Sceptic Links Discussion

[Admiral_Bobski]

While reading through the Y2K Bug thread I saw the link posted by gmoran to the Climate Feedback blog. I looked around the site and found another recent post which I thought would be (or might be, at any rate) relevant to the model discussion. Here it is:

http://blogs.nature.com/climatefeedback/20...of_climate.html

Any thoughts on this?

CB

[Gray-Wolf]

While reading through the Y2K Bug thread I saw the link posted by gmoran to the Climate Feedback blog. I looked around the site and found another recent post which I thought would be (or might be, at any rate) relevant to the model discussion. Here it is:

http://blogs.nature.com/climatefeedback/20...of_climate.html

Any thoughts on this?

CB

I enjoyed the section of the potential for 'medicanes' since Vince's arrival in Spain 2004 or Caterinas landfall in South America.

Once again I'm left wondering how the Skeptics would deal with a Cat 2/3 making landfall in N. Africa/SW Spain seeing as we have no folk history (unlike the Meso-Americans) of such events..........probably put it down to a 1 in a million year event LOL.

[Admiral_Bobski]

I enjoyed the section of the potential for 'medicanes' since Vince's arrival in Spain 2004 or Caterinas landfall in South America.

Sorry, which part is that? (I'm either consistently overlooking that part or I'm reading a different page!)

Once again I'm left wondering how the Skeptics would deal with a Cat 2/3 making landfall in N. Africa/SW Spain seeing as we have no folk history (unlike the Meso-Americans) of such events..........probably put it down to a 1 in a million year event LOL.

Once again, I am left wondering why (or how) an event such as this would indisputably prove Human-Induced Climate Change...

Nonetheless, did you have any thoughts on the models as a result of the article?

CB

[Admiral_Bobski]

Doesn't anyone want to discuss model reliability?

CB

[parmenides3]

Ues, but I've been busy doing other things. I'm goinf to look for a link to the new Stainforth et al paper, which I think you'll appreciate.

:)P

But don't hold your breath!

Seriously; I really have been very busy, so this has passed me by for a few days, but I am onto it...

:)P

Edit; before I go, two questions: what are climate models for, in other words, what it is we expect them to be able to do?

and: what would you want from a climate model, which would demonstrate that it had some use or value in planning for the future or deciding what problems there really are?

:)P

Edited August 18, 2007 by parmenides3

[parmenides3]

Here's a link to the abstract of the Stainforth paper; you are going to love it C-Bob! (no irony). Plenty to discuss there...

http://www.journals.royalsoc.ac.uk/content/n471534w71q29k04/

:)P

[Admiral_Bobski]

Here's a link to the abstract of the Stainforth paper; you are going to love it C-Bob! (no irony). Plenty to discuss there...

http://www.journals.royalsoc.ac.uk/content/n471534w71q29k04/

:)P

Interesting abstract - shame I can't read the paper itself (at least not without forking out money I don't have! ). I would be particularly interested to read the section where "we discuss where we derive confidence in climate forecasts and present some concepts to aid discussion and communicate the state-of-the-art."

In answer to your questions (and apologies for my brevity)...

What do we expect climate models to do?

I would expect models to help scientists refine their understanding of climatic processes, and possibly to give suggestions as to what areas of research to focus on next. In this sense they should be used only for the scientists' benefit, and not to make specific predictions (or even projections) about future climatic change.

What would you want from a climate model, which would demonstrate that it had some use or value in planning for the future or deciding what problems there really are?

This is a tricky question, and one I shall have to return to at a later date to go into detail with a response. (I am lucky enough to be able to go on a quick break to North Wales this week, so I shall mull it over while I am away!)

You see, my big problem with this question is that it is assuming that the models should be used for this purpose. This is where the politics comes into the climate debate, really, because the projections given us by the models are then used by politicians as excuses for their political conduct. Perhaps this is one of the reasons I find the models such a contentious part of the debate - basically, I feel that models should be used by scientists to further their understanding, and not by politicians to further their agendas.

