Discussion > Are Geological Paleo-Climate Records Relevant to The Climate Debate?
Perhaps worth noting how that quotation from the IPCC continues ...
The climate system is a coupled non-linear chaotic system, and therefore the long-term prediction of future climate states is not possible. Rather the focus must be upon the prediction of the probability distribution of the system's future possible states by the generation of ensembles of model solutions. Addressing adequately the statistical nature of climate is computationally intensive and requires the application of new methods of model diagnosis, but such statistical information is essential.
Phil Clarke
Alan,
I thought this was the reason why the IPCC never acknowledges its outputs as predictions.
The main reason the IPCC uses "projection", rather than "prediction", is because the outcome depends on certain inputs that cannot be known in advance, such as our future emissions. Hence the "prediction" is conditional on the assumed inputs being what actually takes place (i.e., our actual emissions matching what was assumed). This is why there are numerous emission pathways that are considered; from high ones, to low ones. This gives information as to what might happen for a number of different possible future pathways.
...and Then There's Physics
Spectator, Alan Kendall
Spectator
What part of " the long-term prediction of future climate states is not possible" do you not understand?
Level 1 thinking. The proper phrasing should have been
" the long-term prediction of future climate states is not possible precisely "
Alan Kendall
You are also struggling to think beyond level 1.
As you said
simulation does not equate to prediction but might encompass a range of different possible outcomes. The example you use -a double pendulum - only illustrates the inability to predict such systems (except in terms of statistical probabilities)
Once you get beyond simple physics systems get too complex and non-linear for a precise mathematical prediction. Hence the projection of possible futures, a probability distribution such as the CMIP5 ensemble.
There is no incompatibility betwwen my two statements, though I should probably have used projection throughout. Prediction is Level 1 thinking. Projection is Level 2 and Level 3 thinking.
To me a prediction is a precise outcome-"The global average temperature in 2030 will be anomaly 1.30C".
A projection includes the uncertainties- " To 95% confidence the global average temperature in 2030 will be anomaly 1.30C+/-0.3C"
Entropic man
Addressing adequately the statistical nature of climate is computationally intensive and requires the application of new methods of model diagnosis...equates to "keep sending the money"
... but such statistical information is essential.Why? We've managed pretty well without it for the last few million years. Why now?
In fact the entire quote says, "there is no way we can possibly tell what the climate is going to be like in 100 years time, but tell us what it is you want to hear and we'll give you a nice sciencey-sounding answer. In fact as long as you keep the grants flowing we'll keep on giving you different answers so you can pick whichever one suits at the time."
Mike Jackson
and Then There's Physics,
Sorry, I might be exposing my lack of understanding even more than I ought.
If what you say is correct and the unknown variable that prevents predictions being made is the emission pathways, then surely if the emission pathways are arbitrarily specified (as IPCC does), are not the modelled outcomes then predictions?
Why does the IPCC and its spokespeople continually deny its products are predictions?
Alan Kendall
ER.
So now you are further refining your definition of prediction to fit your own argument.
If you carefully read what I wrote you will find that it agrees with what you are now writing - yet you have the unmitigated gall to accuse me of not thinking beyond stage 1. It must be so very pleasant to always be so very, very right about everything. Ice mechanics, statistical logic, isotope proxies, climate prediction: is there nothing beyond your grasp? I salute you.
Alan Kendall
Alan,
If what you say is correct and the unknown variable that prevents predictions being made is the emission pathways, then surely if the emission pathways are arbitrarily specified (as IPCC does), are not the modelled outcomes then predictions?
They're conditional predictions and, typically, that is referred to as a projection, rather than as a prediction. There is also the scientific issue of also taking uncertainty into account, which is done by presenting ranges, rather than specific numbers. The outcomes then depend both on the input assumptions (things we can't know in advance like our future emissions) and on various other uncertainties in the modelling.
Why does the IPCC and its spokespeople continually deny its products are predictions?
I think for the reason above. They're not really predictions, because the IPCC is not trying to predict what we decide in the future; it's trying to present information that can inform decisions. Hence, what they present (in the case of the future, at least) are normally referred to as projections, rather than predictions, so as to make this distinction.
...and Then There's Physics
Why does the IPCC and its spokespeople continually deny its products are predictions?
Plausible deniability springs to mind, there will always be some parameter different or forgotten to give a get out when reality forgets to act at it should.
Breath of Fresh Air
Consider the difference between a simple pendulum and a double pendulum.And how close to reality will this simulation be? What if we consider the future behaviour of a multi-pendulum? What if we do not actually know how many pendula are linked, or what the weight or length of each one is? How “simple” would projections based upon simulation be then? That is the level of complexity we have here, EM, and I doubt that your relatively simple mathematical brain (this is not deriding your attempts, but you have admitted that you have trouble with calculus, which at least one other on this site finds easy – “easy” being a relative term, of course) will be able to sort that out – after all, it does seem to be confounding considerably greater mathematical brains than anyone else is displaying on this site.The future behaviour of a simple pendulum is easily projected by t=2π√l/g.
