Coby,
I accept that I should not have said [in my previous blog post, which can be accessed by clicking here] that well mixed GHGs “presumably have a similar warming impact in both hemispheres” (Figure 11 of Hansen et al, third map down in right column refers).
But I don’t agree that ‘there is no significant discrepancy between modeled and observed behaviours.’ For example, on my reading of Figure 18 of Hansen et al (2005), to which Gavin directed me, the observed warming exceeded modelled warming by a significant margin over much the greater part of the northern hemisphere in the 1979-2003 period. In the Antarctic the discrepancy was in the opposite direction.
For the world as a whole, both the observed and the modelled warming in this period was about 0.4°C – i.e., equivalent to less than one-seventh of the vertical axis of the Figure. Visual inspection suggests that, over about one-half of the globe, the DIFFERENCE between the observed and modelled warming for this 24-year period was similar to, or greater than, the observed average warming for the world as a whole for this period. I’m surprised that you think that these aren’t significant discrepancies.
I did not misinterpret the mapping of aerosols and their effects on the ‘Sulphur Cycle Experiment’ page of the climateprediction.net website. The description is quite clear: the map shows ‘the model’s surface temperature response to increasing sulphur emissions from pre-industrial levels.. to present day levels..’. The description also says that ‘The cooling effect of sulphate aerosol can be seen throughout the whole northern hemisphere’ (which isn’t strictly true: there is a warming effect in Northern Scandinavia).
The accompanying text says that ‘a PREDICTION of the climate of the 21st century needs to contain the effects of sulphate aerosol otherwise the warming trend may be OVERestimated’ (EMPHASES added). There are two errors here. First, the climateprediction.net simulations aren’t predictions; and secondly, the statement assumes, contrary to most expectations, that sulphate aerosol emissions will increase in the 21st century.
In fact, nearly all of scenarios project that emissions of sulphur oxides will DECREASE in this century. For the four SRES markers, the projected decreases between 2000 and 2100 are: A1, 60%; A2, 13%; B1, 84%; and B2, 31%. For the two illustrative A1 scenarios, the projected decreases are: A1FI, 42%; and A1T, 71%. These decreases lead to significant positive (negative of a negative) forcing in the 21st century, and concomitant WARMING in the IPCC scenarios.
There are huge uncertainties in relation to aerosols, both in respect of the trends in emissions in recent times and in the effects on climate. Having already tripped myself up once in a confusion of forcings and temperatures, I’ll avoid drawing my own conclusions and will simply draw a contrast between what Hansen et al (2005) say and what is posted on the climateprediction.net website:
(a) Hansen et al (2005) assess the total 1880-2003 negative aerosol forcing, including the indirect effect, as equivalent to more than one-half of the effective forcing for the total of the well-mixed GHGs (CO2, CH4, N2O and the CFCs) over the same period (pps. 5, 7). But they say that ‘empirical data for checking model-based temporal changes of tropospheric aerosol amount.. are meager’, and ‘Our largely subjective estimate of the uncertainty in the net aerosol forcing is at least 50%’ (p. 7).
(b) Hansen et al (2005) also say that: ‘Observed global warming, as well as the global warming in the model driven by all forcings, has been nearly constant at almost 0.15°C/decade over the past 3-4 decades, except for temporary interruptions by large volcanoes. This high warming rate has been maintained in the recent decade despite a slowdown in the growth rate of climate forcing by well-mixed GHGs.. The warming rate in the model is maintained because, BY ASSUMPTION, TROPOSPHERIC AEROSOLS STOP INCREASING IN 1990.. The ASSUMPTION that global aerosol amount approximately levelled off after 1990 IS UNCERTAIN, because adequate aerosol observations are not available.. An implicit well-known conclusion is that future global warming may depend substantially on how the global aerosol amount continues to evolve, as well as on the GHG growth rate’ (EMPHASES added).
(c) By (apparent) contrast, the simulations on the climateprediction.net website for the average of 66 models that had ‘made it to at least 2005’, to which I referred in an earlier post on this thread, show ‘overheating’ as a result of the non-inclusion of sulphate aerosols on what seems to be a much larger scale than implied in Hansen et al. Moreover, the widening of the gap between the ‘without aerosols’ temperature simulations and observations appears to be at least as great between 1990 and 2005 as in the decades preceding 1990.
To me, this suggests that the climateprediction.net estimates of aerosol emissions do NOT level off after 1990, and that a continuing growth in such emissions is reflected in the ‘Sulphur Cycle Experiment’ map.
In your initial post, Coby, you told me that for my follow up questions I might find Chapter 12 of the TAR informative, and provided a link. I am in fact quite familiar with what is said in that Chapter on regional climate projections, for which Australia’s John Zillman was Review Editor. If this post was not already overlong, I’d draw on the conclusions of that chapter, and on some of John’s subsequent statements on this subject, to reinforce some of the points made above.
I’m sorry if you think that I am again raising ‘the most elementary complication’ and that, because I ‘have never done anything but the most cursory research, [I] assume no one ever before has ever thought of it’ – and am ‘triumphantly pronounc[ing] climate science as an ignorant religion.’
