[IDV #RGF-974040]: Re: Reading ECMWF grib files

Hi Xavier-

> Thanks for your quick response. My answers below:
> 
> On 12/14/06, Unidata Support < address@hidden> wrote:
> 
> > Hi Xavier-
> >
> > For future reference, you need to be subscribed to the idvusers list
> > in order to post to it.  You can subscribe from the IDV homepage:
> >
> > http://www.unidata.ucar.edu/software/idv
> 
> 
> 
> Sorry, I thought I sent this message to the mailing list.
> Is this so?
> I believe I am already subscribed to idv users and idv developers.

Sorry, I should have checked into this further before responding.
You are subscribed as address@hidden, but your message
came from address@hidden.  I just tried a message from
my gmail account and it came through as gmail.com, so I'm not sure
where the googlemail.com came in.

I would probably have been the one to respond anyway, so no worries.

> > What version of the IDV are you using?  What are the errors?
> > With version 2.0 or later, it should give you a warning that
> > it doesn't know how to deal with the vertical coordinate (sigma),
> > but still should plot the data.
> 
> 
> Sorry for not being more explicit. version 2.0.
> After selecting 3D grid temperature and selecting temperature
> and hitting "create display" with contour plan view,
> it says in an info box "Unknown vertical coordinate Unit,
> using linear scale to Altitude" and after clickin OK I get
> an Error message: "An error has occurred:
> getFlatField read got IOException
> java.io.IOException" and after clicking OK I get
> "Unable to load field: Temperature". These last two
> messages are repeated again and then nothing happens.
> On the command line I get
> Exception: GeoGrid.getdataSlice() on dataset Temperature
> Exception: GeoGrid.getdataSlice() on dataset Temperature
> Again two times as shown above.

This might be related to a bug that I discovered after the
2.0 release.  If you want to try using the latest release
candidate (2.1RC2), you can get it from the IDV downloads
site:

http://www.unidata.ucar.edu/downloads/idv

to see if that fixed the problem.  It will still not handle
the vertical scale other than to map it directly to the vertical
scale.  You could then set your vertical scale to 0-1 to view
the sigma levels.

> 
> At present, the IDV only plots data with pressure or height
> > coordinates at the correct locations in the map box.  The vertical
> > scale on the box is altitude.  There is no convention for relating
> > sigma to height in the GRIB files.  For data in netCDF files,
> > we use the CF conventions to relate sigma to height.
> 
> 
> 
> What are the CF conventions? I do have the formula
> conversions, how easy would it be to put them in the
> IDV code? I am not familiar with the IDV source code yet.

CF is a convention for netCDF data:

http://www.cgd.ucar.edu/cms/eaton/cf-metadata/CF-1.0.html

which supports adding metadata to the file to describe the
formula.

I'll have to think about whether you could use the formula
system in IDV to do the conversion manually.  Do you have
a sample file that I could test with?

Don

Ticket Details
===================
Ticket ID: RGF-974040
Department: Support IDV
Priority: Normal
Status: Open