New Research

[knocker]

UMass Amherst Climate Researchers Find Evidence of Past Southern Hemisphere Rainfall Cycles Related to Antarctic Temperatures

AMHERST, Mass. - Geoscientists at the University of Massachusetts Amherst and the University of Minnesota this week published the first evidence that warm-cold climate oscillations well known in the Northern Hemisphere over the most recent glacial period also appear as tropical rainfall variations in the Amazon Basin of South America. It is the first clear expression of these cycles in the Southern Hemisphere.

The work by Stephen Burns and his doctoral student Lisa Kanner at UMass Amherst is reported in the current issue of ScienceXpress. Burns says, "The study also demonstrates that rainfall in the Southern Hemisphere of South America is, though to a lesser extent, also influenced by temperature changes in the Antarctic, which has not been previously observed."

http://www.umass.edu/newsoffice/newsreleases/articles/144768.php

[Village]

It would take more effort than I can be bothered to waste to explain to you that there are clear, unambiguous laws that prevent the complexity of weather being vectored up to predict climate. It's rather like the difference between counting to ten in halves and calculating the mean of all of those halves. The point is that these are different things. One is predictable, one isn't.

But, I digress.

I am pleased you didnt try to explain in your own words how you think the mathematics of the models work. I understand why interpolating data renders the whole procedure useless. I have worked with programers for twenty years here in the city to produce mathematical models. These models use the same techniques and formula and we still cannot predict where the financial markets at a future date.

At least we agree on one thing....the models are a waste of time for long term predictions. No wonder the Met Office gave up trying.

Edited January 19, 2012 by Village

[Village]

As I said, as more information is gleaned it is fed into the models, the end result changes.

Would you rather they didn't add new information and simply stuck with old data? Hardly scientific.

You missunderstand my point.

My point was that they are completely wasting their time going about the process in that way they do. One cant create a synthetic mathematical model to predict future conditions in the real world from the inside out. Seeing as chaos does exist in the real world then they dont have a hope in hell.

Edited January 19, 2012 by Village

[Devonian]

If climate is chaotic how come we have seasons?

[VillagePlank]

I am pleased you didnt try to explain in your own words how you think the mathematics of the models work. I understand why interpolating data renders the whole procedure useless. I have worked with programers for twenty years here in the city to produce mathematical models. These models use the same techniques and formula and we still cannot predict where the financial markets at a future date.

At least we agree on one thing....the models are a waste of time for long term predictions. No wonder the Met Office gave up trying.

I am not talking of interpolation.

Interpolation is, essentially, taking two timeseries coordinates and inferring a value that falls between them. That can be a part of mathematical modelling, but it isn't what I'm talking about. Interpolation is a very specific technique. I've kindly added the Wiki leak at the bottom of this post. And if you didn't know that, then maybe your programmers (two 'm's by the way) have been leading you up the garden path, claiming that they're great, taking a huge salary out of the pot, and when it doesn't work, claim that, hey, it's unpredictable, isn't it?

And since it's a waste of time for long term predictions, I will place a £100 bet with you that the December CET, 2014 will post a lower value than the July CET, 2014. You'd do well not to agree to the bet - but, morally, you must, since you claim that neither weather or climate is predictable. I can also tell you where the moon will be, the planets will be, and where the stars will be - with, maybe, the exception of Betelguise (that might not be there, by then) Perhaps, raise the stakes to £1000?

I'm willing to put my money where my mouth is ...

[1] http://en.wikipedia....i/Interpolation

Edited January 20, 2012 by Boar Wrinklestorm

[VillagePlank]

I'm going to give it a try to show why climate is predictable, and weather isn't.

