I don't think it takes much to make scientists political - they all have opinions!
Non scientists should be careful what they ask for though. A lot of the financial
mess was caused by good scientists doing proper math with wrong
assumptions. The people asking for the work did not fully understand the
problem until it hit them between the eyes with a 2x4.
Same thing is true with nuclear weapons - the scientists made a good case
to the politicians at the time but it later got away from them. Just because
we know how to do things doesn't mean we should do it - and that is a
political decision. Using nuclear weapons for mining is an example - yes it
can be done, but it really doesn't make sense.
The earth's climate is changing. What it is changing to won't be known for
1000 years. I would argue humans are too insignificant to be part of the
change, but others argue otherwise. The arguments can be scientific and
polite - but the actions of humans are still based on politics. Usually, the
facts and the science are ignored and people do what they feel like until
hit over the head with reality.
To the people in the UK I'd warn "be careful what you wish for". You just
might get it - and really not like the results. Science is kind of like
religion - it needs to be separated from politics. It's better for political
decisions to be made on more than just science - sometimes emotion can
play a really important role and must be dealt with first.
Global Warming Concensus Broken
The derivatives (Black-Scholes model etc) people are mathematicians, not scientists. And their work is grossly misinterpreted (as might be expected) by people trained as economists. As it stands I do not view economists as scientists - it is the nature of the subject that if treated as science it will be oversimplified to try to find neat causal relationships.
In UK the scientists got into trouble over BSE. They were pretty sure that there would be no problem for human health, and wrong. They caved in to political pressure to turn "highly unlikley given what we know, but can't be sure" into something stronger and more politically acceptable.
Now politicians want scientists to take the (political) responsibility over issues like food safety. they should not. But equally they should speak out clearly without political spin when there are clear scientific statements that affect public policy and are in danger of being ignored by politicians:
over-fishing the sea
ill-heath through bad lifestyle
(in UK) lack of future energy security unless we build more nuclear power stations
etc.
Many of the scientific truths are inconvenient, and politicians do not like to dwell on them.
Best wishes, Tom
In UK the scientists got into trouble over BSE. They were pretty sure that there would be no problem for human health, and wrong. They caved in to political pressure to turn "highly unlikley given what we know, but can't be sure" into something stronger and more politically acceptable.
Now politicians want scientists to take the (political) responsibility over issues like food safety. they should not. But equally they should speak out clearly without political spin when there are clear scientific statements that affect public policy and are in danger of being ignored by politicians:
over-fishing the sea
ill-heath through bad lifestyle
(in UK) lack of future energy security unless we build more nuclear power stations
etc.
Many of the scientific truths are inconvenient, and politicians do not like to dwell on them.
Best wishes, Tom
Tom,
I was thinking of RF circuits with component values specified to the limits of measurement.
Well actually it is not calculable because it derives from chaotic processes which are sensitive to initial conditions. Which are unknown to sufficient accuracy.
On top of that - even assuming you can subtract out the noise (or at least account for it) we have - so far - a "signal" that is under 1/2 the noise level.
==
The PDO and other similar "noise" has been known for 10 years. So far none of the models have subtracted it out.
I wonder why?
Well for one thing it might kill most of the warming.
I was thinking of RF circuits with component values specified to the limits of measurement.
This noise is however in principle calculable, so calling it noise is a misnomer.
Well actually it is not calculable because it derives from chaotic processes which are sensitive to initial conditions. Which are unknown to sufficient accuracy.
On top of that - even assuming you can subtract out the noise (or at least account for it) we have - so far - a "signal" that is under 1/2 the noise level.
==
The PDO and other similar "noise" has been known for 10 years. So far none of the models have subtracted it out.
I wonder why?
Well for one thing it might kill most of the warming.
Engineering is the art of making what you want from what you can get at a profit.
Don't forget the essential element of tying down the safeties - credit scores don't matter because the government is backing the mortgage "securities". Allowing for things like NINJA mortgages. No Income, No Job, No Assets.Non scientists should be careful what they ask for though. A lot of the financial mess was caused by good scientists doing proper math with wrong
assumptions. The people asking for the work did not fully understand the
problem until it hit them between the eyes with a 2x4.
