Thursday
Dec182014
by 
Bishop Hill
Bishop Hill Sans ifs, sans buts, sans everything
Dec 18, 2014 Climate: Models Climate: other Ethics Judith Curry quotes this sentence from Peter Lee's GWPF essay on climate change and ethics
Omitting the ‘doubts, the caveats, the ifs, ands and buts’ is not a morally neutral act; it is a subtle deception that calls scientific practice into disrepute.
I couldn't help but recall the reaction from climate scientists when I said it was "grossly misleading" of Keith Shine to omit any caveats when explaining the efficacy of GCMs to parliamentarians.
I stand by what I said.
Reader Comments (132)
Paul,
Really, you think that's me being funny? Okay, Ha Ha. If only I'd known you've just been joking all this time. Impressive to have such a deep level of humour that noone actually realises. Maybe you should point it out more often, because I think many people are very confused by your behaviour.
I'm assuming the first bit of that is ironic, but I thought we were on the verge of an interesting discussion, until comedian Matthews popped up with his humorous jibe. Well, okay, I wasn't all that optimistic that it would end up being an interesting debate, but I had hopes.
You misunderstand, as usual. I was referring to your joke, where you responded to my suggestion that you should calm down and not have a temper tantrum with "WTF!"
ATTP:
There is only one value for climate sensitivity.
Really? Any sensible physical model I can think of would produce a range of values corresponding to variability in many of the other parameters, and their interactions.
Paul,
Oh, I see. Since, I've been accused of a lack of integrity, I should probably acknowledge that that was not a joke. That probably was more correctly described as "losing my cool" - well, losing whatever cool I might have had.
It doesn't,
I meant in reality. We live on a single planet that will respond in only one way to a change in radiative forcing, given a particular initial climate state. Yes, there will be some variability on decadal timescales, but the range of climate sensitivity that models present is an indication of our uncertainty in what that value probably is, not an indication that the planet will randomly select from that range.
aTTP
I'm beginning to detect the trollism that others seem to have previously experienced. Just man up and admit you were wrong!
Of course "what would really help is for those who aren't in denial" directly contradicts "what would help would be if people like yourself didn't paint others in this simplistic way". Well I say skeptics are the realists and scads of climate scientists are in denial and that wouldn't be at all important except that mother nature agrees with us skeptics thus far and history tells us that will likely continue.
And of course "Climate sensitivity will be whatever physical reality dictates" directly contradicts "Focusing only on the best estimate and ignoring the range is, in my view, not presenting a complete picture". If there is a range we are supposed to consider then blindingly obviously there is no one number that represents 'physical reality' and what you choose then depends on how pessimistic you are - or you could be logical and just choose the 'best estimate'. Clearly in order to tease out manmade warming we first have to know the extent of natural variation and nobody yet from the mainstream seems to be able to predict this. By contrast a few skeptics have been quite accurate in predicting of the plateau before it happened - me included btw.
I fully agree though that people should not make 'hyperbolic comment' but I'd go further - they should stop lying about 'mountains of evidence', 'extreme weather events getting worse', 'hot areas getting hotter and dry areas getting drier' etc, etc. So far there is nothing unusual happening and no end in sight for the temperature plateau and so the best policy is business as usual and we should throw out all catastrophists from anything to do with energy policy.
James,
No, I don't think I'm wrong. You're free to disagree. I chose my words carefully.
No, I disagree. In fact, this is the fundamental point. We don't really know what climate sensitivity is. We do, however, have estimates of the range and hence the likelihood of certain values. Therefore, given a particular emission pathway, we can estimate the likelihood of different amounts of warming and, hence, what the likelihood of reaching levels that might be damaging. Sensible risk analysis then suggests we should balance that with the risk/cost associated with minimising the chance of damaging warming. So, for example, if we want to have a less than 1% chance of reaching levels of warming that might be damaging, that then allows us to determine a future emission pathway that would achieve that. If we discover that doing so would be extremely economically damaging, we could chose to accept a higher level of risk.
Hence, my point is that claiming that we should use the best estimate ignores that there is a better than 50% chance of it actually being higher than this and a not insignificant chance of it being substantially higher than this.
"I chose my words carefully."
