Just to be clear ahead of time, chaos in weather is NOT the same as climate disruption listed below – Anthony
Guest submission by Dr. Andy Edmonds
This is not intended to be a scientific paper, but a discussion of the disruptive light Chaos Theory can cast on climate change, for non-specialist readers. This will have a focus on the critical assumptions that global warming supporters have made that involve chaos, and their shortcomings. While much of the global warming case in temperature records and other areas has been chipped away, they can and do, still point to their computer models as proof of their assertions. This has been hard to fight, as the warmists can choose their own ground, and move it as they see fit. This discussion looks at the constraints on those models, and shows that from first principles in both chaos theory and the theory of modelling they cannot place reliance on these models.
First of all, what is Chaos? I use the term here in its mathematical sense. Just as in recent years Scientists have discovered extra states of matter (not just solid, liquid, gas, but also plasma) so also science has discovered new states that systems can have.
Systems of forces, equations, photons, or financial trading, can exist effectively in two states: one that is amenable to mathematics, where the future states of the systems can be easily predicted, and another where seemingly random behaviour occurs.
This second state is what we will call chaos. It can happen occasionally in many systems.
For instance, if you are unfortunate enough to suffer a heart attack, the normally predictable firing of heart muscles goes into a chaotic state where the muscles fire seemingly randomly, from which only a shock will bring them back. If youāve ever braked hard on a motorbike on an icy road you may have experienced a ātank slapperā a chaotic motion of the handlebars that almost always results in you falling off. There are circumstances at sea where wave patterns behave chaotically, resulting in unexplained huge waves.
Chaos theory is the study of Chaos, and a variety of analytical methods, measures and insights have been gathered together in the past 30 years.
Generally, chaos is an unusual occurrence, and where engineers have the tools they will attempt to ādesign it outā, i.e. to make it impossible.
There are, however, systems where chaos is not rare, but is the norm. One of these, you will have guessed, is the weather, but there are others, the financial markets for instance, and surprisingly nature. Investigations of the populations of predators and prey, for instance shows that these often behave chaotically over time. The author has been involved in work that shows that even single cellular organisms can display population chaos at high densities.
So, what does it mean to say that a system can behave seemingly randomly? Surely if a system starts to behave randomly the laws of cause and effect are broken?
A little over a hundred years ago scientists were confident that everything in the world would be amenable to analysis, that everything would be therefore predictable, given the tools and enough time. This cosy certainty was destroyed first by Heisenbergās uncertainty principle, then by the work of Kurt Gƶdel, and finally by the work of Edward Lorenz, who first discovered Chaos, in, of course, weather simulations!
Chaotic systems are not entirely unpredictable, as something truly random would be. They exhibit diminishing predictability as they move forward in time, and this diminishment is caused by greater and greater computational requirements to calculate the next set of predictions. Computing requirements to make predictions of chaotic systems grow exponentially, and so in practice, with finite resources, prediction accuracy will drop off rapidly the further you try to predict into the future. Chaos doesnāt murder cause and effect; it just wounds it!
Now would be a good place for an example. Everyone owns a spread sheet program. The following is very easy to try for yourself.
The simplest man-made equation known that produces chaos is called the logistic map.
Itās simplest form is: Xn+1 = 4Xn(1-Xn)
Meaning that the next step of the sequence is equal to 4 times the previous step times 1 ā the previous step. If we open a spread sheet we can create two columns of values:
Each column A and B is created by writing =A1*4* (1-A1) into cell A2, and then copying it down for as many cells as you like, the same for B2, writing in =B1*4* (1-B1). A1 and B1 contain the initial conditions. A1 contains just 0.3 and B1 contains a very slightly different number: 0.30000001
The graph to the right shows the two copies of the series. Initially they are perfectly in sync, then they start to divert at around step 22, while by step 28 they are starting to behave entirely differently.
This effect occurs for a wide range of initial conditions. It is fun to get out your spread sheet program and experiment. The bigger the difference between the initial conditions the faster the sequences diverge.
The difference between the initial conditions is minute, but the two series diverge for all that. This illustrates one of the key things about chaos. This is the acute sensitivity to initial conditions.
If we look at this the other way round, suppose that you only had the series, and letās assume to make it easy, that you know the form of the equation but not the initial conditions. If you try to make predictions from your model, any minute inaccuracies in your guess of the initial conditions will result in your prediction and the result diverging dramatically. This divergence grows exponentially, and one way of measuring this is called the Lyapunov exponent. This measures in bits per time step how rapidly these values diverge, averaged over a large set of samples. A positive Lyapunov exponent is considered to be proof of chaos. It also gives us a bound on the quality of predictions we can get if we try to model a chaotic system.
These basic characteristics apply to all chaotic systems.
Hereās something else to stimulate thought. The values of our simple chaos generator in the spread sheet vary between 0 and 1. If we subtract 0.5 from each, so we have positive and negative going values, and accumulate them we get this graph, stretched now to a thousand points.
If, ignoring the scale, I told you this was the share price last year for some FTSE or NASDAQ stock, or yearly sea temperature youād probably believe me. The point Iām trying to make is that chaos is entirely capable of driving a system itself and creating behaviour that looks like itās driven by some external force. When a system drifts as in this example, it might be because of an external force, or just because of chaos.
So, how about the weather?
Edward Lorenz, (1917, 2008) was the father of the study of Chaos, and also a weather researcher. He created an early weather simulation using three coupled equations and was amazed to find that as he progressed the simulation in time the values in the simulation behaved unpredictably.
He then looked for evidence that real world weather behaved in this same unpredictable fashion, and found it, before working on discovering more about the nature of Chaos.
No climate researchers dispute his analysis that the weather is chaotic.
Edward Lorenz estimated that the global weather exhibited a Lyapunov exponent equivalent to one bit of information every 4 days. This is an average over time and the worldās surface. There are times and places where weather is much more chaotic, as anyone who lives in England can testify. What this means though, is that if you can predict tomorrows weather with an accuracy of 1 degree C, then your best prediction of the weather on average 5 days hence will be +/- 2 degrees, 9 days hence +/-4 degrees and 13 days hence +/- 8 degrees, so to all intents and purposes after 9-10 days your predictions will be useless. Of course, if you can predict tomorrowās weather to +/- 0.1 degree, then the growth in errors is slowed, but since they grow exponentially, it wonāt be many days till they become useless again.
Interestingly the performance of weather predictions made by organisations like the UK Met office drop off in exactly this fashion. This is proof of a positive Lyapunov exponent, and thus of the existence of chaos in weather, if any were still needed.
So thatās weather prediction, how about long term modelling?
Letās look first at the scientific method. The principle ideas are that science develops by someone forming an hypothesis, testing this hypothesis by constructing an experiment, and modifying the hypothesis, proving or disproving it, by examining the results of the experiment.
A model, whether an equation or a computer model, is just a big hypothesis. Where you canāt modify the thing you are hypothesising over with an experiment, then you have to make predictions using your model and wait for the system to confirm or deny them.
A classic example is the development of our knowledge of the solar system. The first models had us at the centre, then the sun at the centre, then the discovery of elliptical orbits, and then enough observations to work out the exact nature of these orbits. Obviously, we could never hope to affect the movement of the planets, so experiments werenāt possible, but if our models were right, key things would happen at key times: eclipses, the transit of Venus, etc. Once models were sophisticated enough, errors between the model and reality could be used to predict new features. This is how the outer planets, Neptune and Pluto were discovered. If you want to know where the planets will be in ten yearsā time to the second, there is software available online that will tell you exactly.
Climate scientists would love to be able to follow this way of working. The one problem is that, because the weather is chaotic, there is never any hope that they can match up their models and the real world.
They can never match up the model to shorter term events, like say six months away, because as weāve seen, the weather six months away is completely and utterly unpredictable, except in very general terms.
This has terrible implications for their ability to model.
I want to throw another concept into this mix, drawn from my other speciality, the world of computer modelling through self-learning systems.
This is the field of artificial intelligence, where scientists attempt to create mostly computer programs that behave intelligently and are capable of learning. Like any area of study, this tends to throw up bits of general theory and one of these is to do with the nature of incremental learning.
Incremental learning is where a learning process tries to model something by starting out simple and adding complexity, testing the quality of the model as it goes.
Examples of this are neural networks, where the strength of connections between simulated brain cells are adapted as learning goes on or genetic programming, where bits of computer programs are modified and elaborated to improve the fit of the model.
From my example above of theories of the solar system, you can see that the scientific method itself is a form of incremental learning.
There is a graph that is universal in incremental learning. It shows the performance of an incremental learning algorithm, it doesnāt matter which, on two sets of data.
The idea is that these two sets of data must be drawn from the same source, but they are split randomly into two, the training set, used to train the model, and a test set used to test it every now and then. Usually the training set is bigger than the test set, but if there is plenty of data this doesnāt matter either. So as learning progresses the learning system uses the training data to modify itself, but not the test data, which is used to test the system, but is immediately forgotten by it.
As can be seen, the performance on the training set gets better and better as more complexity is added to the model, but the performance of the test set gets better, and then starts to get worse!
Just to make this clear, the test set is the only thing that matters. If we are to use the model to make predictions we are going to present new data to it, just like our test set data. The performance on the training set is irrelevant.
This is an example of a principle that has been talked about since William of Ockham first wrote āEntia non sunt multiplicanda praeter necessitatem ā, known as Ockhamās razor and translatable as āentities should not be multiplied without necessityā, entities being in his case embellishments to a theory. The corollary of this is that the simplest theory that fits the facts is most likely to be correct.
There are proofs for the generality of this idea from Bayesian Statistics and Information Theory.
So, this means that our intrepid weather modellers are in trouble from both ends: if their theories are insufficiently complex to explain the weather their model will be worthless, if too complex then they will also be worthless. Whoād be a weather modeller?
Given that they canāt calibrate their models to the real world, how do weather modellers develop and evaluate their models?
As you would expect, weather models behave chaotically too. They exhibit the same sensitivity to initial conditions. The solution chosen for evaluation (developed by Lorenz) is to run thousands of examples each with slightly different initial conditions. These sets are called ensembles.
Each example explores a possible path for the weather, and by collecting the set, they generate a distribution of possible outcomes. For weather predictions they give you the biggest peak as their prediction. Interestingly, with this kind of model evaluation there is likely to be more than one answer, i.e. more than one peak, but they choose never to tell us the other possibilities. In statistics this methodology is called the Monte Carlo method.
For climate change they modify the model so as to simulate more CO2, more solar radiation or some other parameter of interest and then run another ensemble. Once again the results will be a series of distributions over time, not a single value, though the information that the modellers give us seems to leave out alternate solutions in favour of the peak value.
Models are generated by observing the earth, modelling land masses and air currents, tree cover, ice cover and so on. Itās a great intellectual achievement, but itās still full of assumptions. As youād expect the modellers are always looking to refine the model and add new pet features. In practice there is only one real model, as any changes in one are rapidly incorporated into the others.
The key areas of debate are the interactions of one feature with another. For instance the hypothesis that increased CO2 will result in run-away temperature rises is based on the idea that the melting of the permafrost in Siberia due to increased temperatures will release more CO2 and thus positive feedback will bake us all. Permafrost may well melt, or not, but the rate of melting and the CO2 released are not hard scientific facts but estimates. There are thousands of similar ābest guessesāā in the models.
As weāve seen from looking at incremental learning systems too much complexity is as fatal as too little. No one has any idea where the current models lie on the graph above, because they canāt directly test the models.
However, dwarfing all this arguing about parameters is the fact that weather is chaotic.
We know of course that chaos is not the whole story. Itās warmer on average away from the equatorial regions during the summer than the winter. Monsoons and freezing of ice occur regularly every year, and so itās tempting to see chaos as a bit like noise in other systems.
The argument used by climate change believers runs that we can treat chaos like noise, so chaos can be āaveraged outā.
To digress a little, this idea of averaging out of errors/noise has a long history. If we take the example of measuring the height of Mount Everest before the days of GPS and Radar satellites, the method to calculate height was to start at Sea level with a theodolite and take measurements of local landmarks using their distance and their angle above the horizon to estimate their height. Then to move on to those sites and do the same thing with other landmarks, moving slowly inland. By the time surveyors got to the foothills of the Himalayas they were relying on many thousand previous measurements, all with measurement error included. In the event the surveyorās estimate of the height of Everest was only a few hundred feet out!
This is because all those measurement errors tended to average out. If, however there had been a systemic error, like the theodolites all measuring 5 degrees up, then the errors would have been enormous. The key thing is that the errors were unrelated to the thing being measured.
There are lots of other examples of this in Electronics, Radio Astronomy and other fields.
You can understand climate modellers would hope for the same to be true of chaos. In fact, they claim this is true. Note however that the errors with the theodolites were nothing to do with the actual height of Everest, as noise in radio telescope amplifiers has nothing to do with the signals from distant stars. Chaos, however, is implicit in weather, so there is no reason why it should average out. Itās not part of the measurement; itās part of the system being measured.
So can chaos be averaged out? If it can, then we would expect long term measurements of weather to exhibit no chaos. When a team of Italian researchers asked to use my Chaos analysis software last year to look at a time series of 500 years of averaged South Italian winter temperatures, the opportunity arose to test this. The picture below is this time series displayed in my Chaos Analysis program, ChaosKit.
The result? Buckets of chaos. The Lyapunov exponent was measured at 2.28 bits per year.
To put that in English, the predictability of the temperature quarters every year further ahead you try to predict, or the other way round, the errors more than quadruple.
What does this mean? Chaos doesnāt average out. Weather is still chaotic at this scale over hundreds of years.
If we were, as climate modellers try to do, to run a moving average over the data, to hide the inconvenient spikes, we might find a slight bump to the right, as well as many bumps to the left. Would we be justified in saying that this bump to the right was proof of global warming? Absolutely not: It would be impossible to say if the bump was the result of chaos, and the drifts weāve see it can create or some fundamental change, like increasing CO2.
So, to summarize, climate researchers have constructed models based on their understanding of the climate, current theories and a series of assumptions. They cannot test their models over the short term, as they acknowledge, because of the chaotic nature of the weather.
They hoped, though, to be able to calibrate, confirm or fix up their models by looking at very long term data, but we now know thatās chaotic too. They donāt, and cannot know, whether their models are too simple, too complex, or just right, because even if they were perfect, if weather is chaotic at this scale, they cannot hope to match up their models to the real world, the slightest errors in initial conditions would create entirely different outcomes.
All they can honestly say is this: āweāve created models that weāve done our best to match up to the real world, but we cannot prove to be correct. We appreciate that small errors in our models would create dramatically different predictions, and we cannot say if we have errors or not. In our models the relationships that we have publicized seem to hold.ā
It is my view that governmental policymakers should not act on the basis of these models. The likelihood seems to be that they have as much similarity to the real world as The Sims, or Half-life.
On a final note, there is another school of weather prediction that holds that long term weather is largely determined by variations in solar output. Nothing here either confirms or denies that hypothesis, as long term sunspot records have shown that solar activity is chaotic too.
Andy Edmonds
Short Bio
Dr Andrew Edmonds is an author of computer software and an academic. He designed various early artificial intelligence computer software packages and was arguably the author of the first commercial data mining system. He has been the CEO of an American public company and involved in several successful start-up businesses. His PhD thesis was concerned with time series prediction of chaotic series, and resulted in his product ChaosKit, the only standalone commercial product for analysing chaos in time series. He has published papers on Neural Networks, genetic programming of fuzzy logic systems, AI for financial trading, and contributed to papers in Biotech, Marketing and Climate.
Short summary: AA discussion of the disruptive light Chaos Theory can cast on climate change, for non-specialist readers. This will have a focus on the critical assumptions that global warming supporters have made that involve chaos, and their shortcomings. While much of the global warming case in temperature records and other areas has been chipped away, they can and do, still point to their computer models as proof of their assertions. This has been hard to fight, as the warmists can choose their own ground, and move it as they see fit. This discussion looks at the constraints on those models, and shows that from first principles in both chaos theory and the theory of modelling they cannot place reliance on these models.
On his Website: http://scientio.blogspot.com/2011/06/chaos-theoretic-argument-that.html
chaos…….
When the computer says one thing, reality does another….
…and you still have to make your house payment
Very clearly put, thanks. The test set performance retardation of models with increasing complexity and increasing performance in the training set is an overspecialization; the models have learned to exploit every wiggle of the data in the training set but “unlearned” more general rules that would have helped them survive with a different data set.
Hi,
Very, very interesting. Dr Edmond’s explanation is simple and elegant, and for an old duffer like me, very enlightening.
I am not qualified to comment on the mathematical details, my engineering maths courses back in the ’60’s did not stretch to Chaos theory but my gut feeling is that it makes sense.
Perhaps this paper should be required reading for all the alarmist journalists, especially that English lit. graduate, Richard Black who purports to write scientific articles for the BBC!
Keep up the good work!
Patrick
Excellent article!
Epiphany at last. Now I can find other interests.
I’ll throw this out again to ponder…
You have to go back to the 1930’s to find as many volcanoes honking as we’ve had since 2008.
http://www.volcano.si.edu/world/find_eruptions.cfm
What was going on back then? Chaos! The great depression and…
http://en.wikipedia.org/wiki/Dust_Bowl
Cold waves, heat waves, droughts and floods. Earthquakes and volcanic eruptions. This is what is going on now, it’s trying to dĆ© jĆ vu 1930’s.
Good post.
Another problem with climate models is that they rely on a theory, GHG effect, that does not exist. There are other explanations for Earth’s average surface temperature being ‘too high’ without need of a theory that violates the laws of thermodynamics.
I took a graduate level course in computer modeling in the late 1980s. The professor was a renowed mathematician with a field of study named after him. Most of the course was actually about how you can’t really model large complex systems and a large part of that had to do with a mathematical understanding of chaos and how chaotic systems can appear appealingly predictable for a time but are not. So, the course was basically about how can can develop models
but why you can’t trust them. The summary is models tell you nothing about nature. That course was taken completed in 1987, so presumably they donāt hold to old-fashioned ideas like that anymore ā even if they are grounded in theoretical mathematics
This is as good a summary of chaos and the influence of chaos on times series computer models as I have seen.
A major problem
with scientific method
is thereās more than one.
Wild hypotheses
and unsubstantiated
conjectures abound,
for the āscientistsā
of the climate-change bubble
just want to have fun
and invent figures;
but bubbles will always burst,
or so I have found.
That last chart looks remarkably like a tree-ring series IMHO (without the “blade” of course!). Many thanks to Dr. Edwards for this “chaos theory for dummies”. He mentions modellers averaging output from different runs, which is dubious enough, but we know they also average output from different models to get a kind of “consensus”. If models produce differing results, then no more than one can be accurate. If we assume that one of them is correct, but we don’t know which one, as they’re making predictions about the future, then averaging their output is meaningless. The averaged output would only be something close to reality if the “accurate” model was somewhere in the middle of the range.