I shall think on this further...

CB

[parmenides3]

Here's a copy of the paper, then (naughty me...): http://www.lse.ac.uk/collections/cats/pape...evance_2007.pdf

:)P

How do I do that?

[Admiral_Bobski]

Here's a copy of the paper, then (naughty me...): http://www.lse.ac.uk/collections/cats/pape...evance_2007.pdf

:)P

How do I do that?

Cheers for that, P3 - I see you are a man of infinite resource!

I'll have a read when I get a chance. (Maybe I'll print it out and read it on holiday - that'll please the wife!)

CB

[jethro]

I'm not sure where to post these two links, so many discussions here, mostly heading down the same road so I guess which thread is mostly irrelevant. Before anyone jumps on me and says this proves nowt in the AGW/sceptic debate; it's not intended to, it's just interesting and one more thing to puzzle over. I hadn't read it before and in case others haven't either I've posted the links.

http://climatesci.colorado.edu/2007/08/14/...by-roy-spencer/

http://climatesci.colorado.edu/2007/08/15/...and-colleagues/

[parmenides3]

I'll be honest and say I have avoided the Spencer stuff on CS for two reasons; first, just look at the rubbish that the commenters were producing, including a discussion of Darwin vs Intelligent Design; WTF? The tone of the debate was either uncritical praise or cynical dismissal, with little in between. Most telling comment for me was number 88, by Spencer himself.

The discussion of Dick Lindzen's 'iris' theory is very odd. I had thought that this had been thrown out by serious scientists a few years ago, as not much kop. Because it has strong links to a certain strand of denialism in the USA, it's hard to separate discussions of the theory from discussions of the motives for publishing the theory. I don't have a high opinion of it, personally.

Do climate models miss important feedbacks? Are they reliable enough to use as predictors of likely future climates? The answer to the first is probably, yes, but it doesn't seem to make much difference to the overal warming signal; to the second...I address this today on the blog, but I'm not really offering the answer you might expect; bottom line, in some ways, yes, in others, don't know, in others still, probably not, but they still have value.

:)P

[jethro]

Any thoughts on this?

http://www.ecd.bnl.gov/steve/pubs/HeatCapacity.pdf

[Devonian]

Any thoughts on this?

http://www.ecd.bnl.gov/steve/pubs/HeatCapacity.pdf

No, but climate modeller James Annan has.

DO I understand the maths? No frankly I don't.

Edited August 21, 2007 by Devonian

[jethro]

The tone of the debate was either uncritical praise or cynical dismissal, with little in between

Much like here then most of the time.

The discussion of Dick Lindzen's 'iris' theory is very odd. I had thought that this had been thrown out by serious scientists a few years ago, as not much kop. Because it has strong links to a certain strand of denialism in the USA, it's hard to separate discussions of the theory from discussions of the motives for publishing the theory. I don't have a high opinion of it, personally.

I don't personally think it should be dismissed out of hand, nor am I aware that it has been. Opinion for motive for publishing is just that, opinion. It comes down the the same old debate of does it matter who provided the funding, so long as the data is correct. If Greenpeace fund a project it can be denegrated by sceptics as being bound to be skewed in favour of AGW, if Exon provide funding it will be denegrated as being skewed in favour on non AGW cause. I believe that type of bias of thought pattern compounds the devision and does little to further understanding.

:)P

[parmenides3]

The tone of the debate was either uncritical praise or cynical dismissal, with little in between

Much like here then most of the time.

The discussion of Dick Lindzen's 'iris' theory is very odd. I had thought that this had been thrown out by serious scientists a few years ago, as not much kop. Because it has strong links to a certain strand of denialism in the USA, it's hard to separate discussions of the theory from discussions of the motives for publishing the theory. I don't have a high opinion of it, personally.