There's a joke right there. You keep saying "maybe I shouldn't have said this, maybe I shouldn't have said that", "maybe could have phrased it better" etc.
Anyone using the "M" word has lost the argument any any respect IMHO. Further postings are to be ignored.
shub,
Maybe I've got confused between "choose" and "chose", but I think I used the one that would be interpreted as referring to what I've said in this particular case, rather than the one that could be interpreted generally. But maybe I chose my word poorly in that particular case :-) What I meant was that I added words like "prominent" because I know that Nic Lewis doesn't never mention it and I have no issues with his papers. It's my opinion that anyone who was to look at the Lewis & Crok report or read anything that has used it (for example Matt Ridley's work) would conclude that climate sensitivity is probably low. I would assert that that is the wrong conclusion and that that ignores caveats. You can disagree.
I will say that I'm starting to discover that there are some who think we need to simply choose a value and that if we are to do so we should choose the best estimate. My claim is that from a policy perspective that is wrong. We can't ignore the possibility that climate sensitivity could be high if we are to consider the policy implications associated with climate change.
Okay, here's a question I've always wanted to ask Nic Lewis, Matt Ridley and our host. Let's say we go ahead with what I think you would all support: assume that climate sensitivity is quite low (TCR about 1.35 degrees) and do little to actually address climate change, given that even a relatively high emission pathway (RCP6.5) would only - if TCR is 1.35 - produce about 2 degrees of warming by 2100. Okay, if you're right then maybe you have saved us from making costly decisions. What about if you're wrong, though? Does that concern you at all? (just to avoid this being addressed by asking me a question - of course I'm concerned that we'll make stupid policy decisions to address climate change. That's why I'd rather we actually engaged in sensible dialogue about it.)
Sandy,
I wasn't really trying to win one nor was I trying to gain any respect.
On the Slingo thread, Gareth was asking for a question to put to her in her upcoming talk. That would be a reasonable start.
Another one might be.
I only need to look back at records and observations to tell me that the climate will probably start to warm again in about 15 years though if I were a politician (or even a real scientist) I wouldn't bet the farm on it. I would be watching the current and up-coming solar cycles very carefully and keeping my options open.
And any scientist or advisor who came along and told me that we had to act now! because otherwise catastrophe was inevitable and tipping points would be reached (especially people who told me that by 2000 half of New York would be under water!) would be presented with his own personal tipping point very rapidly.
That's why I'd rather we actually engaged in sensible dialogue about it.
Indeed, but in this highly politicised arena a sensible dialogue is unlikely to occur.
In risk assessment (not risk management) decisions it is normal practice to err on the side of safety, in other words you would not normally base your control measures on the assumption that the central estimate was the appropriate number for your point of departure. However nobody would base their assessment on the high end of the range unless that decision had no adverse consequences.
In the subsequent risk management decision, i.e. the policy generating discussion there needs to be an understanding of the degree of certainty that can be ascribed to the risk assessment information. How much is based on empirical data rather than on computer predictions, what are the consequences of varying degrees of management action from zero response to complete prohibition. The consequences are of course related both to action and inaction.
In the present circumstances decision makers are faced with a situation where some (most) climate specialists are saying unless CO2 releases are reduced by x% "this" might happen and "these" might be the consequences and we are highly (?) confident that this might happen because we can predict (project) this using our sophisticated computer models. If the scientists are being objective and acting as policy advisers they would say our best estimate of the climate sensitivity suggests that there is nothing to worry about in the next 100years although our worst estimate indicates that the outcome would have severe consequences.
The sensible policy maker should then ask: How good have your models been shown to be so far at prediction (projection) of the situation in the real world?
The next question should be what do I have to do, at what cost and with what consequences to achieve either your extreme projection or your projection based on best estimates?
This is the debate that needs to take place. The costs and implications, especially for developing economies of effectively eliminating the use of fossil fuel by 2050 would have severe implications for society we need more certainty and certainly more empirical data to show that that is necessary before taking such a decision.
ATTP is demoonstrating for us how easily even fairly sharp minds can be entrapped in apocalyptic claptrap.
Is the gist of ATTP’s argument that it’s ok for the might and authority of the AGW machine to lie and cheat because (in his opinion) sceptics don’t stress all the cases put by the warmists? The ‘you started it first’ defence.
