rggibb
commented
3 years ago Roger Lott has a similar query: "It struck me that you had to be one of the cognoscenti to understand what is going on. In principle I find the notion of utilising DGGS technology on a CRS that I understand to be quite appealing, but in what way does it differ from something like the web Mercator tiling beloved by the WMTS community? An explanation of this in an informative annex would I feel be useful (as would fleshing out the existing Annex which gives some insight but lacks clarity on what it is all about."
I read through the description of each abstract tests, and at first glance it seems that many Two Dimensional Tile Matrix Sets, i.e. those based on geographic/geodetic CRS with progressive partitioning of tiles/cells (whether using the variable width conformance class or not) satisfy the requirements of Axis-Aligned DGGS. At least those three examples defined within the 2DTMS document itself appear to qualify:
The latter two (using Variable Widths) actually might currently be excluded if the AxisRatio needs to be constant, but I believe this really should be addressed by #7. They are actually closer to being a full DGGS because they better approximate equal area.
It would be great to clarify the relationship between the two standards, and maybe mention and/or include a reference as examples of Axis-Aligned DGGS?
In particular, please take a look at Annex E where we include a reference to DGGS Part 1 and this upcoming Part 4
As well, the new Annex K discussing extension to additional dimensions might also be of interest.
It also seems that the GEOSOT AA DGGS could easily be described as a TileMatrixSet (and its 3D version as well using the extension). Providing this description in the JSON encoding defined by the 2DTMS standard would be very useful and facilitate interoperability with GeoPackage and OGC API - Tiles for example.