Page 1 of 2

Wind bearing zero when calm

Posted: Tue 09 Jul 2019 7:09 am
by liddelljs
HI all,
I have two fine offset stations and have recently noticed that the wind bearing goes to zero whenever the wind gust is zero. I know there is a check box for this ("Wind bearing zero when calm") but I get the same behaviour whether the check box is checked or unchecked. Wondering whether anyone else has noticed this? I have only noticed it recently since I started exporting the cumulus data to an IOT platform where the charting only allows for line charts rather than the Cumulus point charts. With the line chart, the issue becomes very noticeable.

Cheers
Steve

Re: Wind bearing zero when calm

Posted: Tue 09 Jul 2019 12:28 pm
by Dennisdg
Have you tried shutting down MX and restarting after the change?

Re: Wind bearing zero when calm

Posted: Tue 09 Jul 2019 11:58 pm
by liddelljs
Thanks for the response Dennis. Tried restarting (with the check box checked and unchecked). Also noticed the same behaviour after updating from version 3046 to 3050. The fact it is happening on two independent systems (running on rpi 2) made me wonder whether it was a bug in the software but if others aren't seeing it, I'll need to keep looking.

Cheers
Steve

Re: Wind bearing zero when calm

Posted: Wed 10 Jul 2019 1:27 am
by beteljuice
Silly question ... What do you want the wind bearing to be when it is calm ????

Re: Wind bearing zero when calm

Posted: Wed 10 Jul 2019 8:16 am
by liddelljs
I just want it to keep reporting the current bearing. I assumed (perhaps incorrectly) that would be the behaviour if I left the check box unchecked. If the wind is calm and the bearing remains unchanged on (say) 270 degrees, then it reports a straight line at 270 degrees.

Cheers

Re: Wind bearing zero when calm

Posted: Wed 10 Jul 2019 10:16 am
by beteljuice
... but if it is calm there is no bearing.
If you give it one, then any calculations you may be doing eg. dominant wind direction, will be in error.

Can't you code around what you want to achieve ?

Re: Wind bearing zero when calm

Posted: Wed 10 Jul 2019 11:08 am
by freddie
I don't have a Fine Offset station so I don't know - but perhaps the FO is reporting a direction of 0 if the speed is calm?

Beteljuice is correct in that in the meteorological world if there is no windspeed then the wind direction is reported as 0. Wind directions are reported from 1 degree through to 360 degrees. The value 0 is reserved for those occasions when speed is reported as 0.

Re: Wind bearing zero when calm

Posted: Wed 10 Jul 2019 2:07 pm
by mcrossley
freddie wrote: Wed 10 Jul 2019 11:08 am I don't have a Fine Offset station so I don't know - but perhaps the FO is reporting a direction of 0 if the speed is calm?
I suspect that may be the case. If you do not check the "use zero for calm" box, Cumulus just uses whatever the station reports.

Re: Wind bearing zero when calm

Posted: Thu 11 Jul 2019 11:30 am
by liddelljs
HI all,
many thanks for the responses - all very helpful. Thanks Mark for confirming the Cumulus behaviour when the check box is unmarked. Would agree that it must be the station reporting the zero bearing. I'll need to do some observing of the local screen to confirm. Looks like I will have to take up Beetlejuice's solution and code out the problem in the IOT platform.

Thanks again
Steve

Re: Wind bearing zero when calm

Posted: Thu 11 Jul 2019 11:37 am
by water01
But if you do your site will not, as has been pointed out, follow the meteorological norm.
in the meteorological world if there is no windspeed then the wind direction is reported as 0.

Re: Wind bearing zero when calm

Posted: Fri 12 Jul 2019 11:37 am
by AllyCat
Hi,
liddelljs wrote: Thu 11 Jul 2019 11:30 amWould agree that it must be the station reporting the zero bearing.
No I don't think so. The FO transmitter sends only 4 bits corresponding to the 16 "possible" wind directions (N , NNE, NE, etc..). It definitely cannot distinguish between 0 and 360 degrees. There is another flag for "direction unknown" but AFAIK that is set only when the sensor resistance is invalid, i.e. cable Short-Circuit, Open-Circuit and a few resistance values well separated from those "expected". The Console definitely continues to report a direction when the windspeed is zero.