Question – when did we ever read or hear a warmist speculate about the best case scenario or even mention there was one?
There has only been one viewpoint presented from the start, that AGW would be catastrophic.
"What about if you're wrong, though?" asks ATTP.
Do you really think that anything being done now would pospone thermogeddon? Serious CO2 reduction would require very different levels of commitment and climate science is not fit to spawn that type of response. Nothing sceptics are doing is substantially contributing to the lack of belief. The science evidence itself is weak. Particularly weak because it has no form of quality control. The moment scientists decided to nudge public opinion beyond what the science could prove they lost the right to be trusted.
Science in an Age of Endarkenment
"The past 30 years or so have been an age of endarkenment. It has been a period in which truth ceased to matter very much, and dogma and irrationality became once more respectable".
And Michael Mann sues for damages if someone publicly characterizes such deceptions (omitting the ifs, ands, buts and caveats) as fraud. And cites various whitewash investigations as proof than no deception took place. Legal proceedings have already consumed two years and a million dollars. So much for freedom of speech.
ATTP "Okay, if you're right then maybe you have saved us from making costly decisions."
No, what we've done so far has been costly and has had almost zero impact on emissions. The biggest reductions have been the substitution of gas for coal, which would have happened anyway. To start reducing CO2 we have to spend an unimaginably bigger sum and change our lives radically.
Oh dear, wrong again.
"claiming that we should use the best estimate ignores that there is a better than 50% chance of it actually being higher than this".
The Lewis best estimates are medians, so the chance would be 50%.
But from someone who doesn't even know what a boundary condition is, maybe such an elementary error is not so surprising. Is he really a professional scientist?
Paul,
Jeez, you don't make it easy, do you? And just as I was starting to feel a little bad about the manner in which I had responded to your earlier comment.
Yes, you're quite right. Nic Lewis' numbers are medians. Yes, what I wrote is wrong. Believe it or not, though, I do know the definition of median. Believe it or not, I had not intended to get it wrong. The person who reviews my blog comments is on leave today.
Yes, you're also right that the way I was using the term "boundary condition" was wrong. I will add, though, that I think it has been used in that way by others. Additionally, the point I was getting at is that there are some conditions that bound/constrain the total energy of the system, such as the energy received from the Sun, the amount reflected and the composition of the atmosphere. It's difficult to see how there can be large deviations from what these conditions would requires, simply because the system happens to be chaotic.
Here's a challenge. A typical pseudo-skeptic tactic is to nit pick errors without actually addressing the point that is being made. Given that you are actually an academic, you could - if you wish - try to address the actual point being made, rather than simply choosing to nit pick. If all you want to do is nit pick, though, that's also fine. I don't really care either way, and don't really expect much else these days.
Yes.
Arthur,
Yes, you're probably right. The rest of what you say seems largely consistent with what I was saying earlier in this thread. The real issue is balancing the risks associated with climate change with the risks/costs associated with minimising/mitigating those risks. Maybe we can agree on that, even if we can't actually have a sensible dialogue about it.
Tiny,
No, why would you possibly think that?
The costs at both ends of the estimation are extremely high. Why should the higher end be given any extra credibility?
Mike,
Sorry, should I have made it more obvious that that wasn't serious? :-)
Shub,
Both ends of which estimates, and why do you think one is being given more credibility than the other?
Both ends of a climate sensitivity probability distribution.
Shub,
Seriously? What do you mean? The bottom end means less warming and - as I understand it - fewer risks associated with climate change. The bottom end doesn't mean we'll cool by as much as we'd warm if climate sensitivity is high.
The harm at the higher end happens (a) if the C/AGW hypothesis is broadly correct, (b) we could have cut CO2 but didn't, and (c) sensitivity actually turned out be high.
The harm from the lower end happens (a) if the AGW hypothesis is narrowly correct, (b) we need not have CO2 at all but nevertheless did, and (c) sensitivity actually turned out to be low.
Remember, the climate system is not amenable to experimentation so the risk of post-hoc justification exists at both ends.
