After writing this question about the ecology of a terraformed planet, I realized that I hadn't quite tied up all the loose ends of the terraforming process itself. I intended for a group of aliens scientists to turn a desolate hunk of rock into a somewhat Earth-like planet, but this isn't so simple. In particular, I faced a problem with water, among other things.

Let's say that the scientists have already introduced an atmosphere (sans water vapor) that may help keep temperatures at the right level, good enough for liquid water to exist. One of their next steps will be to add oceans (roughly half the size of Earth's), to establish a water cycle and get things ready for plants to be added.

The scientists don't know if there's any water belowground, though measurements have established that there may be some. Crust analyses show the potential for hydrogen and oxygen, but these haven't been definitive. The poles and other regions have been fully explored, but no liquid ice has been found.

Taking all of that into account, what's the quickest (not necessarily cheapest) way to add the oceans? Other bodies of water will come later.

  • 8
    $\begingroup$ what-if.xkcd.com/53 $\endgroup$
    – Dragonrage
    Mar 30, 2016 at 17:42
  • 2
    $\begingroup$ "I intended ... to turn a desolate hunk of rock into a somewhat Earth-like planet, but this isn't so simple." Indeed. $\endgroup$ Mar 30, 2016 at 18:22
  • 8
    $\begingroup$ @Dragonrage Don't forget part 2 (which is arguably more relevant). $\endgroup$
    – Ajedi32
    Mar 30, 2016 at 19:18
  • 10
    $\begingroup$ For some reason, I'm imagining a packet labeled "Instant Ocean -- just add water!" :P $\endgroup$ Mar 30, 2016 at 19:39
  • 1
    $\begingroup$ Yet I wonder, still I wonder, who'll stop the rain? $\endgroup$ Mar 31, 2016 at 13:25

7 Answers 7


I would say capturing a bunch of comets and burn them up in the atmosphere, leaving plenty of water vapor.

Hydrogen and Oxygen are two of the most common elements in the universe, #1 and #3 respectively. So any 'atmosphere' you have will likely have plenty of both of these elements. As long as there is hydrogen available, you can have microbes or something else breaking down Iron Oxide since Fe is #6 on the most common elements and this will release oxygen and depending on the biological makeup, it can also be creating a soil at the same time.

Jordi recommended SiO2 as the Oxide to separate.

Of course, the absolute simplest would be to find a frozen moon like Europa put it in orbit around the planet and then send large chunks to burn up in the atmosphere. Smashing the moon directly into the planet might postpone any terra forming for a million years or so.

  • 2
    $\begingroup$ I would suggest the bacteria and microbes to break apart Silicate (SiO2) if the crust is Earth-like. Silicon presence in Earth's crust is about 277ppm (about 5 times iron's abundance), most of it in form of silicate, and Iron is so heavy that it tends to sink to the planet's core. $\endgroup$ Mar 30, 2016 at 16:04
  • $\begingroup$ @JordiVilaplana I figured there might be another better oxide, but I knew there would be iron oxide $\endgroup$
    – bowlturner
    Mar 30, 2016 at 16:06
  • $\begingroup$ SiO2 is Quartz, might I just say. $\endgroup$ Mar 31, 2016 at 13:26
  • 5
    $\begingroup$ "The absolute simplest would be to change the orbit of a small moon..." man, I'd hate to see the complicated solution. $\endgroup$ Mar 31, 2016 at 19:44
  • $\begingroup$ There are about 4,000 known comets that come anywhere close to the inner solar system that together don't have enough water to give you lake Michigan let alone oceans. Most comets are in the Oort cloud which is 3 light years away. Assuming you can get technology all the way out there to capture them, it will take about 40,000 years for them to reach your planet. So the fast part of this answer is missing! $\endgroup$
    – Dave
    Apr 8, 2016 at 2:31

Use a Solar Recombinator

Your sun has everything you need to get full-sized oceans in 10 to 100 years.

Finding, capturing comets and bombarding your carefully terraformed planet seems expensive and hazardous and way, way too time consuming. Plus it's too unpredictable. You need a systematic, reliable, predictable process.

Better would be to use technology to extract the oxygen and hydrogen that your planet's sun already has in abundance. Assuming your sun is within the main sequence, it lacks the core pressure to fuse Oxygen, however most Suns are composed of about 1% Oxygen (from previously exploded Suns) which is WAY more than you need. Of course, the Sun is 98% Hydrogen with the remainder being Helium which is being fused in the core....

