[PROJ] Preferred grid format for transformations?

[Kristian Evers]

> 
> For 3D (time,long,lat) indexing, I'd assume the number of values along the 
> time axis not to be too large (that would be great to have some feedback for 
> Brian, Chris or other potential data producers on that). In that case, one 
> TIFF IFD per time slice should be able to address that.
> 

An interesting case here is transformations between ITRFyyyy and the Greenlandic
GR96 reference frame. Due to the mostly melting nature of the ice sheet Greenland
is subject to large amount of elastic deformation. This deformation is not predictable
from year to year, so in one year you may see deformation in the up component of
+5 cm and the next year it may be -1 cm. To capture this fully we need a time-varying
deformation model. Currently we are producing a set of deformation grids (x, y and z
components) per year since 1996. Returning to the question above, that gives us
23 values (and rising) on the time axis. Whether that is too large or not I don’t know.

/Kristian