Wind Speed Correction

35 views
Skip to first unread message

Yotta

unread,
Dec 28, 2014, 5:06:04 AM12/28/14
to wfrog...@googlegroups.com
I have a WMR200A weather station

I've never really thought about it, but when I went from using Weather OS to Wfrog to record data from my weather station there was a marked drop in reported wind speed. Weather OS was reporting around 20km/hr average, with gusts up to 120km/hr, Wfrog rarely has an average over 5km/hr and gusts are rarely over 30km/hr.

Comparing to the Beaufort scale suggests something is not quite right and doing a quick number crunch it seems coincidental that the conversion of m/s to km/hr is a factor of 3.6 which roughly corresponds to the drop in wind speed. I'm suspicious that either a conversion is taking place that shouldn't be (or conversely not taking place that should)

Is there any easy way to check what conversions are being done, and to either disable or add a conversion to the driver/database?

Jordi Puigsegur Figueras

unread,
Dec 29, 2014, 2:54:02 AM12/29/14
to wfrog...@googlegroups.com
Hi,

120 Km/h. is really a lot of wind. I've never seen such a high reading with an Oregon Scientific (WMRS200) and generally they tend to register less widn than more professional weather stations. I maintain a station in a very windy place (http://www.puigsegur.cat/meteolajonquera/24hours.html) where we now have a Davis VantagePro and very rarely we get more than 100 Km./hr. At these wind speeds there are problems with trucks, big vans, etc. 

Isn't there a console or another close by station to compare values?

Jordi.

--
You received this message because you are subscribed to the Google Groups "wfrog-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to wfrog-users...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Yotta

unread,
Dec 29, 2014, 7:30:21 PM12/29/14
to wfrog...@googlegroups.com
120km/hr was a reasonably exceptional case, I believe that particular reading was within a whirlwind (a very small tornado say 3m/10ft across) which are reasonably common in this part of the world through summer.

Cross referencing with the local Bureau of Meteorology readings;
BOM: 20km/hr gusts, 13km/hr average
Wfrog: 4km/hr gusts, 3.6km/hr average

I'm currently getting nothing above 35km/hr even in severe storms ( http://extreme-weather.org )

Jordi Puigsegur Figueras

unread,
Dec 31, 2014, 1:02:39 PM12/31/14
to wfrog...@googlegroups.com
Ok. Just checking to be sure because OS stations use to have a reputation of not reporting properly the wind speed, always giving values below real, specially for high speeds.

I assume you are using the wmr200 driver. If so, this driver receives wind speed (both avg and gust) in m/s, wfrog expects m/s, so no conversion is done. You can check the source code:

All other OS drivers in wfrog receive m/s. However note that OS doesn't publish the protocols, therefore they are all reverse engineered. 
 
Please, check /var/log/wflogger.log where there should be a line for each record received and you can check which data is being saved.

Jordi.

Yotta

unread,
Jan 1, 2015, 3:34:54 AM1/1/15
to wfrog...@googlegroups.com
Ok, yeah checked out the log and it looks like everything is on order from the wfrog side;
- 1m/s logged, 3.6km/hr on Wfrog while Bureau of Meteorology is reporting 13km/hr

So guess that the weather station is reporting low, would changing the formula on lines 559/565 be an effective way to calibrate the readings? (I don't know python, thinking I could calculate what factor I need than modify the "* 0.1" on each of those lines to suit)

Yotta

unread,
Jan 2, 2015, 6:25:03 AM1/2/15
to wfrog...@googlegroups.com
And storm went through today... 37km/hr on wfrog vs 63km/hr of BOM. That's a much smaller difference than previous observations. That would support that the station is reporting low coupled with a difference in how wfrog and Weather OS calculate/report gust speeds perhaps
Reply all
Reply to author
Forward
0 new messages