The world is made of floating islands suspended in the atmosphere. For all intents and purposes, the world doesn't have a solid surface.

Due to a magical influence, islands rise from the core of the planet into the outer atmosphere and then fall back into the core in a cycle that takes thousands of years. So islands move vertically albeit very slowly.

Due to a lack of a strong enough magical influence, all but the largest islands (Ireland-sized and above) will drift in the horizontal plane under the influence of the wind. If the wind is strong enough, it may be able to push smaller islands up or down.

The world is about twice the size of earth, that combined with the technology level means that people haven't traveled around the world and haven't discovered most inhabitable land-masses.

Because of the way islands work, they can be 'tied' together to prevent drifting, but this isn't commonplace and is only done on islands that are very close together due to the difficulty of such practice.

Also consider that vertical distances are substantially larger than those found on the Earth, since the temperature and pressure gradients in this world are much less pronounced.


How would the people in the world go about mapping, charting or otherwise have a consistent means of navigating the airs with airplanes and airships?

Similar questions

  • $\begingroup$ Is any kind of bouy/lighthouse system allowed as a viable answer? Or do you just want charting answers? $\endgroup$
    – SRM
    Jan 21, 2017 at 14:21
  • $\begingroup$ Anything that provides a consistent means of navigation for the people of the world is viable. $\endgroup$ Jan 21, 2017 at 14:22
  • $\begingroup$ That is an interesting premise, but how will you deal with the water cycle in this world? Not trying to debunk it. I'm just interested, since all the water will end up in the surface of the planet after a while because rivers will just lead to the edge of the islands. $\endgroup$
    – user31746
    Jan 30, 2017 at 12:42
  • $\begingroup$ @Masterzagh yeah, I've been thinking about that. In Colombia we have a plant called frailejón which makes water out of air humidity, it is a protected species because it's one of our major sources of water. I was thinking of using similar flora to gather water from clouds. $\endgroup$ Jan 31, 2017 at 14:51
  • $\begingroup$ So a hot spot at the center depths of the atmosphere which evaporates all of the river-runoff water back up into the heights as steam, where the frailejon-like plants can capture it and return it to the rivers. Nice simple cycle. Wonder how it worked before the plants evolved? $\endgroup$ May 29, 2019 at 15:46

6 Answers 6


You've actually stumbled into something pretty similar to the philosophy of Polynesian navigators!

Part of their approach to navigating the pacific ocean was to envision themselves at a center of a circle, with the islands, stars, and ocean rotating around them. This would help them maintain a course in the general direction of where they were heading. Then, when they reached certain points in a journey, they'd switch to more active forms of navigation. These methods meant that they very reliable could both find new land, and adjust their course as they zero'd in on their destination.

It's also important to mention that this perspective is extremely difficult to translate to a map. Which for the Polynesians, wasn't that big of a deal! Instead this knowledge was passed through oral tradition, which was far more useful for this approach to navigating the world, especially because their routes and methods would change depending on time. Because of this, when attempts to translate this knowledge into maps happened, the results were illegible to the western perspective.

What does this mean for you?

Like the Polynesians, navigation would probably be done by expert navigators, rather than having a series of tools that allow anyone to navigate. For the first part of a journey, the navigators would consider wind currents for the time of year, allowing them to generally get a sense of where an island will be. Then they'd head in that general direction, probably using a star compass to maintain their bearing. Additionally, the shifts in elevations may affect the movement of stars. A skilled navigator might note how certain stars are eclipsed by the presence of land.

Once in the general area of a land mass (several 100 kilometres), they'd switch to more active methods.

Cloud formations Clouds tend to form more heavily over the ocean (https://atmos.uw.edu/~rmeast/OceanCloudsweb.pdf) and sit at a lower elevation than above land.

While your world doesn't have an ocean, the shifting elevation of the drifting islands would certainly help. Navigators would keep an eye out for gaps in cloud formations (indicating a landmass beyond the horizon), as well as looking for cloud formations that are more likely to appear above land (I'm not sure how your atmosphere would affect this, but there would still be some differences due to the temperature shifts between land and open air). Additionally they might look for cloud formations at incorrect elevations, which would indicate an island being slightly lower or higher than they expected.