The Cumulus display "Compass Rose" diagram is "weighted" by the wind speed (i.e. priority is given to directions with a higher wind speed and I believe also is the "Average" Wind Direction. So I suspect the issue is related to Cumulus multiplying (weighting) the wind direction by "zero". It's not a "new" (MX) feature, Cumulus v1 behaved in the same way.

Whilst it may be reasonable to report "dead calm" with no direction, the lowest speed reported by the FO transmitter is 0.75 mph (and I believe the Davis is 1 mph and some higher still) at which speed the FO vane is definitely "responsive" and possibly correct. Also, I take the pragmatic view that the Anemometer might be "sticky" (or even temporarily faulty) and I would prefer not to "throw away" any other data which might be of value. Of course the physical design of some sensors, such as the Rotorvane (Derek Weston) and Ultimeter (Peet Bros), cannot resolve any wind direction if the Anemometer is not spinning

As an aside, this "feature" of Cumulus actually scuppered a "Bright Idea" (literally) project of mine a few years ago. It is well-known that the FO (magnetic reeds) Vane generally resolves only the 8 primary directions (N, NE, etc..) and the 8 intermediate directions (NNE, ENE, etc..) are rarely reported (because two reeds must operate at the same time). So my plan was for a small add-on microcontroller system in series with the "Wind" cable, to add 3 features to any of the Original FO stations (1080/1, etc....). The main idea was overnight to transmit only the 8 primary wind directions, but during the day, if it was determined that "The Sun is Shining" then the adjacent secondary wind directions would be transmitted instead. Then a "minor" addition to Cumulus (probably easier said than done) would have separated out the Sunshine hours and average wind direction. The other two features would have been to charge the batteries (from the same PV panel) and damping out (averaging) the acknowledged "lively" nature of the FO vane, such that a genuine average and not just a "snapshot" position would be logged.

Admittedly, the "post processing" of the data might have been an issue, but the real "killer" was that when no wind was detected then no "Vane" (and thus no Sunshine) data would be logged either. I did consider designing the microcontroller software to "invent" a minimum windspeed when there wasn't any, but that seemed a little too much of a bodge.

Cheers, Alan.

Re: Wind bearing zero when calm

Posted: Sun 14 Jul 2019 3:07 am
by liddelljs
Hi Alan,
Great detailed analysis - many thanks. After observing my station for the past couple of days, I think I would agree with you. I have seen two instances where the wind speed has been zero, the station is still reporting a bearing, but Cumulus MX reports a zero bearing.
If I understand your thinking correctly, the issue may be that Cumulus is not properly reporting the station bearing when the check box "Wind bearing zero when calm" is left unchecked. It is somehow still reporting a zero bearing. Is this correct?

Cheers
Steve

Re: Wind bearing zero when calm

Posted: Sun 14 Jul 2019 3:34 am
by beteljuice
From wiki - cumulus.ini ..
Note that the average wind bearing will always display zero bearing if there has been no wind for the last N minutes (where N is defined by previous parameter if present but defaults to 10), as the average bearing takes wind speed into account.

Re: Wind bearing zero when calm

Posted: Sun 14 Jul 2019 6:02 am
by liddelljs
aha - thanks Beteljuice. So the 10min gust is weighted by wind speed which goes to zero if these has been no wind for 10 mins. To get around it I display the reported instantaneous bearing rather than the 10min avge bearing (probably the downside is that there will be more variability).

Many thanks
Steve

Re: Wind bearing zero when calm

Posted: Fri 26 Jul 2019 6:45 pm
by jim_korman
I'll throw a little reasoning in the mix. Currently METAR and Synoptic data reports calm winds as 00000 so it does make sense.

Some older formats reported CALM, note that no direction was implied! Also, before most equipment was digitized, our wind recorders were analog devices, paper and ink, so that the direction of the last time the wind was blowing was recorded until it changed.