misrepresenting” usually means “lie”.

    Klare is a political scientist one of whose interests is geopolitics of energy. He wrote the 1500-word article I referenced (about 2 pages) for TheNation trying to explain the nature of petrostates to the general audience, explaining why Alaska was one, and explaining the huge gaps between Palin rhetoric and reality. He is in no way a Palin fan.

    I don’t argue at all with the bulk of what you say and neither would Prof. Klare. How do I *know* that? In fact, he speaks and writes widely and *passionately* about these issues.

    a) I happened to hear him talk at Stanford and talked to him a few days ago.

    b) His new book, “Rising Powers, Shrinking Planet” is sitting on my desk, and while its focus is on geopolitical energy issues and potential for major conflicts, i.e., so Alaska gets relatively few mentions compared to Africa, Caspian Sea, Russia, etc, it still says:

    “-Canadian tar sands, Venezuelan extra-heavy crude, Rocky Mountain oil shale-but the financial and environmental costs of using these materials are huge, and they are unlikely to rescue us, even briefly, from a dramatic and painful contraction in primary energy supplies.”

    There are many references to global warming and CO2 issues. He gives an extra set of reasons for wanting to “permit a gradual reduction in global military expenditures and thereby free up substantial funds for systematic efforts to tackle the threat of global warming.”

    What worldview is likely to be found in a book whose blurbs include Paul Ehrlich, Bill McKibben, Matt Simmons, Amory Lovins?

    Is your complaint that *he* didn’t spend his 1500 words writing about the topics you’d rather he covered? I’m hard-pressed to understand how the Washington Post or Warren Buffet have any relevance to deciding whether or not *Klare* “persists in misrepresenting.”

    Anyway, the book is well worth reading, and if he happens to give a lecture nearby, he’s well worth hearing, especially because his depth of political science + resource economics offers additional insights.

  28. 178
    Rod B says:

    Hank, it’s probably not a good comparison for the global warming likelihood discourse, but there is nothing confusing about the crime analysis. One must assume that a crime has been committed. Then if the evidence says X did not do it, that same evidence must say/conclude with certainty that someone else did. There are no other possibilities.

    [Response: But actually there might not have been a crime committed at all. It might have been a suicide, or an accident etc. Judgments of confidence need to integrate over all the uncertainty. – gavin]

  29. 179
    Rod B says:

    Ike, I can’t grasp this particular complaint with Palin. Are you saying that as Governor of Alaska she is looking out for Alaska’s economic interest per se, and somehow that is malicious? Or that despite being Governor of Alaska she is none-the-less not looking out for AGW mitigation and its proponents? Or something else?

  30. 180
    Rod B says:

    Ray, a maybe goofy thought question that might help my thinking (which is statistics challenged :-P ) Is it possible for an assessment to have a 90% probability also with 90% confidence while the converse of the same assessment is mathematically a 10% probability but with 0% confidence (or 0+%)?

    I’m also assuming “confidence” here is a strict mathematical/statistical construct as opposed to a collective human judgement. True?

  31. 181
    Hank Roberts says:

    > 112, 154, 157, 178
    Gavin understands. This is important to understand, both for the analogy (and your own civil liberties) and for climate attribution.

    > judgments of confidence need to integrate over all the uncertainty.

    Gavin, is it fair to rephrase that as:

    You must _not_assume_ any part of what you claim you can prove.

    [Response: Yes – I think that gets to the heart of it. – gavin]

  32. 182

    #152, #157, and others RE 90-95% confidence that it is AGW.

    Let’s just think about that for a little while. 90-95% confidence that serious harms are to befall us from our own hands. I’m thinking, even 50% confidence of such, even 20% confidence, even 10%(since the problems could be truly enormous) is enough to start turning off lights not in use, and the myriad of other measures that can reduce our GHGs here in America by 75% without sacrifice.

    Which means everyone in the world should have started reducing back in 1990 (and the poor climbing out of dire poverty should do so in efficient ways), and the world should be at about a 20-30% lower level of GHG emissions than the 1990 level.

    Is this world crazy or something? Are the candidates (all of them, actually) crazy or something?

  33. 183
    Ray Ladbury says:

    Rod, consider what we mean by “confidence”. As Mark said, it has to do with how often we’re likely to be fooled by sampling errors or data conspiring to fool us. So a 0% CL really means it’s so noisy that you can’t say anything about it–your signal looks just like your noise. Then there’s the matter of how you’d calculate a 0% confidence level. Often, you use Monte Carlo methods, so near 0% and 100% you get big errors. And from a practical point of view, CLs less than 50% just aren’t interesting.