Add to this the fact that models don’t include factors which are poorly understood, or indeed unknown (there are many), and “fudge factors” for some others which are poorly quantified, predictions made for time-scales of decades or longer must be highly suspect, if not worthless (as many here have long suspected).
This is one of the best articles this site has seen. If the weather is inherently chaotic at long timescales and can’t be averaged means that climate modelling is much more demanding and unaccurate than generally believed. However there must be some physical stabilizing mechanisms (since the earths temperature has only varied very little on geological timescales) that are not chaotic.
In my opinion the physical constraints set the envelope or conditions where the weather can random-walk. The mainstream climate science has just thought that a minor change in the CO2 undoudably creates a definite rise in the temperature. It just may be the the warming effect may be drowned by the random walk of the climate. Anyways the climate models rely on a mechanistic view of the climate where each action has a certain reaction, based on this article that seem to be an outdated approach.
Chaos is everywhere except the Sun’s output: it is a constant and can’t be a climate driver!!
Blind, and deaf climate modelers… The Sun’s output varies as per the Dalton minimum in 1810 and the Maunder minimum in 1650 and the 11 year cycle from minimum to maximum. But not today!!
Of course, the effects of money, power, politics, and, women, do tend to affect scientific results.
The biggest problem with the climate models is the modellers themselves. They have a biblical faith in the models and treat them as inviolate. The can’t simply accept the fact that they are trying to model a chaotic system and it is a loser’s game. Instead of humility in accepting reality you see only arrogance from these modellers. They demand wholesale changes to the world based on these models which have not proved effective at predicting anything till date.
Excellent article! I hope it stimulates some good discussions.
I also hope it helps put to rest the nonsense that “climate prediction is a boundary value problem” … the GCMs don’t even solve it that way!
So, the GHG effect violates the laws of thermodynamics? Pray tell how…
Evidently, Dr. Edwards is unaware of the three decade old work of Christensen, Eilbert, Lingren and Rans that made it possible to predict surface temperature and precipitation related variables in the western states of the U.S. as much as 3 years in advance. I provide an introduction to the methodology that made this advance possible and compare it to the methodology of modern climatology at http://judithcurry.com/2011/02/15/the-principles-of-reasoning-part-iii-logic-and-climatology/ . In brief, the idea was to construct an information theoretically optimal decoder of a “message” from the future that conveyed the outcomes of weather-related events.
While chaos precludes the success of an approach to long range weather forecasting that is based entirely upon the principles of modern physics, it does not preclude long range weather forecasting. The same might be be found to be true of long range climate forecasting if construction of a climate-forecasing decoder were to be attempted.
Thanks for an excellent article, Andy.
As others have said, it is a shame politicians and so called science journalists do not read the science on this site.
Exactly the kind of discussion I’ve been looking for for the last couple of years. With exactly the common sense results I expected from a lay reading of the field (but with more authority, of course).
Thank you.
It amazes ,me that climate scientist’s can teach so many fields of science, mathmatic’s and Software engineering how not to do it.
And excellent article, top marks to Dr Andy.
Weather is inherently chaotic, but climate (long term averages) is not. The IPCC is quite aware of that, see http://www.ipcc.ch/publications_and_data/ar4/wg1/en/faq-1-2.html for a discussion. Climate is driven, controlled, and limited by energy, and a departure in one direction from balance (up or down) drives the climate back to that balance point.
Your continuing chaos with time is simply the same kind of data repeated, no averaging – the Lyapunov exponent of a chaotic series shouldn’t change – it’s going to vary around the climate mean. You have to look at the data averaged over a sufficient period to see past the effects of fluid turbulence and non-period variations such as the ENSO.
“Once again the results will be a series of distributions over time, not a single value, though the information that the modellers give us seems to leave out alternate solutions in favour of the peak value.”
Really? That’s an amazingly incorrect statement, Mr. Edmonds. See http://www.ipcc.ch/publications_and_data/ar4/wg1/en/ch10s10-5-4-6.html, also http://www.ipcc.ch/publications_and_data/ar4/wg1/en/ch10s10-5-3.html, http://www.ipcc.ch/publications_and_data/ar4/wg1/en/ch10s10-5-4-5.html for examples.
The claim is made that weather is chaotic – yes, it is. But the averages are not. Arguing that we cannot predict the climate is simply a plea to do nothing, to continue with business as usual. You’ve spent a lot of effort on a very deceptive posting.
My apologies, I should have referred to “Dr. Edmonds” in my last post, not “Mr.” – no insult intended, just bad typing.
Dr. Andy Edwards:
Thank you – very interesting.
I am wondering about the difference though between weather and climate.
Clearly, I cannot predict the temperature on a particular day in the winter 6 months hence, as you pointed out.
However, I do know (I think to a very high confidence) that it is going to be cooler on average in the winter than in the summer.
So, does predicting the climate (of a season) for example, differ with respect to chaos, than does predicting the weather?
Climate models aren’t trying to predict the weather years in the future, but the climate, and I was wondering if you think that makes any difference?
PM says:
June 13, 2011 at 8:43 am
In my opinion the physical constraints set the envelope or conditions where the weather can random-walk. The mainstream climate science has just thought that a minor change in the CO2 undoudably creates a definite rise in the temperature. It just may be the the warming effect may be drowned by the random walk of the climate. Anyways the climate models rely on a mechanistic view of the climate where each action has a certain reaction, based on this article that seem to be an outdated approach.
Its not so much a random walk as a Levy Flight as some of the random motions are far more than others. ‘Weather’ is actually a term for a complex multiplicity of interacting chaotic systems each in Levy flight. The bounding conditions are those of the current interglacial ‘attractor’ and the bounds are defined by the current limits of the variations of the inter-reacting systems. If some of these chaotic systems happen to act at the right time or periodocity with the right values then they can move the entire system to a different attactor and we can rapidly enter another ice-age.
I am so sorry to rain on such a happy parade, but the author advertises his cluelessness about scientific method, not to mention Kepler’s work, as follows:
“A model, whether an equation or a computer model, is just a big hypothesis. Where you canāt modify the thing you are hypothesising over with an experiment, then you have to make predictions using your model and wait for the system to confirm or deny them.”
Hypotheses are used for prediction and explanation. If they do not predict phenomenon X then they cannot explain phenomenon X. If they do not explain phenomenon X then they cannot predict phenomenon X. We will see this in Kepler’s Laws below.
“A classic example is the development of our knowledge of the solar system. The first models had us at the centre, then the sun at the centre, then the discovery of elliptical orbits, and then enough observations to work out the exact nature of these orbits.”
Isn’t there a bit of hand-waving about this “discovery of elliptical orbits?” Did Kepler stumble over them on his way to the refrigerator? No, Kepler’s genius created his first hypothesis, that all planetary orbits are elliptical with the sun at one foci, just as Zeus gave birth to Athena from his forehead. Then Kepler applied his hypothesis to the data that had been collected mostly by Tyco Brahe and found that all the data fit the elliptical orbits.
Let’s review the steps. There is some data that needs organization. Existing hypotheses do not cover them. Kepler freely creates a hypothesis, now known as Kepler’s First Law, that exactly covers an important regularity in the data. However, there is more than prediction. The elliptical path will serve in powerful explanations of how the planets actually move and why they are observed to do so. Kepler’s three Laws will explain the observed retrograde motion of Mars, something that had been a puzzle for thousands of years. That comes below.
“Obviously, we could never hope to affect the movement of the planets, so experiments werenāt possible, but if our models were right, key things would happen at key times: eclipses, the transit of Venus, etc.”
No. We can conduct experiments. It’s just that they are passive, not active. Once Galileo has a telescope, he will conduct experiments by predicting the behavior of planetary phenomena and then testing the observations with his telescope.
“Once models were sophisticated enough, errors between the model and reality could be used to predict new features. This is how the outer planets, Neptune and Pluto were discovered. If you want to know where the planets will be in ten yearsā time to the second, there is software available online that will tell you exactly.”
There was no need for more sophistication in Kepler’s hypotheses, though there was a need for Newton’s calculus. After inventing the calculus, Newton was able to deduce Kepler’s Laws from his Law of Universal Gravitation.
Our author drops the topic here.
Kepler’s Second Law is that a planet’s speed varies in direct proportion to the area swept by a line from the sun to the planet as the planet travels in its orbit. His Third Law specifies the size of orbits and is a predecessor of Newton’s Law of Gravitation. Given these three laws, Kepler was able to predict the observed path of Mars, the one true anomaly in the astronomy of that time, and to explain it. He explained that Mars is observed to stop, back up, and move forward again over a period of months because Mars and Earth are on intersecting elliptical paths and are changing speeds. In other words, he could explain to his fellows that Mars’ observed path is analogous to the observed path of a sailboat that temporarily overtakes our own, then seems to slow because our sailboat enjoys a burst of speed, and then enjoys its own burst of speed and moves past us.
The big point to notice is that Kepler’s hypotheses are used both for prediction and explanation. In all of science, hypotheses are useful for both explanation and prediction. The two always go together. In addition, each hypothesis is “fleshed out” in that it contributes content to explanations of planetary motions and predictions about those motions. In other words, each hypothesis has its own integrity. Of course, it follows from the preceding that each hypothesis is testable and falsifiable on its own.
By contrast to scientific hypotheses and sets of them, a computer model is analogous to a system of deduction. As anyone who has studied a system of deduction knows, such systems are not unique. I can create a deductive system that implies all and only the true statements that I specify. For example, Hans Reichenbach wrote a book called “Axiomatization of the Theory of Relativity” in which he identified a set of most basic principles and deduced all of the other consequences of Einstein’s theory from them. Such deductive systems are not unique. Reichanbach, or anyone, could have chosen an alternative set of deductive principles for his axiomatization. It would have yielded all and only the same truths, but the proofs of those truths would have differed. Difference in proof does not matter. If the two systems yield all and only the same truths then they are identical The same reasoning applies to computer models.
There are no hypotheses in computer models. You cannot remove one proof, one set of proofs, or even a heuristic for making equations solve and present it as a hypothesis. It has no more content than a set of deductive inferences in formal logic. For that reason, computer models cannot answer the scientific question Why? Hypotheses can answer scientific questions, as shown in the case of Kepler. Genuine hypotheses always give explanation along with prediction. There is never any scientific reason for coding physical hypotheses into a computer model. If you have the hypotheses, you do not need the model and you have explanations. You might axiomatize a physical theory for the purpose of searching for consequences that you have overlooked but that is a matter of accounting not science.
What can a computer model do? You can program it to produce a particular set of results. It can be used to retrodict (predict the past). Any two models that retrodict the same past are identical regardless of their internal construction. Can it predict? That is a matter of programming it to produce a set of results that you expect to happen in the future. You have to program those in. The computer will never be able to do more that produce results that you program. The model does not contain physical hypotheses and cannot be made to offer explanations for the results you program it to produce.
KR says:
“Arguing that we cannot predict the climate is simply a plea to do nothing, to continue with business as usual. Youāve spent a lot of effort on a very deceptive posting.”
Actually, the IPCC’s position is this:
Not one GCM [computer climate model] predicted the flat to declining temperatures over the past decade, proving that they cannot accurately predict the climate. Readers can decide for themselves who is posting deceptively.
And the most reasonable course of action at this point is to do nothing. There is no evidence whatever of global harm from increased CO2, and much evidence of global benefit. Thus, CO2 is harmless and beneficial. Under the circumstances, doing nothing is the only rational response.
Excellent article, thanks. This article is succinct enough to help explain practical mathematical chaos to the lay person.
KR says:
June 13, 2011 at 9:45 am
“The claim is made that weather is chaotic ā yes, it is. But the averages are not. ”
Averaging is only dampening high frequencies in a well defined way – a lowpass filter operation- the power spectrum of chaos is not the (regular) power spectrum of noise but can have strong (and unpredictable) spikes. You don’t get rid of these spikes with a lame old general lowpass filter – you can only dampen them a little but they’ll still stick out. I wonder how low-passing should help you there.
“Climate scientists would love to be able to follow this way of working. The one problem is that, because the weather is chaotic, there is never any hope that they can match up their models and the real world.
They can never match up the model to shorter term events, like say six months away, because as weāve seen, the weather six months away is completely and utterly unpredictable, except in very general terms.”
We need to be clear about what is being asserted. Someone observed that weather six months away always strikes him as unpredictable. That does not mean that it is unpredictable. More important, it also does not mean that the important features of the weather are being observed. Clearly, then, the assertion that some phenomenon is chaotic requires some evidence beyond the observations of several people that it strikes them as chaotic. Weather might seem unpredicatble. But climate science and meteorology are both in their infancy. What we call weather or climate today is not likely to be called weather or climate in ten years.
John Marshall says: June 13, 2011 at 8:31 am
“Another problem with climate models is that they rely on a theory, GHG effect, that does not exist. There are other explanations for Earthās average surface temperature being ātoo highā without need of a theory that violates the laws of thermodynamics.”
Please state exactly what part of the GHG effect you think violates what law of thermodynamics.
Smokey – You should have posted the next few very relevant lines from that piece: “The most we can expect to achieve is the prediction of the probability distribution of the systems future possible states by the generation of ensembles of model solutions. This reduces climate change to the discernment of significant differences in the statistics of such ensembles.” (http://www.ipcc.ch/ipccreports/tar/wg1/505.htm)
That’s the core of Monte Carlo statistical estimation, Smokey. Predicting the long term average climate, not detailed weather.
As to “no evidence whatever of global harm from increased CO2, and much evidence of global benefit.”, I’ll have to disagree based on every study I’ve seen.
Terry Oldberg says:
June 13, 2011 at 9:14 am
Evidently, Dr. Edwards is unaware of the three decade old work of Christensen, Eilbert, Lingren and Rans that made it possible to predict surface temperature and precipitation related variables in the western states of the U.S. as much as 3 years in advance. I provide an introduction to the methodology that made this advance possible and compare it to the methodology of modern climatology at http://judithcurry.com/2011/02/15/the-principles-of-reasoning-part-iii-logic-and-climatology/ . In brief, the idea was to construct an information theoretically optimal decoder of a āmessageā from the future that conveyed the outcomes of weather-related events.
While chaos precludes the success of an approach to long range weather forecasting that is based entirely upon the principles of modern physics, it does not preclude long range weather forecasting. The same might be be found to be true of long range climate forecasting if construction of a climate-forecasing decoder were to be attempted.
If that worked then it would be in use right now – the users of it would be as rich as Croesus!
Have you any idea what construction companies would give for that kind of information let alone farmers?
You probably should demonstrate it at the ocean front property you have in Kansas.
KR,
Either provide testable empirical evidence of global harm from CO2 per the scientific method, or admit that there is no such evidence. Keep in mind that “studies” are not evidence, and neither are computer models.
And if you want solid evidence of the benefits of increased CO2, just ask.
Well, now I know a little more about chaos. Thank you.
Nevertheless, the physics of the planet demands that angular momentum and energy be conserved, that radiation depend on temperature, that convection in the oceans be constrained by the viscosity of water and changes in density forced by temperature … so all these provide boundaries within which the chaos will be confined. I suppose that is why life has survived for all these aeons.
Ed Mertin says: June 13, 2011 at 8:30 am
“You have to go back to the 1930ā²s to find as many volcanoes honking as weāve had since 2008.”
According to this link of major volcanoes, the 1930’s were hardly an exceptional decade for volcanoes.
http://online.wsj.com/article/SB10001424052748703465204575208412972387390.html
Can you provide statistics from your source (not just the intro page listing a few interesting volcanoes) that show how the 1930’s were any more severe than other decades?
Excellent article and thank you.
@KR
“Weather is inherently chaotic, but climate (long term averages) is not. The IPCC is quite aware of that, see …”
The above statement makes no sense whatsoever. If the climate can change from one extreme to another, Ice Age to Holocene and there is no defined time scale or defined cause then it must be chaotic?
Sorry if I am missing the point.
June 13, 2011 at 10:07 am
KR says:
āArguing that we cannot predict the climate is simply a plea to do nothing, to continue with business as usual. Youāve spent a lot of effort on a very deceptive posting.ā
That is exactly what I am arguing. Don’t touch that dial. Do not pass go and do not collect $200. Resist the urge to think that you can legislate climate when climate is more chaotic than the stock market and look how that worked out for the likes of Long Term Capital.
KR
Since the IPC pretty much stakes all on the value of models (few of which agree with each other, let alone nature) it would be hard to believe that they would utter a word which would in any way question the utility of such models.
In any event, as the author notes:
“So, this means that our intrepid weather modellers are in trouble from both ends: if their theories are insufficiently complex to explain the weather their model will be worthless, if too complex then they will also be worthless.”
Once you have a highly complex system, it can behave very much much like a chaotic system in all ways which matter. Current models have rudmientary simulations of things like cloud cover, and they do not and can not deal with biological feedback systems. In the olden days the carbon cycle was a thing we studied, don’t you know. Running a 16km3 grid simulation might be convenient computationally, but since most features with matter are much smaller than 16km3, you pretty much immediately lose the opportunity to predict anything.
You know, there is a lot more money at stake in financial models and none of those have been shown to be worth anything either. Only climate science presumes what all other real scientific disciplines already know: models are nice and fun, but nature is the teacher.
How did Kepler end up as part of the conversation?
Nice explanation. You should point out that the example given using the logistics formula is the simplest of many, and that the first term (4) varies and equates to being the “driver” of the system. Other values will result in different behavior, as I’m sure you know, and it’s interesting to increment this value over the series.
Excellent article Dr. Edwards.
A couple notes on things that impact the edges:
1) The recurring case of the climate modellers taking results from widely disparate models and combining them – not into a ‘single model ensemble’ but into a ‘all available models ensemble’ – comes from the methodology of the individual models. “Combining into an ensemble” is fundamentally what they’re doing on a per-model basis. But all it does functionally is completely prevent the invalidation of -any- model.
2) The one area that seems fundamental in the methodology is the hindcasting to demonstrate any skill whatsoever. And yet the central theme of ‘historical temperatures are flat!’ of Mann et al. is less and less viable over time.
Ian W says:
June 13, 2011 at 10:00 am
Its not so much a random walk as a Levy Flight as some of the random motions are far more than others.
Thank you for this insight, I just learned something new and interesting. Indeed the some changes have more effect than others. The main problem with currernt climatological research is that conclusions are drawn too fast. I think that the main problem is that they have focused too much on the relatively simple parts as CO2 absorption and emission and not at the fact that there are number of phenomena varying at very different time-scales. The main reason for the poor predictive power of the models is that they have lumped the effects of multiple phenomena and explained them by just one component of the system. It’s a simplistic mechanistic approach to a very complex system. Its just sad that major economic and political decisions are based on predictions based on a science that isn’t mature.