Anyway - The easiest place to find Oxygen in the Sun is within the Sun spots which are cool enough (4500K) to collect it in its molecular form. Given this, the solution needs three primary subsystems:

  1. Solar Oxygen Extractor - This would be a solar satellite which orbits the Sun and can be directed to pass over Sun spots where it could collect the O and using magnetic resonance (think rail gun) shoot a beam of O atoms to an H2O Recombinator that is permanently located at your planet's L1 Lagrange point (the one between the sun and your planet.) Depending on the flow you need, you could have many of these.

  2. Solar Hydrogen Extractor - Like the O extractor, the H extractor orbits the Sun, collects the Hydrogen, and shoots a beam of H to the H2O Recombinator. You will have have twice as many of these as Oxygen extractors (e.g. H2O).

  3. H2O Recombinator -- This is a massive space station, with two receiving pads -- one for the O and one for the H. Within it, it maintains a continuous combination reaction to create water. Once created, the water is immediately sprayed in a long beam (via Ion wave transport) at very high speed towards your planet. The beam of water in the vacuum of space will immediately crystallize, but it will not have time to sublimate before it hits the upper atmosphere where it will immediately melt and add to your water cycle. Because the recombinator station is at L1, your planet will be rotating in place without any relative motion, thus the ice beam will create a nice even coating of moisture throughout your world without any need to for hyper accurate aim corrections for timing, etc. -- EASY! Let the oceans naturally form in the low altitude regions.

In order to create 10^18 tons of water (which is how much the earth has) over a period of 100 years, you would need to send water at the rate of about 300 million tons per second. (And you think comets could get you there? no way!) This seems incredibly high, but its not if you build a large enough recombinator.

Assuming the size of the water dispenser on the recombinator is roughly a 1000m x 1000m square (or 1 million square meters), you only need to produce an ice stream 300 meters long per second -- which is certainly do-able as long as the inputs coming from the solar extractors have sufficient flow rates.

If 100 years is too slow, you can make it less than 10 years if you just ramp up the flow rates or deploy additional recombinator stations.

  • $\begingroup$ Disclaimer - my math might be off a little bit as I just thought this idea up and hammered it out having saw this question on the stack overflow sidebar and thought it was such an awesome question I joined the group to add my 2 cents! $\endgroup$
    – Dave
    Mar 30, 2016 at 23:41
  • $\begingroup$ You still have to worry about melting the lithosphere when all that mass is brought down from space, converting grqvitational potential energy to heat. $\endgroup$
    – JDługosz
    Mar 31, 2016 at 19:13
  • $\begingroup$ Good point, even at -53C ice crystals will generate some heat on entry -- more due to kinetic energy than gravitational potential. However most of that heat will simply radiate into space. Perhaps we need to add a series of low orbit ice de-accelerators that convert solar energy into an electrical field that can absorb some of that energy. I'll put a team on this issue right away! $\endgroup$
    – Dave
    Mar 31, 2016 at 20:12
  • $\begingroup$ Kenetic energy comes from the potential. You can figure out how much by working with gravity potential "at escape". If the object falls from infinity, it will crash with kenetic energy of escape velocity. If you use parachutes, gliders, etc., even rockets within the atmosphere (or firing towards the ground) you still deliver that energy to the planet. $\endgroup$
    – JDługosz
    Mar 31, 2016 at 20:17

There's a couple of ways.

Use What's Already Available

You can reroute comets to crash into the planet. Thousands and thousands of them, over many years. (you did not specify a time frame, so I'm taking some liberties here)

By crash I mean to say that these comets should make contact with, and enter the atmosphere. I would hope that they would simply evaporate and become a part of your water cycle.

You can also set up processing plants on some nearby planet that has polar caps, or water in some form, and either cut up the ice and shoot it out to the planet, or freeze it, then do so. You may want to create a fleet or robots that get started on this ASAP.

Let There Be Water

Alternatively, you can try creating some water by combining oxygen and hydrogen atoms. You've already "imported" an atmosphere, so I assume you have access to some way of creating these gases in large amounts.

Between these two methods you should get the job done in a reasonable time frame.

