It was mainly a consistency exercise as several different combinations of sites have been used in the past 360 years
I remember Philip Eden used to publish his "own" version based on some of the original sites rather then the newer ones
I updated my own database when the changes came out last May and kept a list of the updates
Dec 1786 was missing on the old version (every day was set to 2.8C for the mean)
There were about 100 errors where the mean did not add up to (max+min)/2 and were corrected
There were about 30,000 out of 90,000 ish where fractions of a degree were adjusted but of these about 2,000 were by 0.5C or more not all in the same direction
Originally Posted by: lanky