In the near future, the surface of the Earth becomes magically unlivable. However, luckily for people, there are now magically floating islands in the skies (typically 5000-8000 feet above the surface), where people can live safely. And also, the Earth now has a thicker atmosphere such that air pressure on these sky islands matches what it currently is on the surface of our Earth.

These islands in the sky vary in size, the smallest sky islands are about 1 acre in land area (I only count the upper surface), up to the largest sky islands having 1 million acres. There's enough total land on the sky islands to sustain 100 million people with current-day western agricultural practices and diet, though there might be some technological improvements to practices that increase this number.

Ignore how the islands themselves are floating. There's some magic that keeps them afloat and prevents rocks from falling down. The islands generally stay somewhat fixed above a given point on the Earth's surface at a fixed elevation. They act as though there's a linear restoring force that ensures even the strongest gusts of wind move the island by little more than a few feet. Aside from the magic making the islands float, real-world physics applies (including weather), and your transportation solution has to follow real-world physics.

Obviously, all modern-day forms of aerial transportation (blimps, helicopters, airplanes if you have a long enough runway) can be used to move people between islands. However, I expect people will want to build cheaper, higher-capacity methods of transportation between different flying islands, both for people and for goods.

Limiting ourselves to near future technology (anything reasonably expected by 2035 is fair game), what other methods could these people come up with to move things from one floating island to another? Assume there are people on both ends working together to build this infrastructure, and they have enough time to build.

The surface is unlivable, but we can build things that rest on the surface. These people have robots that can perform construction at roughly human levels, but they're not cheap and they tend to break down after about a day of operation. If we need to build down to the surface for supports, then we can, but we can't be sending people or goods to within 3000 feet of the surface with any regularity. We can also grab building materials from the surface, so we're not limited on building materials.

Society in this world is peaceful, so you don't need to worry about hardening the infrastructure against attacks or sabotage.

Here are some scales that I'm looking into:

  • For anything under 6000 ft, it looks like suspension bridges are totally doable, and we could probably stretch it a little farther than the current world record.
  • How do we connect a city and a suburb which are on the order of 10 miles apart from each other, and at similar elevation?
  • How do we connect two major cities that are about 200 to 500 miles apart from each other (or can we build any infrastructure connecting them at all)? And does this change if it needs to go up or down by potentially a mile or more?
  • What's the farthest distance that we can build infrastructure that's more cost-effective than flying people and goods via plane or blimp, assuming there's enough demand to drive this infrastructure?

Ideally, your answer will address both movement of people/high-priority goods that need to travel quickly, and bulk, durable goods that can travel more slowly, but only addressing one or the other is fine.

Bonus points if you can figure out these challenge problems:

  • There are two sky islands near each other, but there's a giant mountain in the way, and going within 3000 feet of the mountain is a no-go (and the mountain peak is about level with the two cities). Can we make some form of curved bridge connecting them around the side of the mountain (or up over the top), without relying on any floating islands for support in the middle? And how far apart can we have the cities/how wide can the mountains be with this still working?
  • There is a pair of sky islands, one directly above the other, half a mile apart. How do we move people and goods between these two islands?
  • 2
    $\begingroup$ I don't really get your question. You have already answered it, what else are you looking for? $\endgroup$
    – L.Dutch
    Feb 29 at 6:10
  • $\begingroup$ Planes/helicopters/blimps aren't great for moving a lot of people or goods, hence why we don't do that IRL (and we build other stuff instead). I'm asking if the people in this world have any better options they can build instead. $\endgroup$ Feb 29 at 6:11
  • 3
    $\begingroup$ Our civilization is crucially dependent on mining. How is that possible when surface can't be regularly accessed? $\endgroup$
    – Juraj
    Feb 29 at 11:43
  • 1
    $\begingroup$ It's really hard to ignore the giant floating islands when you're asking about how to make floating things. $\endgroup$
    – fectin
    Mar 1 at 20:41
  • 1
    $\begingroup$ "Ignore how the islands themselves are floating" is the most difficult task I've been given in a long time. $\endgroup$
    – Wyck
    Mar 1 at 20:49

9 Answers 9


Aerial tramways

Admittedly not very cool tech, but it's established - we know how to make them reliable and cost effective. To span long distances they need pylons only every few kilometers or so, with advanced materials maybe farther. And the pylons don't have to carry as big load as if there were suspension bridges, to carry similar amounts of payload. They are somewhat adjustable, although adjustment requires interruption of service. They work best when the path is straight but curved tramway around that hill is possible too. For vertical transport use lifts, they can be made as big as needed, again tried and true tech.

Ballistic or rocket launches

Obviously not suitable for humans or live cargo, but they don't require large airports when there's not enough land. You simply shot out the payload, possibly with some computer steering to correct for wind. Near the destination deploy parachutes and catch it into a net.

  • $\begingroup$ I came to write trains after reading the title and doubly sure after reading the description so giving you an upvote instead. When it comes to mass transportation all modelling eventually creates trains... So, suspended trains running between the bigger islands that can afford to run the cables between them for the trains to run on. $\endgroup$ Mar 4 at 3:41

Bulk freight may not be strictly necessary

As of 2017, the breakdown of bulk transportation done by US owned companies looked like the following:

Locally available goods | 9.44 billion tons | 54.86%

Fossil Fuels | 6.80 billion tons | 39.49%

Non-locally available goods | 0.97 billion tons | 5.64%

In our world of really cheap train and ship transit, saving just a few cents on the dollar is enough to encourage someone to stop using a locally available resources, and start importing it; so, if you take away this option, you will still have access to those resources on every major island with favorable conditions, they will just be slightly more expensive. This means that less hospitable places that currently require a lot of imports for basic survival (like deserts) will generally just be avoided instead of even trying to settle them. This will reduce your total shipping requirements by up to 55%.

The other big shipping cost is the fossil fuel industry. Hopefully, as a near future society, they won't need as much fossil fuels as we do anyway, but if they do, they could probably pipe it in. Since you say that raw materials survive just fine on the surface I will assume that simple pipes made of inert materials will be just fine, as well as the raw materials they carry. So these could probably be distributed by ground-level pipelines made by your robots to move them around and then piped up to the islands. Natural gases will probably become your main fossil fuel (as opposed to petrol or coal) because it can be piped up several km without needing to use hundreds of staged pumps.

This means that if you're willing to take a few minor hits to your efficiency, you can reduce your total actual shipping needs down to about 6% of current levels which is small enough to handle by air-freight alone. While this is significantly more expensive than trains or boats, the kinds of things in the last 6% tend to be much more expensive for thier weight anyway like electronics, appliances, industrial equipment, etc so, shipping them by air is not nearly as proportionally expensive as trying to ship bulk goods you can go local with like corn or lumber.



Though similar to many airship answers, I would offer a better perspective.

Boats and submarines work on buoyancy. To simplify, if you have a shape of which the volume is lighter than that exact shape in water, you'll float. Submarines take this a step further, making themselves heavier than water close to the surface, but lighter than crushing water below a certain depth. That way a submarine can control how deep it goes. We can see this witch cargo ships. Fully laden they are deeper in the water, while empty they are higher. Some open doors in this age, but great breakthroughs in earlier times.

We can do the exact same on your world. Gas is just another fluid, as airships demonstrate so well. The higher density air was probably just meant so people wouldn't comment "how can they breathe!" in the comments, but we can use this to our advantage.

Boats, or a new form of airship, could navigate the higher pressure air closer to the surface. They could potentially take many tons of cargo or people. Probably not as many as cargo ships of today, but more than enough to make mass transport economically viable.