Firstly, we have to have a system that is capable of undergoing complex, chaotic behaviour. Perhaps one of the most studied is Robert May's Logistic Map [1] ,

It's often said that as soon as one puts an equation in a post, a book, or an article, you halve the number of readers. I urge you to keep reading; whilst this does look like a gobbledegook of symbols, it's actually rather easy to work through. This is an example of a dynamical system [2] ie we determine values of a system against time against itself. This type of system, regardless of what you might hear, is entirely, deterministic [3]

Let's get on and give it a go. First the letters. x is the value we're looking following, and the subscript t is the time period we want to find x for. So x_t means x for the time, t, and x_t-1 is whatever x was for the immediately previous time period. Robert May wanted to model population growth in an arbitrary ecosystem, so he introduced the term k, which is the growth rate of a population - ie the combined rate reproductive success and death. Let's do a quick run-through. We'll start the growth rate at k=4, with an initial population of x_t-1=0.4

x_t, then, equals 4 * 0.4 * (1 - 0.4)

= 4 * 0.4 * 0.6

= 0.96

If we run this for 40 time periods, we get a chart that looks like this,

Not very interesting unless you are a biologist studying ecosystems populations! But, things do get a little more interesting when we make further runs. Let's change the growth rate by 1/10,000th (0.0001) That's some small amount - what should we expect to see? Perhaps the graph nudged up a little bit, perhaps no discernible difference. Here's that chart overlayed with our primary run,

Now this should seem odd. Why should such a small difference make such a large difference at the end of the 40 time periods? It makes so much difference that after just 40 iterations the series are completely uncorrelated. Using the same maths, using the same model, but offereing a perturbation of some 1/10,000th. This is the colloquially known as the butterfly effect. Small perturbations in the system have massive ramifications later on. So much so that, even though the system is deterministic, it's qualitative nature cannot be predicted with any degree of accuracy.

Clearly, this is analogous to forecasting weather. Indeed, this phenomenon is exactly why weather cannot be predicted. The Navier Stokes equations produce non-linear output in exactly the same manner as this model. This is amply demonstrated by charting six runs all perturbed by the 1/10,000th,

To any avid weather watcher this should be immediately recognisable as being very similar to GEFS ensemble output. What happens if we chart 20 runs? Here we go,

Clearly we are into a complete mess. It's worth restating that the difference in the runs from our first run, to twentieth run is between 4.0000 and 4.002. Not much to get a mess!

But we can still conduct an analysis on this. Let's look at the complete and utter mess between t=20, and t=40, and average the values between for each run. Here's the chart,

Now something is immediately obvious. Why isn't that average a random walk, why is the average over all of those runs restricted to within a range of 0.17 when each of those runs could post any average between 0.0 and 1.0? Obviously, this is the climate analogy; even though each run is thoroughly unpredictable we can say with certainty that overall the averages will post between a (highly) restricted range.

And there you have it - I can make a prediction that any of the runs in the system will, in the future, post an average of approximately 0.52 give or take +/- 0.085.

[1] http://en.wikipedia....ki/Logistic_map

[2] http://en.wikipedia.org/wiki/Dynamical

[3] http://en.wikipedia....i/Deterministic

Edited January 20, 2012 by Boar Wrinklestorm

[Village]

If climate is chaotic how come we have seasons?

I am sure we have seasons because we dont live at the equator, this has nothing to do with chaos theory. What is your point Dev?

[Village]

I am not talking of interpolation.

Interpolation is, essentially, taking two timeseries coordinates and inferring a value that falls between them. That can be a part of mathematical modelling, but it isn't what I'm talking about. Interpolation is a very specific technique. I've kindly added the Wiki leak at the bottom of this post. And if you didn't know that, then maybe your programmers (two 'm's by the way) have been leading you up the garden path, claiming that they're great, taking a huge salary out of the pot, and when it doesn't work, claim that, hey, it's unpredictable, isn't it?

And since it's a waste of time for long term predictions, I will place a £100 bet with you that the December CET, 2014 will post a lower value than the July CET, 2014. You'd do well not to agree to the bet - but, morally, you must, since you claim that neither weather or climate is predictable. I can also tell you where the moon will be, the planets will be, and where the stars will be - with, maybe, the exception of Betelguise (that might not be there, by then) Perhaps, raise the stakes to £1000?

I'm willing to put my money where my mouth is ...

[1] http://en.wikipedia....i/Interpolation

I think you are going off on a tangent with this. As you said earlier...you have digressed.

