[PROJ] Could you please explain me why I can't project RDN2008 (EPSG:6706) to ERTF2000 (EPSG:9067) or ITRF2000 (EPGS:8997)

[Giacomo Cappellini]

Thanks Even, thanks Greg.

I now understand why there are no proposed transformations from RDN2008 to
ETRF2000 or ITRF2000.
To get a better picture on the topic I downloaded the EPSG database and by
grepping the available transformations starting from RDN2008 I've confirmed
that "RDN2008 to ETRS89 (1)" and "RDN2008 to WGS 84 (1)" are the only
available transformations according to EPSG authority.

The only point I miss from the picture is the meaning of "ENSEMBLE". While
I can generally grasp how to move from a source Reference Frame to a target
Reference Frame, I fail to picture how to move from a Reference Frame to a
Reference System ensemble. I'd like to ask if possible a little more
details about the meaning of "we actually do that manually for the WGS 84
datum ensemble vs its realizations". The information I've reached so far
about RS ensemble says that "all but the highest accuracy requirements may
be considered to be insignificantly different from each other" and
"Datasets referenced to the various realizations may be merged without
change of coordinates":
- https://docs.opengeospatial.org/as/18-005r4/18-005r4.html#87
- https://docs.opengeospatial.org/as/18-005r4/18-005r4.html#53

Thanks,

G.C.

Il giorno mar 19 lug 2022 alle ore 13:48 Greg Troxel <gdt at lexort.com> ha
scritto:

>
> Even Rouault <even.rouault at spatialys.com> writes:
>
> > For RDN2008, there are 2 transformations available:
> >
> > - one from RDN2008 to ETRS89 (EPSG:4258), with a null Helmert
> > transformation and a accuracy of 0 m. This is consistant with how you
> > described RDN2008, and this transformation has the comment "RDN2008 is
> > the second Italian realization of ETRS89"
>
> A far more lengthy way forward might be:
>
>   Italian geodetic authorities (whoever defined RDN2008) publish a
>   transformation from RDN2008 to some modern ETRFyyyy, including an
>   error estimate.
>
>   They submit that to EPSG for inclusion.
>
>   A new EPSG version is published.
>
>   A proj release arrives with the new EPSG version.
>
>   Meanwhile, you wait for the previous 4 steps.
>
> I realize it doesn't solve your problem now, but authoritative
> transforms with tight error estimates shorten the transform pipelines
> that are used and lead to better answers.  The above avoids ETRS89.
>
> > Afaik, RDN2008 is ETRS89 realization ETRF2000 epoch 2008.0.
>
> If you believe that precisely, you could just relabel your data as being
> ETRF2000 (assuming the standard epoch is 2008.0).
>
> Or insert a transform similar to Even's sugggestion, perhaps with better
> accuracy.  I would expect RDN2008/ETRF2000:2008.0 to be at the 0.01m
> level -- but I am pretty clueless about ETRF details.
>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/proj/attachments/20220719/663afa1e/attachment.htm>