KR says: “Weather is inherently chaotic, but climate (long term averages) is not. The IPCC is quite aware of that…”
The IPCC was also aware that all the Himalayan glaciers are going to disappear by 2035.
In regards to long term averages and behaviors:
Dr. Edmonds, are you stating that the chaotic attractor of the weather does not have a mean and standard deviation? If so, that would make it unlike just about every chaotic system I’ve seen. The weather cycles around the attractor, but summer is predictably warmer and winter predictably cooler, as the average changes.
Stacy – Long term variations in climate come from changes in forcings (input/output changes in energy balance), such as insolation, orbit, etc.. For the ice ages the Milankovich cycles of orbital precession appear to be the major cause of climate change. And those changes are pretty well defined and timed. Right now, based on our position in the tapering of the interglacial portion of the Milankovich cycle, we should be seeing cooling temperatures. We’re seeing warming, because we’ve added another forcing – extra CO2, to a level currently 100 ppm greater than anything in the last 800K years.
I spent my nuclear engineering career as a modeler. Nuclear systems are simple systems compared to climate and I had a tough enough time with them to immediately be suspicious of climate models when AGW became the fad of the day. So I don’t disagree with the arguments and conclusion made here. But I think we should distinguish between chaos and complexity. Chaotic behaviour can arise from a few simple equations subject to minute changes in initial conditions. In contrast, complex systems are characterized by many equations with many parameters that are not precisely known. Even starting from the same initial conditions with computers of infinite word length, the variability of the parameters is enough to make the results unpredictable. Of course the initial conditions are also imprecisely known, computer word length is not infinite, discretization is limited and so on. Small wonder climate is not predictable. Negative feedback can provide bounds for models, including climate models and my guess is that is the case for us – else we would not have made it this far.
Anyway, my point is that climate is complex enough to not need chaos to ensure unpredictability. If we could restart the planet with exactly the same initial conditions at some point in the past, it would be the influence of small differences in parameters, caused by small external perturbations, that would ensure a different evolution. It may or may not be bounded depending on the path evolution. That is complexity, not chaos.
Semantics perhaps but when I think of chaos, I think of sensitivity to initial conditions and when I think of complexity I think of sensitivity to the details of the equations, parameters and boundary conditions. Of course complex systems can be chaotic and chaotic systems can be complex….
At any rate, it is negative feedback and overall energy considerations that provide climate bounds and provide some level of predictability, in spite of any chaotic or complex nature of the models.
Bill
Dr.Edwards, thank you for a fascinating article. This math challenged individual actually understood most of the concepts raised. Actually tried your excel experiment. Pretty neat. To K.R.: 1) If all our scientific capabilities were available during the LIA, would the current warmer period be “averaged out”? 2) Why are negative feedbacks “averaged out” but not positive ones?
Alan S. Blue
The interesting thing with ‘back casting’ is that any model which cannot ‘back cast’ the data used to craft and groom (‘train’) its algorithms clearly isn’t worth spit, yet they do show up in the climate science world. Hoards of PhDs in Economics have been granted for the development of models which have the sole merit of ‘modeling’ the past. They have no apparent utility in predicting the future (if they did, according ot economic theory, they would not be published!).
It is bizarre beyond belief that a model’s ability to predict the past is considered somehow an indicator of its ability to predict the futre. If it can’t predict the past it is a piece of garbage, but not necessarily any worse at predicting the future than a model which can.
Smokey said “And if you want solid evidence of the benefits of increased CO2, just ask.”
OK, I’m asking. I want t osee the kind of thing you onsider evidence. Then I, or KR, or anyone else, can provide evidence of the same kind. e.g.. if, you post a photo, we can post a photo, if you post a graph, we can post a graph. Does that seem fair?
Bill garland:
I am not arguing with you, however, while perhaps the physical system that is the earth may not be in isolation a chaotic system, I am pretty sure there is no doubt some of the feedbacks are. The easiest one to consider is the biological feedback associated with the carbon cycle. My prof would use population dynamics over and over again as an example of the limitations of modeling. Now, a modeler might be able to assume certain population response but the real response would ultimately depend on starting conditions of everything from potasium in the soil to species’ adaptability.
The way I think of it, since the earth’s climate is a dynamic system, and modeling the impact of CO2 would have to take into account *all* (not a few and not average) biological impacts, and because those are inherently chaotic, that makes the system itself likely chaotic.
So even if you did know all the starting conditions, etc., the biosphere would have a major say and you can’t model that.
Mingy,
Mostly agree. I was really referring to extending beyond the training data. But I do have issue with this one, mostly that it leaves out ‘in a Monte Carlo simulation’:
“It is bizarre beyond belief that a modelās ability to predict the past is considered somehow an indicator of its ability to predict the futre.”
In non-climate science, the ‘training data/test data’ split is applied pretty regularly. In non-Monte Carlo situations, the ‘ability to predict the past’ on the test data is generally a -very- good indicator of something’s ability to predict the future. Think: Models of balls rolling on inclines, pendulums, etc.
The issue in Monte Carlo simulations is that you can fall into the trap of ‘model shopping’. That is “Hmm, this one didn’t work at all well. I’ll raise this parameter a bit, lower that one, and try, try again!” Repeat.
If you’re doing that, (which people are), you’ve fundamentally incorporated the -test-data- into the training period – because you’ve become part of the model.
But in a system where any chaos is swamped by deterministic action, testing models on ‘historical data’ is -the- crucial step.
Dr, Edwards,
A very thought provoking article –thanks for posting!
I do have a few specific comments/critiques
1) “The point Iām trying to make is that chaos is entirely capable of driving a system itself and creating behaviour that looks like itās driven by some external force. ”
I wouldn’t quite say that. Take a simple example of chaotic motion, the double pendulum. The motion is “driven” by forces, not by “chaos”. The chaotic motion is a RESULT of the sensitivity of the motion to the initial conditions.
Chaos itself does not “drive” anything.
2) “What this means though, is that if you can predict tomorrows weather with an accuracy of 1 degree C, then your best prediction of the weather on average 5 days hence will be +/- 2 degrees, 9 days hence +/-4 degrees and 13 days hence +/- 8 degrees, so to all intents and purposes after 9-10 days your predictions will be useless. ”
Doesn’t this assume that you know nothing about the TRUE drivers of weather and are basing your predictions simple on the short-term chaotic behavior? Based on your numbers above, a prediction for the temperatures 1 year from now would be +/- millions of degrees, but I am confident I could predict the high temperature of your location 1 year from now within +/- 20 C.
3) “Each example explores a possible path for the weather, and by collecting the set, they generate a distribution of possible outcomes. For weather predictions they give you the biggest peak as their prediction. Interestingly, with this kind of model evaluation there is likely to be more than one answer, i.e. more than one peak, but they choose never to tell us the other possibilities. …
… the information that the modellers give us seems to leave out alternate solutions in favour of the peak value.”
Of course, it makes sense to select a “typical” result of the calculations to report as the expected result. For example, using your spreadsheet equations, If I wanted to know the “typical” result of the calculation for initial condition of 0.3 after 25 steps, I could try values of 0.29999990. – 0.30000010 in steps of 0.00000001 and find the average (or median or geometric mean or what ever other version of “typical” is appropriate).
Beyond this point, weathermen OFTEN choose to tell us the other possibilities — eg “tomorrow will be warm and sunny, with a possibility of scattered thunderstorms”. I suspect that the papers on climate modeling ALSO give more than the “typical” value — information like the standard deviation or the actual distribution of the results.of the runs. Do you have data to back up your perception of climate science as only presenting one peak value??
4) “Chaos, however, is implicit in weather, so there is no reason why it should average out. ”
I can think of two reasons. Chaos does not mean completely random. There are “attractors” in many systems that lead to somewhat predictable paths. Also, weather and climate are constrained by actual physics. Energy must be conserved; net energy flow must be from warm to cool, … . If a weather system blows a little more north than south, the rain will fall in different, unpredictable places, but the total rainfall is still constrained.
I’d love to hear you response on these comments. There is always so much to learn! š
John B, glad you asked. Here is real world evidence that CO2 promotes plant growth, showing that added CO2 has increased agricultural productivity:
click1
click2
click3
click4
click5
click6
click7
click8
More CO2 is benefitting the biosphere. And there is no evidence of global harm due to CO2. Thus, CO2 is harmless and beneficial. QED
Contrary to what the author wrote it remains a tenet of physics that any system is predictable given enough information about it both forward or backward in time. In point of fact there is no such thing as “time”. Time itself is an illusion created by the law of entropy. Chaos is an illusion created by lack of information. “Quantum uncertainty” is a misleading term. It does not render the universe non-deterministic. Quantum uncertainty is the inability to simultaneously measure two or more properties such as position and velocity. As a measurement of position gets more and more precise simultaneous measurement of velocity gets less and less precise. Thus if we have obtain complete information about the position of a particle we cannot obtain complete information about its velocity. This is an observational constraint not a constraint of nature. The quantum wave formula is unitary and time reversible.
This common misconception that the universe is chaotic by nature (non-deterministic) is well described in a letter from an MIT/Harvard physicist to a science writer at the New York Times who made the same mistaken assumptions as the author of this WUWT article.
http://bemasc.net/wordpress/2008/07/17/decoherence-theory/
“And if you want solid evidence of the benefits of increased CO2, just ask.”
For most of the history of the earth, including most of the past 100 million years, CO2 levels were higher than at present and the oceans were more acidic, and life did very well.
If you accept that CO2 drives temperature, that means the the low levels of CO2 in the past few million years are a contributing factor to the ice ages we have experienced, which are not in general beneficial to life.
Thus, if you believe CO2 drivers temperature, this leads to the conclusion that increasing CO2 makes the next ice age less likely. It also is mo0re natural, as it returns the earth to a state more similar in which life evolved, The low levels of CO2 over the past few million years being unlike most of the past history of the earth, it represents an unnatural state and thus a stress to life rather than a help.
The caustic nature of the present oceans can be traced to the low levels of CO2 over the past few million years, which is also a threat to life as caustic environments dissolve living material. A return to higher CO2 levels and more neutral oceans would therefore likely assist life, because for most of the past 100 million years the oceans have been more acidic than they are now and life has evolved to take advantage of those conditions.
The ice ages were beneficial to polar bears and penguins, but that is not a good reason to keep CO2 levels low enough to bring on the next ice age. Most life forms are not cold weather adapted. Human’s did comparatively better during the ice ages because we domesticated fire, which gave us a survival advantage. If we ban fossil fuels this advantage may not be there for the next ice age.
Dave Springer says:
June 13, 2011 at 11:27 am
“Contrary to what the author wrote it remains a tenet of physics that any system is predictable given enough information about it both forward or backward in time.”
Enough information = the complete state of the system at one moment in time.
“Chaos is an illusion created by lack of information. āQuantum uncertaintyā is a misleading term. It does not render the universe non-deterministic. Quantum uncertainty is the inability to simultaneously measure two or more properties such as position and velocity.”
In other words, you will not get enough information to predict the development of the system – Heisenberg’s uncertainty relationship prevents it.
But besides that, even if you could get the information you couldn’t run the model. See Wolfram’s principle of computational irreducibility – how would you emulate a physical system? You’d need a larger system to run the emulation on; you can’t take shortcuts due to the chaotic (not, i didn’t say nondeterministic) nature of the system. The emulating system must be bigger than the original system because it runs a program on a kind of universal computer; so to emulate the Earth you’d need a computer bigger than the Earth.
Dave, Chaos is not an illusion but a classifier. It classifies a system as having a deterministic but chaotic behaviour.
This article actually contains thoughts, clearly formulated and explained.
What a miracle.
“We know of course that chaos is not the whole story. Itās warmer on average away from the equatorial regions during the summer than the winter. Monsoons and freezing of ice occur regularly every year, and so itās tempting to see chaos as a bit like noise in other systems.
The argument used by climate change believers runs that we can treat chaos like noise, so chaos can be āaveraged outā.”
In a binary chaotic system one can easily, especially during summer vacation, trig the system with a beer, or a six pack or two, and super easy just remove one bit from the equation, thusly leaving only 50%, or one bit if you really must know, that ends up being the complete, and very only, answer.
Thusly, summer, autumn, winter, and spring, hurricane season, indian summer, autumn rains, and a 20 foot snow pack at midsummer, is the noise. It just tend to happen, pretty much always, every year, pretty much around the same time each year, what with us, on puny planet earth, being on an ever merry go round around old “silent” Sol.
The irony, of course, is that you’re a right any how, the hippies thinks they can average it out anyway.
KR: “Any two models that retrodict the same past are identical regardless of their internal construction”
Rubbish!
So with the volcano erupting spewing S02 into the stratosphere we should believe thusly:
1. If there is a drop in temperature that is within “normal variations” for the past millions of years, then the drop
is “explained” by “natural variation”. There is no need to look at the volcano as the source of the cooling. After all the cooling, if it occurs, will be within the range of “natural variability”. Nothing here to see, the earth cools all the time.
2. If a model successfully predicts this cooling, there is also no point to that. Chaos rules and that means we can say nothing about the trajectory that external forcings have on the system. Heck, even if the sun were to double its output we cannot predict that it will get warmer, because well, Chaos rules.
In regards to cost/benefit balances for increased CO2, I’ll point you at http://www.skepticalscience.com/global-warming-positives-negatives-intermediate.htm
This is a summary of some of the benefits (yes, there are some) and the costs (and yes, there are costs too) of CO2 driven climate change, with references to 60+ papers (pictures, charts, data, analysis, sorry, I don’t think there are any cartoons) on the issues. More than I care to type into a blog posting – go read the references.
In regards to the specifics of plant growth and CO2 – we can expect limited increases in productivity of some plants (considerable differences between C3 and C4 cycle plants, mind you), although most of the increase in plant mass will be in the woody/non-food portions of the plants. Plant ranges will change considerably – the California Central Valley is expected to lose ~50% productivity over the next century, and that currently provides 8% of US food production. Pests will also change ranges with temperature.
Total change in plant productivity is expected to be only a slight increase.
This was an excellent post, but I think you might have missed the bigger differences between the nature of chaos in the weather and in the climate. Two very different things that display deterministic chaos in two very different ways and different scales of time and space. This statement you made gets to the point:
“Climate scientists would love to be able to follow this way of working. The one problem is that, because the weather is chaotic, there is never any hope that they can match up their models and the real world.”
Climate scientists (versus meteorologlists) are not particularily concerned with the chaotic nature of weather, so that kind of chaos is not a problem for them. Climate scientists are concerned with long-term forcings that can tip the climate one way or another into a different regime. To see the difference between the chaos of weather and the chaos of climate it’s best to use a simple analogy of the sandpile as way of illustration that many can readily understand.
Sandpiles are often used for illustrating a chaotic system, and are in fact studied directly for what they can teach us about the nature of deterministic chaos. See for example: http://pre.aps.org/abstract/PRE/v52/i6/pR5749_1
http://www.jmu.edu/geology/ComplexEvolutionarySystems/SOC.htm
So when it comes to weather and climate, we could the sandpile analogy to say that the general state of the sandpile could be likened to the climate, and the general state of any indivdual grain in that pile could be likened to a specific weather event or forcing within that overall climate sandpile. Thus as the sandpile changes over time, by the adding of grains of sand for example, the nature of weather patterns within that sandpile change over time. Milankovitch forcings (the primary natural driver of long term climate changes) could be likened to adding grains of sand to the climate sandpile, very slowly, one at time. At some critical point, not random at all, but completely determined by frictional forces working between the grains of sand and gravity, the addition of just one grain of sand causes a collapse of the sandpile until it reaches a new point of equalibrium. This is exactly what happens with Milankovtichs cycles. It is the slow change in solar insolation on the planet, one year at a time that eventually reaches some critical point that there a collapse of the climate sandpile and a new glacial period begins or ends.
We know that there are points of criticality in the climate. All things being equal (i.e. the composition of the atmosphere, the strength of solar radition, etc.) we can be pretty confident that we know what the climate of the earth will generally be at a given point in the Milankovitch cycles. And so, though it may be remarkable it is true that it may be harder to predict the weather two weeks from now than it is to predict the earth’s climate 25,000 years from now based on Milankovtich forcings. What we currently don’t know as well as we know Milankovtich cycles is the sensitivity of the climate to other “grains of sand” that may play a role in climate such as variations if cosmic rays or the addition of small quantities of GH gases each year to the atmosphere.
One mistake that seems almost universal in this forum is that we talk about climate and weather as if they have some inherent meaning that we all understand. I don’t think so. We should focus on some reasonably well-known natural phenomenon, such as La Nina, and create a description of it which embodies descriptions of the several natural regularities that make it up. Each description of a natural regularity is one or more physical hypotheses. These natural regularities are predictable. They are also the physical reality that makes up La Nina. Once the system or regularities is well understood then La Nina will be well understood. Then we can decide whether to call it weather or climate, but I think that question will have become irrelevant.
Here’s some more good reading on quantum determinism. It was a scholarly war between arguably the two greatest living theoretical physicists: Leonard Susskind and Stephen Hawking.
http://en.wikipedia.org/wiki/Susskind-Hawking_battle
In a nutshell Hawing argued that information could be destroyed, or at least permanently removed from the observable universe, when it fell through the event horizon of a black hole. This situation violates quantum determinism and Susskind objected saying that outcome violates a most deeply held tenet of physics – determinism.
http://en.wikipedia.org/wiki/Susskind-Hawking_battle
It took Susskind 28 years to prove it to Hawking who famously conceded to their bet and Hawking paid off giving Susskind an encyclopedia of Baseball. I followed the debate on and off all those years.
My emphasis:
Information appears to follow one of the basic laws of physics most of us will recoginize:
Conservation of Energy – energy may not be created nor destroyed, it can only change form.
This implies determinism and determinism implies that chaos is no more than an illusion created by insufficient information.
Thus chaotic weather and climate are manifestations of human ignorance rather than a consequence of the laws of nature.
“This common misconception that the universe is chaotic by nature”
The three body problem shows that the universe is chaotic at a fundamental level. A deterministic universe implies that the universe is finite. That at some level of scale we will find the end of the inverse. This would then allow us to predict everything. An infinite universe implies that there is no level of scale that is sufficient to make the universe deterministic. Like a fractal, no matter how big or how small our ruler, the universe extends further, leading to a fundamentally chaotic nature.
