Yes, archive_day_rain holds the calendar day rain in the local time. The key dateTime in the main archive (table "archive") is in UTC.
-tk On Thu, Aug 30, 2018 at 1:40 PM Valley <[email protected]> wrote: > Indeed I am using METRIC, and interval of 300 (5 minutes). I was not aware > of that the stored unit is cm! Then I get very close to 25.8 but still not > exactly the same! > Unfortunately I have no idea how Netatmo calculates daily rainfall, it > justs reports it. Of course I could use Netamo's API to retrieve the values > from the Netatmo server (it is impossible to gain access to the local > receiver, everything has to go via the Netatmo server). However I thought > reading the weewx database was a quicker way for me than digging into > Netatmo's coding examples. Netatmo has an example, returning all modules > data from the last 30 days, and among other things I get: > > 27 August: > > sum_rain: 25.8 mm > > So, if I look at table archive_day_rain, field "sum" has a unit of cm. > What about the "unit" of date and time, are we talking UTC or local time? > If I look at the record that has the value closest to 25.8 (25.654) the > dateTime is 1535320800 which corresponds to 2018-08-27 00:00:00, so I guess > the date and time are in local values? > > When it comes to "daily" rain we have another complicating factor: In > official reporting "daily" is NOT the calendar day (00-24). To be more > precise: Officially, the total amount of rain of the 27th of Aug should > actually have been measured starting 27th of Aug at 08:00, and ending on > the 28th (!) of Aug 08:00. This is for Sweden (DST, in winter it is 07:00 > instead). > > This led me to query "archive" with "select 10*sum(rain) as rain_mm from > archive where (dateTime >= UNIX_TIMESTAMP('2018-08-27 08:00:00')) and > (dateTime <= UNIX_TIMESTAMP('2018-08-28 08:00:00')). Result 25.86 which > indeed is closer to Netatmo's own value of 25.8! > Querying "archive" with the calendar day "select 10*sum(rain) as rain_mm > from archive where (dateTime >= UNIX_TIMESTAMP('2018-08-27 00:00:00')) and > (dateTime <= UNIX_TIMESTAMP('2018-08-27 23:59:59')) results in the same > value as archive_day_rain, namely 25,65. > > This means that table "archive_day_rain" is a calendar day? For all other > parameters (pressure, wind etc) this is the same as in official reporting > but in my opinion "archive_day_rain should be calculated between DAY > 06:00:00 (UTC) and DAY+1 06:00:00 (UTC). > > /Valley > > > On Thursday, August 30, 2018 at 2:42:36 AM UTC+2, Thomas Keffer wrote: >> >> First of all, I suspect you are using a database with a unit system of >> METRIC, which measures in centimeters (not millimeters). >> >> As for the difference between 25.654 and 25.8, there can be several >> explanations. Here's one example: say you are using a 5 minute archive >> interval. For WeeWX, the first archive interval of the day is timestamped >> 0005. It measures the rain that fell between 0000 and 0005. However, many >> weather services include the measurement timestamped 0000. >> >> To get any more specific, we would need more information, such as exactly >> what unit system you are using, and how the Netatmo app is calculating >> daily rainfall. >> >> -tk >> >> >> On Wed, Aug 29, 2018 at 2:23 PM Valley <[email protected]> wrote: >> >>> I use weewx with a Netatmo rain gauge. It works well but I cannot figure >>> out how the total amount of daily rain is calculated. >>> >>> Since Python is too hard for me to learn, I use php coding to query the >>> database. >>> >>> The other day we got 25.8 mm of rain. A figure reported by my Netatmo >>> app on the iPhone and on the Netatmo web page. Also very close to what I >>> could read on an old fashioned measuring glass in the garden. So the figure >>> 25.8 mm is definitely ok. >>> >>> The graph in weewx also shows 25.8 (hard to read exactly, but somewhere >>> between 25 and 26) (see image). Now: If I look at the >>> table archive_day_rain for that specific day it says sum=2.5654 (also >>> wsum=2.5654). How is 2.5654 becoming 25.8? Difference between sum and wsum? >>> >>> I also queried the archive table with sum(rain) for all records for that >>> date. As a result I got 2.5654, the same as in the sum column in table >>> archive_day_rain. >>> >>> So the question is: How is the original data from Netatmo actually >>> stored, and how is it that the graph shows me the correct amount? >>> >>> Very puzzling for me! >>> >>> /Valley >>> >>> Here is a picture of the graph >>> >>> [image: raingraph.png] >>> >>> -- >>> 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 [email protected]. >>> For more options, visit https://groups.google.com/d/optout. >>> >> -- > 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 [email protected]. > For more options, visit https://groups.google.com/d/optout. > -- 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 [email protected]. For more options, visit https://groups.google.com/d/optout.
