Mapping the Moon
Previously, I simulated the climate of a terraformed Luna and clustered the results into four biomes. Now, the results have been baked into hexmaps, free for anyone to download and use as they see fit.
Map Projections
Figuring out what map projection(s) to use for the Moon contrasted with the challenges posed by the 6-mile Hex Map of Earth in interesting ways. On the one hand, the Moon's small size greatly constrains the amount of land I can map with minimal distortion. On the other, Lunar geography (selenography?) is considerably less recognizable than that on Earth, so the bar for acceptable distortion is more forgiving -- nobody is going to get on my case if Mare Nubium is a bit warped. (Similarly, I suspect that few users of these maps will have strong pre-existing attachments to specific features, so it seems preferable to make fewer, larger, somewhat distorted maps for people to browse freely and clip to their liking).
Finally, there are technical limitations. QGIS can handle non-earth geography, but it's a bit wonky and the number of projections available is small (though it is possible to define custom coordinate reference systems).
With all that in mind, I settled on making four maps: Lunar North Pole, Lunar South Pole, Nearside Equator, and Farside Equator, following the precedent here:
The distortion around 45 degrees latitude is strong but not beyond what's acceptable for elfgames (imo), and the divisions feel logical & intuitive.
Overlaying the maps with hexes was a bit tricky, given the amount of distortion -- it's not totally clear to me how the QGIS Create Grid tool determines spacing, so I did napkin math to get the rough number of hexes expected for each map and then played with the spacing parameters until I got close enough.
- One 6-mile hex covers 80 km^2.
- The surface area of the moon is 38,000,000 km^2
- Therefore, the total number of hexes across all maps should be 38,000,000/80 = approx. 475,000.
- The surface area represented by the polar maps bounded at 45 degrees latitude is 2*pi*(R^2)*(1-cos(r/R)), where R is the radius of the Moon and r is the arc-length radius of the circle, which works out to be 2pi(R^2)(1-cos(pi/4)) since the radius of the polar maps is 1/8th the Lunar circumference. This works out to 5,555,000 km^2, which passes the sniff test of being a bit less than 1/6th of the total surface area. This means the polar maps should be covered by around 69,000 hexes
- The two equatorial maps each cover half the remaining area, or around 168,000 hexes
This is obviously a ludicrously large number, but it should be noted that only half of the hexes are on land.
Cartographic Stylings
A dilemma I've run into when making hexmaps from real terrain: do I use the terrain as-is, depicting it with a raster basemap & hillshading? Or do I assign each hex a single feature type?. Are hexes rooms or measuring sticks? What is the sound of one hand clapping?
For the Hex Map of Earth, I decided that, given the availability of real-world maps, it was more useful to front-load the work of binning terrain into discrete hex types. But for the Moon, the rough, cratered terrain should really be the star of the show, so I decided to simply overlay a hex grid over the terrain.
(For the Antarctica map, as of recently I have uploaded both approaches).
John Nelson recently put up a neat video about multi-directional hillshades & hypsometry, which I implemented in QGIS to make the pockmarked Lunar contours really pop.
For the biomes, I ended up just applying a green gradient as a 'soft light' layer to tint the terrain according to imagined vegetation density. I may tinker with this in the future.