Scepticism Of Man Made Climate Change

[jethro]

That's good. Scepticism is essential...

 

But arguing with a sceptical person isn't. The two threads were opened so that those of a like mind, could discuss the topics without the interjection or disagreement from other people. A space where both sides of this debate are at liberty to chat and knock around ideas, without the endless bickering. Sceptics who have criticised any aspect of AGW in the manmade thread are few and far between, the same cannot be said of this thread. All those in favour of AGW have made numerous pleas to be able to discuss the topic without the interruption of sceptical comments - you all now have that opportunity. A chance to show that it is the subject which interests you, not just the delight of arguing with someone else. 

[BornFromTheVoid]

We're all sceptical of certain aspects of climate change. If this is a thread for non-climate related posts that attempt to make judgements on climate change, then the name should be changed.

[Methuselah]

But arguing with a sceptical person isn't. The two threads were opened so that those of a like mind, could discuss the topics without the interjection or disagreement from other people. A space where both sides of this debate are at liberty to chat and knock around ideas, without the endless bickering. Sceptics who have criticised any aspect of AGW in the manmade thread are few and far between, the same cannot be said of this thread. All those in favour of AGW have made numerous pleas to be able to discuss the topic without the interruption of sceptical comments - you all now have that opportunity. A chance to show that it is the subject which interests you, not just the delight of arguing with someone else. 

But I'm not disagreeing, J...I'm merely trying to focus scepticism to where it is valid...I am simply not a denialist: both AGW and Natural Cycles are valid lines of research IMO...

 

But I am, and will always remain sceptical...

Edited July 4, 2013 by A Boy Named Sue

[Polar Maritime]

I would not call it an argument Jethro, reads more of a discussion to me...?

[jethro]

The point to this thread was clearly explained in the opening post from Paul. It's not a difficult concept to grasp, nor should it be difficult to adhere to. 

 

Anyone who cannot remember the point of these threads, or what they are designed for, can find the original post from Paul here: http://forum.netweather.tv/topic/76448-scepticism-of-man-made-climate-change/

[Gray-Wolf]

And surely there is a place for correcting 'errors' that folk might make?

 

If you are working on wrong assuptions then you are never going to be 'right' in your final assesment?

 

This was the point of the 'arctic circle' Temps comment? If you are trying to show folk that something is occuring then surely you need be 'correct' in your starting data and , in that case, the 80N temps were not doing that for the 'Arctic'?

[jethro]

And surely there is a place for correcting 'errors' that folk might make?

 

If you are working on wrong assuptions then you are never going to be 'right' in your final assesment?

 

This was the point of the 'arctic circle' Temps comment? If you are trying to show folk that something is occuring then surely you need be 'correct' in your starting data and , in that case, the 80N temps were not doing that for the 'Arctic'?

 

The purpose of the threads are clear. If you wish to change the purpose of the thread, you'll have to contact Paul. 

 

There is nothing to stop you taking information posted here, copying it into an appropriate thread and discussing it there.

[Methuselah]

The purpose of the threads are clear. If you wish to change the purpose of the thread, you'll have to contact Paul. 

 

There is nothing to stop you taking information posted here, copying it into an appropriate thread and discussing it there.

Now that's a thought!

Edited July 4, 2013 by A Boy Named Sue

[VillagePlank]

Just for fun, have a read of this

 

http://www.aussmc.org/documents/waiting-for-global-cooling.pdf

 

Spot any problems?

 

(i) You shouldn't use an unweighted moving average where seasonal patterns and/or trends are known to exist in the data (if CO₂ is causing an underlying trend, we must presume that it does exist) since it will exaggerate the direction of the trend and perpetuate a trend longer than it actually exists.

