20010924: dcgrib2 problem: "bulletin too long"

Brian,

The decoder looks for complete GRIB messages and verifies that
the length of the bulletin obtained is that which is defined in
the Section 0 (IS) block of the GRIB bulletin.
By checking the obtained length against the expected length, the
decoder guards against possible corrupt data.

The bulletin too long message means that the "7777" end sequence found is
at a point greater than the stated length of the product. This would typically
happen if either: 1) The 7777 for a product was ommitted, or 2) at some point,
the bulletin became garbled and you have found the 7777 for the next
product.

In either case, it is not safe to assume that the bulletin is correct.

The same case is discovered when using the gribdump utility included
with the decoders.tar.Z distribution.

Steve Chiswell
Unidata User Support





On Mon, 24 Sep 2001, Brian Jewett wrote:

> Hello,
> 
> We're having problems processing Eta GRIB files with dcgrid2.
> For example,
> 
>   cat us008_gf084_01091000_YxQCx | dcgrib2 -v 1 YYYYMMDDHH_eta211.gem
> 
> resulted in:
> 
> [24956] 010919/1409 [DC 3]  Starting up.
> [24956] 010919/1409 [DCGRIB 0] Opened 2001091000_eta211.gem model:84
> [24956] 010919/1409 [DCGRIB 0] grid:211
> [24956] 010919/1409 [DCGRIB -9] End of Data
> [24956] 010919/1409 [DCGRIB 0] Closing 2001091000_eta211.gem
> [24956] 010919/1409 [DC 5]  Normal termination.
> [24956] 010919/1409 [DC 2]  Number of bulletins read and processed: 132
> [24956] 010919/1409 [DC 6]  Shutting down.
>  Changing center table to cntrgrib1.tbl
>  Changing vertical coord table to vcrdgrib1.tbl
>  Changing WMO parameter table to wmogrib2.tbl
>  Changing center parameter table to ncepgrib2.tbl
> bulletin too long 13828 > 5384
> bulletin too long 13828 > 5384
> bulletin too long 13826 > 5384
> bulletin too long 13072 > 5384
> bulletin too long 13072 > 5384
> bulletin too long 13072 > 5384
> bulletin too long 13071 > 5384
> bulletin too long 13072 > 5384
> bulletin too long 13072 > 5384
> bulletin too long 13072 > 5384
> bulletin too long 13071 > 5384
> bulletin too long 13072 > 5384
> bulletin too long 13072 > 5384
> bulletin too long 12316 > 5384
> bulletin too long 12316 > 5384
> bulletin too long 17606 > 5384
> bulletin too long 14581 > 7650
> bulletin too long 10806 > 5384
> bulletin too long 10806 > 5384
> bulletin too long 11560 > 5384
> bulletin too long 6137 > 5384
> bulletin too long 11560 > 5384
> bulletin too long 6186 > 5384
> 
> Has anyone else encountered this problem?
> 
> I have placed the grib file and output at
>       http://blizzard.atmos.uiuc.edu/rt/GRIB
> 
> Any comments or suggestions are appreciated..
> 
>   Dr. Brian Jewett
>   address@hidden
>   Dept. of Atmospheric Sciences
>   University of Illinois at Urbana-Champaign
> 
> 
>