  • $\begingroup$ My only worry with the comet idea is whether or not they could survive atmospheric entry in good shape. $\endgroup$
    – HDE 226868
    Mar 30, 2016 at 15:17
  • 9
    $\begingroup$ @HDE226868- I think the point is you don't want the comets to survive re-entry. You want them to break up and for their raw materials to blend in with your new planet- especially the water. $\endgroup$
    – cobaltduck
    Mar 30, 2016 at 15:25
  • 6
    $\begingroup$ @HDE226868 You might be better off if they didn't survive entry. Comets hitting the ground are going to hurt the teraforming. You'd be better off if they came apart in the atmosphere, turning into water vapor. Then it would precipitate out as rain, and start the water cycle. If you had large comets, you could cut them apart so the pieces would be small enough to "burn up" or melt. If you had a space elevator or something like that, then you could maybe take them down in one piece. The only reason to crash them is if you need to add heat to the equation. $\endgroup$
    – AndyD273
    Mar 30, 2016 at 15:28
  • 2
    $\begingroup$ Andrei you may want to include the discussion in the comments in your answer...they help flesh it out and explain it quite a bit. $\endgroup$
    – James
    Mar 30, 2016 at 15:33
  • 1
    $\begingroup$ @AndyD273 You get the same amount of thermal energy whether the comet hits the ground or not. - The kinetic energy of the comet is going to be transformed to heat. The difference is if you want it concentrated in a single spot or not. $\endgroup$
    – Taemyr
    Mar 31, 2016 at 12:19

For a cold planet like Mars, simply sending comets on intersecting orbits to collide with your target will both add energy and heat to the ecosystem, and also water (comets are mostly ice, after all).

You will probably want to refine the comet on the way in so that you are not loading the planet with the various other volatile elements frozen in the ice, unless these elements are important to the project. So the comet will have processing station of some sort to refine the ice and essentially extrude a huge ice cube of pure water which will be allowed to crash into the planet, while the rest of the assembly accelerates away to avoid the crash.

One thing to keep in mind is a comet impact is extremely energetic (as in dinosaur killer energetic), and if your planet is small enough, a lot of the atmosphere and water could be blasted back into space, undoing the purpose of your mission. If the orbital parameters are wrong, one of the other tasks of your ice processing mission might be to attach the ice cubes to solar sails and reduce the impact speed so the comets don't blast their water back into space. The speed will have to be calculated separately for each planet, but the upper amount of energy imparted to the molecules of water has to be less than the escape velocity of the planet.


"The secret is to use a really big bucket." - Buck Godot: Zap Gun for Hire, The Gallimaufry series.

If you start with either a Mars or Venus type world, what is really missing is hydrogen Both Venus and Mars have plenty of oxygen bound as CO2 or oxides to create a lot of water if you can add hydrogen. Both planets became barren after loosing their hydrogen.

The importation of hydrogen is the obvious answer and the method depends on your aliens level of technology. If they're slightly more advanced than we are, then comets or dwarf planets are the only choice.

But if they are star-faring with FTL drives of some sort, they likely have control over gravity so the easiest source of hydrogen would be for them to use a controlled gravity field to scoop some hydrogen off the nearest gas giant.

They'd probably want to grab a bunch of helium as well for a martian world, in order to maintain atmospheric pressure in the short term. Mars surface is covered in a layer of superoxide compounds (essentially bleach) so adding anything mildly reactive, such as atomic hydrogen, will cause a violent release of hot oxygen, which in this case will form water. The water in turn will degrade the oxide but not be consumed by it so oxygen will bubble out all over in chain reaction. The released heat will likely cause the the ignition of of more hydrogen.

A cold earth sized planet will have water because it will retain its hydrogen through gravity. A hot earth sized earth can turn into a Venus if it never evolves oxygen producing organism... and so on.


Some of the icy moons have more water than Earth! This alien system might be the same, with some large body of water-ice in one lump that is more than enough.

Some of it will be reaction mass to get the rest moved inward.

The real problem will be landing it without adding too much energy to the planet. That kind of tonnage is a lot of potential energy to release. You'll use more of it up as reaction mass to bring it into low orbit (way below escape velocity) and then kill the orbital velocity so it falls straight down from a meer few hundred miles. The staging area, which would be a ring system, can shade the planet from the sun to compensate for the amount of heat you can't help adding.


Go out to the system's Oort cloud, find all of the objects which have substantial amounts of water ice, and send them inward so that they collide with the planet you wish to terraform.

They will be hitting the planet with a kinetic energy that is not different from escape velocity, which is much higher than the amount of energy required to turn ice into water vapor. I strongly recommend that this be the first stage of your terraforming project, because nothing you have done up to this point will survive adding the water.

The Earth's oceans contain 1.3 x 10^9 cubic kilometers of water, so that will give you an idea of how many ice balls you need to find.


You must log in to answer this question.

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