My e-pal, 'Fallenmonk', a hopelessly soppy, old, 'Confederate' Leftie but a good egg, was unimpressed with 'my' graph in the preceding post. That hurt because I was really proud of it, well not 'it' exactly, because I didn't produce it, but I managed, somehow, to re-produce it over here and regular readers will know that there has been a dearth of decent graphics on this site for some time ... well, for ever, actually. This is because I am to software what the quill pen is to a word processor! Anyway, 'FM' (as I call him), demanded a 'trend line'. This was obviously a "cunning plan" devised by him to throw me completely because he was guessing, shrewdly and correctly, that for all I know a trend line is something you go fishing with. However, with one bound I was free of his HAF (Hot Air Fanatic) machinations and I can now produce not just an old trend line (I mean, sooo last century!) but a super-duper, brand new "loess line" - don't ask! And, not just one new "loess line" but three of the little rascals!
What they show is confirmation that James Hansen, the chief honcho of the IPCC and spiritual leader to every HAF on the planet, made some alarming prophesies concerning global warming, predicting a "perfect storm" of disasters, to the US Congress 20 years ago but the consensus of satellite measurements (generally considered to be the most reliable instruments) shows that at best the global temperature is in a state of hiatus but at worst (from his point of view) it has actually cooled slightly.
Don't take my word for it, just pop over to Anthony Watts's place and read a very learned explanation by one of his regualr contributors, Basil Copeland.
David, while I appreciate your effort to find a statistical method that purports to discount us HAFs. I seriously question the application of loess methodology. The method is very dependent on some of the starting parameters you chose and depending upon data density and how granular you decide to be can show pretty much any answer you want.
There is absolutely no reason to suggest that a straightforward linear regression is not meaningful when describing the average temperature change over the recent past unless, of course, it gives an answer you don't like. You can search out a statistical method that can produce any result desired. I am sure the natives of the northern climes (yes including polar bears) will take some exception the conclusion that things aren't, in fact, getting warmer. When you can explain why this summer will probably see the first time in modern history that the arctic ice pack will be non existent then I'll sign up for your voodoo stats.
Posted by: Fallenmonk | Wednesday, 02 July 2008 at 03:16
"voodoo stats"
As in when is a 'hockey stick' not a 'hockey stick'?
(Unless you're involved in the HAF vs. the Rest of the World debate that might not make a lot of sense to you but all will be revealed later - when I've woken up properly!)
Posted by: David Duff | Wednesday, 02 July 2008 at 08:49