Calculating the global temperature from reanalysis data

[xmetman]

This is a graph of the daily global temperatures that I've calculated for each climatic zone from the NOAA reanalysis temperature data since August 2011. I now realise this is the raw data, and will need to be weighted in some way due to latitude, because of course the values are in a grid of 73 x 144 equally spaced values (2.5°x2.5°), and areas close to the equator would be for a much larger area than those at the poles and will influence the average.

What's interesting to see though is :

• How the Pacific Ocean modifies the temperature in the southern temperate zone.
  
• How much colder the Antarctic is than the Arctic which I suppose is fairly obvious.
• How little variation there is in the tropics again fairly obvious.
• There is a slight seasonality to the global curve (dark line).

I just wonder if this gridded approach is how the CRUTEM and GISS monthly temperature series are arrived at?

[xmetman]

Here are the daily global unadjusted raw anomalies since 1948 which don't look that out of line with the CRUTEM4 and GISS values.