You alleged, referring to me, ‘that it was completely understandable that most climate scientists are not interested in responding to people who come with pre-formed conclusions that imply they are stupid or frauds’, and said that it is ‘a credit to people like Gavin that they understand the importance of this issue and therefore the importance of overlooking, for the most part, such egregious behaviour.’
And you concluded that ‘the primary consequence is that the real experts in general decide not to waste their time with prejudice matched by ignorance and it falls on non-experts like [you] to spend the time.’
Let me assure you that I don’t come to climate change science issues with pre-formed conclusions that imply that climate change scientists are stupid or frauds. I tried for a long time to stick to my knitting, but found that I was being criticised (notably on this blog) for pleading ignorance of climate change science.
I’ve therefore decided that we should all be prepared to ask some dumb questions in areas outside our area of specialisation and that’s what I’m doing. I’ll have some separate questions to address to Gavin, but in the meantime I hope that you are able to take some time to respond to the issues raised above.
Ian Castles
Jennifer Marohasy BSc PhD is a critical thinker with expertise in the scientific method.
Sorry, it’s Chapter 10 of the TAR that Coby believed that I’d find informative, not Chapter 12. Chapter 10 deals with regional climate information, Chapter 12 (for which John Zillman was also Review Editor) deals with ‘Detection of climate change and attribution of causes’, and Chapter 13 (for which John Z was yet again Review Editor) deals with climate scenario development.
In your previous blog entry Ian Castles (and I am paraphrasing you somewhat) you stated that skeptics are exonerated from producing an alternative theory for the warming over the past century ONLY if the prevailing consensus explanation does in fact offer a “satisfactory” explanation of what has happened.
You then went on to address what you felt was a “problem” in the predictions of the models which amounted to a “heresy” (pretty strong stuff). Your basic contention was that if “the prevailing explanation of warming is correct, the greater increase in temperature should be in the southern hemisphere” and since this was not observed then the present consensus view was unsatisfactory (thereby implicitly exonerating skeptics from having to do the hard work of coming up with a reasonable alternative explanation).
Now I presume from your initial comments in this latest posting that you have reassessed the strength of your alleged initial discrepancy concerning the difference in the warming trends in the northern and southern hemispheres. I am glad that you seem to accept also, though not stating it explicitly, that your original “discrepancy” seemed to arise from a basic error in the understanding of climate physics and that the forcefulness of your proposition (which seemed to me to amount to no less than that the current theories were so inadequate that “the causes of climate change are not yet adequately understood (and may never be).” (To me that was again very strong stuff and implied the complete failure of the current generation of climate models and of those such as Gavin Schmidt who labour on these beasts with such apparently derisory results. I was not surprised that your post drew a sharp response from the latter.)
In your current post you seem to have retreated to focus much more upon lower level discrepancies between models and reality. My point now is, at what level is a discrepancy between models and reality “insignificant” to you? When can long-suffering climate scientists finally say to the deniers that their models are adequate for practical purposes and that it is now up to them to “put up or shut up”? When are we justified in taking the first tentative steps (yes, with their economic implications, which implications we will again have to foresee with economic models, which I would suggest are even more “unsatisfactory”) to address the issue of climate change?
Ian K, During the past few years, James Hansen has criticised “the IPCC predilection for exaggerated growth rates of population, energy intensity and pollution”. He has been dismissive of the IPCC’s high emissions scenarios, to the point of commenting on a Figure that “The IPCC scenarios that extend far off-scale (high) are impractical to show in entirety with a linear scale, but they do not need to be shown as they are unrealistic”.
Hansen has listed reasons for believing that the IPCC scenarios are “unduly pessimistic”, and questioned whether they were necessary or even plausible. He has argued that “global warming can be slowed, and stopped, with practical actions that yield a cleaner healthier atmosphere”, and that the focus of international action, at least in the short run, should be the reduction of air pollution through concerted efforts to develop and share clean technologies.
Dr. Hansen has explicitly rejected the claim that the actions needed to avoid a “gloom and doom scenario” were “economically wrenching”, and argued that, on the contrary, they “made sense independent of global warming”.
Gavin has now drawn to our notice a paper by Hansen and more than 40 co-authors (including Gavin himself) in which these attitudes are completely reversed. The scenarios that Dr. Hansen said didn’t need to be shown ‘as they are unrealistic’ have suddenly become ‘business-as-usual’ scenarios (notwithstanding that the IPCC specifically said that they were not).
Dr. Hansen was of course right the first time. The assumptions underlying what he and his co-authors are now calling BAU are highly unrealistic, as has been acknowledged by the authors themselves.
You ask ‘when can long-suffering climate scientists finally say to the deniers that their models are adequate for practical purposes.. ?’ I’ll answer in the words of Garth Paltridge, former Chief Research Scientist of the CSIRO Division of Atmospheric Research, in his presentation to the APEC Study Centre Conference in April 2005:
‘.. The real problem is that the climate models are now so complex that no scientist outside the closed shop of the numerical modelling community can ever really hope to assess whether or not the physical representations within them are acceptable. The normal and necessary process of scientific criticism cannot take place. The very least than needs to be done if the IPCC is to maintain its credibility is to insist that all models used in IPCC assessments must calculate and publish the implicit feedback factors built into their calculations.’
There is also, of course, a need for climate scientists to base their assessments on more realistic emissions scenarios.
