global-workflow icon indicating copy to clipboard operation
global-workflow copied to clipboard

tocgrib2 fails in AWIPS job

Open WalterKolczynski-NOAA opened this issue 1 year ago • 3 comments
trafficstars

What is wrong?

The awips job uses tocgrib2 to add the WMO header to grib files. The executable is currently failing due to missing records.

What should have happened?

tocgrib2 should run successfully, allowing the AWIPS job to do the same and produce the associated products.

When this is fixed, DO_AWIPS should be able to be turned for the C96_atm3DVar_extended test and the test still pass on WCOSS.

What machines are impacted?

WCOSS2

Steps to reproduce

  1. Check out and build develop on WCOSS
  2. Create an experiment with AWIPS on
  3. Observe (almost) all of the AWIPS jobs (gfsawips_20km_1p0deg_*) failing in the first full cycle

Additional information

The records tocgrib2 is having issue with are the four convective precip fields (CRAIN, CFRZR, CICEP, CSNOW).

*******************************************************************************
 Start new record no. =  254
 DESC= CRAIN    Surface                                                        
 WMOHEAD=ZMNB98 KWBC
 GRIB2 DISCIPLINE= -1
 Section 1= -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999
 PDT 4. 8 = 1 192 2 0 96 0 0 1 0 1 0 0 255 0 0 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999
 GETGB2P: ERROR REQUEST NOT FOUND

*******************************************************************************
 Start new record no. =  255
 DESC= CFRZR    Surface                                                        
 WMOHEAD=ZMNB98 KWBC
 GRIB2 DISCIPLINE= -1
 Section 1= -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999
 PDT 4. 8 = 1 193 2 0 96 0 0 1 0 1 0 0 255 0 0 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999
 GETGB2P: ERROR REQUEST NOT FOUND

*******************************************************************************
 Start new record no. =  256
 DESC= CICEP    Surface                                                        
 WMOHEAD=ZMNB98 KWBC
 GRIB2 DISCIPLINE= -1
 Section 1= -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999
 PDT 4. 8 = 1 194 2 0 96 0 0 1 0 1 0 0 255 0 0 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999
 GETGB2P: ERROR REQUEST NOT FOUND

*******************************************************************************
 Start new record no. =  257
 DESC= CSNOW    Surface                                                        
 WMOHEAD=ZMNB98 KWBC
 GRIB2 DISCIPLINE= -1
 Section 1= -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999
 PDT 4. 8 = 1 195 2 0 96 0 0 1 0 1 0 0 255 0 0 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999 -9999
 GETGB2P: ERROR REQUEST NOT FOUND


 ******** RECAP OF THIS EXECUTION ********
     READ     257 INDIVIDUAL IDS
     WROTE    253 BULLETINS OUT FOR TRANSMISSION


 BULLETINS MISSING =  4

The grib file being supplied to tocgrib2 does have those fields:

WCOSS2 (BACKUPSYS) awips_20km_1p0deg.119475> $WGRIB2 awps_file_f003_conus | egrep '(CRAIN|CFRZR|CICEP|CSNOW)'
172:10840359:d=2021122100:CSNOW:surface:3 hour fcst:
173:10843301:d=2021122100:CICEP:surface:3 hour fcst:
174:10843489:d=2021122100:CFRZR:surface:3 hour fcst:
175:10843823:d=2021122100:CRAIN:surface:3 hour fcst:

Do you have a proposed solution?

I wonder if something needs to be updated in the 20km WMO parm files.

WalterKolczynski-NOAA avatar May 03 '24 00:05 WalterKolczynski-NOAA

@WalterKolczynski-NOAA, could you provide the experiment setup to run the C96_atm3DVar_extended test? For example,

./setup_expt.py gfs cycled --app ATM --start cold --resdetatmos 96 --resensatmos 96 --nens 80 --gfs_cyc 4 --idate 2021122018 --edate 2021122018 --comroot $COMROOT --expdir $EXPDIR --pslot test

GwenChen-NOAA avatar May 16 '24 16:05 GwenChen-NOAA

@WalterKolczynski-NOAA can you provide on of the GRIB2 files which are not working on hera somewhere and let me know where it is?

We will then build a test for it and make sure it works.

edwardhartnett avatar May 18 '24 05:05 edwardhartnett

@WalterKolczynski-NOAA can you provide on of the GRIB2 files which are not working on hera somewhere and let me know where it is?

We will then build a test for it and make sure it works.

@edwardhartnett, you can find some sample input files (gfs.tCCz.pgrb2.0p25.fFFF and gfs.tCCz.pgrb2b.0p25.fFFF) on WCOSS2 at /lfs/h2/emc/vpppg/noscrub/Lichuan.Chen/gwDev/COMROOT/test/gfs.20160701/00/products/atmos/grib2/0p25. The CRAIN, CFRZR, CICEP, and CSNOW fields are generated for forecast lead time >= 003. The JGFS_ATMOS_AWIPS_20KM_1P0DEG job ran successfully for f000 and generated AWIPS products. The f000 awips files do not output the CRAIN, CFRZR, CICEP, and CSNOW fields.

GwenChen-NOAA avatar May 20 '24 14:05 GwenChen-NOAA