Thinking some more... 1.5mm of rain so far today and a plot that shows about 40 000 mm/hour. Querying the station (I assume) every 3-5 seconds, lets say about 1 000 times/hr. 40 000/1 000 gives about 40mm/query. 40/1.5 gives a factor of about 26 which given my rubbery figures is about the number of mm in an inch. So maybe you are returning a daily sum in each loop packet and there is a conversion factor error. To diagnose the daily sum possible issue what do successive loop packets show now? If it's not raining do they include a value and what is it? If its 1.5mm then your driver needs to emit the difference between successive queries of the station (some other drivers do that, can't remember which though). Once that is sorted it should be a case of seeing what your driver is/should be emitting for each archive record and then comparing it with what is being stored in the database, obviously needs to be raining or some manual tipping of your station needs to occur to simulate rain. Shoudl be evident then whether there is a simple conversion error or not and where it is (weewx or driver).
rainRate I would worry about separately, if your hardware emits rainRate then you will need to deal with it in the driver. If it does not weewx will calculate it for you (correctly) once the rain issue is resolved. Oh, and you can't edit Google Groups posts, you can delete the post but not change it. Gary On Friday, 25 November 2016 09:27:23 UTC+10, gjr80 wrote: > > Hi, > > No solution but some thoughts. Could be unit conversions getting messed, a > factor of 10 is common when METRIC is involved since METRIC records rain in > cm > not mm as some might expect. But you are using METRICWX so that is not so > likely. If you driver is setting the usUnits correctly and the rain field > value is in these units then weewx should handle the storage/conversions > fine. Another thought, the rain field in a loop packet (or for that > matter an archive record) is the total rain measured during period the loop > packet (or archive record) covers. Could you perhaps be picking up a 'day > rain' value and hence reporting an accumulated rain value in every packet? > Guess you would really need to monitor your loop packets when it > starts/stops raining. > > Gary > > On Friday, 25 November 2016 09:12:41 UTC+10, [email protected] wrote: >> >> (Don't know how to edit my post) >> The WU time scale for the rain is that it started to rain about 0400 and >> finished about 0530. >> Susan >> >
