20000119: LDM/GEMPAK support

>From: "Ka Kit Lai" <address@hidden>
>Organization: .
>Keywords: 200001191719.KAA01084

>Robb:
>   This is Ka Kit from CCNY. Sorry to bother you again. I know you are 
>really busy, but I need help to configure the LDM with the Gempak that I 
>just installed. Please tell me when you will be available. I think I already 
>have all the related files downloaded from Unidata and the correct data feed 
>from albany.
>
>                                      Ka Kit Lai
>______________________________________________________
>Get Your Private, Free Email at http://www.hotmail.com
>

Ka Kit,

In the decoders examples that we provide from the GEMPAK www pages
for the pqact.conf LDM file, you will need to edit the location to the 
GEMPAK tables to relate to where you installed the software on your system. 
Most of the rest should be ready to go.

For example, in the decoders.tbl examples:
# NOAAport ETA grids
# Grid #211 80km CONUS:    ^[YZ].Q.*/mETA
# Grid #212 40km CONUS:    ^[YZ].R.*/mETA
# Grid #215 20km CONUS:    ^[YZ].U.*/mETA
# Grid #214 47.5km Alaska: ^[YZ].T.*/mETA
# Select any/all grids desired from [QRUT]
HRS     ^[YZ].[Q].*/mETA
        PIPE    /usr/local/ldm/decoders/dcgrib -d data/gempak/logs/dcgrib.log
        -g /home/gempak/NAWIPS-5.4/gempak5.4/tables
        PACK data/gempak/hds/YYYYMMDDHH_eta_grid@@@.gem


The path /usr/local/ldm/decoders/dcgrib should point to where the dcgrib
decoder is installed on your system. This would be the $GEMEXE directory
on your system. We generally copy our decoders over to the LDM tree 
so that the decoders for the different packages ate in the same
location for the LDM (but this is not required).

The path to the $GEMTBL directory is defined by the -g flag above.
Again, you will need to modify this location to where $GEMTBL is on your system.

You will need to create the ~ldm/data/gempak/logs directory as well
as the ~ldm/data/gempak/hds directory.

The FILE action for nwx creates the data directories, but the PIPE action
for decoders etc does not create the directories automatically.

The example above shows what grids the ETA model is distributed on.
The pattern: ^[YZ].[Q].*/mETA specifies to pass the grid 211 conus
grids to the decoder. If you also wanter to store the 212 grids
you would modify the pattern to: ^[YZ].[QR].*/mETA
To receive all of the different grid areas, you would use:
[YZ].[QRUT].*/mETA. Each of the lettern in the second [] entry 
represent different grid prodjections as explained above. 

When dcgrib decodes the data, the output file name: YYYYMMDDHH_eta_grid@@@.gem
will be replaced by the decoder with the current year, month, dat, and
hour of the model run and @@@ will be replaced by the grid projection
number. As a result, you will se files in the $HDS director named like
2000011812_eta211.gem (if you store the eta 211 (identifier Q) grids.

If you chose not to decode the ETA grids, you would comment out these entries
with a # character at the beginning of the lines. You should chose what
decoders you want to run based on your data needs and how much grid
space you have available.


Similarly:
DDS|IDS ^S[AP].* .... ([0-3][0-9])([0-2][0-9])
        PIPE    /usr/local/ldm/decoders/dchrly -v 1 -b 9 -m 24
        -d data/gempak/logs/dchrly.log
        -p /home/gempak/NAWIPS-5.4/gempak5.4/tables/pack/hrly.pack
        -s /home/gempak/NAWIPS-5.4/gempak5.4/tables/stns/sfmetar_sa.tbl
        data/gempak/surface/YYYYMMDD_sao.gem

You will edit the location to the dchrly decoder and edit the path to
the $GEMTBL/pack/hrly.pack and $GEMTBL/stns/sfmetar_sa.tbl files.
You should only have to modify the path up to: /home/gempak/NAWIPS-5.4
in order to match your $NAWIPS directory location. Everything else under
that would be the same.

One note on the nwx.tbl actions. All of the entries simply FILE the
data to the $TEXT_DATA tree for use by NWX, except the Zone forecast products
which need to be reassymboled using the ldmConnect perl script provided in
the ~ftp/pub/decoders/decoders.tar.Z distribution on our anonymous ftp
server. The actions look like:
WMO     ^FPUS5[1-6] .... ([0-3][0-9])([0-2][0-9])
        PIPE    /usr/local/ldm/decoders/ldmConnect -e .zone (\1:yy)(\1:mm) 
data/gempak/nwx /

Since the Zone forecasts can be large, the NWS breaks them up into smaller 
pieces.
The ldmConnect script joins the pieces back together before filing them.

Steve Chiswell