Midnight data applied to previous day in NOAA report

[Dan Ciarniello]

I recently noticed that the earliest data record in my DB which happens to be at midnight is listed in the NOAA reports on the previous day. 

To illustrate, the dateTime of the first data record in my archive table is 1513670400 which corresponds to Tue 19 Dec 2017 12:00:00 AM PST but it's included in the NOAA reports on Dec 18.

A side-effect of this is that the Belchertown Records page gives Dec 18 as the day with the smallest daily temperature range with a value of 0 degrees.

This seems a little strange to me so I'm wondering if it's intentional or a bug.  If it's intentional, what's the rationale?

[Tom Keffer]

It's intentional.

Say you have a ten minute archive interval. When a weather station emits a record at, say, midnight, it actually includes observations from 2350 to 0000, that is, from the previous day. Now, you could timestamp this 2350, but most weather stations actually timestamp it with the "current" time, which is 0000.

WeeWX consistently uses this time convention.

-tk

--
You received this message because you are subscribed to the Google Groups "weewx-user" group.
To unsubscribe from this group and stop receiving emails from it, send an email to weewx-user+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/weewx-user/42a334be-8095-43ab-8ce1-d4c72abea926n%40googlegroups.com.

[John Kline]

It’s intentional.  Midnight is considered part of the previous day.  If you think about it, the midnight archive period is for the period from <archive-interval> minutes before midnight (often 23:55) through to midnight.  It belongs in the previous day.

On Feb 15, 2021, at 4:41 PM, Dan Ciarniello <dan.cia...@gmail.com> wrote:



--

[Dan Ciarniello]

Yes, of course, that makes perfect sense. 

Thanks for the explanation.