[Met_help] Problem with MET and accumulation intervals

[John Halley Gotway]

Thomas,

The flags in your GRIB file are set in a way that we've never encountered for precipitation.  Where did you get this GRIB file?

I was able to reproduce the error message you're seeing.  I stepped through the call to PCP-Combine using a debugger and found that unexpected flag value in the Product Description Section (PDS) of
the GRIB file.  I ran the file you sent and a sample GRIB file that's distributed with MET through the "wgrib" utility to dump out info about each record:

File you sent...    4:26640:d=07080100:APCP:kpds5=61:kpds6=1:kpds7=0:TR=0:P1=4:P2=0:TimeU=13:sfc:1hr fcst:NAve=0
Sample GRIB file... 2:31408:d=05080700:APCP:kpds5=61:kpds6=1:kpds7=0:TR=4:P1=3:P2=6:TimeU=1:sfc:3-6hr acc:NAve=0

Looking at these, you can see that the value for the time range indicator (TR) is set to 0 in yours, while it's set to 4 in the other file.  Now take a look at Table 5 of the GRIB spec:
http://www.nco.ncep.noaa.gov/pmb/docs/on388/table5.html

A value of 0 is used to simply define a valid time, while a value of 4 is used to define an accumulation interval between times 1 and 2 (P1 and P2).  MET expects that the time range indicator will be
set as an accumulation interval for precipitation.

Do you think there was an error in defining the flags for this GRIB file, or do you really need to be able to read this file as is in MET?  If you do need to read it in MET, you'll need to make two
source code changes:

(1) Replace the file METv2.0/lib/vx_met_util/read_grib.cc with the version that's attached.
(2) Replace the file METv2.0/src/pcp_combine/pcp_combine.cc with the version that's attached.

Recompile MET with a "make clean" and then rebuild MET.  Please let me know the origins of this file and whether we'll need to support it directly in MET.

Hope that helps,
John


Thomas Schwitalla wrote:
> Hi John,
> 
> currently I have a problem using pcp_combine with a grib dataset for
> precipitation verification. I get the following error message as an
> example:
> "can't find grib code 61 with accumulation of XX hours in grib file". I
> attached a sample grib file, maybe you can reproduce the error.
> 
> Best regards from Hohenheim
> Thomas
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Met_help mailing list
> Met_help at mailman.ucar.edu
> http://mailman.ucar.edu/mailman/listinfo/met_help