[Met_help] [rt.rap.ucar.edu #64811] History for Fwd: Forward:Questions on MET

[John Halley Gotway via RT]

----------------------------------------------------------------
  Initial Request
----------------------------------------------------------------

This transaction appears to have no content

----------------------------------------------------------------
  Complete Ticket History
----------------------------------------------------------------

Subject: Fwd: Forward:Questions on MET
From: John Halley Gotway
Time: Tue Jan 07 13:57:41 2014

Dear met_help,

Sorry for troubling you! I have some questions on MET and luckily I
see your presentations on MET website, so I write this email to ask
you for help! Here is my case:
 I conducted a WRF simulation and would like to verify the results
with mast observations, whose observation height was 10、30、50、70、80m
above ground level. First I applied UPP (WRF post-processor) to
convert WRF output from nc format into grib1 format and also to
interpolate the concerned variables(U、V、TEMP) onto the corresponding
mast observation heights. Then I wrote mast observations into the
required format that could be used by ascii2nc utility in MET to get
prepared for MET verification.
But when I verified my hourly WRF output with mast observation using
point_stat, I found that I could succeed only if variables in each of
the mast observation file was on one observation height , and if one
mast observation file contained several observation heights, MET could
not distinguish the different heights and all the values on different
heights were used and the average statistical value for all the
heights were returned. The second problem was that if the mast
observation file contained several observation locations on one
height, MET could only return the average value for all the locations,
the information for each location wasn't included. The last question
was on the verification of wind direction. I wasn't clear about the
detailed operation.
I attach my configuration file and several mast files in MET Point
ASCII format for the exploration of possible reasons. Just let me konw
if you need other data or material. It will be highly appreciated if
it is possible for you to share me some suggestion on my questions.
Thank you very much!
PS: m0339_10_speed.txt contains wind speed on 10 m above ground level
for mast 0339, m0339_10_30_speed.txt contains wind speed on 10 & 30 m
above ground level for mast  0339, m0339_10_30_m9242_10_30_speed.txt
contains for wind speed on 10 & 30 m above ground level for mast 0339
and mast 9242, PointStatConfig is the configuration file for
point_stat, the last file is the converted WRF output.

Kind Regard,
Yangyang Zhao

------------------------------------------------
Subject: Fwd: Forward:Questions on MET
From: John Halley Gotway
Time: Tue Jan 07 14:14:33 2014

Yangyang,

I spoke with Huiqin, and she forwarded your met_help question to me.
I apologize for the trouble with this support request.  I don't know
why it didn't show up in our system when you originally sent it.  It
may have to do with sending all the attachments.  In the future,
please send us email at "met_help at ucar.edu", but instead of attaching
large data files, you can post them to our anonymous ftp site
following these instructions:
   http://www.dtcenter.org/met/users/support/met_help.php#ftp

I believe the answer to your question is pretty simple.  I see that
you're using the ADPSFC message type which is intended for use only
with surface observations.  Encoding observations with ADPSFC tells
Point-Stat to ignore the vertical level and pressure information and
to use that observation in the verification regardless.  Instead, I
suspect that you only want to verify 30m forecasts with observations
that occur at 30m.  I'd suggest using the ADPUPA message type instead.
Just update your ascii observation files by switching ADPSFC to ADPUPA
and then modify the Point-Stat config file to use ADPUPA.  Then try
re-running Point-Stat to see if the behavior matches what you expect.

If you have more questions, please just respond to this email without
changing the subject line.

Thanks,
John Halley Gotway
met_help at ucar.edu

------------------------------------------------