And for all the screw ups in the American market the Europeans did worse.
Engineering is the art of making what you want from what you can get at a profit.
I need to make a list of the comon misconceptions:
(1) GW=NHUSGW (Northern hemisphere US GW)
(2) Climate scenarios are designed to predict global temp
(3) All the climate modelers are in a club together rooting for higher temperatures
You are right that the other focing factors are not much predciatble. What I meant is that they are measurable, and their consequences calculable. Therefore the effect of CO2 in global temperatures can be better determined by allowing for them.
The climate models which do not yet do this will soon do so. Of course, this will not make for much better prediction of temperature, but it will make for better validation of models.
So if climate models are not meant to predict temperatures what can they do? They can predict the average delta T expected for a doubling of atmosphere CO2. That figure is relevant to a lot of decisions.
Best wishes, Tom
(1) GW=NHUSGW (Northern hemisphere US GW)
(2) Climate scenarios are designed to predict global temp
(3) All the climate modelers are in a club together rooting for higher temperatures
You are right that the other focing factors are not much predciatble. What I meant is that they are measurable, and their consequences calculable. Therefore the effect of CO2 in global temperatures can be better determined by allowing for them.
The climate models which do not yet do this will soon do so. Of course, this will not make for much better prediction of temperature, but it will make for better validation of models.
So if climate models are not meant to predict temperatures what can they do? They can predict the average delta T expected for a doubling of atmosphere CO2. That figure is relevant to a lot of decisions.
Best wishes, Tom
I'm reminded of the anti-nuclear protester carrying a sign "Two too many".Same thing is true with nuclear weapons - the scientists made a good case to the politicians at the time but it later got away from them. Just because we know how to do things doesn't mean we should do it - and that is a political decision.
A nuclear scientist remarked "one not enough".
Engineering is the art of making what you want from what you can get at a profit.
Well no. They can account for changes in heat (Q) caused by CO2.So if climate models are not meant to predict temperatures what can they do? They can predict the average delta T expected for a doubling of atmosphere CO2. That figure is relevant to a lot of decisions.
What screws all that up re: temperatures is that the heat pipes in the atmosphere are not well understood. Nor are clouds.
Clouds can come and go quickly - minutes. Their area can be small - miles. Their type is variable.
This is not much help in models that calculate in hourly increments (at best) and whose area covers a grid square 125 miles on a side. So what do the modelers do? They parameterize. And since all the causes and effects are variable - the parameters are rough estimates. Well fine.
To get the right results to within 1 deg C the parameters have to be known to .3% or better - for all the parameters. Which might mean that of the 5 or 6 things parameterized they must be known to .1% or better. Probably a lot better since the parameters interact in successive steps. The errors accumulate.
Engineering is the art of making what you want from what you can get at a profit.
I remember when butter was socially unacceptable - because of science. Now margarine is out and butter is back in. Transfats.in UK there is now a strong push to make scientists more socially responsible, political, etc. I am not sure I agree with it!
Engineering is the art of making what you want from what you can get at a profit.
Simon -
The heat transport uncertainties all affect how much CO2 warming is amplified or diminished by feedback effects.
Given that the whole calculation is based on difference between two models differing only in CO2 forcing I would expect parameters to affect amplification factor K. A model which has +/-20% error bars for this would be more than good enough. Errors are funny things - & not simple to estimate.
Nor do the errors in different parameters necessarily combine badly. One could reasonably hope for an RMS combination of all errors (true if each error source is Gaussian and independent). Of course if parameters are multiplicative things are worse - but is this typically the case?
I can't give proper quantitative answers to this one - the above argument just makes it seem plausible to me that precise parameters are not needed. But can you do this? The climate modellers, if competent, can. And I prefer not to condemn a whole set of scientists (not operating in collusion) to incompetence without good evidence.
Re clouds what is needed is the average effect, not all of the details. there is no reason to think that we cannot deduce this average effect reasonably well from existing 3D data. and physical models without over-fitting.
