High Wind Speed Discussion

[midcon07]

Been having this High Wind Speed discussion with WU. Not knowing much (nothing) about programing I'm having a hard time understanding the following: on my WU History page, in the Summary my High Wind Speed and High Wind Gust are usually the same figure, occasionally they vary by 1 or 2 mph. However the Wind Graph shows a lower High Wind Speed which I believe is correct and a Wind Gust which is several mph higher then the Wind Speed and which I also think is correct. Switch to the Table and it backs up the graph. Other sites such as findu.com-CWOP, Mesowest and PWS also back up what the WU graphs show. WU says that Cumulus is " is sending the same value for max wind speed and max wind speed gust." First of all is that indeed what is happening and if it is why do I get the variation between the Summary and the graphs and tables?

Jerry

[steve]

First of all, Cumulus does not send max wind speed or max wind gust at all to WU, as those are not part of the protocol, see http://wiki.wunderground.com/index.php/ ... d_Protocol

WU determines the max values itself from the individual values submitted.

Here are the parameters concerning wind/gust speed in the protocol, and what Cumulus sends for each, while running 'live':

windspeedmph - [mph instantaneous wind speed]
By default, it sends the last 3-second wind speed from the console. There is an option "Send average wind" in the WU settings in Cumulus, setting this causes the average speed to be sent instead (this is not normally useful for Davis stations).
windgustmph - [mph current wind gust, using software specific time period]
Cumulus sends the peak gust over the last 10 minutes.
windspdmph_avg2m - [mph 2 minute average wind speed mph]
Cumulus send the average wind speed (normally 10 minutes)
windgustmph_10m - [mph past 10 minutes wind gust mph ]
Cumulus does not actually send this, I believe that it was not part of the protocol when I wrote the code.

While Cumulus is sending archive data to WU as part of the 'catch up' when it first starts, the only wind speeds available to it are the peak speed for the archive interval, and the average speed for the interval. In that case, it uses the peak speed for the interval where running live it would use the latest 3-second wind speed (so that the peak value is not lost) - so that value will get sent to WU as 'windspeedmph'. The peak value of those over the last 10 minutes gets sent as 'windgustmph'.

So if you are using archive data significantly (i.e whether you shut down Cumulus for extended periods), what your station's logger interval is, whether your peak speeds occur during the times that Cumulus wasn't running, and exactly how WU do their calculations, your WU high wind speed could come out close to your WU high wind gust. Whether or not you are using 'rapid fire' will also affect the WU calculations compared to the data shown in the table, while Cumulus is running with live data (not every set of data sent appears in the graph or table, but presumably WU do use every set to determine max wind speed and gust).