Most accurate global temperature?
Further work on the satellite global monthly both RSS and UAH produces
This is the most accurate I can do at the moment.
A discovery was a time offset for the residual annual cycle and harmonics for both datasets, an unexpected and strange effect. Why? It ought to be centred on exactly one year.
I am nervous about allowing for this without knowing what is going on. Perhaps it is an artefact of the quantised satellite orbit mismatching orbital period. (is about 0.97 year instead of 1)
I’ve used datasets computed here from gridded because this avoids the Shannon limitation in the published data. (doubt it makes a material difference)
Data, intermediates, full explanations, ask.