Promoted from the comments, the download of the BBC Radio 4 ‘Now Show’ (Mar 16) is available here (at least for now). Key bit starts at about 18min in, (the rest of the show is pretty funny too).
About Gavin
Reader Interactions
307 Responses to "A much more eloquent rebuttal of TGGWS"
Ray Ladburysays
Re 299: Nick, would you go with stochastic to describe climate? True chaotic behavior would be very difficult to diagnose in a system as complicated as climate, but I would not rule it out at least in some portions of the phase space.
Nick Gottssays
Re #301. Yes, I’d go with “stochastic”. Another term that might be useful, as it’s often applied to complex systems such as ecologies and economies is “path-dependent” – which also implies that where you end up can be very different depending on small initial differences, but doesn’t carry all the technical freight of “chaotic”, at least as far as I know.
If you want to nip this trope of the denialists in the bud, you should yourselves STOP using “causes unknown” if you are fairly sure that Milankovic was correct. Besides, it’s actually a point of agreement – both we and they think that it’s a sun-powered warming to begin with. And no one I know of is disputing that.
Ike Solemsays
Nick and Ray,
You have to be careful when using concepts like deterministic and chaotic especially in relation to global climate, which is the sum of many, many interacting ‘systems’. Take, for example, the widely studied issue of the onset of turbulence in fluid dynamics.
A river flowing slowly has little turbulence, but areas of rapids have high turbulence. Still, you can predict that the net mass transport will be downhill. This issue also crops up in climate models, for example, an equilibrium climate sensitivity estimate produced by a model (for a doubling of CO2) doesn’t depend on the timescale; if you double CO2 in one decade, you get the same T(eq) as if you double CO2 over 150 years, if I understand it correctly. However, the transient climate response in the two cases is clearly going to be different.
If you take the example of forcing water through a pipe, the same amount of water might flow through, but in one case (slow laminar flow) the behavior is quite different from the other (fast, turbulent flow). Also, experimental chaos is only studied in very simple systems like pipes; trying to extrapolate this to the global climate system is quite a stretch, though an increase in extreme weather events might be comparable to the onset of turbulence; extreme weather events are closely linked to the transient climate response, not so much to the equilibrium climate sensitivity. Still, any broad statements about ‘the behavior of the climate system in phase space’ are probably nonsensical.
However, for the denialist camp, what they like to say in their talking points is that “climate is chaotic, so it can’t be predicted, especially at the regional scale, and this means all the models are wrong, and thus we shouldn’t take any action to limit fossil fuel use”. This is the argument that Roger Pielke Sr used over at Climate Science (A New Quote on Regional Predictability, Feb 20 2007).
This is the standard contrarian approach, as far as I can tell – take a very complex scientific discussion, strip it down to simplistic nonsense that fits your agenda, and then repeat, repeat, repeat. It’s true that regional climate predictability is very iffy (which is why people whose idea of responding to climate change is buying property in Alaska and Iceland are behaving foolishly), certain factors can be constrained – for example, regional climate change in the American West will likely lead to more frequent drought (see the Persisent Drought in North America discussion by Richard Seager). Whether or not the drought in the African Sahel is related to anthropogenic climate change? Such issues are complex because they involve both local and global issues – similarly, drought in the Amazon can be linked to local deforestation as well as to global climate change.
The bottom line is that broad statements like ‘the climate system is deterministic’ or ‘the climate system is chaotic’ just don’t make much sense.
Hank Robertssays
Well (I’m just another reader, not a Contributor), seems to me they don’t need to go beyond the published science; the feedback once the process starts is known physics (temperature < -> CO2), and I gather is understood _better_ than whatever causes the nudges at various points when climate has changed in the past. I recall the fits to astronomical cycles are approximate, and the forcing involved is relatively small.
Take a rock rolling down a hill — something caused the rock to start rolling, various things change its velocity over time, and those may be poorly understood —- even though gravity per se is well understood.
Maybe the same story would be ‘continued on next rock’ — but maybe not, something else may kick that one off.
I think the science is focusing on forcings and feedbacks because, whatever happened when previous big changes ran their course, this one’s so much faster that the past is only useful, not determinative.
