Head To Head
Log In
Register
U-Know! Forum »
Climate-Cat's out of the Bag!
Log In to post a reply

81 messages
Topic View: Flat | Threaded
stray
stray
2057 posts

Edited Dec 06, 2009, 08:17
Re: Hi Guys!
Dec 06, 2009, 07:25
Sorry, been busy so only just got round to this.

DarkMagus wrote:

Here's an example of my understanding of data. Say I'm using temperature records in a computer model taken from sites that have seen a change in urbanisation and therefore ambient temperature due to local man made influences. How do I correct for that with any degree of certainty? Do I pick a number that fits the trend I want to see? How do I know I'm right? I can't. Also, say cloud cover is a bigger influence on temperature than CO2, how do I model cloud cover for the last 100 years say? Where's the accurate data for that? You have to make something up. Only you couch it in some nicer language. Getting the picture? We simply do not have reliable data over any significant period of time. When I do any measurements, everything is traceable to national standards & therefore has some chance of validity. Can you say that of the temperature measurements from around the world for the last few centuries? I doubt it. Then factor in weather stations moving & the fact they are measuring at individual points, not integrating over wide areas & the picture starts looking even more unreliable.


Seriously ? Are you really just plucking (really bad, illinformed) scenarios out of the air without really knowing exactly what datasets are actually available in order to defend your position ? well, it isn't defending your position very well, and I say again, you don't seem to have much of an understanding of modelling/analysing complex datasets. This paragraph you have written kinda confirms it.

I'm guessing you've read someone going off on one about the 'aperture effect' as you mention cloud cover. So ? Does that negate the positive correlation between CO2 emissions and global temperature increases ? No, it doesn't of course (assuming here that you do know something of what you're talking about you should agree). The underlying trend/correlation is still there, yes ?

Sure, we don't know why it's happening exactly (remember the causal/statistical link between smoking and cancer here) and as in every other mathematical model of a non-linear, non deterministic, recursively adaptive system we can never, ever be sure we have all the numbers on the table. So ? What you definitely CANNOT say is global warming is BS when there is a definite statistical link between CO2 emissions and global temperature increases, seen ? What you can say, and we probably agree here, is that we have no way of sensibly predicting how bad things are going to get EXACTLY and WHEN.
We can say however things are bad and they are going to get worse.

Also

DarkMagus wrote:

Can you say that of the temperature measurements from around the world for the last few centuries? I doubt it.


You see, if you did know what you were talking about then saying that is being a bit of a cock right ? You do know there are ways of dealing with such inconsistencies mathematically right ? You do know that it is possible to create regions of legitimacy accross disparate datasets yeah ? You do know that an unknown variable at a particular time can be reasonably inferred from a seperate dataset where a calculatable relationship to the unknown exists ? Also, ever heard of regression analysis ? That could also probably be used against a more solid dataset. God there are so many ways of dealing with things like this you know ? Edit : I bet theres even some kind of awesome probability based approach you could use too (Peirce perhaps ?).

I could pull apart every bad example from the above paragraph but I won't, cos I'm taking you at your word that you do actually know something about the subject of modelling and data analysis. Therefore I'm putting this post of yours down to being from the cheap end of philosophy, a poor attempt to point score against a considered uneducated. However, if you do believe what you've said, and honestly think these problems you've raised are the major issues then seriously, read more.

Edit : All datasets have different health warnings attached to them in terms of their consistency and reliability. People like me know this, we know how to cope with it, and publish how the analysis was done, and how confident we are about the results/conclusions. Not only that, but there is this thing called peer review.

You do see that all you're doing is mumbling conspiracy theory with absolutely no real understanding of the science or the method right ?
Topic Outline:

U-Know! Forum Index