[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[GEMPAK #UER-356689]: NAGRIB not decoding AMPS WRF grids



Robert,

When I use your settings for uv, I decode the winds as expected.

                                        

      4370 GRIB messages were read or scanned from the GRIB file:
     2007120300_D5.grb                                                       

       868 grids were written to the GEMPAK file:
     antarctica_uv.gem         


When I use your grdarea for the antarctica.gem file,
GRDAREA = -78.34;156.78;-75.77;174.52
I get the 0 grids decoded as you saw. I notice that the longitudes for LL and 
UR has changed
slightly by .01 each when running dcgrib2 on the data.

When I change this in nagrib to:
 PROJ     = str/-90;180;0
 GRDAREA  = -78.34;156.77;-75.77;174.51
 KXKY     = 159;171

I get:
     4370 GRIB messages were read or scanned from the GRIB file:
     2007120300_D5.grb                                                       

      1890 grids were written to the GEMPAK file:
     antarctica.gem

I haven't looked in to the source of the .01 difference, but since the LL point
is from the data itself, I assume that the post-processing involved in 
converting from the model
domain to grib is just that much different in roundoff error than previous.

So, to fix your decoding, you will have to modify the GRDAREA value.

The other unknown parameters of numbers greater than 127 will be referenced
in ncargrib2.tbl, and if you happen to know those parameters, you can create a
table so that they are decoded as well.