Smokey,
So, I followed yout first link, to a photo of trees growing better under increased levels of CO2. I then googled “more co2 is good”, the 4th result was to to this page:
http://mind.ofdan.ca/?p=2374
It’s contains a youtube video, which goes into some detail about why increased CO2 is not necessarily better for plants, and also describes some harm being done, and expected to be done, in the near future as a result of increased CO2. Please, watch the whole thing. (BTW, I don’t know anything about “Dan”, but the video is by “Peter Sinclair”, whom I suspect you may have heard of.
Conclusive? No, of course not. But it as least as good as your photo. Better, because he cites evidence and doesn’t just appeal to a superficial reaction to a single photograph of a tree grown in unrealistic conditions.
Really, if a grade schooler supported an essay with evidence like your photo, what sort of marks do you think they would get? I suppose it would depend on whether their teacher was a “skeptic”.
How about somoeone else addresses one of Smokey’s other pieces of “QED” evidence?
Reposted from another discussion on this topic
The difference between a chaotic initial value system (weather) and a boundary limited system (climate) is the difference between trajectory details and trajectory averages.
Weather is highly susceptible to initial conditions, and predicting weather and it’s details (rain or not, where will the pressure systems go?) requires detailed information and a lot of computing power to predict even a few days out. This is a very hard problem, as even a small error or approximation of initial conditions will inevitably cause the prediction to deviate from reality a few days out.
Climate, however, is a boundary condition system. We don’t know what days in March 2012 it will rain, but we can predict even this far out what the average temperature is likely to be with high certainty. When a particular bit of weather departs from the averages, it will (statistically) return to the average, and spend some time on the other side as well.
So why is climate a boundary limited system? It depends on total energies. If energy leaving the climate exceeds energy coming in, we’ll get colder, and less heat will leave – back to the average as determined by the insolation and thermal radiation. If we have a hot season, the climate will radiate above the average, and we’ll cool down. The weather will vary around those averages in a difficult to predict way, but it will vary around the averages determined by energy conservation! And those averages are what climate predictions are about.
Boundary conditions drive any deviations back to the averages for that system. So while we cannot state whether it will be sunny on your birthday – we can still note that winters will be colder than summers, and that reducing the amount of energy leaving the climate at any temperature (with GHG’s) will make the average temperatures higher.
Sorry, left out the link to the related discussion on weather, climate, and chaos
http://www.skepticalscience.com/chaos-theory-global-warming-can-climate-be-predicted-intermediate.htm
Chaotic does not imply uncontrollable.
As an example consider the neutron flux at the core of a nuclear reactor. Fission is occurring randomly throughout the core. We cannot say with certainty how many neutrons will be emitted from each fissioning atom, but we can say, on average 2.5 neutrons will emit due to fission. We can now control the flux by adjusting neutron leakage and absorption around a flux setpoint. We can also control the flux shape via neutron suppression in some areas while unsuppressing it in another area (flux tilting). So in this specific case, we can model chaotic and complex systems. GK
so in practice, with finite resources, prediction accuracy will drop off rapidly the further you try to predict into the future
Isn’t this equally true of a false model? How then could one ever prove a given chaotic model fits reality?
Good article, but you should correct some misinformation… Chaos is much older than Lorenz… Bernoulli found it in his work… http://sgtnd.narod.ru/science/hyper/SAW/eng/saw.htm
And… Chaos is currently the topic of research for new areas of communication systems…
Lastly… Chaos is deterministic. Given a particular identical set of initial conditions, the same result will occur. However, if you don’t know the entire set of initial conditions, a chaotic system remains non predicative — not non-deterministic…
BTW there’s a new book coming out on chaotic communication system design…
Synchronization Techniques for Chaotic Communication Systems
ISBN: 3642218482
Sorry for the formatting, I’m very tired today…
A 1960s geography text is a pretty good predictor of today’s climate, and probably won’t be that far off in the year 2100 barring a climate tipping point being reached such as a plunge to a new ice age. That said, climate is a nonlinear dynamic system, but the models don’t even meet basic requirements for linear systems. They have documented errors larger than the phenomenon of interest, namely an energy imbalance of under 1 W/m^2. Beyond that, even when combined into ensembles they have documented correlated biases on the same scale as the change in CO2 forcing. In linear systems the correlated biases would not cancel out like it is hoped that uncorrelated errors would. In a nonlinear system, such cancellation would be a surprising and unlikely coincidence.
Theo Goodwin says:
June 13, 2011 at 11:56 am
One mistake that seems almost universal in this forum is that we talk about climate and weather as if they have some inherent meaning that we all understand. I donāt think so. We should focus on some reasonably well-known natural phenomenon, such as La Nina, and create a description of it which embodies descriptions of the several natural regularities that make it up. Each description of a natural regularity is one or more physical hypotheses. These natural regularities are predictable. They are also the physical reality that makes up La Nina. Once the system or regularities is well understood then La Nina will be well understood. Then we can decide whether to call it weather or climate, but I think that question will have become irrelevant.
We’re taking another look from a few new angles on El Nino at the moment. SOme insightful comment is coming forward. http://tallbloke.wordpress.com/2011/06/12/the-timing-of-el-nino-in-relation-to-the-solar-cycle/
I appreciate this post on chaos theory, and believe there are chaotic elements to the climate system. However, there is a constant danger that it get’s used as a lazy way out of working out what the relationships are, and how they work.
Dave Springer says:
June 13, 2011 at 12:06 pm
“This implies determinism and determinism implies that chaos is no more than an illusion created by insufficient information.
Thus chaotic weather and climate are manifestations of human ignorance rather than a consequence of the laws of nature.”
There are many deterministic yet chaotic systems, like coupled penduluums. Determinism and chaos are no opposites, just like “alkoholic” and “cold” are no opposites.
The unique quality of a chaotic system is that you can determine its future state only by completely simulating it through every timestep without taking any shortcut (which in practice gets impractical very fast). You cannot approximate the future state via a shortcut – you can be arbitrarily wide off the mark when it’s a chaotic system.
Chaos is not an illusion but a property of certain dynamic systems. Whether such a system is deterministic is a different matter. (Indeterminism usually makes it even harder to predict, but indeterminism and a chaotic nature are not the same thing)
Shouldn’t some proof of this be required first…………..
“Our assessment affirms the conclusion that late 20th century warmth is unprecedented at hemispheric and, likely, global”
I mean, that is an assumption and everything after it is an assumption…………….
“There are many deterministic yet chaotic systems”
Good observation. From wikipedia:
Define the error as the difference between the time evolution of the ‘test’ state and the time evolution of the nearby state. A deterministic system will have an error that either remains small (stable, regular solution) or increases exponentially with time (chaos). A stochastic system will have a randomly distributed error.[63]
http://en.wikipedia.org/wiki/Chaos_theory
So, even is climate is completely deterministic, if it is chaotic then the size of the error between the real climate and and anything less than a “perfect” model grows exponentially with time. For all intents and purposes impossible, very quickly the prediction of a climate model will be overwhelmed by the size of the error bars. We predict the temperature will increase over 100 years by 3 C, plus or minus 300 C.
The argument that you can average chaos and arrive a non-chaotic solution implies that if you average exponential error growth over time, it will no longer be exponential. This is not true. The average of an exponential still gives an exponential. Thus it seems unlikely that climate is not chaotic.
Oh my gosh!!!!! Your post was nothing short of eye candy!!!! My brain cells have been delectably fed a gourmet meal.
Chaotic non-linear systems cannot be modelled with computers or anything else.
http://thepointman.wordpress.com/2011/01/21/the-seductiveness-of-models/
Pointman
Great article on a subject that has been greatly overlooked in the Global Warming debate.
I was once responsible for forecasting long term North American natural gas supply and demand for a major oil company. They had spent many millions of dollars to generate the most complete supply model of North American ever created. They had also hired some of the foremost demand modelers in the world.
Late in the project, I was brought into to get the supply/demand model to converge, which sunk funds and invested egos demanded. The only way you could do that was to force a fit by pre-deciding what outcomes you wanted. That totally negated the very expensive honest effort to predict the future natural gas picture in North America. That is a simple task compared to long term global climate forecasting, particularly given that the modelers determined man is the main climate driver.
Management was convinced to dump the effort for three very simple reasons.
1. You will never have perfect and complete data.
2. Even if you have a perfect data, you will never have a perfect and complete model.
3. Even if you had perfect data and a perfect model, if such a thing existed people will see it and change their behavior, negating your efforts.
As the article shows, off by a little means you can be off by a lot. As my Grandfather used to say āClose only counts in horseshoes and hand grenadesā That applies to the global climate models also.
Has anyone ever seen a highly complex, long term forecast that includes human behavior (the ultimate chaotic system) that has been shown to be correct? I would be happy to see just one model that works.
Actually, the idea of constructing an information theoretically optimal decoder of messages from the future does work in practice. Please see http://www.knowledgetothemax.com/Bibliography.htm for citations to the peer reviewed literature.
John B says:
“…if a grade schooler supported an essay with evidence like your photos, what sort of marks do you think they would get?”
If the teacher cared about the scientific method and honest science… A+
It has been demonstrated repeatedly that plants, both C3 and C4, grow faster with increased CO2. Commercial greenhouse growers inject CO2 in order to increase yields. They wouldn’t spend the money involved if it didn’t work. And in a world where one-third of inhabitants subsist on $1 a day or less, it borders on the criminal to try and reduce atmospheric CO2.
At current and projected concentrations, CO2 is a harmless and beneficial trace gas, with no evidence to the contrary. The entire “carbon” scare is based on money and political power, not on science. In the geologic past CO2 has been many thousands of ppm with no ill effects, compared with today’s minuscule 0.00039 of the atmosphere. The “carbon” scare is just that: a scare. Ask yourself: cui bono?
@Pointman,
You have given a nice example of why WEATHER is hard to model. Climate is actually easier. Climate models are more analogous to something like “if you keep hitting balls randomly, the chances of a ball going down will vary as the square root (or some other function) of the number of balls on the table, based on which we can predict the number of balls that will be left on the table after a given period.” Such a model is unlikely to be very accurate over short periods (small number of hits), but will get better over longer periods.
I just made that one up, but there are many other examples like it on the web.
John
Smokey said “In the geologic past CO2 has been many thousands of ppm with no ill effects”
Two problems with that:
(a) In most of the geological past, when that was true, there were no humans around to be affected
(b) It is the rate of current change that is (as far as we know) unprecedented, even in the geological past
Excellent article! Thank you, Dr Andrew
And a third problem:
Even if changes of the same rate did ocur in the past, say after a comet strike, so what? That in no way contradicts AGW. Forest fires have happened in the past, that does not mean there is no such thing as arson.
KR says:
June 13, 2011 at 12:19 pm
“Boundary conditions drive any deviations back to the averages for that system. So while we cannot state whether it will be sunny on your birthday ā we can still note that winters will be colder than summers, and that reducing the amount of energy leaving the climate at any temperature (with GHGās) will make the average temperatures higher.”
Be careful; your boundary conditions can easily be used to argue against a global warming tipping point, as Willis Eschenbach clearly does in his Thunderstorm governor hypothesis. As a warmist, the last thing you want is boundary conditions.
Also, “the amount of energy leaving the climate” is a function of the cloudiness, amongst others, because of the albedo. You didn’t dispute the chaotic nature of weather, so you have admitted that it is impossible to know the cloudiness of a given day in the future (beyond a few days); so you have also admitted that it is impossible to know the amount of energy leaving the planet.
Smokey
Cui bono? Good question. Who benefits from “business as usual”? CO2 is not toxic at any foreseeable concentrations (your position is a strawman argument at best, http://en.wikipedia.org/wiki/Straw_man), but the side effects of resulting climate change are, well, rather severe. Who benefits from not modifying their behavior?
– Perhaps the $$$’d interests in fossil fuels? Hmm?
Who loses?
– Everyone else?
—
Side topic:
I’ve yet to find a clean categorization of those who deny (sorry, but that’s the appropriate word for those who dismiss all the data) global warming, but it appears to include folks who feel that any restrictions on their behavior, or for that matter corporate profits, leads to a dictatorship. Unrestricted free markets lead to sweatshops and 19th century violations of basic human rights, as people are treated as disposable resources. You can certainly go too far the other direction, overly restricting market freedom, but a completely unrestrained capitalism is not a good solution. It really sucks, and we’ve spent a lot of time overcoming that “robber baron” mentality.
Perhaps a middle path? Charge companies the external costs (global warming) of their actions with a carbon tax or something similar (I’m not tied to a particular approach, as long as external costs are directed to those who cause them), while keeping it as simple and minimalistic as possible to not impede innovation?
John B – “It is the rate of current change that is (as far as we know) unprecedented, even in the geological past”
Indeed! As someone pointed out on another blog – surely crashing your car into a tree at 60mph won’t be harmful, will it? After all, uncounted cars have decellerated from 60 to 0?
Or is the rate of change important?
@John B
Great improv, enjoy your mathematically-ignorant post normal science delusion. Ignorance, as Orwell said, is bliss …
Pointman
John B says:
June 13, 2011 at 2:20 pm
(a) In most of the geological past, when that was true, there were no humans around to be affected
================================================================================
John, CO2 makes some people feel drowsy at 10,000 ppm………………
John B says:
June 13, 2011 at 2:20 pm
(a) In most of the geological past, when that was true, there were no humans around to be affected
=======================================================================
and 2000 ppm is common in houses and office buildings………
John B says:
“(a) In most of the geological past, when that was true, there were no humans around to be affected”
So CO2 adversely affects humans? Certainly not at current and projected concentrations. We’ve had this discussion here over the past year; search the archives. The U.S. Navy allows continuous exposure of 5,000 ppm CO2 on its nuclear submarines. There are no ill effects and there is no loss of alertness. So point (a) is deconstructed.
Next:
“(b) It is the rate of current change that is (as far as we know) unprecedented, even in the geological past”
You simply don’t know. CO2 may well have increased more rapidly in the past. There is evidence that it increased more rapidly in the early 1800’s and in the 1940’s. What we do know is that CO2 has been almost twenty times higher than now, and the biosphere flourished.
Finally:
“That in no way contradicts AGW.”
There is no evidence of AGW, John. Sorry to disappoint. There is only conjecture and hypothesis. Further, the fact that CO2 continues to rise, and the global temperature has been flat to cooling for the past decade indicates that CO2 has little effect.
Do try and make stronger arguments, John, these take no effort to deconstruct.
Hello everybody, thanks for all your comments, there are now too many to answer, and when I looked someone had usually answered them better than I could anyway!
You’ve given me some good feedbackm so I can attempt t to tighten the argument up still further.
Thanks
Andy
@Smokey
“So CO2 adversely affects humans?” Yes, by changing the climate – that’s what this is about, remember? It will do that (according to mainstream science) long before it makes anyone drowsy.
“There is no evidence of AGW, John” – er, because you say so. The IPCC reports summarise all the evidence: temperatures, glaciers, sea ice, etc., etc. Oh, but you’ve got a picture of a Christmas tree. “QED”
“ferd berple says:
June 13, 2011 at 1:22 pm
āThere are many deterministic yet chaotic systemsā
What would help with climate is to find properties of the system that are not chaotic eg like with the 3 body problem, the scope of movement of the 3 masses always lies within the total gravitational potential and kinetic energy, overall system momenta are conserved etc An additional problem with climate is that it is not a closed system like the 3 body problem.
I know of no such climate properties equivalent to conservation of energy and momentum that could be measured unless by chance something like global average temperature was such a property which I very much doubt.
@Pointman
“Great improv, enjoy your mathematically-ignorant post normal science delusion. ”
You have no idea how much maths I know. Here is another example: the behaviour of gas molecules is chaotic and cannot be predicted, but the gas laws, which work on the bulk are both accurate and useful. Oftentimes long range properties are predictable, i.e. can be modelled, even when the chaotic details cannot.
I don’t doubt that you know some maths too, but you are using it to obfuscate and confuse those who are only too willing to go along with your inappropriate conclusions.
Smokey – “You simply donāt know. CO2 may well have increased more rapidly in the past. There is evidence that it increased more rapidly in the early 1800ā²s and in the 1940ā²s.”
Really. You’re sure about that? Take a look at this, then, at some actual data:
and stop spouting nonsense. I’m always saddened when someone posts something that could be fact-checked in 5 minutes with a search engine.
I thought it was now Global Climate Instability. We need a special page.
Ed Mertin says:
June 13, 2011 at 8:30 am
What is even more astounding is that human greed and societal upheaval directly preceed times of great chaos in the general climate. It’s as if the general populatino is pre-programmed to put itself into the worst possible straits, just before salt is rubbed into the wounds.
I recently summarize this wordy article in two lines in my Authority graphic poster as:
“Only weather is chaotic. Never climate!”
I wonder if anyone has experimented with one or more of the climate models to see what the impact of increasing floating point size has on the output. Using the same input values one could vary the float size from a base of 64bits to 128bits to 256bits… In a chaotic system of this nature there may never be an adequate length of floating point number such that the results are identical when the float length is doubled??? Another possibility is that an adequate float length exists but it is too expensive (in terms of time) to use for “normal” runs. In either of the above two cases chaos can enter the model calculations at each simulation step thus aggravating the chaotic nature of the output.
Dave W
@John B,
You’ve just demonstrated to me how little math you do actually know. Obfuscation is more your forte I feel.
Pointman
Smokey says:
June 13, 2011 at 2:14 pm
10,000 ppm CO2 directly diplaces 10,000 ppm O2.
I guarantee you that the drop from 21% O2 to 20% O2 is not noticeable.
You might notice it at 20,000 ppm CO2 (19% O2) if you are working physically hard (been there, observed that).
We’ll slap run out of fossil fuels long before 10,000 ppm is reached.
No danger. None.
Climate Modeling has itself generated a lot of chaos and angst.
The problem is that the real world diverged from the predictions, but the chaos of climate modeling continued well into the absurd.
How to guarantee damage to a scientist’s spleen; tell him there are some things he simply cannot know!
A kindergarten teacher was observing her classroom of children while they drew. She would occasionally walk around to see each child’s artwork. As she came to one little girl who was working diligently, she asked what the drawing was.
The girl replied, “I’m drawing God.”
The teacher paused and said, “But no one knows what God looks like.”
Without missing a beat, or looking up from her drawing the girl replied, “They will in a minute.”
Can anyone guarantee that if we drop CO2 levels back to 1850 levels….
….temperatures will not continue to rise?
Nope……and that lag crap don’t cut it
and if no one can do that, they are all talking out of their hats
So Andy Edmonds has just proved that winter, summer, spring and autumn are figments of our imaginations.
Because if the climate is entirely chaotic we should only ever see random variations and physical system inputs such as solar insolation can have no visible effect.
Seriously I think Andy has a bad case of model mania. The climate is not just about the maths of chaos. It is also about the physical processes involved and they put significant constraints on the amount of chaos and what part of the system is chaotic and what part is not chaotic.