I must interrupt here that “observed warmings” etc of 0.15C are NOT observations but statistics.
Current instrumental technology cannot produce such accuracy. However those readings are precisely inaccurate, and those here contemplating critical retorts, think about it.
I just came back from an AIG conference and JORC auditing procedures etc would cause some posters here some discomfort.
‘.. The real problem is that the climate models are now so complex that no scientist outside the closed shop of the numerical modelling community can ever really hope to assess whether or not the physical representations within them are acceptable. The normal and necessary process of scientific criticism cannot take place. The very least than needs to be done if the IPCC is to maintain its credibility is to insist that all models used in IPCC assessments must calculate and publish the implicit feedback factors built into their calculations.’
Computer modelling cannot deal with non-linear funtions. As such, ALL climate models are incomplete, and thus only reflect the bias of the programmers. In other words, GI GO.
Copying Coby response from the previous thread…
Ian, I have to apologize for my initial tone with you, and confess I mistook most of the post as Jennifer’s words rather than it being all yours. You do strike me as more serious minded about the science in this issue than she, I was therefore a bit too abrasive. Who knows, maybe I don’t give Jennifer enough credit, time will tell.
I can’t really add anything to your excerpts from Hansen’s papers, I hope the point is not to prove there are in fact uncertainties because that is a given. As for regional failures of modeled climate, the question of its significance is necessarily a judgement call. After all, to quote someone (?) “all models are wrong, but some are useful”. The failings you point out might lead one to say things are turning out worse than we would have expected, getting more warming where people live and less in the antarctic where it doesn’t matter.
You quoted climateprediction.net:
“a PREDICTION of the climate of the 21st century
needs to contain the effects of sulphate aerosol
otherwise the warming trend may be OVERestimated”
I agree they should use the word projected. I don’t see how this implies aerosols will increase, though have no knowledge of what they do assume. I really can’t say for sure if your inference from roughly equal widening of the non-aerosol vs observations gap in the 1990’s really means they are not using an aerosol scenario that levels off in the nineties. The effects may not be so linear.
I do appreciate the obvious time and effort that you are investing here. There are no dumb questions, only dumb answers. I will do my best to pass on what I know and provide resources where I can for what I don’t.
Coby
It was me that called my questions ‘dumb’, Coby. I said we should all be prepared to ask some dumb questions in areas outside our area of specialisation, as I had done.
You cited a Wikipedia entry presenting a table derived from Meehl et al, in support of your claim that there is no discrepancy between modelled and observed changes in temperatures. It showed that Meehl et al’s modelled GLOBALLY AVERAGED temperature change corresponded closely with the observed changes for three periods in the twentieth century ending in 1994. In all of the time periods shown the difference between the observed and modelled results was less than 0.1 deg. C.
As Gavin pointed out, my statement about the hemispheric differences was easily checkable. He directed me to Figure 18 of Hansen et al 2005. I checked and found that my statement about the hemispheric changes was incorrect but that the point that it was intended to illustrate – that there was no close correspondence between observed and modelled results at the zonal level – was completely valid.
For the 1979-2003 period the difference between modelled and observed results was at least 0.3 deg. C over most of the northern hemisphere, and several times greater than that in the high southern latitudes. These differences are much greater than those that seemed to arise from a comparison of the two hemispheres and amply confirm, to my mind, that ‘the causes of climate change are not yet adequately understood.’
It is not the case, as Ian K suggests, that I am asserting that there is any basic ‘error’ in the understanding of climate physics. My difference with the so-called consensus view is about the implications that arise from what is agreed to be not understood. This is from the conclusions of a recent Australian Academy of Science workshop):
‘For reliable climate prediction we do not know enough of the fundamentals of how greenhouse gases and aerosol accumulation influences atmospheric humidity, cloud formation and cloud properties, solar radiation, whether greenhouse warming involves an intensification or a dampening of the hydrological cycle, and how ocean evaporation responds to greenhouse warming’ (Gifford R et al, 2005, ‘Preface’, in ‘Pan evaporation: an example of the detection and attribution of trends in climate variables’).
And let me quote again from Garth Paltridge’s presentation last year:
‘… it is fairly easy to calculate the likely rise of global average temperature for the purely theoretical situation where atmospherid carbon dioxide is doubled but nothing else about the atmosphere is allowed to change. The increase is about 1.2 deg. C, and it would take a couple of hundred years to complete the change.
‘However, in the real world, all sorts of other atmospheric and oceanic processes that depend on surface temperature are happening. There are so-called feedback processes, and many of them amplify or reduce the original change of 1.2 degrees caused directly by the carbon dioxide.
‘Their effects need to be added up to give an overall value for the total ‘feedback factor’ (F) to calaulate the predicted change in temperature. Thus the calculated temperature change may be greater or less than 1.2 degrees – this depending entirely on the value of F and on whether it is positive or negative.’
As I’ve already noted, Garth Paltridge went on to argue that ‘if the IPCC is to maintain its credibility [it shoud] insist that all models used in IPCC assessments must calculate and publish the implicit feedback built into their calculations.’ It will be interesting to see whether or not this is done in AR4.
Thanks Coby for the comments posted above. As will be obvious, I had not seen them when I posted my last comment.
‘the causes of climate change are not yet adequately understood.’