The future behaviour of a compound pendulum is complex and non-linear. It can only be projected by simulation.
Because of the heat capacity of the system there is a lag.But you seem to discount the possibility of a lag in response to cosmic rays. Hmmm…
To return to the original question, the title of this thread – I suggest that the answer should be: Yes, but with extreme caution.
Radical Rodent
However the surface temperature would continue to increase until the energy of the outgoing radiation matches the incoming radiation.
According to Svenmark the same thing would happen with cosmic rays. (...)
Apr 8, 2016 at 3:59 PM | Unregistered CommenterEntropic man
EM - thanks. If I have understood (not sure that I have) you have conceded that cosmic ray intensity can remain constant while their effect continues to change things.
Pendulums
Note that both simple and compound pendulums are nonlinear for finite amplitudes.
A simple pendulum can be solved exactly - but not easily (think elliptic integrals) for finite amplitudes. Conversely a double pendulum can be approximated quite well by a linear 1st order vector differential equation easily solvable (its state vector having just four elements) for small amplitudes.
Chaos
Despite what you believe, chaotic systems, essentially by definition, *are* "impossible to predict more than a short way ahead". The precision needed for their initial conditions explodes. You find you soon finish up needing more decimal places of initial condition accuracy than there are particles in the universe, to predict more than a short way ahead.
____________________________________________________________________________
You wondered whether the CET record shows acceleration of temperature. It seems possible it does. [E&OE for the following - done very rapidly and not checked]
I divided the CET January temperature record into three segments: 1659 – 1777, 1778-1896, 1897- 2015 (118 years each segment). I computed the mean temperature for each segment T1, T2, T3.
Then I took the differences, getting T2-T1 = 0.08 °C, T3-T2 = 1.1 °C.
I made an estimate of the standard deviation of each temperature difference (each temperature difference contains 2*118 evidently uncorrelated annual temperature fluctuations, whose overall variance can be estimated from the data): standard deviation of each difference estimated at 0.25 °C. [No confidence intervals please.]
So it looks as if the mean January temperature changed by about 0.08 ± 0.25 °C over 118 years in the first part of the period 1659 - 2015 and by about 1.10 ± 0.25 °C over 118 years in the second part of the period. So unless I made a mistake in the calculations, January warming in central England has accelerated to some degree over the period 1659 - 2015.
Note that, as I said, the foregoing has not been checked and may contain errors. If I get the opportunity, I'll check it over the weekend.
Martin A
Dr Kendall
If you carefully read what I wrote you will find that it agrees with what you are now writing - yet you have the unmitigated gall to accuse me of not thinking beyond stage 1. It must be so very pleasant to always be so very, very right about everything. Ice mechanics, statistical logic, isotope proxies, climate prediction: is there nothing beyond your grasp?Good. I am glad we are now in agreement on projections.
Level 1? If you link me to a paper of youurs in which you use statistical methods and confidence limits I will apologise immediately.
The nice thing about being a science teacher was the breadth of knowledge you picked up.
After a lifetime of reading I can understand the basics of a wide range of sciences.In this internet age the knowledge gives me a head start . I know the right questions to ask to easily esearch more detail as required. It also gives me a very sensitive bullshit filter.On the specific subject of climate I started helping paleoclimate researchers as a student in the early 1970s, then found myself living in the Murrins Complex. I have been reading about climate and paleoclimate ever since.
Entropic man
A critique of an expert on confidence intervals (who once stated that he had attended 'statistics 101'):
http://bishophill.squarespace.com/discussion/post/2578881?currentPage=7
And who also stated that he had "medical training" adding to his fields of expertise.
Martin A
Alan
When I use a word,' Humpty Dumpty said, in rather a scornful tone, 'it means just what I choose it to mean — neither more nor less.'Knew a thing or two, did Lewis Carroll!
'The question is,' said Alice, 'whether you can make words mean so many different things.'
'The question is,' said Humpty Dumpty, 'which is to be master — that's all.'
Mike Jackson
Mr K: perhaps you are now beginning to understand the responses of others on this site… However, do note that your protagonist has yet to grasp the finer intricacies of HTML coding.
Radical Rodent
Martin A
EM - thanks. If I have understood (not sure that I have) you have conceded that cosmic ray intensity can remain constant while their effect continues to change things.
Yes, under the specific conditions I discussed.
I wonder if we have been at cross purposes again. I interpreted your earlier calculus as dJ/dT is constant, J is changing at a constant rate. iIf Svenmark was correct, that would produce a temperature change.
Did you intend the more specific case dJ/dt=0, J is constant? That would only produce a temperature change during the lag.
impossible to predict more than a short way ahead".
I find it fascinating how the chaos makes it possible to make short term forecasts in detail over a few days then impossible to make forecasts over three months. Then on a global scale the chaos goes away and you can make climate forecasts.
You are restoring my faith in CET. I hate it when different lines of evidence give contradictory results.
Entropic man
Martin A
Yes, a year's postgraduate Biomedical Science. Unfortunately I damaged my back after 6 months in a hospital Haematology lab and had to retire AGAIN.😕
Entropic man
The idea of EM possessing a bullshit filter has almost made me cough up my dinner. EM is no more than a bullshit sprayer. All those neat but meaningless formulae.
