On Apr 12, 10:22*am, Falcon wrote:
In article d47a02bb-592a-4c02-8917-

, Roger Coppock wrote...

On Apr 12, 7:47 am, Falcon wrote:

It's also time-dependent. Roger's not the only one who can draw pretty
graphs, but in this example http://i55.tinypic.com/iwrg35.png you can see
how useful a linear trend line is, depending on what you want to show.. The
data is from Roger's thoughtfully provided source. I added another trend
line that's probably a little more representative of what's been happening
lately.

You're cherrypicking, again
R^2=0.0075 means nothing.

So are you. You're using an entire record to show "continued warming", when
clearly, the latest records do not show that it's "continuing".

Why are you continuing with this "canard" ? The latest records can
show neither warming, cooling or flat, because the scatter about any
trend line drawn, however robust that might be in statistical terms,
is too great to allow meaningful inference.

Data for the last twenty years allows meaningful inference. It shows
warming is highly likely to have taken place. Data for the last thirty
years shows warming to be so likely, that any proposal it has not
warmed would be unthinkable.