Okay, I have a planet with a global ocean of liquid water, and there are these cities in the ocean, floating a few hundred to a few thousand meters beneath the waves. They have spaceports that allow spaceships to access them directly (good thing that the heroine who visits that planet brought an amphibious ship). Weights and counterweights keep them floating at the right depth, energy is gathered via hydroelectric pumps, the materials used to make them are waterproof, and they have farms for food, and large pots are used to heat saltwater up to evaporate it.

And pumps are used to prevent flooding, with most species living there being amphibious at the very least.

The question is: How do we expand a city like that to adapt to an ever-growing population? We can’t let it rise too high, or else it might capsize or something, or if we’re growing seaweed on the outside, it’ll probably dry out and die when exposed to the air for far too long, nor can we let it sink too low, or else there won’t be enough light for the gardens, and if we go lower still, then the city implodes. And we can’t let it grow too crowded, of course.

  • $\begingroup$ Reality checks: Not much light gets down below 200ft, and seaweed only grows down to a depth of ~165ft because of this. You would also need more area than your bubbles provide for agriculture. Note that you would need full pressure vessels/airlocks; moon pools would require pressurized air to counterbalance the water, which can be toxic at about ~184ft (oxygen toxicity). Note that if your cities are that deep to avoid storms, you're better off making the city close up and dive as needed, rather than staying down there (or just attempting to sail away), which would be less energy intensive. $\endgroup$ Dec 4, 2020 at 8:04
  • $\begingroup$ Clockwork-Muse, for cities below 200 feet, hydroelectric power can be used to generate energy for lights to grow crops. Or chemoautotrophic crops can be used. Seals can be used to keep in air and keep out excess water, and the percent oxygen concentration of the cities doesn’t have to be the same as the surface. People can survive an atmosphere of 2 atm and 10% O2, and a atmosphere of .5 atm and 40% O2. $\endgroup$ Dec 4, 2020 at 14:51
  • $\begingroup$ Sure, but if you're using lights in the first place your comment about depth is a red herring. If you have an airlock and a full pressure vessel (required for anything below a certain depth anyways, because humans wouldn't be able to survive the pressure required to counterbalance the water), then you may as well make it... regular atmosphere and pressure. $\endgroup$ Dec 4, 2020 at 16:22
  • $\begingroup$ Yeah, that can totally work. $\endgroup$ Dec 4, 2020 at 18:48

2 Answers 2


If you can't expand up or down, then you expand on the sides, increasing the footprint of the city as the population grows. That's how cities expanded also on land on Earth until it was possible to build skyscrapers.

In your case you need to extend the platform sustaining the city.

Another alternative is that, once the city has reached a preset limit of population or extension it simply starts a new city on a newly built platform.


There are many ways to expand cities, and many things to consider.

Our current-day cities all have limitations and pragmatic issues that determine the way they expand, and what form that expansion takes. Various issues affect this, in many ways much more than the simple constraints you mentioned:

  • Physical constraints (such as the constraints you've identified already)
  • Transport networks (how to get from A to B, or even if you need to)
  • Economic models (retail, industrial, commercial and residential influences)
  • Property value and density (concentration, density and the desire & movement of people)
  • Governance (how organised is the growth, or is it not planned at all)

Here are figure ground maps of different cities to demonstrate the myriad different forms this expansion could take over time: enter image description here As you can see, some cities are planned (like Los Angeles, New York and Barcelona) and some have evolved without constraint (such as Cairo and Mumbai). It matters if your expansion is ordered or not.

Some use Nodal transport networks (like Shanghai, and LA) while others use 'matrix' transport networks (Barcelona is a good example). This also in many ways affects your Land Use, for instance are your industrial areas centralised, or are they distributed evenly throughout?

In cities like Cairo, where you work is close to where you live, so you end up with very even density and little need for large transport distributors. On the other hand, in Tokyo, it is common to commute so you need a large network to ensure people can get to where they want.

It matters not if your city is underwater or not - the only difference is that you can treat the above images in vertical elevation as much as in plan.

The physical constraints you mention simply contain the city to its boundary, however if the other factors start to take priority, even your physical constraints will not be enough to stop the expansion of your city (hence some cities I have witnessed in my travels in China actually demolish nearby mountains to make way for more city - where there is a will there is always a way).

  • $\begingroup$ So, there's a factor that goes for them. Do you see modules of sort being constructed, shipped down, added onto the existing cities, and drained to keep the city at the proper depth? $\endgroup$ Dec 3, 2020 at 15:22
  • $\begingroup$ @TysonDennis I see many techniques used to add to your city. If there is a centralised planning department / bureau then it would be more organised like modules. But if unplanned then it would be more organic, such as companies creating habitat clusters at their leisure, or individuals adding parts themselves, creating a lot more disorderly growth. The depth just changes according to economic factors then. The other point to consider is transport, and how connected people are. Do they interface with cars/subs? Or walk? These also change what form growth takes. $\endgroup$
    – flox
    Dec 3, 2020 at 23:45

You must log in to answer this question.

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