(ii) You shouldn't use linear regression (straight line trend) unless every point in the series is produced in exactly the same way: a property called homogeneity. None of the global climate series exhibit this property. (I know I've done it before, I just decided to check my work, and found my work to be faulty) Perhaps they might get away with it since in this case they are only using the last 15 years or so (where the statistical properties of the data are much more likely to be the same, but not guarenteed to be)

 

Is it any wonder as people learn more they end scratching their heads and saying 'I just don't understand' .... maybe it's just me.

Edited July 4, 2013 by Sparkicle

[VillagePlank]

Very worth reading the conclusions (p505) (and if you have time, the content) of this paper

 

Don't forget, these criticisms are mainly levelled across the entire blogosphere (of both sides of the debate) rather than peer-reviewed scientific texts. For instance, you'd never catch HadCrut doing such things [ii] which clearly isn't a moving average, nor a linear trend and of course, at the time of typing is clearly showing a reduction in the rate of warming.

 

http://onlinelibrary.wiley.com/doi/10.1002/joc.616/pdf

[ii] http://www.cru.uea.ac.uk/cru/data/temperature/

Edited July 4, 2013 by Sparkicle

[UV-RAY]

And surely there is a place for correcting 'errors' that folk might make? If you are working on wrong assuptions then you are never going to be 'right' in your final assesment? This was the point of the 'arctic circle' Temps comment? If you are trying to show folk that something is occuring then surely you need be 'correct' in your starting data and , in that case, the 80N temps were not doing that for the 'Arctic'?

Can anyone prove anything other than the globe warmed and now it's not. The causation of such cannot be proven one way or the other and on that its time to get back to what this thread is about.

[keithlucky]

2 highest level for Antarctic  sea ice ever  recorded GW causes warming and cooling or is it simply getting colder in the Antarctic

http://nsidc.org/data/seaice_index/index.html

Edited July 5, 2013 by keithlucky

[Gray-Wolf]

2 highest level for Antarctic  sea ice ever  recorded GW causes warming and cooling or is it simply getting colder in the Antarctic

http://nsidc.org/data/seaice_index/index.html

http://arctic.atmos.uiuc.edu/cryosphere/antarctic.sea.ice.interactive.html

[keithlucky]

Lets have a look at the IPPC leading lights take a look at the bunch of charlatans https://www. youtube.com/watch?v=pGPqV_ Fwo4Q …

[keithlucky]

Never mind GW hype GW is at a stanstill a futher quote from a GW supporter honesty

 

Prof Myles Allen from the University of Oxford, interviewed by the BBC this week about global temperatures, has finally admitted that:-

 

“no-one predicted the shorter-timescale lack-of-trend we have seen since 2000â€

 

Finally a bit of honesty.

Thank you, Professor.

[UV-RAY]

Never mind GW hype GW is at a stanstill a futher quote from a GW supporter honesty Prof Myles Allen from the University of Oxford, interviewed by the BBC this week about global temperatures, has finally admitted that:- â€œno-one predicted the shorter-timescale lack-of-trend we have seen since 2000†Finally a bit of honesty.Thank you, Professor.

Further tweaks needed with computer models me thinks. Why oh why will the globes climate not follow the theory I wonder!

[UV-RAY]

Lets have a look at the IPPC leading lights take a look at the bunch of charlatans https://www. youtube.com/watch?v=pGPqV_ Fwo4Q …

Great link and confirms what many of us always felt about the IPCC. The worlds greatest scientific minds my a**.

[VillagePlank]

Never mind GW hype GW is at a stanstill a futher quote from a GW supporter honesty

 

Prof Myles Allen from the University of Oxford, interviewed by the BBC this week about global temperatures, has finally admitted that:-

 

“no-one predicted the shorter-timescale lack-of-trend we have seen since 2000â€

 

Finally a bit of honesty.

Thank you, Professor.

 

Sigh. Now for what he *really* said ...

 

 

Prof Myles Allen from the University of Oxford told BBC News: "We predicted the temperature of this decade using a conventional detection and attribution analysis and data to 1996 (when lots of people were arguing there wasn't even a discernible human influence on global climate), and nailed it to within a couple of hundredths of a degree.

