[Met_help] getting MET stats for wind speed

[John Halley Gotway]

Brian,

That's correct.  The prepbufr files only contain P, Q, T, Z, U, V observation type variables.  I need to update pb2nc to derive wind speed and direction from U and V.  In addition, I'll be adding
derivations for a variety of humidity types.

So even if you were to compute wind speed and direction from the model output, MET would not be able to verify it because those observation types haven't been derived yet.

As for plotting scripts, unfortunately we don't have any available to distribute yet.  I've written a couple of rough scripts in R for working with MODE output.  And I know that Lacey has written a
couple of scripts in IDL for working with the output of grid_stat and point_stat.  One of the tasks on our list after the next release is to work on those scripts, clean them up, and post them to the
MET website.  Right now they're very specific to the datasets with which we're working.

Actually, if you'd like to contribute your gnuplot code, we'd be happy to post it to MET website as well.

Sorry I can't be of more help right now.

John

Brian Gaudet wrote:
> 
> Hi John,
> 
> 
> Thanks for the info.  It sounds like from your response that even if I were
> to modify the WRF postprocessor to dump out wind speed and direction,
> the pb2nc package is not currently set up to extract those particular 
> variables from the PrepBufr files.  Is that correct?
> 
> Also, I have one more question.  We have been interested in obtaining
> some graphical capability for displaying the output of MET (currently
> we are simply using gnuplot to make some basic plots).  I know that
> this capability is in the planning stages for the next release, but Ligia
> Bernardet has told me that you may have some current graphical
> capability.   What kind of capability would this be, and how could
> we obtain it?
> 
> 
> Thanks again,
> 
> 
> Brian
> 
> 
> 
> 
> 
> 
> 
> On Wed, Dec  5, 2007 02:43 PM, John Halley Gotway <johnhg at rap.ucar.edu> wrote:
> Brain,
>> Good question... and a rather timely one.
>>
>> The answer is no.  There is currently no way of verifying wind speed and
>> direction with MET.
>>
>> However, I'm working to include support for verification of some derived
>> quantities in the next release of MET at the end of the month.
>>
>> The WRF postprocessor doesn't currently dump out wind speed and direction -
>> just U and V.  We're hoping to add capability for grid_stat and point_stat to
>> derive wind speed and direction from U and V.
>> Also, we'll add that derivation in pb2nc to derive wind speed and direction
>> for the observations.  From a user perspective, you should just need to modify
>> your config files to indicate that you'd
>> like to verify wind speed and direction.
>>
>> Let me mention an issue regarding winds that's come up in the last couple of
>> days.  Apparently, the WRF postprocessor is dumping out winds that are defined
>> relative to the grid.  The point
>> observations in the prepbufr files contain winds that are earth relative. 
>> Currently, MET is not accounting for this difference.  Within the DTC we're
>> discussing how to address this - likely through
>> changes to the postprocessor or changes to MET.
>>
>> Does that help?
>>
>> John
>>
>> Brian Gaudet wrote:
>>>
>>> Hello,
>>>
>>>
>>> I have been using the Beta version of METv0.9 to verify WRF-ARW model runs
>>> (after having been processed by the WRF postprocessor and put into
>>> GRIB format) against PrepBufr obs.  The tool I've been using
>>> is point_stat.
>>>
>>> I was wondering if there was a way to verify wind speed and direction 
>>> using MET if U and V are provided by the model.
>>>
>>> I have put these variables in succession in the .Config file to get
>>> vector statistics, as recommended by the online Users Guide.
>>> I do indeed get VL1L2  files, as described in the Users Guide.
>>>
>>> However, I noticed that the parameter provided in the VL1L2
>>> file is 'Mean(uf ** 2 + vf**2)'.  What I want is 'Mean(
>> (uf ** 2 + vf ** 2) **
>>> 0.5)',
>>> and I don't believe these parameters are derivable from each other
>>> (or any other parameters provided).  Is there another way to
>>> get this parameter?
>>>
>>>
>>> The only way I know of that should work is to output WIND and
>>> WDIR from the WRF postprocessor; then these fields can
>>> be verified directly using MET.  However, to do this would
>>> require some code changes to the WRF postprocessor.
>>>
>>>
>>> Thanks for any help you can provide.
>>>
>>>
>>>
>>> Brian Gaudet
>>> bjg20 at psu.edu
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> ------------------------------------------------------------------------
>>>
>>> _______________________________________________
>>> Met_help mailing list
>>> Met_help at mailman.ucar.edu
>>> http://mailman.ucar.edu/mailman/listinfo/met_help
>>
>>
> 
>