The problem with this is that it is too ambiguous. We could argue endlessly and never know what the other means by it. Usually when I see someone say that I think they are saying we don’t know enough yet so it is unwise to take mitigating action. I argue with that. I don’t know if you mean that, or if you didn’t here but do think it anyway. I do know that in the context of this discussion the only way I can agree with it is if it means this: the causes of regional responses to global climate forcings are not understood well enough predict what local climate conditions around the globe will be in the future.
As you see, we need to know exactly what climate change means and for what purpose adequacy is sought.
Coby,
Change is a fact of life. Living things adapt to changing physical conditions, and thus change their physical appearance accordingly. I think it is called Evolution.
We need to adapt to climate change, or die.
Climate alarmists seem to wish to die, since they don’t want to adapt but rather want to change everything else, apart themselves.
Only the dead can’t adapt.
Hi Loius,
Seems to me you climate sceptics are the ones afraid of change, anything to avoid changing your energy consumption habits. Funny thing is, you will be forced to change one way or the other as oil is a non-renewable.
Yes, climate has varied in the past and it has varied for many different reasons, some better understood than others. The present day climate change is very well understood and is different. Simply noting that something happened before, change is not unheard of and it happens without humans does not in any logical way show that humans are not causing it today or that it is not a problem.
For example, we see in ice core records from Antarctica and Greenland that the world cycled in and out of glacial periods over 120Kyr cycles. The cause for that climate cycle’s timing is fairly well understood to be the results of changes in the orbit of the Earth, though the mechanism behind the resulting response has not been conclusively established. These orbital cycles are regular and predictable and they are definately not the cause of today’s warming. The other important difference between the glacial-interglacial cycles and today is the rapidity of the current change. The rate of warming is on the order of 10 times faster today than seen in the ice cores.
Such rapid warming on a global scale is in fact very rare in the geological record, and while it may not be unprecedented, there is very strong evidence that whenever such a change has happened, whatever the cause, it was a catastrophic event for the biosphere.
You will find a very sobering historical analogue for what humans are now doing in the Paleocen-Eocene Thermal Maximum event 55 million years ago. Catastrophic climate change is in fact the prime suspect in many of the great extinction events seen in the fossil record.
It’s very true that eventually life recovered and the planet each time has seen even greater biodiversity emerge, but the time scale is far longer that the entire history of the human species, sometimes many millions of years. Evolution is a slow process, extinction is immediate. With a rapid climate change such as we are causing, the migration of many existing ecosystems is not going to be possible so they will instead die. Without a doubt this will mean new niches and new opportunities for evolutionary magic but not until some point so far into the future as to effectively be never.
I find such a cavalier attitude as your comment exhibits to be the height of irresponsibility and a dizzying depth of ignorance.
Ian Castle, I think that Garth Paltridge (could equally have said):
“it is fairly easy to calculate the likely rise of global average temperature for the purely theoretical situation where atmospheric carbon dioxide is doubled (and “static” water vapour feedback is added also, ie without convection, cloud formation, etc) but nothing else about the atmosphere is allowed to change. The increase is (no longer) about 1.2 deg. C, (but an alarmingly large number) and it would take (now considerably less than) a couple of hundred years (to get very seriously warmer, although not yet complete!) to complete the change.
‘However, in the real world, all sorts of other atmospheric and oceanic processes that depend on surface temperature are happening. There are so-called (further) feedback processes, and many of them amplify or reduce the original change of 1.2 degrees (plus an alarmingly large number) caused directly by the carbon dioxide (and “static” water vapour feedback).
Their effects need to be added up to give an overall value for the total ‘feedback factor’ (F) to calculate the predicted change in temperature. Thus the calculated temperature change may be greater or less than 1.2 degrees (plus an alarmingly large number!) – this depending entirely on the value of F and on whether it is positive or negative .”
In short, my view of the models is that they are a necessarily complicated attempt to bring us out of a (theoretical) hothouse by using the dynamic processes of convection and cloud formation, etc. Furthermore as we have added a stew of extra forcings mankind has forced the models to be very complicated indeed. It is now a matter of exactly how precise the standards are to be that we set the modelers. As this post is long enough and I am still ploughing through the Hansen paper I will leave it there.
As for Louis all I can say is: lets hope that within many hundreds of years your descendents will have had your “delusion” gene mutated into a “realism” gene. Fortunately the rest of us have societal action to augment evolution.
Thanks Ian K. I won’t presume to interpret Garth Paltridge’s analysis, which for those who are interested is available at http:www.climatechangeissues.com/files/PDF/conf05paltridge.pdf . I do however draw attention to his comment that:
‘If one has a model with parameters tuned to give an already large position feedback (and therefore large temperature rise) it takes only a very small change in tuning (or the addition of another positive feedback process not yet incorporated in the model) to give enormously frightening figures of potential temperature change. A useful characteristic that!’
Do you have any comment on Professor Paltridge’s proposal that ‘all models used in IPCC assessments must calculate and publish the implicit feedback factors built into their calculations’, so that ‘climate scientists of the outside world will have some understandable physics on which their intuition can work, and perhaps also a design of real-world experiment and observations so as to improve the modeller’s arbitrary selection of tuneable parameters’?