One doesnt need a computer program and in depth mathematics to show that it is on average much warmer on our lattitude in June than December.

However, a dumb computer program will never be able to predict in December the next day that snow will fall in June. If you want to believe that you can predict this with the aid of a computer model, then I will offer you 500 pounds to say that you havent a chance in hell of being right.

Further, drawing a comparisson with totally predictable heavenly bodies and an unpredictable climate I think is not a true representation of the the glaringly obvious differences. Do you really believe the climate models for 100 years forth are as accurate as the clockwork of the heavens? Are you sure that you have thought through your stance on this properly?

Edited January 20, 2012 by Village

[VillagePlank]

However, a dumb computer program will never be able to predict in December the next day that snow will fall in June. If you want to believe that you can predict this with the aid of a computer model, then I will offer you 500 pounds to say that you havent a chance in hell of being right.

I can certainly create a computer model to show that the CET in December will be less than the CET in July. A touch of orbital mechanics, some standard Stefan-Boltzmann law stuff, a little bit here and there, and hey presto - I will be able to produce output that correlates well enough to the seasons to show that the CET differences are, within two standard deviations, certain.

How do I collect my money?

[Village]

You get nothing for predicting its warmer in June than December, theres no skill in that and if you need to produce a computer programe to show that then you have digressed even further.

But if you really believe that computers can predict the future climate to any degree of accuracy you can create a model that will do precisely that. Problem is ....you wont be able to because nobody has managed it to date and I am sure it will never happen using a serial computer and current approach from outside inwards as I stated earlier.

Its no good pretending that computer synthetics can do this with any degree of accuracy. If computer models could have done I would be on a 50 million pound yacht in the Mediteranean with a bunch of beautiful women and all my mates partying each day away at your and everyone elses expence.

Edited January 20, 2012 by Village

[laserguy]

I would be on a 50 million pound yacht in the Mediteranean with a bunch of beautiful women and all my mates partying each day away at your and everyone elses expence.

Al Gore's getting close....

[Village]

Yeah, because the chimps who were in power last bought the daft video and force fed it as scientific fact to our school kids!

Is it no wonder this generation leave school knowing nothing.

[knocker]

Is it no wonder this generation leave school knowing nothing.

Are you sure you didn't mean the last generation?

[Methuselah]

Are you sure you didn't mean the last generation?

[Devonian]

I am sure we have seasons because we dont live at the equator, this has nothing to do with chaos theory. What is your point Dev?

That your words show climate IS predictable. We can indeed all do it!

What you're doing is mudding the water here. We all know prefect prediction of climate isn't possible but equally to write as you have about how trying to predict the climate is a waste of time is nonsense.

Edited January 20, 2012 by Devonian

[VillagePlank]

You get nothing for predicting its warmer in June than December, theres no skill in that and if you need to produce a computer programe to show that then you have digressed even further.

But if you really believe that computers can predict the future climate to any degree of accuracy you can create a model that will do precisely that. Problem is ....you wont be able to because nobody has managed it to date and I am sure it will never happen using a serial computer and current approach from outside inwards as I stated earlier.

Its no good pretending that computer synthetics can do this with any degree of accuracy. If computer models could have done I would be on a 50 million pound yacht in the Mediteranean with a bunch of beautiful women and all my mates partying each day away at your and everyone elses expence.

Frankly, this is nonsense.

You asked if I could, I said I could, I want £500. Thanks, PM when you have the payment method sorted. You haven't countered any science I have offered. Yes, people have managed to predict climate years in advance - look at the CET prediction - that's climate, that's a prediction. Do you want to owe me £1500? You are clearly not skilled in either mathematics (you didn't know what interpolated meant) nor mathematical modelling (you can't distinguish between different systems type) and certainly have only probably managed to write 'Hello World' on a Sinclair Spectrum some decades ago. It's obvious, by your language, you are neither a skilled programmer, nor a computer scientist, nor anyone who would ever know anything about these subjects. Indeed, I am still a student, some 20 years in!

Thanks for trolling, thanks for wasting my time. Hey, I fell for it. Therefore, I am an idiot.