These ships would operate only down below. It would be too much hassle to move them up to the planets. Cargo could be dropped off at fixed infrastructure, or (automated) large drones can take people and cargo up and down. Possibly with cables, allowing them to not use too much energy just for flight. Possibly even generate a lot of energy so the return trip won't cost as much.

This way mass transit and tons of cargo can be moved via the lower dense layers of the atmosphere, using boats buoyant enough to float on the air. Submarines are probably the best to base them off, as they stay upright without the need for a much lower density fluid above it (like water and air). You can have 'open' submarines for cargo.

Keep in mind that you shouldn't expect the ease of transport of our world. Trains would just be unfeasible. People would accept slower transport times if the others are not economical or technologically too advanced.

  • $\begingroup$ "Gas is just another liquid" I think you mean "Gas is just another fluid"? $\endgroup$
    – Nosajimiki
    Feb 29 at 19:52
  • 1
    $\begingroup$ @Nosajimiki yes exactly. The intricacies of the English language often escape me, so I do not always use the right vernacular. I'll edit it. $\endgroup$
    – Trioxidane
    Feb 29 at 19:57
  • 1
    $\begingroup$ That's a cool answer, but I am not sure how well it'd actually work. According to pdblowers.com/tech-talk/elevation-versus-atmospheric-pressure the atmospheric pressure at 0 ft. is 14.7 PSIA and at 8000 ft. it's 10.91 PSIA. The ratio is roughly 15/11 or 3/2, ergo if the pressure shifted, the new pressure at 0 ft. would only be +50% compared to today. I'm not sure airships would have that much lift (and cargo)... especially as the surface is likely not flat and they'd need to be able go up/down a bit to avoid hills... $\endgroup$ Mar 1 at 13:12
  • 1
    $\begingroup$ I actually meant the higher air pressure to make relying on buoyancy or lift function the same as today, not about breathing (I've been to 14k feet and breathed fine, other people live even higher). But that's clever, using the higher pressure near the surface to make airships more effective. $\endgroup$ Mar 2 at 18:04

My idea cant fully answer your question without more detailed information, or exploration of your environment on your part, but in your description the surface air pressure was greater than today. Depending on the diffrence, dirigibles could be an elegant solution.

With denser atmosphere, they could be made easily more ridged / durable. As well as cheaper since more easily obtainable buoyancy gas can be used. With the inhospitable environment imagining them more like submarines than the airships we know may add a flavor that fits into the setting.


Sky lifts and automated cargo trains

Railroads would likely still exist after such a cataclysm, at least where they initially were, wind-based energy would be aplenty at the top, so people can make large wind-based farms, then develop lift-type suspension (8000ft is 2.6 km, we have materials that can withstand the height and tension of that long) off the lower parts of the islands, inland sloped tunnels or lifts on smaller islands to transport stuff within the rock to the surface, effectively using railroad on electric power to do bulk transportation over the land. This would let those sky islands that are over a single landmass to cooperate and coordinate cargo transportation.

For people, airplanes would still be preferred

As much as I am a fan of airships, it's windy at the top, unimpeded winds at 2km height can reach several dozen MPH routinely, not allowing an airship to moor consistently, or just blow it off the island. (The same winds should hamper agriculture, forcing humanity to breed smaller grain cultures, also smaller trees or they would get uprooted by winds). Those magically anchored floating islands allow for cheap wind power generation, especially with denser atmosphere, but winds at that height impede transportation a great deal.

Given the sizes of islands listed at 1 to 1e6 acres, their dimensions (horizontal) should be in range of 60 to 60000 meters, but inter-island distances should then be longer than that, since "the area needed to feed 100M people", according to this question, is 1e8/2350*4 ~= 170 thousand km^2 or 42 million acres (4 here stands for living and extended nutrient base over the figure listed in the accepted answer), quite a few if you ask me. If those islands would all be 1 acre, 42 million of them would be spaced across the Earth at average distance of... 2*sqrt(4*PI*6370^2/42e6) km or 24.28 km (I just divided total square of Earth over the number of islands to find what radius would each of them require to fill that area, and multiplied by 2 to find actual distance) apart. You won't be able to build a bridge covering this distance without touching the ground! With your actual distribution, the average distance would be even greater. So, bridges without ground support are off, thus all kinds of roads, including trolleys, are also off.