I don't personally think it should be dismissed out of hand, nor am I aware that it has been. Opinion for motive for publishing is just that, opinion. It comes down the the same old debate of does it matter who provided the funding, so long as the data is correct. If Greenpeace fund a project it can be denegrated by sceptics as being bound to be skewed in favour of AGW, if Exon provide funding it will be denegrated as being skewed in favour on non AGW cause. I believe that type of bias of thought pattern compounds the devision and does little to further understanding.

:)P

Sorry; to be clearer; two follow-up studies attempted to replicate Lindzen's results and couldn't; in fact they got very different results. They concluded that Lindzen had made some fundamental errors. Lindzen responded to defend his hypothesis, the response was responded to, then the hypothesis just sort of vanished. Googling 'lindzen iris' gives a range of results, but the sources I tend to trust all seem to agree that the iris hypothesis has been sidelined, basically because it is unproven and the foundations of it are questionable - hence the tendency for the vast majority of climate scientists to simply ignore it; as things stand, it has nothing to tell them; it remains a hypothesis and nothing more. Observations done since the original paper in 2001 seem to contradict the theory, too, so the passage of time hasn't made it look any healthier.

I agree that the source should not detract from the truth or falsity of an argument' all I said was that it's hard to separate out, in this case, arguments about whether the idea has any merit from arguments about how 'real' AGW is; this is mainly because it is still cited as evidence against AGW on some sites, even though it self-evidently is not.

Hope that makes things clearer...

:)P

[Admiral_Bobski]

Hi everybody!

I'm not sure if anyone is interested in talking about models any more (and I'm not talking about Kate Moss here, you understand), but I thought I'd have one last pop at the subject.

I confess to not having read the whole paper which P3 kindly linked to, but I've had a scan through some of it and I came across this interesting part:

"We have no hope of confirming the reliability of a predicting system, so confident statements about future climate will be more qualitative than one might wish and conditional on a number of significant assumptions. They may nevertheless by extremely valuable in guiding societal response to the climate change problem."

(End of 1st paragraph, page 2156)

This comment strikes me as being a tad oxymoronic - essentially it is telling us that the problem cannot be substantiated by the models, but the models can encourage people to fix the unsubstantiated problem. Is this not a misapplication of science, or indeed a prime example of science being politicised? ("The models don't prove a thing, but look how bad the models show the future to be...")

Any thoughts?

CB

[Gray-Wolf]

I have a strange feeling of 'Deja-vu' right now.

It is worrisome for some isn't it C-Bob?

The climate system has way, way too many nuances for any 'team' to spot them all for inclusion in any model but should we abandon the pursuit feeling it 'unproductive'? At least it pokes us in a 'direction' for our further ponderance................

Edited September 2, 2007 by Gray-Wolf

[parmenides3]

Hi everybody!

I'm not sure if anyone is interested in talking about models any more (and I'm not talking about Kate Moss here, you understand), but I thought I'd have one last pop at the subject.

I confess to not having read the whole paper which P3 kindly linked to, but I've had a scan through some of it and I came across this interesting part:

"We have no hope of confirming the reliability of a predicting system, so confident statements about future climate will be more qualitative than one might wish and conditional on a number of significant assumptions. They may nevertheless by extremely valuable in guiding societal response to the climate change problem."

(End of 1st paragraph, page 2156)

This comment strikes me as being a tad oxymoronic - essentially it is telling us that the problem cannot be substantiated by the models, but the models can encourage people to fix the unsubstantiated problem. Is this not a misapplication of science, or indeed a prime example of science being politicised? ("The models don't prove a thing, but look how bad the models show the future to be...")

Any thoughts?

CB

I'm not sure it's saying that the problem can't be substantiated by the models, C-Bob, its saying that we can't tell (yet) how good the predictions are. This means that comments are not as quantitative or as absolute as we would like, which is a shame. The problem is substantive; the effects and future emissions pathways are not - they are still on likelihoods and estimates (though still based on 'best available evidence'). This means that, whilst we cannot know for sure what will happen, we can still have a good idea of what might happen, and we can assign a confidence or probablility value to it.

