[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[netCDFDecoders #DDA-176665]: dcgrib2 and DSS MRF data




Kristen,

The data file you downloaded from mass store appears corrupt.
Each of the GRIB products in the file is 32 or 40 bytes to long.
Dcgrib2 checks the bulletin against the stated size which is the error
message you saw. I typically see this when a file is transferred by ftp
as text rather than binary, and the ftp inserts some line junk into the data.
I don't have access to NCAR mass store stuff, so can't tell you whether the 
file you
have is exactly as they archived it, or if it was corrupted on download
or transfer. You might try consulting with one of the operators to
see if the file was clobbered on transfer. The web page you referenced mentioned
that the file was ftp'd from NCEP, so hopefully they didn't clobber the files
originally on download.

The nagrib program will ignore that error and decode it until the needed length,
but when you try to plot the data, it will appear as junk.

Steve Chiswell
Unidata User Support

> Steve-
> 
> Sorry about that.  Yes, I do want to put the GRIB files into GEMPAK
> format.  The web site said the first  address I e-mailed was for netCDF
> and GRIB decoders, but I did debate as to whether I should e-mail the
> GEMPAK address instead.
> 
> In any event, I have attached the output from the 'dd' command.
> 
> Thanks,
> Kristen
> 
> Kristen L. Corbosiero
> ASP / MMM
> National Center for Atmospheric Research
> P.O. BOX 3000
> Boulder, CO 80307-3000
> 
> Office: FL3 Rm.2068
> e-mail: address@hidden
> phone:  (303) 497-8185
> web: http://www.atmos.albany.edu/student/kristen/index.html
> 
> 
> 
> Unidata GEMPAK Support wrote:
> > Kristin,
> >
> > I'm assuming that you want to use the data in GEMPAK since you mentioned 
> > dcgrib2, even though the
> > question was set to the netcdf decoders address. Is that the case?
> >
> > There are many tools that can parse grib messages such as wgrib and dump 
> > out a
> > single grid.
> >
> > However, since your first error message states that you were looking at 
> > byte position 18481
> > when not finding the "7777" message, the easiest thing to do would just be 
> > to
> > send me 100kbytes or so, so that I can see the first 2 or 3 products.
> > The "dd" command is useful for this purpose. Just do:
> > dd if="your_mrf_grib_file" ibs=100000 obs=100000 count=1 > testmrf.grib
> >
> > in the above "if" is the input file, the input and output block size is 
> > 100,000 bytes
> > and just 1 block is dumped out to the testmrf.grib file. You can send that 
> > file
> > as an email attachmet to me, or post via the web support form.
> >
> > Steve Chiswell
> > Unidata User Support
> >
> >
> >
> >
> >
> >> Hi Steve-
> >>
> >> No, I did not ftp them.  As I'm here at NCAR, I just copied them over
> >> from the mass store using 'msrcp'.  The files contain 15 days of
> >> forecasts each and thus are on the large side (3 G).  I'm not very
> >> familiar with handling GRIB files, but is there some way to cut it down
> >> or look at a sample so that I could send you a section to look at?
> >>
> >> The DSS web page says that the 1997 grids are 'List B', which is on a
> >> 2.5 deg grid in GRIB.
> >>
> >> Thanks,
> >> Kristen
> >>
> >> Kristen L. Corbosiero
> >> ASP / MMM
> >> National Center for Atmospheric Research
> >> P.O. BOX 3000
> >> Boulder, CO 80307-3000
> >>
> >> Office: FL3 Rm.2068
> >> e-mail: address@hidden
> >> phone:  (303) 497-8185
> >> web: http://www.atmos.albany.edu/student/kristen/index.html
> >>
> >>
> >>
> >> Unidata GEMPAK Support wrote:
> >>
> >>> Kristen,
> >>>
> >>> Did you FTP these files? If so, there might be a problem
> >>> with needing to transfer them in binary mode. The decoder should
> >>> be able to find the GRIB delimeters otherwise. When you start FTP, make
> >>> sure you thpe "bin" at the ftp prompt. Otherwise, you may need to post a 
> >>> sample file
> >>> for me to look at.
> >>>
> >>> If the data are on a 2.5 degree grid (eg grid #002), you should have no
> >>> trouble using the data in GEMPAK. If however, you have sperical harmonics 
> >>> rather than
> >>> gridpoints, GEMPAK will not handel those.
> >>>
> >>> Steve Chiswell
> >>> Unidata User Support
> >>>
> >>>
> >>>
> >>>
> >>>> Hi-
> >>>>
> >>>> I downloaded a couple of MRF forecast files from DSS at NCAR
> >>>> (http://dss.ucar.edu/datasets/ds084.0/).   The files are from 1997 and
> >>>> thus in GRIB format at 2.5 degree resolution.  I tried to degrib these
> >>>> files using dcgrib2 and got the following error message:
> >>>>
> >>>> [28534] 061011/1245[DC 3]  Starting up. Version 5.9.1
> >>>> [28534] 061011/1245[DCGRIB -60] no 7777 found: bpos [18481] > griblength
> >>>> [18480]
> >>>>
> >>>> repeated about 20,000 times.  I've looked at the Unidata help archives
> >>>> and see that the '7777' is supposed to be a marker to separate between
> >>>> dates, variables or grids.
> >>>>
> >>>> Any help that you could provide as to what the problem is and how I
> >>>> might degrib this data would be much appreciated.
> >>>>
> >>>> Thanks!
> >>>>
> >>>> Kristen
> >>>>
> >>>> --
> >>>> Kristen L. Corbosiero
> >>>> ASP / MMM
> >>>> National Center for Atmospheric Research
> >>>> P.O. BOX 3000
> >>>> Boulder, CO 80307-3000
> >>>>
> >>>> Office: FL3 Rm.2068
> >>>> e-mail: address@hidden
> >>>> phone:  (303) 497-8185
> >>>> web: http://www.atmos.albany.edu/student/kristen/index.html
> >>>>
> >>>>
> >>>>
> >>>>
> >>> Ticket Details
> >>> ===================
> >>> Ticket ID: DDA-176665
> >>> Department: Support GEMPAK
> >>> Priority: Normal
> >>> Status: Closed
> >>>
> >>>
> >>>
> >>>
> >>>
> >>
> >
> >
> > Ticket Details
> > ===================
> > Ticket ID: DDA-176665
> > Department: Support GEMPAK
> > Priority: Normal
> > Status: Closed
> >
> >
> >
> >
> 
> 


Ticket Details
===================
Ticket ID: DDA-176665
Department: Support GEMPAK
Priority: Normal
Status: Closed