Hi Gary, I've spent some time looking at the log files and focusing on the windgust, still unresolved unfortunately.
I’m using the interceptor driver, I looked at the comments in the interceptor driver code which says "*The gust measure has resolution of 1.1 m/s (2.46 mph) - one revolution of the anemometer."* However the lowest gust value in the LOOP logs is “0.683510010315”. There must be a units problem here, I understand that “0.683510010315” usunits=1=mph, displayed in the each log record, and will convert to 1.1 km/h, but the console shows 4 km/h, so I’m assuming that 1.1 is m/s which converts to 3.96 km/h, very confusing, but I can get the figures to match even if they don’t make sense to me. Is there a simple explanation about these figures? I notice that the LOOP & debug data for windgust is consistently the same, however debug data does not show the REC: records, these are only shown from the weewx direct logs. I'm unsure if the interceptor driver is sending the REC: records, do you know? I assume it is as weewx.conf says *“[StdArchive]* *….* *# Set the following to "software" to force software record generation.* * record_generation = hardware”* On Wednesday, 11 March 2020 22:10:40 UTC, Simon Davies wrote: > > Hi Gary, > > From my previous table "DB" showed the windGust value. > > As requested, I've taken readings from the weewx direct LOOP/REC data, > debug data, extract from the DB and I've taken a couple of screenshots > from the console. > Time period would be best to take from 21:03 for approximately 15 > minutes, although the logs cover a larger period. I've attached 5 files > with this information. I hope they contain the information you have > requested. > > I've not checked all of the values myself, I'll will do after posting > here. > > loop_hilo is set to True. > > Many Thanks Simon > > On Wednesday, March 11, 2020 at 4:11:45 AM UTC, gjr80 wrote: >> >> Hi Simon, >> >> I tried setting the debug level to 1 and viewed the syslog. The units >>> displayed in the log are usunits=16 (which I assume is meters per second?). >>> >> >> The usUnits decodes are 1 (or 0x01) is US (aka US customary) and uses >> miles per hour for wind speeds, 16 (or 0x10) is Metric with wind speeds in >> km per hour and 17 (or 0x11) is known as MetricWX with wind speeds in >> metres per second. >> >> >>> I could not understand how these values were used to be stored in the >>> archive database. >>> >> >> A little background on how units are used within WeeWX. First up the >> database, in weewx.conf under [StdConvert] you have a setting >> target_unit that will either be US, METRIC or METRICWX. This sets the >> database unit system used throughout the database. The default is US. Some >> folks insist on having target_unit set to the units they use natively, >> there really is little value in doing this unless you have specific >> requirements as WeeWX automatically converts to/from the units required (in >> fact I have seen it cause issues variously where database units are used >> (for example specifying QC limits without units), folks forget their >> database units are different to the normal and they can get unexpected >> results). Data coming from drivers, whether they be loop packets or archive >> records can be in any of the three unit systems, WeeWX automatically >> converts any data being saved to database to the target_unit unit >> system. Likewise any archive records or loop packets being processed >> elsewhere by WeeWX could be in any of the three unit systems. Data being >> output in a Cheetah report or plotted by the image generator is >> automatically converted to the units specified by the tag, skin.conf or >> weewx.conf or the defaults as applicable. Bottom line is when analysing >> WeeWX data never assume any units, always check the usUnits value. >> >> >>> See the table below, for example 1.1 mps becomes 1.36 mph. it should >>> be 2.46 mph using my calculations. Do you have any ideas why the archive >>> DB has this value? >>> >>> >>> >>> >>> >>> *time* >>> >>> *debug data wind gust (mps)* >>> >>> *debug data usunits* >>> >>> *console minutes* >>> >>> *console wind gust km/h* >>> >>> *DB (mph)* >>> >>> Mar >>> >>> 8 >>> >>> 22:40:57 >>> >>> 1.1, >>> >>> 16 >>> >>> 40 >>> >>> 4 >>> >>> 1.36702 >>> >>> Mar >>> >>> 8 >>> >>> 22:42:49 >>> >>> 0.0, >>> >>> 16 >>> >>> 42 >>> >>> 0 >>> >>> 0.68351 >>> >>> Mar >>> >>> 8 >>> >>> 22:44:57 >>> >>> 1.1, >>> >>> 16 >>> >>> 44 >>> >>> 4 >>> >>> 2.05053 >>> >>> As mentioned above you usUnits=16 data will be km/hr rather than m/s. >> That still doesn't help things tally. So what is 'DB', is that windSpeed >> or windGust? >> >> Appreciate the table may make things easy to read but perhaps it might >> help more if you provide the actual loop/archive record output when WeeWX >> is run directly, a copy of the log (with debug set) over the same period >> and a query from the database again covering the same period and your >> recorded observations of the console gain over the same period. >> >> So when there is some wind around let's run WeeWX directly for about 5 >> minutes. At the same time WeeWX is running directly note your wind speed >> values on the station display. Capture the console output (LOOP: and REC: >> data) for the entire period. Take a log extract that covers the entire >> period that WeeWX was run directly. >> >> Now query the database after finishing the direct WeeWX run. Am assuming >> you are using SQLite as your DBMS. Note you may need to install the sqlite3 >> application using sudo apt-get install sqlite3 if not already installed. >> Also assuming your database is named weewx.sdb and it is located in >> /home/weewx/archive (it may be located in /var/lib/weewx depending on >> your WeeWX install method - adjust following command as necessary): >> >> $ sqlite3 /home/weewx/archive/weewx.sdb >> sqlite> .headers on >> sqlite> SELECT datetime,datetime(dateTime,'unixepoch','localtime'), >> usUnits,windSpeed,windGust FROM archive ORDER BY dateTime DESC LIMIT 20; >> >> that will display the timestamp, human readable date-time, usUnits, >> windSpeed and windGust values for the last 20 minutes (use .quit to exit >> sqlite3). If run within 15 minutes of completing your direct run we should >> have some overlapping data to compare from (1) the loop packets and archive >> records seen by WeeWX (2) the raw data seen by the SDR driver (3) your >> visual observations of the station display and (4) the database. Post these >> four pieces of data. >> >> One final question, is your loop_hilo setting under [StdArchive] in >> weewx.conf set to True or something else? >> Gary >> > -- 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/d6dcd7cb-30dc-4218-b476-9c92aab6de6b%40googlegroups.com.