  34. 184

    Re #182


    We in the rich world should have cut back on our borrowing in the 1990s, but we didn’t because it made us feel rich. Peak oil was inevitable, and now we can’t repay the debts we have run up. Yes, this world is crazy or something! But the canidates aren’t crazy. They know that no one will vote for someone who tells them the good times are over :-(

    Cheers, Alastair.

  35. 185
    paulina says:

    As per RC:

    Whatever else you can say about the candidates, it has been encouraging that both John McCain and Barack Obama favor mandatory action to reduce US carbon emissions.

    We speak of various levels of denial–denial of GW; agreement with GW but denial of AGW; agreement with AGW but denial that AGW is a problem we need to do anything about; agreement that AGW is a problem we need to do something about but denial that we can do anything about it; and then, typically, there is a sense of “and now, for something completely different,” and we, supposedly, enter the stage of “agreeing that we can and must act.”

    This conceptual framework has the unfortunate consequence of tacitly suggesting, among other things, that:

    –no important kind of denial exists at “the action stage,” and

    –there’s some important difference between being in one of these first four stages of denial and being in “the action stage” where we want to “put a price on carbon” or “put a mandatory cap on emissions,” or where we “favor mandatory action to reduce US carbon emissions”

    But important denial does of course exist at the “action stage,” and there isn’t any important practical difference between being an AGW-denier and wanting to put a price on carbon. How about a penny per ton? How about a trading system that benefits polluters (massively) and does not manage to prevent construction of new coal-fired power plants?

    How many hundreds of times have you heard a policy, science, NGO, or scholarly authority on climate destabilization mention how great or at least encouraging it is that all or both candidates “believe in global warming,” “will take action,” “will put in place a cap and trade system,” or “will put a mandatory limit on emissions,” etc?

    But “taking action,” “putting in place a cap and trade system”, and “putting a mandatory limit on emissions,” are all entirely open-ended and therefore, in context, entirely empty.

    “Mandatory action to reduce emissions”? Empty.

    Yes, supporting mandatory action to reduce emissions means acknowledging that government has a role to play in this space. But this would-be moving of the Overton window has an auto-retract leash. This is interesting and worth considering. It’s also a big problem.

    Is it not virtually impossible to do justice to the scale of the problem and the scale of action required (and by so failing, we move the window back), by calling something empty “encouraging”?

    This part of the problem of communicating about climate destabilization needs more exploring.

    By “empty” I do not here mean candidates-contradict-themselves-the-whole-time-empty, or let’s-look-at-voting-records-instead-of-talking-records-empty.

    I’m also not talking about McCain’s attempt to disavow embracing the “mandatory.” Heck, I wouldn’t voluntarily qualify my proposal as involving anything “mandatory,” either (no pun intended). But if my proposal involved a legally imposed cap (on some sector’s emissions, for instance), I wouldn’t deny that I was in favor of mandatory caps, with or without an exculpatory sotto voce qualifier. And if RC really finds anything to be encouraged about in McCain’s “favoring mandatory action,” presumably you should experience at least a little bit of discouragement at his convoluted disavowal of the word ‘mandatory.’

    Finally, I’m also not talking about the candidates’ specific proposals; only about the generic way of talking about these and about the candidates.

    I’m making a separate point specifically to do with the conceptual framework in which we consider various kinds of denier and delayer attitudes pertaining to putative government actions taken to address climate destabilization.

    I understand if people (RCers) don’t want to give specific policy recommendations without having all the expertise they believe is required, but I think there’s plenty of space between, on the one hand, giving specific policy proposals, and, on the other, actually (involuntarily in your case, I assume) contributing to confusion by perpetuating the idea that simply “putting a price on carbon” or “putting in place mandatory emission limits” counts for anything.

    We need a conceptual framework that covers more of the significant denial as well as the understanding, a framework that doesn’t promote 0≠0 math.

    This comment is much too long, but hopefully the main gist will be clear. Thanks.

  36. 186
    Chris G says:

    Probably preaching to the choir here, but…

    From a philosophical perspective, I believe it was DesCarte who gave us the maxim that in the absence of sure knowledge, we should take the best knowledge that we have and act upon it as though we knew it with certainty.

    From a pragmatic perspective, if a large body of electricians told you that, unless something was done to correct it, the wiring in your house had a 95% chance of causing a fire, would it not be worth sacrificing the purchase of a new TV to fix it?

    I just wish the general population had a better understanding of risk management.

  37. 187
    Steve Bloom says:

    Re #167 (Bart): ‘I do find the IPCC statement surprising. (“It is very likely that anthropogenic greenhouse gas increases caused most of the observed increase in global average temperatures since the mid-20th century.”).