Here is a little something from our friends at The Resilient Earth.
āSeven Climate Models, Seven Different Answersā
“Why Climate Models Lie”
and a little something from our other friends. ;O)
Nevermind, let’s make trillion Dollar decisions based on all or some of the above. CRAP!
I was debating the topic concerning extreme weather events. I challenged the consesus opinion on Skeptical Science and was provided a link to a presentation by Stu Ostro who is a meteorologist. He is making the claim that the weather patterns are indeed getting more extreme. In his powerpoint presentation by shows numerous charts of upper level activity and claims that they are most unusual and are proof global warming is shifting climate into very dangerous territory. I watch local weather but they do not go into upper level troughs and ridges or what normal or abnormal pressures are for this region.
I am wondering if anyone on this sight can either verify the powerpoint presentation or explain why the upper level disturbances are not that unique. Thanks.
Stu Ostro presentation proving global warming is shifting weather patterns in dangerous ways.
KR,
Your scary hokey stick video can be easily deconstructed with a simple graph: click Notice that temperature is independent of rising CO2. If CO2 had any more than a minuscule effect on temperature, then temperature would closely track the rise in CO2. It doesn’t. That’s because CO2 has little to no effect on the climate. If there is any effect, it is too small to measure.
CO2 is a function of changing temperature, not a cause [except in the most minuscule sense]. By using a normal zero-axis CO2 chart, the scariness goes away: click Keep in mind that CO2 is a very minor trace gas. Quadrupling CO2 would still make it a very minor trace gas. Here is a graph of atmospheric CO2 with a proper y-axis: click See the CO2? Didn’t think so.
Here is a graph with even better temperature correlation: click It makes as much sense as your silly video.
John B:
You claim there is “evidence” of CO2 changing the climate. That is not true. But feel free to post anything you’ve got, and we will find that it is not measurable, testable evidence per the scientific method, but rather, conjecture and/or always-inaccurate computer models. If there was any real “evidence” of CO2 changing the climate, we would hear it trumpeted 24/7/365. So post what you’ve got, and we’ll deconstruct it here.
Smokey – As I expected, you’re moving the goalposts. CO2 is increasing faster now than in the 1940’s or in the 1800’s, completely contradictory to your earlier statements, so you change the subject. Serious weak sauce there…
Here are the various forcings over the last few hundred years:
http://data.giss.nasa.gov/modelforce/
Temperature decline from 1940-1970 or so was due to a lot of aerosols – the Clean Air act and similar legislation cleaned that up, leading us right back to the CO2 driven temperature rise. Climate is driven by the total of all forcings – nobody claims that it’s a single factor. To state otherwise is a strawman argument (http://en.wikipedia.org/wiki/Straw_man), such as you have presented before.
As to your “Post Office charges drive climate change”, I would point you at: http://www.venganza.org/images/spreadword/pchart1.jpg
Correlation does not equal causation. On the other hand, we know the physics of CO2 and IR – that’s a physical cause that correlates to climate change. The only trumpeting we see is people (like you) loudly denying the evidence in front of you.
Stay on topic, Smokey. You said “CO2 may well have increased more rapidly in the past.” KR’s video shows that was not the case. No mention of temperatures. You are moving the goalposts, as ever. Had you referred to temperatures, I am sure he would have provided different counter-evidence.
Your postal charges graph would be amusing, if you didn’t seriously think it made a point. For the benefit of other readers, who might not be quite so stubborn, here is why it is silly:
Back in the day when the smoking/cancer link was still disputed, people used to show the link between cancer rates and, say, the number of TV aerials. The attempt was to show “correlation does not mean causation”. Yes, that is true, it does not. But there was also a causal link, which was soon demonstrated to be true, between cancer and cigarettes. Same thing with CO2 and warming. Not only is there a correlation, there is also a link – the greenhouse effect. The correlation is not linear beause there are other effects, such as solar cycles, volcanoes, etc., but do some hard math, look at long enough periods and the correlation is clear. Supported, as I said, by the link of the greenhouse effect. Theory says that CO2 should cause warming, observations of many kinds show that it does.
Smokey, you just so desperately want it not to be true, that no amount of evidence will sway you. Why is that?
Me? As I have said before, and KR and others have also said, if temperatures don’t go up in the next 10 years or so, something is seriously amiss with AGW. I will be swayed by the evidence, if and when it appears.
He doesn’t seem to know that, no matter what a computer may preduct about planetary positions in the future, look far enough ahead and there is no doubt at all that they are chaotic. We can’t even solve the equations of the Moon’s motion exactly, despite centuries of effort. We have only a good approximation which is useful for navigating spacecraft and predicting eclipses.
KR the alarmist is getting desperate:
“As I expected, youāre moving the goalposts.”
Not really, KR. That’s why you didn’t quote me verbatim.
And please, dispense with your links to models. They’re models, see? Conjectures. They are not evidence. But I did enjoy the pirate chart, which has been posted here off and on for the past four years. Like the postal chart, it makes a clever mockery of CAGW, like this and this.
And when you say I am “denying the evidence,” my response is: what evidence?! You haven’t provided any evidence. For the umpteenth time, evidence means empirical [real world], testable, measurable evidence subject to falsification per the scientific method. All you’re doing is pasting model-based conjectures. None of it is evidence.
On the other hand, Beck’s peer reviewed paper is based on empirical evidence showing CO2 spikes in the early 1800’s and the 1940’s, and it has never been falsified. The alarmist contingent is desperate to try and falsify it, but they have always failed. Their criticism amounts to: “That can’t be right!” But it is right.
Beck et. al is rigorously documented, with over 90,000 separate CO2 measurements, accurate to within Ā±3%, with the CO2 readings meticulously recorded by Nobel laureates and many other internationally esteemed scientists who took copious notes and made drawings of their test apparatus [which has been duplicated, and which proves their measurements were within Ā±3%].
The CO2 readings were taken by numerous scientists on the windward side of ships during mid-ocean crossings on the Arctic Ocean, the Antarctic Ocean, the Sea of Okhotsk, the South Seas, and the Atlantic and Pacific Oceans, where there was no chance of contamination by urban factories, only fresh sea breezes. Readings were also taken on rural mountain tops and the desolate Ayrshire coast of Scotland. Those scientists knew exactly what they were doing, and they did it right. But the alarmist crowd just can not tolerate the findings, because the rise in CO2 in the 1800’s and 1940’s was more rapid than the current rise.
More empirical evidence is found in the Northern and Southern Hemisphere ice cores. It shows conclusively that rises in CO2 follow rises in temperature by 800 Ā±200 years. Effect cannot precede cause, therefore it is reasonable to conclude that much of the current rise in CO2 was caused by the MWP.
Still waiting for any testable, measurable evidence showing that X amount of CO2 results in Y amount of temperature change. That’s because there is no such evidence. There is only model-based GIGO, conjecture, and unproven hypotheses. None of it is testable and measurable.
In the past, CO2 levels followed temperature changes as a feedback, amplifying them. Now we’re increasing CO2 on our own.
Do natural forest fires in the past disprove arson in the present? Hmmmm.
Now you’re referring to Beck? Published in Energy and Environment? Which ignores the last 50 years of carbon cycle research? Read http://www.realclimate.org/index.php/archives/2007/05/beck-to-the-future/ for a bit of perspective on that. If you want to convince people (like me) who actually read primary references, look at data, look at multiple analyses – you’re going to have to do better. Perhaps by posting some facts.
You appear to be suffering from confirmation bias (http://en.wikipedia.org/wiki/Confirmation_bias) – nothing we can say or do will dissuade you from your confirmed opinions. But perhaps, just perhaps, other, less blindered folks will learn something from this exchange…
In your little spreadsheet example of the diverging series, you may also need to consider the fundamental errors in floating point operations in modern computers. It is easy to show that basic mathematical “laws” do not apply and most people are totally unaware of that fact. It is true that using modern fixed size mantissa and exponent floating point representations that (A-(B-C) does not always equal ((A-B)-C), and this can also be easily shown in Excel.
Best reference for the technical discussion is Donald Knuth’s “The Art of Computer Programming” Vol 2.
As far as I know, all the “models” rely on the underlying floating point hardware and I have not seen one that tries to track or estimate accumulated errors or to use true number representations (eg arbitrary precision rational arithmetic). I believe that in models iterating millions of times over millions of elements this fundamental source of error make the results almost random, hence the need to marshal the hoardes of models and model runs in an attempt to “average” away this chaotic behaviour.
As I have pointed out many times, if it weren’t for psychological projection, the alarmist contingent wouldn’t have much to say. Case in point:
KR ignores the numerous verifiable facts I have posted here, improbably saying “youāre going to have to do better. Perhaps by posting some facts.” To make matters even worse, KR simply cuts ‘n’ pasts realclimate propaganda [and he should be aware that until RC allows uncensored comments by scientific skeptics, realclimate its links are not worth clicking on. KR needs to try and make his own arguments, instead of using the crutch of a propaganda blog like realclimate, home of Michael Mann the climate charlatan.
KR is just frustrated because he has no empirical evidence showing global harm from CO2. None. That’s because there is no such evidence. The whole CAGW conjecture is based on psuedo-science and promoted by religious true believers.
I have a completely open mind, and I’ve stated on a number of occasions that I will begin to change my mind if anyone can produce measurable, testable evidence, per the scientific method, showing global harm due specifically to CO2. That is not confirmation bias, that is the scientific method in action ā something avoided by the alarmist crowd like Dracula avoids the dawn.
Oh, come on now. Surely the GCMs are chaotic enough to match the climate’s chaos to a high degree of precision! /sarc
Smokey – Did you read my post at June 13, 2011 at 11:55 am? It doesn’t seem so. That contains 60+ peer reviewed papers regarding benefits and costs of CO2 driven climate change. As I recall, cost papers outnumber benefit papers by more than 2x.
I don’t want to get into a insult-trading exchange here; I believe my posts stand on their own merits. Others may judge as they may. But that this point Smokey and I are at the “I hit you last” stage (remember that from childhood?), which is just not productive.
—
Back to the topic of this thread – weather (chaotic initial value problem) is not climate (boundary energy problem), and inability to predict weather does not invalidate long term climate averages. Weather is constrained by the average and standard deviation of the chaotic attractor, and if that mean changes due to forcing changes weather will (as an average) change accordingly. Claiming a boundary problem is insoluble based on an enclosed initial value problem is a distinct mislabeling of the situation.
Smokey, just answer me this one question: Why do you trust Beck’s paper, in preference to all the mainstream research that shows that CO2 has risen steadily but at an increasing rate?
Forget plant food for a moment, stick to that question. I am trying to get an insight into how you think.
Gah. Late night typing betrays me.
“Weather is constrained by the average and standard deviation of the chaotic attractor, and if those averages change due to forcing changes, weather will (as an average) change accordingly.” is what I meant to write.
KR,
You’re actually referring me to the disreputable Skeptical Pseudo-Science blog, run by a cartoonist?? Please.
When any of these censoring alarmist blogs list WUWT on their sidebars [like WUWT does for them], I’ll start clicking on their pseudo-science. But not before.
Try making your own resoned arguments like I do, without using propaganda blogs as a crutch. Cutting and pasting something from a Scientology blog is no substitute for a discussion or argument. Especially since those censoring blogs are not worth clicking on. So I don’t. I will read what you post here. But I draw the line at climate charlatans.
John B,
Let me turn the question around, and ask you why you casually dismiss the 90,000 CO2 measurements taken by well known scientists, incuding Nobel laureates [when the Nobel Prize really meant something]?
Is it because you have already arrived at your conclusion, and you don’t want inconvenient facts to get in the way?
KR says:
June 13, 2011 at 7:34 pm
Smokey ā Did you read my post at June 13, 2011 at 11:55 am? It doesnāt seem so. That contains 60+ peer reviewed papers regarding benefits and costs of CO2 driven climate change. As I recall, cost papers outnumber benefit papers by more than 2x.
——————————————————————————-
So why don’t you take any one of those sh!t papers and show conclusively (or partially even), in your own words, the evidence that anthropogenic CO2 has caused any of those effects. You have 60+ to choose from. Should be a breeze right ? Although, if you do, you will be the first person in history to accomplish it.
John B
Why don’t you look at scenario A, B and C of Hansen’s models and tell us where CO2 linked temperatures are now in relation to the projections of the models?
And John B what’s your evidence? IPCC? Don’t make us laugh. Their statements are not science and are not backed by any empirical evidence. They are based on these models.
Why do you think Hansen and Trenberth are desperately searching for the ” missing heat ” in the troposphere and the oceans. That heat is indeed ” missing ” because it never existed in the first place.
Listen, model output is not evidence. Empiricallly measured data is evidence. And if data and models don’t agree, then the models are wrong, not the data. So the projections of these models are worth zilch.
Smokey, I asked first! But anyway…
I am skeptical of Beck’s paper because, among other things and in no particular order:
1. It contradicts the ice core data
2. It suggests wild swings of CO2 that accepted theory says can’t happen
3. If concentrations had been that high, there should be residual C13, which there isn’t
4. The chemical measurements on which the paper is based are known to be problematic
5. It is only one paper, and no one else has published work that supports it or can reproduce its results
Now, you tell me why you discount everything else that contradicts Beck?
John B says:
“1. It contradicts the ice core data”
No, it doesn’t, because the ice core data ends more than a century ago.
“2. It suggests wild swings of CO2 that accepted theory says canāt happen”
There is no “accepted theory.” There is only conjecture at this point.
“3. If concentrations had been that high, there should be residual C13, which there isnāt”
You probably get that misinformation from Skeptical Pseudo-Science. Try reading the WUWT archives, you will get the straight skinny here.
“4. The chemical measurements on which the paper is based are known to be problematic”
Really? “Problematic”? How so, exactly? The test apparatus has been replicated from the original drawings, and shown to be accurate to within Ā±3%. Looks like you’re just winging it with your ‘problematic’ comment.
“5. It is only one paper, and no one else has published work that supports it or can reproduce its results”
As Albert Einstein said, it doesn’t require a hundred scientists signing a letter to falsify Relativity, it only requires one fact. CAGW has been repeatedly debunked. Only CAGW true believers still demonize “carbon”. Wise up, and stop being a true believer. CAGW is pseudo-science. It is more credible than Scientology at this point.
Hi KR,
Just to jump in on someone else’s argument…
āWeather is constrained by the average and standard deviation of the chaotic attractor, and if those averages change due to forcing changes, weather will (as an average) change accordingly.ā
This is just a variant of the It’ll average out argument. Even worse it’s a chaos is amenable to stats argument.
What attractor are we talking about here? Weather seems to have several acting at different scales. I used a software tool that was developed and tested on a variety of time series to look at yearly temperature data and found chaos. This simple result doesn’t begin to tell us the bounds of whatever attractor is operating at this scale.
Nor does it tell us that your simple implied equation is true: fundamentals + chaos = recorded values.
What if it’s fundamentals * chaos = recorded values? You’re making assumptions you can’t justify.
This is the point of the article, climate models are full of assumptions, made for convenience, and let’s assume with good intent, which cannot be tested in the real world.
1 dimension (time) isn’t enough.
Can’t ignore the spatial dimensions.
Spatiotemporal chaos differs FUNDAMENTALLY from temporal chaos. (See Milanovic’s writings at Curry’s blog Climate Etc.)
I’m willing to tentatively entertain the notion of INTERANNUAL SPATIOTEMPORAL chaos, but longer-term TEMPORAL chaos is ELIMINATED from contention by BOUNDARIES, INCLUDING SPATIAL ONES:
http://wattsupwiththat.com/2011/06/08/on-the-amopdo-dataset/#comment-678688
Earth Orientation Parameters (EOP) inform us about HARD boundaries on climate. With absolute certainty, we are NOT dealing simply with temporal chaos.
Thanks for the stimulating article.
Andy; you have just put an event which happened to me 64 years ago into a single word:
“If youāve ever braked hard on a motorbike on an icy road you may have experienced a ātank slapperā a chaotic motion of the handlebars that almost always results in you falling off.”
It was a pushbike, not a motorbike, and it was on the main drag of a hot and dusty little wheat town, Nagambie, in Australia. I was seven, the bike was flying, then the handlebars went rogue…
“Chaos” huh? Thank you.
John B says:
June 13, 2011 at 5:52 pm
“Theory says that CO2 should cause warming, observations of many kinds show that it does. ”
Huh? You say “many”, could you name me , say, two observations that CO2 causes warming? With source if possible?
Thank you Dr. Edwards – that was an excellent introduction into the role of chaos in climatology. When I first started getting into looking at climatology about 3 years ago what surprised me the most was the total absence of chaos in the models. This was even more surprising in view of Lorenz’s discovery of chaos in climate. This alone convinced me that the climate models over a period of many decades are essentially petaflops worth of garbage.
My interest in chaos goes back 20 years when I first read James Gleick’s book Chaos and suddenly a huge number of anomalous electrophysiologic findings from my research career suddenly made sense. My interests have been primarily dabbling in chaos theory as it applies to physiology and medicine and, in physiologic systems, chaos = health. The normal heart rate time series is chaotic and multifractal, based on my amateur analysis, whereas the heart rate time series in heart failure is linear. I worry when I see a Holter monitor of a patient that has essentially straight lines of heart rate for most of the day.
The primary failing of climate models is that assumption that by averaging a large number of runs one will come up with a result that will approximate the true climate. We don’t want to know the average of a large number of theoretical models, what we’re interested is what is going to happen to the earth’s actual climate which is a unique time series. I see this confusion all of the time in patients who assume that if they eat the right foods, exercise and do all of the things that they’re told will keep them healthy that they won’t become ill. When they see me after they’ve had their MI or some serious illness one of the first things I hear from them is “I did everything right and still I got sick”. While medical interventions may demonstrate a population effect, they are for all practical purposes useless in determining whether a particular patient will suffer an adverse event. I suggest that the earth’s climate is more akin to an individual patient’s medical history rather than that of the population.
Another observation I’ve made is that people can’t understand randomness or chaos. Almost inevitably they have the belief system that everything is causal; I’ve given up trying to discuss acausality with patients. If something goes wrong there has to be an external reason for it and I hear all of the fashionable reasons of why people assume they have gotten sick. We don’t throw virgins into volcanoes or burn witches at the stake to deal with crop failures any more, we just blame “pollution” or demonize CO2. When it comes to time series analysis, the majority of physicians are uncomfortable with even non-linear relationships let alone the assumption of chaos in the time series.
