It is correct to take the data for the 09z to 09z period. However, you need to remember that the maximum figure for that period is thrown back to the previous day whereas the minimum is not. So for example the maximum temperature in the 24 hours to 09z today is taken as the maximum for the 6th. Whereas the minimum for the 24 hours to 09z today is the minimum for the 7th.
To get the most accurate data you should use the actual maximum and minimum for the sites rather than just the hourly figures. The maximum and minimum for Stonyhurst are not available publicly so for that station you do have to use the hourly data.
It is normal to get some sort of downward adjustment at the end of the month. Sometimes is can be small and on other occasions it has been as high as 0.5C. When I investigated this some time back it appeared to be because the provisional data reported by Hadley is not based on the actual CET stations but a number of other stations some of which are not even in the CET area. No idea why this is.
Originally Posted by: Global Warming