Wind average question...

[GlennMartin]

The wind AVERAGE readings seem highly suspect.

For example, tonight is windy, consistenly ranging from 10 to 25 miles per hour. But the AVERAGE reading bounces around in the single digits.

I understand that it's not a day average - it's a short-term (10-minute) average. But the averages shown are often lower than even the lowest wind reading in the past 10 minutes.

Yesterday was very windy - almost all day from 20 to 45 mph. But the PERIOD record for yesterday shows it was only about 7 mph.

My unscientific guess is that the reading is often only 1/2 or 1/3 what the actual reading should be.

What am I missing? any ideas?

[steve]

What have you got for these settings:

windsettings.png

Which figures are you comparing - have you seen the FAQ entry about Fine Offset wind speeds? Please confirm that you are talking about 'live' data, rather than data read from the logger. Where are you getting your quoted "10 to 25 mph" from? Do you have any "calibration" settings? Have you used any cumulus.ini settings to change the averaging interval? Please show a graph of the wind speeds.

[GlennMartin]

I did not modify any *.ini files.
I have BOTH the "Calculate 10-min Wind Avg" and "Use 'Speed' for avg calculations" checked.
I have no calibrations set.
And if it matters, I am using 1.9.1 build 964 on Windows 7 64-bit.

LIVE DATA:
The graph has the "red zone" showing the range of the past 10 minutes, with the current speed shown by the pointer within that range.
>>> But what is listed as AVERAGE is the bottom of the red zone, which is the MINIMUM of the past 10 minutes.

The attached file has an example. The average shown is 5.5, but I think it should be 9.5 (half way through the last 10-minute range).

LOGGED DATA:
I can't be scientific about this unless I watch the reading with calculator in hand for 24 hours, but I think it's often about 1/2 of what it should be. But we can leave this go till I figure out the live data first.

What am I missing??

[nking]

I think you need to untick the "use speed for avg" (2nd item from Steve's attachment). Many of us find that averaging the gust value is more accurate.

[steve]

"Use 'Speed' for avg calculations" checked.

So it's your station which is giving the speeds which you say are low (Cumulus is just extending the averaging period). You can prove (or disprove) this by watching the 'speed' values on your console and/or turning off the Cumulus calculation. You may genuinely have very 'gusty' winds (i.e. a fairly low average speed, but with large gusts); if you know that your average wind speed is actually higher than your weather station is reporting, then you could try doing as Neil suggests, and get Cumulus to use the gust figures for the average calculation.

[mcrossley]

I thought the FO 'speed' was too low compared to the 'gust', but when I watched the anemometer cups for a while, I was surprised how slowly they turned much of the time even stopping altogether even when it 'felt' windy. I think this may be fairly typical in suburban/urban environments when the anemometer is often mounted much lower than the recommended 10m, and the buildings/trees etc induce much low level turbulence and drag. Out in open country the wind speed is much less variable - and stronger!

[Ned]

I tried unticking the 10 minute average box for an hour just to see the result on the graph. Went back to averaging

[GlennMartin]

Thank you, folks. You have been very helpful!