Wind and air currents The presence of an island would also be noticeable through the way it impacts air flow around it. Polynesian navigators were able to locate an island based on how its presence would disrupt the wind nearby. An easier way to imagine this process would be like a rock in a river. Even if the rock wasn't visible to you, you could still see the wake and ripples that occur as the water moves past it. Your navigators might look for pockets of dead air, indicating that they're in the 'wake' of a ladnmass. Or, from further away, they'd learn to recognise the turbulence caused by the presence of a landmass. There's some photos on this page that might help to visualise this (https://en.wikipedia.org/wiki/Wake_(physics))

Fauna, and other close indicators As they drew even closer, navigators would then look for signs of animals that indicate a landmass. In the morning, they'd look for birds leaving their homes to forage and head in the opposite direction. While in the evening, they'd follow the direction of birds as they returned to their nests.

However, since there's very little reason for birds to venture out in your world, this method would be slightly adjusted. Instead, navigators would try to align their voyages with the migrations of birds. The idea being that they try to estimate the final leg of their journey to coincide with birds returning to their home islands.

Additionally, there are several minor indicators that could be used to more finely search out a land mass. Polynesian navigators noted that when they were close to an island, they would be able to smell the difference in the air (compared to the smell of the open ocean).

Navigators in your world might be able to tell the difference in light that indicates that an area is in the shade of an island. As well as the shifts in temperature and humidity that indicate a landmass.

Hopefully these methods are helpful, I know this post is a little old but I couldn't help myself from sharing some info on Polynesian navigators!


You say islands move slowly. That means you can produce maps for short-term use, probably pictures of the island with some sort of pitch/roll/yaw notation for how to get to the neighboring islands. For areas that a ship travels commonly, they might have a 3D model, like a mechanical model of the orbit of planets that the ship's navigator can update over time. (It wouldn't be fixed routes like in a planet model... more like positioning the arms of an artificial Christmas tree, bending them to get the right relative positions.)

But for longer-term navigation, a system of "lighthouses" seems useful.

These lighthouses would be some sort of navigation buoy built on the islands that broadcast their positions to nearby ships. Depending upon tech level, this could just be very bright lights (making navigation in foreign lands viable only at night, which would be an interesting story impact) or radio. If each one of the lighthouses has a unique signature, a ship could navigate from point to point by looking for a sequence of lighthouses.

A "harbormaster"-type position would also develop. In our world, lots of ports have local navigators who come aboard ships to do the piloting through the port. These local navigators know where all the underwater rocks hide and where the current of the reef will push a ship. In your world, such local harbormasters might be hired out to do hops or to update maps of ships as they pass through, especially after a hurricane or something that moves the islands more than normal.

As technology improves, you might have ships dedicated to station-keeping the islands -- big tugs that attach to the islands (especially the richest ones) and constantly tug them back into relative position to other islands (especially the richest ones) to keep those navigation routes fixed.

  • 1
    $\begingroup$ This, it's creative and relatively cheap compared to other potential solutions. +1 $\endgroup$
    – Zxyrra
    Jan 21, 2017 at 18:02
  • 1
    $\begingroup$ The mechanical model that can be updated over time is a great idea. $\endgroup$
    – DLosc
    Jan 21, 2017 at 21:16

Allow me to also answer on a 'magical' stand point.

First, the one who want to mapping the world must be someone who is able/mastering the 'wind'. He/She/It must also have knowledge about it. Being able to move from island to island horizontal and vertically. It doesn't matter how he can fly, by magic, steam engined plane, zeppelin or just a big kite.

Then with his arcane knowledge of gravity and wind cycle, combined with his data of the island form, volume, and mass now he can predicted where and when each of his islands moves. The last thing to do is to create the map then.

The map obviously will be much different and more exciting of course than our map. It can be something like a brass globe, with some kind of dials to input date info at the base. When we change the date, it will also move the gear inside and change the islands layout accordingly.

Well you also could make the map on magic parchment, that is touch interactive like our nowadays touch screen.

Of course we should make a small room of error for the sake of the adventures. Be it a big enough unmapped land, with rare creatures and weird indigenous people live on it. Some area that always covered with mist or even other islands near the core or far away above the clouds.

I wish you a clear sky and safe trip.

Kindest Regards

Sky Cartographers Guild Meister

  • $\begingroup$ Similar to this, in a more science vein, consider investigating modern star charts for long periods of time... stars drift. How do we record them? Star movement is less random than wind-blown islands, but maybe the methods apply? $\endgroup$
    – SRM
    Jan 21, 2017 at 14:17
  • 1
    $\begingroup$ Hi, thanks for the answer, I appreciate you taking the time ^‿^. I'd prefer science-based answers rather than magic-based ones since I want to keep magic (wielded by humans) at a minimum. Of course, if it turns out that there's no reasonable non-magical answer, magic is the next best option. $\endgroup$ Jan 21, 2017 at 14:21
  • 1
    $\begingroup$ @Miguel my pleasure. Well basically my answer stated that to map it you must map the wind movement and cycles first (since you stated that wind is the factor), then add the island raw volume and mass count. There you go. As the map should also consider time factor to know where the island be on a specific time, you can create something like a starmap, or solar system globes model, but instead of planets you get islands. $\endgroup$ Jan 21, 2017 at 14:29

In terms of flat maps as we know them, the mapping could be in fact be rather similar to maps in our world. We have longitude and latitude grid lines on maps, and a similar method of charting positions could be explored, using two lines emerging from every mapped island, air and magick. The maps would be circular, with the islands arranged around a center (which could possibly have the same air and magick lines as the other islands, if the map is centered on an island).

