meteohub.de

Hello,

Is it possible to give more detail regarding the two different wind selectors - wind0wind, wind0avgwind?

Specifically, the wind0avgwind, what time period would it be over for a Davis VP2? I have tried to match wind0avgwind-act with a number of wind0wind-avg[n] (where n = 1, 5, 10, etc), but haven't managed to work it out. That's assuming taking wind0wind-avg[n] does actually give the average wind speed over n minutes, and wind0avgwind-act is the correct way to use that selector.

It is important, because if say, I want to export 10 minute average wind data at 1 minute intervals using the data export service, it doesn't appear that you can do it by using something like wind0wind-avg10, only wind0wind-avg will work, and at the temporal resolution set in the service (ie if you set minute, you can then get minute temperature data but then also just get 1-minute average winds). However, if wind0avgwind-avg/act (?) was a 10-minute average, then it could be used as a substitute.

More detailed guidance on how to use these wind selectors would be most appreciated.

Thanks very much.

My understanding is, wind0avgwind is station dependent and would only use this such as wind0avgwind-act, no point averaging an average

Davis typically use the "10-minute average wind speed" for many derived variables but have no detail on what defines Davis 10-minute average wind speed cut-offs, per update, computed end of each minute or what? VP and VP2 measures wind speed in one mile per hour increments but maintains higher resolution for hourly averages, 10 minute average resolution ???

What is the issue with like wind0wind-avg10 not working? Is this only in a export service? I think you will find data export template is a special case? avg# by default is the export file resolution

Hi Matt,

Thanks for all of that. Good point regarding the details of the 10-min average, the way Davis compute it and the way meteobridge would compute wind0wind-avg10 could be subtly different, even if they are both nominally calculating a 10-min average.

Yeah, the issue is just with data export, if I wanted to set up a job that exported a daily csv mirroring my database structure (which is what I'm currently trying to do), then I store at 1-minute intervals in the db a 1-min avg temperature, and a 10-min average wind speed, but with a data export job it doesn't seem possible to output those simultaneously in one line, because the temporal resolution is specified once only, outside of the export template.

Edit: here's a comparison of the last 25 entries for wind0wind-avg10 vs wind0avgwind-act in my db - http://starlingsroost.ddns.net/weather/ ... arison.png

Cheers.

You are right, any min/max/average export is bound to the selected overall time resolution.
When you have a Davis connected and it does the wanted 10 min average using the "wind0avgwind" sensor should do the job, even when 1 minute interval is selected for export.

Lovely stuff, thanks admin.

On a related note then, is there a specific selector for 10-min average wind direction too, something like wind0avgwinddir? Or does one have to use wind0dir-avg10 to retrieve it. What is the recommended practice please.

Thank you

[wind0dir-avg] by itself is not valid but will work in an export template relative to the resolution. [win0dir-avg10] by itself is valid but not valid export template

OK, thanks both, understood now.

Maybe quite a niche point then, but it might be an improvement to be able to specify time resolution for each field inside the export, rather than an overarching fixed resolution. For example, if I wanted the 10 minute avg wind direction, but every minute, I couldn't get to it via export since I would have to export at 10-minute resolution and then I would only get a line every 10 minutes. I guess the method at present would be to calculate a 10-min avg from the 1-min avgs in a separate script, when the export data is made use of further down the line

Is there a method for formally putting forward suggestions such as these, on here?

Thanks again for all your help.