Shub,
Okay, but that's not the same cost at both ends of the sensitivity distribution. There are essentially 4 possible scenarios (broadly at least)
1. Climate sensitivity low - do nothing - no cost.
2. Climate sensitivity low - do something - cost.
3. Climate sensitivity high - do nothing - cost.
4. Climate sensitivity high - do something - cost.
The correct comparison, in my opinion, is between 3 and 4. We don't need to consider 1 and 2 when assessing what to do. In a sense the cost associated with 2 and 4 are the same, since they reflect the economic cost of acting, not the costs associated with climate change. 1 is what we could do if we just want to gamble on a low climate sensitivity, but that would seem to be a poor way of making policy.
ATTP “Omitting the ‘doubts, the caveats, the ifs, ands and buts’ is not a morally neutral act; it is a subtle deception that calls scientific practice into disrepute.
Maybe you should have a quiet chat with Nic Lewis and Matt Ridley?”
Instead of conceding that there have been faults in warmist communication (because you can’t) you pointed the finger at sceptics instead. That’s a strategy learnt in the playground.
“A typical pseudo-skeptic tactic is to nit pick errors without actually addressing the point that is being made.”
Pot and kettles.
I get that you don’t think that AGW message has been skewed towards the extremist end but it doesn’t matter what you think. It matters what everyone else things and I’m talking about the wider world than the climate blogosphere. People now roll their eyes when climate is mentioned. They say ‘how’s that global warming’ when it snows. They know that the evidence was exaggerated. They've stopped listening because they don't trust people who spin messages. There could be a point where you need the public to trust you but they'll have got fed up of you crying wolf.
Before we proceed to 'costs' as in penny counting a la Stern, we can and should think whether the expense is required. 4 and 1, for example, are equivalent. In the former, we spent money and effort but controlled the climate and in the latter we did not spend money and there was no need to. That 4 is greater than 1 in absolute dollar/pound terms is secondary to the fact that the money was well-spent (or well saved).
In addition, the option to be able to do a 'time-out', or a spot-check should be kept alive. The option to walk back from foolish and costly measures, if, and as and when they are shown to be white elephants and albatrosses, should be kept open.
Knowing history and the structure of democratic governments, the possibility of the above happening is highly unlikely. Costs and pork-barrels added on by governments hang around forever eventually sinking all involved. It is an easier burden to prevent stupid people from doing their favourite things than to be cleaning up after them.
More than quarter of the comments in this post to date are from ATTP.
Frankly, I'm bored. It is quite apparent that he is not open to argument, so why bother?
Please wake me up when things have gone quiet again.
I've only just spotted this thread. I won't bother defending myself against ATTP's attacks on me as others have already done so.
However, I'd like to put in a good word for Keith Shine - one of my fellow participants at the Royal Society climate feedbacks workshop at Chicheley Hall last week. I've met Keith several times now, and he strikes me as a basically honest, very decent guy. I simply don't think he's the sort of person who would set out deliberately to mislead policymakers. That's not to say he doesn't have a biased view of climate models and their reliability, or that he may have failed to spell out the uncertainties involved and to include adequate caveats.
Tiny CO2:
Actually, what I think you really meant was; “…has hadalmostzero impact on emissions.BTW, shame on you all! Do you not realise that, if you cannot understand what aTTP has written, then it is your fault for not reading it correctly, not his for not making it clear, despite any errors that it might contain.
(p.s. Messenger is right; he has succeeded again, taking this discussion way off thread and talking about how great a person a certain aTTP is, and you are just chaff, to be blown away in his threshing mill…. Meh…)
aTTP, care to give us the stats on traffic to your backwater blog since you started trolling here?
Nic,
If it seemed that I was attacking you, I apologise, that wasn't my intent. I must admit, though, that I do find it rather frustrating when people seem so ready to cast something as an attack. Let's remind ourselves of this post. It is our host using what - in my opinion - is an embarrassingly poor report on research ethics, to justify his criticism of someone else and to imply that this other person's behaviour reflects badly on science as a whole. I think that is rather appalling and I'm pleased that you've at least partly stood up for Keith Shine.
Whether you like it or not, from what I've seen you seem reluctant to acknowledge the caveats with your own work, you seem reluctant to accept that even your work does not rule out that warming could be quite a bit higher than your best estimate suggests, and you seem to have a habit of nitpicking when others comment on your work, rather than engaging with the point they're trying to make. I may, of course, simply have seen a small selection of what you've said and so, maybe, I'm wrong. If so, you could always try harder to illustrate that I am so that I can feel guilty for having criticised you in this way.