Magick lines go through the center of the map and through the center of each of the islands at their current position, air lines map to the wind eddies and patterns around the island, and the island travels along them just as it does the magick lines. The mapping of wind lines would probably be done with some sort of balloon system, otherwise knowledge of the weather, probably through magic, would have to be obtained. Along every air or magick line would have to be placed some sort of formula to be able to calculate, from the difference in date of creation of the map to the present, how far along the lines the island would have moved, and in which direction, it's highly likely that some sort of special calculus or shorthand would evolve over the years to write this. It goes without saying all maps would be dated.

When it comes to air lines, what might be important to consider is that as an island moves across some air lines, it may encounter separate air lines and be whisked away in those wind currents, in which case, wind lines wouldn't be restricted locally to islands, but rather all global wind currents would be drawn on the map, together with their strength, and all islands would be labelled with some form of weight metric, with a formula to decide whether it will be affected by a certain current.

Globes could be very interesting structures, made of spheres and poles. A number of poles would originate from a central sphere, representing magick lines, and on each pole would be a sphere, an island, from which would spring another pole representing air lines. There would be some sort of dial which would enable you to enter the date and the globe would mechanically move all the spheres to the correct place, or possibly, the globe would move automatically every day.

As SRM mentioned, it is highly likely that a more permanent form of navigation would be implemented, using lights and angles, similarly to old nautical navigation. One such implementation might be a sort of sundial placed on an island, however, the "sun" is not in fact the sun, rather, some arcane light emitter placed on a nearby island, set up so that its rays will fall on the sundial on another island, and the precise angle and distance of those two islands can be measured. Of course, as tech level improves, this method will improve, but this is a low tech level example.

Some form of 4D (or even 5D) coordinate system would develop, 3 spacial dimensions and an extra value, correlating to our form of calculus which defines how much an island moves by wind or magick. As SRM has stated, a role of "harbourmaster" or cartographer would become incredibly important, as they know the ins and outs of how each island moves. It is quite likely that these "elders" would be the silent power of your world, and should they organise, could become a formidable political body.


Assuming that the technology level prevents the use of GPS, the problem will be working out where each island actually is, because there is no solid surface for a fixed point of reference.

A compass is not sufficient and may not even work, as a planet needs a rotating liquid metal core to generate a magnetic field. It might be necessary to give your world a magnetic core, or a magical alternative, as you can only use the position of the sun and stars to find North when they aren't covered by clouds.

For position they'll have to use celestial navigational techniques. They can determine their latitude by measuring the angular height of the sun at midday, and allowing for the season.

Determining longitude is much harder as it involves measuring the difference between local solar time and the time at a reference point. For the last couple of hundred years, accurate clocks have simplified this - but it has to be accurate as every second of error creates an inaccuracy of a few miles. Slightly less accurate clocks can be used if you have long-range radio to broadcast a time signal but otherwise a navigator needs to watch for an astronomical event such as a lunar eclipse or a particular configuration of the moons of Jupiter that can be calculated in advance.

This requires a civilisation with fairly advanced mathematics and an understanding that the world is round, and that planets follow predictable orbits around the sun. It also requires full-time astronomers to calculate the necessary tables, and a way to publish these tables to each astronomer.

Humans didn't manage this until we have nation states with fleets of hundreds of ships regularly undertaking long passages out of sight of land, and I imagine only people travelling between islands that were out of sight would need accurate navigation, so I would expect that to be the same.

Given the above, I'd expect each island to have a navigator who takes regular fixes and maintains a clock, and for islands to communicate via signal lamps (and possibly telescopes). They'd send their current location, course and speed to other islands, and pass on the details of others. They could use that to estimate the position of islands for a while, depending how predictable they were.


Radar with Air Traffic Controllers. It's done today

The only difference needed would be multiple networked radar stations presenting a 3D map to the Air Traffic Controllers (ATC) or self managed flight control systems in their air ships/planes.



You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .