19990305: problems using oabsnd

>From: Robert Mullenax <address@hidden>
>Organization: National Scientific Balloon Facility
>Keywords: 199903052045.NAA17847

>I am trying to grid upper air files for use in producing u/a analyses in
>Gempak.
>I create a grid using oagrid with GRDAREA  = 45;150;80;-100 and DTAAREA=
>
>The proram runs and creates a grid.  When I use oabsnd to do an analysis
>on the
>grid I get the following error:
>
>
>[GEMPLT -16]  NIPBND - Invalid projection bounds.
> [GR -7]  Error defining grid navigation in GEMPLT.
> [OABSND -9]  No valid parameters have been entered
>
>Everything works fine if I don't define grdarea using the lat, lon, for
>example if
>I use grdarea=fai--, all is well.  What have I done wrong?  Thanks, and
>sorry
>for the barrage of questions.
>
>Robert Mullenax
>
>
>

Robert,

Assumimg you have:
 DELTAN   = 
 DELTAX   = 
 DELTAY   =
for the above GAREA, oabsnd reports too few stations to calculate
DELTAN at least for the IDD data I have.

I can create the file with gdcfil, and I created a file for the above
region as:
 GDOUTF   = test.gem
 PROJ     = ced
 GRDAREA  = 45;150;80;-100
 KXKY     = 20;20
 MAXGRD   = 1000
 CPYFIL   = 
 ANLYSS   = 
 GEMPAK-GDCFIL>r

 GDCFIL PARAMETERS: 

 New grid file:         test.gem                                                
                
 Grid projection:       CED                     
 Lower left corner:         45.00   150.00
 Upper right corner:        80.00  -100.00
 KX, KY:                       20       20
 X/Y grid spacing:           1.84     5.79
 DELTAN =     8.592       EXTEND =   2;  2;  2;  2
 Maximum number of grids that can be added to file 1000

Since oagrid would use a CED projection, that is what I specified
in gdcfil.

Running Oabsnd for 500mb hght, gives:
 SNFILE   = $UPA/990304_upa.gem
 GDFILE   = test.gem
 SNPARM   = hght
 STNDEX   = 
 LEVELS   = 500
 VCOORD   = pres
 DATTIM   = 0000
 DTAAREA  = 
 GUESS    = 
 GAMMA    = .3
 SEARCH   = 50/EX
 NPASS    = 2
 GEMPAK-OABSND>r

 OABSND PARAMETERS: 

 Output grid file:        test.gem
 Number of x, y points:    20    20
 Grid bnds:     45.00   150.00    80.00  -100.00 Grid xtnd:    2   2   2   2
 Station spacing:           8.59
 No first guess file is requested. 
 Sounding file:           $UPA/990304_upa.gem
 Data area:               40.00;-180.00;85.00;180.00
 Level parameters:        HGHT
 Levels:                   500
 Vertical coordinate:     pres                                                  
                  
 Station parameters:    
 Time:                   990304/0000         
 GAMMA: 0.30   SEARCH:  50.00  NPASS:  2
 Data extrapolation may occur. 
Enter <cr> to accept parameters or type EXIT:

 Processing data at time 990304/0000         

   RMS values for pass  1
 PARAMETER:     HGHT
 LEVEL:          500
 RMS:          48.29
 #STN:            31

   RMS values for pass  2
 PARAMETER:     HGHT
 LEVEL:          500
 RMS:          26.87
 #STN:            31


So, there is nothing intrinsically wrong with the grdarea you
want to use.  

I think the problem exists with the EXTEND area. Depending on the KXKY
that OAGRID ends up suggesting, it could be wrapping the north pole-
which is very bad in a CED projection. Have you tried setting
EXTEND = 0;0;0;0? 

can you send me the output from oagrid so I can see what the suggested 
station spaceing for the grid is, and if the extended area will
wrap over the pole?

Steve Chiswell