While experimenting with BoM raw Canberra Airport data I was surprised to find the UKMO have corrected the data here and there. It is noteworthy that Jones et al 1986 rejected Canberra (WMO 949260) with an 80 code which translates as “Non-homogenous and uncorrectable” – after comparing it with Sydney.
Canberra was “pardoned” and included in Jones 1994 station data – but oddly it never made the cut for the 1226 station 1996 update, which is confined to stations still reporting.
For larger graphic.
In the newly discovered Jones 1999 station data CANBERRA-AIRP 949260 is included again, with different “corrections” – all undocumented now of course and secret until ClimateGate.
Then the UKMO version just released under pressure of ClimateGate has a different set of “corrections” again and of course their reasons for the corrections are so far – secret.
The trends in degrees C per decade vary as follows: UKMO09 0.1, BoM 0.13, Jones94 1951-80 0.1, Jones99 0.057