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

20010315: 20010315: 20010314: From netCDF to gempak



Bill,

This data set seems to use a different convention than the reanalysis
files I had access to which identified themselves as COARDS in the global
attributes.

Where dcreanal expects "lat" and "lon" and the variable names,
your data has "X" and "Y". For time, dcreanal expects "time", and
your file has "T". For "level", you file just has "P". And instead of
"GRIB_id", your file has "GRIBgridcode".

I can update the dcreanal code to handle this. What is the source of these
files? Is it the new reanal2 files? The file looks almost identical except the
dimension and variable ids, except there is no way to identify the vertival
level as pressure unless you happen to be looking for "P".

Steve Chiswell
Unidata User Support


>From: Bill Ramstrom <address@hidden>
>Organization: UCAR/Unidata
>Keywords: 200103152020.f2FKKnL03434

>Here is the output from ncdump -h
>
>===============================================
>
>
>netcdf z {
>dimensions:
>        Y = 73 ;
>        X = 144 ;
>        T = 5 ;
>        P = 17 ;
>variables:
>        float Y(Y) ;
>                Y:long_name = "Latitude" ;
>                Y:gridtype = 0 ;
>                Y:units = "degree_north" ;
>        float X(X) ;
>                X:long_name = "Longitude" ;
>                X:gridtype = 1 ;
>                X:units = "degree_east" ;
>        int T(T) ;
>                T:long_name = "Time" ;
>                T:calendar = "standard" ;
>                T:defaultvalue = 19427 ;
>                T:gridtype = 0 ;
>                T:units = "days since 1948-01-01" ;
>        int P(P) ;
>                P:long_name = "Pressure" ;
>                P:gridtype = 0 ;
>                P:units = "mb" ;
>        float phi(T, P, Y, X) ;
>                phi:scale_max = 32480.f ;
>                phi:calendar = "standard" ;
>                phi:units = "gpm" ;
>                phi:PTVersion = 2 ;
>                phi:PDS_TimeRange = 113 ;
>                phi:GRIBgridcode = 2 ;
>                phi:long_name = "Geopotential height" ;
>                phi:process = "62 wave triangular, 28 layer Spectral model
>from \"Medium Range
> Forecast\" run" ;
>                phi:center = "US Weather Service - National Met. Center" ;
>                phi:grib_name = "HGT" ;
>                phi:subcenter = "NCEP Reanalysis Project" ;
>                phi:scale_min = -594.f ;
>                phi:gribleveltype = 100 ;
>                phi:gribparam = 7 ;
>                phi:missing_value = 9.999e+20f ;
>                phi:Compression.compressiontype = "null" ;
>}
>
>
>
>
>
>===============================================
>
>
>On Thu, 15 Mar 2001, Unidata Support wrote:
>
>> 
>> Lodovica, 
>> could you send me the output from "ncdump -h" so that I ca see what
>> the cdl looks like?
>> 
>> dcreanal doesn't use udunits, so no trouble there. The program is set up
>> according to the coards conventions that were used in the reanalysis
>> files I had access to. The parameter variables were 3 or 4 dimensional.
>> 
>> Steve Chiswell
>> Unidata User Support
>> 
>> 
>> 
>> 
>> 
>> >From: Lodovica Illari <address@hidden>
>> >Organization: UCAR/Unidata
>> >Keywords: 200103151842.f2FIglL27392
>> 
>> >Hi Steve,
>> >
>> >I have installed dcreanal in my gempak distribution, but when we tried to
>> >use it no gempak files were created.
>> >
>> >This is the output:
>> >
>> > dcreanal z.cdf zwind.gem
>> > Changing center table to cntrgrib1.tbl
>> > Changing vertical coord table to vcrdgrib1.tbl
>> > Changing WMO parameter table to ncarncep1.tbl
>> > Changing center parameter table to ncarncep1.tbl
>> >---------------------------------------------------
>> >We checked the netcdf file with ncdump and it is fine.
>> >
>> >What do you think is going wrong?
>> >
>> >By the way I have compiled dcreanal without the udunits lib, because
>> >there were not available on our system.  Could this create some problem?
>> >
>> >Thanks again
>> >
>> >Lodovica
>> >.
>> >
>> >
>> >On Wed, 14 Mar 2001, Unidata Support wrote:
>> >
>> >> >From: Lodovica Illari <address@hidden>
>> >> >Organization: UCAR/Unidata
>> >> >Keywords: 200103141850.f2EIoYL09674
>> >> 
>> >> >Hi,
>> >> >
>> >> >I have some reanalysis data in netCDF format.  I would like to make a
>> >> >gempak file from that? We are running gempak5.4 and we don't have
>> >> >dcreanal. Can you give us a patch? 
>> >> >
>> >> >Thanks for your help
>> >> >
>> >> >Lodovica Illari
>> >> >
>> >> 
>> >> 
>> >> Lodovica,
>> >> 
>> >> There is a source tarfile for 5.4 in the gbuddy ftp account:
>> >> ftp to ftp.unidata.ucar.edu
>> >> login: gbuddy
>> >> password: XXXXXX
>> >> 
>> >> cd nawips-5.4/contrib
>> >> get dcreanal.tar.Z
>> >> 
>> >> 
>> >> The installation instructions are in the archives at:
>> >> http://www.unidata.ucar.edu/projects/coohl/mhonarc/MailArchives/gempak/ms
> g02
>> > 840.html
>> >> 
>> >> You will need to have NetCDF library installed on your system and make su
> re 
>> > that
>> >> the NETCDF and NETCDFINC definitions in Makeinc.common are set appropriat
> ely
>> > .
>> >> 
>> >> 
>> >> The dcreanal program does not know how to deal with the reanalysis data
>> >> on Gaussian grids.
>> >> 
>> >> Since there can be many times in a NetCDF file, and you are limited to
>> >> 1000 times in a grid file, I generally suggest monthly GEMPAK output
>> >> since 365 days x 4 analysis per day exceeds 1000.
>> >> 
>> >> Execution example:
>> >> dcreanal slp.1948.nc YYYYMM_slp.gem
>> >> 
>> >> The above reads from the NetCDF file slp.1948.nc and writes to monthly
>> >> Gempak files using the YYYYMM template.
>> >> 
>> >> Steve Chiswell
>> >> *************************************************************************
> ***
>> >> Unidata User Support                                    UCAR Unidata Prog
> ram
>> >> (303)497-8644                                                  P.O. Box 3
> 000
>> >> address@hidden                                   Boulder, CO 80
> 307
>> >> -------------------------------------------------------------------------
> ---
>> >> Unidata WWW Service                        http://www.unidata.ucar.edu/  
>    
>> >> *************************************************************************
> ***
>> >> 
>> >
>> 
>> ****************************************************************************
>> Unidata User Support                                    UCAR Unidata Program
>> (303)497-8644                                                  P.O. Box 3000
>> address@hidden                                   Boulder, CO 80307
>> ----------------------------------------------------------------------------
>> Unidata WWW Service                        http://www.unidata.ucar.edu/     
>> ****************************************************************************
>> 
>