Flight routes: Straight path or actual route?

I brought my GPS unit along on a flight from one city to another. It was a reasonably short flight of about thirty minutes, so the plane didn’t take the direct path between the cities. Also, the flight back went along a different route.

I wasn’t able to locate a best practice for entering flight routes into OSM. Should I simply mark the source and the destination of the flight as a straight line (with intermediate notes for each tile to help rendering), or should I track the actual route taken by the plane?

I wonder how highly variable entities like air routes could be of use in OSM? The only 2 givens here are departure and destination airports, and the flight inbetween is pretty hard to map consistently. With emerging air traffic control systems, and the departure of fixed flight corridors, it gets even harder to map.

Also, why not draw a fully meshed network between every conceivable airport in the world? In the end, that’s what it’s gonna look like anyway.

But if I were to add a flight path, I would it as a straight line.

I agree that flight paths should be entered as straight lines, not only because of the varying paths taken, but primarily because it’s really a binary choice: you either go by plane, or you don’t. Then again, the same thing might be said for ferry routes, which I’m not sure are currently entered as straight lines. It does make sense to add flights, however: for routing purposes, flights might be a desirable option in some cases.

As long as we restrict ourselves to mapping passenger airlines, then we won’t have to create a complete mesh between every conceivable airport in the world, because even major airports cover a reasonably manageable selection of possible destinations. While it’s possible to send an airplane from any airport to virtually any other airport, passenger airlines don’t actually do so (that’s why we have to change flights).

Forgive me for not being an airline pilot, but I understand that there exists “highways for the air”. Air trafic controllers will guide the airplanes to use these standard routes. Using those highways probably makes monitoring the planes easier. Drawing those highways are a nice addon to OSM. I have seen once an airmap from around Shiphol (the biggest airport of the Netherlands). On that map you can clearly see the different routes the planes take.
Drawing straight lines between airports sounds a bit silly and does not give any information that is not already present in the OSM database.

So they should be mapped as very wide “roads” in the 3D space above the normal “groundlevel” of OSM (and I assume they have a "height as well, which normal roads don’t).


Eurocontrol Beek received a traffic control system update quite recently, that makes it possible, in the future, to make more dynamic flight paths possible, and not take these ‘highways for the air’. These ‘flightpaths’ are also not 10 meters wide, like a road, and you’re bound to only map the centerline. Approach and departure from the airport may also take different runways at different times, and planes can go into holding patterns. So, when near the start or end of the flight, there are a lot of variables.

I wasn’t actually advocating really doing that (full mesh). Sorry if my intention wasn’t clear.

Id love to see people submit air chart data they collect. There is alot of stuff to map like the above mentioned airways AWY. Control Zons (CTR), Terminal areas (TMA), Traffic information zones (TIZ). All kinds of beacons, stations, And much, much more really fun stuff that the aviation geeks would love to add. It rather belongs to an air navigation chart then a street map but whatta heck! :slight_smile:

It would never be used for anything else then flight simulations tho. (And we probably would need to put a warning text “Not to be used for rl air navigation” like it it is on the MS flightsimulator charts… hehe). There’s a lot of air nerds out there just waiting for something like this to put their teeth into… :smiley:

Let’s quantify the width of the air “highways”. The image found at http://gallery.blazingangles.net/v/diverse/dokumentar/osm-aalborg-copenhagen.png.html shows the recorded flight routes that I tracked with my GPS unit. The yellow line shows the flight from Aalborg, Denmark to Copenhagen, Denmark. The green line is the return flight. (The tighter curves at the beginning and at the end are car navigation.)

The flight covers a distance of about 200 km in each direction. The track shows how the outgoing flight was delayed and had to fly in a pattern before the landing permission was obtained. You can also see that the routes were quite different; the distance between the outgoing corridor and the return corridor is about 60 km at the widest point.

Obviously a straight line doesn’t show the actual route taken, and I don’t know whether the next several flight will use the same paths (I do intend to map these flights, however, as I expect to ). The question is whether it matters, so I’m still curious about a best practice in case it exists.

I know that some Germans were adding flight routes to OSM last year, and that there was a long and heated discussion going on pro vs. con. You should be able to find those on the talk-de mailing list archives. Hope you can read German.


I’m not sure I understand this. In what form is this information already present in the OSM database–that is, how do I locate this information in, e.g., JOSM?

Sure it could. I routinely need to travel from my home town to the capital of our country. I could go by car: five hours. Or, I could go exclusively by public transportation (train and bus): six hours. Or, I could go by car and ferry: four hours. Or, I could go by car and plane like the other day: two hours including airport check-in.

For route planning purposes, it would be worthwhile having an option to go by plane, assuming that limited data such as typical check-in waits and airplane speed (or, alternatively, flight time) is available.

It might make sense to not include flights on the rendered map as they would probably clutter the view; some specialized maps might render the flight routes instead. It also doesn’t answer the question whether a straight line should be entered or whether the actual routes taken should be entered.

Might as well webscrape this:

And the local Aviation regulating bodies tend to have flight data available in some way.

It’s interesting data, but it would probably be better to have a separate site for flight data since this kind of data is very different from street data. But I understand that using th OSM model is probably the best since it’s so well supported…