NOTICE: This version of the NSF Unidata web site (archive.unidata.ucar.edu) is no longer being updated.
Current content can be found at unidata.ucar.edu.

Re: [netcdf-java] Longitude unit degrees_west not recognized

By convention, McIDAS datasets all use degrees_west for their longitude values. I think that the netCDF-Java should support this (just like it supports "CF" files without the required lat/lon coordinates when there is a addition to the grid_mapping). I also think it's reasonable for CF to support this since it is a valid udunit for longitude. 

Don

Joe Sirott wrote:

Hi John,
There are a lot of legacy oceanographic in-situ datasets out there that use the EPIC netCDF convention <http://www.epic.noaa.gov/epic/document/convention.htm> which in turn recommends use of "degree_west" (note the missing "s") for the longitude units attribute.I've found that you need support the 8 combinations (degree_west, degrees_west, ...) of lat/lon unit attributes in order to reliably handle these kind of datasets. In practice, EPIC datasets are even more complicated as there is an "epic_code" attribute that independently specifies the units for the variable. Some datasets only have the epic_code attribute, some have both, and some (as yet another proof of Murphy's Law) have an epic_code attribute that conflicts with the units attribute.
It would be great (at least for me!) if netCDF-java supported all of these attributes. However, I don't think they should be added to CF, as this would make CF even more complicated than it already is and I think that most data providers should be encouraged to use the degrees_east/degrees_north convention. 

- Joe

John Caron wrote:

Tom Whittaker wrote:

We have some point data that has positive-west longitude values.  It
appears that netcdf-java (and perhaps the CF conventions) do not
support a unit of "degrees_west".  This seems inconsistent, since
degrees_east is supported.  While I assume we can work arround this by
adding a scale_factor attribute with a -1.0 value, I would encourage
you to consider this as a valid unit.  (I also wonder if
"degrees_south" are recognized?  I did not check this.....)

Thanks.....



Hi Tom:
CF does not recognize degrees_west (nor degrees_south). You could propose adding that to CF, which would be the best easy to solve it from our POV.
One thing that complicates it is that degrees_north, degrees_east units is the way to recognize lat/lon coordinates. so now degrees_west, degees_south would also be.
Can you give me some background on what datasets are likely to be using these units? 

_______________________________________________
netcdf-java mailing list
netcdf-java@xxxxxxxxxxxxxxxx
For list information or to unsubscribe, visit: http://www.unidata.ucar.edu/mailing_lists/ 



------------------------------------------------------------------------

_______________________________________________
netcdf-java mailing list
netcdf-java@xxxxxxxxxxxxxxxx
For list information or to unsubscribe, visit: http://www.unidata.ucar.edu/mailing_lists/ 

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