Plausible geographical world map

[chateauferret]

I quite like the results that some of these flat-world tectonics programs produce, like Nortantis and WorldEngine, but they do fall down because they operate on a finite flat plane instead of a sphere - or rather, since they roll over opposite pairs of edges, they work on a torus. It occurs to me that one way of partly dealing with this is to consider a spherical cube, i.e. a cube whose six faces are projected onto the surface of a sphere, and to generate a "world" which is four times as wide as it is high; e.g. 4096 x 1024. Map that to four consecutive faces of the cubic sphere, either around the equator or around both poles; you are then in the business of merely filling in the other two faces, one of which will possibly be ocean; you might also have to move some bits around so that the bits that wrapped around over the long edges of the map are in sensible places.

[Maxicus]

I quite like the results that some of these flat-world tectonics programs produce, like Nortantis and WorldEngine, but they do fall down because they operate on a finite flat plane instead of a sphere - or rather, since they roll over opposite pairs of edges, they work on a torus. It occurs to me that one way of partly dealing with this is to consider a spherical cube, i.e. a cube whose six faces are projected onto the surface of a sphere, and to generate a "world" which is four times as wide as it is high; e.g. 4096 x 1024. Map that to four consecutive faces of the cubic sphere, either around the equator or around both poles; you are then in the business of merely filling in the other two faces, one of which will possibly be ocean; you might also have to move some bits around so that the bits that wrapped around over the long edges of the map are in sensible places.

That is actually a very good method to try out.

[chateauferret]

That is actually a very good method to try out.

Well I tried it out and it didn't work very well

It looks as if Platec (which produces the tectonics WorldEngine uses) doesn't like it if the grid isn't square. It generates a "world" on the torus as you expect, but the grid seems to be stretched to fit into the dimensions you give it. Now the 4096 x 1024 world isn't just a stretched version of the square (1024 x 1024) one, but it looks stretched in the x direction; the coastlines are quite angular and square, and there aren't enough mountains.

Now WorldEngine does a nice job of enhancing the Platec tectonics into a more mature map using fractal and erosion algorithms, and the square output looks mostly very nice, you might need to tinker with some of the island chains and things to make them a bit less square and so on. But to make a spheroidal planet out of it looks more like a case of generating six tiles (not necessarily all with the same settings) and chopping them around to arrange the continents and other features. I foresee two challenges: firstly, getting the results to join up, you would need to produce content over the joins to get rid of discontinuities; and secondly, you could lose global-scale tectonic features such as coastlines matching up on opposite sides of an ocean.

Oh well.

[Maxicus]

Well, I actually did some playing around. and did this:

created a plane equatorial map indicating longitude and latitude cells using Excel.
World Scale.png

generated a few maps using Nortantis and placed imported them into my excel sheet.
TEST 1.png

For the Pole, I generated another map using Nortantis, imported it into photoshop - Filter - Distort - Polar coordinates - Polar to rectangular. and applied to the excel map sheet.
Finished.png

Once I did that, I applied the map to "map to globe" (https://www.maptoglobe.com/)
Globe 1.PNG
Globe 2.PNG

Now I can work with the map to make it appear natural and place the plates using photoshop.