Steve Chiswell
Unidata User Support






> 
> Steve,
> 
> Last year you helped me write NAGRIB scripts that would handle the AMPS MM5 
> and WRF grids.  The GRIB files provided by AMPS have the U/V winds offset by 
> one grid point.  Decoding with dcgrib2 will allow  you to display everything 
> but winds fine.  Winds either display in the wrong place or won't display at 
> all (and causes gdplot2 to dump core).  The solution was to decode the grib 
> messages with nagrib2 and set the proper geog for the U/V, then decode the 
> rest then merge back together:
> 
> #!/bin/csh
> set DATE=`date +%Y%m%d`
> #set DATE=20061106
> set HR=$1
> source ~gempak/GEMPAK/Gemenviron
> nagrib << EOF
> GBFILE   = /data/ldm/EXP/mm5/${DATE}${HR}_D5.grb
> INDXFL   =
> GDOUTF   = /usr2/config/grdtmp/mm5d5/antarctica_uv.gem
> PROJ     = str/-90;180;0
> GRDAREA  = -78.34300;156.7770;-75.74857;174.6063
> KXKY     = 160;172
> MAXGRD   = 10000
> CPYFIL   =
> GAREA    = dset
> OUTPUT   = t
> GBTBLS   =
> GBDIAG   =
> PDSEXT   = NO
> OVERWR   = NO
> r
> 
> e
> EOF
> gpend
> nagrib << EOF2
> GBFILE   = /data/ldm/EXP/mm5/${DATE}${HR}_D5.grb
> INDXFL   =
> GDOUTF   = /usr2/config/grdtmp/mm5d5/antarctica.gem
> PROJ     = str/-90;180;0
> GRDAREA  = -78.34;156.78;-75.77;174.52
> KXKY     = 159;171
> MAXGRD   = 10000
> CPYFIL   =
> GAREA    = dset
> OUTPUT   = t
> GBTBLS   =
> GBDIAG   =
> PDSEXT   = NO
> OVERWR   = NO
> r
> 
> e
> EOF2
> gpend
> cd /usr2/config/grdtmp/mm5d5/
> /usr2/config/scripts/gempak/antarctica_merge.csh
> cleanup
> mv antarctica.gem /data/ldm/gempak/model/mm5_ice/${DATE}${HR}_d5.gem
> rm antarctica_uv.gem
> 
> 
> where the merge script is:
> /usr2/config/scripts/gempak% more ant*merge.csh
> #!/bin/csh -f
> 
> 
> @ SHOUR=6
> @ EHOUR=36
> 
> set LEVELS=("1000" "925" "850" "700" "600" "500" "400" "300" "250" "200" 
> "150" "100" "10" "300")
> while ( $SHOUR <= $EHOUR )
> 
> if ( $SHOUR < 10 ) then
> set FHOUR=F00${SHOUR}
> else
> set FHOUR=F0${SHOUR}
> endif
> 
> @ LCNT=1
> 
> foreach LEVEL ($LEVELS)
> if ( $LCNT < 13 ) then
> set GVCORD="PRES"
> else
> set GVCORD="HGHT"
> endif
> 
> echo LEVEL $LEVEL $GVCORD
> 
> 
> gddiag<< EOF_GDDIAG
> GDFILE   = antarctica_uv.gem
> GDOUTF   = antarctica.gem
> GFUNC    = UREL
> GDATTIM  = $FHOUR
> GLEVEL   = $LEVEL
> GVCORD   = $GVCORD
> GRDNAM   =
> GRDTYP   = S
> GPACK    = none
> GRDHDR   =
> PROJ     =
> GRDAREA  =
> KXKY     =
> MAXGRD   =
> CPYFIL   =
> ANLYSS   =
> run
> 
> GFUNC    = VREL
> run
> 
> exit
> EOF_GDDIAG
> @ LCNT = $LCNT + 1
> end
> 
> @ SHOUR = $SHOUR + 1
> end
> gpend
> 
> exit 0
> 
> 
> This is and has been working fine for the MM5.  However, the WRF decoding 
> does not work after I turned it back on this week.  All the AMPS guys are on 
> the ice right now so it's difficult to ask if they have changed anything.  I 
> do not believe they have.  It seems like it MAY be a problem with nagrib.  
> The reason I say that is that I can decode the file properly with dcgrib2, 
> and can display anything other than winds.  Winds as always won't display and 
> gdplot2 core dumps.  When the nagrib script runs it seemingly finds nothing 
> to decode.  See below:
> 
> 
> [NAGRIB 3]  No valid parameter found for 170.
> [NAGRIB 3]  No valid parameter found for 171.
> [NAGRIB 3]  No valid parameter found for 153.
> [NAGRIB 3]  No valid parameter found for 178.
> [NAGRIB 3]  No valid parameter found for 170.
> [NAGRIB 3]  No valid parameter found for 171.
> [NAGRIB 3]  No valid parameter found for 153.
> [NAGRIB 3]  No valid parameter found for 178.
> [NAGRIB 3]  No valid parameter found for 170.
> [NAGRIB 3]  No valid parameter found for 171.
> [NAGRIB 3]  No valid parameter found for 153.
> [NAGRIB 3]  No valid parameter found for 178.
> [NAGRIB 3]  No valid parameter found for 170.
> [NAGRIB 3]  No valid parameter found for 171.
> [NAGRIB 3]  No valid parameter found for 153.
> [NAGRIB 3]  No valid parameter found for 178.
> [NAGRIB 3]  No valid parameter found for 170.
> etc..etc..etc..
> 
> 
> 4368 GRIB messages were read or scanned from the GRIB file:
> 2007120112_D5.grb
> 
> 0 grids were written to the GEMPAK file:
> tr.gem
> 
> 
> GRID FILE: tr.gem
> 
> GRID NAVIGATION:
> PROJECTION:          STR
> ANGLES:               -90.0   180.0     0.0
> GRID SIZE:          160 172
> LL CORNER:             -78.35    156.77
> UR CORNER:             -75.75    174.61
> 
> GRID ANALYSIS BLOCK:
> ANALYSIS TYPE:        BARNES
> DELTAN:               4.000
> DELTAX:           -9999.000
> DELTAY:           -9999.000
> GRID AREA:           -80.00  156.00  -74.00  176.00
> EXTEND AREA:         -80.00  156.00  -74.00  176.00
> DATA AREA:           -80.00  156.00  -74.00  176.00
> 
> Number of grids in file:     0
> 
> Maximum number of grids in file:  10000
> 
> First time in file: ^L\233^G^Hp/
> ^Hh/
> ^Hl/
> ^Hd/
> ^H^V\307        ^H\200
> Last  time in file: \200
> 
> 
> 
> You can grab a file of the GRIB data to test with from:
> 
> http://psnwx.balloonfacility.org/csbf/2007120300_D5.grb
> 
> username is csbfmet, password is p!bal.
> 
> Thanks a bunch,
> Robert Mullenax
> CSBF Meteorology
> 
> 
> 
> 


Ticket Details
===================
Ticket ID: UER-356689
Department: Support GEMPAK
Priority: Normal
Status: Closed