In the interests of possibly ending this on a more positive note (although maybe not) I have learned something while engaging on this thread that I hadn't appreciated. It seems (and feel free to correct me if I'm wrong) that you (and others) think that what we should be doing is determining a best estimate for climate sensitivity and should be basing our policy decisions on that best estimate. Again, if I'm wrong feel free to correct me. If I'm not, that might explain the impression I have. If I have correctly characterised your position, then it may explain why I hold the view I do, but I think it's the wrong policy position to hold. I don't think that the correct policy option is to work from the best estimate. I think we need to consider the range and the risks associated with the higher estimates and balance that with the risks/costs associated with minimising/mitigating those risks. If you disagree, that may be a more interesting discussion than a discussion about climate sensitivity itself.
Shub,
Can you explain what you mean here. They may result in the same climate state, but only one of them is real.
joanna,
Almost exactly the same as before I started commenting here (maybe slightly lower as I think I've ended up spending more time here than on my blog).
For what it's worth, I am someone who had a quick look at the Lewis and Crok report. This was (iirc) back when it came out, and I also saw some of the things that (iirc) Nic Lewis said about it here and/or at CA. I firmly remember being left with the clear impression that Lewis' range was similar to the IPCC's while the central case was lower. I was also left with the impression that the authors had clearly brought this to my attention, and that they had presented an argument as to why the lower central case (along with a general shifting of the probabilities in that direction) was decisive while the similar range was not. (To be sure I was also left with the impression that the authors would have been happier if they could have said that their range was narrower too. But not that they'd suggested, by act or omission, that it actually was.) Again, to emphasise, I wasn't some adversarial expert combing the fine print of the report, I'm just an ignorant galoot who cast his eyes over it—and someone who's probably more predisposed emotionally to believe Lewis and Crok than to look for reasons to doubt them—and the IPCC-like range was completely evident to me, plain as day. If the report ignores caveats or insinuates the wrong conclusion it has to be in some other way.
What I find funny is that the physics dude(tte) is all the time saying things like "If all you want to is X/Y/Z/be a denying denialistic denier/nitpick/whatever, than go ahead BECAUSE I AM NOT EXPECTING MUCH ELSE."
Or stuff like "I wasn't all that optimistic that it would end up being an interesting debate, but I had hopes."
Or something along the lines of "I don't care what you do/say/think, and I am not expecting any of you to actualy do x/y/z."
then what are u still doing here?
Weeks I have seen you say stuff like that here. One would think you would eventually give up if you really felt that way.
Now it just looks like, well, TROLLING. Trying to get a reaction by being an obnoxious dillhole. Sad really.
What does "almost exactly the same" mean? How about some independently verifiable numbers?
Typical, just trust us. Never mind the facts, feel the length.
joanna,
Why? You asked for what had happened to my stats since I started commenting here. Nothing has happened, so why would actual numbers help? If it makes you feel better, my stats are probably a lot less than our host's, and that's fine by me.
ATTP:
To be blunt, when you say "I still maintain that's he reluctant to discuss [his model's ranges] in a manner that I would regard as reasonable" the impression I'm getting is that you're trying to blur Lewis' assertions that the IPCC-like range doesn't undermine his position with the suggestion that he's been trying to hide its existence. But these two things are radically different in the context of the suggestion that Lewis is dishonestly hiding uncertainties. The latter would amount to dishonestly hiding uncertainties; the former does not, even if for the sake of argument we assume that Lewis' justification for that position is as mad as a bag of cats. (There would be a reasonably separate issue of dishonesty if he didn't believe that justification, or if he were deliberately hiding things about it.)
"aTTP, care to give us the stats on traffic to your backwater blog since you started trolling here?
Dec 18, 2014 at 8:21 PM | Registered Commenterjohanna"
If, and I say if, this troll really is an academic, then I hope his/her/its paymasters do a serious time and motion (money) study as to how much of their time and money (and our money) he/she/it is spending blogging on his/her/its site, this site and others?
wijnand, I agree - wants to discuss things with people and call them idiots at the same time. Kind of like having the cake and eating it. I don't know why.
attp, what do you mean "only one of them is real". I was comparing your 1 and 4 - both scenarios in which our anticipatory response turns out appropriate.
