On Feb 12, 10:24*am, Peter Franks wrote:
The land surface temperature record is dirty and must be constantly
corrected, often after the fact using 'best guesses' at correction, not
to mention the numerous stations that should be corrected, but aren't.
All of this stinks of a useless dataset.

For historical analysis, why not just use the ocean surface temperature
record? No correction needed.

?

OK, it's certainly a good sized statistical sample.
However, classic global warming theory says sea
surface temperatures rise slower than land surface
temperatures. (The water mixes, hiding the solar
heating with colder water from below, preventing
that energy from re-radiating,) Although the rates
are different, the global land and sea surface data
do show strong correlations. They both show a
warming planet.

The Sea Surface Temperature, or SST, Record

Here, from Hadley Centre, are the global sea surface
temperatures from 1850 to 2009. Please see:
http://www.cru.uea.ac.uk/cru/data/temperature/
http://www.cru.uea.ac.uk/cru/data/te.../hadsst2gl.txt
The yearly means of these data are graphed he
http://members.cox.net/rcoppock/HadSST2gl.jpg

Here are the slope data:
http://members.cox.net/rcoppock/hadSlope1850-2008.jpg