Inconsistent usage of false_easting and false_northing in grid mappings definitions and in examples #212
Description
Summary
Title
Inconsistent usage of false_easting and false_northing in grid mappings definitions and in examples
Moderator
Moderator status review
Explicit support from
No objections expressed in this discussion.
SciTools/iris#3626 notes that the current inconsistency sometimes leads to problems; @erget has proposed that solving the ambiguity would at least make the intent of the Conventions clear. Impact: Software would need to be aware of the defaults.
Requirement Summary
Definitions of grid mappings / projections and their usage in examples should be consistent with respect to required attributes.
Technical Proposal Summary
As in cf-convention/discuss#225.
Add false_easting and false_northing to example 5.7 or make them optional in the definition of all projections where applicable (e.g. Lambert Conformal Conic projection).
Benefits
Consistency
Status Quo
The attributes false_easting and false_northing are required in the grid mapping variable for a Lambert Conformal Conic projection. Example 5.7 does not use false_easting and false_northing.
Detailed Proposal
Either false_easting and false_northing should be made optional in the definition of the projections (assumed to be zero if not set) or they should be added to example 5.7. I would prefer the first option as it is already the case for the attribute north_pole_grid_longitude in the grid mapping for rotated poles:
north_pole_grid_longitude- This parameter is option (default is 0).
Hence, I would suggest to replace
false_easting
false_northing
in all grid mapping definitions by
false_easting- This parameter is option (default is 0).
false_northing- This parameter is option (default is 0).