It is quite evident, given that our set-up is based on what would happen in the future, the confusion matrix should involve three dimensions: sensitivity - high or low, our response: yes or no, and, validity of prediction: true or false.
anonym,
I'm not trying to suggest anything like that and am certainly not even trying to imply any dishonesty. In fact, the latter part of my response to Nic was an acknowledgement that maybe I have a better understanding of his position than I had prior to engaging here. Having said that, if Nic does indeed think that the best estimate is what we should really be focusing on, then I disagree, but that may explain why I have the impression that I have. Of course, if I do misunderstand Nic's position, he's free to correct me as I have no interest in misrepresenting anyone.
If you did read the Lewis & Crok report and did indeed come away with a good understanding of the caveats and ranges, then the opinion I expressed was clearly wrong. However, that's why I express opinions, rather than state facts. I still think that a report titled "over-sensitive" and which has a prominent table that doesn't include the range of warning is quite likely to be interpreted as I suggested. I may, of course, be wrong.
Shub,
Do I? I thought I'd only done that once on this thread :-)
Maybe you can answer my question which was why you think 1 and 4 is equivalent?
Shub,
Okay, you did answer. It's getting late. I don't agree that they're equivalent. There is only one reality. In the simple scenario we're considering, climate sensitivity is either high or low (okay, it's not binary, but let's ignore that for now). If it's high and we do something, we can minimise the risks and everything's fine (okay, I know that's simple). If it's low and we do nothing, we're also fine. But until we know that it's low, we can't really ignore that it could be high. If we do, we're essentially gambling on it being low and hoping that we're not wrong. So, again, I don't understand how you can suggest that 1 and 4 are equivalent.
"t I'd only done that once on this thread :-)"
No, that has been your pattern of interaction on this blog.
"which was why you think 1 and 4 is equivalent?"
I thought I already did - twice.
In any two predictive scenarios where the actions are undertaken prior to the outcome, and the actions turn out appropriate, they are equivalent. In 1, we did not embark on expensive mitigation actions and harm the present poor. In 4, we undertook expensive mitigation and saved the future poor.
EDIT: Saw your response above. I see your point and do get it. But you are refusing to see the other side. When the possibility of action confronts us - remember, that action today will harm today's poor. I think you are missing this in your framing.
Shub,
I fail to see how me commenting on the likely outcome of a discussion here is calling people idiots. If you think my view of how discussions here are likely to progress is wrong, prove it, rather than complaining about my impression.
They're equivalent in the sense that they produce the same outcome if they both turn out to be the most appropriate response. However, there is only one climate sensitivity and so I fail to see how the fact that they are equivalent in some hypothetical scenario where both could be the most appropriate response, is actually relevant. Either climate sensitivity is low enough for us to not bother doing anything, or it's high enough that we should do something (okay, again, in our hypothetical scenario - our current reality has a continuum). Deciding which is the most appropriate response is the crucial issue.
That's right - which is why I asked why low-probability values at one end of the sensitivity distribution be weighted than those at the other end - they are both low-probability, and a certain degree of harm to the present, a substantive one in all likelihood, is inherent in acting on either.
The questions I pose are simple enough - but a graphical representation of the three axes would perhaps be better and beneficial.
Shub,
Well because I don't see how that is how a risk assessment works. There's no climate risk associated with the low-end of the climate sensitivity distribution. The climate risk is all at the high end. Therefore we should be balancing the risk associated with high climate sensitivity with the risk/costs associated with minimising/mitigating those risks. If the damage associated with minimising these risks is greater than the potential damage associated with climate change, you might not bother. If it damage/cost/risk associated with minimising the climate risks was small by comparison you might go ahead.
I don't really like analogies but I'll try one. If your tyres on your car are going bald you might consider paying to replace them. You'd consider how much that would cost and would consider the risk of a serious accident if you didn't. If replacing them now meant not feeding your kids for a week, maybe you'd wait till next month. If, however, it just meant reducing your savings a little you'd probably go ahead. It would seem highly unlikely that you'd go "it'll probably be fine, I won't bother".