"There were plenty of solar enthusiasts back in the 1990s who were attributing the observed warming since the 1970s to a brightening sun - which didn't really work out when we moved into an extreme solar minimum and still saw the warmest decade on record.

He added: "It's only a single data point (and no-one predicted the shorter-timescale lack-of-trend we have seen since 2000) but it's still worth noting. Let's see what the next decade will bring."

 

http://www.bbc.co.uk/news/science-environment-23154073

 

Please post links, next time!

Edited July 6, 2013 by Sparkicle

[VillagePlank]

Presumably ensembles and statistics aren't going to solve the problem that your forecasts are only as good as your data and your formulas. It's surely a matter of timescales and how much evidence has gone into your formulas. If you're modelling a system with an infinite number of expressions, surely the only way to advance is trial and improvement.

 

Not really. Ensembles and statistics are the only way to getting close to solve a problem because the data is known to be incomplete. Also, it is the understanding that measurement error has and will always occur, and given sensitivity to intial conditions that measurement error, even down to 1/100,000/unit, can turn out to be significant as we predict further and further ahead.

Edited July 6, 2013 by Sparkicle

[UV-RAY]

Not really. Ensembles and statistics are the only way to getting close to solve a problem because the data is known to be incomplete. Also, it is the understanding that measurement error has and will always occur, and given sensitivity to intial conditions that measurement error, even down to 1/100,000/unit, can turn out to be significant as we predict further and further ahead.

But the conclusions are only as good as the data fed into them!

[VillagePlank]

But the conclusions are only as good as the data fed into them!

 

Not as simple as that. Regardless of the quality of data fed into them, the conclusion of one model run is virtually worthless. So you can have the best possible data and even though the model might be perfect, indetectable and minute differences will throw the end result. It is therefore essential that many runs are completed articificially nudging the data this way and that for each run where a view and hopefully a conclusion can eventually be attained.

[VillagePlank]

But the conclusions are only as good as the data fed into them!

 

For instance (as a corrolary to my last post) consider the equation commonly referred to as the logistic map ,

 

x_t = Rx_t-1(1-x_t-1)

 

This is a remarkably simple dynamical equation. Essentially, what we do is we say that x has a relation to a previous time period. So we choose a value of R (in this case we'll choose 4) and we'll plug in what x was the previous time period into the x we're currently interested in. Clearly, we're going to need as starting point for t=0; I'll use 0.2. Here's what the chart looks like:

 

 

Pretty unremarkable apart from the observation that the pattern appears to be complete unpredictable (no reoccuring cycles etc etc). Now, let's say we have a way of measure units to one 10 billionth of a unit ie 0.0000000001. That's a pretty small change, I'm sure you'd agree. Here's the chart with a starting condition of 0.2000000001 along with 0.2,

 

 

Now, if you're still awake, that's pretty amazing. It follows the pattern for, say, the first third, but then goes and does it's own thing a little later on. Maybe a fluke, so let's add the next one 10 billionth of a unit. Here's the chart for the three of them,

 

 

Same thing: the third attempt is nothing like the first or second. Here's what it looks like for,

 

 

So what can we say about this? Well we *know* that these series' are produced in a deterministic way, after all we know the equation that produces them. The only thing that has changed is the starting condition which we have perturbed (adjusted) by exceptionally small amounts which leads us, to effectively be unable to predict the final outcome no matter how accurate the initial conditions happen to be. Of course, a wild outlier as in incoming parameter will lead to a wild outcome, too. It is just like being between a rock and hard place.

 

It is therefore an obvious conclusion, in the case of the logistic map, that higher quality data does not necessarily lead to more accurate predictions. Weather and climate models certainly behave in this manner [ii] Take this morning's GEFS 850hPa ensemble output,

 

 

This should now look familiar with the classic signature of initial conditions vastly affecting the outcome. This is all obvious with the weather. When the runs bifurcate, we know that confidence in a forecast rapidly drops away; but what happens with climate predictions? Well, they publish ranges, too, based on ensemble runs, and statistical analysis of those runs.

