6

I wonder why the Authority and the SRID are often both used together to refer to a CRS, like in EPSG:4326. Isn't the SRID unique?

Wikipedia says it is a "unique value", but also that spatial vendors may have created "[...] their own SRID implementation or refer to those of an authority, [...]". What does that mean, creating an own SRID implementation?

After browsing through spatialreference.org it seems that there are no duplicate SRIDs for different authorities.

If this is true and SRIDs are truly unique, why the OGC spatial_ref_sys metadata table (see Wikipedia link above; you may find this table in databases like PostGIS) is defined with both, an SRID column as well as an AUTH_SRID column? In a default PostGIS setup (6184 EPSG and 2315 ESRI entries), both columns always have the same value per row. If I wanted to add an user-defined entry, why would I want to choose different values for those columns?

Or if I were an authority, what could be the rationale for a decision that would cause me to define an SRID that has already been defined by another authority?

8
  • 4
    Obviously, the same authority doesn't assign the same identifier (number) to different entities. But different authorities may have, historically. EPSG is just one of them, and the most common. Across all combinations, implementations use SRID identifiers that are unique across their library, wh proj4 being the most commonly used in FOSS GIS applications.
    – geozelot
    Nov 26, 2021 at 23:16
  • The most problematic case would be the new srid. Until the autorities recognize it, you have to create a temporary SRID, so the definitive SRID can be different in the end, and you can have overlaps if someone use the same number at the same time. An example is the SRID 3857 that was the 900913 when created by google (it's google in leetspeak). Google decided its number so people can use it until it was officially added by EPSG, so there potentially was other people creating a projection at the same time with the same number without knowing. Nov 29, 2021 at 15:02
  • The story above is not true, 900913 was not invented by Google but by an open source developer Christopher Schmidt crschmidt.net/blog/archives/243/google-projection-900913.
    – user30184
    Jan 8 at 14:22
  • The story is doubly not true, 900913 was never added to the EPSG registry
    – nmtoken
    May 13 at 11:09
  • @geozelot not that obvious, EPSG does assign the same identifier to different entities
    – nmtoken
    May 13 at 11:11

1 Answer 1

2

EPSG:4326 isn't unique even for EPSG codes.

Within the EPSG Geodetic parameter dataset, EPSG:4326 can either refer to a CRS or an Extent.

For uniqueness you need the full identifier (URI)

Codes 2315 and 6184 are also both used by EPSG to define CRS inter al.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.