Untangling Wind

[Oz Greg]

Tom,

Most of the Wind (Reporting) Templates Report Wind Measurements as the following..

A) Wind Gust (Max Speed & Direction Generally for a Day )

B) Wind Speed (Max Speed & Direction, Min Speed & Direction & Avg).

At a Archiving Level weewx Records this perfectly..

windSpeed
windDir
windGust
windGustDir

However at a Stats level the wind data (Gust & Speed) is mashed together resulting in loss of the ability to report the individual elements of the wind measurements (other than $current tags) (ie $day = You cannot derive a Max Wind Speed and a Max Wind Gust measurement)

I know from our conversations that you would prefer to remove stats (not add to them) but in this case would you consider splitting out the windGust stats into a windGust stats group otherwise we will need to build a reporting subclass to replicate all wind stat measurements which kind of goes against what we would like to achieve but we are heading down this path in order to be able to offer the level of granularity most of the other weather reporting templates require.. 

Hoping you could consider this..

weewx-wd Team

[Thomas Keffer]

One thing I don't understand here: what's the difference between, "Max Wind Speed'', "Wind Gust", and "Max Wind Gust." 

To me, they are all the same. 

Or, am I missing something?

-tk

--
You received this message because you are subscribed to the Google Groups "Weewx user's group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to weewx-user+...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

--

Tom Keffer

kef...@threefools.org
+1 541-386-8891 (h)
+1 541-490-9507 (c)

Skype: tkeffer

[Oz Greg]

Tom,

I don't believe it is a get it or not discussion, the underlying recording of wind I believe well represented in weewx so I believe you get it.. :)

From my understanding of what is a wind measurement..

Wind Speed = Constant (Avg) Recorded Speed of the wind during a measurement interval.

Wind Gust = Maximum Recorded Speed of a wind during a measurement interval.

Thus for a day I see the following requirements to be reported..

Wind = Max Speed and Direction

Wind Gust = Max Speed and Direction

If you look at weather underground: you will see this clearly illustrated Wind Speed: Current, Avg and High and Wind Gust: Current and High

http://www.wunderground.com/weatherstation/WXDailyHistory.asp?ID=IVICTORI324

Either Weewx (wind) stats $day just don't support this requirement as I only have one Max to work with unless you got a neat trick in the way you have flattened the stats that will enable this kind of reporting..  

Thanks :)

[Thomas Keffer]

OK, I think I 'get it'. :-)

You divide the day up into archive intervals (say, 5 minutes). Then you average the winds in those intervals. The max of those averages over the course of a day is the "Max Wind Speed." Right?

I'm not sure what the value is in such a statistic, as it would be highly sensitive to the averaging interval. For example, as the interval goes to zero, then there would be no difference between max wind speed (as defined above) and the gust speed. As the interval goes the other way, towards one day, the max wind speed becomes essentially the average wind speed. 

Another way to think of it: suppose I told you that yesterday had an average wind speed of 5 m/s, a max wind speed of 10 m/s, and a max gust of 15 m/s. What would you have learned? Now suppose that today the average wind speed and max gust were the same, but max wind speed was 8 m/s. What have you learned? That yesterday was windier? Probably not.

If there was a physical basis for the averaging interval, such as a dominant time in the wind frequency spectrum (something like a "set" in ocean waves) then I could see it --- you'd want to know how much energy or speed there was in each interval. 

In any case, the customer is always right. :-) The problem is that, right now, recording something like that in the stats database would require changing the schema --- something that I'm not very enthusiastic about because it would inevitably be a huge support headache. 

So, bottom line: this becomes another good reason to do away with the stats database and extract everything from the archive database. Or, at least, just use the stats database as an optimization strategy. 

As always, your thoughts are welcome.

-tk

[Andrew Milner]

Yes, Wunderground has indeed got current speed and gust and the weekly, monthly and yearly have got high, avg and gust so one must indeed assume that avg == (total averages / # periods); high == highest avg and gust == highest within a period.  In the case of wunderground the period is therefore fixed to day or week or month or year - they do not do the same figures where the reporting interval is variable and dependant on the reporting stations.

Your assertion is therefore correct Tom in the need for knowing the period - and for wunderground the minimum period is 1 day.

[Oz Greg]

Tom,

I understand your point of view on yet more stats thus I am starting to think I am going to use the archivedb.getSqlVectors and untangle the readings this way..  

This way I can get the MAX on the windSpeed and windGust and let the person running the report figure out what value they are getting from these stats...  (To me the only stat that makes sense is WindSpeed avg  to answer the Q was it winder today than yesterday..

Guessing maybe a broader discussion on stats might also be in order ;)  (Pity this one cannot be in your hot tub with a cold drink or three) ;)

Greg