Best wishes, Tom
The heat transport uncertainties all affect how much CO2 warming is amplified or diminished by feedback effects.
Given that the whole calculation is based on difference between two models differing only in CO2 forcing I would expect parameters to affect amplification factor K. A model which has +/-20% error bars for this would be more than good enough. Errors are funny things - & not simple to estimate.
Nor do the errors in different parameters necessarily combine badly. One could reasonably hope for an RMS combination of all errors (true if each error source is Gaussian and independent). Of course if parameters are multiplicative things are worse - but is this typically the case?
I can't give proper quantitative answers to this one - the above argument just makes it seem plausible to me that precise parameters are not needed. But can you do this? The climate modellers, if competent, can. And I prefer not to condemn a whole set of scientists (not operating in collusion) to incompetence without good evidence.
Re clouds what is needed is the average effect, not all of the details. there is no reason to think that we cannot deduce this average effect reasonably well from existing 3D data. and physical models without over-fitting.
Best wishes, Tom
The trouble is: the average effect is determined by the previous state. Cumulus clouds or cirrus? Well it depends.Re clouds what is needed is the average effect,
If the previous states are drifting from reality future states will be further in error.
The big argument now is: is the water vapor amplification factor high (in the range of 3 to 5) or low (close to 1 and slightly positive).
And the effects of the various ocean oscillations have not been subtracted out of the input data. And they are not even estimated for future results.
Engineering is the art of making what you want from what you can get at a profit.
Simon -
I believe that the parameters relevant to WVA factor (which all relate to atmospheric thermal transport) are not determined from global temperature data, but from other data. This is crucial to understanding what is uncertain and what not. In this context if PDO is not included it is not a big deal since though it affects current and medium-term temperatures (especially regionally) it is not relevant to the crucial WVA factor.
But maybe I am wrong - do you have evidence to the contrary?
Best wishes, Tom
I believe that the parameters relevant to WVA factor (which all relate to atmospheric thermal transport) are not determined from global temperature data, but from other data. This is crucial to understanding what is uncertain and what not. In this context if PDO is not included it is not a big deal since though it affects current and medium-term temperatures (especially regionally) it is not relevant to the crucial WVA factor.
But maybe I am wrong - do you have evidence to the contrary?
Best wishes, Tom
Its a bit more than that. For instance, high altitude clouds cool by blocking sunlight above the CO2 greenhouse altitude while low altitude clouds warm by blocking radiation of ground heat to space. Sounds simple enough, EXCEPT that the effective altitude that CO2's greenhouse effects happen will vary based on how swelled the atmosphere is from solar wind. Currently, the upper boundary of the atmosphere is at the lowest altitude ever measured because the sun is so quiet right now. This means that all other altitude related relationships are also at the lowest possible altitude right now. When the solar wind picks up again, the atmosphere will swell up again, and those altitude thresholds will rise as well.MSimon wrote:The trouble is: the average effect is determined by the previous state. Cumulus clouds or cirrus? Well it depends.Re clouds what is needed is the average effect,
If the previous states are drifting from reality future states will be further in error.
The big argument now is: is the water vapor amplification factor high (in the range of 3 to 5) or low (close to 1 and slightly positive).
And the effects of the various ocean oscillations have not been subtracted out of the input data. And they are not even estimated for future results.
This means that right now, medium altitude clouds are more effective at reflecting sunlight back to space, etc. because the CO2 is less effective at those altitudes than before. I would put money on climate models doing NOTHING to model for this effect.
intlibber -
This is one of the (many) arguments for future global cooling.
This is the opposite argument to Simon's. Simon thinks that sunspot minimum => lower temperatures due to lower solar forcing. You think that sunspot minimum => higher temperatures due to cloud changes.
And both of you see your effect as likely to be important in ensuring future cooling.
You cannot both be right!
GW models are gradually incorporating more and more of the subsidiart effects. Given this lack of agreement it is perhaps not surprising that they do not yet incorporate this (but they are now starting to include solar forcing).
