I'm new at this (studying for a masters of GIS) and trying to understand the WGS84 geographic coordinate system as presented in ArcPro

WGS84 is a datum and reference ellipsoid EPSG 4326 is WGS84 using latitute and longitude as a coordinate system The data is have is being represented in ArcPro in WGS84 with lat/lon coordinates, so really EPSG 4326.

However, WGS84 is an ellipsoid, while the representation I see in ArcPro is flat. So the coordinates are being projected in some way. My lecturer assures me that there is no projection. When I use the "measure distance" tool the line between two points is curved since it's a geographic system. I get that, that part is fine. But the map is still flat, so I keep insisting there has to be some sort of projection. If there was no projection I'd be looking at an ellipsoid. What am I missing?

  • There have been similar questions here, e.g. gis.stackexchange.com/questions/387517/… - also search for plate carree projection on this site and see: desktop.arcgis.com/en/arcmap/latest/map/projections/… – Babel Mar 18 at 9:15
  • From Gabor Farkas: Practical GIS: "Technically, EPSG:4326 is the QGS84 ellipsoid (or datum) alone as a geodetic CRS. However, as every ellipsoid is mapped with a linear projection when used alone, the fact that EPSG:4326 uses a Plate Carrée projection still stands.", see books.google.ch/books?id=Qng5DwAAQBAJ&pg=PA90 – Babel Mar 18 at 9:22
  • 2
    There is no projection, not in the stricter sense of the term; there is no mathematical procedure involved to transform geographic coordinates into a planar reference here; they are literally only put into a coordinate grid where the numbers dictate them to be (which is congruent to the Plate Carrée 'projection'). This is purely presentational, while internally ArcPro is aware you are working on geographic coordinates. – geozelot Mar 18 at 9:28
  • As per comment on answer below I understand that what I'm seeing is a mapping from WGS84 to a rectagular grid, which is equivalent to Plate Caree, but not actually a projection. – GlenS Mar 18 at 20:59

Do those lat-lon coordinates that you see make a perfectly square degree grid everywhere? In other words, is 1 degree of longitude (E-W) the same visual size as 1 degree of latitude (N-S) at the equator and at the poles? Then the projection is "Plate Caree" (aka "equirectangular" https://en.wikipedia.org/wiki/Equirectangular_projection). The EPSG:4326 coordinates are "projected" to Equirectangular by simply scaling the value in degrees up or down by a constant to fit the size of your map canvas.

If you look at Google Maps (in 2d mode, switch off the 3d sphere view) or OpenStreetMap and zoom out you'll see the lat-long boxes aren't square because the data is being shown in a Web Mercator projection. If your ArcPro map looks more like those, then ArcPro is projecting your data from its stored EPSG:4326 Coordinate System to EPSG:3857 Web Mercator on the screen. This is a non-constant stretching of the latitude - with the poles getting distorted and stretched so much that you can't show the N or S pole in Web Mercator - they're at +/-Infinity.

  • Okay, so what I'm hearing (from your answer and comments above) is that there is a one-to-one mapping from lat/lon coords on the reference ellipsoid to lat/lon coords on the rectangular grid on the screen and that this does not count as a projection, though it is equivalent to Plate Caree. I need to do more reading. Thanks for the answer! – GlenS Mar 18 at 20:57
  • I'd say it definitely is a projection. There are equations that turn your 3d lat, long, height (where the height is going to be a constant of the earth radius) to X-Y coords on your screen. That's a map projection. – Spacedman Mar 18 at 21:35
  • Yeah that's what I think! Your quotation marks around "projection" confused me a bit. There's a reduction in the number of axes, which is a projection. What I'm thinking (because people keep trying to tell me WGS84 in ArcPro is not a projection) is that there's a difference between a general mathematical projection and a geographic/cartographic projection, which must be a subset of the general idea of projection. I just can't put my finger on what exactly it is yet. – GlenS Mar 19 at 20:18

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.