On the question of ‘exactly how precise the standards are to be that we set the modellers’, from my perspective I’d want to see the modellers brief themselves more adequately on the assumptions underlying the input scenarios they are using.
For example, Hansen et al say that ‘Regional climate change also yields a clear distinction between BAU scenarios with global warming ~3°C and an alternative that keeps global warming <~1°C’, and go on to argue that ‘Ecosystems, wildlife, and humans thus would be subjected in the BAU scenarios to conditions far outside their local range of experience.’
It is surprising that none of these 46 US climate scientists knew that the IPCC SRES specifically states that ‘There is no business-as-usual scenario’ (see http://www.grida.no/climate/ipcc/emission/012.htm , third paragraph). Hansen et al also say that only the Alternative scenario of Hansen and Sato (2004) keeps warming below 1°C, and that ‘Global warming in ALL other scenarios already exceeds 1°C by 2100, ranging from 1.1°C in scenario B1 to 2.7°C in scenario A2’ (EMPHASIS added). But projected cumulative CO2 emissions under the B1T MESSAGE scenario are ~25% less during this century than under the B1 marker scenario, and the temperature increase under this scenario would therefore, almost certainly, be less than <1°C.
When I pointed this out on the RealClimate site, I was told that the IPCC low emissions scenarios could only be realised if stringent climate-change related policies were introduced. This contradicts a specific statement in the SRES study, which was required under its Terms of Reference to ensure that NONE of the scenarios included any future policies that explicitly address additional climate change initiatives: see http://www.grida.no/climate/ipcc/emission/010.htm , second paragraph.
I find it interesting that I, a sole unfunded researcher, have been taken to task for making an incorrect assumption in a post on this blog. I acknowledged the error after having my attention drawn to a Figure in a paper by Hansen et al, the link to which was provided by one of the co-authors of the paper, Dr. Gavin Schmidt. In my response I pointed to incorrect assumptions that had been made in this major paper by 46 scientists from 12 research institutions – for example, I noted that some IPCC scenarios were repeatedly but wrongly described as ‘business-as-usual’ (see link above, but I made the point some days ago). As the paper has been submitted to Geophysical Research Letters and some of the analyses therein have been provided to the IPCC, it seems particularly important that this issue be addressed.
————————————————
[This comment got blocked by my spam filter and was not uploaded until 10am Saturday. Sorry. I’ve been traveling. Jennifer]
Ian K,
Your last paragraph is a nonesense since it is not based on fact.
It is essentially an “ad hominem” of the classical type.
Your concluding sentence is illuminating – you think brainwashing will alter evolution?
Who washed your pea ?
Louis, as I see the situation with regard to climate change there are three lots of people.
One lot are prone to getting alarmed (perhaps too easily). These people would, if we were gnu be most alert to the presence of the lion. I would say that their tendency to alarmism is pretty useful in an evolutionary sense. In the context of climate change they are so alarmed by the prospect of climate change that they want to react to it now. They are willing to adapt by changing their lifestyle as necessary and by convincing governments to take action also. How you can denigrate a tendency to alarmism is beyond me. As a survival skill it is clearly of great evolutionary benefit to the herd, although it comes at a cost to the individual.
A second group (the great majority in the gnu herd, are too busy eating grass, and doing what gnus do, to notice the approach of any lion. Of course, perhaps subconsciously, they know that they don’t have to worry because the alarmists will warn them in time. Of course perhaps also they know they are also the faster runners and don’t need to worry so much. These people are probably happier than the alarmists, but are they more evolved? I don’t know.)
Then there is the third group (which as far as I am aware don’t exist in the gnu herd) who keep saying to the alarmists and to any one else that will listen; “calm down there is no lion there, it is only a mirage”. Perhaps this group is useful too, if they are correct in their assessment. These soothsayers however had better be pretty certain of their ground if they are to be of any evolutionary benefit to the herd.
By the way my joke about changing genes (I am sorry if you took it the wrong way) is not one to brainwashing. It is a commentary on your seeming reliance on future evolutionary change to get us out of rapid climate change. I believe that evolution is far too slow a mechanism for us to adapt to this problem in an evolutionary sense. To cope with climate change now, we have to use our present endowments which are the result of meeting past evolutionary challenges. One of the most important of these is mankind’s highly developed powers of reasoning. Reasoning, which I am sure you have in abundance, is however of little benefit in an evolutionary sense if it is not backed up by an alertness to danger. Only then will it allow us to react, even to obscure and distant threats to the survival of the species.
Sorry about the preachy parable. I don’t know what came over me! Oh no, I have just read your last two lines to Coby. I should have just reciprocated your abuse! Anyhow what does: “who washed your pea” mean?
Please note comment by Ian C in response to Ian K, only just posted, but inserted higher in this thread … as it had been submitted yesterday, but got stuck in the spam filter.
Ian Castles,
Regarding Professor Paltridge’s proposal that “all models used in IPCC assessments must calculate and publish the implicit feedback factors built into their calculations”, so that “climate scientists of the outside world will have some understandable physics…” I think this is one of those red herrings, like “Mann et al should make all their data available”. AFAIU, all the major climate models are fully documented and this documentation as well as, quite often, the actual source code, is available online. When papers are published and model runs are involved, the models and scenarios and parameterizations are explicit.
