19991210: Getting Numerical Products (cont.)

>From: Erick Lorenz (address@hidden) <address@hidden>
>Organization: UC Davis
>Keywords: 199912100033.RAA06506 ldm-mcidas

Erick,

re: LDM-MCIDAS?
>Sorry I was just trying to refer to the whole system of LDM and McIDAS
>rather than to a specific package.

No problem.

re: 'xcd_run HRS' entry in ~ldm/etc/pqact.conf
>This was already not commented

OK.

re: decinfo.k SET DMGRID ACTIVE
>I did this as you suggested

>I see some GRIDnnnn files showing up with today's date.  Is this what I 
>should be looking for?

Yes.  I took a quick at your GRID files (via ADDE) and saw that the
the XCD decoders are working.

>I have tried using the F-Keys menu to display model data but so far 
>I am getting no results.

There are a couple of things here.  The Fkey menu actions are setup to
access gridded data in ADDE datasets RTGRIDS/NGM-UW and RTGRIDS/AVN-UW.
The GRID files that would compose these datasets are the ones that
were removed from the Unidata-Wisconsin broadcast back on July 1.
The latest version of the Fkey menu allows you to change the GRID
datasets you point at for these menus, but not all of the grid fields
that were accessible from the UW broadcast files will be available
(since some like the 250 mb wind speed grids were created by PostProcessing
after the broadcast GRID was received).

There are two options here:

o point at XCD-produced datasets

o recreate the NGM-UW and AVN-UW datasets from XCD-decoded grids on
  your system

To point at XCD-produced GRID datasets, you would do:

TE ?MODEL1 "RTGRIDS/NGM
TE ?MODEL2 "RTGRIDS/AVN

from the McIDAS session for each/every user.  Then when you go to put
up loops of gridded fields from the Fkey menu F3 and F4 actions, you
should get plots.  What will be missing, however, is the ability to
do the cross sections in Fkey menu action F5.  The reason for this
is that the grid cross section plotter in McIDAS has not been ADDEized
yet.

To get information on how to create the NGM-UW and AVN-UW datasets, you
should read the information header of the script 'uwgrid.sh' which
will be located in the ~mcidas/workdata directory.  The tricky part
in running this script from cron is setting the times that it runs
so that the needed input GRID files are available to produce the
new output GRID files.


>I thought to use GRDLIST to check for available
>data but I don't know how to name the dataset.

You can see what is setup on a Unidata ADDE by looking at:

o LSSERVE.BAT (a copy of DSSERVE.BAT) in ~mcidas/data
o RESOLV.SRV in ~mcidas/workdata (entries created by LSSERVE.BAT entries)

If you knew that the grid dataset was named RTGRIDS (it is in the default
setup we send out in DSSERVE.BAT), you could see the various possible
types of grids by running:

DATALOC ADD RTGRIDS atm25.ucdavis.edu
DSINFO GRID RTGRIDS

You could then list grid file information for a particular gridded
dataset using the GRDLIST command.  For example, here is the output
from my session looking at your machine:

DSINFO GRID RTGRIDS

        Dataset Names of Type: GRID  in Group: RTGRIDS

Name         NumPos   Content
------------ ------   --------------------------------------
ALL            500    Real-Time Grids
AVN             80    Real-Time AVN Grids
ECMWF           10    Real-Time ECMWF Grids
ETA             40    Real-Time ETA Grids
MDR             10    Real-Time MDR Grids
MISC            10    Other Real-Time Grids
MRF            100    Real-Time MRF Grids
MRF-UW           2    MRF Grids in UW stream
NGM             40    Real-Time NGM Grids
NGM-UW           2    NGM Grids in UW stream
RUC             80    Real-Time RUC Grids

DSINFO -- done

GRDLIST RTGRIDS/ETA.ALL FORM=FILE
DATASET NAME: RTGRIDS/ETA
 Dataset Position  Creation Date Max Grids         Directory Title
 ----------------  ------------- --------- -------------------------------
       24             1999344       2000   ALL  18Z ETA 0  HR<=VT<=24 HR
       34             1999344       2000   ALL  18Z ETA 24 HR< VT<=48 HR
GRDLIST - done

The above shows that you have 18Z ETA runs for forecasts from 0 to 48 hours.
Actually, it shows that the files exist; it doesn't show what is actually
in the files.  For this you can use a GRDLIST invocation like:

GRDLIST RTGRIDS/ETA.24 NUM=ALL
Dataset position 24     Directory Title= ALL  18Z ETA 0  HR<=VT<=24 HR
PAR    LEVEL       DAY        TIME    SRC FHOUR     FDAY       FTIME  GRID  PRO
---- --------- ------------ -------- ---- ----- ------------ -------- ----- ----
Z    SFC       10 DEC 99344 18:00:00  ETA     0 10 DEC 99344 18:00:00     1 LAMB
Z    1000 MB   10 DEC 99344 18:00:00  ETA     0 10 DEC 99344 18:00:00     2 LAMB
Z     975 MB   10 DEC 99344 18:00:00  ETA     0 10 DEC 99344 18:00:00     3 LAMB
Z     925 MB   10 DEC 99344 18:00:00  ETA     0 10 DEC 99344 18:00:00     4 LAMB
Z     900 MB   10 DEC 99344 18:00:00  ETA     0 10 DEC 99344 18:00:00     5 LAMB
Z     875 MB   10 DEC 99344 18:00:00  ETA     0 10 DEC 99344 18:00:00     6 LAMB
Z     950 MB   10 DEC 99344 18:00:00  ETA     0 10 DEC 99344 18:00:00     7 LAMB
Z     850 MB   10 DEC 99344 18:00:00  ETA     0 10 DEC 99344 18:00:00     8 LAMB
Z     750 MB   10 DEC 99344 18:00:00  ETA     0 10 DEC 99344 18:00:00     9 LAMB
Z     825 MB   10 DEC 99344 18:00:00  ETA     0 10 DEC 99344 18:00:00    10 LAMB
 ...

(there were 343 grids in this list, so I truncated the output).

>If this procedure is all I need to get model files then I don't see
>the need to go over the other stuff for now.

OK.

>I didn't know about the
>SET DMGRID command and in poring over the documentation about decoders
>I thought I saw something that suggested that that was what I needed.

This information is contained in the online documentation.

>so I downloaded the decoders file and tried to install it with the
>problems that I mentioned.

Right.

>Thank you

You are welcome.

Tom