I am reading the weather data using the UDP API and noticing that the rapid wind message frequently gives several consecutive “zeroed out” readings. The app also shows the zero reading then goes back to the normal wind reading. I read somewhere that this is normal. Is that true?
Why the zero reading? I am averaging the wind over time and am concerned that these zero reading are skewing the results. Also I am not sure I can exclude the zero readings because the wind might actually be zero speed and zero angle at times.
during my testing of a Sky in my impromptu wind tunnel over several days of constant wind I would encounter several zero readings. When I discussed it with David he called them glitches.
It appears to be an accepted issue.
I now allow for that in my analysis and display of my important wind data by not displaying the minimum and by using two adjacent average periods and using the highest average. I realise this is not true. But in my opinion it is tidier and my graphs appear more correct.
And most personal weather stations are too low which causes more gusty winds. I find the gusts are more similar to actual wind speed than the stations average.
I might include a couple of graphs that display the issue to this message when I am on my computer. Look for the edit later.
Cheers Ian
This one comes up a lot in various forums. Basically they have to pick whether in zero wind they show no direction, or whether they show the last known direction. Folks tend to be 50/50 on the answer they pick.
Personally I’d code in something where I’d ignore direction if speed = 0 (or null, depending on what the WF gear returns)
I am referring to incorrect zero wind speed readings from the Sky. It has come up a few times on this forum that some times a zero is measured when it is not zero wind speed. I wanted to prove if it was an incorrect reading so I set up a fan and ran it for a few days. There were occasional 3 sec values recording zero speed when it was not zero speed. David explained the issue as a glitch. So I assume from my experiment that it may always occur perhaps a few times per day.
And there is another problem which may have been fixed in a recent firmware update that records an entire minute as zero speed even though when I watched the 3 second values they were reading strong winds.
For this reason to display my average wind speed I am using the higher of some average wind speeds to prevent the random false zeros reducing the average. It does mean my average is lifted.
I am still experimenting with what size and number of periods to average together and take the maximum, but you get the idea.
Here are 2 graphs of a windy day with a southerly change showing how I was displaying it. And the new version which has lifted the false average green arrow at 11:40. (And there were 4 periods during that day with false zero’s.)
The average including the false zeros:
Yes I am talking about the readings dropping to zero when there is actually wind present. I may need to figure out an algorithm to exclude those readings.