If you think there is a model result where this is not the case let’s have a look at it and see if we can’t get all the details.
About the “business as usual” complaint I think you are making alot out of a really minor quibble. In the reference paper it says “… the IPCC A2 and A1B scenarios, commonly called “business-as-usual” (BAU) scenarios…”
So it is perfectly explicit about which scenarios they are talking about, so what if it is commonly, if incorrectly, reffered to as BAU? The IPCC is more rigorous, as they should be, while the climate scientists are apparently a little sloppy in their economic thinking. This is why they are using rather than designing the scenarios. A rose by any other name…
As for scenarios in general, I think I have acknowledged before that I don’t know that much about them, though I will definately try to pay more attention as the AR4 comes out. Is there no scenario that you think is plausible, and if not what do you think emissions in this century will look like?
Thanks.
Coby, I’m not going to go over this all over again. I’ve said repeatedly on this blog, and also at Real Climate, that I believe that the IPCC B1T MESSAGE scenario is plausible, and that would mean emissions in this century that would look like the GISS Alternative scenario.
Professor Paltridge’s proposal was made only a year ago. He is a distinguished scientist whose paper ‘Climate and thermodynamic systems of maximum dissipation’ was published in Nature, 14 June 1979, pps. 629-630. Many of his papers have been published in peer-reviewed journals.
Garth Paltridge has participated actively in several workshops on aspects of climate change science which I have attended in recent years. In his entry in the Fellowship list of the Australian Academy of Science (to which he was elected in 1980), Professor Paltridge gives his areas of specialisation as ‘Fluid earth sciences, climate and Antarctica.’
You say that ‘All the major climate models are fully documented and this documentation as well as, quite often, the actual source code, is available online. When papers are published and model runs are involved, the models and scenarios and parameterizations are explicit.’
Do you think that Professor Paltridge, who was Director of the Antarctic Cooperative Research Centre and the Institute for Antarctic and Southern Ocean Studies at the University of Tasmania for 12 years, is unaware of the documentation that is published on climate models? Why do you suppose that he made the suggestion?
Ian,
I meant no harassment with my question about a plausible scenario, and apologize if you have said that before somewhere I should have seen. I will use that as a starting point for learning about the scenarios, thanks.
1979 was a long time ago…but I thought we didn’t wave credentials around, surely that is a losing strategy for anyone who has reservations about the IPCC WG1 reports. Regardless of Professor Paltridge’s qualifications:
Is he “unaware of the documentation that is published on climate models?”
It seems very likely that he is unfamiliar with it based on his comments as presented by you.
“Why do you suppose that he made the suggestion?”
I don’t care to comment on the motivation of anyone I know so little about.
But might I turn the tables and enquire as to why you put such faith in one man’s public doubting, especially given the overwhelming maority of at least equally qualified individuals who have satisfied themselves that things are not so impenetrable?
Coby, The implicit feedback factor F and its components, as shown in Garth Paltridge’s illustrative example, must be either
(1) Impossible to calculate;
(2) Capable in principle of being calculated, but not in fact calculated by the modellers (hence ‘implicit’);
(3) Calculated by the modellers and reported in the modelling documentation, but not in such a format that enables non-modellers (who may nonetheless be highly-qualified in one or more of the disciplines relevant to climate change science) to make their contribution to the reduction of uncertainty (which, bear in mind, is still very large even in the formulation in the IPCC draft report); or
(4) Calculated by the modellers and reported in a standard format, enabling both dumb economists like me and able atmospheric scientists such as Professor Paltridge to observe the proximate reasons why the climate sensitivity of model A is twice as great as that of model B (I think AR4 uses 14 different models).
If this comparison has been done in AR4, well and good. If it has not been, is it because of reason (1), (2) or (3) above?
I am not a modeller myself, but that did not prevent me observing that the IPCC models were conceptually unsound, and then to find that leading theorists agreed with me when they got round to looking at it.
You are saying, in effect, that there is no analogy to this in the field of climate science. Perhaps that is true, but as a first step towards allowing me to decide that to my own satisfaction, could I know your answer to the question raised above?
Ian K
So what, precisely, is your point, then?
Coby,
could you be a tad more concise in your,er, statement or was it a question. Hard to work out which was which. Idiotic of me to presume, to thay the leatht.
Ian,
There are different kinds of feedbacks of course, and this is all crossing over the limits of my familiarity with the details. But, for example, with water vapor (as opposed to clouds) the feedback is going to be a result, not a set parameter, and as such the way to understand it is to delve into the details of the physics and how it is represented internally. It will depend on calculations of radiative absorbtions and emissions, temperature lapse rates (also a result not a parameter), air pressure gradients, all kinds of things. The way to put a number to what this feedback effect is is to do an experiment specifically designed to find out, generally by doing two seperate runs of the same scenario (ie 2x CO2 run to equilibrium) one with H2O concentrations held constant and then quantify the different results. Going just by memory of past discussions I believe the effect is to triple, or thereabouts, the warming from CO2 alone. The key point being that H2O feedback is an end result of the minutia of the model that has been constructed based on theoretical physics and unless you do an experiment such as I outlined you can not really look at a model run and say X warming is CO2, Y is H2O etc etc.
