Problem with this years highs and lows records

[colinpb]

Whilst correcting temperatures in the log files I noticed I have a another problem, which I believe came about due to a console communications glitch so that cumulus downloaded the days 1 to 5 May again whilst downloading the 6 to 8 of May. Also 2 spurious records appeared, 1 Jan 2000 and a date in 1999! I even had a log file Jan00. So I spent a little time removing duplicates, but didn’t notice the following problem until now.

The “highs and lows” for the year are not the same as view “averages and extremes for the 2012”. When I looked in the “this year’s records editor” I could see why, see screen shot. I no longer have any records before 28 March and I presume this year’s records should have the same values as the dayfile and need amending to match dayfile.

Are the “averages and extremes for the 2012” correct because the data comes from the dayfile?

Colin

[steve]

Are the “averages and extremes for the 2012” correct because the data comes from the dayfile?

They're calculated 'on demand' from the data in dayfile.txt. The fact that you don't have any 'this year' records before 28th March suggest that there was some kind of problem preventing year.ini being read on that date (assuming you used to have records before that date) so they started again from that point.

The odd dates suggest that there was a problem with the clock on your PC resetting briefly; I've seen this happen on other systems when looking at the diags files.