Re: [netcdf-java] problem recognising variable

Adit Santokhee wrote:
Hi John,

Thanks, changing the Convention to CF fixes the problem.

Did you also have to add a time unit?


Cheers,
Adit

John Caron wrote:
Adit Santokhee wrote:
Hi,

We recently upgraded ncWMS to add new functionality and for that we are using the java netcdf-4.0.48.4 library. However, we are now having difficulty displaying "wind stress" variable from a certain dataset; the library thinks that it's a vertical axis with units of pressure. Previously, we didn't have this problem with netcdf-4.0.

The main problem is that youve got an unknown convention (GDT 1.3), so it gets handed to the default Convention parser, which is trying to do all sorts of things for all sorts of files. GDT was superceded by CF. Are these legacy files, not actively being written? There is an attribute
  :TimeStamp = "2008-OCT-15 10:29:42 GMT+0100";
If you have a huge stash of these files, we might have to write a Convention parser for them.

You can fix the problem by adding a unit to the time coordinate, not sure what the units are, but NcML like this

<?xml version="1.0" encoding="UTF-8"?>
<netcdf xmlns="http://www.unidata.ucar.edu/namespaces/netcdf/ncml-2.2"; location="C:/data/work/adit/ORCA025-R07_y2007m03d01_icemod.nc">
 <variable name="time_counter">
   <attribute name="units" value="secs since 1970-01-01" />
 </variable>
</netcdf>

fixes the coordinate system, but the time coords look wrong.

Im not sure how this is playing out inside ncWms, i would have assumed that there would be no grids found at all.




  • 2009 messages navigation, sorted by:
    1. Thread
    2. Subject
    3. Author
    4. Date
    5. ↑ Table Of Contents
  • Search the netcdf-java archives: