[ncl-talk] reading s2s ecmwf forecast grib file

[David Brown]

Hi Ousmane,
I finally was able to get the complete file and it appears to be as I
suggested earlier, that NCL is not reading all the records because it
does not yet handle GRIB2 product definition template 61. This one
should be relatively easy to add to NCL so I think we should be able
to have this fixed in time for the next NCL release (6.5.0) which is
going to be available soon.
  -dave

On Mon, Sep 4, 2017 at 5:10 AM, ousmane ndiaye <ondiaye70 at gmail.com> wrote:
> Thanks dave
> for your answer and investigation
> The size is actually :
>
> -rwxrwxrwx   1 macbook  staff  1298838 Jul 20 16:36 1Sep2009.grib
>
> I ftp it while changing the name to "file1Sep2009.grib"
> Ousmane
> thanks
>
>
> On Thu, Aug 31, 2017 at 9:37 PM, David Brown <dbrown at ucar.edu> wrote:
>>
>> Hi Ousmane,
>> The second error message is pretty minor. It simply indicates that the
>> table used to give a value to the "production_status" attribute is not
>> quite up-to-date. If it were up-to-date. you would see
>>
>>  production_status : S2S Operational Products
>>
>> instead of
>>
>> production_status : 6
>>
>> (6 is the code for "S2S Operational Products" in the table 1.6.table)
>>
>> The first error is more serious, but I am unable to duplicate it with
>> any version of NCL that I have.  One possible reason for this is that
>> the file you uploaded may be truncated. It only contains one complete
>> GRIB record, but it also contains the beginning of a second record.
>> The file as we received it is 24020 bytes:
>> ls -l 1Sep2009.grib
>> -rw-r--r-- 1 dbrown ncar 24020 Aug 31 15:15 1Sep2009.grib
>>
>> If your file is bigger than that, then the file was not successfully
>> uploaded to the ftp site. Can you check the size of your file and if
>> necessary upload it again? If it is the same size, then your file is
>> incomplete and maybe you can retrieve it again from wherever it came
>> from.
>>
>> I can tell you that this file uses a GRIB2 Product Definition Template
>> number (see
>> http://www.nco.ncep.noaa.gov/pmb/docs/grib2/grib2_table4-0.shtml),
>> 61, that NCL does not yet fully support. That means that the error you
>> are seeing would not be surprising assuming the file is complete. We
>> will be updating our code to handle this template as soon as possible.
>>
>> Thanks for your report and hopefully you can send a complete file to us.
>>  -dave
>>
>>
>>
>> On Thu, Aug 31, 2017 at 6:50 AM, ousmane ndiaye <ondiaye70 at gmail.com>
>> wrote:
>> > Dear All,
>> >
>> > I downloaded forecast file from ecmwf s2s data base in grib2 format.
>> >
>> > When I try to read it I got this error :
>> >
>> > ncl 3> ff=addfile("1Sep2009.grib","r")
>> >
>> > warning:NclGRIB2: TPRATE_P61_L1 contains records that NCL cannot
>> > currently
>> > differentiate. One or more records will be ignored.
>> >
>> > warning: Entry (6) not found in code table file
>> >
>> > /Users/macbook/Downloads/ncl-6.4.0/lib/ncarg/grib2_codetables/ncep/1/1.3.table
>> >
>> >
>> >
>> > ==
>> >
>> > any help
>> >
>> > Ousmane
>> >
>> >
>> >
>> >
>> > _______________________________________________
>> > ncl-talk mailing list
>> > ncl-talk at ucar.edu
>> > List instructions, subscriber options, unsubscribe:
>> > http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>> >
>
>