    ‘To me, these words mean that there is a less than 10% chance that GHG did not cause most of the warming. Based on the scientific literature, and on this and other websites, I think that one of the two stronger qualifications would have been better suited in this case.’

    Looking closely at the IPCC language, it seems to me that it’s a nod to uncertainty about the magnitude of temperature increases due to other anthropogenic effects, i.e. land use change and aerosols. The level of certainty seems consistent with the literature.

  38. 188
    Ray Ladbury says:

    Chris, Given what the current financial crisis doing to people’s 401Ks, they probably wish they understood risk assessment better as well. At this point, I just wish we had a way to limit the risk we had to mitigate.

  39. 189
    Mark Bahner says:

    “Yep. I don’t see the logic in Mr. Bahner’s”

    ‘If the defendant has not committed the crime, then there is evidence that *someone* else has committed the crime.’

    “Does not compute.”

    OK, there’s a dead body in a morgue. It was found with a knife in the back. The defendant didn’t commit the crime. How is it not completely logical that *someone* else must have committed the crime? The evidence is the body with a knife in its back!

    [Response: Maybe he fell backwards on to it by accident; maybe he was dead already and someone put a knife in his back for a laugh; maybe it’s a set up and the body isn’t who everyone thinks it is; maybe the mortician has been paid to say that there is body there when there isn’t… etc. etc. The point being that in the real world (as opposed to probability textbooks), there are huge numbers of other uncertainties which need to be judged (in some kind of Bayesian fashion). That means always that confidence levels are always less than the formal calculation of a likelihood. – gavin]

  40. 190
    Mark says:

    re: 189

    Who said there was a body with a knife in the back? You’re making things up to prove your point when you’re ignoring what is right in front of you:

    If someone is possibly innocent, it doesn’t mean someone else is guilty.

  41. 191
    Mark says:

    Rod’ #180.

    Yes. Both are possible. Probability isn’t entirely orthogonal to confidence. An example:

    Naively, heads or tails: probability of head:50%.

    Confidence level: 99.9% or more. ‘cos it COULD land on the edge, fall, be miscounted, roll away…

    Rolling a dice: probability of a 4: 16%.

    Confidence in the dice being fair: 0%.

    Rolling a dice 100 times: probability of a mean within 3.5 +/- 0.7: 91%

    Confidence that the dice is fair when the average is 2.47: less than 8% (i think, someone do the maths, that’s a rough figure).

    See how it goes?

    Hope that helps.

  42. 192
    Ray Ladbury says:

    Mark Bahner, Logically, if, just hypothetically, you found someone with their throat slit, a bloody glove…, and the defendant did not commit the crime, someone else must have. Confidence has to do, however, less with logic and more with what EVIDENCE allows you to say. OK, so, we know we have a corpse with a knife in its back. Have we established that the cause of death was loss of blood consistent with a stab wound? Have we established that the wound in question was sufficient to cause death? Have we established that there is no other cause that could explain the death. There is a reason why legal burden is only “beyond reasonable doubt”. Reasonable doubt is probably somewhere below 85% CL in most cases.

  43. 193
    Hank Roberts says:

    Mark Bahner — your black-and-white thinking is the restricted imagination that creates lynch mobs. Okay, stay on the civil liberties point for the moment (sigh). I had thought basic civics was taught.

    Gavin gives you a handful of possibilities. Miss Marple could also.

    “Someone must be responsible” –> “someone must die for this” –> the police must find someone –> someone must be found guilty.

    Requiring someone be found to punish puts state power, and fear, above individual freedom and liberty — “pour encourager les autres.”*

    Do you understand?

    Once you understand this — it was a revolutionary idea a few centuries ago, now it’s playground-level common sense — you can move on from civics to science.
    * See

  44. 194
    David B. Benson says:

    Could we stick to climatology and go play Clue somewhere else?

    [C aptcha suggests “Dover state”.]

  45. 195
    RichardC says:

    186 Chris, it depends on how it is phrased. “There is a 90% or greater chance that the faulty wiring will burn your house down sometime over the next 10 million years, and it may cost so much to fix that you will lose your house to foreclosure.”

    There, even though in reality fixing the wiring will probably save money since the losses due to arcing have been costing you greatly and the wiring will probably burn your house down in a few years, I haven’t lied and the logical path taken by those who don’t ask further questions is to not fix the wiring.

