Re: WRF grib2 data

This is a multipart message in MIME format.
--=_alternative 000CD081852572D9_
Content-Type: text/plain; charset="US-ASCII"

Don,
Thanks for your reply. If I understand correctly, following the NcmL 
example, I could create the file manually to combine the separate files 
into one and use IDV to read it in. 
What is not clear from your comments is whether this will or will not work 
with netCDF, if not will it work with grib files? Not having read the 
updates, is this only possible in version 2.2?
Thanks

Anthony P. Praino
ACTC/Deep Computing Systems
IBM T.J. Watson Research Center
914-945-1953 (office) 
apraino@xxxxxxxxxx
http://www.research.ibm.com/weather



Don Murray <dmurray@xxxxxxxxxxxxxxxx> 
05/11/2007 05:53 PM
Please respond to
dmurray@xxxxxxxxxxxxxxxx


To
Anthony P Praino/Watson/IBM@IBMUS
cc
idvusers@xxxxxxxxxxxxxxxx
Subject
Re: WRF grib2 data






Hi Anthony-

Anthony P Praino wrote:
> I assume this may be a newbie question but I have checked the user 
manual, 
> tutorial and scanned archives but have not run across an answer. 

Well, you didn't miss anything.  It's just not there.

How does
> one 'catalog' or group gridded data files together. I am trying to 
create 
> animation sequences of WRF gridded fields from a locally run model. When 
I 
> use the data chooser to select a group of WRF output files which are 
> timestepped files (either netCDF or grib2), I cannot group them in order 

> to create a display based upon a selected field of multiple times. 

The nightly release of the IDV allows you to aggregate WRF
GRIB2 files together from disk or from a remote server.  You can 
download this from:

http://www.unidata.ucar.edu/software/idv/release/nightly

 From disk, use the file chooser and set the Datasource Type to be
"Aggregated Grid files".   This will not work with the WRF netCDF files.
Basically, the grid reader generates a netCDF Markup Language (NcML) 
aggregation file on the fly.  You could also just create an NcML file
manually and read that in.  For information on creating NcML files for
aggregation, see:

http://www.unidata.ucar.edu/software/netcdf/ncml/v2.2/Aggregation.html

The NcML that the Aggregated Grid Data uses looks something like:

<netcdf xmlns="http://www.unidata.ucar.edu/namespaces/netcdf/ncml-2.2";>
<aggregation type="joinExisting" dimName="time">
<netcdf location="file:/I:/rtmodel/07051012_nmm_alt1.GrbF000" />
<netcdf location="file:/I:/rtmodel/07051012_nmm_alt1.GrbF001" />
<netcdf location="file:/I:/rtmodel/07051012_nmm_alt1.GrbF002" />
</aggregation>
</netcdf>

This is still not in the documentation, though. ;-)

I have
> tried this with cataloged data sources and it works fine (what are the 
> details of gridded  catalog sources such as NCEP NAM 12km). 

This has all timesteps in one file.

I have also
> tried to manually create a display out for each time which I can see in 
> the dashboard but the display visual does not change even though the 
files 
> appear to be sequencing.  Any help of pointers to such information is 
> appreciated.

If there is only one time associated with the data, it gets shown on all
times.  So, in this case, each file has one timestep and they all get 
shown no matter what time the animation is on.

We hope to make a full release of version 2.2 on Monday.

Don
*************************************************************
Don Murray                               UCAR Unidata Program
dmurray@xxxxxxxxxxxxxxxx                        P.O. Box 3000
(303) 497-8628                              Boulder, CO 80307
http://www.unidata.ucar.edu/staff/donm
*************************************************************




--=_alternative 000CD081852572D9_