[Met_help] Re: MET FEEDBACK from Dan Schaffer (fwd)

[John Halley Gotway]

Dan,

Ah, I think I understand now.  You were trying to read gridded data in NetCDF format.  Currently, we really only support gridded data in Grib format.  In future releases however, we do plan to support
NetCDF format more fully.  The pcp_cobmine tool does write out data in NetCDF format which can be read in by Grid_Stat or MODE.  However, they expect that any NetCDF files are the output of
pcp_combine with all the variables and dimensions included as such.  In a future release, we plan to select a NetCDF convention to support so that we'll be able to handle NetCDF much better.

For the time being, your stuck with either using Grib data or piecing something together to handle your flavor of NetCDF, as you did.

Thanks for following up on that.  I appreciate it.
John

Daniel Schaffer wrote:
> Thanks for your quick response John.
> 
>> 4. It should actually be pretty easy to run MET on variables other than
>> precip.  That was our goal in using configuration files.  Take a look in
>> the MET User's Guide for a discussion of how to edit the configuration
>> files for each tool to run it on whatever field(s) are in your input Grib
>> files.
>>
>> Point_Stat: See section 4.3.2 of the MET User's Guide and look at the
>> vx_grib_code parameter.
>> Grid_Stat: See section 5.3.2 of the MET User's Guide and look at the
>> vx_grib_code parameter.
>> MODE: See section 6.3.2 of the MET User's Guide and look at the
>> fcst_grib_code and obs_grib_code parameters.
> I did take advantage of these configurable parameters (although since it appeared that my particular field, Vertically Integrated Liquid as a representation of convection does not appear to be one of the GRIB codes, I had to add a grib code).
> 
> However, what was more concerning is that the code calls read_pcp_combine_netcdf.cc which appears to hard-wire in the fact that the variable in the netcdf file must be called "precip" and that there must be an "accum_time" attribute.  Is there something I'm still missing here?
> 
> All that said, I'm really in good shape right now so these are more of academic questions/observations that you may want to think about for MET 2.0.
> 
> Dan
> 
> _______________________________________________
> Met_help mailing list
> Met_help at mailman.ucar.edu
> http://mailman.ucar.edu/mailman/listinfo/met_help