Knwing these uncertainties about the future, nonetheless we can still consider a whole range of 'what if' scenarios, and develop an informed evaluation of which of these have the greatest chance of coming about. This is aside from certain known consequences of certain known physical realities, such as the relationship between CO2 and Global Mean Surface Temperature.

What needs to be done, then, is to look at how the changes which have already occurred are already affecting the system, ot parts of it, and consider what this tells us about both the likelihood of these things continuing and the predictive skill of the models which showed this before it happened.

:)P

[The Penguin]

For shame 3P, C-Bob is correct; the quote clearly offers the excuse, 'sod the science, get with the programme', and that's not right whether it's the models or the predictions that are unfounded.

[parmenides3]

For shame 3P, C-Bob is correct; the quote clearly offers the excuse, 'sod the science, get with the programme', and that's not right whether it's the models or the predictions that are unfounded.

have you read the paper? Do you know why S & al say this? (the context). Because 'sod the science' wouldn't be right, it isn't what these scientists are saying. And there is big difference between foundation and substantiation. The reason they cannot substantiate the models is because to do so requires that the future has happened. All they are saying is that, as 'proof' is dependent on the passage of time, so waiting for 'proof' is rather pointless; we have to go with what we have, including the uncertainties.

If you are so willing to believe that scientists are inept, dishonest or stupid (I'm not saying that you are, but it's how this reads), then nothing I am going to say is going to make any difference. As I have said elsewhere, is not so much what we do know that scares the scientists, its what we don't know; how bad it could get being one key uncertainty.

Regards,

:)P

[Admiral_Bobski]

have you read the paper? Do you know why S & al say this? (the context). Because 'sod the science' wouldn't be right, it isn't what these scientists are saying. And there is big difference between foundation and substantiation. The reason they cannot substantiate the models is because to do so requires that the future has happened. All they are saying is that, as 'proof' is dependent on the passage of time, so waiting for 'proof' is rather pointless; we have to go with what we have, including the uncertainties.

If you are so willing to believe that scientists are inept, dishonest or stupid (I'm not saying that you are, but it's how this reads), then nothing I am going to say is going to make any difference. As I have said elsewhere, is not so much what we do know that scares the scientists, its what we don't know; how bad it could get being one key uncertainty.

Regards,

:)P

I got the distinct impression that they were saying that the models cannot be substantiated because of 1) the qualitative, rather than quantitative, nature of said models and 2) the "number of significant assumptions" upon which they are based (and "assumptions" are, by their very nature, unproven).

So, there are a number of assumptions involved which cannot (as yet) be verified, which is a pretty shaky foundation for a scientific analysis. Furthermore, the models are qualitative by virtue of the fact that, under certain circumstances, they appear to show reasonable, physically plausible, results. Due to our inability to quantitatively assess the models, though, we cannot be certain - to any reliable degree - that these results have any place in the real world, despite their theoretical plausibility.

That said, how can (and why should) we act on the basis of what the models show us?

CB

[The Penguin]

If you are so willing to believe that scientists are inept, dishonest or stupid (I'm not saying that you are, but it's how this reads), then nothing I am going to say is going to make any difference. As I have said elsewhere, is not so much what we do know that scares the scientists, its what we don't know; how bad it could get being one key uncertainty.

Regards,

:)P

Certainly not stupid. But why is the emphasis on ignorance being a case for fear of worse to come, rather than joy at a potential improvement?

[Stratos Ferric]

...

So, there are a number of assumptions involved which cannot (as yet) be verified, which is a pretty shaky foundation for a scientific analysis. Furthermore, the models are qualitative by virtue of the fact that, under certain circumstances, they appear to show reasonable, physically plausible, results. Due to our inability to quantitatively assess the models, though, we cannot be certain - to any reliable degree - that these results have any place in the real world, despite their theoretical plausibility.

That said, how can (and why should) we act on the basis of what the models show us?

CB

Minor quibble: "qualitative" as opposed to "quantitative" concerns the presence of numeric data or not. The extent to which the numeric data is robust simply determines the robustness of the model, NOT it's quantitative nature.

