New GeoTIFF model CRS for engineering CRS

[desruisseaux]

Pull request #117 proposes the addition of a GTModelTypeGeoKey value for geocentric CRS associated to a two-dimensional spherical coordinate system. This pull request proposes to add an engineering CRS, but also associated to a two-dimensional spherical coordinate system. The definition would be similar to #117 except for the datum:

Section to edit

User-defined Model Coordinate Reference Systems with the addition of the following sub-section:

For a user-defined engineering CRS (spherical 2D) the user is expected to provide:

• coordinate reference system name (through the EngineeringCitationGeoKey);
• engineering datum through the EngineeringDatumGeoKey:
  • user-defined engineering datum name and other defining information:
    • the geodetic datum name (through the EngineeringCitationGeoKey);
  • axis unit through the GeogAngularUnitsGeoKey, either:
    • angle unit code (if available through standard EPSG code), or
    • user-defined angle unit name (through the EngineeringCitationGeoKey) and scaling from SI base unit of radian (through the GeogAngularUnitSizeGeoKey).

Note: GeogAngularUnitsGeoKey is not an appropriate name. Maybe this key should be renamed only AngularUnitGeoKey. Such renaming would not be a compatibility break if the numerical value does not change.

New GeoKeys

• EngineeringCitationGeoKey
• EngineeringDatumGeoKey

Open questions

• Should we support a spherical 3D case?
• Should we support a Cartesian 3D case? That one seems hard to implement because GeoTIFF images are 2D. For the shperical case, we can said that the third-dimension (the radius) is the pixel value. For the Cartesian case there is no obvious axis to use as pixel values.