 

https://en.wikipedia.org/wiki/Logistic_map

[ii] http://judithcurry.com/2011/02/10/spatio-temporal-chaos/

Edited July 6, 2013 by Sparkicle

[UV-RAY]

 

For instance (as a corrolary to my last post) consider the equation commonly referred to as the logistic map ,

 

x_t = Rx_t-1(1-x_t-1)

 

This is a remarkably simple dynamical equation. Essentially, what we do is we say that x has a relation to a previous time period. So we choose a value of R (in this case we'll choose 4) and we'll plug in what x was the previous time period into the x we're currently interested in. Clearly, we're going to need as starting point for t=0; I'll use 0.2. Here's what the chart looks like:

 

x0_20.png

 

Pretty unremarkable apart from the observation that the pattern appears to be complete unpredictable (no reoccuring cycles etc etc). Now, let's say we have a way of measure units to one 10 billionth of a unit ie 0.0000000001. That's a pretty small change, I'm sure you'd agree. Here's the chart with a starting condition of 0.2000000001 along with 0.2,

 

x0_1tb.png

 

Now, if you're still awake, that's pretty amazing. It follows the pattern for, say, the first third, but then goes and does it's own thing a little later on. Maybe a fluke, so let's add the next one 10 billionth of a unit. Here's the chart for the three of them,

 

x0_2tb.png

 

Same thing: the third attempt is nothing like the first or second. Here's what it looks like for,

 

xo_10tb.png

 

So what can we say about this? Well we *know* that these series' are produced in a deterministic way, after all we know the equation that produces them. The only thing that has changed is the starting condition which we have perturbed (adjusted) by exceptionally small amounts which leads us, to effectively be unable to predict the final outcome no matter how accurate the initial conditions happen to be. Of course, a wild outlier as in incoming parameter will lead to a wild outcome, too. It is just like being between a rock and hard place.

 

It is therefore an obvious conclusion, in the case of the logistic map, that higher quality data does not necessarily lead to more accurate predictions. Weather and climate models certainly behave in this manner [ii] Take this morning's GEFS 850hPa ensemble output,

 

t850Kent.png

 

This should now look familiar with the classic signature of initial conditions vastly affecting the outcome. This is all obvious with the weather. When the runs bifurcate, we know that confidence in a forecast rapidly drops away; but what happens with climate predictions? Well, they publish ranges, too, based on ensemble runs, and statistical analysis of those runs.

 

https://en.wikipedia.org/wiki/Logistic_map

[ii] http://judithcurry.com/2011/02/10/spatio-temporal-chaos/

Any prediction regardless needs quality data in order to make quality predictions otherwise we end up with such data continuously being tweaked. This is the case of all climate models as all of those predictions made a decade ago are already worthless, as any warming has remained static for sixteen years and counting. It's a case of junk in junk out and until we reassess the way our climate works then any future predictions will be binned in another decade.

[VillagePlank]

 It's a case of junk in junk out

 

No: it's *not* just a case of junk-in-junk-out. You can have 'perfect'-in-junk-out! Whilst, of course, we need to get the best data we can, crucially, it's still not enough, and reliance upon such idioms, within non-linear dynamical systems, is the perfect way to get faulty reasoning.

[UV-RAY]

No: it's *not* just a case of junk-in-junk-out. You can have 'perfect'-in-junk-out! Whilst, of course, we need to get the best data we can, crucially, it's still not enough, and reliance upon such idioms, within non-linear dynamical systems, is the perfect way to get faulty reasoning.

Time will tell Sparkicle , but the odds do not favour a rosy outlook for climate models in their current guise. Much more work is needed in understanding the bits we don't really understand as of yet. Not having a go at you by the way and thanks for the excellent replies by your good self.