In terms of reliability and robustness, you're correct at one level, but all of scientific (and much economic) endeavour is founded on precisely this kind of analysis. You talk as if because we don't know precisely what will happen we cannot model. The modellers way around this is to run the model many many times, with a range of ever wilder assumptions and interdependencies.

The simple fact is, with all current climate modelling, and after throwing in even fairly obviously wide variables, the total predictive corridor points north: the oft quoted middle ground is around 2-3C of warming over the next century, but the range around this is something like 1-6C. I'm not aware of any models that point south.

The other process of validation of models is to retrorun them. Feed in known data from, say, 1980, and see whether 27 years on the model produces what we now see in 2007. My understanding is that the models are gettting to be pretty good at this. This being the case it takes some fairly warped logic, or the presumption that something sudden, and as yet unknown, will happen to undermine all forward projections based on the current logic.

I suppose this latter cannot be ruled out completely, however, I am reminded of a client of mine a few years ago that spent 18 months waiting for the reservoirs to replenish, all the while sitting on their corporate hands and not taking alternative (more expensive) action. The price for sitting on their hands was that standpipes were erected on the streets, many lost their jobs, and the required investment far exceeded the cost of earlier pre-emptive action.

The history of the world is littered with examples of people who were blinkered regarding probability because they chose, for a range of reasons, to hang blindly onto possibility.

[Admiral_Bobski]

Thank you for your reply, SF. I do agree with you to a certain extent - I am not saying that climate scientists shouldn't use models, as they are (as I have said before) an extremely useful tool in any scientific discipline. What I am saying is that the models should be used in the the right way and for the right reasons. Using models to help climatologists better understand the climate system is laudable and is precisely the way the science needs to go (in part, although nothing beats a good 'ol piece of down an' dirty research!).

What irritates me about the models is the way in which they are used. Whether it's the media, or the politicians, or the scientists themselves, I don't think so much attention has ever been paid to a bunch of scientific models before (excepting weather forecasting models, of course, but these are specifically used to predict weather and are not used for scientific research). Further, proposing potentially significant policy changes on the basis of the climate models is patently a misapplication of them. (There's nothing like a good bit of alliteration )

Minor quibble: "qualitative" as opposed to "quantitative" concerns the presence of numeric data or not. The extent to which the numeric data is robust simply determines the robustness of the model, NOT it's quantitative nature.

That was supposed to be my point, but perhaps (on re-reading my post!) what I said was a bit vague. A qualitative analysis focuses on the qualities of the subject being studied - a qualitative analysis of the climate models can determine whether the outcomes of those models are scientifically plausible. I agree that the models results are plausible, but are they correct? That's a whole other issue.

A quantitative analysis will, effectively, accept the outcome no matter what so long as the numbers are correct (or accurate to within an allowed degree of certainty). When some (or many) of the numbers which are input are based on incomplete understanding and perhaps even a little guesswork then a quantitative analysis is not so easy to do. The impression I got from the paper was that there were too many assumptions and gaps in understanding with regards to the climate models to make them assessable in a quantitative manner.

Which only leaves qualitative analysis - scientists look at the model results and figure out whether those results are at least possible.

In terms of reliability and robustness, you're correct at one level, but all of scientific (and much economic) endeavour is founded on precisely this kind of analysis. You talk as if because we don't know precisely what will happen we cannot model. The modellers way around this is to run the model many many times, with a range of ever wilder assumptions and interdependencies.

As I say above, I have no problem with this at all - models are an extremely useful scientific tool. It's the emphasis that's placed on the models' predictive power, and how those predictions (or projections) are used to support policy decisions. (Dang, it's hard to separate the science and politics sometimes!)

The simple fact is, with all current climate modelling, and after throwing in even fairly obviously wide variables, the total predictive corridor points north: the oft quoted middle ground is around 2-3C of warming over the next century, but the range around this is something like 1-6C. I'm not aware of any models that point south.

