Re: netcdf output: defining co-ordinate arrays for multiple dimensions

Hi Chris,
Ideally you will want to follow the CF conventions for creating your NetCDF file. Unfortunately the regular CF convention website is down right now so I cannot point you to it.
But you can get it at this URL: https://github.com/cf-convention/cf-documents/blob/master/cf-conventions/1.6/cf-conventions.pdf?raw=true

The section on 2-D lat/lon coordinates gives this example:

Example 5.2. Two-dimensional coordinate variables

dimensions:
  xc = 128 ;
  yc = 64 ;
  lev = 18 ;

variables:

  float T(lev,yc,xc) ;
    T:long_name = "temperature" ;
    T:units = "K" ;
    T:coordinates = "lon lat" ;

  float xc(xc) ;
    xc:axis = "X" ;
    xc:long_name = "x-coordinate in Cartesian system" ;
    xc:units = "m" ;

  float yc(yc) ;
    yc:axis = "Y" ;
    yc:long_name = "y-coordinate in Cartesian system" ;
    yc:units = "m" ;

  float lev(lev) ;
    lev:long_name = "pressure level" ;
    lev:units = "hPa" ;

  float lon(yc,xc) ;
    lon:long_name = "longitude" ;
    lon:units = "degrees_east" ;

  float lat(yc,xc) ;
    lat:long_name = "latitude" ;
    lat:units = "degrees_north" ;

=======

The data variable T and the lat and lon variables have the same 2 rightmost dimensions which are often just index dimensions without a coordinate
variable. In this example cartesian 1D coordinates are supplied in meters, but they are not necessary for CF-compliance. The main point here is that a "coordinates"
attribute is attached to the data variable that points to the lat and lon variables used as coordinates.

Hope this helps,
 -dave

On Apr 1, 2014, at 7:47 AM, Christopher Steele (ENV) <Christopher.Steele@uea.ac.uk> wrote:

> Hi everyone,
>
> I'm trying to create an output in netcdf for variables which are on a Lambert Conformal projection. The trouble is, from looking at the documentation, I'm not sure how to define a variable so that it has coordinates lat and lon only (where lat and lon are 2D). Does anyone have any ideas as to how I go about achieving this?
>
> cheers
>
> Chris
>
>
>
>
> Christopher Steele,
> Academic Researcher,
> Centre for Ocean and Atmospheric Sciences,
> School of Environmental Sciences,
> University of East Anglia,
> Norwich,
> _______________________________________________
> ncl-talk mailing list
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/mailman/listinfo/ncl-talk

_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Tue Apr 1 16:35:09 2014