Here's the most easily understood picture:
Its CERES Top-Of-Atmosphere (TOA) radiation balance plotted against surface temperature, for ocean-only grid points. I'm not desperately familiar with CERES, but lets take it literally, as radiation inbalance, averaged over a year. Fine. What do we expect to see?
[At this point, I recommend you to stop reading for a while, and see if you can work out what you would expect to see: what relationship do you expect to see, pointwise, between surface temperature and TOA radiation balance?
Back now? Good. Onwards:]
Knowing the good old "atmospheric heat engine" type analogy, we expect to see an excess of incoming radiation in the tropics (with heat transported to the poles by the atmosphere, not shown in this pic of course (but its fig 3 in this pdf)) and then an excess of outgoing radiation at the poles. Which is to say, the tropics are colder than you'd expect, from incoming radiation alone; and the poles are warmer than you'd expect.
And this is exactly what you do see: there's a positive balance for warm temperatures, and a negative balance for cold temperatures. There's a fair degree of scatter, of course, because the planet is far from simple; the land-ocean differences and mountains and vegetation differences all complicate the atmospheric circulation. Still, you see the basic picture. And its relatively simple, because we're looking at ocean-only. Note that there's a major complication over the tropics from the ITCZ, and from the sub-solar point changing over the year, so I think you'd expect the "slope" to be less there.
Now lets look at land and ocean:
This is, of course, the same picture as before but with land points added; and not all the points are see-through. The main difference is a pile of points on land below the freezing point of seawater. These, interestingly, go "backwards" - the colder it gets, the smaller the radiative inbalance. Those parts are pretty well all over Antarctica - there isn't much of the world that can get annual average temperatures less than -20 oC - and they're strongly over-weighted in the picture, because the dots are on a 1x1 grid - so they get a far bigger visual impression than the area they cover. The reason for the "reverse slope" is (I think) simply that there's less radiation about at lower temperatures: less in, less out; so while the "relative" inbalance would be even greater for these points, the absolute inbalance declines.
So there's interesting stuff to be seen in these datasets. Interesting, but basic: none of this is new, and you'd find it in textbooks if you looked, I'm sure. The shame is that the septics are so keen to find their fantasies that they can't see the interesting reality.
This is the same sort of stuff he got taken to task for by Spencer a few months ago. Please Willis, read a textbook!
The imbalance spatial pattern on the NASA page is very similar to that seen on an Antarctic topographical map. I would suggest the trend reversal is mostly due to the strong influence of altitude on surface temperature and, therefore, Planck response.
Another factor might be the correlation between distance from coast and latitude on Antarctica, particularly East Antarctica. The heat engine is partly ocean-driven so it's influence should generally be weaker the further you go inland. The less an area gets handouts from the heat engine the more it must live within its own means, resulting in a smaller imbalance (There's a metaphor for the Daily Mail to run with). By itself that means lower temperatures but the reduction in annual solar radiation by latitude (altitude too) cools those inland, less imbalanced, areas even further.
I saw that post and considered writing about it, but have kind of had enough of focusing on WUWT - for the moment at least. Glad you've done this though. WE virtually always misses the point.
[Agreed, just reacting to someone else all the time is a bit dull. Someone has to do it, though, and I did wait a bit to give you or Sou a chance -W]