Have you read Ray Pierrehumbert’s paper on ‘Science Fiction Atmospheres’? (It’s linked at his website, his name’s listed under Contributors). It gave me a glimpse of how to think about what might be possible, and what we don’t know.
Bob Walkersays
There seems to be a lot of discussion about climate models and whether they are deterministic or chaotic.
Reading this thread it sounds as if on one hand the laboratory physics is very simple and fully understood while on the other no single climate model follows the climate with any great accuracy implying that once the physics gets out of the lab and into the atmosphere it is not really well understood.
Although computers do a wonderful job of simulating airplane wings this is surely different. In that case the model is being used to explore different wing types using the extremely well understood mechanisms that wings undergo. The fact that these models produce better wings shows that they are accurately modelling the real item.
In climate science it seems the computer models are actually being used as a/the primary tool to better understand the climate mechanisms themselves because these are still not well understood.
Regardless of how impressive they are a software developments and how much people talk them up, the only real check on their validity is their ability to actually model our climate and so far no single model has managed this.
Until a single model (as opposed to a mean of models) actually starts being accurate can we assume that climate science is really not as well understood as the consensus would like us to believe.
I have read than most of Antarctica is cooling something not predicted by Global Warming or the models � is this correct.
Ray Ladburysays
Re 306. Bob, First, where did you get the idea that the models are inaccurate. They have gotten the trends right for over 20 years–and trends are what climate is about. They have accurately predicted the effects of volcanic eruptions.
Second, modeling is a very reasonable way of exploring a system that cannot be realized or simulated in the laboratory. If there are uncertainties, you conduct simulations over the range of uncertainties and report the robust results and perhaps the range of results.
Third, the models are all giving pretty much the same trends. Yes, climate is complicated. Yes there are things we don’t understand. However, the field is sufficiently mature that it is unlikely we will see any major revolutions in our understanding that change the conclusions.
If there are climate scientists out there who really disagree with the hypothesis of anthropogenic causation, they sure aren’t publishing. Even Lindzen acknowledges it is happening. He just thinks we’ll be saved by some miraculous restorative mechanism that keeps the effects from getting outside our comfort range.
Re 299: Nick, would you go with stochastic to describe climate? True chaotic behavior would be very difficult to diagnose in a system as complicated as climate, but I would not rule it out at least in some portions of the phase space.
Re #301. Yes, I’d go with “stochastic”. Another term that might be useful, as it’s often applied to complex systems such as ecologies and economies is “path-dependent” – which also implies that where you end up can be very different depending on small initial differences, but doesn’t carry all the technical freight of “chaotic”, at least as far as I know.
If you want to nip this trope of the denialists in the bud, you should yourselves STOP using “causes unknown” if you are fairly sure that Milankovic was correct. Besides, it’s actually a point of agreement – both we and they think that it’s a sun-powered warming to begin with. And no one I know of is disputing that.
Nick and Ray,
You have to be careful when using concepts like deterministic and chaotic especially in relation to global climate, which is the sum of many, many interacting ‘systems’. Take, for example, the widely studied issue of the onset of turbulence in fluid dynamics.
A river flowing slowly has little turbulence, but areas of rapids have high turbulence. Still, you can predict that the net mass transport will be downhill. This issue also crops up in climate models, for example, an equilibrium climate sensitivity estimate produced by a model (for a doubling of CO2) doesn’t depend on the timescale; if you double CO2 in one decade, you get the same T(eq) as if you double CO2 over 150 years, if I understand it correctly. However, the transient climate response in the two cases is clearly going to be different.
If you take the example of forcing water through a pipe, the same amount of water might flow through, but in one case (slow laminar flow) the behavior is quite different from the other (fast, turbulent flow). Also, experimental chaos is only studied in very simple systems like pipes; trying to extrapolate this to the global climate system is quite a stretch, though an increase in extreme weather events might be comparable to the onset of turbulence; extreme weather events are closely linked to the transient climate response, not so much to the equilibrium climate sensitivity. Still, any broad statements about ‘the behavior of the climate system in phase space’ are probably nonsensical.