When I was an electrophysiology researcher I noticed we stayed away from the chaotic in what is a highly non-linear system and attempted to deal with the non-linearity by using very small perturbations of the system in a linear realm. I’ve also noticed that engineers tend to stick to analytically tractable areas and avoid the chaotic. Awareness of chaotic properties of nature is not new and Hurst was the first to actually look at fractal nature of river flows in his measurements of the Nile river. His calculations on reservoir sizes for dams date from the 1940’s. Given that the earth’s climate is chaotic, the primary response to this should be to setup the infrastructure of civilization in such a manner that it is capable of withstanding the adverse climates that result from living in a chaotic environment. Unfortunately LENR power plants aren’t sufficiently developed yet so that we can each have our own decentralized home power production but, when LENR actually produce power then this will make for a more robust society. Such a system is much less vulnerable to CME’s than the current centralized electrical production system. Similarly, where one has extreme variability in river flows constructing dams to level out the fluctuations is useful.
Given that humans have survived on the earth this long means that we have some means of dealing with chaotic systems. What is needed is to make people far more aware of the role of chaos in climate so that there is an intellectual capacity to appreciate it rather than the intuitive capacity that seems to be the norm.
one of the most common chaotic systems people encounter is the dripping tap or spigot.
With a constant pressure or head of water and a constant spout size the flow is very predictable, to the point that the system has been used to make ‘water clocks’ where the rate of flow is used to meause time.
But each drip varies in size and timing around an average, each drip is chaotic and inherently unpredictable in size and timing.
Like many natural systems there is a driving energy, the head or pressure or water in this case, and a pathway for that energy to be expended, the spout. While the way in which the energy is dissipated may be chaotic, the rate at which it is expended is constant over many drips even though each drip is chaotic. Changing the energy input, the pressure or changing the spout size, the energy dissipation process, will change the drip behavior, it will still be chaotic but the average rate of flow will change to reflect the new conditions.
The implications for climate, a similar system constrained by its energy input and energy dissipation processes, are obvious.
Climate system is not a chaos, but on decadal or century scale unpredictable, which I think is a different matter. On such scale climate change is driven by natural causes with an unpredictable future time line. This can be clearly demonstrated in case of CET:
http://www.vukcevic.talktalk.net/CET-NAP.htm
Correlation between two is by no way exceptional, but it is indicative of high degree of a cause-consequence relationship.
Boris Gimbarzevsky says:
June 13, 2011 at 10:36 pm
Great comment. As a whole scientific research communities across biology, physics, engineering etc. pay lip service to nonlinear dynamics and chaos but stick to linear systems for the vast majority of their working models. Its something akin to the rabbit-in-car-headlights phenomenon, a freezing in terror at the apparent violation of a tidy (but illusory) linearity lying at the heart of assumptions about how the world – and the scientific method – works. However such numinous dread of chaos / nonlinearity is unneccesary – there are ordered principles and logical structure within nonlinear chaotic systems also – just of a different kind. Science remains too compartmentalised. The field of study of chaotic / nonlinear systems is well developed and such systems despite their “chaos” label display conformity to a number of rules and patterns and types of analysis that are well understood. Such insights need to be communicated to fields where they are needed, e.g. climate, physiology, biology etc. in a qualitative manner. FWIW my own attempt to do so was posted here on WUWT in January this year.
Mingy,
You wrote “while perhaps the physical system that is the earth may not be in isolation a chaotic system, I am pretty sure there is no doubt some of the feedbacks are….So even if you did know all the starting conditions, etc., the biosphere would have a major say and you canāt model that.”
Good point. I agree. Thanks.
Bill
Anthony,
thank you for the introduction. There is a nice appendix for your equation:
http://en.wikipedia.org/wiki/Logistic_map
KR says:
June 13, 2011 at 6:42 pm
“In the past, CO2 levels followed temperature changes as a feedback, amplifying them. Now weāre increasing CO2 on our own.”
Yes, the ice cores clearly show that CO2 levels followed temperature changes. But, as far as I know, the data shows no sign of any feedback at all. If increases in CO2 caused a corresponding increase in temperature then it should show up in the data, but it doesn’t.
This is what the empirical data says: when temperatures go up the CO2 goes up, and vice versa – and that changes in CO2 have essentially zero effect on the climate.
Forget the models and look at the data.
Chris
Chaos theory is a rich and powerful branch of mathematics, and it can not be summarized in any short popular article. Still I want to add some considerations to the text above to dismantle widespread misconceptions:
1) Chaotic systems can display different levels of chaoticity, from barely detectable to fully developed. There is a known universal rout to chaos by period duplication. At first we see quasi-periodic dynamics, and as parameter of chaoticity grows, more and more different periods emerge being powers of 2 of original period.
2) Statistics of chaos is quite different from that of noise, so looking for signatures of chaotic behavior in Fourier transform of time series is a standard procedure in this field. Unfortunately, it requires rather long time series, often unobtainable in actual measurements. But model outputs can be always tested this way for detecting chaotic behavior of the models. Actually this is how the first chaotic attractor (Lorentz attractor) was discovered.
3) Mathematical models which display chaotic behavior are not robust. That means that their reliability as true representation of real-world phenomena can not be proved by comparison to measurements, and their parameters can not be reliably identified. This severely restrict their usefulness to prove or disprove any hypotheses about real world. They are merely a basis for speculations and insights, but hardly anything else.
Chaos. There seems to be a purist argument here versus something theoretical? On one side the purist believes we can never know everything (Heisenberg etc), BUT, we certainly can determine *some* things. The others argue vehemently against chaos on principle, and they appear to believe it is theoretically possible to obtain all parameters and be deterministic, perhaps perfectly?
Isn’t it merely a question of exactly what can be determined? In my opinion, certainly not everything. It would require hubris to state we can obtain *all* existing information (clouds, wind, location of electrons too?). Someone said (cannot remember) that to store all the required information of a system in order to be perfectly deterministic it would exceed the mass of all the matter in the system (or something along those lines, someone help!).
I don’t understand why the two sides here, chaos vs determinism simply cannot agree that Chaos == Incomplete Information, and move on. It seems like semantics. I’ll stick my neck out and say: we’ll never know everything. But that does not mean we cannot try.
It went right over your head. How about this, just pick a day, a famous one, the NH Summer Solstice, usually June 21. Now pick a time, high noon. Now compare the conditions of that date/time in the same place for all the years you can find data for. We’ll see 100 ° F and 55 ° F and pouring rain and drought and if you go back far enough (1816 perhaps, in the NE), snow flurries. That is the real world Lazy Boy and it is chaotic, from our point of view. The sum total of countless forces (parameters/variables) interacting for a *minimum* of 364 days (think about it, years are our invention) preceding June 21, give a result that will never be perfectly determined in advance.
Now your model, due to lack of information, at best will only be a poor approximation of that gigantic system that created those various outcomes found at high noon on June 21. So the model can only suck. The question is, just how much can you reduce the suckiness of it as to not waste our time, and our money chasing the un-catchable. I say you cannot, but you are free to try, just use your own money, okay?
Am I the only one who barfs when math is written as maths? I must be getting old.
I find it humorous that KR brought up chaotic attractors without knowing what it meant. There’s a reason for attractors. It means physical mechanisms exist that keep the system close to the attractor state. What are these mechanisms called? Well, in climate they are called negative feedbacks.
Thanks KR for mentioning that the climate feedbacks must be negative for us to exist in an attractor state.
As for comments that CO2 increases are dangerous. Pure poppycock. The current low values are the dangerous situation. We even have empirical evidence that increases promote plant growth and make MOST plants more resistent to drought.
Dr Andy Edmonds
Even if we don’t know the analytic forms for the chaotic attractor(s) of weather, we still have the observational data to determine limits on them. If we didn’t, farmers could not make even a wild guess as to whether or not next year’s crop would even be possible in their area, or whether they should up and plant it 100 miles north or south. We know the rough outlines of how warm or cold next year will be within some limits of variation.
The major issue I have with your post is that you are taking an initial value problem, predicting weather, assigning its difficulties to a boundary condition problem, climate, and stating that therefore climate prediction and analysis is impossible. Which leads right into “We can’t know, so let’s not do anything about it…”
The boundary condition problem of climate is in terms of energy balances. Over a period of time, despite internal variation (fluid dynamics, ENSO, lead/lag responses, moving pressure zones), any imbalance, positive or negative, will correct itself through basic conservation of energy.
In terms of climate modeling, estimations can be made will full coupled General Circulation Models (http://en.wikipedia.org/wiki/Global_climate_model) running with a variety of initializations to perform a Monte Carlo sampling of the interaction space (and yes, while the one and two sigma bounds are often left off graphic representations of the results, they are available, contrary to your post).
Alternately, you can get essentially the same results using a zero dimensional climate model (same wiki page) that simply looks at the energy imbalances due to radiative physics, insolation, and GHG levels. A zero dimensional boundary condition model like this cannot tell you the geographic spread of temperatures, obviously, but it will tell you the average temperature based on those conditions.
Boundary condition problems are.not.chaotic. Their solutions cannot give you daily weather – that’s the wrong question. But they can be solved for long term values that the chaotic system will average to based on energy considerations.
The only remaining question is the time scale of weather, of chaotic variation, how long for the attractors to cycle. People have spent a great deal of time on this, and 30 years sampling appears to be long enough to cover observed variations. See http://tamino.wordpress.com/2011/03/02/8000-years-of-amo/ and http://tamino.wordpress.com/2011/02/26/mathturbation/ for some numeric analysis of this issue. A much shorter term period can be used to look at trends if you account for the larger known variations, such as the ENSO, and for forcing changes, such as total solar intensity, volcanic aerosols, and the like. But 30 years as a simple running average appears sufficient to encompass multiple +/- swings of weather variations.
In summary, your “We can’t know, so let’s not do anything about it…” conclusions (whether that was your intention or not, that’s how it will be used) are based upon a mischaracterization of the the problem.
John B says:
June 13, 2011 at 2:20 pm
1) Why do you believe that humans will be more affected by CO2 than other animals?
2) According to the geological record, CO2 rates often change quite fast.
KR:
There are no costs to enhanced CO2.
There are many benefits to enhanced CO2.
CO2 production should be subsidized, not taxed.
Since our CAGW believers are once again focusing on CO2, I will once again challenge them to explain why the cooling effects of GHGs like CO2 are never mentioned in any discussions. For some reason they always run away and avoid the topic.
John B says:
June 13, 2011 at 3:09 pm:
And what pray tell is this so called evidence that the IPCC sumarizes? Is it the loss of Himalayan glaciers? Oops, that has been disproven.
Was it the loss of the Kilamanjaro glaciers? Oops, that was disproven as well.
The only thing the IPCC has is conjecture and supposition backed by either unproven or disproven “facts”.
LazyTeenager says:
June 13, 2011 at 4:36 pm
No, he proved that there is no way to predict with any certainty how warm summer will be, or how cold winter will be.
Of course winter and summer involve changes in energy flow that are huge compared to what CO2 is capable of generating.
KR says:
June 13, 2011 at 5:20 pm
The claim that the cooling of the 60’s and 70’s was caused by [aerosol] cooling is trivially easy to disprove.
The places with the highest loads of aerosols saw the least cooling.
Aerosol’s are just another wild guess used by modelers to explain away the failings of their models.
In the past, CO2 levels followed temperature changes as a feedback, amplifying them.
—
Prove it.
DirkH says:
June 13, 2011 at 10:15 pm
John B says:
June 13, 2011 at 5:52 pm
āTheory says that CO2 should cause warming, observations of many kinds show that it does. ā
Huh? You say āmanyā, could you name me , say, two observations that CO2 causes warming? With source if possible?
—
Dirk, he’s trying to claim that correlation is proof positive of causation.
IE, we have measurements shonwing that CO2 has increases.
We also have very poor and problematic data shows that temperatures have probably increased during the same time.
Therefore, CO2 must be the cause.
Blade says:
June 14, 2011 at 4:21 am
Am I the only one who barfs when math is written as maths? I must be getting old.
If you ever come to the UK, bring plenty of barf bags.
KR,
The point is that the models you cite assign all residual warming to man-made CO2. However, chaos can produce the same warming without man-made CO2. Moreover, as climate science advances, we find other explanations for for the residual warming that the AGW community assigns to man-made CO2. As it now stands, the cost and coordination problems required to reduce the small amount of unexplained warming are so large as to render the entire AGW proposition moot.
Here is a question for you Andy.
Can your software determine if the amount of chaos in the temperatures or other climate variables such as wind speed is changing with time and if so what does it show?
The appearance that the problem being addressed is a matter of mere semantics is created by the device of equating chaos to incomplete information. It would be more accurate to state that Chaos “implies” Incomplete Information. That it implies incomplete information has implications for climatology.
It follows from the non-linearity of the equations which describe a chaotic system that the associated system cannot be fully described as an interaction among the parts of this system. Thus, when an attempt is made at fully describing a chaotic system as an interaction among its parts, this attempt fails from the incompleteness of the information about the whole of this system. Through competently performed scientific research, the incompleteness may sometimes be reduced to a level that is sufficient for policy making.
Climatologists of the “consensus” argue, in effect, that the whole of the global climate system can be described as an interaction among the parts of this system. If the global climate system can be described in this way, the evolution of the climate is susceptible to being computed (e.g. by an atmosphere-ocean general circulation model). By his essay, Dr. Edmonds is making the important point that this argument is false.
As the climate models provide policy makers with incomplete information about climate outcomes, the question arises of the degree of the error. An answer to this question is unavailable in the literature of climatology. In effect, climatologists have based their claim to anthropogenic global warming upon a false argument.
Amazing how many ignore the spatial dimensions. No wonder they buy the notion of temporal climate chaos wholesale. A string of spatiotemporally BIASED samples (a time series sampled from an eddy field) may appear chaotic, but oscillations run up against each other in space, which WRAPS on itself spherically. There are GLOBAL CONSTRAINTS on that sphere, as FIRMLY INDICATED by the Earth Orientation Parameter (EOP) record.
Is there regional interannual turbulence? Of course:
1) http://upload.wikimedia.org/wikipedia/commons/6/67/Ocean_currents_1943_%28borderless%293.png
2) http://wattsupwiththat.com/2011/05/15/interannual-terrestrial-oscillations/
But adjust the microscope across scale and look what aggregates into focus:
1. http://wattsupwiththat.files.wordpress.com/2011/04/vaughn_lod2_fig7.png
2. http://wattsupwiththat.files.wordpress.com/2010/09/scl_northpacificsst.png
3. http://wattsupwiththat.files.wordpress.com/2010/12/vaughn_lod_fig1b.png
4. http://wattsupwiththat.files.wordpress.com/2010/08/vaughn_lod_amo_sc.png
SCLā = solar cycle deceleration
What one sees is a function of the location & scale of one’s samples. Turbulence at one scale. Strict periodicity at others. (And some are still refusing to adjust their scope location & zoom…)
Boris Gimbarzevsky, thanks for the stimulating comments.
Yet (NH) summer is always warmer than winter
It would appear at first look that Dr. Andy Edmonds has defined the the anatomy and functional operation of any government program and particular those programs conceived by committee.
Tamino discusses this very post right here: [snip. When Tamino lists WUWT on his blogroll you can provide free advertising. ~dbs, mod.]
As I and others have noted, a chaotic system can certainly possess stable averages and deviations – and those are climate. Changing the boundary conditions (changing the constants in those chaotic attractors) changes the climate.
Edmond’s second illustration, incidentally, is just a random walk, which is not related to either climate or weather.
This particular error seems to recur every once in a while on various skeptic blogs – I’ll have to see if the distribution of recurrence fits a chaotic attractor š
@Theo Goodwin. Let’s just say that your long dissertation yielded nothing. It’s a confusing mess.
“There are no hypotheses in computer models.” Would you care to elaborate on that? Explicit assumptions are hypotheses. That means that fudge factors are hypotheses. The entire model is built on assumptions, especially the one about the amplification of CO2 effects. There is an assumption that solar variation doesn’t count. The whole point of the models is the hypothesis that they can predict future climate!
You posted a lot of words but frankly, they made no sense as a coherent argument..
Terry Oldberg said: ” [It is] possible to predict surface temperature and precipitation related variables in the western states of the U.S. as much as 3 years in advance. I provide an introduction to the methodology that made this advance possible and compare it to the methodology of modern climatology at http://judithcurry.com/2011/02/15/the-principles-of-reasoning-part-iii-logic-and-climatology/ .”
The referenced article is the most unreadable bunch of claptrap I’ve seen in some time. And it purports to tell us about reasoning! ModĆØle vs model? Give us a break!
KR says:
June 14, 2011 at 10:31 am
“…As I and others have noted, a chaotic system can certainly possess stable averages and deviations ā and those are climate. Changing the boundary conditions (changing the constants in those chaotic attractors) changes the climate…”
No, no, NO – your statement is complete and utter rubbish! Should you wish to understand the real story about how chaos (not randomness) affects our climate you need to spend some time learning about it, not listening to ignorant people on Tamino’s alarmist blog. A good place to start is here…
http://judithcurry.com/2011/02/10/spatio-temporal-chaos/
Perhaps when you understand a bit more you will be able to make a useful contribution to this insightful post.
But there are no negative feedbacks in a dripping-tap scenario. Output can be computed from input factors when averaged over a long enough time scale. This is what is really objectionable about Warmism. We scorcher-scoffers claim there are such feedbacks in a warming climate, such as increased cloud formation and increased thunderstorm activity.
Here’s a WUWT article by Willis decrying āsimple physicsā thinking:
http://wattsupwiththat.com/2009/12/27/the-unbearable-complexity-of-climate-2/
And we claim that the warmistsā posited positive feedbacks, which are required to make CO2 alarming, are speculative and unlikely. (Else climate would have exhibited runaway warming in the past.)
Chris Colose says:
June 14, 2011 at 8:04 am
Yet (NH) summer is always warmer than winter
______
As obvious as this is, it is also the perfect way to see the difference in spatio-temperal scales when speaking about the chaos in the weather and chaos in the climate. You can in fact, be pretty certain the 25, 250, or 2500 years from now, NH summers will be warmer than winters. Yes, you can’t tell me what the temperature will be on Jan 1st next year versus July 1st for any given city. Climate, in his regard is easier to forecast than weather as the scales and forcings are long-term large events that are reliable (once you’ve got them all identified, which we may not yet).
Richard M says:
June 14, 2011 at 6:22 am
Since our CAGW believers are once again focusing on CO2, I will once again challenge them to explain why the cooling effects of GHGs like CO2 are never mentioned in any discussions. For some reason they always run away and avoid the topic.
_____
Maybe because the net effect of GHG’s (taken over the whole atmosphere) is one of warming, not cooling…i.e. take away the GHG’s and the world would be so much colder.
Smokey says:
June 13, 2011 at 8:43 pm
John B says:
ā1. It contradicts the ice core dataā
No, it doesnāt, because the ice core data ends more than a century ago.
Well, how come the very picture you link to shows “CO2 ice core Antarctica” going right up to 1960? Check your sources!