More generally, your question can not be answered as it is written as it based on some misunderstandings IMO and is ambiguous as to what feedbacks you are talking about. If you changed it to be about parameterization, which seems to be getting conflated with feedback, then the answer would probably be 3, though you might have to follow a chain of references to get the original numbers and their genesis.
As an example of what is publically available, here is the home page of the GISS model
http://www.giss.nasa.gov/tools/modelE/
From there you can get the extensive documentation and even all the source code. There is a reference manual, instructions on system requirements, a HowTo and a FAQ, data from past runs and references to the papers published on its development.
I appreciate the desire of laypeople to know how it works and to check a few things themselves, I share it. But the fact is these things represent hundreds of man-years of sophisticated effort so it is just not practical for most of us and the goals of transparencey and simplicity are in conflict. It is impossible to meet both in the same document.
Here is an introduction to AOGCM’s by William Connolley that may shed some light on how they work:
http://mustelid.blogspot.com/2005/11/how-coupled-ao-gcms-work.html
You said:
“I am not a modeller myself, but that did not prevent me observing that the IPCC models were conceptually unsound, and then to find that leading theorists agreed with me when they got round to looking at it.”
Eh? Sorry, I must have missed that. What is conceptually unsound in the IPCC models and which leading theorists agree with you?
Hi Coby, I meant to say that I’d observed that the IPCC SRES models – the ones that produced the emissions scenarios – were conceptually unsound. I haven’t and wouldn’t reflect on the CLIMATE models.
Thanks for your detailed response to my question. Your understanding as stated in your first para. is in line with mine. I hadn’t interpreted Prof. Paltridge as suggesting that you could ‘look at a model and say X warming is CO2, Y is H2O, etc.’
I thought that he was recognising that some ‘feedbacks’ are results, not parameters, by using the word ‘implicit.’ Is your ‘paremetization’ of 3 the same as the 3.47 derivable from his example?
I’m feeling my way a bit, but your distinction between results and parameters seems similar to the econometricians’ distinction between exogenous and endogenous variables. That doesn’t mean that there’s always agreement about whether a particular variable is one or the other, and there will be argument about whether particular multipliers (‘feedbacks’, more or less) could in principle be negative.
So far as climate models are concerned, I don’t have any interest in checking a few things for myself (I couldn’t), but I took Garth Paltridge to be suggesting some form of tabulation that would quantify and expose for scrutiny the influences that lead to climate sensitivity being higher in some models than others.
You misunderstood the reason why I mentioned Prof. Paltridge’s qualifications, and I only cited the 1979 paper to show that he’d been in the climate science business for a long time. Of course he’s had many more papers in learned journals since – I notice that there’s one in the Journal of the Royal Meteorological Society in 2001.
My reason for giving these details was to show why I wouldn’t expect that a climate scientist of Prof. Paltridge’s calibre would make a suggestion that was incapable of implementation. It’s not a matter of trusting him more than scientists no less qualified than he is, because I haven’t yet heard it said that other scientists believe that his proposal couldn’t work.
Having just read the letter from Jan Veizer that’s been posted today on another thread, I found myself asking whether it is possible (leaving aside the question of the validity of the Veizer/Shaviv position) to devise some form of schematic presentation that brings out why different models give different answers. I infer from your answer that you believe that this is not possible.
Thanks Ian Castle for your reference to the article by Garth Partridge which helps flesh out his views. Perhaps we could get some Australian expert to comment on his critique of the models, Jennifer? Could we get someone from RealClimate to comment perhaps, Coby?
Getting back to what I think was your original criticism, Ian.Your main point was, it seems to me, that you questioned the validity of the models in the light of their inaccuracy in reflecting aggregate values of a particular variable, albeit a very relevant one to global warming, average temperature.
With respect, this misses the point of global climate models. They are, in essence, (although aggregates and averages can be derived from them), designed to reveal the ANTITHESIS of global, hemispheric or even regional aggregates. The role of the models is a dynamic one. Judging models, is a much more exacting, and no doubt somewhat subjective, task than looking at aggregates. They should give a holistic representation of the climate at each point on the globe as these climates develop through time. So they must represent not only temperature but how temperature varies through the seasons, and likewise for rainfall patterns, wind speeds, the extent of ice formation or melting, etc, etc. These climate representations need only be sufficiently accurate so that we can assess the impact of climate change on people and living systems at particular places. For instance a model which accurately predicts a change in a particular region from moderate to more desert-like rainfall patterns may be valuable even if it is somewhat inaccurate in its representation of temperature changes.
This is just from the top of my head, I hope it is not teaching you how to suck eggs but I wanted to get it out, having not had time to read Coby’s references yet.
Hi Ian K,
I have posted a comment at RC inviting some participation here. Here is the comment link http://www.realclimate.org/index.php/archives/2006/05/ipcc-draft-no-comment/#comment-13089 in case it gets response in line.
‘the causes of climate change are not yet adequately understood.’
This is a akin to a call to imply nothing;do nothing. It dismisses all we know in favor of knowing nothing.
Thanks Ian K. The Annan & Hargreaves paper recently discussed at RC began by referring to climate sensitivity as being SUBJECTIVELY estimated as being in the range 1.5-4.5 C, and I recall that another recent paper referred to 8 of the 14 GCMs used by IPCC as having a climate sensitivity in the range 2.6-4.0C. There’s also been discussion about whether simple climate models can provide a ‘better’ representation of average global temperature than GCMs.
