The following temperatures are read out by hand from this visualization. In principle one could of course code that and make curves etc.....
Anyways I wanted to have unaveraged temperature data from Antarctica stations, here are three stations.
Antarctica looks like a head with a Pinocchio kind of nose. Halley is on the forhead and Syowa and Novolazarevskaja are on top. Novolazarevskaya more to the front and Syowa more to the back of the head. All values are January.

Halley Station

1980 -5.6

1981 -3.7

1982 -3.9



1998 -4.7

1999 -4.6

2000 -6.4

2001 -5.6

2002 -4.5

2003 -4.3

2004 -5.2

2005 -4.4

2006 -4.9

2007 -3.6

2008 -6.6

2009 -5.3

2010 -3.9

2011 -3.5

Novolazarevskaja Station

1980 0.1

1981 0.0

1982 -2.8



1998 -0.3

1999 -0.2

2000 -1.8

2001 -0.7

2002 -0.4

2003 -0.2

2004 0.4

2005 0.5

2006 -0.6

2007 0.0

2008 -1.6

2009 -1.3

2010 0.6

2011 1.1

Unfortunately I found even a webcam of Syowa and the CH4 data but sofar no temperature curves.

Syowa Station

1980 0.6

1981 -1.2

1982 -1.6


1998 -0.2

1999 -1.2

2000 -2.3

2001 -1.0

2002 -1.8

2003 0.0

2004 0.3

2005 -0.5

2006 -1.0

2007 0.2

2008 -1.8

2009 -1.9

2010 0.4

2011 0.0

As a trend 2010 and 2011 look warmer than within an expectable average of the "plateau" years before, but especially by looking at the 8o's values not really out of fluctuation range.

This is certainly no statistical sound argument, but a hint that it may make sense to "look" (i.e. rather time consuming programming) alone even at the temperature of that stations together with their methane values.