Trolleys, while they can reliably cross distances over 1 km per flight, the longest single trolley flight is 2200 meters at Seilbahn Reiteralpe (German), and one of the largest height-wise flights is 848 meters at Ai-Petri trolley (Russian), are still about an order of magnitude shorter than required for such a great project. With each supporting pillar having to be about 1 km high, it would be cheaper to lower the road down to the surface instead of suspending serious mass with higher winds than current trolleys sustain (mass wise), and use railroads. Anyway in case of a disaster on such a suspension all people would have to be evacuated over the land, which is "magically unlivable", so trolleys are off also for safety reasons.

This all leaves airborne transport for humans, which would too be hampered by high winds, but it's designed to withstand wind, and can also be designed to be sheltered at landing and takeoff, to not break apart during crucial parts of flight. But VTOL is off even in such conditions, as any shelter against such high winds would cause enough turbulence to drop an aircraft that does not maintain its own air lift, and/or enough drag to blow it off or crash against land-based structures while traversing the wind barrier.

If automated ships would also be able to remain online after the cataclysm, operating cargo freighters would cover the majority of cargo haul overseas

Ships, however, are afloat over water, and currents both close to the shores/ports and in the high seas would make automated docking very difficult without direct human control. Past that, cranes are already almost fully automated and would allow unloading ships into trains to achieve cargo transfer over the land.

Maintenance will be a GREAT issue

I don't believe that in 2035 we would be able to produce fully autonomous repair drones to fix up various incidents overland, such as train crash, flooding of rail, track displacement and that sort of things, and RC might not be sufficient in case of some serious wreckage, so people would have to go long ways to maintain the railway system in such conditions (unless there would also be magic to fix automatons handy :D), down to making maintenance as hard as today's underwater, over the entire land used for transportation system. I'm not sure if a hundred million people locked way up there could succeed in retaining operability of such enterprise, even assisted by increased renewable energy supplies and a little extra technology edge. Especially if considering that metals and other material would be in short supply due to similarly increased costs to operate mining machinery and whatever processing industry they would have to also maintain.

  • $\begingroup$ I think your math is slightly off for the average distance between islands. If we ignore the Earth's curvature and approximate it as a hexagonal grid, then my math finds $d = 4 R_E \sqrt{\frac{2\pi\sqrt3}{N}}$ (with $R_E$ as the Earth's radius, and $N$ as the number of islands). Plugging in your count of 42M islands gives an average distance of 13km (though also I imagined the distribution as random, not uniform, so there'll be some closer than that and some farther). $\endgroup$ Mar 2 at 19:17
  • $\begingroup$ With airships being hard because it's windy, it seems to me like you could build airship hangars that block out the wind and let it moor stably. You might have difficulty entering/exiting if there's a cross-wind, but this seems to allow some amount of using them for transportation/logistics. $\endgroup$ Mar 2 at 19:34
  • $\begingroup$ @JarredAllen re math, it's possible as it was out-of-the-head napkin calculation. But even 13 km is longer than you're able to cross with a sky trolley, and there would be isolated islands with random distribution, also their quantity would be fewer. Re hangars for airships - maybe again, however they would have to be titanic in size, and turbulence won't be predictable, it's possible that even with a ready-made hangar an airship won't be able to enter it because of buffeting by the wind. $\endgroup$
    – Vesper
    Mar 3 at 6:08

The advancement of technology is rarely in the form of entirely new things, and very commonly in the form of new-ish things being applied to formats they're not already in or massively scaled up. Computer networks had existed for decades, but when the internet came along, it wasn't because they had created something new it was because A. Computers had become cheap enough that anyone could have them and B. Some folks just took the time to spread that infrastructure.

So, ideally, your civilization wouldn't create some magical new method of transportation, but they would build infrastructure and mate technology together in new ways to make what we have significantly more effective.

Cable-based infrastructure

For one: The inhabited islands are floating separately, but are there smaller chunks of land between islands?

Then you can use enormously long, but small train tracks anchored to these little rocks floating around. Think not locomotives, but cable cars.

It's much less plausible, but these cables could be lifted by hot air balloons or some other mechanism that holds them up in the middle. If the cable's relatively taught, sure the angle might change, but it would always stay relatively consistent, and anyways, the start and end are the same, so who cares if the path curves a different way than last time?

For altitude changes, enormous elevator platforms hanging off the sides of islands or dug down through them can be used to give people access to and from straight bridges.

Air-based infrastructure (large scale)

Things like the Lockheed CL-1201 act as massive nomadic airports with planes constantly flying to and fro. What I'm picturing would be a scale such that this can hold 30 mid-sized planes like a Dash-8. People take a short helicopter flight to it as it flies by their island, then from there they can use these planes to take their long journeys. If you look up a Gerald Ford class aircraft carrier, you'll see on their Wikipedia page "Endurance: 25 years. Range: Unlimited". These would only need refueling every few decades and a huge number of them (like the current number of international airports) would be in the skies at all times.

In a smaller scale, airships (or again, massive fusion-powered vessels) look like real-world Landing Helicopter Docks, a form of smaller aircraft carrier built around helicopter operations as opposed to jets. Helicopters are used for short flights to and from the ship, but the bulk of travel is done with all of the helicopters docked and all of the people living on the ship. Things like the Prydwen from Fallout act as cruise ships, cargo vessels, or even surface craft. Lumbering airships carrying rotorcraft specifically designed to dock with it can send these little things out to pick people up, drop off supplies, and whatnot. Remember, airships crossed the Atlantic before flying boats become popular. These things can be used for any kind of distance as long as they don't mind being a little slow.

Personal vehicles (and private infrastructure)

For spread out, small islands, (I'm thinking situations parallel to Alaskan bush pilots having to fly from their house to the post office) small commuter aircraft can be used as dropping planes. There's a great scene in the 2021 Dune where an ornithopter loses two of four wings and can't generate enough lift, so they pull them in, drop out of the sky, lose probably half their altitude, and in doing so get enough lift to get back to level flight and climb back to the altitude they were at earlier, just at a much faster speed now. They get a little bit of speed (say 40% of takeoff speed) then just drop right off the side of the island and climb right back up. Arrestor cables (like those used on aircraft carriers) would be a popular, cheap bit of infrastructure that anyone can invest in building in their backyard. For that same point: in situations where people don't want to live like crazy Alaskan bush pilots diving off the side of an island, the modern steam and electromagnetic catapults used by aircraft carriers could also be something people invest in, but these would be significantly more expensive.

Skyhook systems would be used very often for picking up things. When you want to buy something, the easiest option would be to call ahead to the store, they set it out on a dedicated skyhook platform, send up a wire, and you catch it, then winch it into your plane.

STOL bush planes and VTOL craft like tilt-rotors would also be quite popular.

Last aircraft option: sailing. People track the wind patterns, and when it's pointed where they'd like to go, they head out in a very small airship or glider with perhaps a small emergency prop that stays off most of the time.

Powering all these aircraft

With the exception of the sailing, all of these aircraft options would require a lot of power. You could make electric engines more powerful. This may seem crazy, but you've got a lot of extra space around all of these islands. Huge solar arrays sticking out of the side like the ISS, or just wind turbines along the sides of the islands would be used to generate the electricity for people to run their personal aircraft.