ā2. It suggests wild swings of CO2 that accepted theory says canāt happenā
There is no āaccepted theory.ā There is only conjecture at this point.”
Just because you don’t accept it, doesn’t mean it is not “accepted theory”
ā3. If concentrations had been that high, there should be residual C13, which there isnātā
You probably get that misinformation from Skeptical Pseudo-Science. Try reading the WUWT archives, you will get the straight skinny here.”
Now we are getting somewhere. You, a “skeptic”, are telling me I should get the “straight skinny” from… a skeptic blog. Not a paper, not research, but a skeptic blog. THAT is confirmation bias! And no, I didn’t get it from skepticalscience, rebuttals of Beck are widespread.
ā4. The chemical measurements on which the paper is based are known to be problematicā
Really? āProblematicā? How so, exactly? The test apparatus has been replicated from the original drawings, and shown to be accurate to within Ā±3%. Looks like youāre just winging it with your āproblematicā comment.
This is from Keeling’s (of Mauna Lao fame) autobiography, talking about the wet chemical method, “This Scandinavian program, started by Rossby in 1954, had been a major factor in triggering interest in measuring CO2 during the IGY. Nevertheless it was quietly abandoned after the meeting, when the reported range in concentrations, 150ā450 ppm, was seen to reflect large errors.” Again, this kind of criticism is widespread.
ā5. It is only one paper, and no one else has published work that supports it or can reproduce its resultsā
As Albert Einstein said, it doesnāt require a hundred scientists signing a letter to falsify Relativity, it only requires one fact. CAGW has been repeatedly debunked. Only CAGW true believers still demonize ācarbonā. Wise up, and stop being a true believer. CAGW is pseudo-science. It is more credible than Scientology at this point.
And now to my main point: Yes, indeed, one fact would debunk CAGW. But it has to (a) be a genuine “fact” and (b) be a relevant “fact”. If I say I have measured temperatures on my front porch for the last 30 years and seen no rise in temperatures, does that mean I have a “fact” which will debunk AGW? Probably not. I probably have a broken thermometer. How would you check? You would look at other people’s thermometers. Get it? Replication! Just the thing you skeptics are always calling for – except when it suits you. If one set of measurement is anomalous, first you should be skeptical of the measurement. If it holds up, then go ahead and be skeptical of the theory it questions.
The real problem here is that Beck is one paper, probably a flawed paper, but you accept it because it seems to say what you want to hear. And that is not skepticism, it is, well, “skepticism”. A true skeptic will look at all the data, and be skeptical of all the data.
But it’s not really about skepticism, is it? I see that now.
I see JohnB keeps dodging the problem of the missing evidence of AGW. Without evidence its all just opinion and you know what they say about opinions.
Its warmer and brighter in the day and darker and colder at night.
Therefore chaos and nonlinear / nonequilibrium pattern dynamics play no role in climate (and it can be only CO2) š Therefore atmosphere and ocean are always at or close to equilibrium, and the earth is not an open system receiving energy from outside (only CO2 warms us, we get not a single joule of heat energy from the sun). Therefore there are no positive or negative feedbacks, the interplay between which might otherwise cause nonlinear dynamics. No – its all simple, all linear, all equilibrium and all CO2! Hooray!
ummm … FAIL
DCC:
While I’d be pleased to debate the quality of the article entitled “The Principles of Reasoning: Part III” with you, the most appropriate place do to so is not in the comments section for Dr. Edmunds’ article but rather in the comments section for “The Principles of Reasoning: Part III.” You could start the debate going by sharing the basis in facts and logic, if any, for your characterization of “The Principles of Reasoning: Part III” as “a bunch of claptrap.”
“No climate researchers dispute his analysis that the weather is chaotic.”
Except long range forecasters using solar factors.
RockyH said “I see JohnB keeps dodging the problem of the missing evidence of AGW. Without evidence its all just opinion and you know what they say about opinions.”
I can provide lots of evidence, but I have come to realise that it means nothing to most people here, whereas a picture of a Christmas tree or an out of context graph is treated as the “fact” that single handedly debunks AGW.
Here’s a whole bunch of studies for example:
http://ams.confex.com/ams/Annual2006/techprogram/session_19150.htm
The conclusion from paper 1.7 states:
“In comparison, an ensemble summary of our measurements indicates that an energy flux imbalance of 3.5 W/m2 has been created by anthropogenic emissions of greenhouse gases since 1850. This experimental data should effectively end the argument by skeptics that no experimental evidence exists for the connection between greenhouse gas increases in the atmosphere and global warming.” (emphasis mine)
You don’t like that one? There are literally hundreds more where that came from (i.e. mainstream science). It’s all there if you open your eyes. But Smokey will tell you that CO2 is plant food, and he has a picture to prove it!
While paper 1.7 certainly provides evidence of the downward pointing radiative flux at selected points in space and time, it fails to provide evidence of the causal relationship between an increase in the atmospheric CO2 concentration at the Mouna Loa observatory and an increase in the equilibrium temperature at Earth’s surface that is assumed by the theory of AGM. In fact, this evidence is unobtainable for the equilibrium temperature is not an observable. As this evidence is unobtainable, the theory of AGM is non-falsifiable thus lying outside science.
Weather arises from the random cycling of sticky variables (Mean temperature/precipitation/etc. values on some date) around their attractors. The mentioned acute sensitivity to initial conditions exhibited by Earthās atmosphere is what makes forecasting individual weather events far into the future computationally intractable. However predictions regarding Climate, as Tamino states in his post, is not concerned with the exact values of the afformentioned variables but instead with the distributions of these variables. The distributions are dependent on the location of the attractors which shouldnāt change absent an imbalance in the ingoing and outgoung fluxes of energy to the Earth. There is a measured global heat imbalance of .75Wm^(-2) which is causing various attractors to change location. Accurate predictions on what effects these will have are in principle doable.
http://tamino.wordpress.com/2011/06/14/chaos/#comment-51525
John B says:
“And now to my main point: Yes, indeed, one fact would debunk CAGW.”
I keep posting those inconvenient, pesky facts, and JB keeps ignoring them. Temperature moves independently of CO2, another pesky fact. And there is zero evidence of any global harm resulting from CO2 ā another pesky fact ignored by the cognitive dissonance-afflicted alarmist crowd.
The reason WUWT is a much better place to learn the facts, rather than reading grant-trolling papers, is because “studies” cannot be cross examined. Faulty pal reviewed papers are routinely hand-waved through peer review, while skeptical papers have to jump through flaming hoops before they’re disallowed by heavily biased referees and editors. Here, we can have a conversation ā if the alarmist contingent stops ignoring the scientific method, and concedes that there is no evidence of global damage from CO2. But so far, they have simply avoided the topic, rather than honestly admitting that they have no such evidence.
Another reason to discount peer reviewed papers is because they are usually wrong. Particularly climate-related papers, which are routinely debunked here, at Climate Audit, and at other scientifically skeptical sites. Keep in mind that scientific skeptics are the only honest kind of scientists, which leaves out the alarmist crowd entirely. Michael Mann has ignored the scientific method’s transparency requirement for 13 years and counting, with no sign yet of willing cooperation. That’s the kind of scientific charlatanism that indicates the alarmists have plenty to hide.
Finally, you can prove anything with contrived assumptions, which form the basis of the alarmist arguments. But that doesn’t make for correct conclusions. It tortures the alarmist crowd that the planet itself is proving them wrong. They’ve believed in Mann’s debunked Hokey Stick chart for so long that they own that falsified belief. Sad for them, the more we learn the more ridiculous the demonization of harmless, beneficial CO2 is.
I’m willing to discuss facts, but first John B needs to either put up testable, measurable evidence per the scientific method, showing global damage caused specifically by CO2, or admit that he has no evidence.
JohnB
In comparison, an ensemble summary of our measurements ..
“Ensemble”? Ensemble refers to multiple model runs. No-one talks about ensembles of actual data. I think you might have shot yourself in the foot.
Is your AGW indoctrination so deep that, in seeking a data example, all you can come up with are model runs?
It looks like for many AGW devotees, what it will take to persuade them to recognise the existence of real world data, as opposed to computer models, is a process similar to that which Leonardo di Caprio had to undergo in the film “Shutter Island”.
Skeptical as I am of the results of GCMs run over climatic time-scales, there seems to be an overdependence upon chaos theory in explaining the workings of the climate system on a planetary scale. Mathematical possibilty is not the same same as physical reality. I believe there is much to be learned from adopting the same viewpoint as in ocean wave studies. While the actual time-history of wave motion at any point in time and space is is unpredictable, the spectral characteristics can be fairly well estimated from knowlege of relatively few parameters: wind speed, duration, and fetch dimensions. Unfortunately, climate science’s obsession with GHGs has led to fundamental confusion concerning the factors that set surface temperatures and no comparably skilled forecasting/hindcasting methodolgy has been developed.
R. Gates says:
June 14, 2011 at 12:02 pm
Richard M says:
June 14, 2011 at 6:22 am
Since our CAGW believers are once again focusing on CO2, I will once again challenge them to explain why the cooling effects of GHGs like CO2 are never mentioned in any discussions. For some reason they always run away and avoid the topic.
_____
Maybe because the net effect of GHGās (taken over the whole atmosphere) is one of warming, not coolingā¦i.e. take away the GHGās and the world would be so much colder.
OK, show me the computations. The truth is I’ve never seen anyone look at the “net effect”. Since you are so sure you understand it all, show me how it’s computed.
Richard M – “OK, show me the computations. The truth is Iāve never seen anyone look at the ānet effectā. Since you are so sure you understand it all, show me how itās computed.”
I would suggest starting at the beginning, with Arrhenius 1896 (http://www.globalwarmingart.com/images/1/18/Arrhenius.pdf), in particular page 265, where he discusses nights warming faster than days, polar amplification, increased IR at ground level, land warming faster than days, etc. (GHG signatures), combined with numeric computations of the level of the CO2 greenhouse effect and H2O feedback on temperature. In other words, the “net effect”, fully presented as the initial discussion of the topic.
Stratospheric cooling (a signature of only greenhouse gas effects), is not discussed, as the stratosphere had not been discovered yet.
Of course, we’ve now obtained more accurate numbers on various aspects, but Arrhenius did a great job given the measurement limitations of the time.
Follow that with http://www.aip.org/history/climate/co2.htm for a history of the theory, with branches to whatever aspects of the theory you wish to look at.
Theo Goodwin says “Hypotheses are used for prediction and explanation…If they do nor explain phenomenon x then they cannot predict phenomenon x….In all of science, hypotheses are useful for both prediction and explanation.(eg Kepler’s 3Laws.) ….it follows from the preceding that each hypothesis is testable and falsifiable..” ” A computer model is analogous to a system of deduction….The computer will never be able to do more than produce results that you program.The model does not contain physical hypotheses and cannot be made to offer explanations for the results you program it to produce.”
Thank you for this lucid critique of climate science based on modelling, Theo.
@terry, The paper says “ensemble summary of measurements”. It compares them to models, but it is summarising measurements, not model runs. The measurements are of downwelling radiation at frequencies attributable to GHGs. Conservation of energy says that that energy must go somewhere. Get it? Next you will be telling me that we haven’t measured temperatures, all we have seen is the level of mercury in a glass tube, and correlation does not equal causation. But if you are more swayed by a picture of a Christmas tree, I can’t help you.
@Smokey, your “facts” are just out of context (in this case cherry picked, short timescale) graphs. But you think they are better than any study ever done, unless that study agrees with you. And all the old drivel about “pal-review”. Gimme a break! Write a paper that makes sense, it will get published. Show a graph of the last 8 years and claim that “warming has stopped, FACT”, you will not, except maybe in one “journal” (E&E).
Look, Lindzen, for example, questions sensitivity and asserts that negative feedback from clouds will damp down warming. He has a point, though many disagree with him. You are just silly. Your pictures appeal to folk who lack either the abilitiy or desire to look deeper. Sadly, there are many such people.
OK, we are off Anthony’s front page now, see you all on another thread.
PS. If you don’t like Mann’s hockey stick (for which the data IS available), there are plenty of others. Replication, get it? Even Wyner and McShane produced a hockey stick, though, bizarrely, most here seemed not to see it.
@KR, I suspect few here will follow your links, so I have transcribed part of Arrhenius, p265, as it is so mind bogglingly prescient for 1896!!!
“The influence [of increased CO2] has a minimum near the equator, and increases from this to a flat maximum near the poles. … The influence is in general greater in the winter than in the summer … also greater for land than for ocean … [in the] Southern hemisphere, the effect will be less there than in the Northern hemisphere. … [the effect of increased CO2] will of course diminish the difference in temperature between day and night. ”
@everyone else. KR’s other link is good, too. Follow it and be enlightened!
John B fawns over KR’s link and discounts every one of the numerous links I’ve posted in this thread. That’s cognitive dissonance in action, folks. Harold Camping has nothing on the blinkered JB, who says:
“Show a graph of the last 8 years and claim that āwarming has stopped, FACTā, you will not, except maybe in one ājournalā (E&E).”
Wrong again, and glad to oblige: clickA, clickB, clickC clickD clickE clickF clickG clickH clickI clickJ clickK clickL clickM clickN clickO clickP
Prediction: John B will never accept any of those charts, because to admit that even one is valid will undermine the basis of his belief in catastrophic AGW. JB’s cognitive dissonance will not allow him to admit that the planet itself is falsifying his belief in the evil “carbon” demon, which is putatively causing runaway global warming ā even though the real world evidence clearly debunks that nonsense. Unlike skeptics, who have nothing to prove, the believers in runaway global warming are forced to cherry-pick only what supports their improbable belief system. Too bad about that pesky reality, eh?āŗ
As we see in many of the charts above, the rise in temperature since the LIA is simply coincidental with the rise in CO2. It’s a coincidence, see? The planet is warming naturally from the LIA, and that warming trend has not accelerated despite the 40% increase in CO2. If rising CO2 caused rising temperatures, the trend line would be accelerating. But it’s not. In fact, temperatures are declining, which debunks the CO2 nonsense. And of course, none of the wild-eyed climate catastrophe predictions have come to pass. But when someone is afflicted with cognitive dissonance, they respond just like true believer Harold Camping: the end of the world didn’t happen as predicted, but that can’t possibly mean that Harold was wrong, it only means that the end of the world has been postponed until October. Cognitive dissonance in action; Orwell’s “doublethink”. And since the rise in harmless, beneficial CO2 has not caused the predicted climate catastrophe, it only means that doom is somewhere down the road. It couldn’t possibly mean that they were wrong about their CO2=CAGW conjecture. See how it works? True believers can never admit that they were wrong about CO2=CAGW, even when Planet Earth is decisively falsifying their belief system. That’s cognitive dissonance in action, and it is rarely curable. The earth could descend into the next great Ice Age, and John B would still believe that runaway global warming is coming.
John B:
One of several barriers to generalizing to AGW from measurements of the downwelling radiation that were made at discrete points in space and time is that it is the divergence of a heat flux vector and not the divergence of a radiant energy flux vector that warms Earth’s surface. Heat is transferred by conduction and convection as well as by electromagnetic radiation. Climatologists have not yet gotten around to predicting the consequences of the various fluxes for temperatures at Earth’s surface and comparing the predicted to the observed temperatures. In AR4, the comparisons are of projected to observed temperatures. While climatologists often confuse projections with predictions, unlike predictions, projections lack the property of falsifiability. Thus, while the climatology of AR4 can appear to confused climatologists to be a science it is in fact a pseudoscience from its lack of falsifiability.
says:
June 14, 2011 at 10:31 am
āā¦As I and others have noted, a chaotic system can certainly possess stable averages and deviations
—
Noting something and proving something are not the same thing.
Beyond that, even if you have proven the previous point, you haven’t proven that climate is one of those chaotic systems with this property.
(oh, I see we are still on the front page, so let’s continue)
Smokey said: “Prediction: John B will never accept any of those charts, because to admit that even one is valid will undermine the basis of his belief in catastrophic AGW. “
I accept that they are charts! I accept that they are valid in the sense that they do not actively lie (though some are so badly labelled that it is hard to tell). But my overriding view is, “it’s the trend that matters”. Yes, temperatures have been flattish for the last few years, but not long enough to conclude that the trend on late 20C warming has stopped.
Now, you tell me what you think of Arrhenius’ work, over 100 years ago, where he predicted the effects of increasing CO2 levels in such detail and every one of those predictions has now been observed. Lucky guess? Coincidence? Both of those things are possible (hence words in AR4 such as “highly unlikely” rather than “fact”), but do you really think so?
AGW rests on the following:
The characteristics of CO2-driven warming were predicted by theory (Arrhenius and others) and have been observed
No other plausible mechanism explains those observations
“Internal variablity” and all this chaos stuff are just obfuscation, amounting to saying “scientists don’t know everything, so they don’t know anything”.
Have you read KR’s link on the history of climate science? It’s a good read.
Here are quite a few charts of the *mild*upward trend from the LIA. This particular set of charts ends in 2000, but since then temps have been declining.
That debunks the claim that CO2 causes any measurable warming. CO2 has risen rapidly since WWII, but global temperatures haven’t changed frrom the trend beginning in the mid-1600’s. Rational folks will understand that CO2 has not caused the predicted accelerated warming. Just the opposite has happened, in fact. See the comment in red at the bottom of the chart.
I suspect the alarmist crowd will fall back on Trenberth’s ‘hidden heat in the pipeline’. But of course, that is simply baseless conjecture.
We need to look at the entire hemisphere, not just a few cities. So the chart I will fall back on is something like this:
http://www.ncdc.noaa.gov/paleo/ei/ei_image/highlat.gif
Interestingly, it comes from the same source, NCDC/NOAA, as some of yours. It also shows that warming has been greater at higher latitudes.
Stop the name calling for a moment and tell me, honestly, what you make of that chart.
This argument misrepresents the semantics of the word “predict.” The subject of a prediction is the outcome of a statistical event and not (as alleged by John B) the characteristics of this event. For example, in the event of a coin flip the subject of a prediction is whether the outcome is heads or tails and not whether the coin is bent.
Terry Oldberg – “* The characteristics of CO2-driven warming were predicted by theory (Arrhenius and others) and have been observed
* No other plausible mechanism explains those observations”
Terry – This argument misrepresents the semantics of the word āpredict.ā
So – it quacks like a duck, has white feathers like a duck, has webbed feed like a duck. Obviously, by your logic, it’s an elephant…
KR:
It sounds as though you missed my point. My point is that the entity predicted by a prediction is not a set of characteristics but rather is the outcome of a statistical event. To speak of predicting characteristics is to misuse the word “predict.” In the search for truth, it is crucial for all parties to use words properly for when they are used improperly a possible result is for a falsehood to seem true.
