Bishop Hill In which computer models collide with the real world
Mar 17, 2015 Climate: WG3 Energy: coal Greens Yesterday's post on the trade-off between the need to expand use of fossil fuels in Africa and the wish to restrict carbon dioxide emissions seems to have stirred up a bit of a rumpus. Most commenters from the other side of the debate apparently deemed my question over the wisdom of access restrictions as entirely illegitimate, although the reasons why are somewhat unclear to me.
Firstly, as Roger Pielke Jr pointed out, in the real world there are trade-offs that have to be made.
@ClimateOfGavin Not really-see: 1 http://t.co/oOu02cRHeD 2 http://t.co/WkxXfuwpJK Real world tradeoffs are actually being made @mattwridley
— Roger Pielke Jr. (@RogerPielkeJr) March 16, 2015
The first of the papers contains this:
...under US Senate Bill S.329 (2013) the Overseas Private Investment Corporation – a federal agency responsible for backstopping U.S. companies which invest in developing countries – is essentially prohibited from investing in energy projects that involve fossil fuels, a policy that may have profound consequences in places like sub-Saharan Africa that are seeking to develop oil and gas resources to help alleviate widespread energy poverty.
I have heard no arguments that there isn't a trade off, so this is presumably not the reason why my questions are being declared off limits.
Ken Rice, of AndThenTheresPhysics says that my framing is malign although his allegation displays his normal attention to facts. Readers may recall that in the early days of his visits here I wrote a long piece explaining why climate science could only rely on physical models because of the difficulty in choosing a statistical model. To this, Rice responded, in effect, that I was an idiot and that climate scientists should be using physical models. Something similar seems to have happened here. I carefully framed my case as how to weigh deaths in the present against deaths, albeit hypothetical ones, in the future. This is the essence of the trade-off that has to be made and which is, according to Pielke Jr's paper, being made in favour of those not yet born and at the expense of those alive today. Rice says my framing is that:
...those who might be concerned about the risks associated with climate change [are not] concerned about the fate of poor people in the developed world.
But I specifically said this was not the case. My words were:
The accusation is not...that greens are callous about deaths in Africa.
As I explained, the choice is between real deaths now or hypothetical deaths later and all points in between. It is the choice that politicians are making right now. And, advised by climatologists and economists of the Stern/Fankhauser genre of the horrors to come and the costs to be borne, they have decided to do what they can to keep fossil fuels out of the hands of Africans. Who knows, it might even be the correct decision.
This is where the computer models of climatologists and the discounting choices of economists bang right up against the real world. The projections and predictions are no longer academic playthings to be bickered over at conferences and seminars, they are the tools with which our leaders make life-or-death decisions.
I hope scientists have the right caveats in place.
Bishop Hill
In the comments to the original article, Ken Rice explained my argument as follows:
Let me try and lay this out again, as this is what I thought the point of this post was. Framing this whole discussion as "my solution is the only way to help the poor in the developing world...anyone who disagrees with this is arguing for the death of people in the developing world" is appalling.
This, he says, was his way of "paraphrasing" my argument. You know, the one that said:
The accusation is not...that greens are callous about deaths in Africa.
I don't think he can help himself.
Reader Comments (104)
But sure it is the case that greens ARE callous about deaths in Africa - how else explain their drive to keep carbon energy away from them ?
"So are you going to tell us that all the claims by skeptics here and elsewhere that there has been a pause in warming are also just an ironic joke based on an argument that they all understand is nonsense?"
That depends on what definition of 'pause' they are using. Some are using purely descriptive definitions that don't assume models. For example, if you want to know whether the temperature has "gone up" over a particular time period, just subtract the final temperature from the starting temperature. Some are using the mainstream's definitions and models, because their aim is to falsify them. (Falsification being the method of science, according to Popper.) Some are doing it ironically - deliberately using the same methods used by the mainstream during the 90s to claim unstoppable global warming to show that by their own chosen methods it's stopped. I've done that myself, because I think it's amusing to watch the wriggling - and yes, it was intended ironically.
Quite a lot, I'm sure, are equally unaware of the problems with it, and are making the same error that the mainstream does. It's likely that many of them think that OLS trends and AR(1) tests are meaningful. Given the example the mainstream has set, one can hardly blame them, but it's not correct.
For that matter, I'm sure there are people here on my side who are falling for the same one-sided 'think-of-the-children' argument pointing the other way. Matt Ridley and the Bishop are not among them, but some here probably are. However, while I might think they're incorrect, I don't think it's beyond the pale of civilised conversation. If somebody doesn't understand, it's our job to explain. It isn't always obvious to everyone, and it's much better to discuss such 'unfashionable' opinions than try to suppress them. It's the only way to fix people's errors and misunderstandings.
As I've said numerous times here, the 'pause' is a serious problem for the climate models, most of which predict that pauses of such length ought to be very rare, but is not a problem for the global warming hypothesis itself. Because we don't know what the distribution of background variation is, we don't know that the variability isn't much bigger than climate scientists think, and that this isn't just a very big but still temporary excursion from a longer-term systematic rise. We don't know that it is, either.
It does strongly indicate that the magnitude of the natural variability must be at least comparable with the rate of rise, to be capable of cancelling it. And that raises the possibility that the rise during the 1990s might have been enhanced by natural variability, but we don't know. We certainly can't say the length of the pause is 'significant' with regard to the real weather. It is only 'significant' with respect to falsifying the models.
"No, you've not got it"
How so?
"But sure it is the case that greens ARE callous about deaths in Africa - how else explain their drive to keep carbon energy away from them ?"
By considering the possibility that they're trading one set of harms against another. Sometimes you're faced with a choice between options *all* of which will kill people. Which do you choose? Usually, the one that kills fewest.
If someone *genuinely* believes that the climate apocalypse is imminent, and I expect some people really do, then it's arguably the 'right' choice. They might well be wrong about that apocalypse, but given what they believe their position is at least understandable. And none of us can be right about *everything* we believe in - judge not, lest ye be judged yourself.
Heh, very nice exposition. Another example is all the ways that 'hide the decline' is used, ironically, incorrectly, pejoratively, polemically.
Well, now I hope he gets it, by my judgement, anyway. Or his.
==============
Nullius,
Thanks for your interesting response.
'Some are using the mainstream's definitions and models, because their aim is to falsify them. (Falsification being the method of science, according to Popper.) Some are doing it ironically - deliberately using the same methods used by the mainstream during the 90s to claim unstoppable global warming to show that by their own chosen methods it's stopped.'
'Given the example the mainstream has set, one can hardly blame them, but it's not correct.'
I'm interested in this. Where did the idea of using OLS come from? Was it really the mainstream? In particular, I have the impression that the idea of a 'pause' based on the 'significance' of the OLS trean was started by Lindzen (of whom I would have expected better), as a throw away remark. Probably it was encouraged post-1998 El Nino by Patrick Michaels and Chip Knappenberger, again almost more as a rhetorical talking point. I think Lindzen, Michaels and Knappenberger probably had / have a reasonable grasp of it's limits.
But then it was popularised by Robert Carter in the Telegraph ('There is a problem with global warming ... it stopped in 1998') and then, via Lubos Motl, landed up as a question put to Phil Jones by Roger Harribin. Apart from anything a major confusion starts to become common in the media from that point on between 'no warming' and 'no statistically significant warming'.
When I look, for example at the Bishop Hill discussion of that interview:
http://www.bishop-hill.net/blog/2010/2/13/wow.html"
I don't get a strong sense that commentators have a subtle understanding of OLS or significance, or that they are making point ironically.
On the other hand, while I don't think the IPCC is above criticism by any means, but I don't see them putting a big emphasis on OLS and significance testing. In chapter 12 of ths AR3 WGI on Detection and Attribution it is only a peripheral and supplemental discussion, appropriately qualified (section 12.4.1).
http://www.grida.no/climate/ipcc_tar/wg1/pdf/TAR-12.pdf"
(Chapter 2 on observations does show tables (not graphs) with linear trends, but these look like maximum likelihood trends, not OLS. Although the tables do show significance levels, I think linear trends are less problematic as summaries of data.)
Again, in the Attribution chapter of AR4 (chapter 9) linear trends get only a cursory initial mention. The entire chapter is an attempt (not necessarily successful) at a more sophisticated approach:
http://www.ipcc.ch/publications_and_data/ar4/wg1/en/ch9.html
(Also similarly, in chapter 3 there are tables of trends based on AR(1) models, but again as data summaries I don't think this is so objectionable.)
Moving on to AR5, I don't see the emphasis on OLS there either. The Attribution chapter is chapter 10 of WG1. Again, the discussion of time series methods (10.2.2) seems at least miles ahead of most (non-ironic) skeptic discussions of the pause:
http://www.ipcc.ch/pdf/assessment-report/ar5/wg1/WG1AR5_Chapter10_FINAL.pdf
The presentation in the chapter on observations is a bit more subtle in AR5 than previously, probably benefitting from the public discussion of the pause. I would be interested to know if people here think that Box 2.2 is helpful, or what it's strengths and weaknesses are:
http://www.ipcc.ch/pdf/assessment-report/ar5/wg1/WG1AR5_Chapter02_FINAL.pdf
I don't know if the Met Office put a strong emphasis on OLS and significance of trends before Lord Donoughue asked his questions, but I'd be interested to hear what people know.
It would be great to get more examples of how OLS was really promoted by the mainstream, the ones Nullius has in mind or others.
'If somebody doesn't understand, it's our job to explain. It isn't always obvious to everyone, and it's much better to discuss such 'unfashionable' opinions than try to suppress them. It's the only way to fix people's errors and misunderstandings.'
Yes, I would encourage that approach, much more than 'ironically' going along with wrong arguments.
There are other interesting points that you raise in your comment, but my reply is already overlong, so I'll give it a break for now.