19990708: garp questions

Lynn,

You might need to check the number of grids you stored in the
gempak file from the grib file.

Did you use dcgrib or nagrib?

The problem seems to be that a vertical coordinate number
used in the  grib file did not exists in the gempak table.
If you run gdinfo, you might see a blank instead of
a vertical coordinate.

If you use dcgrib with -v (and probably need -m 10000 to allow for
up to 10,000 grids if you want all 384 hours of the mrf),
you might see a message about a coordinate not found in
the vcrdgrib1.tbl file.

I'll send you the most recent  table so that you can compare. 

The second problem you describe does sound as if the key were
finding a duplicate entry....which you have checked. It might
be the case that f00 is used both in the 00 file and in the 12 file,
eg TIME1 in the first file and TIME1:TIME2 in the 12hour forecast file.

I have not seen this problem before- I store all of the forecast hours
from the 104 ETA and MRF in the single model run files.

Steve Chiswell



>From: Lynn McMurdie <address@hidden>
>Organization: .
>Keywords: 199907081902.NAA03817

>Hi...
>
>I've got an error message that I need help decifering.  We recently
>obtained global MRF data from a contact at NCEP for a recent case that
>occurred last march.  We downloaded the grib data from them, then ran
>our own script to create gempak files that can then be used by GARP.
>I got the configure file all arrange correctly and can actually plot
>the data (however, I haven't tried all the fields), however, I've
>gotten two messages.
>
>First, when I go to modelplanprojection key and select the button for
>our data, I immediately get an error that says:
>
>44 skye% G A R P - v2.01 starting...
> lv_ccrd error   233 lev   0  -1 jvcord   -1
>  CreateGridInfoObject: getnxtgrd ret -3
>
>Then, I can still manage to plot the data as if nothing happened.
>But, then I tried to look at the data file itself using the "info" key
>and as soon as I selected the time I wanted, I got this error:
>
> lv_ccrd error   27 lev   0  -1 jvcord   -1
>  CreateGridInfoObject: getnxtgrd ret -3
>Segmentation Fault
>
>Clearly, garp crashed after that.  So, I'm wondering what these error
>messages mean.  I suspect something went wrong in the translation from
>grib files to gempak files, but am not clear what.
>
>Also, somewhat unrelated.  I can't figure out why in the
>modelplanprojection pop-up window that these files show the available
>times twice.  For example, the time is shown twice:
>
>99030100f00 
>99030100f00 and I can't plot anything using the second time.  The
>files are named:
>99030100_mrf00.gem
>99030100_mrf12.gem
>99030100_mrf24.gem
>
>etc., and the model key is:  mrf.  There are no other files in the
>directory with mrf in their names and all the files are gempak files
>in my directory.  The same problem happens for my ETA104 grids.  The
>files are named 99030100_meso_12.gem, 99030100_meso_18.gem, etc. and
>the model key is: meso, with no other meso to be seen in the
>directory.  So, I'm wondering if I'm screwing up the naming of the
>files somehow.
>
>Thanks for your help.
>       Lynn
>
>-- 
>Lynn McMurdie          INTERNET: address@hidden
>(206) 685-9405         UUCP:     uw-beaver!atmos.washington.edu!mcmurdie
>Dept of Atmospheric Sciences
>University of Washington
>Box 351640
>Seattle, WA  98195
>