[ncl-talk] Trouble with GFS grib table

[David Brown]

The messages are just trying to be informative. I do think that NCEP
table 133 parameter 191 is not yet incorporated in the NCL code.
Currently the GRIB 1 tables are compiled into the code as C header
files, so they cannot be updated by the user. You can set up a
user-defined table to override the current version of the table if you
like. There are instructions at
http://www.ncl.ucar.edu/Document/Manuals/Ref_Manual/NclFormatSupport.shtml#GRIB1-user-defined-parameter-tables
 -dave

On Fri, Jun 26, 2015 at 12:43 PM, Walter Hannah <walter at hannahlab.org> wrote:
> Thanks Dave and Kyle,
>
> I'm running version 6.3, so maybe I need to update the tables. Where would I
> find the files to do this?
>
> My code seems to be working ok otherwise, so I'll just ignore it for now.
>
> Thanks,
> Walter
>
> On Thu, Jun 25, 2015 at 11:15 PM, Kyle Griffin <ksgriffin2 at wisc.edu> wrote:
>>
>> Walter,
>>
>> I'm not sure of the version you are running, but NCL 6.3.0 generally
>> doesn't produce any such errors for me when I process real-time GFS files
>> (including GFS analysis, I believe). I did see similar warnings in versions
>> of NCL prior to 6.3.0 and after the January upgrade of the GFS changed a
>> number of different things. Each version usually includes a fresh set of
>> GRIB tables, so even if you can't/don't want to update NCL (probably the
>> easiest solution), you should be able to at least grab the updated tables
>> from a new version.
>>
>>
>> Kyle
>>
>> ----------------------------------------
>> Kyle S. Griffin
>> Department of Atmospheric and Oceanic Sciences
>> University of Wisconsin - Madison
>> Room 1421
>> 1225 W Dayton St, Madison, WI 53706
>> Email: ksgriffin2 at wisc.edu
>>
>> On Thu, Jun 25, 2015 at 5:41 PM, David Brown <dbrown at ucar.edu> wrote:
>>>
>>> Hi Walter,
>>> Well the GFS analysis is not static. New parameters are added from
>>> time to time. NCEP updates their GRIB tables as the need arises as
>>> well. The tables that NCL uses are updated periodically to incorporate
>>> changes that NCEP makes, but there is necessarily a lag since for now
>>> at least this is a manual process.
>>> In this case, the message about parameter 191 in NCEP table 133 does
>>> indicate a new parameter that NCL has not yet incorporated. For the
>>> record, the docs
>>> (http://www.nco.ncep.noaa.gov/pmb/docs/on388/table2.html#TABLE133) say
>>> this parameter represents "Sunshine Duration" with units of seconds.
>>> The other messages document a record in the file that will never have
>>> a name. GRIB 1 generally uses the number 255 to mean "missing". So in
>>> this case there is both a table with the "missing" designation and in
>>> that table there is a parameter that is "missing". I have no idea what
>>> this record is supposed to represent or why it appears. NCL is just
>>> letting you know about it. Perhaps the GFS folks could explain its
>>> purpose.
>>>  -dave
>>>
>>> On Thu, Jun 25, 2015 at 3:51 PM, Walter Hannah <walter at hannahlab.org>
>>> wrote:
>>> > Hi,
>>> >
>>> > I downloaded some analysis data from the NOMADS server, specifically:
>>> >
>>> > http://nomads.ncdc.noaa.gov/data/gfsanl/
>>> >
>>> > But when I try to read the files with NCL, I get this message:
>>> >
>>> > warning:NclGRIB: Unrecognized parameter table (center 7, subcenter 0,
>>> > table
>>> > 255), defaulting to NCEP operational table: variable names and units
>>> > may be
>>> > incorrect
>>> > warning:NclGRIB: Unknown grib parameter number detected (255, center 7,
>>> > table version 255 grib record 3), using default variable name (VAR_255)
>>> > warning:NclGRIB: Unknown grib parameter number detected (191, center 7,
>>> > table version 133 grib record 240), using default variable name
>>> > (VAR_191)
>>> >
>>> > Looking at some previous ncl-talk tthreads, I think I understand the
>>> > issue,
>>> > but I'm confused that the GFS analysis wouldn't be consistent with the
>>> > default NCEP grib tables.
>>> >
>>> > Can anyone shed some light on what's going on here?
>>> >
>>> > Thanks,
>>> > Walter
>>> >
>>> > _______________________________________________
>>> > ncl-talk mailing list
>>> > ncl-talk at ucar.edu
>>> > List instructions, subscriber options, unsubscribe:
>>> > http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>> >
>>> _______________________________________________
>>> ncl-talk mailing list
>>> ncl-talk at ucar.edu
>>> List instructions, subscriber options, unsubscribe:
>>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>
>>
>