I have combined GHCN V3 data with HadSST3 to calculate the temperature anomaly using the same timebase as HADCRUT4, using my Spherical Triangulation algorithm. The resulting global average value is 0.69C. The spatial temperature distribution is shown below.
The global V3C/HadSST3 average is 0.69C which is up from the HadCRUT4.6 value for September (0.54C). The CRUTEM4 station data for October are not yet available. However any discrepancy will likely be rather small, as demonstrated below.
The annual average for 2017 based on the first 10 months is 0.74C which would make 2017 cooler than 2016 and roughly the same temperature as 2015.
I’ll calculate the HadCRUT4.6 equivalent value when the station data become available.
Update: As requested – Here is the temperature distribution viewed from the Southern Indian Ocean.