[PROJ] Setting the GEOIDMODEL in the CRS short id

[Greg Troxel]

"Lesparre, Jochem" <Jochem.Lesparre at kadaster.nl> writes:

>> This explains why the Danish height system DVR90 is now in EPSG as a datum ensemble
>
> I'm not sure of the Danish situation, but for the Netherlands the only
> official realisation of the height system is the levelling
> benchmarks. The geoid provides only an approximate
> transformation. Making it two different CRSs is an interesting tweak,
> but I'm afraid it would only create more confusion.

I think the the Netherlands situation is the same as the US (different
orthometric height definitions, different underlying datum for HAE,
different geoid models of course, but structurally the same).  In both
NL and US cases, I think it is incorrect to add a CRS for [orthometric
height estimated from HAE and GEOID X].

(I believe, after off-list conversations with Kristian and trying to
read a report in Danish, that in Denmark, there really are additional
height system realizations.)