However, for the denialist camp, what they like to say in their talking points is that “climate is chaotic, so it can’t be predicted, especially at the regional scale, and this means all the models are wrong, and thus we shouldn’t take any action to limit fossil fuel use”. This is the argument that Roger Pielke Sr used over at Climate Science (A New Quote on Regional Predictability, Feb 20 2007).
This is the standard contrarian approach, as far as I can tell – take a very complex scientific discussion, strip it down to simplistic nonsense that fits your agenda, and then repeat, repeat, repeat. It’s true that regional climate predictability is very iffy (which is why people whose idea of responding to climate change is buying property in Alaska and Iceland are behaving foolishly), certain factors can be constrained – for example, regional climate change in the American West will likely lead to more frequent drought (see the Persisent Drought in North America discussion by Richard Seager). Whether or not the drought in the African Sahel is related to anthropogenic climate change? Such issues are complex because they involve both local and global issues – similarly, drought in the Amazon can be linked to local deforestation as well as to global climate change.
The bottom line is that broad statements like ‘the climate system is deterministic’ or ‘the climate system is chaotic’ just don’t make much sense.
Well (I’m just another reader, not a Contributor), seems to me they don’t need to go beyond the published science; the feedback once the process starts is known physics (temperature < -> CO2), and I gather is understood _better_ than whatever causes the nudges at various points when climate has changed in the past. I recall the fits to astronomical cycles are approximate, and the forcing involved is relatively small.
Take a rock rolling down a hill — something caused the rock to start rolling, various things change its velocity over time, and those may be poorly understood —- even though gravity per se is well understood.
Maybe the same story would be ‘continued on next rock’ — but maybe not, something else may kick that one off.
I think the science is focusing on forcings and feedbacks because, whatever happened when previous big changes ran their course, this one’s so much faster that the past is only useful, not determinative.
Have you read Ray Pierrehumbert’s paper on ‘Science Fiction Atmospheres’? (It’s linked at his website, his name’s listed under Contributors). It gave me a glimpse of how to think about what might be possible, and what we don’t know.
There seems to be a lot of discussion about climate models and whether they are deterministic or chaotic.
Reading this thread it sounds as if on one hand the laboratory physics is very simple and fully understood while on the other no single climate model follows the climate with any great accuracy implying that once the physics gets out of the lab and into the atmosphere it is not really well understood.
Although computers do a wonderful job of simulating airplane wings this is surely different. In that case the model is being used to explore different wing types using the extremely well understood mechanisms that wings undergo. The fact that these models produce better wings shows that they are accurately modelling the real item.
In climate science it seems the computer models are actually being used as a/the primary tool to better understand the climate mechanisms themselves because these are still not well understood.
Regardless of how impressive they are a software developments and how much people talk them up, the only real check on their validity is their ability to actually model our climate and so far no single model has managed this.
Until a single model (as opposed to a mean of models) actually starts being accurate can we assume that climate science is really not as well understood as the consensus would like us to believe.
I have read than most of Antarctica is cooling something not predicted by Global Warming or the models � is this correct.
Re 306. Bob, First, where did you get the idea that the models are inaccurate. They have gotten the trends right for over 20 years–and trends are what climate is about. They have accurately predicted the effects of volcanic eruptions.
Second, modeling is a very reasonable way of exploring a system that cannot be realized or simulated in the laboratory. If there are uncertainties, you conduct simulations over the range of uncertainties and report the robust results and perhaps the range of results.
Third, the models are all giving pretty much the same trends. Yes, climate is complicated. Yes there are things we don’t understand. However, the field is sufficiently mature that it is unlikely we will see any major revolutions in our understanding that change the conclusions.
Regarding cooling in the Antarctic, this has been treated in-depth on this site–most directly here:
https://www.realclimate.org/index.php/archives/2004/12/antarctic-cooling-global-warming/
If there are climate scientists out there who really disagree with the hypothesis of anthropogenic causation, they sure aren’t publishing. Even Lindzen acknowledges it is happening. He just thinks we’ll be saved by some miraculous restorative mechanism that keeps the effects from getting outside our comfort range.