A link was posted on one of these threads a while ago in which the author claimed that any "south-pointing" models were discarded, on the basis that they were "clearly" wrong (and by "wrong" they mean "don't agree with the concensus viewpoint"). There were references to back up this claim too - I'll see if I can find out which paper it was and re-link to it. I still think that the models are, to some extent, geared towards giving the results the climatologists "want" to see.

The other process of validation of models is to retrorun them. Feed in known data from, say, 1980, and see whether 27 years on the model produces what we now see in 2007. My understanding is that the models are gettting to be pretty good at this. This being the case it takes some fairly warped logic, or the presumption that something sudden, and as yet unknown, will happen to undermine all forward projections based on the current logic.

Again, retrorunning the models (or making "retrodictions") is all well and good and a valuable scientific tool. I have no complaints there. But recreating the broad climatological changes of the past 100 years doesn't necessarily make the models correct. To go back to a simplified analogy I have made a few times (and at the risk of getting boring and repetitive!), if you know the outcome of the equation x+y is 10 then you can easily piece together what the equation could have been to give that result. But that doesn't necessarily mean that you have actually selected the correct values for x and y.

I suppose this latter cannot be ruled out completely, however, I am reminded of a client of mine a few years ago that spent 18 months waiting for the reservoirs to replenish, all the while sitting on their corporate hands and not taking alternative (more expensive) action. The price for sitting on their hands was that standpipes were erected on the streets, many lost their jobs, and the required investment far exceeded the cost of earlier pre-emptive action.

The history of the world is littered with examples of people who were blinkered regarding probability because they chose, for a range of reasons, to hang blindly onto possibility.

I agree again - action is generally better than inaction. I fully advocate trying to do something about the way we pollute our planet. But having dramatic policy changes foisted on us by a government which wants to be seen to be doing something NOW! is not what I had in mind. It is shortsighted and potentially damaging to society as a whole. There's something wrong with making the consumers change their ways but not the producers. There's something wrong with telling everyone to cut their energy consumption as opposed to finding an alternative means of energy.

No Regrets, I say! No Regrets!

CB

[parmenides3]

C-Bob, if the models were the only tool in the arsenal of policy making, then your point would be quite strong. I actually agree with you that sometimes, too many assumptions about future conditions are based on the (unsubstantiated) predictive skill of models, especially RCMs and single or short-scale runs.

But the government (or the people advising the government) are looking at more than simply the model runs. they are looking at what the world was like when it was last about as warm as we expect it to get - the palaeo stuff; they look at the observations from a range of tools which measure variations, and look at the trends. Many of these are connected to the idea that, under warming conditions, certain artefacts of warming should appear in the data. They also look at the scenarios and try to estimate how emissions and other influences might change; these estimates are largely based on economic activity, rather than weather or climate. Population dynamics also gets a look-in.

Yes, it looks as if the current round of policies assume both AGW and the role of CO2, but they have good reasons to assume both of these, which aren't dependent on the models for their status. As to what is then done about this; along with most environmental policy, the burden of action (and cost) is placed on the end-user, the taxpayer, or, occasionally, the developer.

The government has developed a nominal energy policy to address this; it's the nuclear option (with enviro-friendly add-ons). Like it or lump it, there is no other way of getting enough energy to enough people in the next 20 years to have an impact on emissions. This holds, even if we all make a big effort and increase our personal energy efficiency.

The only way to have no regrets in twenty (maybe even ten) years' time is to face up to the unpleasant reality of our situation, both with the environment and the climate, and work very, very hard to develop a collective effort, then to encourage everyone else to do the same.

In the face of the difficulties this line of reasoning suggests, it is easy to understand why many people feel discouraged, even defeatist about the whole thing. It does look like an almost impossible challenge. But so was getting to the moon on the back of the computing power of a mobile phone. So was circumnavigating the globe. So was building the Great Wall, or any one of hundreds of examples of collective human endeavour.

Speech over.

:)P