This is just a quick post to get some charts up for a future discussion. Some of you might find it interesting though.
I have been playing with a static model that needs a good deal of work yet, but since there was a discussion on ocean oscillations and how best to use what for which, I thought I would get back to the battle of the hemispheres which is always going on and creates a lot of noise. Using the NOMAD3 Optimally Interpolated Sea Surface Temperature data, (click the link if you want to play). I broke the global SST up by zonal bands. Since I have a fairly good break down of ocean versus land area by 5 degree latitude, I used the two to create my own "Global" average SST, with "Global" meaning from 65S to 65N. Of course all the data sets have different bands they like to call regions, so this is not quite equivalent to the GISS regionals that use 24 degree zones, but fairly close. This gives me a "mean" SST of around 19.5 C which is lower than AQUA, but then AQUA uses 60S to 60N. There is no consistency it seems. Anyway, "global" average SST appears to be higher than the 16 to 17 C thrown around if you consider only ice free oceans as oceans.
Because of the internal lags, there is a little frequency modulation going on which is why I tend to spend a good bit of time on harmonics and RMS values instead of the whole "average" thing most prefer.