Terry:
Predicting a change with a particular set of characteristics is a prediction, especially when those characteristics can be used to distinguish between various causes of change. If I predict that dropping a vase will cause it to shatter, that’s a prediction of the effects of some change.
In this case, the characteristics of nights warming faster than days, polar amplification, increased IR at the surface, stratospheric cooling, etc., are all “fingerprints” of warming driven by greenhouse gas increases. And not the characteristics of solar increases, albedo changes, cosmic rays, or any of the alternative hypotheses for the warming.
We see the changes (global temperatures), the changes match the characteristics predicted for the greenhouse gas increases we have also measured – cause matches effect in all particulars.
KR, a night time slow down in cooling over land must be explained via the mechanics. These would be increased water vapor and cloud cover preventing nocternal radiative cooling (such as you experience in a dry desert at night). Are you saying that no other driver but an anthropogenic increase in CO2 can cause this night time slow down in cooling?
Terry Oldberg says:
June 15, 2011 at 9:37 pm
“For example, in the event of a coin flip the subject of a prediction is whether the outcome is heads or tails…”
Not so! When flipping coins, a reasonable prediction would be “1 out of 2 tosses will land heads”, as an individual toss cannot be predicted. You could also predict “10 heads in a row will be seen in 1 out of 1024 sequences of 10 tosses”. These are statistical predictions, based on what is known about coins. If they are not met, you can reasonably suspect something fishy is going on. But notice how we can predict that short term runs against the prevailing trend will happen, but that eventually the trend will prevail. Statisticians would not be surprised that in a long enough run of tosses, 10 heads turn up occasionally, in fact they would expect it. Sound familiar?
That is analogous to my use of the word predict, and to the way it is used in climate science. BTW, Arrhenius did provide numbers.
Smokey, I really would like to know what you think of my chart vs. yours. I think a discussion of that kind of evidence, what we each accept as evidence, and why, might be enlightening for both sides. But please stick to those charts to keep the discussion focussed. Anyone else like to weigh in?
Pamela Gray
A night-time reduction in cooling means less IR leaving to space. Clouds (of the appropriate type, namely high clouds entrapping IR) would do the job in this case, although satellite measures seem to indicate cloud cover decreasing with rising temperatures over the last 25 years. Low cloud cover increases should cool the earth by increases in albedo. Low cloud decreases alone would cause daytime warming but nighttime cooling.
So no – aside from stratospheric cooling, which is GHG effect only, any one or two of the characteristics could be caused by something else. It’s the group of characteristics that indicate AGW, not a single one or two. Nights warming faster than days, winters faster than summers, polar amplification, increased IR at the surface and decreased at the top of the atmosphere, cooling stratosphere (as IR is blocked lower in the atmosphere), faster warming over land than water – increased GHG’s are the only known cause that matches all the characteristics. And since we know we’re increasing GHG’s, that shouldn’t be a big surprise.
The Arrhenius link I posted earlier to his 1896 paper on “atmospheric carbonic acid” calls these characteristics out – except for the stratospheric cooling, which isn’t surprising since they hadn’t discovered it in 1896.
John B:
This thread has drifted off the track on which I tried to place it. I’ll try to put it back on this track.
You said (June 15, 2011) “The characteristics of CO2-driven warming were predicted by theory (Arrhenius and others) and have been observed. My objection was to your usage in this sentence of the word “predicted.” This usage makes “characteristics” the entity that is predicted by a scientific theory when it is the outcomes of independent statistical events that are predicted. I used “heads” and “tails” as an example because the set {heads, tails} contains a complete set of outcomes of a statistical event that is a coin flip.
If we agree that it is the outcomes of statistical events that are predicted by a theory and not a set of characteristics then the two of us are in a position to examine the scientific basis for AGM, for it is a set of independent, observed statistical events that provides the basis for testing the claims that are made by a theory. In a theory that is “scientific,” these claims are testable.
AR4 fails to identify the set of independent observed statistical events by which AGW can be tested or to reveal the results from such testing. Having searched without success for a description of these events, I believe they do not exist. If my belief is correct, AGM is not testable thus lying outside science. The year 2007 study of Green and Armstrong establishes that, at about the time AR4 was written, climatologists confused “projections” with “predictions” thus, perhaps, reaching the erroneous conclusion that AGW was testable and tested.
Terry Oldberg
I’m afraid your definition of testable does not match any I have run across in science, as it is far too limited.
Does the precession of Mercury’s orbit, and deflection of light near the sun, match the predictions of Einstein’s theory of relativity? Yes or no? (Yes.) Did the Michaelson-Morley experiment detect the difference in lightspeed expected due to the everpresent ether? Yes or no? (No.) And do the characteristics of warming match greenhouse gas increase and not other possible causes? Yes or no? (Yes.)
You can do the same thing in the old game of Clue – was it the butler in the parlor with the candlestick? If so, you would predict certain answers to various tests, and finding out the answers to those predictions lets you identify the culprit.
Predicted consequences of an action are entirely testable, Terry, your (re)definition notwithstanding.
Hey Mr Lynn,
Perhaps you could explain why Smokey’s chart of specific cities is “ample evidence” while mine, of an entire hemisphere, is merely “hypotheses stemming from said conjecture”. I really don’t understand what constitutes “evidence” around here.
John
Oops, wrong thread, but the request still stands,
This psuedo-mathematical post is not only misleading, but fundamentally wrong. See the ACTUAL mathematics here:
http://tamino.wordpress.com/2011/06/14/chaos/#more-3883
KR:
By “testable” I mean “falsifable” or “refutable.” Your analogy to Einsteinian relativity fails from the fact that relativity makes predictions while AGW makes projections and while predictions are testable projections are not.
@terry Oldberg
You are correct in that the IPCC reports contain projections, but you are not correct in saying that climate science does not make predictions. “GHG-induced warming will cause stratospheric cooling, warmer nights, etc.” are predictions. They are not exactly like the predictions of physics, but they are just like the predictions of, say, geology or evolution – in the sense that they predict things that we should find in the world if the theory is correct and not find if it is incorrect. e.g. evolution theory predicts that Australian mammals will be weird, and at the genetic level how weird. A prediction does not have to be to 10 decimal places to be valid, it just has to be distinguish between hypothesis and null hypothesis. “AGW will cause nights to warm more than days” is just such a prediction.
John B:
Thanks for taking the time to respond.
I am unable to reconcile your claim that “AGW will cause nights to warm more than days” is an example of a prediction with the elementary ideas of statistical reasoning. As I understand these ideas in relation to global climatology they are as follows:
* As only a single object (the Earth) is under observation, a study of the climate is an example of a “longitudinal study.”
* A longitudinal study divides the time into non-overlapping segments;
* Each segment is associated with an independent statistical event.
* A statistical “sample” is a subset of these events, each of which is observed.
* A “theory” is a procedure for making inferences.
* Inferences are prone to being incorrect.
* Predictions are associated with the kind of inference that is called a “predictive inference.”
* A predictive inference is a conditional prediction.
* In making a predictive inference, an extrapolation is made from the state of nature that I’ll call the “condition” to the state of nature that I’ll call the “outcome.”
* A “condition” is a condition on the associated theory’s independent variables such as “the CO2 concentration exceeds 400 ppm.”
* An “outcome” is a condition on the associated theory’s dependent variables such as “the global temperature anomaly exceeds nil.”
* A pairing of a condition with an outcome such as “the CO2 concentration exceeds 400 ppm and the global temperature anomaly exceeds nil” provides a partial description of an event.
* A “condition” is defined at the starting time of an event
* An “outcome” is defined at the ending time of an event.
* A “prediction” is a proposition that assigns a numerical value to the probability of each of the several possible outcomes of a particular event at the time the condition of this event is observed but before the outcome is observed. For example, given the condition that “the CO2 concentration exceeds 400 ppm” a prediction assigns a numerical value to the probability of the outcome “the global temperature anomaly exceeds nil” and a numerical value to the probability of the outcome “the global temperature anomaly does not exceed nil.”
* The probabilities of the conditional outcomes are called “state transition probabilities.”
* A theory is tested by comparison of the numerical values that are assigned to the state transition probabilities by the theory to the relative frequencies of the same conditional outcomes in a sample that is reserved for testing.
Perhaps you could enlighten me on the relationship of these ideas to the entity that you describe as a “prediction.” Also, if the purpose of regulation of CO2 emissions is to control the global temperature anomaly don’t we need the outcomes to be defined on this anomaly?
Now apply, your logic to the predictions of geology or evolution. Like I said, climate science is not quite like physics because, as you said, we only have one world.
It just occurred to me, that there are people who doubt the scientific nature of evolution and geology, too. Creationists! Hmmm.
Terry Oldberg
Let’s take this down to the basic, a testable assertion that informs our knowledge.
If A then B
If Not A then Not B
B is testable
The state of B tells us whether A is true or not.
The prediction is a consequence (or a set of them, in the case of greenhouse gas fingerprints) that can be tested – and if those consequences turn out to be true, we’ve tested the prediction. That was the point of my earlier examples of Einstein (orbit of Mercury, light deflection near the edge of the sun) and Michaelson-Morley (light speed difference due to passage through the ether).
As John B said earlier, the IPCC report certainly contains some projections. But the climate science as a whole contains a great number of predictions – teh characteristics of CO2 absorption, temporal and spatial distributions of temperature change, temperature change versus greenhouse gas concentrations, etc.
You have overdefined yourself into a corner, I believe, with a definition that simply does not include a great deal of how science is done.
KR:
Your description of the ideas surrounding the idea of a scientific theory is incomplete. Toward the end of eliminating ambiguity of description, I’ll complete the description before proceeding.
Let A and B designate propositions. The following 3 sentences provide the instructions for making a kind of inference:
If A then B
A
Therefore B
Assume that these instructions are cast into the form of a computer program. A running instance of this program on a computer is called a “process.” Each such process concludes that B is true.
By stipulation that A is the observed state of a physical system at time tA and that this system has an observable state at time tB, the conclusion that B is true may be made testable. A computer program that has been made
testable in this way is an example of a scientific theory. It is “scientific” because it is testable by reference to the observable states A and B.
Let tB exceed tA. Then the conclusion that B is true is an example of a “prediction.” A single prediction is made by every process.
Associated with every such process is a statistical event. In each such event the the state A is observed at time tA and then the state becomes observable at time tB. A subset of these events share the property that at time tB the state is observed; these events are called “observed statistical events.” The subset of the statistical events that are observed and in addition are statistically independent form an example of a “statistical sample.”
A sample provides the basis for testing the theory. In a test, the observed state at tB is compared to the predicted state. As the theory predicts B in each of the observed statistical events, if the observed state is not B in at least one of them the theory is falsified by the evidence. Otherwise, it is said to be “statistically validated.”
In a predictive theory that plays a role in the functioning of a control system, the state A is replaced by the set of states { A1, A2…} and the state B is replaced by the set of states { B1, B2…}. Given that A1 is observed, B1 is predicted and so forth. The aim of the control system is to place the system in that state in { A1, A2…} for which the state in { B1, B2…} is most advantageous. A requirement for the development of a control system is for { B1, B2…} to be identified and for a theory to be developed with reference to this set of states. That a theory has been developed with reference to some other set of states is irrelevant. In particular, that a theories have been developed that predict aborption by CO2, temporal and spatial distributions of temperature change, temperature change versus greenhouse gas concentrations is irrelevant unless the associated states are in { A1, A2…} and { B1, B2…}.
If such a theory had been developed for the purpose of regulating greenhouse gas emissions, this development would leave in its wake a description of the statistical sample in which this theory had been tested plus an account of the results of the testing. I don’t find citations to these items in AR4. Where are these citations?
And, like I said, try applying your logic to evolution or geology. Not all sciences work the way you suggest. Climate science is not physics, but it IS science.
Terry Oldberg
You are demonstrating a basic error here. The test form is actually a modus tollens, not a modus polens (http://en.wikipedia.org/wiki/Modus_tollens), in that the set of consequences (not a single consequence, but rather multiple consequences/fingerprints) must be present for the consequences to be true. And each hypothesis (antecedent) has multiple consequences/predictions, not just one.
In other words:
If A then {B, C, D, E…}
If not {B, C, D, E…} then not A
The set of consequences matches greenhouse gas emissions, and invalidates other explanations
Your formulation is strictly one of modus ponens, starting with the observation of the antecedent “A” and checking for “B” as a tied result, in other words testing the relationship between antecedent and consequent. What I am describing, and in fact what is present even in Arrhenius 1896, are a set of consequences driven by the physics that will be observed if the antecedent becomes true – and the presence or absence of those consequences are very very testable.
I will note that this is not an absolute proof of “A”, but rather a disproof of various alternatives. If another antecedent could be found that also matched the set of consequences/fingerprints, other tests would have to be found to distinguish between those cases. So far, though, I haven’t seen any such hypotheses that match the observed fingerprints of GHG’s and are remotely physically plausible – “Just So Stories” notwithstanding.
Having read the site your name links to (and having some experience in deductive and inductive logic myself), I believe you are simply over-complicating matters. Predictions based upon physics are eminently testable – you’ve just overdefined yourself as part of the framework you put forth on your (overly complex) website.
All that said, considering the amount of work you put into your website, I suspect you are unlikely to be convinced. My hope is that other (less ‘locked in’) readers will find something useful in this discussion.
Adieu
My apology, in the previous post, first paragraph:
…must be present for the consequences to be true
should be:
…must be present for the antecedent to be true
Perhaps, just perhaps, I will learn to type properly some day…
Whoops. I left out an HTML operator in my June 19 at 4:25 post. This post should have read:
KR:
It sounds as though you are bailing out of our debate to avoid admission of defeat. If this is the case, adieu.
KR says:
June 19, 2011 at 2:52 pm
Terry Oldberg
You are demonstrating a basic error here. The test form is actually a modus tollens, not a modus polens (http://en.wikipedia.org/wiki/Modus_tollens), in that the set of consequences (not a single consequence, but rather multiple consequences/fingerprints) must be present for the consequences to be true. And each hypothesis (antecedent) has multiple consequences/predictions, not just one.
In other words:
If A then {B, C, D, Eā¦}
If not {B, C, D, Eā¦} then not A
The set of consequences matches greenhouse gas emissions, and invalidates other explanations
Your formulation is strictly one of modus ponens, starting with the observation of the antecedent āAā and checking for āBā as a tied result, in other words testing the relationship between antecedent and consequent. What I am describing, and in fact what is present even in Arrhenius 1896, are a set of consequences driven by the physics that will be observed if the antecedent becomes true ā and the presence or absence of those consequences are very very testable.
What you are describing is linear Catholic logic, the foundation of western scholarship. e.g.
Death came through sin
Sin came through the law
Christ’s propitiation has met the righteous requirements of the law
Thus while in Adam all died
Even so in Christ shall all be made alive
This is for some purposes appropriate and useful of course.
But in scientific research into complex nonlinear systems, archetypally climate, linear Catholic logic is inadequate and cannot find a solution.
What is needed is structured logic. Essentially linear Catholic logic is along the lines of:
if A then B if B then C if not C then D if D then E etc…
While structured logic is something like:
if A and if B and if C and if not D then WORD
phlogiston
Not at all – I was pointing out that Terry was arguing about the relationship between the antecedent and the consequent(s) (modus ponens) rather than using sets of predicted consequences as test as to what the initial cause of warming is (modus tollens, as I had presented it). That’s the logical misinterpretation I was referring to.
As to the logic I was discussing, this is simple first order logic, equivalent in both forms you presented.
I’ll note that there’s plenty of statistical arguments and likelihood analysis in climate science as well – my examples are just some of the simplest predictions made by physical hypotheses that can be tested in an informative manner, contrary to Terry’s mistaken assertion that climate theory doesn’t make predictions. Terry’s logic, for some reason, doesn’t include these.
—
Terry – Yep, I’m leaving this thread. I believe I’ve stated my opinion pretty clearly over the course of these posts, and I’ll let readers make up their minds. You appear quite invested in your logical formulations and complications (an unpublished/unreviewed rewrite of inductive logic? Hmmm…), and aren’t going to be convinced, or for that matter listen – it’s not worth my while to repeat myself in the face of your rather immovable position.
Where I live (central England) the weather is highly unpredictable. The third day of the 3-day forecast is often nothing like what actually transpires.
The climate is pretty much the same from year to year, even though we always complain that the summer is disappointing – it’s temperate maritime. In over 60 years I’ve never seen it Mediterranean, Arctic, humid tropical or hot desert.
I can see this for myself. This article is totally misconceived and out of touch with reality.
By this response, KR fails to produce the citation I had requested to the statistical sample by which AGW might be refuted; having failed to produce this citation, he and resigns from our debate. Nonexistence of the citation would prove AGW to be irrefutable, thus lying outside science. After a diligent search stretching over more than 2 years, I’ve been unable to uncover any such citation.
KRās discursion on modus ponens and modus tollens is flawed and misleading and does nothing to diminish the requirement for refutability by reference to a statistical sample. His statements
and
match the pattern of neither modus ponens nor modus tollens unless the set
of propositions represents a proposition of some kind but if it represents a proposition this proposition can be represented by a single symbol such as
Sorry, in my previous post I omitted a couple of HTML tags.
By this response, KR fails to provide the citation I had requested to the statistical sample by which AGW might be refuted. The unavailability of such a citation proves AGW to be irrefutable, thus lying outside science.
KRās discursion on modus ponens and modus tollens is flawed and misleading and does nothing to diminish the requirement for refutability by reference to a statistical sample. His statements that
and
and
match the pattern of neither modus ponens nor modus tollens unless the set
unless this set represents a proposition but if it represents a proposition this proposition can be represented by a single symbol such as
; replacement by a
leaves my argument intact.
While KR claims AGW to be obvious to all who view the observational data with an open mind, what he has succeeded in doing is helping me to prove AGW to be a pseudoscience.
Maybe Iāll get it right on the 3rd try!
By this response, KR fails to provide the citation I had requested to the statistical sample by which AGW might be refuted. The unavailability of such a citation would prove AGW to be irrefutable, thus lying outside science.
KRās discursion on modus ponens and modus tollens is flawed and misleading and does nothing to diminish the requirement for refutability by reference to a statistical sample. His statements that If A then {B, C, D, Eā¦} and If not {B, C, D, Eā¦} then not A match the pattern of neither modus ponens nor modus tollens unless the set {B, C, D, Eā¦}represents a proposition (such as B AND C AND D AND E but if {B, C, D, Eā¦} represents a proposition this proposition can be represented by a single symbol such asB; replacement by a
B leaves my argument intact.
While KR claims AGW to be obvious to all who view the observational data with an open mind, what he has succeeded in doing is helping me to prove AGW to be a pseudoscience.