Yes, the station in question is still the WS28xx, yes, there is a rainRate = prefer_hardware setting in the config. Thanks, that explains it. My best guess how the WS28xx driver calculates the rainRate is in some kind "hour-based rolling average": the Super Cell event brought about 25mm in one hour, a couple of weeks later another event brought ~38mm in one hour:
dateTime rain rainRate 1626545100 4.12799999999879 37.173 1626544800 3.09599999999955 36.3272727272727 1626544500 7.48199999999997 32.737 1626544200 3.8700000000008 27.9744444444444 1626543900 2.83799999999974 25.2572727272727 1626543600 3.86999999999898 22.004 1626543300 3.8700000000008 17.978 1626543000 1.29000000000087 15.296 1626542700 1.54800000000068 13.953 1626542400 1.03199999999924 12.716 1626542100 3.09599999999955 10.96 1626541800 2.57999999999993 8.253 gjr80 schrieb am Samstag, 7. August 2021 um 03:41:49 UTC+2: > I believe the answer will depend on your WeeWX configuration and the > station/driver you use. A default WeeWX install is set to obtain rainRate > from the driver but if the driver does not provide rainRate then WeeWX > will calculate it (rainRate = prefer_hardware under [StdWXCalculate] > [[Calculations]] in weewx.conf). I see in a previous thread you referred > to 'your WS28xx' - if you are still using the WS28xx then the answer is > likely in the WS28xx driver as the WS28xx driver is one of the few that > does provide rainRate. I had a brief look at the WS28xx driver and it is > not evident to me how it derives rainRate, but it is clear it does not > use the same algorithm as WeeWX - that is also evident in the archive data > you presented. > > If you wanted to use the WeeWX algorithm then you could change the rainRate > = prefer_hardware to rainRate = software, I'm not saying the WS28xx is > wrong, it just appears to use a different algorithm. On the face of it I > don't really see anything wrong with your figures, though granted the 25.02 > values are quite high given there was no rainfall for at least five minutes > for the second 25.02 value, that could be because the WS28xx tapers the > rainRate at a different rate to WeeWX. > > Gary > On Saturday, 7 August 2021 at 05:47:40 UTC+10 [email protected] wrote: > >> Thanks Gary. Do you have an explanation for this: >> >> I got hit by a Super Cell thunderstorm some weeks ago which delivered >> ~25mm in about 15 mins. Max rain rate for this month was those 25mm: >> >> archive_day_rainrate >> dateTime min mintime max maxtime sum >> count wsum sumtime >> 1624312800 0.0 1624312803 25.28 1624395078 376.134682539682 >> 288 112840.404761905 86400 >> >> select datetime, rain, rainrate from archive where datetime between >> 1624384800 and 1624399200 ORDER by datetime DESC >> dateTime rain rainRate >> 1624393200 0.0 25.02 >> 1624392900 0.0 25.02 >> 1624392600 1.03199999999924 24.737 >> 1624392300 6.19200000000092 21.512 >> 1624392000 17.2860000000001 8.48545454545454 >> 1624391700 0.515999999997803 0.076 >> 1624391400 0.0 0.0 >> >> gjr80 schrieb am Mittwoch, 28. Juli 2021 um 14:30:52 UTC+2: >> >>> WeeWX calculates rainRate by summing the rain seen in the last 900 >>> seconds (15 minutes) and scales this to a per hour figure by multiplying by >>> 3600/900. So your 21.8 mm appearing in one packet/record would have >>> resulted in 21.8. * 4 = 87.2mm/hr; exactly what you saw. Due to the 900 >>> second window being used the 87.2 mm/hr value would have persisted for up >>> to 15 minutes. The 34.9 value may have been the result of when the 21.8mm >>> value arrived, impossible to say without seeing the exact loop >>> packet/archive record history. You can change the 900 second window to some >>> other value, say 600 seconds (10 minutes) and WeeWX would then take the >>> rainfall seen in the last 10 minutes and scale that figure to a per hour >>> figure (in that case multiply by 3600/600). >>> >>> This is covered in the [[RainRater]] >>> <http://www.weewx.com/docs/usersguide.htm#%5B%5BRainRater%5D%5D>stanza >>> in the User’s Guide. Note the Userks Guide states that 1800 seconds (30 >>> minutes) is used by default for the window but the code actually uses 900 >>> seconds. Must fix the documentation. >>> >>> Gary >>> >>> On Wednesday, 28 July 2021 at 22:10:21 UTC+10 wysiwyg wrote: >>> >>>> Hi ! >>>> >>>> I have a situation where I'm not sure about weewx behavior and how to >>>> deal with it :-). >>>> >>>> I have made my own rain sensors: It reports rain every 2min30 and >>>> reports 0mm when there's no rain. >>>> Rainrate calculation is not done and not emitted by the sensor: Weewx >>>> has to handle it. >>>> >>>> Now come a situation: My sensor crashed yesterday and did not reported >>>> anything over the night (it was raining). >>>> I rebooted it a few minutes ago, without power cycle, so I don't loose >>>> the rain counter of the night => the sensor sent 21.8mm from the whole >>>> night in one shot. >>>> >>>> I would expect that the rainrate shall not be calculated...or maybe >>>> shall be calculate with a time span between last and previous sample >>>> (yesterday before it crashed). >>>> >>>> it seems not working this way as weewx reported: >>>> - 34.9mm/h at next report following the 21.8mm record. >>>> - 87.2mm/h the following report 5min later. >>>> >>>> I assume maybe weewx consider "not rain data" = "0 mm of rain" ? >>>> >>>> I have a couple of questions: >>>> - Is this behavior expected or maybe it's something that could be >>>> improved ? >>>> - Can I delete those rainrate sample from database ? >>>> of course my daily rain graph is a bit weird, but I want to keep the >>>> 21.8mm (so monthly or yearly records are ok), but I don't want to have my >>>> yearly/alltime max rainrate using this wrong data. >>>> >>>> best regards, >>>> >>>> >>>> >>>> >>>> >>>> -- 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]. To view this discussion on the web visit https://groups.google.com/d/msgid/weewx-user/37619713-c3af-4059-bb42-88bff1f7f798n%40googlegroups.com.