You could also make it so biofuel methods have been made more efficient so many farms would be dedicated to making fuel for combustion aircraft in huge quantities. Remember, jet fuel is basically just real clean diesel (NATO nation militaries use "F-34" or "JP-8" as a universal fuel for jets, helicopters, cars, tanks, camping stoves, boats, basically anything but large ships), and making bio-jet fuel (as well as biofuel for regular old diesel piston engines) is quite practical. In fact, many are currently exploring diesel engine conversions to run on pure ethanol, but primarily in the case of land-based trucks. https://en.wikipedia.org/wiki/Aviation_biofuel

I personally toyed with the idea of a post-apocalypse sci-fi universe where a form of moss that can have its photosynthesis stimulated with artificial radiation sources (or the fungi that grows on it) is used as an extraordinarily common biofuel and having a virtually unlimited fuel source becomes key for restarting civilization.


Rock Pontoon bridges.

If the rocks really have a righting force then you should mine them and build pontoon bridges using them. dragging them into place will be a pain in the ass but they should stay there forever. I assume it has to be only certain rocks, otherwise a lot of things are going to get a lot harder.

Scientists are going to be all over the source of this force so you could even have some refined form used for bridge making.

  • $\begingroup$ This might not work, because the magic involved might just not allow dragging smaller rocks into correct position, or a rock would gain instability, etc etc., up to having a rock drop down to the surface and causing various collapses on other floating areas. Also this solution depends on "magic" and not real-world physics as requested. $\endgroup$
    – Vesper
    Mar 1 at 7:28
  • $\begingroup$ As I understand the righting force, each rock wants to be in a certain location (i.e. in a certain island). Trying to build a bridge out of these would just have them move back. $\endgroup$ Mar 2 at 2:18

Style over substance: It's bubbles. Glued to foamyw hite cloud zeppelin, which reshape themselves into sail like structures. Bubbles can merge to form add hoc parties. How? Floating islands thats how.


Airships (aka zeppelins).

Ships are the least expensive way to move things around today. If the only mode of transportation was by air, I'm certain people wouldn't be too squeamish about using hydrogen airships as the least expensive transport option.

Currently, only helium is generally allowed, mostly thanks to the Hindenburg accident. Helium is expensive and less buoyant, which is probably why airships don't see much use today. With modern technology, the risks of hydrogen would most certainly be quite manageable.

  • $\begingroup$ Airships, even hydrogen ones, make really bad cargo carriers. Apart from much higher upstart costs and maintenance from gas loss, they also cost a lot more per ton over distance than an airplane because of drag. $\endgroup$
    – Nosajimiki
    Feb 29 at 19:48
  • $\begingroup$ @Nosajimiki denser atmosphere somewhat compensates for upstart costs, and it's possible it would also allow them to scale up better, as drag raises over square of size, while lift over cube of size, also allowing higher propulsion via stronger (more massive on top of that) engines. At least it's worth consideration $\endgroup$
    – Vesper
    Mar 1 at 7:26
  • $\begingroup$ @Vesper The place where people live is still 1 atmosphere; so, you still need the same size airship, and even at sea level, it's only going to be about 1.5 atmospheres; so, only a marginal savings to fly under the islands. But airships are many times more expensive for shipping than airplanes, let alone trains, trucks, or boats. $\endgroup$
    – Nosajimiki
    Mar 4 at 20:04
  • $\begingroup$ @Nosajimiki my bad, read meteres instead of feet initially, and yes with this small relative increase, airships won't really "suddenly" become better than airplanes. Yet I first considered them tho. $\endgroup$
    – Vesper
    Mar 5 at 15:34
  • $\begingroup$ There are studies that show airships being more cost-efficient than airplanes per tonne-kilometre (tkm), and this is with helium and current technology. (One such here: link.springer.com/article/10.1007/s13272-016-0188-1). Hydrogen cost virtually nothing, so I'd imagine a society that would develop the technology further wouldn't have much issue with leakage. Also utilizing favourable wind patterns might make a significant difference. $\endgroup$
    – TrayMan
    Mar 21 at 6:17

You must log in to answer this question.

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