[wxqc] quick question

[Gerry Creager N5JXS]

Mark Wyman wrote:
> Just a guess that 1 second updates are probably great for plume
> calculations. My UP24 pressure meter shows fluctuations in the 1Hz and
> longer time frame, but not much faster than that unless a door closes or a
> vehicle goes by. These events are very low frequency sound waves anyway and
> over time neutralize. The <1Hz (>1 sec) pulses tend to be wind related and
> don't neutralize, and trends can be developed from them for short and long
> term changes.
> 
> http://www.ramseyelectronics.com/cgi-bin/commerce.exe?preadd=action&key=UP24

I decided to do 15 sec sampling, almost arbitrarily.  And, having seen 
how things are sensed and used in the "real world" I suspect this is the 
finest-grained temporal resolutions they've ever legitimately gotten.

Intersting thought that anything higher-frequency than 1 hz is sound. 
I'm not sure that holds when one is looking at eddy currents in rough 
local terrain.

> This indicates that anything shorter than 1 second unless in severe winds is
> probably meaningless for what you are doing as it is sound waves.
> 
> I wonder if you could spread the range of the ultrasonics and place at a
> 45-degree angle to ground to try and get an average wind speed vs. elevation
> rather than wind speed at a given height. Perhaps you can hit more than 3
> feet at once. Probably just have to adjust the reading based on the changes
> you make in distance.

Using a different frequency set (there's nothing really magic about 90 
khz) one could go with longer baselines and attempt to sense the ground 
and profile the wind, but that's effectively what SODAR does, and to a 
much higher level.  We're considering borrowing a SODAR for future tests.
gerry
> -----Original Message-----
> From: wxqc-bounces at lists.gladstonefamily.net
> [mailto:wxqc-bounces at lists.gladstonefamily.net] On Behalf Of Gerry Creager
> N5JXS
> Sent: Thursday, April 20, 2006 8:52 AM
> To: Discussion of weather data quality issues
> Subject: Re: [wxqc] quick question
> 
> I'm using a pair of the Vaisala WXT-510's, which incorporate an 
> ultrasonic sensor.  I prefer the lack of inertia in my wind measurements:-)
> 
> The models I work with (WRF, MM5) are perfectly happy with hourly or 
> even 3-hr updates for data assimilation.  You're correct: The weather 
> models don't require the really frequent updates.  I can't (yet) speak 
> to the plume models done as CFD on a finite element grid.  Worse, the 
> folks I am helping here can't answer that question pool yet, either.
> 
> Thanks for the comments.
> gerry
> 
> Mark Wyman wrote:
> 
>>I'm not coming from the weather side of things, but the engineering side.
>>The wind measurement rate is going to be dependant on the mechanical
> 
> nature
> 
>>of the sensors. It takes some time for them to accelerate and decelerate.
>>Large wind cups will take a lot longer than say a hot-wire anemometer to
>>respond to a change in wind speed simply due to their mass vs. wind
>>friction. It will also depend on your needs. 
>>
>>Weather models may not require sub-30 second measurements but you may
>>require 1 second to determine the volume of air to pass through a certain
>>area in a given time.
>>
>>Probably the fastest response time air measurements will be the ultrasonic
>>variety, which would also be the safest in a combustible environment. They
>>are also mighty expensive.
>>
>>-Mark
>>
>>-----Original Message-----
>>From: wxqc-bounces at lists.gladstonefamily.net
>>[mailto:wxqc-bounces at lists.gladstonefamily.net] On Behalf Of Gerry Creager
>>N5JXS
>>Sent: Wednesday, April 19, 2006 8:29 AM
>>To: Discussion of weather data quality issues
>>Subject: [wxqc] quick question
>>
>>OK, so this is a little off topic, but I'm trying to get a lot of things 
>>done quickly and this group tends to have some good expertise...
>>
>>I'm instrumenting a site for some liquified natural gas dispersion 
>>experiments, and we are gonna do it tomorrow.  I'm putting up a pair of 
>>Vaisala WXT-510s, one at 2 meters, one at 10 meters, in as close to an 
>>unobstructed area as I can get near the test site ("prop") as possible 
>>while staying out of the anticipated plume (if the plume ignites, and 
>>the WXT-510s are in it, they'll become molten pieces of thermoplastic 
>>slag...).
>>
>>OK, the question: Does anyone have a good idea of the Nyquist frequency 
>>for wind measurements?  I've got a request for 1-sec measurements to 
>>capture all the dynamics.  I can do this but suspect the noise will 
>>overcome the data and I'll end up having to decimate the data to get a 
>>reasonable value.  I guess I can look at an FFT in the frequency domain 
>>and see if anything drops out on the floor, but that's introducing a lot 
>>of work.
>>
>>I seem to recall from the depths of my brain, that wind data should be 
>>collected at a rate not greater than once per 30 sec (twice/min), but 
>>now I can't find documentation on that.
>>
>>Adding insult to injury, the folks I'm working with on this don't know 
>>what sort of continuous data will really aid their analysis.  They do 
>>know they need to initialize the FEM3A codes they're running with 
>>surface met data (temp, pressure, humidity, anemometry) but there's more 
>>than a little confusion as to whether it takes continuous data while 
>>it's running or if it's an isolated measurement.
>>
>>I'll be looking at data for weeks after this, so there's no real time 
>>constraint on answers, if someone's got something!
>>
>>Thanks, Gerry
>>Texas Mesonet
>>AP009
> 
> 

-- 
Gerry Creager -- gerry.creager at tamu.edu
Texas Mesonet -- AATLT, Texas A&M University	
Cell: 979.229.5301 Office: 979.458.4020 FAX: 979.862.3983
Office: 1700 Research Parkway Ste 160, TAMU, College Station, TX 77843