Update annex-d-geopose-frame-specs.adoc per comments from @rouault

[3DXScape]

@rouault Please review this.

[rouault]

Looking closer, there are 2 things I'm wondering for the NED CRS:

• do the abbreviations U, V, W still apply when it is a NED coordinate system. The U, V, W are nominally designed for the ENU coordinate system. It has never been clear to me how the logic of attributing abbreviations worked.
• I'm not sure if the "Geographic/topocentric conversions" EPSG:9836 is expected to be used with a NED coordinate system. That implies that implementations of the conversion should look at the declared axis order to do the easting/northing axis swap and invert the sign of the height to be a depth where needed. Generally in the EPSG dataset, when there are projections that require "unusual" axis order, like Transverse Mercator South Oriented, they have their own code.

You may want to ask about that on https://lists.ogc.org/mailman/listinfo/coordtran.wg

[3DXScape]

• I think that U, V, W are Cartesian coordinate placeholders that imply that there is some transformation from X,Y,Z involved. But that’s just a feeling based on my physics background. I had the same question for ENU and decided that naming the axes “east”, “north”, and “up” was the clearest editorial choice. I followed U, V, W for NED to follow your contribution. I have seen (X,Y,Z), (xNorth, yNorth, zDown), (N, E, D), (U, V, W). In the end, these are just parameters with arbitrary placeholder names. If I controlled the Universe, I would use (north, east, down).
• I think that 9836 is probably not designed for NED. I could not find a better fit that was not “EPSG:9837 with the third axis reversed”. Maybe one way to approach this would be in the WKT we are discussing here, but I am such a non-expert with WKT-speak that I do not know how to say “use EPSG:9837 transformation but reverse the third coordinate”. I am hoping that you or one of your colleagues have a suggestion. …steve

From: Even Rouault @.> Sent: Wednesday, June 5, 2024 10:31 AM To: opengeospatial/GeoPose @.> Cc: Carl Smyth @.>; Author @.> Subject: Re: [opengeospatial/GeoPose] Update annex-d-geopose-frame-specs.adoc per comments from @rouault (PR #84)

Looking closer, there are 2 things I'm wondering for the NED CRS:

• do the abbreviations U, V, W still apply when it is a NED coordinate system. The U, V, W are nominally designed for the ENU coordinate system. It has never been clear to me how the logic of attributing abbreviations worked.
• I'm not sure if the "Geographic/topocentric conversions" EPSG:9836 is expected to be used with a NED coordinate system. That implies that implementations of the conversion should look at the declared axis order to do the easting/northing axis swap and invert the sign of the height to be a depth where needed. Generally in the EPSG dataset, when there are projections that require "unusual" axis order, like Transverse Mercator South Oriented, they have their own code.

You may want to ask about that on https://lists.ogc.org/mailman/listinfo/coordtran.wg

— Reply to this email directly, view it on GitHubhttps://github.com/opengeospatial/GeoPose/pull/84#issuecomment-2150593997, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ABAQF3U4GCMR3LL26WQCQ6LZF5DODAVCNFSM6AAAAABI3E2GZWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDCNJQGU4TGOJZG4. You are receiving this because you authored the thread.Message ID: @.**@.>>

[3DXScape]

Actually “Use EPSG:9837, swap the first two axes and reverse the third axis”. …

From: Carl Smyth Sent: Wednesday, June 5, 2024 11:32 AM To: opengeospatial/GeoPose @.>; opengeospatial/GeoPose @.> Cc: Author @.***> Subject: RE: [opengeospatial/GeoPose] Update annex-d-geopose-frame-specs.adoc per comments from @rouault (PR #84)

• I think that U, V, W are Cartesian coordinate placeholders that imply that there is some transformation from X,Y,Z involved. But that’s just a feeling based on my physics background. I had the same question for ENU and decided that naming the axes “east”, “north”, and “up” was the clearest editorial choice. I followed U, V, W for NED to follow your contribution. I have seen (X,Y,Z), (xNorth, yNorth, zDown), (N, E, D), (U, V, W). In the end, these are just parameters with arbitrary placeholder names. If I controlled the Universe, I would use (north, east, down).
• I think that 9836 is probably not designed for NED. I could not find a better fit that was not “EPSG:9837 with the third axis reversed”. Maybe one way to approach this would be in the WKT we are discussing here, but I am such a non-expert with WKT-speak that I do not know how to say “use EPSG:9837 transformation but reverse the third coordinate”. I am hoping that you or one of your colleagues have a suggestion. …steve

From: Even Rouault @.**@.>> Sent: Wednesday, June 5, 2024 10:31 AM To: opengeospatial/GeoPose @.**@.>> Cc: Carl Smyth @.**@.>>; Author @.**@.>> Subject: Re: [opengeospatial/GeoPose] Update annex-d-geopose-frame-specs.adoc per comments from @rouault (PR #84)

Looking closer, there are 2 things I'm wondering for the NED CRS:

• do the abbreviations U, V, W still apply when it is a NED coordinate system. The U, V, W are nominally designed for the ENU coordinate system. It has never been clear to me how the logic of attributing abbreviations worked.
• I'm not sure if the "Geographic/topocentric conversions" EPSG:9836 is expected to be used with a NED coordinate system. That implies that implementations of the conversion should look at the declared axis order to do the easting/northing axis swap and invert the sign of the height to be a depth where needed. Generally in the EPSG dataset, when there are projections that require "unusual" axis order, like Transverse Mercator South Oriented, they have their own code.

You may want to ask about that on https://lists.ogc.org/mailman/listinfo/coordtran.wg

— Reply to this email directly, view it on GitHubhttps://github.com/opengeospatial/GeoPose/pull/84#issuecomment-2150593997, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ABAQF3U4GCMR3LL26WQCQ6LZF5DODAVCNFSM6AAAAABI3E2GZWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDCNJQGU4TGOJZG4. You are receiving this because you authored the thread.Message ID: @.**@.>>