    And the 90+% number’s subject is *not* about whether AGW is true, but whether the amount is over 50% of the estimate of recent warming. Too bad they didn’t split up the equation into “GHGs contributed over 50% of warming that actually occurred” and “actual warming was over 50% of estimated warming.” Folks are focused on the more high confidence side – whether it is GHGs, while glossing over the less high confidence side of how much actual warming there was, essentially transferring doubt from one to the other.

  46. 196
    David B. Benson says:

    RichardC (195) — I am not following what you are attempting to state. The actual warming is measurable and IPCC AR4 states that the global temperature increased 0.6 K in 100 (and also 140) years, ending in 2000 CE, if I remember correctly.

    The difficulty is in attributing the proportion of the increase to anthropogenic factors.

  47. 197
    Mark Bahner says:

    “Who said there was a body with a knife in the back?”

    As long as Gavin Schmidt was making up analogies, I figured I should be able to make up my own analogies.

    But let’s return to ***the actual situation***.

    The IPCC wrote, “Attribution studies show that it is very likely that these natural forcing factors alone cannot account for the observed warming.”

    They also wrote (Technical Summary, page 23) that “very likely” means a “likelihood of occurrence/outcome” of 90-95% probability.

    Roger Pielke Jr. has observed that given that statement, it can be concluded that the IPCC is saying that there is a 5-10% probability that natural factors alone CAN account for the observed warming.

    I don’t see how this observation is in any way controversial. If there is a 90-95% probability that natural factors alone CANNOT account for the warming, then there must be a 5-10% chance that natural factors alone CAN account for the warming. The two situations have to add to 100% probability, and there is no possibility beyond those two situations.

    [Response: sigh… the whole point of the analogy was so that that you’d be able to see the point more clearly. i.e. there are always more possibilities: attribution methodologies might not be valid; our knowledge of the natural forcings might be incomplete; the models might be wrong; the data against which we are validating might be wrong; the whole thing could be a gigantic conspiracy…. (just kidding about the last one). If instead of looking just at the summary statement and look at the actual studies upon which it is based, you will see that the effect of natural forcings doesn’t come anywhere near explaining what we need to have explained. Indeed it is likely that the natural forcings would have actually induced a cooling. Therefore the 5-10% uncertainty is associated predominantly with methodological uncertainty, not the uncertainty associated with different forcings or internal variability. As I asked all the way above, if you think that there is any decent study that shows that natural forces can explain the recent trends at even 10% likelihood, perhaps you’d care to point it out? – gavin]

  48. 198
    Mark Bahner says:

    Sigh. Well, don’t worry, Gavin, at least you’re giving 110 percent.

    So Roger Pielke and I (and just about every English major in the world, I’d think) say that if the IPCC says that there is 90-95% percent chance that natural forcings alone can’t accout for the warming, the IPCC is also saying that there is a 5-10% chance that natural forcings alone CAN account for the warming.

    What do YOU think the IPCC is saying? Note that I’m NOT asking you whether you think the IPCC is right or wrong, I’m just asking you what you think the IPCC is saying.

  49. 199
    RichardC says:

    196 David says, “The actual warming is measurable and IPCC AR4 states that the global temperature increased 0.6 K [actually 0.74]”

    No, they say the observed increase. That’s different from the real increase in that it includes errors in both measurements and the calculations which massage the collection of measurements into an estimate. What was observed might not be real.

    So when they talk of a 90-100% chance that the majority of the observed warming was AGW, that means that in some alternative universes the measurement was off by 0.37C, in some a yet-unknown natural forcing added 0.1C, the data was off by 0.17C, and the summation of data introduced a 0.1C error. Confidence level says that all those alternative universes that fail the test of 0.38C AGW warming add up to 0-10% of the total possible universes, even as measured by the most devout pessimist.

    Besides, reverse-engineering from a sound-bite is dishonest, especially since it was done with an axe. He ignored the variable with the most deviation (how much it REALLY warmed) and attributed all uncertainty to the variable with the least deviation (what caused it). Also, your source likely knew about the discussions Mark talked about above (that the politicians essentially forced the scientists to degrade the confidence). Then there’s the very next paragraph, which speaks directly about what caused it, “It is likely that increases in greenhouse gas concentrations alone would have caused more warming than observed because volcanic and anthropogenic aerosols have offset some warming that would otherwise have taken place.” Put the three together, and the conclusion is inescapable.

  50. 200
    Mark says:

    Mark #197.

    So you made it up. Why? How did that changed analogy remove the fact that if all you had was “someone dead” that if the person in the dock didn’t do it, it doesn’t mean you have evidence that someone else did?

    It didn’t.

    So it was pointless.

    Rather like your continued harping.