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

Re: [Fwd: Question about Coordinate System]



Hi, John,
 
I forwarded the idea as you suggested. Looking forward to the future discussions.
Thanks for the prompt reply.
 
Cheers
Shawn

 
On 06/02/07, John Caron <address@hidden> wrote:
Hi Shawn:

John Caron wrote:
>
>
> -------- Original Message --------
> Subject: Question about Coordinate System
> Date: Tue, 6 Feb 2007 16:03:36 +0100
> From: Hsiang Yun Chen <address@hidden>
> Reply-To: Hsiang Yun Chen <address@hidden >
> Organization: UCAR/Unidata
> To: <address@hidden>
>
> Hi, Listers.
>
> I have a question of coordinate system after going through available
> documents. Correct me if I am wrong.
> Accodring to the CF-Convnetions, parameters such as projection
> approaches (eg. Lambert_Conformal_projection) are specified. However, I
> find no information about the DATUM.  Although it is assumed to be
> WGS84, there is possiblity for data in other datum such as  Bessel-1841
> specified in RT90. My question is, why not make the DATUM as a standard
> parameter? With all the effort being made by ncML, the inconsistency
> with other correctly geo-referencing dataset is still possible due to
> the lack of DATUM parameter. Am i naive or missing/misunderstanding
> something on the issue?

this is definitely a shortcoming in CF-Conventions. Would you like to post your comments to that mailling list? you have to join first: http://www.cgd.ucar.edu/mailman/listinfo/cf-metadata

>
> ps. some ideas about ncML and netCDF, please correct me if I am wrong
> One crucial point is that the ncML is the metadata of netCDF files.
> Information in ncML is orginated from netCDF. It's true that users can
> edit ncML to make it more complete, however, it still depends on the
> header of netCDF.  Suppose, a user trying to create a data in his
> national coordinate system (eg. RT90) using CF-Conventions.  The user
> specified the projections and relevant parameters. It seems fine at
> first. However, when the data is distributed and people who download the
> data can't be able to know that the datum is Bessel-1841 which specified
> in RT90. The system would transfer the coordinate system using WGS84 as
> datum.  That, eventually, would lead to inconsistency with other data.


You can add/change metadata using NcML. The Java library can read the NcML file, which points to the original netcdf file. The C library will eventually be able to do this also. To distribute the modifications as a netcdf file, you will have to rewrite the netcdf file, which can be easily done with the Java library.

John

===============================================================================
To unsubscribe netcdf-java, visit:
http://www.unidata.ucar.edu/mailing-list-delete-form.html
===============================================================================




--
I have a thing about first impressions.