[gembud] dcreanal bug?

• To: support-gempak@xxxxxxxxxxxxxxxx
• Subject: [gembud] dcreanal bug?
• From: pmanousos@xxxxxxxxxxxxxxxxxxx
• Date: Tue, 3 Jun 2008 16:43:39 -0400

Hi..

The dcreanal utility is yielding unexpected results.  Here are the 
details:

I am using dcreanal from the binary distribution of NAWIPS 5.10.4 (running 
on a 64 bit system with the RHEL 5.1 OS)
The file to be decoded to GEMPAK grid format is air.1993.nc
Contains temperature data in 6 hour increments at the mandatory levels for 
all of 1993 at 2.5 x 2.5 deg resolution 
This file was retrieved from 
ftp://ftp.cdc.noaa.gov/Datasets/ncep.reanalysis/pressure/air.1993.nc
The command "dcreanal air.1993.nc YYYYMM_test.grd" produces 12 grid files 
(199301_test.grd through 199312_test.grd)
However, via gdinfo (gdattim.. glevel.. gvcord.. gfunc set to "all"), each 
grid file's VCORD and PARM listings are empty (gdattim and level info are 
stored properly however) - example output below:
NUM     TIME1      TIME2 LEVL1 LEVL2 VCORD PARM
  1   930101/0000         1000 
  2   930101/0000          925 
  3   930101/0000          850
etc.
dcreanal from older versions of GEMPAK (back to 5.10.1) provide the same 
result
But an even "older" compiled version of dcreanal predating 5.10.1 
(probably from 5.9.x or earlier, actual version unknown) DOES work - that 
is it produced gridded files containing the proper date/time glevel, 
gvcord, and parm listings - example output below:
NUM     TIME1      TIME2 LEVL1 LEVL2 VCORD PARM
  1   930101/0000         1000        PRES TMPK
  2   930101/0000          925        PRES TMPK
  3   930101/0000          850        PRES TMPK
etc.


Therefore, the air.1993.nc file is not corrupt and the required 
ncarncep1.tbl file is properly detected and utilized.

So although dcreanal isn't crashing, it is not properly storing the vcord 
and parm data properly in the output grid.

Pete