I didn’t question the VALIDITY of GCMs. I observe that the increase in average global temperature derivable from different GCMs from a given forcing patterns differs. And, like it or not, the headline numbers emerging from the model inputs are presented (mainly) as global aggregates. I would have thought that there is a way of characterising different models that looks to the proximate reasons for the differences in sensitivity between one and another.
I take the point in your main substntive paragraph, Ian K, but again I see a parallel in econometric models. This is from Warwick McKibbin’s paper presented to the Joint Academies Conference on Climate Change in April 1997, eight months before Kyoto:
‘Economic models have a valuable role to play in the policy debate on greenhouse gas abatement not because of the numbers they generate per se but because they can help understand the potential weaknesses and costs of alternative policies. There are also a range of economic models available, each of which can provide insights particular to the type of issue the model was designed for. There will never be an all purpose encompassing model that can answer every question.
‘For any given policy issue a range of models should be used to give a range of insights and some measure of the range of possible quantitative outcomes for any given policy change.’
Some global models may be ‘better’ at capturing international trade effects, others may be ‘better’ at capturing interindustry transactions or financial flows. Some (most) are focusing on the short-term, others on medium and (now) long-term developments. This doesn’t preclude the provision of standardised comparisons of averages or aggregates.
Coby, I forgot to respond to your statement that ‘it is perfectly explicit about which scenarios [Hansen et al] are talking about, so what if it is commonly, if incorrectly, referred to as BAU?’
The scenarios that Hansen et al are talking about are designated ‘A2′,’A1B’ and ‘A1FI’. These designations may be explicit, but they’re meaningless. Hansen et al attempt to give them meaning by describing them as ‘Business-as-Usual’.
This is not just an incorrect FORMAL description: it is a patently false description of these scenarios.
A2 assumes a global population of 15 billion at the end of the century, which exceeds the assessment of the consensus of demographers by a factor of 2 or thereabouts. To the best of my knowledge, no recognised demographer would venture an estimate that remotely approaches 15 billion as their ‘business-as-usual’ projection.
It is equally absurd to characterise A1B as ‘business-as-usual’. Have any of the scientists who so describe it actually looked at the detailed underlying assumptions? For example, the A1B scenario projecte that the SRES ASIA region (which contained 53% of the world’s population in 1990 and a projected 41% of the population in 2100), will increase its consumption of electricity per head over the projection period by a factor of 84.
As a proportion of electricity use per head in the industrialised OECD90 region, electricity use in ASIA is projected to increase from 5.4% in 1990 to 95.4% in 2100.
Over the same period the average use of electricity per head in the OECD90 regions is projected to increase by a factor of 4.73.
Thus in 2100 the average use of electricity per head in the whole of the ASIA region (including China, India, Pakistan, Bangladesh, Indonesia, the Philippines, Korea and a number of smaller countries) is projected to reach 4.5 times (95.4% of 4.73) the level that had been achieved in the rich OECD countries in 1990.
In the REF region (former Soviet Union and Eastern Europe), electricity use per head in 2100 is projected to be even higher – nearly twice as great as the level that would be reached in the ASIA and OECD90 regions by that time.
It would be reasonable to assume that the growth in productivity and average incomes that made these huge increases in electricity use possible would be accompanied by a considerable increase in the efficiency with which electricity was used.
One cannot state categorically that such increases in electricity use are impossible. One can state categorically that, if it happened, it would not be ‘business-as-usual.’ Do any of the 46 co-authors of Hansen et al believe that that a scenario that makes these assumptions can fairly be described as BAU?
Ian,
At the moment I am highly inclined towards accepting your arguments that the so called “BAU” scenario is incorrectly named, even if it was explicitly tied to an actual SRES scenario in the paper under discussion. It seems a reasonable criticism that Hansen et al. should not refer to it this way and if they do so only because that is common in the climate science community then that reflects a sloppiness on the part of that community and it should be rectified.
Thanks for your patience.
Thanks Coby, I appreciate your comment. I haven’t seen the draft of AR4, but the Hansen et al paper states that the projections therein were done at the request of the IPCC so I can reasonably guess that the draft incorporates projections of concentrations, forcings and temperatures based on the A1B, A1FI and B1 scenarios.
All of these scenarios are based on ‘convergence’ assumptions, and it is important to know (although not easy for an intergovernmental body to say when its developing country members are in a large majority) that the empirical evidence supporting such assumptions is not strong.
The following is from the abstract of a recent paper by ANU economists Alison Stegman and Warwick McKibbin:
‘This paper explores the historical experience of a range of variables related to climate change projections with the goal of examining if there is any evidence historically of convergence. The focus of the paper is on per capita carbon emissions from fossil fuel use because this is the basis of many projections as well as a variety of policy proposals.. We find strong evidence that the wide variety of assumptions about “convergence” commonly used in emissions projections are not based on empirically observed phenomena.’
What if one does not have a model with parameters tuned to give an already large position feedback? how much time should one waste on Prof. Partridge?
The entire argument is a version of if pigs were horses cows would fly. Since pigs are not horses, it is completely reasonable not to waste energy avoiding falling cow pats. At least I would have thought so before reading this thread.