Diogenes
Then on a global scale the chaos goes away and you can make climate forecasts.
Apr 8, 2016 at 8:47 PM | Unregistered CommenterEntropic man
Glendower:
I can call spirits make global climate forecasts from the vasty deep unverified climate models using 'parameterisation' to model the bits that I don't understand and pretending that, over the long term, the system is no longer chaotic, notwithstanding that such a claim is ridiculous.
Hotspur:
Why, so can I, or so can any man even the Met Office Hadley Centre;
But will they come when you do call for them match reality decades ahead ?
Glendower:
Well obviously not, except by pure chance. But who gives a shit? We won't be around by then. They will have served their purpose.
William Shakespeare
"[..]As long as the factors affecting the system remain constant the steady state/equilibrium persists."Apr 8, 2016 at 12:21 PM | Entropic man
No it doesn't.
Categorically.
We are experiencing an El Nino effect right now that demonstrates the point.
Previously-warmed water from the Western Pacific Ocean has sloshed Eastwards, pissing out huge amounts of heat to both the atmosphere and space, warming global surface/tropospheric temperatures. Thus a significant disparity between incoming and outgoing radiation. This has happened over a period of months while solar input is constant.
The climate system does not need any “forcings” at all to undergo such large changes. This is usually called internal variability. These changes can happen over any time-scale. Application of 'equilibrium thinking' does not work for your argument.
There is no specific shame in saying you have a biology background, as opposed to engineering. Thermodynamics applies equally to both. But kinetics are often as important, if not more so. As the chemistry joke goes: "Old Chemists never die. They just reach equilibrium."
michael hart
... I hate it when different lines of evidence give contradictory results.
Apr 8, 2016 at 8:47 PM | Unregistered CommenterEntropic man
EM - that is termed 'intolerance of ambiguity'.
As I think Paul Dennis suggested, what was needed was to look at a smoothed version of the temperature data. If you look carefully at the smoothed version of the January data http://s18.postimg.org/itxgntd6x/smoothed_41_yr.gif the left half of the red trace is more or less horizontal overall; the right half has a more of less steady gradient. So the acceleration was there visible to be seen.
Martin A
EM - thanks. If I have understood (not sure that I have) you have conceded that cosmic ray intensity can remain constant while their effect continues to change things.
There will be a lag in delta-temperatures due to the huge thermal inertia of the planet, if external forcings became net zero, it would take decades to reach equilibrium.
However the effect on clouds would be visible in a much shorter timeframe, and this has not been observed, so we are left with some as yet unobserved mechanism. Cosmic leprechauns perhaps.
Phil Clarke
Michael hart
Biological equilibrium states tend to be maintained by considerable effort. Your body temperature is kept approximately constant by sweating, shivering, metabolic adjustments and a variety of blood flow adjustments.
A climate in equilibrium will also show variation around the equilibrium state due to ENSO and other cycles. The long term trend may be zero but, as you indicate, there will be short term deviations. This is why I prefer dynamic equilibrium to steady state.
Entropic man
Excellent, EM! Now, tell us how large these climate variations might be, and over what time period they might extend.
Radical Rodent
EM - even quite simple nonlinear dynamic systems can have multiple equilibrium states and which one they settle into depends on starting conditions. Or they can hop from one to another (so far as an external observer is concerned) in an apparently random manner.
Apart from your imaginings (which we all know, for you, conflate with reality), there does not seem to be any reason to think that the climate system would behave in the same way as a very simple object with a single unique equilibrium condition. It's in the same category as your belief that, in the long term, its chaotic behaviour disappears.
Talking about "a climate in equilibrium" and doing calculations on that assumption is the sort of thing that Diogenes was talking about when he used the phrase "Micky Mouse pseudoscientists calculations".
It may be interesting to do such calculations and to see what they say, but confusing them with reality is self delusion.
"I have seen calculations for climate sensitivity between 1 and 6"
Is there any more convincing evidence that climate science does not know its arse from its elbow have even a basic understanding of the behaviour of the climate system?
Martin A
Martin A
No. Because of the heat capacity of the system there is a lag.
The decrease in OLR due to the greenhouse effect would stop changing immediately. The imbalance between the incoming radiation and outgoing radiation would also stop increasing.
However the surface temperature would continue to increase until the energy of the outgoing radiation matches the incoming radiation.
According to Svenmark the same thing would happen with cosmic rays. The cooling effect would vary with intensity J. If the intensity varied with time, so would the cooling effect. The temperature would change too, lagging by an amount depending on the rate of change in intensity. Only when the rate of change in intensity is zero for a period of time can the temperature catch up and stabilise.
While dJ/dt is positive or negative, even if it if constant, the temperature would continue to change.
Only if dJ/dt =0 can temperatures stabilise, and then after a lag while the radiation budget adjusts.
Thus dJ/dt ="constant" will only produce dT/dt=0 in the specific case of dJ/dt=0