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

Re: 20050727: IDV - Error trying to plot local GRIB file 2-D and 3-D f ields (Mac OS X 10.3.9) (cont.)

Hi Dave,
Well, it's not a hardware problem. This is a thin grid (the x dimension decreases to 1 as the latitude increases) and our software does not support direct access of the GRIB file. However, you can run the gribtonc decoder (http://my.unidata.ucar.edu/content/software/decoders/) to convert these to NetCDF files that the IDV can then read.
As I understand it, as things move to GRIB2 there will no longer be these thin grids to deal with so we have no plans to support thin grids in their raw GRIB format. 

Thanks,
Doug

Dave Dempsey wrote:

On Thu, 28 Jul 2005, Doug Lindholm wrote:



I wasn't able to access your data file. I got a "ERROR 404: Not Found".
Could you double check that link?



Doug,

Sorry about that; I created a link to the file from the wrong server's
home directory. The link should be:

   http://virga.sfsu.edu/GRIB/05072712_gfsK.grb



I have been running the IDV on my PowerBook G4 with great success so I'd
say there is hope. I'll confirm whether your file works on my PowerBook
and point you to our latest release candidate if that appears to solve
the problem.



I just did some more tests, and I was able to plot the following:

   -- Level III radar data;
   -- a Mollweide composite IR image;
   -- METAR point data;
   -- a skew-T plot of RAOB data;
   -- a color-filled contour plot of mean sea-level pressure from the
latest NAM run from Unidata's Model Data catalog (presumably in NetCDF);
   -- a contour plot of sea level pressure from a NAM forecast, in GRIB
format (a local file).

However, I once again couldn't plot mean sea level pressure from a local
GFS model output GRIB file like the one referred to above (or several
other, similar GFS model output GRIB file that I tried).

We seem to be narrowing the problem to at least one type of GFS GRIB files
that we've got locally via IDD.  There files contain data on hemispheric
octants, by the way, and the border between two of the octants (labelled
with L and K in the file names, respectively; the file referred to above
is a "K" file, which covers the eastern Pacific and western U.S.) splits
the US pretty much in half--I wonder if that's part of the problem?

I might also add that I'm viewing this on a LINUX PC, connected to a Mac
G4 Xserve via VNC, which in turn is connected to the Mac G4 laptop running
the IDV via Apple Remote Desktop (which I think is a fancy wrapper for
VNC). Doing this make for some curious but apparently harmless graphical
behaviors. Most notably, when I make some button selections in IDV, such
as "Add Source", the cursor becomes a column of half a dozen or more
distinct little blue globes with a black arrow (the cursor) sticking out
of each one. Like the cursor itself, this column of cursor clones moves
around as I move the mouse. (Couldn't get a screen capture of it because I
had to move the cursor out of the vncviewer window to start the screen
capture.)

-- Dave