[PROJ] [EXTERNAL] Re: Proj 6 API questions

Kristian Evers kreve at sdfe.dk
Wed Mar 27 04:23:15 PDT 2019


A further point to consider, ITRF2014 is a Cartesian coordinate system and
has no knowledge of the shape of the earth. It is defined by the mass center
of the Earth and not much else is related to the planet we live on - it is a
purely geometric system. You can use any ellipsoid you like to convert
ITRF coordinates to geodetic coordinates. The GRS80 ellipsoid, however,
is recommended by the IERS.

Just to clarify, the thing QGIS is doing is purely a mathematical exercise:
Conversion between different coordinate systems and determining
distances and areas between known points. This can be done without
loss of accuracy (well, except for miniscule floating point rounding errors
in the computer). The points are moved from projected space to the ellipsoid
to avoid distortions in the projection.

When datum shifts are involved we are in the realm of physics where we
have to take a whole range of other errors into account. This should be avoided
in this case.

/Kristian


-----Oprindelig meddelelse-----
Fra: PROJ <proj-bounces at lists.osgeo.org> På vegne af Kristian Evers
Sendt: 27. marts 2019 12:09
Til: Lesparre, Jochem <Jochem.Lesparre at kadaster.nl>; Nyall Dawson <nyall.dawson at gmail.com>; Even Rouault <even.rouault at spatialys.com>
Cc: PROJ <proj at lists.osgeo.org>
Emne: Re: [PROJ] [EXTERNAL] Re: Proj 6 API questions

This approach assumes that a sufficiently accurate transformation between the legacy CRS
and the modern CRS exists. That cannot be guaranteed and it most cases I would argue
that such a transformation comes with a fair amount of uncertainty. Not doing a datum shift
and simply working within the original datum avoids the introduction of errors and ensures
consistency with the coordinates as they were originally measured.

I am most familiar with Danish systems and for those I would not be comfortable adding a datum
shift into the mix. The transformations are simply not accurate enough across the entirety of the
country (near control points everything is good, between them... who knows?).

/Kristian

-----Oprindelig meddelelse-----
Fra: Lesparre, Jochem <Jochem.Lesparre at kadaster.nl> 
Sendt: 27. marts 2019 11:54
Til: Kristian Evers <kreve at sdfe.dk>; Nyall Dawson <nyall.dawson at gmail.com>; Even Rouault <even.rouault at spatialys.com>
Cc: PROJ <proj at lists.osgeo.org>
Emne: RE: [PROJ] [EXTERNAL] Re: Proj 6 API questions

Kristian is right that just using a modern ellipsoid for distances/areas in an old but still used national CRS would give errors.
However, I think computing distances/areas using the old ellipsoid of that CRS is also not correct for most users. Therefore, I suggest:

1. Back-project the projected coordinates to geodetic coordinates using the Ellipsoid parameters of the CRS datum.
2. Perform datum transformation to the most recent realisation of the currently accepted scientific CRS of that planet (currently ITRF2014 for Earth) with its ellipsoid (currently GRS80 for Earth).
3. Perform geodesic calculations using this ellipsoid parameters of this CRS datum.

Kind regards, Jochem



Disclaimer:
De inhoud van dit bericht is uitsluitend bestemd voor geadresseerde.
Gebruik van de inhoud van dit bericht door anderen zonder toestemming van het Kadaster 
is onrechtmatig. Mocht dit bericht ten onrechte bij u terecht komen, dan verzoeken wij u 
dit direct te melden aan de verzender en het bericht te vernietigen. 
Aan de inhoud van dit bericht kunnen geen rechten worden ontleend.

Disclaimer:
The content of this message is meant to be received by the addressee only.
Use of the content of this message by anyone other than the addressee without the consent 
of the Kadaster is unlawful. If you have received this message, but are not the addressee, 
please contact the sender immediately and destroy the message.
No rights can be derived from the content of this message
_______________________________________________
PROJ mailing list
PROJ at lists.osgeo.org
https://lists.osgeo.org/mailman/listinfo/proj


More information about the PROJ mailing list