Arguments
Chaos theory and global warming: can climate be predicted?
What the science says...
| Select a level... |
Basic
|
Intermediate
| |||
|
Weather is chaotic but climate is driven by Earth's energy imbalance, which is more predictable. |
|||||
Climate Myth...
Climate is chaotic and cannot be predicted
'Lorenz (1963), in the landmark paper that founded chaos theory, said that because the climate is a mathematically-chaotic object (a point which the UN's climate panel admits), accurate long-term prediction of the future evolution of the climate is not possible "by any method". At present, climate forecasts even as little as six weeks ahead can be diametrically the opposite of what actually occurs, even if the forecasts are limited to a small region of the planet.' (Christopher Monckton)
One of the defining traits of a chaotic system is 'sensitive dependence to initial conditions'. This means that even very small changes in the state of the system can quickly and radically change the way that the system develops over time. Edward Lorenz's landmark 1963 paper demonstrated this behavior in a simulation of fluid turbulence, and ended hopes for long-term weather forecasting.
However, climate is not weather, and modeling is not forecasting.
Although it is generally not possible to predict a specific future state of a chaotic system (there is no telling what temperature it will be in Oregon on December 21 2012), it is still possible to make statistical claims about the behavior of the system as a whole (it is very likely that Oregon's December 2012 temperatures will be colder than its July 2012 temperatures). There are chaotic components to the climate system, such as El Nino and fluid turbulence, but they all have much less long-term influence than the greenhouse effect. It's a little like an airplane flying through stormy weather: It may be buffeted around from moment to moment, but it can still move from one airport to another.
Nor do climate models generally produce weather forecasts. Models often run a simulation multiple times with different starting conditions, and the ensemble of results are examined for common properties (one example: Easterling & Wehner 2009). This is, incidentally, a technique used by mathematicians to study the Lorenz functions.
The chaotic nature of turbulence is no real obstacle to climate modeling, and it does not negate the existence or attribution of climate change.
Last updated on 8 September 2010 by chuckbot. View Archives
nickels was also resorting to a genuine fallacy of argument from authority - on the sole basis of nickels' asserted PhD, we were supposed to take nickels' criticisms of climate modelling at their word.
To which the response is, not going to happen.
TWFYSYWDI - Weather is certainly chaotic, a non-linear phenomena strongly dependent on initial conditions and difficult to predict more than a week out. Climate, on the other hand, is a boundary condition problem where long term averages are driven by energy balance, and is not chaotic. Exact future weather can't be predicted. But those averages can.
Apples and oranges, as they say.
KR @102,
Indeed, had This-Will-Frighten-You-So-You-Will-Delete-It read the remainder of the paragraph he quotes fron IPCC TAR 14.2.2.2 he would have learned that long-term averages and probability density functions are predictable, as the line following his quote reads "The most we can expect to achieve is the prediction of the probability distribution of the system's future possible states by the generation of ensembles of model solutions." And that is what is being delivered.
And what fun, his nom-de-web was chosen especially for us.
TWFYSOWDI elsewhere illustrates a desire to use one of the standard tools of pseudoscience, out of context quotation. Typically understanding the full context requires not just quoting the full paragraph from which the quote derives, but sufficient background information to understand what is meant by it. So in this case, TWFYSOWDI should have quoted the full paragraph, which reads:
Had he done so, however, he would have undercut his message by showing that the authors of Chapter 14 of the IPCC TAR believed that there where some things that could be predicted about future climate, ie, the probability distribution (aka, the statistics) of future climate states. He would also have shown that the authors believed a particular strategy was needed to make such predictions of the statistics, a strategy they in fact followed.
However, full context requires understanding what the climate is, specifically:
So not only are the statistics of climate states in principle predictable, but climate itself is just the statistics of those states so while weather is not in principle predictable beyond a few days into the future, climate is. At least according to the IPCC TAR. Despite this, by unscrupulous (I would say dishonest) selectivity in quotation, TWFYSOWDI makes them appear to say the opposite.
Note to the moderator: Every now and then we get some insecure individual who is so lacking in confidence that their message can stand on it own that they feel they must give it a boost in their "nom de web". I find such attempts at persuasion outside of reason annoying. I particularly find them annoying when the consist of caclulated insults to their hosts. May I suggest that such names be banned from SkS.
The moderator (RH) responded to TWFYSOWDI elsewhere by saying, in part
That is not quite true, and certainly not true by definition, in that cases of hysteresis open up the possibility of chaotic responses in climate. Indeed, as hysteresis in climate typically involves a threshold effect, and as individual climate states are not predictable, in principle with some levels of forcing, the climate response must be chaotic in that the crossing of the threshold may only occur (with that level of forcing) with the occurence of a particularly improbable realizable state given that forcing. Of course, with a steadilly changing forcing, the threshold will be crossed at some time, though the actual timing of crossing the threshold may vary considerably given a forcing history.
Of course, deniers will not entertain this possibility. Any broaching of the possibility is dismissed as "alarmist" and "hysterical". Nor can they consistently allow for such possibilities in that for such responses to be chaotic, they must involve some combination of large climate sensitivity and/or significant lags in climate responses, both of which they deny (and both of which would be very bad news). So, it may be the case that future climate response to anthropogenic forcing is unpredictable - that we may suddenly transition to an entirely different base climate at an unknown threshold of warming. But that would be a far more alarming situation requiring a far more rapid mitigation response than that indicated by standard projections of future climate found in the IPCC.
[RH] I'm thinking in terms of a very broad definition, as in, deserts don't become rainforests in the short term. Climate is weather averaged over 30 years or more. It can change but it isn't chaotic in the sense that you don't know what the weather is likely to be in coming years.
Tom Curtis - Keep in mind that system hysteresis does not directly imply chaotic dynamics. Many physical systems show hysteresis that is quite predictable, with consistent thresholds, irrespective of previous state trajectory histories.
And while chaotic weather variations may kick global climate energy levels about a fairly small range, initiation of a transition is still based upon rather fixed if unknown thresholds. The chaotic weather simply adds a probablistic blur to those thresholds - long term average climate isn't going to show variegated wandering paths due to initial state dependence.
Tom Curtis - Continuing: I suppose this comes down to the differences between immediate weather states (including aperiodic fluctuations such as ENSO) and long-term climate averages. If the averaging period of climate is long enough to encompass and average multiple chaotic weather variations then the bounds of variability are non-chaotic, no matter how non-linear. A boundary question of averages is very different from an initial state question of precise trajectories.
I suppose that D-O events and their uncertain timing may reflect some chaotic climate behavior, changing long term climate averages; but if they're truly cyclic phenomena they then aren't chaotic by definition.
KR @106, while I agree that hysteresis does not directly imply chaotic dynamics, hysteresis plus chaotic weather does open up that possibility. As a simple case, we could imagine a system where the equilibrium level of the climate for a given forcing lies close to but above a threshold point, such that annual short term variability (forced plus internal) can kick it below the threshold with a return period fo 1/1000; and such that after crossing the threshold the same forcing results in an equilibrium level just below the threshold to return to the prior state, with short term variability again having a similar return period to kick it just above the threshold. In such a system, with stable base forcing, the climate will kick between the two states at unpredictable intervals, but with a mean duration in each state of a thousand years. It would be chaotic.
In fact, some climate scientists at least believe the Earth is in a similar state (with larger perturbations required to cross the threshold and unequal return intervals), and that that accounts for the glacial/interglacial cycle. (Science of Doom had a recent series of blog posts expounding, and linking to relevant papers, for just such a view.)
Of more concern as it is more likely to impact us in the near future, a steady increase in forcing over time may cross unknown thresholds which may result in changes in climate unpredictable from the emperical data prior to the crossing of the threshold. As I understand it, is is a common view of climate scientists that such thresholds do exist, but that the level of the thresholds is essentially unknown for most cases. While not strictly chaotic, this does introduce a level of uncertainty in projections which is the fundamental point of denier arguments about the supposed chaotic nature of climate. Of course, if it is in fact the case, it is bad news rather than good news for it significantly increases the probability of large impacts from AGW.
Anyway, just to be clear about what I am claiming, hysteresis introduces the possibility of chaotic dynamics in climate. That in turn means it is not true "by definition" that climate is not chaotic. Further, there is evidence that hysteresis has introduced some level of chaotic dynamics to climate in the past, most notably with snow-ball earth episodes, and potentially other ice ages. Consequently it is possible but not know to be the case that chaotic dynamics could be introduced at some threshold passed by warming in the next century or two from AGW. Therefore, it is not true in all realized climate states that the climate is not chaotic. I am not, however, claiming that we will experience chaotic dynamics in climate under current conditions or over the next few centuries. I am inclined to think that we will not. We just cannot rule out the possibility.
KR @107, for the sake of discussion, assume that ENSO is in fact random. If that is the case, it is possible that we could experience a long period of more frequent La Nina (or El Nino) states purely by chance. Such an occurence would reduce the long term average of the GMST. It follows that a threshold keyed of long term averages rather than immediate weather states can also cross a threshold for hysteresis randomly. The difference is that the probability of crossing the threshold will be smaller when keyed of long term averages (and hence the return interval larger).
Technically speaking hysteresis doesn't induce the possibility of chaotic behavior - chaotic dynamics require sensitivity to starting conditions, which requires non-linearity. And even non-linear systems can be largely or wholly deterministic - for example many power amplifiers are non-linear yet deterministic over their entire operating ranges.
It's important to distinguish between hysteresis states and chaotic bifurcation attractors. Hysteresis states are separated by unmatched thresholds - the threshold from state A to state B is closer to B than the threshold from state B to state A. But bifurcated chaotic attractors (and the bifurcation itself breeding additional attractor regions) vary with the state of the system, and describe where a chaotic system may range while in that particular state. They are not the same thing at all.
Summary; hysteresis alone doesn't induce starting condition or state history dependent chaotic dynamics. But a non-linear system with hysteresis may in addition exhibit chaotic dynamics.
"...a threshold keyed of long term averages rather than immediate weather states can also cross a threshold for hysteresis randomly"
Indeed. But that is true of any system with both hysteresis and noise, and is what I meant when talking about the probablistic blurring of hysteresis thresholds.
TWFYSYWDI - Let's add some context, shall we? IPCC AR4, FAQ 1.2, :
Which is exactly what climate models are used for, projecting averages. You are conflating weather with climate - and based on this, appear not to have actually read the opening post.
[TD] When TWFYSYWDI persists in pushing the same "points" without incorporating any of the original posts or the responding comments, or the external resources that responders are pointing to, those comments by TWFYSYWDI's are being deleted.
KR, you quote the IPCC as saying "The chaotic nature of weather makes it unpredictable beyond a few days." However, "when weather is averaged over space and time, the fact that the globe is warming emerges clearly..."
This means the climatologists fully understand that predicting weather beyond a few days with a computer model is exactly as effective as predicting it with, say, chicken entrails.
Knowing this, they go ahead and consult their entrails, examining them carefully to learn what the weather might be like in 50 years, if only entrails had predictive value. But since they know this is silly, they don’t stop there. They go on to examine the entrails of a million chickens.
They average the results of the million chicken-entrails predictions together and, voila, pronounce the result “scientific.”
It is amazing what nonsense people will allow themselves to believe.
[PS] You are simply ignoring the points made to you, and continuing with worn-out sloganeering, strawmen, and argumentative language. If you are not prepared to read the actual science with view to understanding, then you are in no position to comment on it. Commentators here have attempted to explain the difference between weather and climate what is predictable or not, but apparently to no avail.
Please note that posting comments here at SkS is a privilege, not a right. This privilege can and will be rescinded if the posting individual continues to treat adherence to the Comments Policy as optional, rather than the mandatory condition of participating in this online forum.
Moderating this site is a tiresome chore, particularly when commentators repeatedly submit offensive or off-topic posts. We really appreciate people's cooperation in abiding by the Comments Policy, which is largely responsible for the quality of this site.
Finally, please understand that moderation policies are not open for discussion. If you find yourself incapable of abiding by these common set of rules that everyone else observes, then a change of venues is in the offing.
Please take the time to review the policy and ensure future comments are in full compliance with it. Thanks for your understanding and compliance in this matter.
@113, I do not believe such gross misrepresentation of the claims in the IPCC should be allowed to stand. Whether they were made due to deceit or stupidity, it does not matter. More than enough has been discussed above that the author of the post should now understand the distinction between predicting precise values and predicting statistical spreads (means and standard deviations) and should no longer be confusing weather and climate. Given that, he is clearly now only sloganeering. He is also clearly not worth discussing issues with.
[PS] I agree on all counts. Comprehension of the comments policy at this site also appears to be problem.
I'm sure that repeating a comment word-for-word, even one that has been zapped by the moderators; that is still contravenes the SkS Comments Policy in that "Comments should avoid excessive repetition." The humour content aside, and the efforts of this deluded troll are droll, I think the chicken entrails have spoken @115.
[TD] Quite correct. Excessive repetition is banned, because the comments are for discussion. Supporting discussion requires as well that the commenter actually read the original post and material that other commenters point them to, think about those, and respond to them in specifics. It is telling that This Will Frighten You so You Will Delete It has never mentioned "boundary conditions." Obviously he or she has not read the Intermediate tabbed pane of this post. Nor has he or she bothered to use the Search field at the top left of the page to search for "weather," else he or she would have found (and of course truly read) The Difference Between Weather and Climate. There is even a video about Weather Vs Climate that he or she would have seen this week if taking the (free!) Denial 101x course.
There are even easy analogies in several comments by various people on several threads, such as why it is possible to predict accurately how high the tide line will be on a beach (climate), despite the difficulty of predicting precisely how high on the beach an individual wave will come in exactly 10 minutes 51 seconds from now (weather more than 5 days from now), compared to the relative ease of predicting how high up the beach will come this wave that already is racing up the beach (weather within 5 days from now). TWFYSYWD could have read those, asked for clarification, or even challenged those analogies, all of which would have been fine.
Science of Doom has a good series on weather versus climate, and related topics.
[DB] TWFYSYWDI has recused themselves from further participation in this venue for a variety of multiple infractions of the Commenting Guidelines of this forum, not the least of which being in a line of sock puppet fake accounts (a serial-sock).
I study physics and statistical analysis. An easy analogy to explain concepts like chaos would be the random processes involved by individual atoms or molecules in an object. That would be like weather. Much more complex then weather. Weather is at least predictable with confidence. It tends to normalize fairly quickly so a butterfly flapping its wings in China does not have any effect on a weather event in the future. That's a part of chaos theory. You have to show that the dynamic system is actually sensitive to minor changes in initial conditions. I would not consider weather let alone climate to be that sensitive to minor fluctuations.
Statistical mechanics makes use of all the individual random probabilities associated with each particle (a very crude definition but the best one in simple terms I can think up). When you include all the probabilities everything normalizes so you can't do a good job of describing the individual particles but you can define the entire object, region, whatever your calculating. That could be kind of like climate.
I believe this was attempted to be explained earlier. In physics and statistics stochastic processes have been validated in great depth and high degree of confidence. I am not a climatologist but the concept applies to a broad range of sciences so when they use some of these methods I generally accept the results if there were no errors in their calculations. I am not sure what the relevance of a lot of the statements on chaos being made so I did not get the best impression that chaos theory and statistical mechanics was well understood by some of the commenters. I'm not naming anyone and some people do grasp it here. But in the future a good guideline to use is if you understand something you then you can generally explain it without having to take the explanations straight out of a book. You also would not take a small part of a science theory in this case statistics or thermodynamics and try to make it sound more important to the overall results then it really would be.
I have an undergraduate degree in physics and mathematics. And I'm currently studying to be a statistician. So I am only commenting on specific instances here. I won't pretend to know climate models very well but I can assure you the methods being applied as far as the physics and math is not in question. if anyone would wish to argue against anthropogenic climate change these are not the areas you want to try and dispute.
The weather has, is and always will change and I think, there is nought that is in the human dynamic, that will change this.
As science is a thing in progress and has not yet reached its zenith, then predicting long term cycles in weather, remains an aspiration.
I sense, yet I cannot prove, there are major parameters that affect weather of which science is not yet aware of.
I could be wrong, its just the way I see it.
Interesting, Alnero2 @118 , that the way you see it is that you could be wrong. You are quite correct in that!!
Interesting also, that you sense there's more than one "major parameter" science is not yet aware of. Your claim is not just extraordinary, but more than extraordinary! How reliable is your ability to sense the unknown — for instance, can you reliably sense the results of future horse races?
I sense, yet I cannot prove, there is a sock missing from your sock drawer.
I certainly hope you do not think that policy-makers should be guided by what you "sense" as opposed to scientists reasonable skill at predicting future climate (a cycle is not climate by definition really).
You need to be clearer about the difference between weather and climate. It would be very hard to predict the temperature for 24Dec 2017 where you live. However, I think you could nail the average temperature for Dec now with reasonable accuracy.
Furthermore, you can state with very high confidence that the average temperature for June will be warmer (if in NH temperate area) than average temperature for Dec. Now think for a moment. Why is that true? What is your answer?