Edited January 20, 2012 by Boar Wrinklestorm

[Devonian]

I think you are going off on a tangent with this. As you said earlier...you have digressed.

One doesnt need a computer program and in depth mathematics to show that it is on average much warmer on our lattitude in June than December.

However, a dumb computer program will never be able to predict in December the next day that snow will fall in June. If you want to believe that you can predict this with the aid of a computer model, then I will offer you 500 pounds to say that you havent a chance in hell of being right.

Further, drawing a comparisson with totally predictable heavenly bodies and an unpredictable climate I think is not a true representation of the the glaringly obvious differences. Do you really believe the climate models for 100 years forth are as accurate as the clockwork of the heavens? Are you sure that you have thought through your stance on this properly?

Village, climate models don't predict the weather they predict the climate. The two are very different, please don't confuse them.

Edit: or do you really expect climate models to predict the weather?

Edited January 20, 2012 by Devonian

[Devonian]

Frankly, this is nonsense.

You asked if I could, I said I could, I want £500. Thanks, PM when you have the payment method sorted. You haven't countered any science I have offered. Yes, people have managed to predict climate years in advance - look at the CET prediction - that's climate, that's a prediction. Do you want to owe me £1500? You are clearly not skilled in either mathematics (you didn't know what interpolated meant) nor mathematical modelling (you can't distinguish between different systems type) and certainly have only probably managed to write 'Hello World' on a Sinclair Spectrum some decades ago. It's obvious, by your language, you are neither a skilled programmer, nor a computer scientist, nor anyone who would ever know anything about these subjects. Indeed, I am still a student, some 20 years in!

Thanks for trolling, thanks for wasting my time.

Years ago, trolls used to get banned. Hey, I fell for it.

Therefore, I am an idiot.

Right on the money until the last line. Edited January 20, 2012 by Devonian

[BornFromTheVoid]

RE: Village

[VillagePlank]

Yes, yes, BFTV, I am sure I will get edited down ...

[Village]

Frankly, this is nonsense.

You asked if I could, I said I could, I want £500. Thanks, PM when you have the payment method sorted. You haven't countered any science I have offered. Yes, people have managed to predict climate years in advance - look at the CET prediction - that's climate, that's a prediction. Do you want to owe me £1500? You are clearly not skilled in either mathematics (you didn't know what interpolated meant) nor mathematical modelling (you can't distinguish between different systems type) and certainly have only probably managed to write 'Hello World' on a Sinclair Spectrum some decades ago. It's obvious, by your language, you are neither a skilled programmer, nor a computer scientist, nor anyone who would ever know anything about these subjects. Indeed, I am still a student, some 20 years in!

Thanks for trolling, thanks for wasting my time. Hey, I fell for it. Therefore, I am an idiot.

ROFLMAO@you

Any idiot can say they can predict the climate hundreds of years forth. And saying white is white isnt a prediction. So far all the climate models have been wrong to date (thats why the IPCC downgraded their estimate of average global temps 100 years forth) and there is no hope for the future because the approach is wrong. Thats why they have to keep updating and fiddling because the new real data is always different from that previously input. Making out you can do it and nobody else can may wash with some but it doesnt wash with me.

You have come with no prediction of climate whatsoever, but you get a digestive biscuit for being a cheeky chump.

Edited January 20, 2012 by Village

[jethro]

Village - please may I borrow your time machine?

[Village]

Village, climate models don't predict the weather they predict the climate. The two are very different, please don't confuse them.

Edit: or do you really expect climate models to predict the weather?

You missed the point again Dev.

The point is that BW was maintaining that climate modelling can predict the seasons! Thats the nonsense of it all because he shouldnt need a computer and climate model. it should be obvious.

[Village]

Village - please may I borrow your time machine?

You sound like you dont trust the IPCC time machine. I thought you had fallen for that one hook line and sinker years ago! LOL

[jethro]

You sound like you dont trust the IPCC time machine. I thought you had fallen for that one hook line and sinker years ago! LOL

I see it for what it is - a rough approximation, likely to change as new information is fed into the system.

I want to use your time machine to visit the future you say has been wrongly predicted by the climate models.