"Roger Coppock" wrote in message
...
On Apr 6, 6:05 pm, "BobLl" wrote:
"Roger Coppock" wrote in message
...
Latest Satellite Data Show A Warming Global Climate

The satellite record, in all its current interpretations,
shows that the air near the surface is warming.
For background on the satellite temperature proxy please see:

http://en.wikipedia.org/wiki/Satelli...e_measurements

The URL below is one of the more conservative records
from the University of Alabama at Huntsville.

http://vortex.nsstc.uah.edu/public/m.../tltglhmam_5.2

The global data given above are graphed he

http://members.cox.net/rcoppock/UAH-MSU.jpg

The regression statistics for the line in the graph above
are below.

Coefficients:
Estimate Std. Error t value Pr(|t|)
(Intercept) -25.39055 2.13541 -11.9 2e-16
YEARMON 0.01277 0.00107 11.9 7.27e-28
---

Residual standard error: 0.179 on 362 degrees of freedom
Multiple R-Squared: 0.282
F-statistic: 142 on 1 and 362 DF, p-value: 7.27e-28

Yes I got the same result 0.0128 deg C warming per year for the full
global
data set since 1988. However if you do the same analysis for 1998 through
the present you get a cooling trend of 0.0044 deg C per year - going on
for
the last 10 years.

Yes, we all know that there was a warm spike a decade
ago. Fossil fool love to cherry pick data to place
that spike at the start of their biased interval.

Now be a good first year statistics student and compute
R squared a p-value for your cherry pie. A R-squared
of 0 and p-value of 0.359 is not significant. These
data do not support your statement about a cooling trend
over the last 10 years.

Coefficients:
Estimate Std. Error t value Pr(|t|)
(Intercept) 9.09247 9.62200 0.94 0.35
YEARMON[K:length(YEARMON)] -0.00442 0.00480 -0.92 0.36

Residual standard error: 0.181 on 133 degrees of freedom
Multiple R-Squared: 0.00634, Adjusted R-squared: -0.00113
F-statistic: 0.849 on 1 and 133 DF, p-value: 0.359

*************************************************

Looking at the data for the last 10 years is hardly cherry picking and the
data is noisy whether you use 20 or 10 years. Your statistics are valid
only if the variations about the line are gaussian but I agree the spike
hurts them. If you avoid the spike and use the data from 2002 onward you
get 0.023 deg C per year cooling with a r square value of 0.1 - not too
shabby given the limitations of the data.

However, I'm happy to accept your statistical argument that it cannot be
determined whether the global temperature has been warming or cooling for
the last 10 years. Make sure to tell the politicians.

How about the hemisphere disparity? No answers for that one I see.