From my point of view, it is irrelevant. Sure - it is possible that quite coincidentally we are raising CO2 levels and solar forcing changes mean that we rapidly exit the current Holocene maxium. Unlikely. It is possible that we move to a regime a degree or two cooler. That would give us a btreathing space.
Neither changes the important issue which is how large is the sensitivity of global temperature on log(CO2) concentration.
The fact is tehre are very many possible effects on global temperature - some modelled, some not. Anti-GW people would like to think
(1) The un-modelled effects are large
(2) the unmodelled effects push us towards a cooler future.
They may be right - but I have not seen convincing evidence of this yet. And I think there is some cherry-picking in how effects are interpreted (only the cooling ones are highlighted). i have more sympathy with the argument that GW models do not yet have enough skill. But this is a highly complex scientific debate. And when I read the detailed debates (and the challenges and responses) so far the scientists who are pro-GW seem to have the evidence on their side.
Best wishes, Tom
This is one of the (many) arguments for future global cooling.
This is the opposite argument to Simon's. Simon thinks that sunspot minimum => lower temperatures due to lower solar forcing. You think that sunspot minimum => higher temperatures due to cloud changes.
And both of you see your effect as likely to be important in ensuring future cooling.
You cannot both be right!
GW models are gradually incorporating more and more of the subsidiart effects. Given this lack of agreement it is perhaps not surprising that they do not yet incorporate this (but they are now starting to include solar forcing).
From my point of view, it is irrelevant. Sure - it is possible that quite coincidentally we are raising CO2 levels and solar forcing changes mean that we rapidly exit the current Holocene maxium. Unlikely. It is possible that we move to a regime a degree or two cooler. That would give us a btreathing space.
Neither changes the important issue which is how large is the sensitivity of global temperature on log(CO2) concentration.
The fact is tehre are very many possible effects on global temperature - some modelled, some not. Anti-GW people would like to think
(1) The un-modelled effects are large
(2) the unmodelled effects push us towards a cooler future.
They may be right - but I have not seen convincing evidence of this yet. And I think there is some cherry-picking in how effects are interpreted (only the cooling ones are highlighted). i have more sympathy with the argument that GW models do not yet have enough skill. But this is a highly complex scientific debate. And when I read the detailed debates (and the challenges and responses) so far the scientists who are pro-GW seem to have the evidence on their side.
Best wishes, Tom
No, you misread me. My argument is that while some pro-AGW scientists have conceded that up to 45% of 20th century warming was solar cycle related, they have only accounted for some of the phenomena. As the atmosphere contracts due to lower solar wind speeds etc, a greater percentage of the cloud cover is more effective at reflecting solar radiance back to space. This is an additional forcing toward cooling, NOT toward warming.tomclarke wrote:intlibber -
This is one of the (many) arguments for future global cooling.
This is the opposite argument to Simon's. Simon thinks that sunspot minimum => lower temperatures due to lower solar forcing. You think that sunspot minimum => higher temperatures due to cloud changes.
Sorry -
But in that case you are arguing that CO2 forcing is likely to be stronger than they predict, since an unmodelled cooling forcing => other forcings must lead to more warming for the same effect?
Or if you agree with me that the climate models are physical, and not fitted at all to the global temperature then an unmodelled cooling would cause estimates to revise downwards. The problem with this is that broadly they think their models match current temperatures, and this cooling influence has been around for some time, so we are left then with the models consistently underestimating the real temperature.
The treassuring point here is that any significant extra forcing will get added in to the models as they become more sophisticated, so the mater will resolve itself.
Best wishes, Tom
But in that case you are arguing that CO2 forcing is likely to be stronger than they predict, since an unmodelled cooling forcing => other forcings must lead to more warming for the same effect?
Or if you agree with me that the climate models are physical, and not fitted at all to the global temperature then an unmodelled cooling would cause estimates to revise downwards. The problem with this is that broadly they think their models match current temperatures, and this cooling influence has been around for some time, so we are left then with the models consistently underestimating the real temperature.
The treassuring point here is that any significant extra forcing will get added in to the models as they become more sophisticated, so the mater will resolve itself.
Best wishes, Tom