I'm Not Sure I Understand This

One of the difficulties that climate scientists face is that it is not that easy to come up with a single global temperature.  Before satellites, with limited measurement points and 75% of the world under water, global temperature is not much more than a guess.  With satellites, the job is easier but not wholly straight-forward.

Spencer and Christy have been using NASA data for a while to try to compute a global temperature, and have released new results  (the top graph is the whole earth, the second is the northern hemisphere, the third is the southern hemisphere):


The first oddity is one that the climate community struggles with but downplays in public.  It is that increases in these tropospheric measurements should be, if the theory of CO2-based anthropogenic global warming is correct, higher than temperature increases observed on the ground.  In fact, just the opposite is true.  Why ground temperatures increases should be higher than troposphere increases is something no explained by the standard greenhouse models (but is explained by alternatives).

The second oddity is the difference between the northern and southern hemispheres.  As you can observe, there really has not been any warming in the last decades in the south.  Why should that be?  One might assume it is because CO2 is produced mainly in the northern hemisphere, but my understanding is that scientists a while back determined that there was incredibly good mixing in the atmosphere and that CO2 concentrations don't vary that much around the globe.  I know that the northern hemisphere tends to have more temperature variability at the ground, since it has more land and land heats and cools faster than over the sea, but I am not sure this is sufficient to explain the difference.