This diagram is oddly selective about which airports it includes. It has some class D airports (Modesto, Santa Maria) but not others (Palo Alto, El Monte, Fullerton). There is no pattern that I can discern other than eliminating airports that are too close to each other, which kind of defeats the purpose of this exercise.
This is my favorite instance of a voronoi diagram: <a href="https://imgur.com/a/zjNWL" rel="nofollow">https://imgur.com/a/zjNWL</a><p>I actually wrote into the Reese's company to find out what that was happening to their giant cups, and got no reply unfortunately.
I did a similar visualisation a few months ago, but less polished, for my own curiosity: <a href="https://twitter.com/iandioch/status/968938231550107649" rel="nofollow">https://twitter.com/iandioch/status/968938231550107649</a>
Interesting to compare with ETOPS (good maps: gc.kls2.com) which governs how far away from acceptable alternate airports a commercial plane can be. It corresponds to safe flying time with a single functional engine. So a lot of the area in some of the larger Voronoi regions, especially in the Southern Hemisphere, is actually unreachable by any commercial flight from the closest airport.
Voronoi diagrams are so fun. I messed around for months with them after I saw Amit Patel's procedural map generator that used them.<p><a href="http://www-cs-students.stanford.edu/~amitp/game-programming/polygon-map-generation/" rel="nofollow">http://www-cs-students.stanford.edu/~amitp/game-programming/...</a><p>Fortune's Algorithm is pretty simple, but it's tricky to get right, particularly if you're also maintaining the information in a form that lets you do something useful with it besides spitting out the points of the Voronoi centers.<p><a href="https://en.wikipedia.org/wiki/Fortune%27s_algorithm" rel="nofollow">https://en.wikipedia.org/wiki/Fortune%27s_algorithm</a>
I wonder what happens if you "give weight" to each point relative to size of the airport (measured e.g. by number of take-offs/landings per year)?
I've flown in and out of Mataveri and I know it sounds superstitious or whatever, but I felt that isolation. It's amazing to me that the Polynesians were able to colonize so widely.<p>Also, I love Voronoi diagrams. I ran across them many moons ago when I was building a wayfinding application and was looking for ways to generate map meshes—this was not that—but I thought they were super interesting.
Discussed from 2015: <a href="https://news.ycombinator.com/item?id=10161326" rel="nofollow">https://news.ycombinator.com/item?id=10161326</a>.<p>(Posted in 2014 too but there weren't really comments: <a href="https://news.ycombinator.com/item?id=7557438" rel="nofollow">https://news.ycombinator.com/item?id=7557438</a>)
I cant see Baltra Airport in the Galapagos Islands (<a href="https://goo.gl/maps/x9PW5i4VWMH2" rel="nofollow">https://goo.gl/maps/x9PW5i4VWMH2</a>) - it is "near" the most remote Mataveri Airport on Easter Island and may skew the most remote result to another airport?<p>I guess not all airports are included (e.g there is one I've flown into as a tourist to the antarctic peninsula that is missing (<a href="https://goo.gl/maps/u54LnTDYb8N2" rel="nofollow">https://goo.gl/maps/u54LnTDYb8N2</a>)<p>Great visualisation though :)
Nice one, actually all Voronoi diagrams are cool.<p>If only there was a webpage/software where someone could click/select points on a map (or even better enter coordinates) and a user Voronoi diagram would be created ;-)
I did some work on Voronoi a few years ago. I made an overlay in Voronoi transparent colored cells that would show for example where the nearest convenience store was to a given location in the city and this would overlay the regular map of the city. I was going to try to sell these to the convenience store as they always had a map of the city they were in somewhere in each store. An improvement would be to run a google map from each store to each house and actually use the mileage to the store the store. This would account for rivers etc that would block the shortest path. There is probably an idea for a startup somewhere there....
I've been grinding on an observation about the distribution of people on Earth. It can be reduced to the observation that the city of Ushuaia, AR at 55 degrees south is the most southerly city (FWIW the other two contenders for that title are near Ushuaia), while the city of Copenhagen, DK at 55 degrees north is definitely not the most northerly city. Although there is a great deal of land below the Antarctic Circle, there are only research stations there. There isn't as much land above the Arctic Circle, but civilians live above it.
If you want to know more about those Antarctic airstrips, there's a great Wendover video on them:<p><a href="https://www.youtube.com/watch?v=-s3j-ptJD10" rel="nofollow">https://www.youtube.com/watch?v=-s3j-ptJD10</a><p>Most of them are just hard packed ice.
Interesting - was looking at some of the larger area ones in the US and learned that Santa Fe has a tiny airport - smaller than the one here near Bend, Oregon in Redmond. Seems Bend is bigger than Santa Fe for that matter... didn't realize that.<p>Which means that the Albuquerque one ends up as the largest, it looks like, followed (eyeballing it) by Elko?
<a href="http://voronoi.surge.sh/" rel="nofollow">http://voronoi.surge.sh/</a><p>Recently, I was working on a WebGL version that renders all 54,000+ airports and got it working again today after seeing this.It's quite pretty and runs well but needs some love.
When I did work on Voronoi, i used the Mathematica built in function. I would get the gps coordinants from Google maps and use them as input. The output put into a google function that could make rectangular colored transparent regions on the map. But this was all 2d.
I bet the vertices are correct, but shouldn't the lines in between them look like curves? They are projected solutions for the points that are equidistant by great circle distance to two airports, so I think they should themselves be great circle arcs.
I assume it's using the great circle distance metric. I wonder how it would look like if the metric was "shortest land/sea travel time".
This is awesome. Now I'd love to be able to visualize which aiports are flown to/from the one under the mouse... Where can one get this data?
Am I missing something, there are a number of airports in Antartica, I assume that none of these are commercial though:<p><a href="https://en.wikipedia.org/wiki/List_of_airports_in_Antarctica" rel="nofollow">https://en.wikipedia.org/wiki/List_of_airports_in_Antarctica</a>
This is interesting because normally I'd assume this would just follow population density (Europe, the NA, Japan, see xkcd.com/1138) but India and China barely have any airports relative to their populations. That's a lot of traveling just to get to an airport.
This gives a lot of importance to small airports in small islands. Big international airports are surrounded by other smaller airports. They are associated with a tiny surface. I can not see how this representation may be of any use for airports.